@prefix vivo: . @prefix edm: . @prefix ns0: . @prefix dcterms: . @prefix skos: . vivo:departmentOrSchool "Arts, Faculty of"@en, "Political Science, Department of"@en ; edm:dataProvider "DSpace"@en ; ns0:degreeCampus "UBCV"@en ; dcterms:creator "Beauvais, Edana"@en ; dcterms:issued "2017-08-16T16:55:31Z"@en, "2017"@en ; vivo:relatedDegree "Doctor of Philosophy - PhD"@en ; ns0:degreeGrantor "University of British Columbia"@en ; dcterms:description """To count as democratic, social systems must empower the inclusion of people affected by collective endeavours to participate in practices that contribute to self-development and self- and collective-rule. As a political practice, talking is important because it is an essential tool for enacting social identities and enabling self-development. Talking is how people think through their preferences, and helps people relate private preferences to those collective opinions and agendas that enable collective rule. However, formal barriers (such as legal restrictions) can entail exclusions that prevent disempowered social group members from participating in, or influencing practices – including talk – that contribute to self-development and self- and collective-rule. Furthermore, even in the absence of formal barriers to social and political participation, the historical legacy of structural inequality can pattern social cognition and contribute to internal exclusions that engender asymmetries in political participation and influence, including asymmetries in discursive participation and influence. I address the empirical question of whether inequality shapes social cognition to engender asymmetries in social group members’ discursive participation and influence in two analyses. In my empirical chapters, I turn my attention from a broader concern with social inequality and narrow my focus to gender inequality. In my first empirical chapter, I use Canada Election Studies (2015) data to show there is an ongoing gender gap in discursive participation. In my second empirical chapter, I use data from an original vignette experiment to show that when women do talk politics, they have less influence than men. Finally, I suggest practices and institutions to help neutralise discursive inequalities, so democratic systems can come closer to the ideal of discursive equality, or equal participation and influence in communicative processes of self-development and self- and collective-rule."""@en ; edm:aggregatedCHO "https://circle.library.ubc.ca/rest/handle/2429/62658?expand=metadata"@en ; skos:note "DISCURSIVE EQUALITY by Edana Beauvais B.A., The University of Toronto, 2008 M.A., The University of British Columbia, 2011 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Political Science) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) August 2017 © Edana Beauvais, 2017 ii Abstract To count as democratic, social systems must empower the inclusion of people affected by collective endeavours to participate in practices that contribute to self-development and self- and collective-rule. As a political practice, talking is important because it is an essential tool for enacting social identities and enabling self-development. Talking is how people think through their preferences, and helps people relate private preferences to those collective opinions and agendas that enable collective rule. However, formal barriers (such as legal restrictions) can entail exclusions that prevent disempowered social group members from participating in, or influencing practices – including talk – that contribute to self-development and self- and collective-rule. Furthermore, even in the absence of formal barriers to social and political participation, the historical legacy of structural inequality can pattern social cognition and contribute to internal exclusions that engender asymmetries in political participation and influence, including asymmetries in discursive participation and influence. I address the empirical question of whether inequality shapes social cognition to engender asymmetries in social group members’ discursive participation and influence in two analyses. In my empirical chapters, I turn my attention from a broader concern with social inequality and narrow my focus to gender inequality. In my first empirical chapter, I use Canada Election Studies (2015) data to show there is an ongoing gender gap in discursive participation. In my second empirical chapter, I use data from an original vignette experiment to show that when women do talk politics, they have less influence than men. Finally, I suggest practices and institutions to help neutralise discursive inequalities, so democratic systems can come closer to iii the ideal of discursive equality, or equal participation and influence in communicative processes of self-development and self- and collective-rule. iv Lay Summary Democracy means we have a say over how we live. This means we can develop who we are, such as by deciding what to do with our lives or who we spend time with. It also means we have a say over the rules that guide our actions, including social expectations and laws that govern what we can and can’t do. Talking is important for democracy because talking helps us develop our capacities, and talking helps us negotiate with others about what rules should govern our actions. Group-based inequality is bad because it can exclude people from participating in, or influencing, conversations. My research shows that, because of ongoing gender inequalities, women participate in political talk at lower rates than men, and have less influence than men when they speak. But I also suggest ways to address inequality so everyone can participate in and influence political talk, so that talking enhances democracy. v Preface I identified and designed this research program in consultation with my supervisory committee. The research conducted this dissertation was approved by the UBC Behavioural Ethics Board, Certificate Number H15-02776 (and the renewal Certificate Number H15-02776-A001). Portions of Chapter 2 (parts of section 2.2.) and 9 (parts of section 9.1) have been published as: Beauvais, Edana. 2017. “Deliberation and Equality.” In The Oxford Handbook of Deliberative Democracy, edited by André Bächtiger, John Dryzek, Jane Mansbridge, and Mark Warren. Oxford University Press. Portions of Chapter 9 (from section 9.1.1.) have been published as: Beauvais, Edana, and Bächtiger, André. 2016. “Taking the Goals of Deliberation Seriously: A Differentiated View on Equality and Equity in Deliberative Designs and Processes.” Journal of Public Deliberation, Special edition: Equality, Equity, and Deliberation, 12 (2). I was the lead investigator, responsible for all major areas of concept formation and manuscript composition. Bächtiger, André was the supervisory author on this project and was involved throughout the project in concept formation and manuscript edits. vi Table of Contents Abstract .......................................................................................................................................... ii Lay Summary ............................................................................................................................... iv Preface .............................................................................................................................................v Table of Contents ......................................................................................................................... vi List of Tables ............................................................................................................................... xii List of Figures ............................................................................................................................. xvi Acknowledgements .................................................................................................................. xviii Dedication .....................................................................................................................................xx Chapter 1: Introduction ................................................................................................................1 1.1 Research Questions and Background.............................................................................. 4 1.1.1 Why is More Democracy Better, and More Oppression and Domination Worse? .... 4 1.1.2 What Kind Of “Equality” Empowers Inclusion in Democratic Processes? ............... 7 1.1.3 What are the Consequences of Inequality for the Democratic Potential of Talk? ...... 8 1.1.4 What are the Consequences of Discursive Inequality for the Democratic Potential of Talk? .... ............................................................................................................................... .10 1.1.5 Where do Discursive Inequalities Come From? ....................................................... 12 1.1.6 How can we Achieve Discursive Equality? .............................................................. 14 1.2 Chapter Summaries ....................................................................................................... 16 1.2.1 Chapter 2 Equality and Democracy .......................................................................... 16 1.2.2 Chapter 3 Talk and Collective Opinion and Agenda formation ............................... 17 1.2.3 Chapter 4 Practical Knowledge ................................................................................ 17 vii 1.2.4 Chapter 5 Spaces of Public Appearance ................................................................... 19 1.2.5 Chapter 6 Inequality, Oppression, and Domination .................................................. 20 1.2.6 Chapter 7 An Analysis of Gendered Asymmetries of Discursive Participation ....... 21 1.2.7 Chapter 8 An Analysis of Gendered Asymmetries of Discursive Influence ............ 23 1.2.8 Chapter 9 Achieving Discursive Equality................................................................. 24 1.2.9 Chapter 10 Conclusion .............................................................................................. 25 1.3 Conclusion .................................................................................................................... 25 Chapter 2: Equality and Democracy..........................................................................................28 2.1 A Systems Approach to Democratic Theory ................................................................ 29 2.1.1 Benefits of the Systems Approach ............................................................................ 33 2.2 Empowered Inclusion ................................................................................................... 35 2.3 Collective Opinion and Agenda Formation .................................................................. 40 2.4 Collective Decision-Making ......................................................................................... 42 2.5 The Problem of Inequality ............................................................................................ 44 2.5.1 Interactional (Including Discursive) Inequalities and Internal Exclusion ................ 49 2.6 Conclusion .................................................................................................................... 53 Chapter 3: Talk and Collective Opinion and Agenda Formation ...........................................55 3.1 Forms of Talk ................................................................................................................ 56 3.1.1 Talk Versus Social Interaction .................................................................................. 56 3.1.2 A Shared Background of Implicit Mutual Understandings ...................................... 58 3.1.3 Resolving Conflict and Misunderstanding Through Deliberation ............................ 59 3.1.4 The Importance of Everyday Talk for Democracy ................................................... 64 3.2 Using Talk to Achieve the Collective Opinion and Agenda Formation Functions ...... 70 viii 3.2.1 Democracy-Supporting Social Integration Functions ............................................... 70 3.2.2 Reciprocity Functions ............................................................................................... 73 3.2.3 Democracy-Supporting Personality Functions ......................................................... 79 3.2.4 Learning Functions ................................................................................................... 80 3.2.5 Legitimation Functions ............................................................................................. 83 3.3 Inequality Prevents Talk from Achieving Collective Opinion and Agenda Formation .. . ………………………………………………………………………………………………..84 3.4 Conclusion .................................................................................................................... 89 Chapter 4: Practical Knowledge.................................................................................................92 4.1 The Sensorimotor Experience of Intersubjectivity ....................................................... 94 4.2 Inscribing Shared Intentions in Things: Language and Institutions ............................. 97 4.3 Inscribing Shared Intentions in Minds: Habitus ......................................................... 102 4.4 Isomorphism, Homology, and the Common-Sense World ......................................... 112 4.5 The Counterweight of the Common-Sense World and Human Agency..................... 117 4.6 Conclusion .................................................................................................................. 122 Chapter 5: Spaces of Public Appearance.................................................................................124 5.1 Human Plurality and the Need for Revelation and Recognition ................................. 127 5.2 Social Differentiation through Social Identities ......................................................... 130 5.3 Social Differentiation through Fields.......................................................................... 135 5.4 Preserving Power Relations ........................................................................................ 141 5.5 The Democratic Aim of Creating and Preserving Symmetrical Power Relations ...... 146 5.6 Conclusion .................................................................................................................. 147 Chapter 6: Inequality, Oppression, and Domination .............................................................151 ix 6.1 The Relationship Between Inequality and the Harms of Oppression and Domination ………..................................................................................................................................... 153 6.1.1 Five Faces of Oppression ........................................................................................ 155 6.1.2 Domination ............................................................................................................. 157 6.2 When Habitus Engenders Interactional Inequality and Internal Exclusions .............. 159 6.3 The Example of Patriarchy ......................................................................................... 163 6.4 Discursive Inequality Contributes to Harms of Oppression ....................................... 170 6.4.1 Discursive Inequity and the Harm of Powerlessness .............................................. 170 6.4.2 Norm-Conditional Communication and the Choice Between Cultural Imperialism or Marginalisation ................................................................................................................... 175 6.5 Discursive Inequalities Contribute to Domination when they Block Deliberation .... 180 6.6 Conclusion .................................................................................................................. 182 Chapter 7: An Analysis of Gendered Asymmetries in Political Participation .....................185 7.1 Literature ..................................................................................................................... 186 7.1.1 A Gender Gap in Political Participation ................................................................. 186 7.1.2 Discursive Participation (Or Political Talk) ........................................................... 191 7.2 Methods and Data ....................................................................................................... 196 7.3 Findings....................................................................................................................... 201 7.3.1 Main effects ............................................................................................................ 201 7.3.2 Interaction effects.................................................................................................... 207 7.4 Discussion ................................................................................................................... 219 7.5 Conclusion .................................................................................................................. 221 Chapter 8: An Analysis of Gendered Asymmetries in Communicative Influence ..............223 x 8.1 Literature: Discursive Inequity and Women’s Internal Exclusion ............................. 224 8.2 Methods....................................................................................................................... 228 8.2.1 Research Design...................................................................................................... 228 8.2.2 Data and Analysis ................................................................................................... 231 8.3 Findings....................................................................................................................... 233 8.4 Discussion ................................................................................................................... 243 8.5 Conclusion .................................................................................................................. 245 Chapter 9: Achieving Discursive Equality ..............................................................................249 9.1 Structures and Practices that Engender Symmetrical Empowerments ....................... 250 9.1.1 Engineering Micro-Institutional Communicative Processes .................................. 256 9.2 Circumventing Normal Procedures: Agonism and Escape ......................................... 263 9.3 Conclusion .................................................................................................................. 266 Chapter 10: Conclusion .............................................................................................................268 10.1 Limitations .................................................................................................................. 269 10.2 Contributions............................................................................................................... 271 10.3 Future Research .......................................................................................................... 278 References ...................................................................................................................................282 Appendices ..................................................................................................................................315 Appendix A Extra Material for Chapter 7 .............................................................................. 315 A.1 Distributions of Independent Variables .................................................................. 315 A.2 Main Effects ............................................................................................................ 316 A.3 Interactions .............................................................................................................. 322 Appendix B Extra Material for Chapter 8 ............................................................................... 349 xi B.1 Experimental Treatment.......................................................................................... 349 B.2 Self-Esteem Measure .............................................................................................. 350 B.3 Implicit Association Test ........................................................................................ 351 B.4 Predicted Marginal Effects ..................................................................................... 353 xii List of Tables Table 7-1 Gendered variation in communicative political participation hypothesis list ............ 195 Table 7-2 Distributions of dependent variables .......................................................................... 198 Table 8-1 Distribution of the dependent variable ....................................................................... 231 Table 8-2 Distributions of independent variables ....................................................................... 233 Table 8-3 Mean likelihood of mind change, by counterargument gender condition .................. 234 Table 8-4 Ordered logistic regression coefficients and odds ratios for likelihood of changing one’s mind after receiving a counterargument (three models) ................................................... 236 Table 8-5 Self-esteem scores, by respondent gender .................................................................. 245 Table A-1 Distributions of independent variables ...................................................................... 315 Table A-2 Baseline model (main effects) logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk ...................................................................... 316 Table A-3 Resource model (main effects) logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk ................................................. 317 Table A-4 Full model (main effects) logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk ...................................................................... 319 Table A-5 Predictive margins for gender (main effect) .............................................................. 320 Table A-6 Discreet change in gender (effect of being a woman) ............................................... 321 Table A-7 Full model with gender and ethnicity interaction logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk ....................................... 322 Table A-8 Predictive margins for the interaction between gender and ethnicity ....................... 324 Table A-9 Discreet change in gender (effect of being a woman), by ethnicity .......................... 325 xiii Table A-10 Discreet change in ethnicity (effect of being a visible minority), by gender .......... 325 Table A-11 Full model with gender and ethnicity interaction logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk ....................................... 326 Table A-12 Predictive margins for the interaction between gender and poverty ....................... 328 Table A-13 Discreet change in gender (effect of being a woman), by poverty .......................... 328 Table A-14 Discreet change in poverty (effect of being poor), by gender ................................. 329 Table A-15 Full model with gender, working for pay, and young kids interaction logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk............................................................................................................................................... 330 Table A-16 Predictive margins for the interaction between gender, employment status, and the presence of young children in the home ..................................................................................... 332 Table A-17 Full model with gender and civic engagement interaction logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk ....................... 333 Table A-18 Predictive margins for the interaction between gender and civic engagement (volunteering for a civic association) .......................................................................................... 335 Table A-19 Discreet change in gender (effect of being a woman), by civic engagement .......... 335 Table A-20 Discreet change in civic engagement (effect of volunteering for a civic association), by gender ..................................................................................................................................... 336 Table A-21 Full model with gender and social trust interaction logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk ................................. 337 Table A-22 Predictive margins for the interaction between gender and social trust (most people can be trusted, or you can’t be too careful) ................................................................................. 339 Table A-23 Discreet change in gender (effect of being a woman), by social trust ..................... 339 xiv Table A-24 Discreet change in trust (effect of agreeing that most people can be trusted), by gender .......................................................................................................................................... 340 Table A-25 Full model with gender and political interest interaction logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk ....................... 341 Table A-26 Predictive margins for the interaction between gender and political interest .......... 343 Table A-27 Discreet change in gender (effect of being female), by political interest ................ 343 Table A-28 Discreet change in political interest (effect of high interest), by gender ................. 344 Table A-29 Full model with gender and political efficacy interaction logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk ....................... 345 Table A-30 Predictive margins for the interaction between gender and political efficacy ........ 347 Table A-31 Discreet change in gender (effect of being female), by political efficacy ............... 347 Table A-32 Discreet change in political efficacy (effect of high efficacy), by gender .............. 348 Table B-1 Initial policy question and counterarguments ............................................................ 349 Table B-2 Rosenberg self-Esteem scale wording and coding .................................................... 350 Table B-3 Mean difference on Rosenberg Self-Esteem (RSE) scale, by respondent gender ..... 350 Table B-4 Implicit association test stimulus materials ............................................................... 351 Table B-5 Mean differences on d-scores between male and female respondents ...................... 351 Table B-6 Mean D-scores, by respondent gender and measure of explicit gender attitude (mean responses to the statement: “society would be better off if more women stayed home.”) ......... 352 Table B-7 Marginal predicted probabilities for main model: five categories of the dependent variable, by counter argument condition .................................................................................... 353 Table B-8 Marginal predicted probabilities for model 2: five categories of the dependent variable, by gender and counterargument interaction ................................................................. 353 xv Table B-9 Marginal predicted probabilities for model 3: five categories of the dependent variable, by gender, counterargument, and self-esteem interaction (for low self-esteem) ......... 354 Table B-10 Marginal predicted probabilities for model 3: five categories of the dependent variable, by gender, counterargument, and self-esteem interaction (for average self-esteem) ... 355 Table B-11 Marginal predicted probabilities for model 3: five categories of the dependent variable, by gender, counterargument, and self-esteem interaction (for high self-esteem) ........ 355 xvi List of Figures Figure 7-1 Full model (main effects) logistic regression results, predicting group deliberation (95% confidence intervals) ......................................................................................................... 223 Figure 7-2 Full model (main effects) logistic regression results, predicting everyday political talk (95% confidence intervals) ......................................................................................................... 224 Figure 7-3 Full model (main effects) logistic regression results, predicting Internet political talk (95% confidence intervals) ......................................................................................................... 225 Figure 7-4 Predicted probabilities men and women participate in group deliberation (main effects)......................................................................................................................................... 226 Figure 7-5 Average marginal effects of ethnicity (the effect of being a visible minority on the predicted probability of participating in discursive politics) by gender, for three forms of discursive politics (with 95% confidence intervals) ................................................................... 208 Figure 7-6 Average marginal effects of poverty (the effect of being poor on the predicted probability of participating in discursive politics) by gender, for three forms of discursive politics (with 95% confidence intervals) ................................................................................................. 210 Figure 7-7 Average marginal effects of civic engagement (the effect of civic engagement on the predicted probability of participating in discursive politics) by gender, for three forms of discursive politics (with 95% confidence intervals) ................................................................... 212 Figure 7-8 Average marginal effects of social trust (the effect of having high social trust on the predicted probability of participating in discursive politics) by gender, for three forms of discursive politics (with 95% confidence intervals) ................................................................... 214 xvii Figure 7-9 Average marginal effects of political interest (the effect of having high political interest on participating in discursive politics) by gender, for three forms of discursive politics (with 95% confidence intervals) ................................................................................................. 216 Figure 7-10 Average marginal effects of political efficacy (the effect of having high political efficacy on participating in discursive politics) by gender, for three forms of discursive politics (with 95% confidence intervals) ................................................................................................. 218 Figure 8-1 Contrasts of predictive margins of man and woman counterargument..................... 238 Figure 8-2 Predicted probabilities for outcome “somewhat likely” to change mind, for man and woman counterargument, by respondent gender ........................................................................ 240 Figure 8-3 Predicted probabilities for outcome “somewhat likely” to change mind, for man and woman counterargument, by respondent gender and self-esteem .............................................. 243 xviii Acknowledgements I have many colleagues and friends to thank for making the dissertation writing process possible, and even – on occasion – fun. First and foremost, I would like to thank my supervisor, Mark Warren. Mark inspired and encouraged my work, helped me think through even the most half-baked ideas until they came out clever, financed me through research assistance positions, found me funding for conferences and summer schools, collaborated with me and helped me publish, and reminded me to take breaks and look after my health. I could not have asked for a more perfect supervisor. I would also like to thank my committee members, Andrew Owen and Simone Chambers, for their comments. Andrew’s help designing and implementing my survey experiment and analysing the data was invaluable. My research was financially supported by a Social Sciences and Humanities Research Council Doctoral Fellowship and by UBC. Many UBC colleagues helped me with my dissertation, and with the side projects that kept my attention deficit brain busy. First and foremost, my friend and co-author, Sule Yaylaci, whose passion, strength, and brilliance inspires me. I am grateful to Fred Cutler, who supervised Sule’s and my study of tutorial deliberation, and found us funding. I thank Alison James, Grace Lore, Elizabeth Mann, and Dave Moscrop for collaborating with me on publications. Other fellow grad students helped me develop my ideas through discussions and friendship, particularly Afsoun Afsahi, Jen Allan, Justin Alger, Katrina Chapelas, Alicia Luedke, Eric Merkley, Miriam Matejova, Spencer McKay, Andrea Nuesser, Dom Stecula, and Dan Westlake. I thank the participants and discussants at the Canadian-Comparative Workshop for their comments and feedback, especially Alan Jacobs. I also thank Paul Quirk and the graduate students who kept the Public Opinion Lab running (Sule, Eric, Go Murakami, and Dom). Other xix professors who helped me during informal chitchat include Barbara Arneil and Richard Johnston. I am grateful to undergraduate research assistants Arthur Nozak and Scott Fletcher, and to the undergraduate students who participated in the tutorial deliberation study and my survey experiment. Finally, I owe a huge debt of gratitude to the department’s administrative staff, particularly Josephine Calazan, Dory Urbano, Amy Becir, and Rebecca Monnerat. I am also thankful for the network of colleagues outside of UBC who helped me with their feedback, comments, and encouragement. Special thanks to André Bächtiger, whose co-authorship and feedback helped develop some of the ideas in this dissertation. I received feedback from many conference discussants, notably Jason Barabas and Tali Mendelberg (whose comments improved what become Chapter 8 of my dissertation). John Dryzek also provided feedback on a book chapter that informed part of this dissertation. I am grateful to the members of the Participedia team, whose conversations and debates at meetings and over dinners were thoughtful and fun. I would also like to thank all the Grandview-Woodland Citizens’ Assembly members who inspired me with their dedication to local democracy (and patiently took my long surveys), as well as MASS LBP for supporting Mark’s and my study of the process. I could not have done this without the love and support of my spouse, Andrew Sage. Andrew was essential for reminding me there is life – love, family, fitness, the great outdoors – outside academia. I also thank my fam in-law, Sandra, Dan, Anthony, Connor, and Brody Sage. Finally, I’m grateful for the village of strong, inspiring women whose friendship over the years got me to, and through the Ph.D.: Magdalena Abassi, Raniela Bovey, Breanne Cravagna, Keely Dakin, Christina Eng, Stefania Evangelista, Mirela Gakovic, Manie Ho, Kari Payne, Danielle Rulens, Krista Silverwood, and Glynda Tentes. And of course, my dog, Sophie, who never gets tired of me. Sophie, I promise I will never grow weary of being compassionate towards you. xx Dedication In memory of my grandmother, Edith, who – although she never made it past primary education – made me promise to “stay in school.” 1 Chapter 1: Introduction To count as democratic, social systems must empower the inclusion of people affected by collective endeavours to participate in political practices that contribute to self-development and self- and collective-rule (Young 2000, 2011; M. E. Warren 2017). I focus on the importance of talk for democracy, and show how everyday communication – particularly mundane talk – is essential for building the kinds of relational ties and facilitative attitudes that allow speakers to withstand the strain of disagreement endemic to deliberation. Democratic processes always begin with symmetrical empowerments that entail inclusions, because the aims of forming collective opinions and making collective decisions are undermined when those affected by collective endeavours are excluded from participating in, or influencing, them (Fung 2013; Goodin 2007; Young 2000). Because structural inequalities can prevent disempowered social group members from participating in or influencing social and political practices (including talk), structural inequalities can undermine democratic processes and contribute to harms of oppression and domination. Focusing on gender inequality, I show that asymmetrical empowerments contribute to an ongoing gender gap in discursive political participation. Furthermore, I show that even when women are formally included in talk, asymmetrical empowerments undermine women’s discursive influence. My dissertation has three broad aims. First, building on Mark Warren’s (2017) problem-based, systems approach to democratic theory, I consider the general problems social systems need to solve to count as “democratic”: empowered inclusion, collective opinion and agenda formation, and collective decision-making. I develop the democratic systems’ approach by clarifying the relationship between equality and inclusion, and articulating why empowered 2 inclusion must functionally precede other democratic goals. I also expand the approach by considering how the practice of “talk” – including, but not limited to deliberation – can achieve democratic functions. Focusing on using talk to achieve democratic functions, I build-out the collective opinion and agenda formation function. I identify five interrelated, but distinguishable components comprising this general function: the social integration, reciprocity, personality, learning, and legitimation aims. I describe how talking helps achieve each component function. My second goal is to clarify the challenges that structural (group-based) inequality present for using talk to achieve democratic goals. Inequality and exclusions can be thought of as “systematic constraints” on social group members that prevent them from developing or exercising their individual capacities, and prevent them from participating in political practices to influence the norms and laws that affect them (Young 2011, 41). When inequality acts as disabling constraints that preclude social group members from participating in or influencing communicative practices, inequality and exclusion prevent talking from achieving the five, interrelated collective opinion and agenda formation functions. Furthermore, these disabling constraints can contribute to, or reinforce harms of oppression. Finally, inequality and exclusion can prevent misunderstanding or disagreement from being solved through deliberation, and so contribute to domination. My third goal is to empirically illustrate the presence and extent of discursive inequalities, or asymmetries of communicative participation and influence. Focusing on gender, and considering the ways gender intersects with ethnicity and poverty, I use data from the Canada Elections Study (2015) to show there is an ongoing gender gap in discursive participation. Even controlling for structural factors (such an income, working for pay, education, and the presence of young children in the home), social capital (civic engagement and 3 social trust), and political capital (political interest and political efficacy), women are significantly less likely than men to participate in the most demanding form of discursive participation: group deliberation. Of course, even if historically disempowered group members – such as women –participate in informal conversations about matters of collective concern or group deliberations, they may not have the same discursive influence as historically empowered social group members. That is, listeners may ignore or give less weight to their utterances. Using data from a novel vignette experiment, I show that there is an ongoing gender gap in discursive influence. I find that, ceteris paribus, people are more likely to accept a man’s counterargument than a woman’s identical counterargument. This means that even when women are nominally included in political talk, gendered asymmetries in communicative influence can undermine women’s contributions and influence. In the first section of this introductory chapter, I state my research questions and sketch the background information required to understand them. In this section, I outline the relationship between equality, inclusion, and practices that serve democratic functions, and describe the problem of inequality for talk and democracy. I then outline each of my chapters. I conclude my introductory chapter by highlighting my dissertation’s most important contributions to democratic theory. My contributions include, firstly, identifying the importance of talk – and particularly the importance of everyday talk – for democracy. Secondly, building on Pierre Bourdieu’s work, I offer a theoretical account of how material and social inequalities structure agents’ cognitive and motivational dispositions in ways that engender internal exclusions, even in the absence of formal prohibitions to participation and influence. Finally, I offer novel 4 evidence of ongoing gender gaps in discursive participation and influence in Western democracies. 1.1 Research Questions and Background 1.1.1 Why is More Democracy Better, and More Oppression and Domination Worse? Democracy is defined by the degree to which people are empowered to participate in self-development and self- and collective-rule. Drawing on Young (2011, 37), I argue that democracy is characterised by two general values: (1) the ability to develop and exercise one’s capacities and express one’s experiences; and (2) the ability to participate in determining one’s actions and the conditions of one’s actions. These universal values are connected to the notion of justice derived from Jürgen Habermas’s communicative ethics which stipulates that – for norms or social conditions to be just – everyone affected by them must, in principle, have “an effective voice” in their creation, and be able to agree to them without coercion (Young 2011, 38). Democracy is linked to this conception of justice because democratic conditions – conditions that promote the universalist values of self-development and collective-rule – can be expected to produce opinions, decisions, norms, and social conditions that are more attentive to the needs and preferences of those affected by them. In short, democratic conditions can be expected to produce functionally better (e.g., more responsive), and normatively better (i.e., more just) opinions, decisions, norms, and social conditions. What I identify as two universalistic democratic principles map onto the problems Young (2011) identifies as oppression and domination. Following Young (2011, 38), I define 5 oppression as institutional constraints on self-development, “which inhibit people’s ability to play and communicate with others or to express their feelings and perspectives on social life in contexts where others can listen.” Domination refers to institutional constraints on self- and collective-rule, “which prevent people from participating in determining their actions or the conditions of their actions.” Oppression and domination are structural concepts, in that they refer to institutional constraints that prevent certain social group members from participating in self-determination and self- and collective-rule. Oppression does not always entail domination: institutional conditions may prevent agents from using skills in socially recognised settings, while leaving agents some opportunities to determine their actions and the conditions of their actions. But domination almost certainly entails oppression. When institutional conditions prevent agents from determining their actions or the conditions of their actions, their capacity for self-development is circumscribed. Social systems characterised by oppression and domination can be expected to produce functionally worse (e.g., less responsive), and normatively worse (i.e., less just) opinions, decisions, norms, and social conditions. 1.1.2. What must a social system do to count as “democratic”? Warren (2017) identifies three broad, “normatively necessary” functions social systems must achieve to count as democratic: empowered inclusion, collective opinion and agenda formation, and collective decision-making. The notion of a “function” is both normative and systematic. The concept of a function is normative “because identifying a democratic function is the same as claiming that a system should function in ways that support democratic ideals” (M. 6 E. Warren 2017, 42). The notion is systematic in that it frames the question in terms of practices’ context-specific “normatively desirable consequences or outcomes (functions).” Functions can be distinguished from generic political practices that serve democratic functions. Generic political practices are ideal-type social actions (in the Weberian sense of socially meaningful actions), commonly “organized or enabled by institutions that serve democratic functions: recognising, resisting, representing, deliberating, voting, joining, and exiting” (M. E. Warren 2017, 43). In my dissertation, I focus on the functions of inclusion and the collective opinion and agenda formation. It is important to note that democratic processes always begin – functionally and normatively – with inclusion. This is because, as my discussion of democracy, oppression, and domination starts to show, the aims of forming collective opinions and wills, and making collective decisions, are undermined if those affected by collective endeavours are excluded from participating in them (Fung 2013; Goodin 2007; Young 2000). A degree of equality means that empowerments are distributed in ways that enable those affected by collective endeavours to communicatively link their personal preferences and needs to collective opinions and agendas, and to link self-government with collective governance. To assess political systems’ democratic problem-solving capacities related to the collective opinion and agenda formation, I expand the collective opinion and agenda formation function. I identify five component functions comprising this broader function: social integration, reciprocity, personality, learning, and legitimation goals. As Warren (2017) notes, talking (including deliberating) is the most important and effective tool for connecting individual judgments to collective agendas. Under the rubric “talk” I include everyday talk about matters of collective concern (Jacobs, Cook, and Delli Carpini 7 2009; J. Mansbridge 1999), mundane talk (not about matters of collective concern), as well as deliberation. Deliberation refers to a rarer form of communication that involves trying to temporarily suspend pre-existing judgments to solve disagreement and misunderstanding through mutual justification (Habermas 1990). 1.1.2 What Kind Of “Equality” Empowers Inclusion in Democratic Processes? Democratic processes begin with inclusion, and inclusion requires a degree of equality. Equality, broadly understood, can be thought of as symmetrical empowerments that enable people to participate in and influence individual and collective judgments and decisions by safeguarding personal and collective freedoms (such as freedom from government prohibitions on speech and other forms of social and political participation), as well as through positive supports (such as welfare supports) to help develop individual and collective capacities. Equality contains two distinct values (B. Williams 1972; see also Beauvais 2017; Beauvais and Bächtiger 2016). On the one hand, universal equality (or universal moral equality), requires abstracting from social circumstances, and recognising the fundamental sameness of common humanity. The value of universal equality involves treating people as if they share a universal starting point (such as the same baseline of moral worth) and have the same fundamental needs (such as the need for life and freedom). However, group members do not “arrive at life’s starting lines” with equal resources (M. S. Williams 2000, 60). The second value of equality is equity, which involves attending to social circumstance, and recognising group-based differences between people. 8 When I refer to equality I am referring to “structural equality,” or equality between salient social group members (rather than, say, idiosyncratic differences between individuals) (Harell and Stolle 2010). Social groups are an expression of social relations, since “a group only exists in relation to another group” (Young 2011, 43). As such, structural equality between social groups denotes egalitarian social relations. Salient social groups might include gender, class linguistic, and ethnic or cultural, racial, or religious groups. Both the value of universal equality and the value of equity can – and should – be accommodated in democratic systems, through different political practices and institutional arrangements. For instance, democratic systems can enable empowerments through a baseline of universal equality enshrined as freedom from government interference in political practices, such as the freedom to organise, speak, and engage in other political acts. But in political systems where historical inequalities result in continuing asymmetries in social group members’ abilities to use these universal empowerments, it may be necessary to pursue the value of equity to promote the empowered inclusion of historically disadvantaged group members. For instance, targeted “get out the vote” campaigns may be designed to motivate the poor, new immigrants, or other historical marginalised groups to vote in general elections. 1.1.3 What are the Consequences of Inequality for the Democratic Potential of Talk? Clearly, gaps in political rights between social group members entail asymmetrical power relations and formal political exclusions. For instance, limiting political participation rights to white, property-owning men obviously enhances white, property owning men’s political power relative to other social group members’ political power. And the prohibitions preventing 9 disempowered group members from legally participating in politics clearly entail formal (external) exclusions from voting, running for office, or other political activities. Large income-gaps or resource-gaps between social group members also entail asymmetrical power relations and external exclusions from political activities. For instance, the poor, who are unable to afford the cost of “pay-for-access” fundraisers, are excluded from attending functions where they could meet politicians and other influential members of society. Another example might be working mothers, whose double-day and finite free time precludes them from volunteering for a political party, or joining a civic association. The exclusions that unequal political rights, and economic or material inequalities, entail are “external” in that they preclude disempowered social group members from participating in politics at all (Young 2000). These inequalities prevent disempowered social group members from legally casting a ballot, affording a donation, or having time to associate with like-minded people. Most democratic theories consider the problem of external exclusion, and “call for limiting the influence of wealth or position on the ability to participate in a democratic process” (Young 2000, 55). However, inequality can also structure social cognition to engender internal exclusions. Internal exclusion refers to when, even in the absence of formal political or economic asymmetries, socio-cognitive features shape how people view themselves and other social group members to engender asymmetries in political participation and influence. Consider how those who have historically been excluded from the franchise – women, the poor, or non-whites – may internalise a sense of “feeling out of place” in politics (e.g., Bourdieu 2000). People who feel out of place in politics are less likely to participate in practices such as voting or political talk, even after formal prohibitions to their participation have been lifted. For instance, research shows that 10 young women are far less likely than young men to envision themselves running for office, and this impacts their political behaviour as adults (Fox and Lawless 2011, 2014). Inequality poses at least three interrelated problems for talk’s democratic potentials. First, inequality can simply prevent talking from achieving the collective opinion and agenda formation functions. Second, under conditions of inequality, talking can contribute to harms of oppression. And third, inequality can block deliberation. It is problematic when inequality prevents misunderstandings or disagreement from being solved through deliberation, because raising complaints about unjust (or unjustifiable) norms and laws is essential to self- and collective-rule. When asymmetrical empowerments block the “forceless force” of the better argument from solving communicative breakdowns – such as misunderstandings or disagreements – or impede efforts to problematize normative assumptions and bring them to the level of explicit disagreement in the first place, inequality contributes to systems of domination. 1.1.4 What are the Consequences of Discursive Inequality for the Democratic Potential of Talk? I refer to unjustifiable, structural asymmetries of communicative participation or influence as discursive inequality, and focus on how cognitive schema can engender discursive inequality. When material inequalities and external exclusions structure cognition, cognitive schemes of perception and appreciation act as informal inequalities that engender internal exclusions that prevent disempowered social group members from taking advantage of formally equal opportunities to participate in talk, and can generate unjustifiable asymmetries in 11 disempowered social group member’s communicative influence. I identify two distinct forms of discursive inequality: discursive inequities and norm-conditional communication. Discursive inequities are disabling constraints that entail internal exclusion, undermining disempowered social group members’ communicative influence when they do participate in talk. A review of the literature on small-group deliberation suggests that members of historically disadvantaged groups are less likely to speak in discussions, and feel less influential (Mendelberg 2002). When women and men talk politics, both women and men are more likely to perceive men conversation partners as more knowledgeable about politics than women, regardless of the conversation partners’ objective political knowledge (Mendez and Osborn 2010, 270). Social stereotypes and structural variation in self-esteem are particularly problematic discursive inequities that undermine disempowered social group members’ communicative influence. The second type of discursive inequality I identify, norm-conditional communication, refers to situations when the terms of successful communication (such as norms regulating what makes an utterance successful or a claim persuasive) are harmful to social group members’ identities. Norm-conditional communication discourages disempowered social group members from participating in communication at all (producing unjustifiable asymmetries of communicative participation), and typically entails discursive inequities (and unjustifiable asymmetries of communicative influence) when disempowered social group members do participate. Discursive inequalities can reinforce existing systems of oppression, since the disempowered are less likely to influence or participate in communicative practices. 12 1.1.5 Where do Discursive Inequalities Come From? To understand how social and material inequalities can structure cognitive schemes of perception and appreciation to engender internal exclusions and asymmetries of communicative participation and influence, it is important to understand humans’ orientation to sharing intersubjective understandings with others, and their capacity to internalise, and so practically understand, the social and political world. Like many other animals, humans have cognitive hardware that enables a neural experience of intersubjectivity (to share automatic, sensorimotor understandings, and participate in emotional contagion) (e.g., Gallese 2001; Gallese, Eagle, and Migone 2007; Gallese and Goldman 1998; Iacoboni 2009; Iacoboni et al. 2005; Kohler et al. 2002; Rizzolatti and Craighero 2004). However, humans are both capable and motivated to go beyond this temporary, corporeal experience of intersubjectivity. Agents expand and give permanence to the capacity for sharing intersubjective understandings and intentions by inscribing them in language games and other social institutions. Inscribing shared understandings and intentions in language games enables more cognitively-oriented mind reading skills (attributing the mental states of others) and more expansive collaborative activity (e.g., Ardila 2008; Burkart, Hrdy, and Van Schaik 2009; Carpenter 2006; Hrdy 2011; Tomasello 2010; Tomasello and Carpenter 2007). Not only do agents inscribe mutual understandings in language games and other institutions, they inscribe them in “bodies,” including the mind’s cognitive and motivational structures, thus internalising a practical comprehension world in which they have been inserted (Bourdieu 1977, 1990, 2000). Human plurality – that humans are both similar to one another, and yet distinct from every other person – means that humans are both capable of understanding one another, but also 13 need to make themselves understood (Arendt 2013). Human plurality motivates two fundamental psychological needs: belonging/similarity (on the one hand) and distinctiveness/differentiation (on the other) (Brewer 2003, 2011; Brewer and Pickett 1999; Leonardelli, Pickett, and Brewer 2010). The simultaneous needs for belonging and distinctiveness motivates agents to create and participate in spaces of public appearance, including collective social identities and what Bourdieu (1977, 1984, 1990, 2000) describes as “fields” of social play. Spaces of public appearance allow agents to simultaneously be the same (e.g., as fellow group members, or fellow field participants) and different (e.g., from out-group members, or from other field participants and non-participants). Spaces of public appearance also create contexts for mutual understanding and recognition. However, the desire for mutual understanding can motivate agents to forcibly co-opt others into participating in their social systems, to expand and give greater permanence to the shared intentionality inscribed in their social systems’ language games, institutions, and spaces of public appearance. This is particularly true under conditions of inequality and asymmetrical empowerments, when forcibly co-opting new participants reinforces or increases the co-opting group’s power. Forcibly co-opting new participants into a social system’s language games, institutions, and spaces of public appearance always entails norm-conditional communication and discursive inequities. This kind of forcible cultural and linguistic co-optation always entails forcing new recruits’ participation into the co-opting group’s language games, according to the terms inscribed in the co-opting group’s social systems. Furthermore, because experiences in the social and material world structure semantic associations and personalities, structural inequalities often entail the association of empowered group members with positive attributes that reflect their empowerments (such as authority, 14 intelligence, strength, and purity), and the association of disempowered group members with negative attributes that reflect their disempowerments (such as subordination, stupidity, weakness, and impurity) (Cuddy, Fiske, and Glick 2008; Fiske et al. 1999; Macrae, Stangor, and Hewstone 1996; Operario and Fiske 2001). In other words, social and material inequalities become embedded in cognitive structures as discursive inequities, which include harmful stereotypes and distorted personalities. These discursive inequalities – created through structural inequalities (such as patriarchy, racial and cultural hierarchies, or class inequality) – act as disabling constraints that undermine disempowered social group members’ capacity for self-development, which contributes to harms of oppression. Discursive inequalities also act as disabling constraints that undermine disempowered social group members’ capacity to act or choose the conditions of their actions – preventing social group members from influencing social systems’ language games, spaces of public appearance, and other institutions (including political institutions) that affect them – which contributes to harms of domination. 1.1.6 How can we Achieve Discursive Equality? Different practices and institutional arrangements can be used to address structural inequalities, and bring democratic systems closer to the regulative ideal of discursive equality, or symmetrical empowerments that entail the universal inclusion and equitable influence of all social group members in everyday talk and deliberation. Universal equality – freedoms from unjustifiable constraints and positive political participation rights – should be consecrated in law. And institutional practices must be designed to protect the judiciary’s independence in law and practice, such as through independent judiciary appointment processes. And mechanisms to 15 achieve equity – such as affirmative hiring practices, or needs-based social welfare redistributions – should also be inscribed in law. Exactly which freedoms (and what constitutes an unjustifiable constraint), participation rights, and institutional practices and social policies to promote equity cannot be determined in advance by social scientists. Rather, they should be determined by those affected by them. Efforts to achieve or safeguard symmetrical empowerments help create a context in which talk can achieve its democratic opinion and agenda formation functions in everyday contexts, or what James Tully (2002) calls “normal” activity. This backdrop of empowerments even helps when ongoing structural inequalities (including discursive inequalities) prevent talk or other political practices from functioning democratically, because agents can use the “regular procedures” to address disabling constraints and injustices (Tully 2002). For instance, agents can address discrimination in hiring practices by filing suit in a court of law or human rights tribunal, or can address problematic normative assumptions inscribed in language games or other cultural artifacts in social systems by making them explicit and challenging them through public deliberation. When this backdrop of empowerments is absent (or does not go far enough to address disabling constraints or ongoing injustices), agents can work outside the normal procedures, such as through confrontation and protest, or through escape. 16 1.2 Chapter Summaries 1.2.1 Chapter 2 Equality and Democracy In Chapter 2, I explain what I mean by democratic systems, and explain why the “systems” approach is superior than the “models” approach for studying democracy. I explain what I mean when I say that social systems are democratic to the extent that people are included in political practices (such as voting or deliberating), can communicatively link personal preferences into collective opinions and agendas, and are empowered to turn collective agendas into collective decisions (M. E. Warren 2017). Drawing on Young (2000), I elucidate the relationship between structural equality and inclusion in social and political practices, and explain why democratic processes begin with empowered inclusion. I explain and build-out the collective opinion and agenda formation function, describing five, interrelated component functions: the social integration, reciprocity, personality, learning, and legitimation component functions. I also discuss the collective decision-making function. I then review different tools – resisting, talking, representing, voting, joining, and exiting – that can be used to achieve these interrelated democratic functions. Finally, I consider the problem of inequality for talk and democracy, and introduce the idea that different forms of inequality – political, material, and interactional (including discursive) inequalities – can prevent political practices from achieving democratic functions. 17 1.2.2 Chapter 3 Talk and Collective Opinion and Agenda formation In Chapter 3, I focus on how talk achieves collective opinion and agenda formation functions, and describe the problem of situations in which inequality prevents talking from achieving this set of democratic functions. I first explain what I mean by the generic practice of “talk,” distinguishing between deliberation and everyday talk (including everyday talk about matters of collective concern, and everyday talk about mundane issues). I elucidate the idea that the collective opinion and agenda formation is comprised of at least five distinguishable but interrelated component functions, namely: the social integration, reciprocity, personality, learning, and legitimation component functions. I elaborate why each of these components are necessary for democratic collection opinion and agenda formation, and explain why talk is so important for achieving each of them. I make the case that everyday talk (particularly everyday mundane talk) is best for achieving the democracy-supporting social integration, reciprocity, and personality goals, while deliberation is best for achieving democracy-supporting learning and legitimation aims. I also explain how different forms of inequality can prevent talk from achieving each of the collective opinion and agenda formation functions. 1.2.3 Chapter 4 Practical Knowledge In Chapter 4, I draw on Pierre Bourdieu to explain the “self” as “practical knowledge” of the material and social world. I discuss how the mind is conditioned by material social experiences, to produce a “self” who internalises and so practically understands the world into which they have been inserted. Borrowing from Bourdieu, I describe the practical knowledge of 18 the world embedded in the mind’s cognitive and motivational structures as “habitus.” Drawing on the “Leipzig School” of evolutionary anthropology, I explain how agents’ capacity for mutual understanding – the capacity and motivation to share understandings and intentions with others – allows them to inscribe mutual understandings into language, institutions, and habitus’ cognitive and motivational structures, and so to internalise a practical comprehension of the material and social world (e.g., Hrdy 2011; Rakoczy, Warneken, and Tomasello 2008; Tomasello 2015, 2009a, 2009b, 2010; Tomasello and Carpenter 2007; Tomasello et al. 2005; Tomasello, Kruger, and Ratner 1993). Understanding how the world conditions bodies is central for the aim of empowering people to participate in and influence communicative processes of collective agenda formation. This is because conditions of structural inequality can systematically engender selves with weakened capacities to participate in, or influence communicative processes. I also explain the development of a common-sense world that acts as backdrop of mutual intelligibility to enable smooth social interaction and everyday talk. Two factors contribute to the development of a common-sense world: isomorphism (congruence) between habitus and the social and material world, and homology (similarities) between the habitus of similarly situated people (Bourdieu 1977, 2000). Finally, I describe how the common-sense world can act as a conservative counterweight to social change, and discuss the role of human agency for changing social systems’ structures (including the structures defining the common-sense world). 19 1.2.4 Chapter 5 Spaces of Public Appearance In Chapter 5, I describe how social systems become differentiated into spaces of public appearance (into social identities and fields), and outline the implications of societal differentiation for distributing and maintaining shared intentions – particularly, for distributing and maintaining power relations – inscribed in language games, institutions, or other cultural features of social systems. I draw on social psychology to explain the motivational and cognitive bases for social identification. I draw on Bourdieu to describe another form of societal differentiation: differentiation in fields (overlapping, rule-bound systems of social play). I consider how power relations can be created and maintained by inscribing them in social systems’ structures (language games, institutions, spaces of public appearance, and habitus). In Chapter 5, I also describe how sets of government institutions (which collectively comprise political systems) are created to direct and protect the shared intentionality inscribed in social systems’ structures. I elucidate the notion that social group members may be motivated to forcibly co-opt new agents into participating in the shared intentionality inscribed in their social systems, to expand and give permanence to their social space. As I discuss, this is particularly likely when preserving and expanding social space preserves or expands asymmetrical power relations that benefit co-opting group members. I also discuss the ideal of political systems that direct and protect social systems’ language games, spaces of public appearance, and habitus to engender symmetrical empowerments that entail inclusions and enable human self-development and self-rule. 20 1.2.5 Chapter 6 Inequality, Oppression, and Domination The purpose of this chapter is to show how, under conditions of inequality, talk can contribute to harms of oppression, and show how inequality can block deliberation from occurring at all, which contributes to domination. I first discuss the relationship between inequality and exclusion, and then I discuss how exclusions contribute to oppression and domination. I also explain in more detail what I mean when I say that inequality can become embedded in thought and perception, describing how features of the social and material world practically structure subjectivity to engender interactional inequalities, or asymmetries in agents’ capacities to participate in or influence social, economic, or political practices. Because of my focus on using talk to achieve democratic collective opinion and agenda formation functions, I focus on the form of interactional inequality I call discursive inequality, which refers to subtle and often subconscious asymmetries in opportunities to participate in, or influence, communication. I distinguish between two forms of discursive inequality: discursive inequity and norm-conditional communication. Discursive inequities refer to asymmetries in communicative influence, structuring how people perceive themselves, others, and social interactions, contributing to the harm of oppression Iris Young (2009, 2011) refers to as powerlessness. Norm-conditional communication refers to when the terms of mutual understanding are harmful to an agent’s self-identity. I explain how norm-conditional communication presents agents with an unenviable choice: participate in norm-conditional communication that is harmful to their self-identities and risk the harm of oppression Young refers to as cultural imperialism (and risk discursive inequity and powerlessness), or exclude themselves from talk, and risk the harm of 21 marginalisation. I also discuss how structural inequalities – including discursive inequalities – can block deliberation from occurring at all (which contributes to domination). Of course, whether norm-conditional communication and discursive inequity engender asymmetries in social group members’ discursive participation and influence are empirical questions. I address these questions in Chapter 7 and Chapter 8, focusing more specifically on gendered variation in discursive participation and influence. 1.2.6 Chapter 7 An Analysis of Gendered Asymmetries of Discursive Participation In today’s Western democracies, most formal prohibitions on political participation – such as legal prohibitions preventing women, minorities, or poor people from voting or running for office – have been removed. As such, my theoretical discussion of the problems of inequality – of the way inequality prevents talking from achieving its functions, and contributes harms of oppression and domination – only really matters for Western democracies if I am correct to suggest that inequalities can become internalised, and can continue to prevent historically disempowered groups from participating in and influencing political practices even in the absence of formal prohibitions. In my empirical chapters, Chapter 7 and Chapter 8, I narrow the broad question of whether of whether the historical exclusion of social group members from political practices – including the practice of talk – manifests even in the absence of formal exclusion, by considering whether there are ongoing gender gaps in communicative participation and influence. Narrowing my focus to gender inequality in Chapter 7, I consider whether women participate in discursive politics at the same rate as men, using data from the Canadian Election 22 Studies (2015). I also consider whether gender intersects with other identities and ascribed attributes – such as ethnicity or socioeconomic status – to impact discursive participation in important ways. Although there is a substantial literature on the gender gap in various forms of political participation (voting, protesting, and other political acts) and political capital (political interest, political efficacy, and political knowledge), this chapter is the first systematic study of the gender gap in discursive political participation. I find that women participate significantly less than men in the most demanding form of discursive political participation: group deliberation. I also find that ethnicity intersects with gender in some important ways, although not always in the predicted direction. For instance, I find that minority status significantly reduces the predicted probability visible minority men participate in everyday talk about matters of collective concern (but not the probability that visible minority women participate in everyday political talk). My analysis reveals that other structural factors – such as poverty or working a “double-day” (working for pay and raising young children) – have surprisingly small effects on women’s discursive participation. I find that social and political capital explain more of the variation in discursive participation. However, controlling for social and political capital does not completely account for the gender gap in group deliberation in part because – although civic engagement and political interest are important for increasing everyone’s participation in group deliberation – the effect of civic engagement and political interest on group deliberation is greater among men than it is among women. 23 1.2.7 Chapter 8 An Analysis of Gendered Asymmetries of Discursive Influence Even when historically disempowered groups are present and speaking in informal conversations about politics or more formal group deliberations, there is still a question of whether listeners give weight to disempowered group members’ utterances. That is, can inequalities become internalised, and continue to prevent historically disempowered groups from influencing political practices even in the absence of formal prohibitions? Continuing with my focus on gender inequality in Chapter 8, I present results from the original vignette experiment I created to test whether, else being equal, women have the same communicative influence as men when they speak. Experiment participants were asked their opinion on a non-salient policy question, and regardless of their initial position, were randomly assigned to read a vignette with a counter argument, from either “Michael” or “Jessica.” I find that, ceteris paribus, respondents who received Michael’s counterargument indicate significantly more willingness to revise their initial positions, as compared to respondents who received Jessica’s identical counterargument. There is also some evidence of an interaction between respondent gender and the gender of the speaker presenting the counter argument: Female respondents are more likely to revise their opinions after hearing a man’s counterargument than male respondents. Self-esteem seems to be part of the mechanism driving this effect, as, for instance, females with lower self-esteem are more sensitive to learning about a man’s counter position and females with higher self-esteem are more resistant to hearing a man’s counter position. 24 1.2.8 Chapter 9 Achieving Discursive Equality In this chapter I describe practices, and ways of structuring practices through institutions and spaces of public appearance, to better approximate the ideal of discursive equality. That is, practices, and ways of structuring practices through institutions and spaces of public appearance, that promote equal opportunities for communicative participation and equitable influence in communicative processes. I draw on James Tully’s (2002, 540) description of how social systems’ structures open “a diverse field of potential ways of thinking and acting in response” to those structures. These responses include practices in regular activity (playing by the “rules of the game”) inscribed in social systems’ structures, and challenging or questioning social systems’ structures using “the available procedures of negotiation, deliberation, problem solving, and reform with the aim of modifying” them (Tully 2002, 540). I consider both macro-systemic and micro-institutional approaches for achieving or safeguarding discursive equality. I also consider the problem of when structural inequalities prevent political practices from achieving their democratic functions in systems of domination where there are insufficient or no procedures for changing social systems’ structures. Under conditions of domination, agents have to go outside the available procedures to change social systems’ structures through agonistic practices of confrontation or escape (Tully 2002, see also 2008a, 2008b). 25 1.2.9 Chapter 10 Conclusion In my conclusion, I reflect on the research questions I posed in my introduction, and discuss my dissertation’s contributions to political theory and political science. I also outline my dissertation’s strengths and limitations. Finally, I describe two possible future research directions. In my first future research project, I propose exploring the relationship between talking’s democratic reciprocity and social integration functions in greater to detail. My second future research project will consider gender dynamics in Indigenous agents’ agonistic responses to norm-conditional inclusion in colonial/ post-colonial contexts. 1.3 Conclusion Few researchers have carefully examined the importance of talk – particularly everyday talk – for democratic and deliberative theory. My dissertation represents the most thorough and complete examination of this topic to date. I show how everyday communication, particularly what I refer to as “mundane talk” (talk that is not concerned with social or political issues), is essential for building the kinds of reciprocal attitudes and social ties that allow speakers to withstand the strain of disagreement endemic to deliberating social and political issues. Another important contribution is to show how structural inequalities, including discursive inequalities – asymmetries of communicative participation and influence – entail exclusions. Under conditions of inequality and exclusion, talking is less likely to promote democracy-supporting social ties, or reciprocal attitudes such as mutual respect, trust, empathy, and tolerance. Under conditions of inequality and exclusion, talking is less likely to promote democracy-supporting (healthy, non-26 debased) personalities. Under conditions of inequality and exclusion, talking is less likely to promote learning, and talking is less likely to incur faith in the acceptability of collective endeavours. Worse yet – under conditions of inequality and exclusion, talking can do harm. Whereas under the right conditions talking is an essential tool for achieving democratic aims, under the wrong conditions talking can be used as a tool for oppression, and to help maintain systems of domination. Understanding how inequality threatens talk’s democratic functions is important in countries around the world, including today’s nominally egalitarian, Western democracies, for at least two reasons. First, it is important to understand how inequality impacts talk because an important form of inequality – economic inequality – is rising; leading scholars to dub our era as the “Second Gilded Age” (Bartels 2008; see also Piketty 2014). Second, as I endeavour to illustrate in my dissertation, material and social inequalities can structure agents’ cognitive and motivational dispositions in ways that engender internal exclusions, resulting in ongoing asymmetries of communicative participation and influence even after formal prohibitions to participation have been removed. Cognitive and motivational structures that discourage historically disempowered groups from participating in discursive politics, and undermine their discursive influence even when they do participation, help keep the legacy social inequality and exclusion – and the resulting harms of oppression and domination – alive. I offer empirical evidence for my theoretical discussion of how the legacy of formal, structural inequality and exclusion can shape historically disempowered social group members’ social cognitive and motivational structures to engender internal exclusions that entail asymmetries of discursive participation and influence. Focusing on gender inequality (and the intersection of gender with ethnicity and poverty), I offer the first systematic study of the gender 27 gap in discursive political participation. I also offer evidence from a novel survey experiment which reveals that even when women are included in talk, they have less discursive influence than men. My two empirical studies are important contributions because they offer convincing evidence for what many historically disempowered social group members (and allies) intuitively know: that even though disempowered group members are formally – legally – allowed to participate in talk and politics in Western democracies, we often find ourselves feeling muted. Even when we find our voices, we notice our utterances are often discounted. By offering evidence of ongoing gaps in communicative participation and influence, my empirical research does important political work by convincing skeptics that discursive inequality is an ongoing problem that needs to be addressed. 28 Chapter 2: Equality and Democracy My dissertation starts with the normative precept that “more democracy is a good thing” (M. E. Warren 1999, 60). Democratic systems – characterised by institutions and procedures that promote self-development and self- and collective-rule – can be expected to produce functionally better (e.g., more responsive), and normatively better (i.e., more just), opinions, decisions, norms, and social conditions. Social systems vary in the degree to which they are democratic, and more democracy is always better, and always possible. Social systems are democratic to the extent that people are included in political practices (such as voting or deliberating), can communicatively link personal preferences into collective opinions and agendas, and are empowered to turn collective agendas into collective decisions (M. E. Warren 2017). I begin this chapter by explaining what I mean by “democratic systems,” and explain why the “systems” approach is superior to the “models” approach for studying democracy. The next three sections are dedicated to the three general democratic functions political systems must achieve before they can be considered “democratic.” In the second section of this chapter, I discuss the democratic function of empowered inclusion. Drawing on Iris Young (2000), I elucidate the relationship between structural equality and inclusion in social and political practices, and explain why democratic processes begin with the empowered inclusion of those affected by collective outcomes. I expand Bernard Williams’ (1972) notion that equality is comprised of two, at times contradictory, values, which I refer to as “universal equality” and “equity.” In the third section, I build-out the collective opinion and agenda formation function. I describe five, interrelated component functions that fall under the general rubric of collective 29 opinion and agenda formation: the democratic social integration, reciprocity, personality, learning, and legitimation functions. I turn my attention to the collective decision-making function in the fifth section. Throughout my discussion of these three general democratic functions, I review different tools – resisting, talking, representing, voting, joining, and exiting – that can be used to achieve these interrelated democratic functions. Finally, in the fifth section, I turn my attention to the problem of inequality. I introduce the idea that different forms of inequality – political, material, and interactional (including discursive) inequalities – can prevent political practices from achieving democratic functions. 2.1 A Systems Approach to Democratic Theory What are social systems? On the one hand, social systems are comprised of socially- recognised – mutually comprehensible, and at least tacitly acknowledged – structures organising social life (e.g., J. Searle 2010; J. R. Searle 1995). These structures include language games, institutions, spaces of public appearance (including social identities and fields of social play), and other cultural artifacts that help define agents’ subjectivities, and enable and constrain agents’ practices. I discuss these structures in detail in Chapter 4 and Chapter 5. On the other hand, social systems are comprised of agents themselves, and agents’ responses to their social systems’ structures and to the material world. Bourdieu (2000, Chapter 4, pp. 128-164) describes agents as the embodiment of a “practical knowledge,” because they are shaped by their experiences in the world, and embody accumulated knowledge of personal and collective (cultural) life experiences. I return to this point in Chapter 4, “Practical Knowledge.” 30 Agents respond to social systems’ features through their practices, including through generic political practices that can serve democratic functions. Generic political practices are ideal-type social actions (in the Weberian sense of socially meaningful actions), commonly “organized or enabled by institutions that serve democratic functions: recognising, resisting, representing, deliberating, voting, joining, and exiting” (M. E. Warren 2017, 43). Agents maintain or modify social systems’ structures through their everyday activity, through available procedures for changing social systems’ structures, and through confrontation (Tully 2002, 2008a, 2008b). I describe how different tools can serve democratic functions in the next section of this chapter. In Chapter 3, I explain why talking, in particular, is essential for achieving democratic collective opinion and agenda formation functions. Within social systems, sets of political institutions – which I collectively refer to as social systems’ political systems – are designed to coordinate, organise, and direct the shared intentions and understandings inscribed in social systems’ mutually comprehensible and at least tacitly acknowledged language games, institutions, spaces of public appearance, and other cultural artifacts. Institutions are “rule-based, incentivized, and sociologically stable combinations of social actions that assign roles to individuals (e.g., voter, representative, etc.)” (M. E. Warren 2017, 43). Political institutions refer to institutions relevant for governing, or exercising a directing or restraining influence over social systems’ language games, social institutions, spaces of public appearance, and cultural artifacts. As such, political institutions play an essential role in distributing social and political empowerments. Because political systems direct and coordinate social systems’ structures, they enable and constrain agents’ choices for responding to these structures (Tully 2002, 2008b, 2008a). For instance, when political systems record equal rights into law, they symbolically consecrate 31 symmetrical power relations that entail inclusions in language games, institutions (including political institutions), spaces of public appearance, and so on. In so doing, political systems enable more equal opportunities to participate in and influence social and political practices through everyday activity and normal procedures for challenging social systems’ features. However, political systems can also record distinctions between social groups into law, and symbolically consecrate asymmetrical power relations that entail the exclusion of the disempowered. In so doing, political systems create asymmetrical opportunities for participating in, and influencing, social and political practices. Social systems – and the sets of political institutions that direct social systems’ structures – range in the degree to which they are democratic or oppressive. Democratic social systems exist when relatively equal conditions engender symmetrical empowerments so that those affected by social systems’ structures have opportunities for self-development, and to participate in and influence self- and collective-rule. More specifically, we can say that social systems and their sets of political institutions are democratic to the extent that they achieve three interrelated functions. First, by definition, social systems’ structures must engender symmetrical empowerments that entail inclusions in practices of self- and collective-rule. Second, practices of self- and collective-rule must involve forming individual preferences into collective opinions (collective understandings about the nature of issues of collective concern) and agendas or wills (shared intentions about what should be done about issues of collective concern). Third, practices of collective-rule must involve empowering collectives – for instance, political communities (municipalities, regional governments, countries, etc.), or other governance bodies (condominium strata councils, school boards, unions, etc.) – to execute collective agendas as collective decisions. 32 Oppression and domination, by contrast, refer to situations in which inequalities engender asymmetrical empowerments that entail exclusions, preventing the disempowered from participating in self-development or using political practices to challenge or change social systems’ structures. Social systems are “oppressive” to the extent that asymmetrical empowerments entail structural (group-based) exclusions that perpetuate systematic constraints on social group members, preventing them from participating in language games and processes of mutual recognition. Social systems are characterised by “domination” to the extent that asymmetrical political empowerments entail structural (group-based) exclusions that perpetuate systematic constraints on social group members, preventing them from participating in or influencing political practices through normal activity. Social systems are also characterised by domination when there are no available procedures for modifying social systems’ structures, or available procedures fail to address systematic constraints on social group members’ capacity for self-development and self- and collective-rule. These systemic constraints are “not necessarily the result of the intentions of a tyrant” (Young 2011, 41). Rather, structural oppression is a property of social systems, “embedded in the unquestioned norms, habits, and symbols, in the assumptions underlying institutional rules and the collective consequences of following those rules.” Structural oppression cannot easily be eliminated by simply changing the rulers or drafting new laws, because oppressions are systematically reproduced in social systems’ language games, institutions (including social, political, and economic institutions), spaces of public appearance, and other structures. I return to this topic at length in Chapter 6, “Inequality, Oppression, and Domination.” As I mentioned in my Introduction (Chapter 1), when I refer to equality (or inequality) I am referring to “structural equality,” or equality between salient social group members (rather 33 than, say, idiosyncratic differences between individuals) (Harell and Stolle 2010). Social groups are an expression of social relations, since “a group only exists in relation to another group” (Young 2011, 43). Salient social groups can include class, gender, language, and ethnic, racial or religious groups. Note that social group memberships and collective social identities are not the same thing. Social group membership is involuntary: material conditions (e.g., the economic conditions delineating classes) and collective understandings or assumptions inscribed in social systems’ structures (e.g., the social and political conditions delineating racial categories) ascribe social group memberships. People can “recognize that they belong to any number of social groups without adopting those classifications as social identities” (Brewer 1991). The worker without class consciousness is still defined as a worker by their position in a capitalist economy, and Rachel Dolezal is still defined as a white woman by her privileged position in a racist social system that empowers white people relative to people of colour (Oluo 2017). I return to the topic of collective social identities at length in Chapter 5, “Spaces of Public Appearance.” 2.1.1 Benefits of the Systems Approach As Warren (2017) explains, democratic theorists typically think in terms of “models of democracy,” defining democracy in terms of a single feature or mechanism (such as deliberative democracy, aggregative democracy, pluralist democracy, and so on). And while the models’ approach has some benefits – such as clarifying normative presuppositions, and helping democratic theorists think through better or worse forms of democracy – the models thinking has several problems that hamper democratic theory’s progress. For one thing, centering thinking on 34 one defining feature of democracy (e.g., deliberation or elections) leads to overextended claims for that feature. This can lead to treating a specific feature of democracy as if it were self-sufficient, and organise other problems of democracy out of the picture. Warren (2017, 40) gives the example of how deliberative democratic theorists’ focus on deliberation spurred a productive discussion of how individuals can form individual preferences into collective agendas. However, this comes at the expense of other problems of democracy, such as the “problems of distributions of power and voice, and actionable decision mechanisms” (2017, 40). As Warren (2017, 40) notes, “a model of deliberative democracy, insofar as it is centred on deliberation, is not a theory of power, nor a theory of distribution, nor a theory of inequality, nor of political decision making.” Ultimately, the problem with the models approach is that it provides “the same answers (e.g., deliberation or elections) to different problems of democratic political organization (in particular, empowered inclusion, collective opinion and agenda formation, and collective decision making)” (M. E. Warren 2017, 41). The systems approach avoids conflating single institutions (such as deliberation) with “democracy,” so we can “ask the normative question as to whether any particular institutional organization of practices serves democratic functions within its context” (2017, 43). In the next three sections of this chapter, I describe the three broad functions – empowered inclusion, collective opinion and agenda formation, and collective decision-making – in more detail, and describe different political practices that can be used to achieve them. 35 2.2 Empowered Inclusion Political systems are democratic to the extent that people are included in political practices (such as voting or deliberating), can communicatively link personal preferences into collective opinions and agendas, and are empowered to turn collective agendas into collective decisions (M. E. Warren 2017). It is important to note that democratic processes always begin – functionally and normatively – with inclusion. This is because the aims of forming collective opinions and agendas, and making collective decisions, are undermined if those affected by collective endeavours are excluded from participating in them (Fung 2013; Goodin 2007; Young 2000). As I began to explain in my Introduction, the norm of inclusion requires a degree of equality. “Equality,” broadly understood, can be thought of as empowerments that enable people to participate in forming individual and collective judgments and decisions through political liberties (i.e., equality entails the absence of constraints), as well as through positive supports to develop individual and collective capacities. Equality contains two distinct values (B. Williams 1972; see also Beauvais 2017; Beauvais and Bächtiger 2016). On the one hand, universal equality (or universal moral equality), requires abstracting from social circumstances, and recognizing the fundamental sameness of common humanity. The value of universal equality involves treating people as if they share a universal starting point (such as the same baseline of moral worth) and have the same fundamental needs (such as the need for life, liberty, and security of the person). However, group members do not “arrive at life’s starting lines” with equal resources (M. S. Williams 2000, 60). The second value of equality is equity, or when justice demands attending 36 to social circumstance and recognising group-based differences between people. “Justice” here refers not only to recognising class or economic differences when redistributing wealth, but more generally to any enabling conditions necessary for “the development and exercise of individual capacities” (Young 2011, 39). Unlike universal moral equality, the value of equity involves treating social group members as if they have different starting points – different access to wealth and power, or different physical, cognitive, or linguistic styles and abilities – and so have different needs for developing and exercising individual capacities to their fullest extent. Various political theorists have described the concept of equity in different ways. Melissa Williams (2000, 61) describes this concept as “difference-conscious equality,” which draws attention to social diversity. Jack Knight and James Johnson (1997, 292) describe the value as “substantive equality,” which draws attention to the fact that realising equality of some outcome (such as equal political influence) may require efforts “that treat individuals unequally” (p. 280). I prefer “equity” – which comes from the Latin aequitas, and denotes fairness or justice – because the term connotes attending to diverse social circumstances and treating people differently based on claims to justice. When Aquinas wrote of aequitas in the Summa, he defined it as “the quality which pertains to moderating the letter of the law” (quoted in Skinner 2002, 49). Recall that when I refer to equality’s twin values I am referring to structural universal equality or structural equity between salient social group members (Harell and Stolle 2010). Both the value of universal equality and the value of equity can – and should – be accommodated in democratic systems, through different political practices and institutional arrangements. For instance, democratic systems must enable empowerments through a baseline of universal equality enshrined as rights and freedoms, such as the universal freedom to organise, speak, and engage in other political acts. But in political systems where historical inequalities 37 result in continuing asymmetries in social group members’ abilities to use these universal empowerments, it may be necessary to pursue the value of equity to promote the empowered inclusion of historically disadvantaged group members. For instance, targeted “get out the vote” campaigns may be designed to motivate the poor, or new citizens to vote in general elections. One of the most important set of institutions for guaranteeing the condition of universal equality is the recognition of universal rights, including political participation rights and political freedoms, typically enshrined in constitutional law and protected through the doctrine of state neutrality. Preserving political equality through a system of rights entails moral inclusion, and expands “deontic commitments” that “support rights and duties of citizenship” (M. E. Warren 2017, 46). Law and the judiciary can also be used to promote equity. For instance, by making legal provisions for promoting the economic and political participation of disadvantaged or historically disempowered group members. The aim of Canada’s “Employment Equity Act” is to “correct the conditions of disadvantage in employment experienced by women, aboriginal peoples, persons with disabilities and members of visible minorities” through “special measures,” which includes preferential hiring systematically underrepresented social group members (Government of Canada 1995 see “Purpose of Act,” paragraph 2 ). The condition of equality provided by the recognition of universal rights and group-based differences is requisite for achieving the empowered inclusion of those affected by collective agenda formation and decision-making. Voting and representing are two other useful generic practices that can be used to achieve the twin values of equality and – ultimately – the empowered inclusion of those affected by collective outcomes in collective agenda formation and decision-making. The value of universal 38 equality informs the cornerstone of fair representation in liberal democracies through the doctrine of “one person one vote,” or the equal opportunity for every enfranchised citizen to influence the outcome of an election. Voting is a useful way to distribute empowerments across populaces, particularly large, heterogeneous populaces (since voting preserves expressions of dissent) (M. E. Warren 2017). However, the drawback for inclusion is that voting requires a defined populace – the demoi must be preformed, typically delineated by the rules of citizenship or enfranchisement. Representation is another particularly useful tool for empowering inclusion because it can expand inclusion over space and time (e.g., to include considerations that affect future generations). Representation can be used to achieve either universal equality or equity, for instance by reserving seats in decision-making bodies for specific group members to guarantee some notion of just (equal or equitable) representation. For instance, efforts to ensure universal equality include using the principle of “representation by population” to elect a federation’s lower legislative house’s representatives, to ensure each voter has equal (the same) influence. Or efforts may be made to ensure the over-representation of certain groups, to guarantee some notion of equitable representation and influence. For instance, institutions may be designed to ensure an equitable representation of all voters through “representation by region,” to achieve the equal representation of regions (and thus over-representation of voters in sparsely populated sub-national regions) in a federation’s upper legislative house. Other efforts to achieve equitable representation in the public sphere might include reserving seats for women on a company’s or university’s board of directors. Other efforts in civil society might include selective recruitment techniques to achieve the over-representation of women in feminist civic associations, to create 39 spaces where feminist women can invent “new terms for describing social reality” (Fraser 1990, 67). Note that talk (including deliberation) is often less effective at achieving equality and equity-based empowered inclusion. Perhaps most obviously, talking privileges speech (verbal communication, but also reading and writing), which potentially excludes non-speaking agents, and those who cannot read or write in the language of conversation, including people with certain cognitive or neurological disorders, or foreign-language speakers. Talking, and especially deliberation, may also privilege forms or habits of speech (articulateness, pitch of voice, and so on) that correlate with class, race, gender or socioeconomic status to produce internal exclusions – a point I return to throughout my dissertation. There are also practical reasons for why talking is often less effective for achieving equality and empowered inclusion. For one thing, talk – particularly face-to-face talk – may require making decisions about who is included and who is excluded in communicative exchanges, since it is impossible for an unlimited number of people to participate simultaneously in a single conversation. Talk that is aimed at an indefinite or unlimited number of people (such as radio broadcasts or podcasts, or social media posts such as tweets), is often unidirectional, and thus may be less effective at achieving talk’s strengths related to inclusion. For instance, unidirectional talk may be less responsive to persons and groups, and those with greater resources (such as the wealthy, politically powerful, or famous) will have greater access to media for broadcasts, and will reach a far greater audience through social media. Finally, unidirectional, mass public talk may be only weakly characterised by the moral commitments that arise during face-to-face social interaction and talk (e.g., Goffman 1967). 40 2.3 Collective Opinion and Agenda Formation When a baseline of universal equality and equity engenders symmetrical empowerments that entail inclusions in political practices, agents can effectively participate in forming individual preferences and interests into collective opinions and wills (M. E. Warren 2017, 44). The intuition informing this general democratic function is the idea that individual preferences need to be communicatively linked to collective judgements to link self-government with collective governance. As Warren (2017) notes, talking (including deliberating) is the most important and effective tool for connecting individual judgments to collective agendas. To assess political systems’ democratic problem-solving capacities related to the collective opinion and agenda formation function, it is useful to build-out Warren’s description of connecting individual judgements to collective agendas, and ask what distinct aims (or sub-functions) must be achieved to form individual opinions and wants into collective opinions and wills. In other words, what do political practices need to accomplish to effectively link individual preferences to collective opinions and wills? For one thing, for people to communicatively link their preferences to any kind of larger collectivity, they need opportunities to talk to others – they must have links to a broader collective of people. That is, social ties are required to create opportunities for communicatively linking individual preferences to collective agendas. More specifically, non-hierarchical social ties that cross-cut salient social group boundaries are required for people to democratically form individual wills into collective wills through communicative processes. Borrowing terminology from Allison Harell and Dietlind Stolle (2010), I refer to this component of the collective agenda formation function as the social integration function. 41 Second, people need to be willing or even motivated to link their preferences with others to form collective agendas, which means that they must be willing and motivated to listen to others, and afford others the respect of taking their preferences into account. Furthermore, they must be willing and motivated to express themselves to others, and because self-disclosure entails vulnerability, they must have faith in others to respect them enough to take their preferences into account. In other words, facilitative attitudes (including mutual respect, trust, mutual liking or at least tolerance, and empathy) that lubricate social interaction and communication by promoting reciprocity are required for people to democratically form individual wills into collective wills through communicative processes. I refer to this component of the collective agenda formation function as the reciprocity function. Just as people must have confidence in others before they can engage in self-disclosure and communicatively express their preferences to form collective agendas, people must have a healthy degree of confidence in themselves. A healthy, non-debased degree of self-confidence and self-efficacy is required so speakers can weather the strain of vulnerability to express themselves and their preferences to others. A healthy, non-narcissistic (that is, not unrealistically exaggerated) degree of self-confidence and self-efficacy is also helpful, since narcissists may be unwilling to reconsider or modify their own preferences to form them into collective agendas. I refer to this component of the collective agenda formation function as the personality function. When agents have the opportunity and willingness to talk to one another, and basic capacity to express themselves (and listen) to others, they should learn from one another. Through the process of communicatively linking their preferences with others to form collective agendas, people should discover and share new information and considerations, and sharpen (and 42 possibly change and improve) one another’s personal understandings and preferences. I refer to this component of the collective agenda formation function as the learning function. Finally, discursively generated mutual understandings and agreements should underwrite commitment to communicatively constituted collective agendas. People who have the opportunity, motivation, and capacity to participate in communicative processes can then learn from one another and form their individual preferences into collective wills. And because they were empowered to participate in communicative processes, and refined their own preferences with others, they develop confidence or faith in the resulting agendas. I refer to this component of the collective agenda formation function as the legitimation function. I return to these five necessary democratic aims constituting “collective opinion and agenda formation” – the democratic social integration, reciprocity, personality, learning, and legitimation functions – in Chapter 3, where I focus my attention more narrowly on the generic practice of talking, and how talking can be used to achieve the interrelated collective opinion and agenda formation functions. I also discuss the problem of when inequality and exclusion prevent talking from achieving these normatively necessary democratic aims. 2.4 Collective Decision-Making In addition to empowered inclusion and collective opinion and agenda formation, democracies must accomplish a democratic collective decision-making function. Like empowered inclusion, collective decision-making is about empowerment. However, while empowered inclusion pertains to individual empowerment, collective decision-making pertains to collective empowerment, or governance. As Warren (2017, 44) describes, this occurs “when 43 collectives have the capacity to make and impose binding decisions upon themselves; it is about ‘getting things done’.” The decision-making function is implied in the very definition of democracy, since democracy is about participating in self- and collective-rule. Perhaps the best tools for achieving democratic collective decision-making are voting and representing. Voting enables clear decision-rules, and preserves dissent, and representative bodies can function usefully as accountable decision-making bodies. Exit is also useful, because it can offer varied and proximate responsiveness (although suffers the drawback of low collective agency). Talk, however, is less useful as a collective decision-making device. Although talking – particularly deliberation – can generate mutual understandings and agreements that underwrite decisions, talking is not inherently a decision-rule. While talk can be a used as a consensus decision-making device, in today’s modern mass democracies collective decision-making is usually achieved by voting (and majority rule), even in deliberative contexts such as parliaments and courts. As Robert E. Goodin (2008, 107) suggests, there are good reasons to use talking as a “discovery procedure” preceding decision-making, rather than as a decision-method in and of itself, or in Goodin’s words, there are good reasons to “first talk, then vote.” As Goodin (2008) explains, talking gets different options on the table and allows discussants to better understand the benefits and drawbacks of different options, thus serving a learning function before decision-making takes place. However, as a decision-method procedure sui generis, talking suffers from the drawbacks of indeterminacy and path-dependency. And, of course, in modern, pluralised mass democracies, talk-based, consensual decision-making is often not a practical option for many or most collective decisions. For these reasons, I do not focus on using talk as a device to achieve the democratic decision-making function in my dissertation. Instead, I focus my 44 discussion on using talk as a device for achieving democratic collective opinion and agenda formation, a point I return to in Chapter 3. 2.5 The Problem of Inequality As I have explained, a degree of structural equality is required for political practices to achieve their democratic inclusion, collective opinion and agenda formation, and collective decision-making functions. Harell and Stolle (2010) identify three structural gaps among salient social groups that can prevent practices such as deliberating and voting from functioning democratically: political gaps, economic gaps, and gaps in resources that otherwise should facilitate economic and political equality. Political and material (economic and resource-based) structural inequalities are “objective” inequalities in the sense that they entail external exclusions constraining disempowered social group members’ (or enable empowered social group members’) participation in political, economic, and social practices, regardless of agents’ subjective understandings, desires, or motivations. Clearly, structural gaps in political rights between social groups entail asymmetrical power relations and formal political exclusions. For instance, limiting political participation rights to white, property-owning men obviously enhances white, property owning men’s political power relative to other social group members’. And the prohibitions preventing disempowered group members from legally participating in politics clearly entail external exclusions from voting, running for office, or other political activities. Structural political inequalities and external exclusion contribute to harms of oppression, rendering people of colour, the proletariat, 45 and women powerless, and marginalising them from the deontic commitments that support the rights and duties of citizenship. Another example of when structural gaps in political rights entail asymmetrical empowerments and formal political exclusion occurs when group members are unfairly represented (unequally or un-equitably, depending on the demands of justice) in political decision-making bodies. When women are underrepresented in parliament, and there are few (or no) women to bring up matters particularly salient to women constituents, then debates are less likely to consider women’s issues. Furthermore, in such situations representation is less likely to extend inclusions. For instance, when parliamentary debates fail to consider women’s issues here and now, it is even less likely that parliamentary debates will consider issues that affect future generations of women. Economic inequalities and resource-gaps also entail asymmetrical power relations and external exclusions from political practices. Recall my examples from Chapter 3, of when poor people who are unable to afford the cost of “pay-for-access” fundraisers are excluded from attending functions where they could meet politicians and other influential members of society. By economic inequality I am referring to structural income-gaps between social groups, whereas by resource-gaps I am referring to structural differences in the ability to access resources that facilitate economic and political participation (Harell and Stolle 2010). Perhaps most obviously, resource-gaps include things such as access to housing and leisure time. However, another category of important resources are materially-important social ties. I define materially-important social ties as social ties to politically or economically powerful actors. Aggregate asymmetries in materially-important social ties between social group members can exacerbate economic and political inequality, and generate external exclusions. For instance, 46 in many societies men maintain disproportionate control over the economy through ownership or management of land and natural resources, financial capital, companies, and so on. Men’s disproportionate economic control in and of itself represents an economic-gap between men and women. However, social ties to those who have control over resources is a resource-gap that facilitates economic participation and influence (or potentially political participation and influence). When men have more social ties to people who control the economy through ownership and management of the means of production (for instance, because men tend to have more ties with other men than women do), then this represents gaps in materially-important social ties. Like political inequality, economic and resource-based inequalities generate asymmetrical power relations that entail the external exclusion of the disempowered, and so prevent generic practices from serving democratic functions. People and groups with greater economic wealth and control over resources have greater capacity to resist, and exert greater influence through joining (in the sense that they are over-represented in well-organised and well-financed groups and associations), excluding less organised social group members from fair influence in collective communicative and decision-making processes. Finally, because of their organisational strength, social group members with greater economic wealth and control over resources may exert disproportionate power through exit, again depriving less well-organised, disempowered social group members from fair influence in collective communicative and decision-making processes. The exclusions that unequal political rights, and economic or material inequalities entail are “external” in that they preclude disempowered social group members from participating in politics at all (Young 2000). These inequalities prevent disempowered social group members 47 from legally casting a ballot, affording a donation, or having time to associate with like-minded people. Most democratic theories consider the problem of external exclusion, and “call for limiting the influence of wealth or position on the ability to participate in a democratic process” (Young 2000, 55). However, inequality can also structure social cognition to engender internal exclusions. Internal exclusions refer to when, even in the absence of formal political or economic asymmetries, socio-cognitive features shape how people view themselves and other social group members to engender asymmetries in political participation and influence. Consider how those who have historically been excluded from the franchise – women, the poor, or non-whites – may internalise a sense of “feeling out of place” in politics. People who feel out of place in politics are less likely to participate in practices such as voting or deliberating, even after formal prohibitions to their participation have been lifted. And when people internalise a feeling that women, the working class, and non-whites are “out of their league” in political activity, they may dismiss or ignore these disempowered group members’ contributions when they do participate in politics. Inequality and the external and internal exclusions it engenders can be thought of as “systematic constraints” on social group members that prevent or exclude them from developing or exercising their individual capacities (contributing to harms of oppression), and prevent them from participating in political practices to influence the norms and laws that affect them (contributing to domination) (Young 2011, 41). Both the absence of universal moral equality (e.g., when slavery or apartheid precludes social group members from moral personhood, and from enjoying life, liberty, and security), and the absence of equity (e.g., the absence of enabling 48 efforts to help disadvantaged social group members) act as disabling constraints that entail exclusions, contributing to, or reinforcing systems of oppression and domination. Under conditions of inequality and exclusion, political practices such as talking or voting can reinforce oppressive political systems. Continuing with my example of a restricted franchise, consider how, until the 20th century, legal prohibitions precluded women from voting or running for office in most jurisdictions, and clear normative expectations regarding women’s place in society – as caregivers (often in the private sphere) – contributed to their external exclusion from political and economic participation and influence. These legal and normative prohibitions on women’s participation in politics and often in the paid labour market were designed to maintain asymmetrical empowerments (men’s domination of women), that entailed external exclusions (the exclusion of women from political practices that contribute to self-development and self- and collective rule). And while women of colour, such as black women in the United States, appeared to achieve greater equality relative to black men because the legacy of slavery forced black women to work on par with black men, systems of white supremacy worked in conjunction with patriarchal systems to exclude women and people of colour from participating in the (white, male) public sphere (e.g., see A. Y. Davis 1981). Men’s and women’s participation/non-participation in politics and the economy contributed to a system of racialized patriarchal domination, since women could not determine their actions or the conditions of their actions. It also contributed to harms of oppression, binding women in relations of dependence and powerlessness, marginalising them from public life, and enabling the exploitation of their unpaid sexual and emotional labour (Pateman 1988; Young 2011). The exploitation of women’s unpaid sexual and emotional labour was also racialized. After the Industrial Revolution (when mechanisation made many of women’s jobs obsolete), 49 women (particularly white women) became venerated for their roles as wives and mothers – as caregivers in the private sphere – opening them to the exploitation of their emotional labour, and to domestic sexual violence (A. Y. Davis 1981). However, because white racism denies the equal moral equality of non-whites, including the moral equality of women of colour “as women,” women of colour were particularly vulnerable to sexual violence. 2.5.1 Interactional (Including Discursive) Inequalities and Internal Exclusion The problems of political and material inequality I have described so far – political participation rights gaps, economic gaps, and resource gaps – mostly engender external exclusions. But today, formal prohibitions on women’s participation in the paid labour market and politics have been removed: employers are not allowed to openly discriminate against women when hiring, and men and women enjoy the same political rights (to vote, associate, run for office, and so on). But the historical memory of patriarchal power relations – the formally legal and explicit distinction between women’s roles as caretakers, particularly (at least among white women) in the private sphere, and white men’s roles as providers and decision-makers, particularly in the public sphere – remain as implicit associations in schemes of thought and expression that entail internal exclusions. Democratic systems not only benefit from structural political and material equalities, they benefit from interactional equalities. Interactional equality refers to the ideal of equal opportunities for participation and equitable influence in social interaction across and within salient social groups in democratic systems. Interactional inequality is not an “objective” 50 inequality in the sense that it does not only involve external structures (a system of political rights, the distribution of economic opportunities and resources) that shape agents’ life chances regardless of their thoughts and feelings. Rather, interactional inequalities refer to when objective inequalities structure agents’ subjectivities – their thoughts and feelings – to shape their interactions with others. While objective inequalities entail external exclusions (or constraints) that prevent agents from participating in or influencing practices of self-development and collective rule, interactional inequalities entail internal exclusions (or constraints) that prevent agents from participating in or influencing practices of self-development and collective rule. To understand the relationship between inequality and exclusion – and the distinction between objective/interactional inequalities and external/internal exclusions – let me consider the example of women’s political engagement. A restricted franchise that prohibits women from voting entails women’s external exclusion from electoral participation. Restrictive electoral laws are an external barrier preventing women from participating in and influencing elections, irrespective of women’s (or men’s) thoughts or motivations. However, even after women’s enfranchisement – after formal, external legal barriers to their participation were dismantled – women did not begin voting or participating in politics at the same rate as men. The norm (or intersubjectively held belief) that politics is a man’s game (particularly, in Western democracies, a white man’s game), or that women are not suited for politics, are interactional inequalities that entails women’s internal exclusion from political engagement. For instance, the gender gap in self-perceptions of political understanding – the tendency for women to be significantly more likely than men to agree that “politics is too complicated for a person like me to understand” – which reduces women’s belief that they 51 should or can participate in politics, is an interactional inequality (Gidengil, Giles, and Thomas 2008; Thomas 2012). This form of inequality clearly does not entail “external” exclusions, since it is not an external barrier preventing agents from participating in social and political practices irrespective of their thoughts and motivations. Rather, it refers to when intersubjectively held understandings or assumptions become “internal barriers” in cognitive and motivational structures, preventing agents from participating in social and political practices. Interactional inequality and internal exclusion not only affect all generic practices that serve democratic functions, but affect all practices, including ostensible non-political, social practices such as dating and marrying, learning and getting accreditation, job searching and working, to name a few (e.g., Bourdieu 1977, 1990, 2000). However, the primary aims of my dissertation are to understand how talk contributes to collective opinion and agenda formation, and to offer a better understanding of the problem that inequality presents to talk and democracy. Given the purpose of my dissertation, I focus my analysis on one form of interactional (in)equality: discursive (in)equality. Discursive inequalities engender unjustifiable, structural asymmetries of communicative participation and influence. Discursive inequalities are disabling constraints that entail internal exclusions, undermining disempowered social group members’ participation and influence in communicative processes. Discursive inequalities reinforce existing systems of oppression and domination, since the disempowered cannot participate in (and may not even be motivated to participate in) language games and self-expression, or political practices to modify their social and political conditions. Discursive inequalities contribute to the internal exclusion of disempowered group members by influencing how people perceive speakers from social outgroups, and how they perceive themselves and other speakers from social ingroups. 52 As a point of clarification, note that both objective inequalities and interactional inequalities generate exclusions that prevent talking (and other practices) from functioning democratically. However, because the problem of objective inequalities and external exclusion has been more thoroughly theorised and researched, most of my dissertation focuses more how internal exclusions prevent talking from functioning democratically. Note also that the concept of “discursive equality” – of equal opportunities for communicative participation and equitable communicative influence across salient social groups in a democratic system – is a regulative ideal. Even the best functioning, most egalitarian utopias will always stop short of this ideal, because (at the very least) of the challenge of including those with severe cognitive or certain neurological disabilities in communicative processes. These inclusions can be achieved in part through relations of care and representation, but expressions of preferences might still be lost, misunderstood, or misrepresented. As Jane Mansbridge (2015) argues, power is always unavoidably at play when we use language, but the closer we get to an absence of power, the better discursive political practices are on normative grounds. The point of a regulative ideal is that we aim to approximate it, even if we can never perfectly achieve it. This means that the closer we approximate to the ideal of discursive equality, the better our always imperfect practices achieve their democratic functions. And to reiterate, more democracy is always possible, and always better. 53 2.6 Conclusion Political systems are democratic to the extent that people are empowered to participate in political practices – voting, representing, deliberating, resisting, and so on – that contribute to self-development and self- and collective-rule (M. E. Warren 2017; Young 2011). Equality and inclusion functionally precede the democratic aims of collective opinion and agenda formation, and collective decision-making. Equality means that empowerments are distributed in ways that enable those affected by collective endeavours to participate in, and influence all social practices. Equality means that empowerments are distributed symmetrically, so that all agents are included in participating in and influencing cultural reproduction, social identifications, and fields of social play. Most importantly for my dissertation topic, equality means that empowerments are distributed symmetrically to entail inclusion in generic political practices that serve democratic functions, such as voting, representing, talking, and so on. Participating in and influencing political practices is necessary for agents to form individual preferences into collective wills, and – ultimately – for communities to make collective decisions. This, perhaps obviously, includes collective decisions about which social systems’ features should be recorded into law, and how law should coordinate social systems’ language games and other institutional and cultural artefacts, spaces of public appearance, and power relations. The problem of inequality (including discursive inequality) for the practice of talk is threefold. I introduced the first problem in this chapter, and will elaborate my discussion in Chapter 3, “Talk and Collective Agenda Formation.” Specifically, that under conditions of inequality, political practices – including talk – are less likely to function democratically. The second problem is that under conditions of inequality, talk can contribute to harms of oppression. 54 The third problem is that under conditions of inequality, disagreement and misunderstanding are less likely to be solved through deliberation – which, as I will explain, contributes to domination. I address these final two problems in Chapter 6, “Inequality, Oppression, and Domination.” As I explain in Chapter 6, the problem of structural political and material inequalities generates asymmetrical power relations that exclude the disempowered from participating in and influencing communicative processes, and so bring democratic systems further from the ideal of discursive equality. Even after large objective (political and material) inequalities have been addressed, inequalities can remain as systems of perception and evaluation, structuring everyday interaction and communication to produce asymmetries in communicative participation and influence, bringing democratic systems further from the ideal of discursive equality. 55 Chapter 3: Talk and Collective Opinion and Agenda Formation Equality means that empowerments are distributed in ways that enable those affected by collective endeavours to participate in and influence social and political practices. Inequality is a problem, because inequality means asymmetrical power relations entail the exclusion of the disempowered from participating in and influencing political practices. In this chapter, I describe how talk achieves collective opinion and agenda formation functions, and the problem of when inequality prevents talking from achieving this set of democratic functions. In the first section of this chapter, I explain what I mean by the generic function of “talk,” distinguishing between the concepts of talk and social interaction. I also identify two forms of talk, differentiating between deliberation and everyday talk. In the second section of this chapter, I extend my discussion from Chapter 1 on the collection opinion and agenda formation functions. I review the idea that collective opinion and agenda formation is comprised of at least five distinguishable but interrelated components: the social integration, reciprocity, personality, learning, and legitimation functions. I elaborate why each of these components are normatively necessary for democratic collection opinion and agenda formation, explaining why talk is so important for achieving each of them. In the second section, I also specify whether deliberation or everyday talk is best for achieving each of the different collective agenda formation component functions. In the third section of this chapter I return to a central idea from Chapter 2, that equality engenders symmetrical empowerments that entail inclusion, and empowered inclusion is required before talking functions democratically. I elucidate how different forms of inequality, 56 including discursive inequality, can prevent talk from achieving each of the component collective opinion and agenda formation functions. 3.1 Forms of Talk 3.1.1 Talk Versus Social Interaction I am not only interested in the political practice of deliberation, but am more broadly interested in using talk to achieve democratic functions. But what is “talking”? Talking or communication refers broadly to using signs or signifiers to express something to someone. Often, these signs or signifiers are words or utterances – either spoken, written, or signed – but communication can include other signifiers used to connote and exchange meanings between people, such as a loud sigh or a raised eyebrow. Although much of what we think of as communication takes place in face-to-face interaction, communication also takes place over time, through stories and texts passed down inter-generationally, and across geographical space through a variety of communication and telecommunication formats. Communication can be highly mediated and unidirectional, as is the case of radio or television broadcasts and podcasts, public speeches, and government or corporate statements and news releases in the mass public. Much of what I refer to as communication overlaps with what sociologists refer to as social interaction, which refers to an exchange (a two-way effect) between two or more people, often occurring face-to-face (David Jary and Jary 1995). I eschew the term social interaction in my analysis because this term is simultaneously too specific and too unspecific. It is too specific in the sense that social interaction requires an exchange between multiple parties, whereas 57 talking does not. Ideally, people “think through” issues and potential solutions themselves, before they speak to others (essentially silently talking through the issues themselves). And as anyone with a graduate degree knows, people regularly write and publish knowing their words will likely go unread. As these two examples demonstrate, communicative activities might only involve one person, 1 and so are not included in the sociological definition of social interaction as an exchange between two or more people. Similarly, the mediated, unidirectional nature of communication in the mass public sphere (for instance, from media broadcasters to the public) is less of an “exchange,” and so is not obviously a social interaction. The term social interaction is too unspecific in the sense that an exchange between two or more people may be largely meaningless, and I am only interested in exchanges imbued with meaning, or communication as social action. If a woman sits next to a stranger on a bus, and the seated stranger, without looking up, moves over slightly to accommodate the woman next to them, we could say they have interacted. But I am less interested in this kind of reciprocal effect between two or more agents, and more interested in social actions, when interactions are used to (or are interpreted to) signify meaning. Consider again the example of a woman sitting next to a stranger on the bus, only imagine now that the woman sitting down has a facial disfigurement. The stranger, noticing the disfigurement, might become stiff with discomfort and press themselves as far away from the woman as they can, quickly averting their wide-eyed gaze. This interaction is communicative in that it expresses something – discomfort, perhaps surprise or disgust – which can be perceived and understood by the disfigured woman and others. 1 Other than in the abstract sense of involving communication with “generalised others.” 58 3.1.2 A Shared Background of Implicit Mutual Understandings The bulk of human communication and language games take place against social systems’ backgrounds of shared attitudes, competencies, and practices, including the implicit understandings or assumptions about what makes a face or body normal, a reaction acceptable, an utterance correct or valid, or an argument adequate. Following Bourdieu, I argue that much public talk takes place in fields of public opinion (including public spheres), which can be sites for the “confrontation of competing discourses” (Bourdieu 1977, 16:168). However, the bulk of human conversations do not involve this clash of competing discourses, but rather take place against a shared backdrop of unproblematized mutual understandings, and are characterised by a shared naïve adherence to the world. I refer to talking that takes place against a shared backdrop of unproblematized mutual understandings as everyday communication or everyday talk. Everyday communication includes both communication about matters of collective concern as well as the more common and ubiquitous mundane communication. Mundane communication refers to everyday talk that is unrelated to issues of collective concern, and may be purely expressive (“I love baking!”), or functional (“please pass me that bag of sugar”). Everyday talk about matters of collective concern involve communication about issues that affect collective life (Jacobs, Cook, and Delli Carpini 2009; J. Mansbridge 1999). But, like mundane talk, everyday talk about matters of collective concern occurs within social systems’ background of shared attitudes, competencies, and practices. 59 Everyday talk is often (but not always) delimited by an orthodoxy in fields of public opinion. Orthodoxy refers to “straight, or rather straightened, opinion” (Bourdieu 1977, 16:169). Orthodoxy “delimits the universe of possible discourse,” constituting that which is “undiscussed, unnamed, admitted without argument or scrutiny” (Bourdieu 1977, 16:169–70). Orthodoxy exists in opposition to heterodoxy, which represents the choice of competing “possibles” and critique of an established order. Heterodoxy always implies the sum of alternatives not chosen, revealing the arbitrariness of political orders (in the sense that they are one option among many). Heterodoxy is created and expanded through the practice of epoché, or the “the deliberate, methodological suspension of naïve adherence to the world.” Epoché describes the state of suspending pre-existing judgements and treating matters as non-evident. Everyday communication – including everyday talk about matters of collective concern – is communication that lacks this element of suspending naïve adherence to the world. 3.1.3 Resolving Conflict and Misunderstanding Through Deliberation The overwhelming bulk of communication is mundane talk, and the rarest form of communication is deliberation. Deliberation refers to when a special form of rational communication characterised by epoché is required to re-establish or establish understanding when the shared backdrop of mutual understanding inscribed in social systems’ language games, rites of institution, and spaces of public appearance that normally scaffolds communication is disturbed or is absent. Ideally, deliberation is used to establish or re-establish understanding when social systems’ background assumptions are disturbed through disagreement or misunderstanding, such 60 disagreement or misunderstanding over the assumptions about what makes a face or body normal, a reaction acceptable, an utterance correct or valid, or an argument adequate. Deliberation should be used to establish or re-establish understanding in the face of disagreement or misunderstanding about norms, decisions, or collective understandings and action, such as whether and which collective understandings should be consecrated in law. Ideally, deliberation (with the help of translation) is also used to establish understanding when speakers do not share the same social system’s frame of reference, such as during inter-cultural or inter-linguistic communication. Deliberation is “rational” in the sense that communication partners must establish or re-establish mutual understanding through reference to reasons, and not through custom and habit (such as that which is provided by social systems’ shared frames of reference), nor through force, coercion, or any other form of physical or symbolic violence. While the first normative precept guiding my dissertation is that democracy is a good thing, the second normative precept guiding my dissertation is that it is always better to solve disagreement or misunderstanding through deliberation, rather than through unthinking habit, force, or coercion (Chambers 1996). Deliberation at both the macro-level and at the level of face-to-face interaction is essential for challenging and modifying social systems’ structures, whether formalised laws, or informal norms, habits, and assumptions inscribed in social systems’ language games, institutions, spaces of public appearance, and habitus. Deliberation also plays an essential role in fostering belief in the legitimacy of laws and norms. To understand what I mean by this, consider the example of interracial marriage in North America. There is a long history of regulating interracial intimacy in North America, through the United States (U.S.)’s state-level anti-miscegenation laws, Canada’s Indian Act, and through informal norms dissuading interracial 61 intermixing. And yet, today there is overwhelming support for interracial marriage in both countries (Thompson 2009). Both macro- and micro-level deliberative processes were required for this shift toward acceptance. At the macro-level, discourses among U.S. state legislatures prohibiting interracial sex and marriage and the courts resulted in the landmark Loving v. Virginia (1967) Supreme Court case, which declared anti-miscegenation laws unconstitutional (Thompson 2009). This played an important role in changing the U.S. social system’s structures. Most obviously, the ruling changed the institution of marriage. But it also changed spaces of public appearance and agents’ cognitive and motivational structures. Because the ruling consecrated this aspect of marriage equality by inscribing it in law (or more accurately, it consecrated this aspect of marriage equality by erasing the legal prohibition of inter-racial marriage), it created new space for discourse in the fields of public opinion. By publicising arguments for both sides, and ultimately communicating that the case for the Lovings’ union was stronger than the state of Virginia’s case against their marriage, the U.S. Supreme Court’s ruling helped push back against the orthodoxy of anti-miscegenation norms, and expand the field of heterodox public opinion. And agents in interracial marriages would no longer face the risk of criminal prosecution, and internalising damaged (i.e., criminalised) practical identities. Norms are more enduring than laws, and even after the 1967 ruling there were certainly many who still ascribed to the orthodoxy of anti-miscegenation norms and questioned the rightness of interracial marriage. Interracial couples legally allowed to marry in 1968 would still face the exhausting endeavour of trying to draw friends, family, and colleagues into epoché – into a willingness to suspend pre-existing judgements – long enough to listen to their rational 62 justification for the legality and normative rightness of interracial unions. But each interracial couple defending their union would be playing a small but important role in gradually changing social systems’ structures through micro-level processes of rational justification and opinion-change. The distinction between everyday communication about matters of collective concern and deliberation is that everyday talk about matters of collective concern always relies upon the naïve acceptance of shared framework of understanding the world that remains unchallenged. Even in modern, pluralised, and diverse democratic social systems, where aggregate public opinion is characterised by a high degree of heterodoxy, most communication is everyday talk (and more specifically, most communication is mundane talk). This is because even though aggregate public opinion is diverse, people tend to talk about matters of collective concern with those who already share their opinions, and so do not have to hear challenges to their naïve assumptions about the world, or raise challenges to other people’s naïve assumptions (R. Huckfeldt et al. 1995; Mutz 2006). Consider two feminists watching a 2016 news clip where the then-presidential hopeful Donald Trump responded to his infamous “grab her by the pussy” comment by claiming that no one respects women as much as he does, causing one of the two feminists to snort with derision. This snort is a politically expressive act, and one that the snorting feminist’s communication partner immediately understands. This is because both feminists share the same underlying assumption – that making casual comments about sexually assaulting women is definitively disrespectful to women – making the expressive snort immediately meaningful to both parties. If the snorting feminist had been watching the news clip with a misogynist who regularly jokes about sexually assaulting women, the meaning of the feminist’s snort might be lost on their 63 communication partner (or the misogynist might disagree with the assumptions that make the snort meaningful). If the misogynist replied by saying, “Why are you snorting?” and the feminist had to explain what her snort means, the communicative snort about a matter of collective concern would – ideally – develop into deliberation, with the feminist and misogynist offering reasons for why comments about unwanted sexual touching are (or are not) harmful. Note, firstly, that the distinction between communication about matters of collective concern and deliberation is primarily whether the shared backdrop of assumptions inscribed in social systems’ structures that scaffolds linguistic communication is problematized. That is, the distinction between everyday talk and deliberation is whether naïve adherence(s) to the world are brought into explicit discussion, which typically occurs in the face of disagreement or misunderstanding. And secondly, note that when the shared backdrop of mutual understandings is challenged, deliberation is defined as resolving the disagreement or misunderstanding through recourse to reasons, and not force or violence. Political scientists tend to be interested in deliberation because of its importance for what we often denote by “politics.” Politics is often concerned with making group decisions, such as decisions about distributions of power and resources, or decisions about what constitutes normatively acceptable and unacceptable conduct, and – often – consecrating those decisions into law. When there is ambiguity or disagreement about the proper course of action, or about the laws or norms scaffolding everyday speech and action, deliberation at some stage of collective action and decision-making processes contributes to more fully-informed, thoughtful opinions, and helps ensure that collective outcomes (collective opinions and agendas, or collective decisions) are perceived as legitimate. 64 Even on those rarer occasions when political scientists study everyday communication, they often limit themselves to the small fragment of everyday talk related to public or collective concerns. For instance, Jane Mansbridge’s (1999, 214) conception of “everyday talk” is limited to “reasoning on issues that the public ought to discuss.” Similarly, Delli Carpini, Cook, and Jacobs (2004; see also Jacobs, Cook, and Delli Carpini 2009) limit their definition of “discursive politics” to those moments where “citizens talk about matters of collective concern.” But democratic theorists should also pay attention to everyday mundane talk. Together deliberation and communication about matters of collective concern only constitute a small fraction of the talking going on in social systems – the bulk of human communication is mundane chitchat. As I explain in the next section, mundane communication is essential for achieving democratic functions, particularly the reciprocity, personality, and social integration functions. And achieving these functions is essential for protecting speakers from the strain of conversations about matters of collective concern and deliberation, which are essential for achieving the learning and legitimation functions. 3.1.4 The Importance of Everyday Talk for Democracy In this section I make the case for paying greater attention to everyday mundane talk, by drawing the link between mundane talk and deliberation. Specifically, by showing how everyday communication, and particularly mundane talk, is essential for building the kinds of relational ties and facilitative attitudes that allow speakers to withstand the strain of disagreement endemic to deliberation. Talking involves raising and responding to utterances, which always requires making and responding to requests or bids for connection (Driver and Gottman 2004; Gottman 65 and Driver 2005). Reciprocally affirming bids for connection fosters social ties, develops norms of reciprocity, and contributes to the intersubjective construction of healthy identities precisely because mundane talk tends to be affirming, and, unlike deliberation, is less likely to involve conflict. Consider the following chitchat after a church bake sale: Sandra says to Cheryl: “I loved Dan’s shortbread.” Cheryl replies to Sandra with a wink and says: “Almost as good as mine!” Sandra smiles and answers: “Almost!” When Sandra made the comment about Dan’s shortbread, she is doing more than commenting on someone else’s baking. She is requesting a response from her conversation partner, hoping to connect – even if only briefly – over shared experience or knowledge. Cheryl could have rejected Sandra’s request to connect. For instance, Cheryl could have ignored Sandra and walked away in silence, or she could have responded with hostility, shouting: “EW! DAN’S COOKIES ARE DISGUSTING!” But that kind of response would be socially unacceptable and unexpected. Instead, Cheryl does the expected thing and accepts Sandra’s request for connection by replying to her utterance. By replying to Sandra’s utterance, Cheryl shows Sandra that she recognises and respects Sandra’s request for connection.2 Research shows that relationships are more likely to last when interaction partners respond to these requests, or “bids,” for connection by responding to, or “turning toward” the request, as compared to when interaction partners ignore the request or respond with hostility, or 2 Note that accepting someone’s request for connection – showing interest in their “bid” for connection – does not require agreeing with what they say. Cheryl might not love Dan’s shortbread, and she does not have to be disingenuous to accept Sandra’s request to connect (for instance, she does not have to lie and agree that the shortbread is good). Even if she hates Dan’s baking she could respond: “That’s very nice of you to say, you should tell Dan, he would be delighted!” 66 in the psychologists’ lingo, “turn away” from the request (Driver and Gottman 2004; Gottman and Driver 2005). Mundane talk may not be particularly important for promoting learning or opinion-change – mundane talk is often merely expressive and is aimed toward nothing more than talk itself – but because mundane chitchat involves reciprocally offering and responding to bids for connection, mundane chatting fosters social ties. Furthermore, mundane chitchat relies upon and develops expectations of communicative reciprocity: Sandra offers a bid for connection expecting that Cheryl will affirm it; Cheryl affirms Sandra’s bid; Cheryl now expects that when she makes a bid for connection (when she says something) Sandra will reciprocally affirm it by replying or acknowledging the utterance. Mundane communication also plays a central role in developing personal and social identities. Consider how Cheryl, in addition to turning toward Sandra’s bid for connection, goes on to make her own request for connection by making a comment that Dan’s cookies are almost as good as hers, revealing her own self-perception related to her baking abilities (Cheryl reveals that she thinks she’s good at baking shortbread too). This sequence in the communicative interaction serves a personality function: Cheryl makes a self-revelation, and Sandra recognises and affirms Cheryl’s expression of self. This interaction is mundane in the sense that shortbread cookies are unimportant for collective life, and baking efficacy is probably a relatively small part of Cheryl’s global self-evaluation. But these kinds of communicative sequences are important because, by making and responding to small requests for connection and recognition, speakers accomplish personality functions that are essential for developing a healthy sense of self, and accomplish reciprocity functions related to developing mutual respect and liking. Healthy (non-debased, non-narcissistic) personalities and strong interpersonal ties will help Cheryl and Sandra withstand the strain of disagreement if their conversation should turn to matters of collective 67 concern, or become deliberation. This kind of mundane talk – and the democracy-supporting social integration, reciprocal attitudes, and personality traits it supports – is essential for self-development. Mundane talk and the democratic functions it achieves also play an important role in preparing speakers for self- and collective-rule, by preparing them for the potential strain of conflict that can arise when conversing about matters of collective concern or when deliberating. Imagine now that, after reading about the recent police shooting of an unarmed black man, Cheryl – a black woman – mentions to Sandra – a white woman – that she thinks police officers should wear body cameras. Sandra – who is also a police officer – takes personal offense and immediately disagrees, telling Cheryl that it is hurtful to automatically assume people like her need to be monitored with body cameras. Thankfully, Cheryl and Sandra are not approaching this disagreement as two, irreconcilable solitudes – a white police officer and a person of colour – but also, as two people who trust one another, and who rely on one another for friendship and connection. Sandra’s kneejerk response might be to reject Cheryl’s claim, but Sandra respects her friend and will ultimately be willing to listen to why Cheryl thinks police officers need body cameras; and Cheryl respects and cares for Sandra enough to want to understand her friend’s personal offense. The resulting discourse might test their friendship, but their mutual respect and care for one another will help prevent deliberation from breaking down into a fight. And although discourse may be difficult, the deliberative process will help Cheryl and Sandra learn about different perspectives, and sharpen or possibly change their minds. Cheryl and Sandra’s friendship is what social psychologists interested in intergroup relations refer to as a “cross-cutting” social tie, or “intergroup contact” (e.g., Allport 1954; 68 Brewer and Gaertner 2001; R. Brown and Hewstone 2005; Christ et al. 2014; Gaertner, Dovidio, and Bachman 1996; Pettigrew 1979, 1998). A sociologist would describe their tie – or describe the common church attendance or church bake sale that gave rise to the friendship in the first place – as “bridging social capital” (e.g., Hooghe and Stolle 2003; Putnam 1995a; Stolle 2003; Stolle and Rochon 1998; M. E. Warren 2001). The social scientist’s fascination with these kinds of relationships stems from the fact that Cheryl and Sandra’s communicative interactions at church events achieve functions that generalise to interactions with other people, including strangers (Pettigrew and Troop 2011; Pettigrew 1997). If Sandra had never met Cheryl, and had only ever interacted with other white people – if Sandra had only ever made and responded to small requests for connection and recognition with other white people – Sandra might not respond very well when a black stranger challenges her sense of self with a racially-charged issue such as police brutality (either through face-to-face communication, or a mediated message in a newspaper, novel, or televised interview). If Sandra only had social ties with other white people, Sandra and the black stranger’s communicative disagreement might result in conflict (such as a shouting match, or Sandra’s unthinking rejection of a claim she receives through the media or popular culture), and so never make it to the level of deliberation. But because Sandra has developed mutual respect and trust with black friends, her sense of who deserves mutual respect, empathy, and trust is broadened to include (at least) both white and black speakers (Pettigrew and Troop 2011; Pettigrew 1997). And as such, if a black stranger challenges Sandra’s implicit understandings or beliefs related to a racially-charged issue that is deeply personal for Officer Sandra, Sandra should be more likely to extend the speaker the courtesy of epoché – to deliberately suspend her naïve adherence to the world at least 69 momentarily – and to listen and consider the speaker’s justifications. The non-hierarchical cross-cutting interpersonal ties, healthy self-identities, and reciprocal attitudes that Cheryl and Sandra foster through everyday chitchat help them withstand the strain of disagreement if their conversation turns to matters of collective concern, and helps ensure their communicative disagreement becomes deliberation rather than a shouting match. And furthermore, the protective force that social connection, and reciprocal facilitative attitudes such as empathy and trust developed through chitchat provides against the strain of disagreement is generalised to conversations and disagreements with others, even strangers. Political scientists and democratic theorists tend to focus on how deliberation can clarify perspectives and contribute to perceptions of legitimacy for deliberative outcomes. By contrast, social psychologists interested in how intergroup contact can reduce prejudice, and sociologists concerned with how crosscutting associations (and interactions) contribute to bridging social capital, are interested in everyday interactions and conversations – the bulk of which is chitchat on mundane topics – rather than those rarer instances when people talk about matters of collective concern, or deliberate. My dissertation pulls together the concerns and findings in political science and political theory, as well as in social psychology and sociology to offer a broader, integrated understanding of how communication – everyday mundane chitchat, everyday talk about matters of collective concern, and deliberation – contribute to collective opinion and agenda formation. 70 3.2 Using Talk to Achieve the Collective Opinion and Agenda Formation Functions As I explained, talking is less effective for achieving the democratic empowered inclusion and decision-making functions. However, talk is indispensable for achieving collective opinion and agenda formation. As I discussed in Chapter 2, to assess a political systems’ democratic problem-solving capacities related to the collective opinion and agenda formation function, we should ask: What does talk (or other generic tools) need to accomplish to effectively link individual preferences to collective agendas? I identify five distinguishable components of collective opinion and agenda formation: democracy-supporting social integration, reciprocity, personality, learning, and legitimation aims. In this section, I show that everyday talk is best for achieving democracy-supporting social integration, reciprocity, and personality aims, while deliberation is best for achieving the epistemic and legitimation functions of democracy. 3.2.1 Democracy-Supporting Social Integration Functions The first step to effectively linking individual preferences to collective agendas is creating opportunities to communicatively link individual preferences to a larger collectivity. In other words, people must be linked to a broader collective of people to communicatively form agendas with others. Social ties are required to create opportunities for communicatively linking individual preferences to collective agendas. In other words, collective opinion and agenda formation always requires satisfying a democratic social integration function. 71 Talking promotes social integration because talking is the primary tool by which people create and maintain relationships.3 People typically play multiple roles in society and belong to multiple networks, including constellations of friends, family, colleagues, neighbourhoods, and associations or other social activities. Democracy-supporting social integration refers specifically to forming non-hierarchical relationships with others both within and across salient social group boundaries. Conditions of structural equality between salient social groups creates opportunities and motivates people to develop non-hierarchical social ties with others, particularly across social group boundaries (Harell and Stolle 2010, 2015; Putnam 2000). Non-hierarchical, crosscutting social ties function democratically because they create opportunities for communication and motivate people to find common ground, which is essential for self-development (expressing one’s self, and one’s thoughts or preferences to others), and participating in collective self-rule under conditions that provide relatively equal opportunities to influence collective judgements and decisions. Everyday talk, and particularly mundane talk, is best for developing social ties in the first place. Recall that mundane talk is often aimed at nothing other than talk itself, at reciprocally affirming bids for connection. The act of reciprocally affirming bids for connections fosters social ties, precisely because it tends to be affirming, and, unlike deliberation, rarely involves conflict. Once democratic social integration – non-hierarchical social ties to a diverse range of communication partners – has been achieved through everyday talk, democratic social integration helps people withstand the strain of diversity and disagreement. For example, a 3 It is no accident that people who cannot talk or have difficulty talking (either because of physical or cognitive disabilities prevent them from speaking or hearing, or because they are facing language barriers) have trouble developing new social ties and are at greater risk for social isolation (Nawyn et al. 2012; Weinstein and Ventry 1982). 72 homophobic person and their homosexual cousin may find themselves in disagreement, but because they are “stuck together” by their familial tie they may be more likely to keep trying to make themselves understood and to understand the other, and to treat one another as moral equals in conversation, even conversations marred by disagreement. Perhaps just as importantly, because social ties create new opportunities (or iterated opportunities) for talking and motivate agents to resolve conflict and find common ground, non-hierarchical relationships that cross intergroup boundaries catalyse talking’s other functions. For instance, because cross-cutting social ties create diverse communication networks, these diverse, non-hierarchical social relations can catalyse talking’s democratic learning function, by creating opportunities for speakers and listeners to learn about different views and opinions. Because cross-cutting social ties create diverse communication networks, they can catalyse talking’s reciprocity functions related to incurring mutual respect for a range of diverse discussants. Consider the example of the Grandview-Woodland Citizens’ Assembly on municipal planning, which took place in Vancouver, British Columbia in 2015. The Grandview-Woodland neighbourhood is a mostly bureaucratic designation, comprised of several smaller subareas that residents normally think of as their neighbourhoods. Early assembly conversations revealed that most assembly members’ pre-existing social activities and neighbourhood ties were localised within the smaller sub-areas where they resided, rather than in the broader Grandview-Woodland neighbourhood. Participating on the Grandview-Woodland Citizens’ Assembly helped integrate assembly members into a somewhat artificial community by creating non-hierarchical social ties with a diverse range of other Grandview-Woodland residents. Assembly organisers started the processes by focusing on talking about “easier” topics, such as broad, common values marked by overwhelming consensus (common values such as 73 “community,” “health,” and so on), before deliberating through more contentious issues, such as neighbourhood density. During this initial phase, as well as during social events (during lunch breaks, or at beers after the assembly meetings) assembly members could get to know one another, and develop social ties by chatting about common experiences, their families, their jobs, and so on. Through these largely unproblematized conversations about matters of common concern (e.g., their common values) and through mundane chitchat (often about personal affinities), assembly members fostered social ties to one another, and thus to people from different neighbourhood subareas who approached the assembly with different local experiences and concerns. These social ties helped protect conversations when topics turn to more contentious issues, and helped motivate speakers to work through disagreements by deliberating, rather than by fighting. 3.2.2 Reciprocity Functions The second component feature of the agenda and collective agenda formation function I identify is the democracy-supporting reciprocity function. Mansbridge and colleagues (2012, 11) define democracy’s “ethical” function as promoting “mutual respect among citizens.” Mutual respect refers to understanding and recognising another agent’s moral status “as a self-authoring source of reasons and claims.” The moral basis for mutual respect in democracy is the idea that citizens “should be treated not merely as objects of legislation… but as autonomous agents who take part in the governance of their society” (Gutmann and Thompson 2004 quoted in Mansbridge et al. 2012, p. 11). Or to put it more broadly, people should be treated as ends in and of themselves, and not merely as means to some other end (Kant 2008). 74 Building on the idea that the moral basis for mutual respect in democratic social life is that mutual respect motivates people to treat others as ends, I define this “ethical function” more broadly in terms of a reciprocity function. The reciprocity function refers to when talking contributes to a range of reciprocal values and attitudes that function to encourage people to treat others as ends, and to allow people to assume others will treat them as ends in return. These attitudes – or what Harell and Stolle (2010, 2015) alternatively refer to as “facilitative attitudes” – include encouraging mutual respect, promoting positive interpersonal and intergroup affect (or at least tolerance), developing interpersonal and generalised trust, and encouraging empathy. Collective opinion and agenda formation always entails realising reciprocity aims because people need to be willing or even motivated to link their preferences with others to form collective agendas, which means that they must be willing and motivated to listen to others, and afford others the respect of taking their preferences into account. Furthermore, they must be willing and motivated to express themselves to others. Because self-disclosure entails vulnerability, agents must have faith that others will respect them enough to take their preferences into account. In other words, the facilitative attitudes (including mutual respect, trust, mutual liking or at least tolerance, and empathy) that lubricate social interaction and communication by promoting reciprocity are required for people to communicatively form individual wills into collective wills. As I explained, mutual respect refers to reciprocally recognising and respecting others as self-authoring agents. In practice, this means supporting the basic rights of others to participate in self-development and self- and collective-rule, including the rights of others to participate in 75 conversations and decisions that affect them.4 Positive affect refers to actual liking (warm feelings) toward other individuals or social groups, whereas tolerance refers to a willingness to accept others even in the absence of positive feelings toward them. Generally, positive affect is a strong motivator for treating others as ends – people often treat their loved ones with more kindness than they would show for themselves. And even though basic tolerance might not motivate people to treat others better than they would treat themselves, tolerance should at least safeguard against a willingness to use others as means. Interpersonal trust refers to having confidence in other people (for instance, confidence in their honesty or decency), whereas generalised trust refers to when this confidence extends beyond face-to-face interactions – beyond friends, family, and acquaintances – and is extended to strangers (Stolle 2002). It may not be immediately obvious why trust is considered a reciprocal function; after all, social scientists often value trust because it reduces transaction costs, including the effort associated with having to monitor and guard against others’ opportunistic behaviour (Stolle 2002; M. E. Warren 1999). But trust reflects people’s faith that others will not treat them as means – that is, interpersonal and generalised trust reflect people’s faith that others will not use or take advantage of them – and this faith in others increases peoples’ willingness to take risks for productive social exchange and collective endeavours, including solving collective problems and disagreements. By empathy I mean participating in emotional perspective-taking, which refers to emotional engagement combined with perspective-taking (understanding the goals, intentions, 4 Harell and Stolle (2015) refer to this facilitative value as “democratic equality,” but to avoid confusion I do not adopt this term. 76 and experiences of others) (F. B. M. De Waal 2008; Beauvais and Yaylaci 2017a). Empathy is not synonymous with positive affect – empathy involves both an affective and cognitive component, and expands and motivates the human capacity to intuit and understand the feelings, motivations, intentions, and experiences of others. Empathy is “essential for the regulation of social interactions, coordinated activity, and cooperation toward shared goals” (De Waal, 2008, p. 282). Because empathy enables and motivates people to understand and recognise others as intentional agents like themselves, empathy motivates people to reciprocally treat one another as ends. These reciprocity functions are of course interrelated – for example, empathy can deepen feelings of mutual concern among fellow citizens (Morrell 2007, 2010), reduce out-group prejudice (M. H. Davis 1994; Finlay and Stephan 2000; Galinsky and Moskowitz 2000), and promote tolerance for opposing views (Mutz 2002a). But what should be obvious by now is that mutual respect, positive feelings, tolerance, trust, and empathy all involve norms of reciprocity. When people respect and have warm feelings for others, they generally assume the objects of their respect and affection do (or would) return the sentiment. Even in the absence of positive feelings, people tolerate others – even those they do not particularly like – in part because they expect to be tolerated in return. Similarly, when people extend their trust to others, they generally expect to be trusted in return. Even empathy is evolutionarily linked to reciprocity: Many evolutionary theories use the concept of “reciprocal altruism” – which refers to the expectation that altruistic acts will be reciprocated – to explain empathy’s evolutionary fitness and origins (Axelrod 1984; Hamilton 1964; Trivers 1971). The Golden Rule’s edict to “treat others as you would like to be treated,” 77 which is reflected in at least seven of the world’s major religions (Browne 1961),5 is not simply a moral prescription. Rather, the Golden Rule is a description of the attitudes that motivate people to treat others as ends, and that allow people to assume that others are motivated to treat them as ends in return. Talking accomplishes these interrelated democratic reciprocity functions, in part because talking always implies a presupposition of reciprocity, and participating in reciprocal interactions iteratively deepens norms of reciprocity. Talking implies norms of reciprocity in the sense that people make utterances with the aim of being understood – they elaborate enough to be understood, tell the truth, make relevant utterances, and speak clearly and not obscurely – and expect that others do the same in return (Grice 1975; see also Bourdieu 2000, 122). Of course, this presupposition of reciprocity is constantly transgressed in real conversations: people over-explain or say too little, withhold information or lie, make irrelevant statements to change the topic, use non-literal language (metaphor, simile, meiosis, or hyperbole), and so on.6 But the 5 In alphabetical order, by religion: ▪ Brahmanism: “This is the sum of duty: Do naught unto others which would cause you pain if done to you” (Mahabharata 5:1517). ▪ Buddhism: “Hurt not others in ways that you yourself would find hurtful” (Udanavarga 5:18). ▪ Christianity: “All things whatsoever ye would that men should do to you, do ye even so to them: for this is the Law and the Prophets” (Matthew 7:12). ▪ Confucianism: “Is there one maxim which ought to be acted upon throughout one’s whole life? Surely it is the maxim of loving kindness: Do not unto others what you would not have them do onto you” (Analects 15:23). ▪ Islam: “No one of you is a believer until he desires for his brother that which he desires for himself” (Sunan) ▪ Judaism: “What is hateful to you, do not do to your fellow man. That is the entire Law; the rest is commentary” (Talmud, Shabbat 3 I a). ▪ Toaism: “Regard your neighbour’s gain as your own gain, and your neighbour’s loss as your own loss” (T’ai-Shang Kan-Ying P’ien). 6 Grice (1975, 75) refers to this as the “Cooperative Principle” (CP), which is summed up as: “Make your conversational contribution such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged.” The CP has four “maxims” (not laws, because these maxims are constantly flouted or violated): quantity (say only what you need to make yourself understood), quality (tell the truth), relevance (say what is appropriate to the conversation or situation), manner (be clear and not obscure). 78 presupposition is always implied in the sense that the norm of reciprocity (that we speak and listen with the aim of achieving mutual comprehension) can be invoked at any time: a conversation partner can ask, “What do you mean?” In these instances, the norm of reciprocity must be respected – the speaker must clarify their meaning – for communication to work.7 In the right context (e.g., egalitarian contexts), social interaction and communication are the most important mechanisms for developing reciprocal norms of trust (Stolle 2003; e.g., Putnam 1993, 1995b). Talking also contributes to mutual respect (J. Mansbridge et al. 2012), and, under the right conditions, social interaction and communication promote tolerance and even positive affect for outgroups (Allport 1954; R. Brown and Hewstone 2005; Pettigrew 1998). Despite the fact that empathy is an innate trait that is present at birth (F. B. De Waal 2008; Singer 2006a), empathy it is not a static predisposition but can be learned and developed through communication (Beauvais and Yaylaci 2017a; Morrell 2007; Stepien and Baernstein 2006). As Morrell (2007, 381) argues, “increasing citizens’ empathetic predispositions should be an important part of democratic education.” Democratic reciprocity outcomes help prepare people for participating in self- and collective-rule. Developing a solid baseline of mutual liking or at least strongly internalising the values of mutual respect and tolerance, engraining trust, and developing healthy levels of empathy through everyday talk is required so that these facilitative attitudes can help speakers reach individual and collective judgements and decisions. And, as my earlier example of Sandra 7 When I say the norm “must” be respected, I mean for communication to be successful (communication partners understand one another). Obviously, a speaker can still fail to make themselves understood (they remain obtuse, or simply have not mastered the language they are speaking and cannot offer a clarification), or they repeat a lie. But in these situations, communication tends to (eventually) fall apart. Even in the case of an ongoing lie, repeated social interaction and communication will only last until the untruth is discovered, in which case the repeated violation of the maxim of truthfulness (an violation of the norm of reciprocity) will damage or even end the relationship. 79 and Cheryl illustrates, facilitative values can help speakers withstand the strain of disagreement endemic to deliberation. 3.2.3 Democracy-Supporting Personality Functions The third component feature of the agenda and collective agenda formation function I identify is the development of democracy-supporting personalities. Collective opinion and agenda formation always entails achieving a democratic personality function because, just as people must have confidence in others before they can engage in self-disclosure and communicatively express their preferences to form collective agendas, people must have a healthy degree of confidence in themselves. A healthy, non-debased degree of self-confidence and self-efficacy is required so speakers can weather the strain of vulnerability to express themselves and their preferences to others. A healthy, non-narcissistic (unrealistically exaggerated) degree of self-confidence and self-efficacy is also helpful, since narcissists may be unwilling to reconsider or modify their own preferences to form them into collective agendas. As Erving Goffman (1967, 45) paradoxically noted, “human nature is not a very human thing.” People’s sense of self is developed through the interplay of interactions with others (and imagined others) in social settings (Bourdieu 2000; Giddens 1984; Goffman 1967, 2009, Habermas 1984, 1998; Mead 1934; M. E. Warren 1993). The idea that that we should expect egalitarian settings to produce individual capacities necessary to democracy has a long history in democratic theory (M. E. Warren 1993). Talking with others achieves a democratic personality function when it contributes to personalities that enhances people’s capacity to participate in 80 individual and collective self-rule under conditions that provide relatively equal opportunities to influence collective judgements and decisions. I limit the scope of my analysis of democratic personality dispositions to include self-efficacy and self-esteem.8 Self-efficacy refers to a person’s belief in their ability to succeed in activities, accomplish tasks, and accomplish goals. The interrelated notion of self-esteem refers to an individual’s global evaluations of self-worth or value. Efficacy relates more to questions of capacity, or “can” (“can I ski down black diamond slopes?”), whereas esteem relates more to “feeling” (“do I like myself?”).9 3.2.4 Learning Functions The fourth component feature of the agenda and collective agenda formation function I identify is the democratic learning function. The democratic learning (or “epistemic”) function is to “produce preferences, opinions, and decisions that are appropriately informed by facts and logic and are the outcome of substantive and meaningful reasons” (J. Mansbridge et al. 2012, 11). Collective opinion and agenda formation should entail satisfying a democratic learning function because, once people have the opportunity and basic willingness to talk to one another, 8 Of course, there are many other features of personal personality beside self- and collective- efficacy and esteem produced by deliberative “self-transformation,” including norms of reciprocity and capacity for critical reflection. However, I wish to maintain a conceptual distinction between talking’s democratic reciprocity, learning, and “personality” functions. The personality functions listed here are essentially personality traits that are not clearly related to norms of reciprocity (which, as I have explained, involve the reciprocity aim of treating others as ends), and which are not clearly related to learning or other learning aims. In other words, self- and collective efficacy and esteem are the personality traits that are “leftover” once I sort the reciprocity and epistemic (learning) traits into their respective categories. 9 Note that although the terms self-esteem and self-confidence are often used interchangeably, “confidence” typically indicates trait specific attributes (such as “confidence playing baseball,” or “confidence in knowledge”), rather than global self-evaluations. 81 and once people have the basic capacity to express themselves to others and be open to the expressions of others, they should be able to learn by engaging in communicative processes with one another. Through communicative processes – by developing preferences and opinions (participating in self-rule) and communicatively linking their preferences with others to form collective agendas (participating in collective rule) – people discover and share new information and considerations, and sharpen (and possibly change and improve) one another’s personal preferences. The relationship between democracy and epistemic quality is neatly summed up by Delli Carpini and Keeter (1997, 8), who state that “political information is to democratic politics what money is to economics: it is the currency of citizenship.” Unfortunately, it seems that citizenship has a liquidity problem: among voters, ignorance of basic political information is endemic (Bartels 2005; Delli Carpini and Keeter 1997), misinformation abounds (Kuklinski et al. 2000), and voters are often fail to link their preferences to policy choices (Bartels 2005). Discourse achieves its learning function when it counterbalances the informational shortcomings of electoral politics. For instance, when deliberative polls increase knowledge and help participants link preferences to policy-choices (Fishkin 1997; Luskin, Fishkin, and Jowell 2002), and, with respect to elite and interpersonal discourses, when exposure to conflicting perspectives successfully guards against framing (and, by extension, potentially guards against manipulation and misinformation) (Chong and Druckman 2007; Druckman and Nelson 2003). Consider again my example of the Grandview-Woodland Citizens’ Assembly on municipal planning: when assembly members talked amongst one other, and talked with experts, advocates and members of the public, they learned about available decisions and their consequences, refined their opinions with others, and reached more informed individual and 82 collective decisions. The learning function is “democratic” in the sense that developing preferences and opinions that are informed by facts, logic, and meaningful reasons better prepares people for participating in individual and collective self-rule. There is overwhelming evidence to suggest that talking – and especially deliberation – helps members of the public develop policy preferences that are “informed, enlightened, and authentic” (Page 1996, 1). Discourse is said to help members of the public improve their understanding of their own preferences, and strengthen their ability to justify their positions with arguments (Chambers 1996; Gutmann and Thompson 1996). When discourse precedes decision-making, the resulting decisions should be more informed by relevant perspectives and evidence (Chambers 1996; Goodin 2008). Disagreements or misunderstandings clarified by debate or other contestatory discussion formats (such as dialectical inquiry or devil’s advocacy) produce more critical evaluation of underlying assumptions and can produce better quality collective decisions (Beauvais and Bächtiger 2016; David M. Schweiger, William R. Sandberg, and Ragan 1986). Through debate and deliberation, speakers challenge one others’ arguments and positions, which can unravel inconsistencies, unearth new and unconsidered facts, and force participants to provide more, and more robust, reasons for their positions, and ultimately advance learning aims (Manin 2005). Deliberation is the most effective form of talk for achieving learning functions. Because mundane talk is less likely to involve disagreement, and more likely to involve consensual communication formats, mundane talk is less effective for achieving these democratic learning functions. 83 3.2.5 Legitimation Functions The fifth component feature of the agenda and collective agenda formation function I identify is legitimation. What makes norms – laws, rules, principles, and standards – acceptable? Following the revisionist deliberative democratic definition of democratic legitimacy (Benhabib 1992; Chambers 1996; McCarthy 1992), I define democratically legitimate norms as those reached through processes that could meet with the approval of all affected. Collective opinion and agenda formation should entail achieving a legitimation function because “discursively generated agreements” (and understandings) should underwrite commitment to communicatively constituted collective agendas (Warren, 2017, p. 46). People who have the opportunity, motivation, and capacity to participate in communicative processes can then learn from one another and form their individual preferences into collective wills. And when people are empowered to participate in communicative processes, and refine their own preferences with others, they are more likely to develop confidence or faith in the resulting collective agendas. Different tools may be used in conjunction with talking to realise democratically legitimate norms. For instance, a law passed by parliament – which typically involves using the tools of representation, talking, and majority vote – would be perceived as legitimate when the different processes (related to selecting representatives, communication processes, and finally, the decisive vote) meet with the approval of those who are affected by the resulting laws. If any one of these tools – and thus the process generating the law – lacks general approval, the democratic legitimacy of the law is undermined. For instance, consider how Canadian voters are represented by non-elected senators in the Canadian parliament’s upper house. If the senators exercised their constitutional powers to significantly change legislation that successfully passed 84 the elected lower house, the democratic legitimacy of the resulting law would likely be questioned by Canadians dissatisfied by the kind of non-electoral representation embodied by their senate. Communication plays an important role in realising the goal of democratic legitimacy largely because talking is the vehicle by which reasons or justifications can be articulated, challenged, and refined. Legitimation straddles both the collective agenda formation function and the collective decision-making function, and talking under relatively equal conditions also helps ensure other democratic tools that are better suited for achieving democratic decision-making function legitimately. Consider representing and voting, which are important democratic decision-making tool in mass democracies. But the features of representing and voting – how representatives are selected (whether appointed or elected), or how votes are counted – should be justified through communicative processes, so that representing and voting can help achieve the goal of democratic legitimacy. For instance, Canadian senators must be able to communicatively justify why their appointment to the senate empowers them to make democratically legitimate decisions on behalf of constituent citizens. If senators are unable to communicatively convince Canadians that they can legitimately author laws on their behalf, the legitimacy of the legislation they author suffers. 3.3 Inequality Prevents Talk from Achieving Collective Opinion and Agenda Formation As I discussed in Chapter 2, empowered inclusion requires equality broadly understood: not only political equality, but also a degree of material, and interactional (including discursive) equality. Different forms of inequality – asymmetries in political rights, asymmetries of wealth, 85 or asymmetries of participation or influence in social and communicative practices between social group members – generate power asymmetries between social group members that often entail politically meaningful exclusions. Most importantly for my discussion here, inequality and exclusion can prevent each of the collective opinion and agenda formation function components from functioning democratically. Equality and the empowered inclusion of those affected by collective outcomes is requisite for talking to achieve democracy-supporting social integration. Under conditions of extreme inequality (such as in totalitarian political systems), communication may not occur at all or may fail to foster social formation, contributing to a socially exclusive society suffering from anomie, or lack of social ties (Harell and Stolle 2010, 27). More often, when a society is structured by group-based asymmetries in political rights, asymmetries of wealth, or asymmetries of communicative participation or influence between social groups exist, talking may reinforce “isolated pockets of networks that do not interconnect,” contributing to a socially exclusive society suffering from structural fragmentation (Harell and Stolle 2010, 27). In other words, under conditions of structural inequality talking may contribute to a kind of social integration, but it is less likely to achieve the democracy-supporting social integration function of promoting non-hierarchical ties both within and across intergroup boundaries. Equality and the empowered inclusion of those affected by collective outcomes is requisite for talking to achieve democracy-supporting reciprocal attitudes: facilitative values that orient people to treat others (including members of other social groups) as ends, and that give people confidence that others (including members of other social groups) will reciprocate. As I mentioned, asymmetrical political rights, economic-gaps, or resource-gaps between social group members generate asymmetries in power relations between social group members that reduce the 86 likelihood that cross-cutting talk happens and all. And when people in heterogeneous societies do not talk to one another, diversity can undermine facilitative attitudes. For instance, in diverse communities where people do not talk to one another, diversity has a negative impact on trust (Stolle, Soroka, and Johnston 2008). And it may be impossible (and inapt) to expect disempowered social group members to feel generalised trust for dominant, empowered group members (Arneil 2006; see also Harell and Stolle 2010). Under conditions of inequality, when talking does occur between social group members it is less likely to achieve the reciprocity aims of promoting facilitative attitudes. Under conditions of inequality, talking is more likely to contribute to negative outgroup feelings (such as explicit dislike, or implicit prejudice) (Christ et al. 2014; Marschall and Stolle 2004; Stolle and Harell 2013; Stolle and Rochon 1998). Under conditions of equality, talking contributes to democracy-supporting personalities in the sense that the social contexts should contribute to healthy evaluations of self-worth and self-efficacy, and should distribute esteem in an egalitarian manner. To understand what I mean by this, consider how conditions of significant structural inequality generate asymmetries in opportunities and influence in ways that distort a person’s or group’s judgements of efficacy or esteem. With respect to efficacy, consider the example of when societies are marred by significant, institutionalised gender inequalities that generate asymmetries in men’s and women’s participation and influence in economic and political life. In such societies, people may systematically overestimate men’s capacities, and underestimate women’s capacities (as individuals, as well as collectively) (Kiefer and Sekaquaptewa 2007a, 2007b; Rudman, Greenwald, and McGhee 2001). With respect to esteem, consider the example of large economic inequalities. At the individual level, the child of a powerful billionaire who has experienced a lifetime of influence 87 and wealth may develop an unhealthily high sense of self-worth (e.g., a narcissistic personality). Conversely, the child of a pauper who has experienced a lifetime of disempowerment and poverty irrespective of their hard work and accomplishments may develop an unhealthily low sense of worth (e.g., chronically low self-esteem) (Twenge and Campbell 2002). These problems should be less prevalent under conditions of structural equality, since equality should enable talking to contribute self- and collective identities that reflect a healthy evaluation of a person’s or group’s abilities, and should distribute esteem symmetrically between social group members. Personalities are largely developed through primary socialisation in the family, although they can change in adulthood. Consider again the example of the Grandview-Woodland Citizens’ Assembly, which brought neighbourhood residents together and, through egalitarian conditions, empowered them to discursively form judgments and make collective decisions, and to communicate their judgments and decisions directly to local decision-makers. This process should boost assembly members’ sense of collective (assembly) efficacy, and perhaps even each members’ self-efficacy related to speaking and decision-making. Empirical research suggests that deliberation can increase self-efficacy (Knobloch and Gastil 2015). At the very least, deliberation and collective decision-making increases “situationally specific” internal efficacy, which refers to “measures of efficacy specific to the deliberative process” (Morrell 2005, 60). Equality and the empowered inclusion of those affected by collective outcomes is also requisite for talking to achieve democratic epistemic outcomes. Again, when social systems’ structures engender group-based asymmetries in political rights, asymmetries of wealth, or asymmetries of communicative participation or influence among social groups, societies are more likely to be characterised by structural fragmentation, by isolated pockets of networks that 88 do not interconnect, and members of different social groups will be less likely to communicate at all. When members of a society do not talk to one another, they cannot learn about one another’s preferences, or hear challenges to their own presuppositions. Structural fragmentation contributes to epistemic shortcomings in public opinion and collective wills. Furthermore, such conditions could contribute to “echo chamber effects,” and group polarisation (Cacciatore, Scheufele, and Iyengar 2016; Iyengar and Hahn 2009; H. T. Williams et al. 2015). Finally, equality is crucial for ensuring that talking contributes to collective and individual judgements and decisions that are perceived as legitimate, because when people are empowered by egalitarian conditions to participate in collective self-rule under conditions that provide relatively equal chances to influence collective judgements and decisions, they are more likely to perceive the outcomes that they helped author (for instance, resulting judgements, norms, and laws) as legitimate. This is particularly true when the learning and reciprocity functions are satisfied, and people can feel confident that their judgements are based on facts and an awareness of diverse (including opposing) views, are attentive to others, and that norms, laws, and decisions are based on good reasons and an awareness of how they affect others. Returning to my example of the Grandview-Woodland planning process, the assembly members had greater confidence in the assembly’s proposals because they largely reflected the opinions assembly members developed while talking things through with community members, stakeholders, and experts, and after considering how proposals would impact other community members (Beauvais and Warren 2017). Inequality presents a problem for democracy because it prevents generic tools from achieving normatively essential democratic functions. As my discussion shows, inequality prevents talking from achieving a range of collective opinion and agenda formation functions, 89 including the democratic social integration, reciprocity, personality, learning, and legitimation functions. But this is not the only problem of inequality for talk and democracy. As I explain in Chapter 6, “Inequality and Oppression,” under conditions of inequality, talk can contribute to harms of oppression. In Chapter 6, I also explain how inequality also blocks disagreements and misunderstandings from being resolved through rational discourse (deliberation). 3.4 Conclusion In Chapter 2, I introduced the idea that democracy begins with empowered inclusion, and that the other functions – including collective opinion and agenda formation – cannot function democratically if those affected by the outcomes of collective agenda formation and decision-making are excluded from discursively linking private preferences into collective agendas (and ultimately, collective decisions). My discussion in this chapter extends the problem-based, systems’ approach to democratic theory in two main ways. First, I consider talking more generally, rather than focusing on deliberation specifically. “Talking” encompasses both everyday talk (both everyday talk about matters of collective concern, and mundane chitchat) and deliberation. Second, I build-out the description of the collective opinion and agenda formation function, identifying five components necessary for creating opportunities, motivations, and capacities for, as well as and satisfaction with collective opinion and agenda formation: democracy-supporting social integration, reciprocity, learning, personality, and legitimation. Of course, these are not hard distinctions, and some of the component functions I identify as part of the collective opinion and agenda formation function may also overlap with other 90 broader democratic functions (for instance, legitimation is a component part of both collective agenda formation and collective decision-making). Furthermore, this list is not meant to be exhaustive – there may be additional normatively necessary components comprising collective opinion and agenda formation not listed here. But my discussion in this chapter helps answer the question of what talk needs to accomplish to effectively link individual preferences to collective agendas. My discussion also illustrates how substantive equalities – equal political rights, material equality, and discursive equality – empower inclusions, and help ensure that communicative processes engender non-hierarchical social ties that crosscut salient social group boundaries, encourage people to reciprocally treat one another as ends, engender healthy, non-debased personalities, produce more informed and enlightened opinions, and develop confidence in the legitimacy of communicative outcomes. With the right opportunities, motivation, and capacities, people can democratically form their individual preferences into collective opinions and agendas. As I explained, everyday talk is better than deliberation for achieving the democratic social integration, reciprocity, and personality functions, while deliberation is best for achieving the democratic learning and legitimation functions. Inequality prevents talking from achieving a range of collective opinion and agenda formation functions – but this is not the only problem of inequality for talk and democracy. As I will discuss in Chapter 6, under conditions of inequality, talk can contribute to harms of oppression, and inequality also blocks problematized communication from being solved through discourse (contributing to domination). But before I turn my attention to the relationship between inequality and oppression and domination, I must first address an essential question that I have so far neglected: who is this agent, this self, with claims to empowerment? In the next chapter I 91 describe the agent with claims to empowerment as a “practical comprehension” or “bodily knowledge” of the material and social world, a system of dispositions inscribed by their social and material conditions of existence (Bourdieu 2000, 135). And in Chapter 5, I discuss how social spaces create contexts or underlying frameworks that give meaning to agents’ speech and action, so that agents can reveal who they are through speaking and doing. My discussion in these two chapters is essential for understanding how structural political and material inequalities become inscribed in agents’ cognitive and motivational structures to engender interactional inequalities – in particularly, asymmetries of communicative participation and influence – which contribute to harms of oppression and domination. 92 Chapter 4: Practical Knowledge My dissertation begins with the democratic ideal that people should be empowered to participate in self-development and self- and collective-rule, and considers the problem of when structural inequalities engender asymmetrical empowerments that entail exclusions from practices that achieve these aims. Yet so far I have glossed over the “question of the subject” plaguing all disciplines taking the subject as its object (Bourdieu 2000, 128). Who is this self with claims to empowerment? In this chapter I explain the self as a “practical comprehension” of the material and social world (2000, 135). The body is conditioned by material social experiences, to produce a “self” who internalises and so practically understands the world into which they have been inserted. Agents’ capacity for mutual understanding – for participating in intersubjectivity (the capacity and eagerness to share experiences and emotional states with others), and more cognitively-oriented mind reading skills (attributing the mental states of others) – allows them to inscribe mutual understandings into language, institutions, and habitus’ cognitive and motivational structures, and so to internalise a practical comprehension of the material and social world. Understanding how the world conditions bodies is central for the aim of empowering people to participate in and influence communicative processes of collective opinion and agenda formation. This is because conditions of structural inequality can systematically engender “selves” with weakened capacities to participate in, or influence communicative processes, which prevents talking from achieving democracy-supporting social integration, reciprocity, personality, learning, and legitimation goals. My primary concern in this chapter (and the next), is with humans’ cumulative cultural evolution, which refers to the way agents inscribe socially-93 shared intentions in things (in language games, institutions, spaces of public appearance, and other cultural artifacts comprising social systems) and minds (in cognitive and motivational durable dispositions). Particularly, my goal is to understand how agents inscribe structural inequalities in things and minds to engender asymmetrical interactional (and especially discursive) participation and influence. Although I am primarily interested in cultural evolution, I begin with a discussion of the mind; of brains and neural processes. I begin my discussion of cultural evolution with a discussion of minds, because the human body, “thanks to its senses and its brain, has the capacity to be present to what is outside of itself, in the world” (Bourdieu 2000, 135). Agents’ physiological qualities make them open to the world, to being “impressed and durably modified by it” (Bourdieu 2000, 135). In the first section of this chapter, I explain the most primal, sensorimotor experiences of intersubjectivity. Specifically, when the same or similar neurons fire both an actor engaging in goal-oriented behaviour’s brain and an onlooker’s brain to produce goal-oriented, sensorimotor comprehension, as well as when the same or similar neurons are activated in an actor and an onlooker’s brains to produce emotional contagion. In the second section, I describe how humans are both capable and intrinsically motivated to expand and give permanence to these fleeting sensorimotor experiences of intersubjectivity by inscribing shared intentions into things – language games, institutions, spaces of public appearance, and other cultural artifacts – comprising social systems. I then describe how humans also expand and give permanence to intersubjectivity by inscribing shared intentions in bodies, in cognitive and motivational structures. In the fourth section, I describe the two factors that give rise to the experience of a common-sense world, to a shared backdrop of mutual understanding that enables smooth social interaction and everyday talk. These factors are: isomorphism 94 (congruence) between habitus and the social and material world, and homology (similarities) between the cognitive and motivational dispositions of similarly situated people (Bourdieu 1977, 2000). Finally, I describe human agency’s role in changing social systems, even though the common-sense world can act as a conservative counterweight to social change. 4.1 The Sensorimotor Experience of Intersubjectivity As Bourdieu (2000, 135) explains, the human body, “thanks to its senses and its brain, has the capacity to be present to what is outside of itself, in the world, to be impressed and durably modified by it.” Where does the body’s openness to the world come from? In particularly, where does the body’s attentiveness to others, and tendency to attune itself with sociality come from? Humans, like many other non-human social mammals (including non-human primates) experience a primal, sensorimotor experience of intersubjectivity, a physiological capacity to share affective or cognitive states with others by recruiting their own neural systems (Di Pellegrino et al. 1992; Gallese et al. 1996; Iacoboni 2009; Rizzolatti and Craighero 2004; Rizzolatti et al. 1988). By “sensorimotor” intersubjectivity I am referring to the experience of intersubjectivity as automatic, neural brain impulses. Mirror neuron systems allow humans (and other social animals, but I limit my analysis to humans) to corporeally represent the actions and intentions of others, by recruiting their own motor systems. The mirror neuron system in humans’ and other social non-human animals’ prefrontal cortexes allow social animals to embody the states of others and so immediately come to non-conceptual, social understandings. Mirror neurons fire when an animal perceives another 95 performing a goal-oriented action (reaching for something, moving their lips or bodies in an intentional way), and internally simulates the goal- and emotion-oriented behaviours of others (Di Pellegrino et al. 1992; Gallese et al. 1996; Iacoboni 2009; Rizzolatti and Craighero 2004; Rizzolatti et al. 1988). To understand how this works, imagine two acquaintances, Estella and Pip, having tea. When Estella lifts her cup to her lips, neurons are activated in her brain; when Pip watches Estella lift her cup to her lips, the same (or similar) neurons are activated in his brain. Mirror neurons are the primal mechanism enabling the shared understanding of the perceptual and motor aspects of goal-oriented actions across the divide of self and other (e.g., the goal of sipping tea from a cup). In short, the mirror matching system provides the neurological scaffolding for a basic, sensorimotor experience of intersubjectivity (Gallese, Ferrari, and Umiltà 2002; Gallese 2001, 2003; Iacoboni 2009; Rizzolatti and Craighero 2004). Motor neurons code the goals of motor acts, rather than merely the movements forming them (Fogassi et al. 2005; Rochat et al. 2010; Umiltà et al. 2008) (Alexander and Crutcher 1990; Crutcher and Alexander 1990; Kakei, Hoffman, and Strick 1999, 2001). Rather than simply coding observed (and performed) actions, these neurons are coding the intention associated with the action. Neurons coding specific acts (such as grasping) show different activations when the act is part of different goal-directed sequences, such as grasping for placing a cup, versus grasping for drinking from a cup (Fogassi et al. 2005). Estella’s neurons fire differentially when she grasps her cup to place it in on a tray, versus when she grasps her cup to drink from it. This is as true of observing goal-directed sequences performed by another. And Pip’s neurons fire differentially when he observes Estella grasp her cup to place it down, versus observing Estella grasp her cup to drink from it. Context appears to be important in shaping goal-96 coding: picking up a cup stimulates different neurons depending on whether objects are displayed just before the tea is consumed (the “drinking” context) or after tea time (the “cleaning” context) (Iacoboni et al. 2005). The mirror system attunes agents – it brings them into accord with, it adjusts or accustoms them – to social contingency (awareness of social context). In addition to (and perhaps in conjunction with) the mirror neuron system, other neurons simulate automatic, emotional responses that allow humans to corporeally share affective states through emotional contagion. Emotional contagion works by a process whereby observing the emotional states of others triggers the associated somatic and automatic responses in the observer (Gallese, Eagle, and Migone 2007; Gallese 2003; Preston and De Waal 2002; Singer 2006b). Examples of emotional contagion include the welling of tears when people see others crying, or yawning in response to seeing other people (or other animals) yawn. These neural impulses enable a practical experience of intersubjectivity, corporeally bridging the divide between self and other. But these sensorimotor experiences of intersubjectivity are limited in scope. Because these corporeal experiences require physically observing the goal-oriented actions or affective states of others, they require visual – often face-to-face or small-group (and possibly video) – interactions. A limited number of bodies can participate in these practical experiences of intersubjectivity at a given time. Furthermore, these forms of intersubjectivity last only as long as the interaction, and end when agents part ways and can no longer perceive each other. These sensorimotor experiences of intersubjectivity are fleeting, impermanent. Humans expand and give permanence to this fragile intersubjective space, and to participate in a broader, more permanent shared conceptual understandings. Humans seem intrinsically motivated to inscribe shared intersubjective understandings in things (language 97 games, institutions, spaces of public appearance, and other cultural features of social systems), and in bodies (the cognitive and motivational structures of habitus). There is no consensus on the extent other, non-human social animals participate in sharing intentions, and inscribing shared intentions in things and bodies to participate in language games and shared cultural scripts, or participate in the symbolic achievements language makes possible (such as mind reading, or intuiting others’ thoughts, intentions, and goals). But this question goes beyond the scope of my dissertation. What matters for my purposes is that humans are motivated to expand and give permanence to shared understandings and intentions in language and other social institutions, and in bodies, and do participate in the symbolic achievements – mind reading and emotional perspective-taking – that language makes possible. 4.2 Inscribing Shared Intentions in Things: Language and Institutions The biological property of being open to the world goes beyond the immediately and temporary sensorimotor impulses that allow agents to corporeally comprehend interaction partners’ emotions and goal-oriented behaviours. It also includes the ability to inscribe shared conceptual space in social systems, through the language games and other institutions. Evolutionary anthropologists describe the socio-cognitive skills that make inscribing shared conceptual space into social systems possible as “shared intentionality” (Burkart, Hrdy, and Van Schaik 2009; Hrdy 2011; Tomasello 2010; Tomasello and Carpenter 2007).10 Shared 10 For an excellent discussion of the evolutionary development of shared intentionality, see Sarah Blaffer Hrdy’s (2011) book, Mothers and Others: The Evolutionary Origins of Mutual Understanding, which convincingly argues that human ancestors’ prosocial tendencies manifested as “cooperative breeding,” which refers to when 98 intentionality refers to the cognitive ability and motivation to “participate with others in collaborative activities with shared goals and intentions” (Tomasello et al. 2005, 675). There are a few distinct socio-cognitive shared intentionality skills that allow agents to inscribe shared intentions and understandings in language, institutions, and other cultural artifacts in social systems, but the most important skills for my discussion are: triadic attention, cooperative communication, and imitative learning. Triadic attention refers simply to interactions that involve the “referential triangle” of (1) self, (2) other (interaction partner), and (3) some external focal point (e.g., an object, a third party, or an underlying referent) that both interaction partners are attending to (Tomasello 2000). This seemingly basic socio-cognitive skill is an essential stepping stone for inscribing shared understandings into the language games and institutions that help comprise the shared perception of a common-sense world. At the most basic level, triadic attention enables social referencing (using other people as emotional reference points) and gaze following (looking where others are looking, to attend to the same external focal point). Triadic attention entails going beyond dyadic (person-to-person) interaction, and sets the stage for cooperative communication and imitative learning. “Cooperative communication”11 “alloparents” (or conspecifics besides a birth mother, including fathers, aunts, grandparents, or others) collectively help raise infants and children, giving these pre-human, hyper-social primates a selective advantage. 11 Cooperative communication (and language games) are only “cooperative” in the Gricean sense. Grice’s (1975, 75) “Cooperative Principle” (“Make your conversational contribution such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged”) describes the way that, in order for conversations to function successfully conversation partners must make utterances with the aim of being understood, and assume that their interaction partners are doing the same. The Cooperative Principle (CP) functions even when interlocutors are not cooperating socially, such as during an argument. The CP is not a sociological law, but rather an implicit presupposition of language games that encourages mutual participation in an underlying framework of conceptual space. The CP is “a specific variant of the principle of reciprocity, which, although it is constantly transgressed, can be invoked at any time as” (Bourdieu 2000, 122). Refer to my discussion in Chapter 3. 99 refers to communication that relies on, or references shared conceptual ground (Tomasello and Carpenter 2007). This kind of communication is triadic in the sense that it always involves (1) a speaker, (2) listener(s), and (3) a shared frame of attention (e.g., the context within which communication takes place) that communication partners can mutually reference and understand. Clearly, the capacity for multiple minds to share information about referential intentions is required for developing symbols and language. The ability to share common ground (to have a joint attentional frame) is necessary for both comprehending and designating symbols, and for understanding utterances in conversation. That is, the capacity to participate in triadic (self, interaction partner, referent) interaction is essential for inscribing shared referential intentions in symbols and language games. Symbols and language games expand and give permanence to the corporeal experience of intersubjectivity, by expanding and giving permanence to a space of mutually accessible common ground. Imitative learning is another shared intentionality skill that helps agents expand and give permanence to shared conceptual space by inscribing it in language, institutions, and other cultural artifacts in social systems. Imitative learning differs from other forms of social learning (learning through others) in the degree to which it is culturally-mediated (Carpenter 2006; Tomasello 2004; Tomasello and Carpenter 2007). Imitative learning involves trying to reproduce others’ intentional activities or behavioural strategy. Imitative learning contrasts with emulation learning, a form of social learning that involves copying another’s behaviour to change the state of the environment, rather than to learn a conspecific’s intentional activity or behavioural strategy (Tomasello 2004). Researchers have devised experiments to distinguish between imitative learning and emulative learning, to see if children and non-human primates will copy a clearly inefficient 100 goal-oriented behaviour, such as using a rake inefficiently to get an out of reach object (Nagell, Olguin, and Tomasello 1993), or inefficiently illuminating a toy box containing a lightbulb by pressing on the box with the top of their head instead of simply using their hands to illuminate the lightbulb in the box (Meltzoff 1988). The researchers find that non-human primates tend to learn by emulation: if they see a researcher use a rake inefficiently to obtain an object, they copy the researcher’s tool use, but they use the rake efficiently, to more effectively change the state of the environment (and get the out of reach object). Apes do not try and replicated researchers’ intentions by imitating the inefficient method for obtaining the object. By contrast, human children try and reproduce the researcher’s behavioural strategy, imitating the inefficient behaviour to obtain the out of reach object. This is less effective for changing the state of the environment and getting the out of reach object, but the children’s imitation signifies to the researchers that they are “in tune” with the situation, that they are playing by the rules of the game (Carpenter 2006). As Tomasello (2004, 53) explains, imitative learning is “not a ‘higher’ or ‘more intelligent’ learning strategy than emulation; it is simply a more culturally-mediated strategy.” Clearly, the capacity to participate in triadic (self, interaction partner, referent) interaction is essential for participating in language games, which entail referencing mutually available conceptual space in ways that are meaningful to all conversations partners. Perhaps less obviously, but just as importantly, triadic interaction motivates imitative learning. Just as the shared knowledge that an utterance signifies an underlying referent gives that utterance meaning, the shared knowledge that an action might signify an underlying referent gives that action a surplus (social) meaning. Agents’ act purposefully not only to change the state of the environment, also to signify an underlying referent or meaning. 101 When children imitate their elders’ actions, they are not simply trying to achieve the identical outcome (i.e., getting an out of reach object, or illuminating a lightbulb in a toy box). Rather, they are trying to discover the underlying conceptual space that makes the action meaningful by participating in the action in the “right” way. They know that they are signifying by doing, even though they might not know what they are signifying. They are learning the rules by playing the game. A child imitates their parents because they perceive their parents’ actions – the way their parents walk, speak, sit, or engage in any other intentional activity – as “constitutive of the social being of the accomplished adult” (Bourdieu 2000, 154). Children do not just want the ability to achieve certain ends, to move physically through space, or make vocalizations. Children also imitatively learn to be an “adult” (or a “woman,” “Canadian,” etc.), who signifies who or what they are by walking, speaking, sitting and doing other acts in a certain way. Before I turn my attention to the conditioning processes that inscribe intersubjective understandings (e.g., understandings of how a woman should walk, talk, and speak) in bodies, I wish to highlight imitative learning’s normative aspect. Children not only imitatively learn to perform the activity when they see an adult perform the activity for the first time, but they also seem to see that activity in normative terms – how “we” do things. In a study of the normative structure of conventional games, experimenters taught young children how to play a simple game (Rakoczy, Warneken, and Tomasello 2008). After the experimenter and a child had played the game a while, a puppet (controlled by another experimenter) asked to join the game, and played the game “wrong.” Almost all the children protested, and the older children could protest very explicitly. 102 Imitating the actions of social referents (such as parents) is not simply about learning to change the state of the environment, it is about learning and incorporating the rules of the game, and playing the “right” way. Imitative learning helps agents to share meanings with others by teaching them to recognise and participate in social acts in the same way, expanding shared conceptual space across bodies. It gives permanence to a shared background of mutual understanding by ensuring social actions’ can be inscribed in language games and institutions (the “rules of the game”) faithfully imitated by subsequent generations. The ability to participate in language games and other social games in socially recognised settings is essential for human self-development (Young 2011). 4.3 Inscribing Shared Intentions in Minds: Habitus So far, I have discussed how shared intentions become inscribed into language games and institutions. In this section, I try – as much as possible – to describe in general terms how these features of social systems act back upon bodies to conform bodies to the conditions of their existence. The processes of primary conditioning that inscribe shared intentions in cognitive and motivational dispositions (in “habitus”) is something humans, in general, participate in. Of course, as my discussion shows, it is difficult to discuss primary conditioning in general terms (as something all humans do) without discussing social groups, and the social distinctions that maintain power relations and social hierarchies. Rest assured, in Chapter 5 I discuss social differentiation – in collective social identifications, in fields (comprised of institutions and distributions of resources and power), and in processes that consecrate social categories and 103 distinctions into law – at length. And in Chapter 6, I discuss the consequences of inscribing group-based inequalities in social systems for democratic or oppressive governance. In this section, I define habitus, and discuss (as generally as possible) how shared understandings and intentions inscribed in language and institutions work back upon the body, structuring habitus’ systems of cognitive and motivational dispositions. Habitus are schemes of thought and expression that enable “the intentionless invention of regulated improvisation” (Bourdieu 1977, 16:79). The concept of habitus is essential for understanding people as socialised beings, influenced by material conditions and social relations. As Bourdieu explains, “The ‘I’ that practically comprehends physical space and social space… is not necessarily a ‘subject’ in the sense of philosophies of the mind, but rather a habitus, a system of dispositions.” Habitus is a practical or embodied comprehension of the world – a “corporeal knowledge” – that provides an “immediate understanding of the familiar world” (Bourdieu 2000, 135). Habitus are the persisting effect of “primary conditioning,” acquired mostly through primary socialisation in the family, but also developed through the interaction between primary socialisation and education or vocational training later in life (Bourdieu 1990, 62, see also 2000, 164). Evolutionary or cultural anthropologists studying humans’ capacity and motivation to share intentions and understandings – to share and understand thoughts and feelings with others, and participate in intersubjective engagement – stress the degree to which participating in shared intentionality skills’ products (language, institutions, and other cultural artifacts in social systems) is cooperative (e.g., Burkart, Hrdy, and Van Schaik 2009; Tomasello 2009b, 2010; Warneken, Chen, and Tomasello 2006). Language games are “cooperative” in the Gricean sense that speakers must make utterances (or signals) with the aim of being understood: signifiers and referents must be 104 intelligible to listeners.12 If the utterances are unintelligible, listeners should be able to request and receive clarification, or else language games fall apart.13 Similarly, actions signify underlying meanings and agents must act cooperatively – they must figure out and follow the rules of the game – for social actions to be mutually comprehensible. However, to say participating in language games and other social systems features is cooperative suggests interlocutors already have the same linguistic and cultural knowledge and skills required to reference a mutually understood framework of meaning. But this is never immediately the case: benighted agents (e.g., children) must be brought into (assimilated), or must appropriate linguistic and cultural competences to make use of an underlying framework of meaning. Agents must co-opt, or be co-opted into participating in language games, institutions, spaces of public appearance, and other cultural artifacts in social systems; they must co-opt or be co-opted into playing by the rules of the game. Co-optation can be conceptually distinguished as taking two interrelated forms: as instruction, and as an expression of power. Teaching and learning linguistic and cultural competences is the most innocuous and common way of co-opting naïve participants into the shared intentionality inscribed in social systems’ features, and of naïve participants themselves actively co-opting spaces of shared intentionality inscribed in language games and social 12 Grice (1975, 75) refers to this as the “Cooperative Principle” (CP), which is summed up as: “Make your conversational contribution such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged.” The CP has four “maxims” (not laws, because these maxims are constantly flouted or violated): quantity (say only what you need to make yourself understood), quality (tell the truth), relevance (say what is appropriate to the conversation or situation), manner (be clear and not obscure). 13 Of course, a speaker can still fail to make themselves understood (they remain obtuse, or simply have not mastered the language they are speaking and cannot offer a clarification), or they repeat a lie. But in these situations, communication will (eventually) fall apart. Even in the case of an ongoing lie, repeated social interaction and communication will only last until the untruth is discovered, in which case the repeated violation of the maxim of truthfulness will damage or even end the relationship. 105 systems’ other structures. Perhaps most obviously, this occurs when language-speakers and culture-knowers – typically caregivers – introduce their benighted young to the signifiers and cultural scripts the young have not yet learned. A caregiver, such as a father, might try to teach the association between an utterance and a thing in the world by repeating the signifier in the presence of the signified, such as by repeatedly uttering “papa” while pointing to himself, until the child responds to, and eventually says, “papa.” Of course, children – being innately inclined to share mutual understanding – also actively co-opt caregivers’ language games, as any parent learns when they first hear their child utter the parent’s favourite curse word. Recall from my discussion in the previous section, how imitative learning motivates children to seek out the rules of the game, to show they are in tune with the other players and can play the right way. As I explained, a child imitates their parents because they perceive their parents’ actions – the way their parents walk, speak, or engage in any other intentional activity – as “constitutive of the social being of the accomplished adult” (Bourdieu 2000, 154). They are capable and motivated to participate in social actions, to play the game the “right way.” Co-optation through teaching and learning aspects of the shared intentionality inscribed in features of social systems also happens when those who are already proficient in a language or culture try to adopt another. When an agent decides to take language classes in Québec, out of a love for the French language, she is co-opting the French language, by adopting it as her own. Co-optation by sharing or exchanging signifiers across languages also occurs when, for instance, two members from different language groups, who are competent in their respective languages, meet, and – to communicate with one another – begin to exchange signifiers. When Manie, a Cantonese speaker, meets Mirela, a Serbian speaker, and asks: “Bējáu?” Mirela does not 106 understand. Manie makes a gesture like she is lifting something to her mouth to drink. “Ah!” Mirela says, “pivo!” and orders two beers. They cheer, each repeating their word and the newly learned word for beer, “bējáu/pivo.” In this instance, the two members of the different language communities mutually co-opt signifiers from each other’s language, and the repository of known signifiers expands for both parties. Participants in shared language games, institutions, spaces of public appearance, and other cultural artifacts inscribed in social systems may also try to forcibly coerce non-participants into participating in, and expanding the shared intentionality inscribed their social systems’ structures. For instance, by forcing others to speak their language, practice their traditions, and take part in their institutions and spaces of social play. This motivation is driven in part by the normative element that people seem to attach to cultural products – recall even young children taught to play a new game automatically adopt a normative attitude of “this is how we do things.” Furthermore, as I will discuss at greater length in Chapter 5, increasing the number of people who participate in social systems’ spaces of shared intentionality preserves and expands spaces of mutual intelligibility. Recruiting new members (even forcibly) – inscribing social systems’ structures in more bodies (bodies being inclusive of the brain and nervous system) – helps expand the scope and guarantee the longevity of social systems’ spaces of mutual understanding. Relatedly, forcibly co-opting others to participate in the shared intentionality inscribed in social systems’ structures increases the co-opting agents’ ability to participate in easy communication with minimal effort. This is because the burden of learning new signifiers is shifted onto the shoulders of those who are forced to learn – there is no mutual “exchange” of signifiers and norms when a more powerful group forcibly assimilates a disempowered group. 107 Ease of communication is necessary for achieving mutual understanding, which may be particularly valuable to empowered group members in hierarchical societies: empowered group members (such as colonial rulers and settlers) will be able to issue directives or commands, and be understood, without having to learn the disempowered group members’ (such as colonised peoples’) languages.14 Efforts to draw others into one’s shared intentionality inscribed in features of social systems are not just limited to obvious majorities or empowered groups. The Québécois in Canada, minority speakers on a majority English-speaking continent, have made special efforts to preserve the Québécois social system’s language games, institutions, and other cultural artifacts (for instance, by requiring that the children of parents educated in any language other than English attend French school, or else attend private school). In these examples of co-optation as an expression of power, agents actively try to draw others into their shared intentionality. But the process of being co-opted into a shared intentionality inscribed in features of social systems does not always involve deliberate efforts to co-opt new recruits. Indirect co-optation into the shared intentionality inscribed in language, institutions, and other cultural features occurs through what Jacques Derrida (1998) describes as “the interdict.” The interdict refers to when socio-economic forces – for instance, that English is 14 Another form of coercion is when participants in empowered social groups try to prevent the disempowered (or minorities) from being able to co-opt the shared intentionality inscribed in their forbearers’ social systems. As I discuss in greater detail in Chapter 5, eliminating the disempowered unique linguistic and cultural competences is often an important goal for empowered or majority cultural group members who wish to co-opt subordinated or minority group members into the dominant shared intentionality inscribed in the symbolic and institutional features of social systems. Participating in a subaltern shared intentionality – such as by speaking another language that members of the empowered group do not understand – opens-up avenues for subversion. Eliminating the languages and cultural competences of disempowered or minority groups decreases their capacity to subvert existing hierarchies. Alternatively, empowered groups may prevent the disempowered from adopting certain linguistic or cultural competences, to ensure inter-group boundaries remain intact. For instance, in the United Kingdom, social class is made audible through dialects and speech, and ideological systems demarcating acceptable and unacceptable behaviour prevent lower class members from simply affecting a higher-class accent. 108 the language of commerce in Canada, and most of Canada’s prestigious universities offer instruction in English – ensure that most people will need or prefer to learn English. Today, English-speaking authorities do not need to institutionalise Ojibway-speaking children to ensure they learn English. The institutions and fields of education (including the education system, but also other cultural products and fields, such as music, television, and radio) and the realities of the economy largely ensure that Ojibway children will speak English. I return to this idea in Chapter 5, where I discuss social differentiation (into collective social identities and fields), and as the motivations for colonialism and co-optation as an expression of power. In the remainder of this section, I describe techniques for co-opting agents, and humans’ practical knowledge’s consequences – bodies shaped by the social and material conditions of their existence’s consequences – for social and political life. The social order inscribes itself in bodies through confrontation, and agents are co-opted into participating in language games, institutions, spaces of public appearance, and other cultural artifacts in social systems through everyday pedagogic action and rites of institution. Everyday pedagogic action refers to imposing the culture of a social group on new inductees (e.g., children) whether through family education, institutionalised education, or diffuse education (through fellow social group members), to reproduce the social order (Bourdieu and Passeron 1970, 4). For instance, pedagogic action helps maintain differences between the sexes, which are taught and learned through clothing and appropriate dress, and even through appropriate walking, talking, sitting, and speaking. A parent engages in pedagogic action when they admonish their daughter to “sit like a lady.” Rites of institution refer to the acts of “performative magic” whereby a person assumes their “social fiction,” where the “social image or essence that is conferred on him [or her/them] 109 in the form of names, titles, degrees, posts or honours, and to incarnate it as a legal person, the ordinary or extraordinary member of the group” (Bourdieu 2000, 243). Rites of institution establish the “definitive differences between those who have undergone the rite (for instance, circumcision) and those who have not (for example, women),” through a collective and public performative act (Bourdieu 2000, 175). As Bourdieu (2000, 218) stresses, rights of institutions are simply an extension of pedagogic action. Rites of institution are “an extra-ordinary” performative instance of “the continuous, infinitesimal and often unnoticed actions that every group exerts on its members… addressed to the child and help to shape his [their] representation of his [their] (generic or individual) capacity to act.” Every small pedagogic action, including acts of nomination – terms of address (Miss, Mrs., Dr.), or terms of reference (such as gendered pronouns) – inscribes a person’s social being, perceived value, and capacity to act in habitus. Both pedagogic action and rites of institutions naturalise social classifications, such as the male/female division, “in the form of divisions in bodies [inclusive of brains]” (as male/female predispositions). These predispositions may take the form of sports preferences (hockey/dance) or colour preferences (blue/pink), or gendered distinctions in speaking, walking, and sitting (Bourdieu 2000, 141). However, because rites of institution play such a clear role in assigning rights and duties (in distributing power), they are also “rights of investiture” (Bourdieu 2000, 243), “favouring initial investment in the game” (Bourdieu 2000, 141). By reassuring the group member of their legitimacy as a full member of the group, rites of institution help guarantee the group member’s commitment to the group. This benefits the group, because it ensures the group is “known and recognized”, which helps maintain the group’s existence (Bourdieu 2000, 243). 110 The automatic, sensorimotor affective responses that attune agents to social interactions, and make them especially suited to cooperation also promotes co-optation into social systems. Bourdieu describes the primary conditioning of pedagogic action and rites of institution as “psychosomatic action,” that that imprints itself in the “memory pad” of the body, through “emotion and suffering” (Bourdieu 2000, 141). Practical knowledge takes the form of emotion, “the unease of someone who is out of place, or the ease that comes from being in one’s place” (Bourdieu 2000, 141). Embarrassment in social situations often stems from “unfulfilled expectations” (Goffman 1967). Goffman (1967, 105) describe how interaction participants, “given their social identities and the setting… will sense what sort of conduct ought to be maintained as the appropriate thing,” and how embarrassment ensues when these expectations go unfulfilled. Here, Goffman is describing how primary conditioning inscribes a practical, bodily knowledge of potential and present positions in social space, and how the gap between the expectations inscribed in body and the reality of the world in which the body finds itself is experienced as painful emotions (shame, embarrassment, feeling out of place). This affective response has the effect of reinforcing social classifications, and bringing the body back into accord with the social and material conditions that shaped it. Practical knowledge takes the form of emotion that motivates “behaviours such as avoidance or unconscious adjustments” (Bourdieu 2000, 184). Returning to my earlier example of how the male/female social classification is naturalised in divisions in the body as male/female dispositions, consider a boy who loves dance, or a girl who loves hockey. In societies where dance is considered a feminine pastime, the boy (who identifies as male) may be mocked for his passion. The resulting feelings of shame may dissuade him from pursuing dance, contributing to 111 the underrepresentation of boys and men in dance, and the ongoing association of dancing with femininity. In a society where hockey is associated with masculinity, the girl (who identifies as female) who loves hockey may find few girls and women play, watch, or talk about the sport. Feeling out of place, she may avoid hockey games and conversations dominated by boys and men, contributing to the underrepresentation of females in hockey culture and the ongoing association of hockey and masculinity. As this discussion shows, the implication of inscribing the social order in bodies, in habitus, is that someone who has incorporated social structures “finds” their place in the world right away, and can feel at home in social interaction. And, as I started to explain, habitus is “isomorphic” to the external (political and material) world in the sense that objective factors – including political rights and material inequalities – shape schemes of thought and expression. This ensures the agent’s habitus is roughly adjusted to their world’s structures and tendencies, and thus guarantees the agent is spontaneously attuned to their world’s structures and tendencies (Bourdieu 2000, 139). In Bourdieu’s (2000, 140) words, “we are disposed because we are exposed.” Because agents developed dispositions through exposure and confrontation with the world’s structures and tendencies, habitus dispositions become “the accomplices of the processes that tend to make the probable a reality” (Bourdieu 1990, 64). For instance, those – such as members of the working class – who are objectively less likely to be able to be able to attend university may be taught that a university degree is a frivolous waste of money, creating a disposition (aversion to university) that makes the probable (not attending university) a reality. No two agents have identical habitus (since no two agents have identical life experiences), but the convergent experiences of the same social group (including social class) members unites group members in a relationship of homology. Homogeneity of conditions of 112 existence homogenises group or class habitus, which can homogenise practices without any direct interaction, calculation, or explicit co-ordination. The homogeneity of group habitus is reflected in the form of statistical regularities, where social group membership (age, gender, residence, social class, ethnicity, etc.) helps predict attitudes (such as political attitudes and opinions), behaviour (such as political behaviour), and taste (such as cultural consumption). This brings me to another, related consequence of habitus: that habitus comes to define and durably maintain a common-sense world. I explain the practical comprehension of the common-sense world at length in the next section, but I want to briefly note here that the common-sense world’s self-evidence comes from harmonising agents’ experiences. That is, the common-sense world arises and is maintained when agents mutually reinforce one another’s experiences, for instance through the expression of similar or identical experiences (common festivals, common sayings, etc.). This harmonisation of experience allows agents to experience the world as “sensible” and “recognisable” without either explicit reason or signifying intent (Bourdieu 1977, 16:80). Habitus tends to favour experiences that reinforce the self-evidence of the common-sense world by motivating people to seek-out exposure to information that reinforces habitus’ self-evidences, and to avoid exposure to information that calls the common-sense world into question. 4.4 Isomorphism, Homology, and the Common-Sense World As I started to explain in Chapter 3, the bulk of human communication or language games take place against a set of shared attitudes, competencies, and practices comprising a “common-sense” or “familiar” world. The common-sense world refers to the “stock of self-113 evidences shared by all, which, within the limits of a social universe, ensures a primordial consensus on the meaning of the world, a set of tacitly accepted commonplaces which make confrontation, dialogue, competition and even conflict possible” (Bourdieu 2000, 98). The experience of the world as self-evident arises from two correlated factors: the isomorphism (congruence) of habitus and the world that structures it, and the homology (similarity) between habitus among similarly socially-situated agents. The first factor, isomorphism, refers to the “quasi perfect coincidence between habitus and habitat,” which produces an “agreement between the dispositions of the agents and the expectations or demands immanent in the world into which they are inserted” (Bourdieu 2000, 147). Recall that habitus’ dispositions become accomplices to the processes that make the probable a reality. For instance, a working-class youth might internalise a feeling of belonging that is isomorphic to their life chances (they are less likely to be admitted to a prestigious university), and so feel embarrassed at the idea of applying to university, and not apply in the first place. A gap between the expectations inscribed in habitus and experience in the world is felt as positive or negative surprise. The second factor contributing to the experience of the world as self-evident is the tendency for people situated in similar situations to have similar, or homologous habitus. Recall that although no two people have identical habitus (since no two people have identical life experiences), people living in similar conditions of existence tend to have similar cognitive and motivational dispositions inscribed in habitus that are reflected as statistical regularities. The common-sense world arises and is maintained when agents mutually reinforce one another’s experiences, for instance through the expression of similar or identical experiences (common festivals, common sayings, and so on). As I explained in the previous section, this harmonisation 114 of experience allows agents to experience the world as “sensible” and “recognisable” without either explicit reason or signifying intent (Bourdieu 1977, 16:80). Sharing a common backdrop of mutual understandings and assumptions is essential to human social and cultural life, because when interaction partners share the same tacit knowledge, they can dispense with analysing the nuances of one another’s practices. In ordinary life occasions, they do not have to tacitly or explicitly inquire: “what do you mean?” A shared stock of self-evidences makes “causes practices and works to be immediately intelligible and foreseeable, and hence taken for granted” (Bourdieu 2000, 80). This has enormous consequences for social interaction and collective endeavours, because it allows large numbers of people to participate in common collective endeavours and shared intentions, over long periods of time (including intergenerationally), without conscious effort or direction. Participating in a common-sense world enables smooth social intercourse and unproblematized talk, which – particularly everyday talk about the mundane – is essential for achieving democratic social integration, reciprocity, and personality functions. As I discussed in Chapter 3, mundane chitchat should enable speakers to form horizontal social ties, both within and across social group boundaries. And the presupposition of reciprocity implied by communication should foster the reciprocity aims of developing and deepening mutual respect, care, empathy, and trust (e.g., Grice 1975; see also Bourdieu 2000, 122). The kind of mundane chitchat that takes place against a shared stock of self-evidences is particularly effective for strengthening feelings that promote reciprocity because this kind of mundane talk largely occurs without disagreements and misunderstandings, which can strain communication and relationships, and (in the case of chronic disagreement or conflict, or lack of social understanding 115 and belonging) can contribute to personality problems such as low self-efficacy or self-esteem (Baumeister and Leary 1995; Schunk 1984). The common-sense world not only makes mutual understanding over agreements possible, but also mutual understandings over disagreements. The common-sense world’s classificatory schemes “paradoxically unite those whom they divide” (Bourdieu 2000, 100). The world’s visions and divisions – inscribed in things (language games and institutions) and in bodies (habitus’ classificatory schemes) – create “position-takings which are immediately recognized as pertinent and meaningful by the very agents whom they oppose and who are opposed to them” precisely because they are common to all agents participating in the social order (Bourdieu 2000, 100). Because this harmonisation of experience allows agents to experience the world as “sensible” and “recognisable” without either explicit reason or signifying intent, agents tend to favour experiences that reinforce the self-evidence of the common-sense and so seek-out exposure to information that reinforces habitus’ self-evidences, and to avoid exposure to information that calls the common-sense world into question (Bourdieu 1977, 16:80). Evidence of this comes from agents’ preference for homophily, or preference for social interaction with similar people (McPherson, Smith-Lovin, and Cook 2001), including agents’ preference to seek out communication partners who share their political views (Bourdieu 1990; R. Huckfeldt et al. 1995; R. R. Huckfeldt, Johnson, and Sprague 2004; Mutz 2002b, 2006). As Bourdieu (1990, 61) explains, “the schemes of perception and appreciation of the habitus which are the basis of all avoidance strategies are largely the product of a non-conscious, unwilled avoidance, whether it results automatically from the conditions of existence (for example, spatial segregation) or has been produced by a strategic intention (such as avoiding ‘bad company’ or ‘unsuitable books’).” 116 Although a shared backdrop of mutual intelligibility is essential for human self-development, and, by enabling everyday talk, can help talk serve democratic functions, the common-sense world can also undermine democratic functions and contribute to harms of oppression and to domination. The common-sense world is the product of struggles to define familiarity. That which is “self-evident, established, settled once and for all, beyond discussion” today has not always been so, but rather has “only gradually imposed itself as such” (Bourdieu 2000, 174). As I alluded in my discussion of co-optation as an expression of power, and will elaborate in Chapter 5, the dominant viewpoint – the viewpoint of the most powerful social group or groups – presents itself as universal, and comes to be experienced as familiar and commonsensical. The “historical evolution” producing the common-sense world functions to “abolish history” by relegating it to the subconscious (Bourdieu 2000, 174). Of course, the familiar world could be the product of democratic struggles to define this familiarity: everyday assumptions might include a deep-seated belief all humans are equally valuable regardless of social group membership or nationality, and they might induce people to judge other people by the content of their words and their actions, without attention to their ascribed social group attributes. Indeed, democratic struggles typically aspire to create contexts wherein everyday, unthinking chitchat, and everyday social actions promote the use of skills in socially recognised settings, and participation in social play and communication (to participate in self-development), and as well as agents’ ability to determine their actions and the conditions of their actions (to participate in self-rule) (Young 2011). I return to this idea in Chapter 8, “Achieving Discursive Equality.” Unfortunately, much of recorded history has not been characterised by democratic social and political systems that empower all social group members’ inclusion in political practices that 117 contribute to self-development and self- and collective-rule. Much of (written) human history has been characterised by oppressive systems of governance, and inequalities that entail exclusions from social, economic, and political practices. As I explained in Chapter 4, struggles to challenge a dominant vision of social space involve struggles to bring the implicit, mutual assumptions comprising orthodoxy the level of explicit statement “to break out of the silent self-evidences” of orthodoxy (Bourdieu 2000, 188). The problem, however, is that the symbolic capital required to bring orthodoxy to the level of explicit statement – and to expand the space of heterodox public opinion, and open the possibility of resolving disagreements and rectifying injustices through rational deliberation – are unevenly distributed. Asymmetrical empowerments and exclusions mean that the disempowered are less likely to be able to push back the self-evidences of orthodoxy imposed by dominant social groups, and as a result the “primordial political viewpoint is a particular viewpoint, that of the dominant, which presents and imposes itself as a universal viewpoint” (Bourdieu 2000, 174). I return to this point in Chapter 5, in my discussion of social differentiation into social groups and fields. I also take this up in Chapter 6, where I discuss how talking can contribute to harms of oppression when structural inequalities engender social group members’ exclusion from communicative participation and influence. 4.5 The Counterweight of the Common-Sense World and Human Agency Imitative learning plays a central role in maintaining the shared intentions inscribed in language games, institutions, spaces of public appearance, and other cultural artifacts comprising social systems. Imitative learning motivates children to seek out the rules of the game, to show they are in tune with the other players and can play the right way. As I explained, a child imitates 118 their parents because they perceive their parents’ actions as constitutive accomplished adults. As Tomasello (2000, 38) explains, “only cultural learning leads to cumulative cultural evolution in which the culture produces artifacts, such as tools, and symbolic artifacts, such as language and Aramaic numerals – that accumulate modifications over historical time.” When one person learns something, or creates a new tool, others can learn by imitation and new knowledge and tools are then passed on to subsequent generations, who can improve and pass on what they have learned. Being born into, and growing-up in social systems is like growing-up in “the context of something like the accumulated wisdom of its entire social group, past and present” (Tomasello 2000, 38). This produces the paradoxical effect of rapid, cumulative changes and a peculiar continuity. Being gifted with, and so being able to innovate upon an entire repertoire of existing cultural artifacts (rather than each generation literally re-inventing the wheel) produces rapid, cumulative changes. At the same time, the tendency to imitate predecessors’ shared intentions inscribed in social systems even if are not helpful – and even if they are harmful – produces peculiar continuity. Consider, for example, innovations in information and communication technologies: because accumulated wisdom is preserved, generations do not have to start from scratch when they invent tools or practices. Steve Jobs did not have to invent either the telephone or the computer to create the iPhone. Jobs could build on what already existed. However, despite adding innovations, we are constantly replicating what came before, even when it does not serve us. To take a much-cited example, consider the inefficient QWERTY keyboard that most English-speakers use for typing. The keyboard is deliberately inefficient, with the most frequently used letters placed furthest away from one another. The initial reason for this was to keep the keys on typewriters from becoming entangled. On today’s iPhone, this is clearly not a 119 concern – and yet the QWERTY keypad remains, because it is what our parents and teachers learned to type on, and taught us to type on, and what we will likely teach our children to type on. Not only are we innately prone to mimicking that which our forbearers already do, but studies of the diffusion of innovations across social systems show that some ideas are more likely to be spread than others. As Richerson and Boyd (2008) point out, the most important predictors of whether ideas spread between societies are: first, whether the innovation is easy to see; and second, whether the innovation is easy to try. And many important features of social systems’ language games and institutions – such as family structure, gender relations, and other social hierarchies – are not directly observable or easy to try out. As such, we can expect the shared intentions inscribed in the things and bodies comprising social systems to be preserved and passed down faithfully across generations, even when agents affected by social systems’ structures do not benefit from them, and even when agents are aware of alternative ways to form and inscribe collective intentions and wills in ways that would serve them better. As Richerson and Boyd (2008, 162) say, “our propensity to adopt dangerous beliefs is part of the price we pay for the marvelous power of cumulative cultural adaptation.” Because we are innately predisposed to learn by imitation, and because alternative cultural habits are difficult to observe or try out, social learning becomes like a system of inheritance, where much of a person’s behaviour is “a product of beliefs, skills, ethical norms, and social attitudes that are acquired from others with little if any modification” (Richerson and Boyd 2008, 161). This is what I refer to as the “conservative counterweight” of the common-sense world. 120 Of course, even though social systems’ structures, including habitus, determine the conditions under which humans act, they do not determine human actions. The habitus comprising the common-sense world are “structuring structures”. Although habitus are part of social structure (they are the internalisation of the social and material world) and help to set the context of human action, they also act back upon the world – agents can change the contexts of their existence. This is a moral imperative for democracy, for empowering agents to participate in self- and collective-rule. When symmetrical empowerments are inscribed in social systems (e.g., equal political rights are consecrated in law, political institutions redistribute wealth equitably) agents should be empowered to challenge the taken-for-granted assumptions of the common-sense world through normal procedures, including deliberation. Recall my example of anti-miscegenation laws from Chapter 3, where – after the U.S. Supreme Court’s Loving v. Virginia ruling legally empowered inter-racial couples to justify their relationship – agents in interracial relationships could try and draw conversation partners into epoché (suspend their naïve assumptions) and participate in deliberation to change their minds about the normative rightness of interracial marriage. Although children tend to imitate adults formed under the same conditions, the experience of oppression and injustice can stimulate efforts to change those conditions. Most young adults imitate their parents and marry people like themselves (a principle known as homogamy) (Blackwell 1998; Blackwell and Lichter 2004; McPherson, Smith-Lovin, and Cook 2001). In social systems where normative expectations include prohibitions against interracial interaction and intimacy, and these prohibitions are consecrated in law as formal prohibitions, most people would marry people of the same race, and never feel a need to question the normative rightness of these assumptions and laws. But when two people from different racial 121 categories meet, fall in love, and want to marry, a social structure’s injustice makes itself felt: the couple must choose between pursuing a forbidden romance (and suffering the consequences of marginalisation, and even physical violence), or complying with the expectations inscribed in social structures (and suffering heartbreak). This tension can motivate efforts to change social structure, such as by using the available procedures (legal challenges and deliberation). Agency is also called upon when social structures ensconcing agents change, and agents’ taken-for-granted understandings comprising the backdrop are shaken, such as during economic or socio-political revolutions, colonisation, or shorter-term crises (e.g., economic crises, wars, etc.). However, structural inequalities (including inequalities embedded in social systems as discursive inequalities) engender asymmetrical empowerments that prevent agents from changing, or influencing the social systems that structure the cognition and action. In the absence of constitutional rights, the Lovings would not have been able to challenge U.S. anti-miscegenation laws in the Supreme Court. And even after the successful legal challenge, implicit assumptions inscribed in habitus likely undermined the discursive influence of racial minorities’ utterances in everyday talk, slowing their ability to draw others into deliberation. Inequality and asymmetrical empowerments block disempowered social group members’ capacity to agentically or autonomously influence the social systems’ structures that affect them. I take this up in Chapter 6, where I discuss the problem as to how, under conditions of inequality, generic political practices (including talking) can contribute to harms of oppression. And I describe how inequality can block epoché and deliberation, contributing to domination. 122 4.6 Conclusion The self is a practical comprehension of the material and social world. The mind is conditioned by material, and – perhaps most importantly – by social experiences, to produce a “self” who internalises and so practically understands the world into which they have been inserted. Agents’ capacity for mutual understanding – collaborative activities with shared intentions and goals – allows them to inscribe mutual understandings in language, institutions, and habitus’ cognitive and motivational structures, and so to internalise a practical comprehension of the material and social world. Humans are born into existing social systems, they co-op and are co-opted into language games, institutional practices, and other cultural practices. The social world inscribes itself into their bodies through pedagogic action and rites of institution, instilling social divisions into their habitus’ cognitive and motivational structures, constructing bodies to conform to the conditions in which they find themselves. This isomorphism between habitus and the world, and the homology between similarly situated agents’ habitus creates the experience of a common-sense world. The experience of a space of mutually shared understandings and assumptions, that enable smooth social interaction and unproblematized everyday talk. Understanding how the world conditions minds is central for the aim of empowering people to participate in and influence communicative processes of collective agenda formation. This is because conditions of structural inequality can systematically engender “selves” with weakened capacities to participate in, or influence communicative processes, and prevent talk from functioning democratically. In Chapter 5, I describe how human plurality, characterised by a twofold need of belonging and distinction, contributes to social differentiation, to collective 123 social identifications as well as the proliferation of fields of social play. I also describe how structural inequalities contribute to the development of discursive inequalities. In Chapter 6, I develop this discussion and describe how, under conditions of structural inequality, talking can contribute to harms of oppression and to systems of domination. 124 Chapter 5: Spaces of Public Appearance The agent with claims to empowerment is a practical self, a system of dispositions inscribed by their social and material conditions of existence. But as my discussion of co-optation through pedagogic action and rites of institution in Chapter 4 suggests, social life does not simply entail living in the presence of other individuals. Rather, it entails living embedded in overlapping social spaces: in overlapping social identifications (with recognised fellow social group members and non-group members) as well as rule-bound social subsystems (the fields of art, law, education, etc.). Social spaces are the contexts or underlying frameworks that give meaning to agents’ speech and action, so agents can reveal who they are through speech and action. The purpose of this chapter is to describe how social systems become differentiated into spaces of public appearance (into social identities and fields), and outline the implications of societal differentiation for distributing and maintaining shared intentions – particularly, for distributing and maintaining power relations – inscribed in language games, institutions, or other cultural features of social systems. That social systems are always differentiated by overlapping social identities and by fields drives my concern with the problem of when structural inequalities engender asymmetrical empowerments that entail the systematic exclusion of disempowered social group members from political practices that contribute to self- and collective-rule. As I discussed in Chapter 2, social systems benefit from structural political, material, and interactional (including discursive) equalities. Structural political inequalities refer to gaps in political rights between social groups, whereas material inequality refers to economic or resource gaps between social groups. Interactional inequalities refer to when objective inequalities 125 structure agents’ subjectivities – their thoughts and feelings – to engender asymmetries in economic, social, and political participation or influence. Discursive inequalities are a kind of interactional inequality that refers more specifically to when objective inequalities structure agents’ subjectivities to engender asymmetries in communicative participation or influence. Objective inequalities entail external exclusions (or constraints) that prevent agents from participating in or influencing practices of self-development and collective-rule regardless of agents’ thought, beliefs, or motivations. And interactional inequalities entail internal exclusions (or constraints) that prevent agents from participating in or influencing practices of self-development and collective-rule by shaping agents’ thought, beliefs, or motivations. Let me review the example of women’s political engagement I introduced in Chapter 2, to clarify the relationship between inequality and exclusion, and the distinction between objective/intersubjective inequalities (on the one hand) and external/internal exclusions (on the other). Gendered gaps in political participation rights – such as a restricted franchise that prohibits women from voting – entails women’s external exclusion from electoral participation. Restrictive electoral laws are an external barrier preventing women from participating in and influencing elections, irrespective of women’s (or men’s) thoughts or motivations. But even after women’s enfranchisement – after formal, external legal barriers to their participation were dismantled – women did not begin voting or participating in politics at the same rate as men. This is because interactional inequalities – norms (or intersubjectively held beliefs) that politics is a man’s game, or that women are not suited for politics – shape gents’ social cognition and motivation to entail women’s internal exclusion from political engagement. 126 In the first section of this chapter I explicate a notion implied by my discussion of practical knowledge, which Hannah Arendt (2013) describes as “human plurality.” Human plurality refers to agents’ simultaneous similarity and difference from each other, which motivates social action. In the second section, I draw on social psychology to explain the motivational and cognitive bases for social identification. The simultaneous psychological need for belonging/inclusion and distinctness/separation, which social psychologists call “optimal distinctiveness,” is the motivational basis for social group identification (Brewer 1991, 2003, 2011; Brewer and Pickett 1999; Leonardelli, Pickett, and Brewer 2010). The cognitive underpinnings of social group formation stem from the need for “cognitive economy” and cognitive capacity to participate in categorisation (Gaertner et al. 1993; Hogg 2006; Hornsey 2008; Macrae and Bodenhausen 2000; Perdue et al. 1990; Turner 1987). In the third section, I describe another form of societal differentiation, what Bourdieu (1977, 1990, 2000) describes as differentiation in fields, defined as overlapping, rule-bound systems of social play. The rise of capitalism increased the number of fields and the relations between fields in social systems, contributing social systems’ growing complexity and differentiation. In the fourth section of this chapter, I turn my attention to the question of power relations, and how power relations can be created and maintained by inscribing them in institutions’, habitus’, and fields’ shared intentionality. In this section I describe how sets of government institutions, or political systems, are created to direct and protect social systems’ space of shared intentionality. Political systems differ in the degree to which they succeed in (dis)empowering those affected by them to participate in self- and collective-rule (ranging from democratic social systems to those characterised by oppression and even domination). Political 127 systems act an “organized remembrance” (Arendt 2013, 198) or “containers” (Giddens 1984, 261), storing shared intentionality and power distributions. In this fourth section, I elucidate a notion I introduced in Chapter 4: that social group members may be motivated to forcibly co-opt new agents into participating in the shared intentionality inscribed in their social systems, to expand and give permanence to their social space. This is particularly likely when preserving and expanding social space preserves or expands asymmetrical power relations benefitting the co-opting group members. In the fifth section, I discuss the ideal of political systems that direct and protect social systems’ language games, spaces of public appearance, and habitus to engender inclusions and symmetrical empowerments. 5.1 Human Plurality and the Need for Revelation and Recognition In Chapter 4 I described humans’ sensorimotor capacity for intersubjectivity, and the shared intentionality skills that enable and motivate people to inscribe intersubjective understandings in symbolic and cultural achievements such as language. I also discussed how, although agents with similar life experiences will share many similarities (they will have homologous habitus), no two agents’ experiences are ever identical, and neither are their habitus. Hannah Arendt (2013, 220) describes this simultaneous human sameness and distinctiveness as the condition of plurality, and describes human plurality as “the sine qua non for that space of appearance which is the public realm.” Human plurality is characterised by what Arendt (2013, 175) describes as “the twofold character of equality and distinction.” By “equality,” Arendt (2013, 175) means a fundamental 128 “sameness” that allows agents to “understand each other and those who came before them,” as well as to “plan for the future and foresee the needs of those who will come after them.” The similarity refers to the capacity to internalise shared understandings in habitus. By contrast, human distinctness refers to the perhaps obvious fact that humans are different from one another. Note that human distinctness does not refer to the fact that sexual reproduction ensures almost all humans are genetically different from each other (except for identical twins). Human distinctness refers to the fact that no two people have identical life experiences, and so no two people have the same practical knowledge of the world; each practical self is distinct. Even “identical” (monozygotic, or maternal) twins sharing identical genetic properties with similar life trajectories will be distinct, since they cannot have identical interactions with the same people at the same moments in time, nor identical experiences in the physical world at the same moments. When genetically identical (maternal) twins are old enough to articulate themselves – to express who they are through speech and action – they will be revealing distinct selves. While agents’ fundamental similarity makes mutual comprehension possible, agents’ fundamental distinctiveness makes efforts aimed at mutual comprehension necessary. If humans were not “distinguished from any other who is, was, or ever will be,” humans would not need to signify underlying referents or meanings using mutually comprehensible words and meaningful action. Making sounds to “communicate immediate, identical needs and wants would be enough” (Arendt 2013, 175–76). But humans do more that communicate things, they communicate their practical selves, their lifetimes of experiences. This simultaneous sameness and difference drives two fundamental psychological needs – the need for belonging and distinction – which together motivate social group identification. I 129 return to this point in the next section. But before I discuss how human plurality drives motivational basis for social identification, I must highlight the fact that practical selves must be communicated to other people. A person’s words and deeds only acquire meaning when they are understood and recognised by others. Bourdieu (2000, 239) describes the “link between three indisputable and inseparable anthropological facts” driving humans’ need for meaning and recognition from others. First: people know they are mortal. Second: the thought of death is unbearable. And, following from this, third: man is haunted by the need for recognition and justification for his temporary existence. Only other people – the social world, or what Arendt calls the space of appearances – can satisfy this quest to justify our temporary existence.15 Only other people can recognise one’s existence as something (worthwhile, great, even terrible). Without this recognition from others, humans are abandoned to “absurdity” of a life without meaning (Bourdieu 2000, 239). Living a meaningful life, having one’s words and deeds recognised by others, requires a “space of appearances” for revealing oneself to others, and recognising others in turn. Arendt’s discussion is deeply influenced by the Greek polis, and she describes the space of appearances as the public realm. In today’s pluralised societies, it makes more sense to discuss spaces of appearance, or publics. In the next section I explain why, although all healthy democracies include a broader public sphere (and we can even speak of an international public sphere), the search for optimal distinctiveness motivates agents to seek out specialised groups of social peers for everyday processes of mutual revelation and recognition that contribute to self-development. 15 Bourdieu references Pascal, who suggests meaning comes from other people (the social world), or God. I restrict my focus in this analysis to the profane. 130 “Peers” can be identified through social identities, by collectively identifying with a social identity (or social identities). Peers can also be identified through shared participation in fields, through the workplace (in the fields of law, academia, medicine, etc.), or through shared participation in leisure and consumption (in fields of cultural production such as art and music, or through fields of sports). In the subsequent two sections I describe the differentiation of societies into social identities and in fields, which act as distinct but overlapping spaces of public appearance. As my discussion below shows, societal differentiation does not mean “the” public sphere is under siege. By separating social spaces into fields that follow different logics, societal differentiation protects public spaces from the means-oriented logic of the economy, opening venues for other goal-oriented actions (such transforming private interests into collective agendas). And in healthy democratic systems, the broader public sphere should be undergirded and communicatively linked to disempowered agents’ subaltern spaces of appearances, such as the “feminist subaltern counterpublic” (Fraser 1990). 5.2 Social Differentiation through Social Identities One essential space for appearances – for self-revelation and recognition – is within collective social identities. As I noted in the previous section, human plurality (fundamental human sameness and difference) drives the psychological needs for belonging and distinction, which motivate social identification processes. This motivation, combined with the cognitive need to simplify information by organising it categorically, contributes to social differentiation. 131 In this section, I define what I mean by “social identity,” and describe the motivational and cognitive bases for social group identification. Social or collective identities refer to the parts of “an individual’s self-concept which derives from his [their] knowledge of his [their] membership of a social group… together with the value and emotional significance attached to that membership” (Tajfel 1981). Note that although social identities are often formed around social group memberships, social group (or category) memberships and social identities are not interchangeable. Group memberships may be imposed, but identities are chosen (Brewer 1991). People can “recognize that they belong to any number of social groups without adopting those classifications as social identities” (Brewer 1991). Optimal distinctiveness theory is the leading theory explaining the motivational underpinnings of identity-formation (Brewer 1991, 2003, 2011; Brewer and Pickett 1999; Leonardelli, Pickett, and Brewer 2010). This theory posits that “social identity derives from a fundamental tension between human needs for validation and similarity to others (on the one hand) and a countervailing need for uniqueness and individuation (on the other)” (Brewer 1991, 477). The psychological need for security or belonging is well-established in the psychology literature (see Baumeister and Leary 1995 for a review). And the idea that people need a degree of individuation (e.g., Codol 1984; Lemaine 1974; Maslach 1974) or uniqueness (Snyder and Fromkin 1980) is not novel either. The optimal distinctiveness theory recognises that both these needs are always present, and posits that efforts to balance these countervailing needs motivate social identification processes. Because social identities simultaneously provide inclusion (within the salient in-group) and differentiation (between group comparisons), they “allow us to be the same and 132 different at the same time” (Brewer 1991, 477). Brewer (1991, 477) gives the example of adolescent peer groups, where each cohort “develops styles of appearance and behavior that allow individual teenagers to blend in with their age mates while ‘sticking out like a sore thumb’ to their parents.” Social identities vary in their degree of inclusiveness, from very broad social groupings that include many agents with few common characteristics (such as nationalities, or gender groupings), to more exclusive categories with fewer people or many overlapping features (Vancouver’s queer women of colour). The optimal level of inclusiveness shifts when different needs are aroused. For instance, a deindividuation experience (e.g., an agent is told they are average, or indistinguishable from a larger social group) arouses the need for differentiation and motivates a preference for more exclusive social identities (Brewer, Manzi, and Shaw 1993). For instance, a woman who told she is a “typical woman” may feel inclined to highlight an aspect of her identity that distinguishes her from the socially-defined, prototypical “woman” (e.g., she may point-out that she is a computer programmer). By contrast, over-individuation leaves agents open to stigmatisation and isolation, motivating “a selective need for assimilation to a distinct in-group” (Brewer 1991, 479). When agents are highly individuated (e.g., being a queer woman of colour in Vancouver), they are motivated to form social identities based on what makes them stand out. By creating a social identity around the attribute(s) that make them stand out, they create a space of appearances where they fit in. Balancing the needs for belonging and distinctiveness explains the motivation for social identification (the “why?” question), but it does not answer the “how?” question. The cognitive underpinnings of social identification stem from the need for “cognitive economy,” and the 133 cognitive capacity for categorisation. The principle of cognitive economy asserts that we create categories to increase cognitive efficiency. Categorisation is one of the most important and basic processes of human cognition, preventing cognitive overload by simplifying large quantities of information, and organising our perception of the world (David L. Hamilton and Tina K. Trolier 1986; Hinton 2000). Understanding how agents come to see themselves and others categorically – to recognise themselves and others as social group members – is important, because social categorisation has consequences for how people think and interact. The leading approach for understanding social categorisation is the social cognitive approach. Within this approach there are two compatible theories for understanding category structure: prototype models and associative networks. The prototype model proposes two general principles for the formation of categories. The first of these principles – cognitive economy – is familiar. The principle of cognitive economy asserts that we create categories to increase cognitive efficiency. The prototype model’s second general principle is referred to as “perceived world structure.” The principle of perceived world structure asserts that objects in the world are perceived to possess a high correlational structure. Rosch (1999) gives the example of a person who, upon perceiving the attributes of feathers, fur, and wings, realises that feathers co-occur more often with wings than with fur. Of course, as Rosch points-out, our association of feathers with wings is not merely a product of the physical environment, but also the social environment; specifically, that we have already possess a cultural and linguistic category called “birds.” A prototype refers to the “most typical”’ member of a category (Operario and Fiske 2001). The prototype theory of categorisation follows Wittgenstein’s (1953) insight that humans deal with categorisation on the basis of “clearest cases,” defined as “peoples’ judgments of 134 goodness of membership in the category,” and not by reference to category boundaries (Rosch 1999, 36). Thus, category membership is determined by goodness-of-fit in the category (which is subjective, but – although there are variations across individuals – these perceptions tend to be inscribed in culture, and so consistent across members of the same cultural group), and not by strict boundaries delineating category membership. The more prototypical a category member is rated, the “more attributes it has in common with other members of the category and fewer attributes in common with members of the contrasting category” (Rosch 1999, 37; see also Rosch and Mervis 1975). Thus, for instance, a sparrow is a more prototypical (representative) member of the category “birds,” while a penguin is a less prototypical category member. Sparrows have many attributes in common with other birds (they chirp, fly, nest, etc.), while penguins do not. By contrast, penguins have many attributes that are not shared with other birds (penguins do not fly, they are excellent ocean divers, etc.), while sparrows do not. The associative network model is a second model for explaining category structure within the social cognitive approach. Associative network model is compatible with the prototype model, and helps clarify some of the basic assumptions of the prototype model. According to associative network model, information is stored in mental structures called nodes, and each node corresponds to a single concept (a name, object, personality trait, place, emotional response, etc.) (Operario and Fiske 2001; see also Carlston and Wyder 1994). Each of these nodes is interconnected by links, which map out meaningful associations between the concepts contained within each individual node. These inter-nodal links are what structure people’s mental representations – stronger links denote more significant associations between concepts, while weaker links denote less meaningful associations. The nature of nodal linkages varies with 135 perceivers’ experiences: “links increase or decrease in strength depending on the perceived correlation between concepts, and new nodal links can develop according to new associations between previously unpaired concepts” (Operario and Fiske 2001, 29). Understanding how the cognitive underpinnings of social identification functions is important for my discussion of discursive inequities in Chapter 6. Stereotypes, or cognitive categories containing assumptions about the attributes of social group members, are a particularly ubiquitous discursive inequity. As I explain in Chapter 6, experiences in the natural and social world structures the linkages between nodes, which can include links between semantic concepts, such as between social groups and human attributes (e.g., the link between the social group “women,” and the attribute of being “nurturing”). With greater exposure to the co-occurrence of semantic concepts, the link between these concepts becomes stronger, denoting a more significant association between these concepts. In Chapter 6, I discuss how stereotypes embedded in habitus’ cognitive and motivational structures can structure communication, engendering internal exclusions that entail asymmetries of communicative influence as discursive inequity. 5.3 Social Differentiation through Fields Societies can also be differentiated into fields, which, like social identities, can act as spaces of public appearance and recognition. Fields are spaces of social play, defined by a set of (often implicit) rules. Social identification is motivated by the simultaneous psychological impetuses to fit in and stand out, revealing a social self who simultaneously belongs to a group (and experiencing security, solidarity, and intimacy) and who is distinct from others (and 136 experiencing validation and recognition) are the ends in themselves. By contrast, fields are intention-oriented spaces of social action, organised around a common aim or dedicated to “the pursuit of specific goals” (Bourdieu 2000, 11). Of course, participating in intention-oriented understandings with others can bring a sense solidarity or comfort. Recall that even very young children seem intrinsically motivated to feel out and play by the “rules of the game,” developing a normative attachment to the structure of the game (“this is how we play”) that may satisfy a need to belong (Rakoczy, Warneken, and Tomasello 2008). And of course, games create spaces for validation and recognition, to be – and be recognised – as something (the best, the worst, or whatever). But fields are ostensibly organised around some other, primary aim: practicing law or medicine, the pursuit of philosophy, the creation of music, an open market for the exchange of private property for profit, and so on. An agent reveals their self, and receives recognition and glory – validation for their temporary existence – by pursuing a goal ostensibly aimed at something other than belonging or distinction. The pianist’s goal is, ostensibly, to master Chopin’s Études or Franz Liszt’s compositions. But in so doing, the pianist demonstrates their mastery of the most technically difficult piano pieces, earning them recognition and acclaim. Taking part in a field means taking the stakes of the game seriously. The stakes of the game arise from the logic of the game itself, and “establish its seriousness” not only to participants, but also “lay people” involved in other fields (e.g., even non-pianists should be able to recognise the seriousness of mastering a technically difficult piece of music) (Bourdieu 2000, 11). The field’s logic (the “feel for the game”) is established in habitus (e.g., expressed as an artistic “spirit”). Bourdieu genealogically describes the rise of Western philosophy as a distinct field. Bourdieu (2000, 18) identifies the philosophic field’s origins in Ancient Greece, describing how 137 “a universe of argument governed by its own rules” developed, wherein “everyone acted as an audience for everyone else, was constantly attentive to the others” (Bourdieu 2000, 18). The institutionalization of the scholastic order was clarified in the Middle Ages, when philosophy ceased to be a way of life and became “a purely theoretical and abstract activity… articulated in a technical language reserved for specialists” (Bourdieu 2000, 18–19). The process of differentiating philosophy from other academic fields reappeared in Renaissance Italy, and then accelerated with rise of capitalism. With capitalism, the economy was constituted “as such, in the objectivity of a separate universe, governed by its own rules, those of self-interested calculation, competition, and exploitation” (Bourdieu 2000, 19, see also 1977, 16:171–83). This discussion shows that fields can develop in any society. However, the rise of capitalism breaks off and reserves a specialised sphere for economic activity, which freed other fields to operate by logics besides the naked exchange of economic capital. Separate fields can more easily “constitute themselves as closed, separate microcosms,” in which “thoroughly symbolic, pure and (from the point of view of the economic economy) disinterested actions” can be performed (Bourdieu 2000, 19). The emergence of these distinct “universes” offers positions for perceiving the world, and for developing distinct world views (Bourdieu 2000, 20). Agents belong to multiple, overlapping fields. Fields may be larger with more amorphous boundaries, with many diverse agents participating in a wide-ranging goal or set of goals (e.g., economic exchange in a capitalist economy). Or fields can be smaller with more determinate boundaries, with fewer agents participating in a very specific goal of set of goals (e.g., Western philosophy). Fields’ boundaries are marked by where their effects end. Fields encompass social positions structured internally in terms of power relationships, such as the relationships between property owners and employees, or between tenured philosophy professors and sessional 138 instructors. Fields are also organised hierarchically in social space in relation to other fields; for instance, non-economic universes of social play may be subsumed under the economic field. Fields prescribe or favour certain actions through both formal, explicit rules as well as more implicit constraints. These constraints include the “conditions of exchange (form and forum discussion, legitimate problematic, etc.),” and, as I discussed in Chapter 4, habitus, or “the dispositions of the agents which are the product of this set of effects” (Bourdieu 2000, 112). These constraints also include the “institutionalized procedures regulating entry into the game (selection and co-option).” Entrance (co-optation and selection) into a field is regulated, particularly entrance into smaller fields with more determine boundaries and specific purposes. Regulating entrance into fields is essential because, for fields to effectively offer spaces of public appearances – so agents can “be understood and recognized” – they must exclude aspirants “who lack the necessary competence to compete effectively” (Bourdieu 2000, 112). Consider how academic disciplines try to restrict participation by including competent/ excluding incompetent peers through graduate studies admissions and exams, conference and journal rejections/acceptances, hiring practices, and so on. The overt aim of this gatekeeping is to ensure that the discipline’s peer group can competently produce, review, and critique or praise discipline-specific academic work. The primary distinction between social identities and fields is social identities allow for the expression of belonging and distinction (the end), whereas that fields are rule-bound systems ostensibly organised around other ends (although they create opportunities for belonging and distinction while pursuing other goals). Of course, social identities can be the basis of fields, and inscribed in rule-bound systems oriented to achieving specific goals. For instance, a religious identity may be inscribed in a rule-bound system oriented to achieving God’s will, or God’s 139 glory on earth, through sacred and profane practices. And fields may form the basis of identities. For instance, the field of class relations may become the basis for social identification; in Marx’s words, when a class in itself becomes a class for itself. A central thesis in Bourdieu’s corpus (e.g., Bourdieu 1977, 1984, 1990, 2000; Bourdieu and Passeron 1970) is that – even though capitalism had the effect of constituting the economy as a separate field, governed by its own rules, and enabled the proliferation of other fields, governed by logics ostensibly unrelated to naked economic exchange – all fields remain concerned with distributions of capital and power. This is implied by the fact that fields are rule-bound, goal-oriented spaces of social play that create space for mutual revelation and recognition. Knowledge of the rules (e.g., properly internalising and playing by a fields’ rules), and mastering the required skills for achieving the field’s goals, are a form of symbolic capital that give possessors the power to play the game, and reveal themselves in that field’s space of appearances. That fields remain concerned with distributions of capital and power without agents’ recognition or awareness that they are concerned with distributions of capital and power is problematic. Economic capital can be converted into symbolic capital in fields to try and obfuscate overt domination, to transform overt domination into misrecognised, or “socially recognized” domination and oppression (Bourdieu 1977, 16:192). For Bourdieu (1977), symbolic capital is simply a “disguised form” of economic capital (p. 183), a kind of credit that is “readily convertible back into economic capital” (p. 179). To phrase it another way, symbolic capital arises from assigning status functions – or to assigning functions on entities that cannot perform their functions without the imposition – to non-economic goods (e.g., J. Searle 2010; J. R. Searle 1995, 2013). The status functions assigned to symbolic capital exercise deontic powers, which refers to powers that exist because they are acknowledged, recognised, or accepted. 140 Procedures regulate entry into fields, and only some actors will be recruited or co-opted into playing. In differentiated societies, the fields of economic production have ostensibly been separated from the fields of cultural production. However, only certain social group members will be recruited into participating in – will be taught to play by the rules of – certain cultural scripts. Bourdieu (1984, 225) describes how, in differentiated societies, the means for appropriating esteemed cultural heritages is unequally distributed, and so culture can function as cultural capital, as misrecognised tool for oppression and domination. By contrast, in undifferentiated societies, the means for appropriating cultural scripts is “fairly equally distributed, so that culture is fairly equally mastered by all members of the group and cannot function as cultural capital, i.e. as an instrument of domination [and oppression]” (Bourdieu 1984, 225). For instance, dominant class members or empowered social group members will learn the upper class or dominant cultural scripts (haute culture, the dominant language, proper etiquette, etc.), whereas lower class or disempowered social group members will not. Fields’ rules are internalised in habitus as dispositions, as a kind of “practical recognition” of social space and one’s place in it (or not). It is the recognition that a working-class person is “out of place” in a fine dining restaurant where they are under-dressed and unsure of how to use the cutlery, since they do not possess the tacit knowledge embodied by upper class diners. It is the feeling that a woman or other historically politically excluded social group member is “out of place” in politics, since they look different from the rest of the politicians, and are unfamiliar with the norms regulating conduct. It is through this “practical recognition” that “the dominated, often unwittingly, contribute to their own domination by tacitly accepting, in advance, the limits imposed on them” 141 (Bourdieu 2000, 169). That is, disempowered social group members avoid participating in spaces of social play where they do not know the rules. When they do participate, they risk revealing a practical self who is ill-fit for the field, who may be disparaged by the field’s other participants. Since fields distribute and regulate different forms of capital, fields help preserve power relations. 5.4 Preserving Power Relations Agents partially overcome the transience of corporeal, sensorimotor intersubjectivity by inscribing shared intentions in languages and institutions, and – ultimately – in social spaces of public appearance. However, as Arendt (2013, 197) notes, these spaces of public appearance, for “the sharing of words and deeds,” are still characterised by transience and futility: words might be forgotten, and deeds may leave no impression. Agents can design a set of government institutions – to which I collectively refer to as government systems – to direct and protect the broader social systems’ spaces of shared intentionality. Arendt (2013, 198) describes how the ancients countered the futility of speech and action by creating the polis to give permanence to speech and action, and act as a kind of “organized remembrance.” The polis was one example of a government system, where a set of government institutions not only directs, preserves, and expands the shared intentionality inscribed in a social systems’ features, but can also symbolically consecrate asymmetric power relations by recording and legitimating social identities’ and fields’ distinctions into law. Anthony Giddens (1985) describes political systems as “containers” that store authoritative resources; political systems are tasked with determining distributions of power within social systems. Giddens (1984, 261) 142 describes the city (polis) as the original form of container, and – like Arendt’s description of the polis as an organised remembrance – describes the city-state as a “medium of ‘binding’ involving, on the level of action, the knowledgeable management of a projected future and recall of an elapsed past.” Giddens describes how the advent of modern capitalism broke-up the polis, and replaced with a new type of container: the state. Containers, or the sets of institutions comprising government systems help preserve social systems’ shared intentionality (including power relations) by recording – or symbolically consecrating – shared intentions and the structure of power relations between social groups and fields into law. Law records and coordinates social systems’ features, including language games, spaces of public appearance, and practical identities. Under conditions of inequality, empowered social group members may forcibly try and harmonise agents’ experiences, to impose their vision of social space and enforce the self-evidences of empowered social group members’ orthodoxy. Total institutions are extreme examples of efforts to harmonise agents’ experiences and impose a dominant orthodoxy, such as when Canadian-European settlers subjected Indigenous children to residential schooling, or when totalitarian governments send political dissidents to internment camps. But even following the underlying institutional rules of more benign institutions – including public schools, welfare assistance programs, or the language games of dominant social group members – can contribute to harmonising agent’s experiences and imposing dominant orthodoxy. For instance, laws record and legitimate “official languages.” Recording official languages into law has consequences for social systems’ language games, practical identities, spaces of public appearance, and power relations, since designating a language “official” assigns it deontic powers, and converts language-use into symbolic capital. With respect to practical 143 identities, agents who speak the official language – and speak it masterfully, as or like a native-speaker – will have greater status, and will be the most desirable school applicants, the most hirable, and the most electable. This can create variation in official-language speakers’ and other-language-speakers’ internalised perceptions of life chances, and thus their choices about whether and which schools or jobs to apply to, and whether and which level of politics to participate in. With respect to social spaces, official-language public spheres and official-language-speaker social identification may have greater status than subaltern language spheres and other-language-speaker identifications. Like economic capital, this symbolic capital constitutes an interdict that can dis-incentivise other-language-speakers from identifying as such, and discourage them from participating in a subaltern language public sphere (e.g. Derrida 1998). In addition to using official designations to create symbolic capital and the interdict’s subtler force, government systems can actively suppress or prevent shared intentions inscribed in social systems from being learned and replicated. For instance, consider Canada’s government-sponsored, mass institutionalisation of Indigenous children in residential schools, which represented an explicit and deliberate attempt to “kill the Indian in the child” (Harper 2008). Students were physically punished for “speaking Indian,” and were told that their cultural practices were sinful and evil (“Truth and Reconciliation,” n.d., 84–86). John Kitabish, an Algonquin survivor from a residential school in Amos, Quebec, testified to the Truth and Reconciliation Commission that when he left the school he could no longer speak Algonquin, and his parents could not speak French. John Kitabish explained that, “we were well anyway because I knew that they were my parents, when I left the residential school, but the communication wasn’t there” (“Truth and Reconciliation,” n.d., 85). The 144 government program not only co-opted Mr. Kitabish into speaking an official Canadian language (French), it prevented his co-option of Algonquin. Furthermore, by preventing the Kitabishs from communicating with their son, the government program impeded the Kitabishs’s ability to co-opt their son into learning and participating in other shared intentions inscribed in the Algonquin social systems’ cultural practices. Recall from Chapter 4 that, in ordinary life occasions, when interaction partners share a common backdrop of implicit and embodied knowledge, they do not have to tacitly or explicitly inquire: “what do you mean?” A shared stock of self-evidences makes “causes practices and works to be immediately intelligible and foreseeable, and hence taken for granted” (Bourdieu 2000, 80). The government-sponsored residential school system that separated Indigenous children from their parents and forced them into total institutions prevented people like John Kitabish from enjoying the same backdrop of implicit and embodied knowledge as their parents and elders. The residential school system also aimed to ensure that if residential school survivors explicitly asked their parents clarifying questions (“what do you mean?”), their parents would have trouble communicating a mutually comprehensible explanation. As this discussion suggests, social group members (e.g., English or French settlers) may be motivated to forcibly co-opt new agents (e.g., Indigenous peoples) into participating in the shared intentionality inscribed in their social systems, to expand and give permanence to their social space. The motivation to co-opt new members is particularly powerful when preserving and expanding social space preserves or expands asymmetrical power relations that benefit the co-opting power. When empowered groups forcibly co-opt participants into their social systems, they co-opt new agents on their norm-conditional terms. Co-opted recruits must participate in the co-opting group’s language games and institutions, and even participate in the co-opting groups’ 145 sanctioned spaces of public appearance, including sanctioned social identities and fields of social play. Through their participation, the co-opted group members help reconstitute, expand, and give permanence to co-opting group’s social system, including the asymmetrical power relations that empower co-opting group members (see also Fanon 2007). As Bourdieu (1984, 248) explains, dominated social group members and classes are often condemned to being defined as “low” or “vulgar,” and “anyone who wants to ‘succeed in life’ must pay for his [their] accession… by a change of nature, a ‘social promotion’ experienced as an ontological promotion, a process of ‘civilization’.” When a disempowered or colonised social group member internalises a dominating group’s shared intentionality, they often internalise shame for their “language, his [their] body and his [their] tastes, and everything he was [they were] bound to, his [their] roots, his [their] family, his [their] peers, sometimes even his [their] mother tongue, from which he is [they are] now separated from a frontier more absolute than any taboo” (Bourdieu 1984, 249). Government systems direct, preserve, and expand the shared intentionality inscribed in a social systems’ features, and can symbolically consecrate asymmetric power relations by recording distinctions into law. Recall from the previous section that because social systems’ features (such as fields) distribute and regulate different forms of capital, they help preserve power relations. Law records and coordinates social systems’ features, and so records and coordinates power relations. 146 5.5 The Democratic Aim of Creating and Preserving Symmetrical Power Relations Bourdieu (1977, 16:188) suggests that the only purpose of law is to “symbolically consecrate… the structure of the power relationship between groups and classes,” that law’s purpose is to legitimate oppression and domination. Bourdieu is correct in identifying this as a potential function. However, as I introduced in Chapter 1, and discussed at length Chapter 2, government institutions can be designed to consecrate either asymmetrical power relations (reproducing social systems characterised by oppression and domination), or symmetrical power relations (reproducing democratic social systems). Although I have introduced the problem that conditions of structural inequality generate asymmetrical power relations between social groups that entail exclusions and contribute to harms of oppression (a point I elucidate in Chapter 6), I must remind my readers that social systems vary in the degree to which they are democratic or oppressive and characterised by domination, and that more democracy is both better and possible. Equality means that empowerments are distributed in ways that enable those affected by collective endeavours to participate in, and influence all social practices. Equality means that symmetrical empowerments entail the inclusion of all social group members in participating in and influencing cultural reproduction, their social identifications, and fields of social play; that the capacity to participate in self-development does not vary by social groups. Equality means that symmetrical empowerments entail the inclusion of all social group members in participating in and influencing generic political practices that serve democratic functions, such as voting, representing, talking, and so on; so that the capacity to participate in self- and collective-rule does not vary by social groups. Participating in and influencing political practices is necessary 147 for agents to form individual preferences into collective wills, and – ultimately – for communities to make collective decisions. This, obviously, includes collective decisions about which social systems’ features should be recorded into law, and how law should coordinate social systems’ language games and other institutional and cultural artefacts, spaces of public appearance, and power relations. 5.6 Conclusion Social life entails living embedded in overlapping social spaces that give meaning to agents’ words and deeds. Social systems are differentiated into spaces of public appearance, including overlapping, variable social identifications. Human plurality – humans’ fundamental sameness and difference – means that humans are both capable of understanding one another, and need to make themselves understood. This creates the fundamental psychological needs for belonging and distinction, which are satisfied through collective social identifications. Social identities allow for people to be the same (as fellow group members) and different (from out-group members) at the same time (Brewer 1991, 2003, 2011; Brewer and Pickett 1999; Leonardelli, Pickett, and Brewer 2010). The cognitive capacity for categorisation and need for cognitive efficiency catalyses social differentiation (Gaertner et al. 1993; Hogg 2006; Hornsey 2008; Macrae and Bodenhausen 2000; Perdue et al. 1990; Turner 1987). Fields provide another space for validation and recognition, to be – and be recognised – as something (the best, the worst, anything at all). The rise of capitalism increased the number of fields and the relations between fields in social systems, and, by bracketing off a specific social space for economic exchange, opened spaces that could be governed by logics ostensibly 148 unrelated to naked economic exchange. However, all fields remain concerned with distributions of capital and power. Knowledge of the rules (e.g., properly internalising and playing by a fields’ rules), and mastering the required skills for achieving the field’s goals, are a form of symbolic capital that give possessors the power to play the game, and reveal themselves in that field’s space of appearances. This is problematic when economic capital is converted into symbolic capital to transform overt oppression and domination into misrecognised, or “socially recognized” oppression and domination (Bourdieu 1977, 16:192). Societal differentiation into social identities and fields has consequences for distributing and maintaining shared intentions inscribed in social systems, particularly, for distributing and maintaining power relations. Government apparatuses within social systems can be created as an “organized remembrance” (Arendt 2013, 198) or “containers” (Giddens 1984, 261), to coordinate social systems’ features, and to store shared intentionality and power relations. Using government apparatuses to store and direct shared intentionality and power relations within social systems is normatively and functionally good when government apparatuses create or preserve symmetrical empowerments that promote all social group members’ capacities to participate in self-development, and all social group members’ capacities to determine their actions and the conditions of their actions (participate in self- and collective-rule). But using government apparatuses to store and direct shared intentionality and power relations within social systems is normatively and functionally bad when government apparatuses create or preserve asymmetrical empowerments that promote only certain social group members’ capacity for self-development and self- and collective-rule, but block other social group members’ capacities for self-development and self- and collective-rule. Systems of governance that store and direct asymmetrical empowerments (for instance, polities that rely on 149 slave labour) are obviously problematic because disempowered group members in these societies experience harms of oppression and domination. But furthermore, they are problematic because empowered social group members are motivated to forcibly co-opt new agents into participating in the shared intentionality inscribed in their social systems, to expand and give permanence to their social space and to their relative power. The Greek colonial motto was: “Everywhere you go, you shall be a polis” (Arendt 2013, 198). The Greeks colonists embodied polis – the language games and institutions, social identities and fields, habitus, and power relations symbolically consecrated in law – and brought it everywhere they went. Expanding their shared intentionality meant going more places, co-opting more participants into the polis. According to Arendt, this expression was “not merely the watchword of Greek colonialization,” but the conviction that the symbolically consecrated space of speech and action “can find its proper location almost any time and anywhere” (Arendt 2013, 198). This benefitted the Greek colonists because they could expand their shared intentionality and protect and expand their position of dominance in asymmetric power relations, but it undermined the democratic aims of universal self-development and self- and collective-rule by subjecting colonised peoples to harms of oppression and domination. Conditions of inequality enable empowered social group members to forcibly try and harmonise agents’ experiences, to impose their vision of social space and enforce the self-evidences of empowered social group members’ orthodoxy, by forcing them to participate in norm-conditional practices, including norm-conditional communication. Norm-conditional practices refer to social and political practices that take place within a dominant orthodoxy, according to the terms of social play determined by the majority or empowered social group. Norm-conditional communication refers more specifically to norm-conditional practices of talk. 150 Norm-conditional practices (including norm-conditional communication) contribute to the harm of oppression called cultural imperialism. Cultural imperialism refers to when empowered or majority social groups’ cultural expressions become normalised or universalised and disempowered or minority social groups’ cultural expressions are marked as “Other,” undermining disempowered or minority group members’ capacity to participate in self-development. I return to this problem in Chapter 6, where I describe how, under conditions of inequality, talking can contribute to harms of oppression. In Chapter 6 I also explain how struggles to challenge a dominant vision of social space involve struggles to bring the “self-evidences” of orthodoxy to the level of explicit statement, “to break out of the silent self-evidences” of orthodoxy (Bourdieu 2000, 188). The problem, however, is that the symbolic capital required to bring orthodoxy to the level of explicit statement (and to expand the space of heterodox public opinion) are unevenly distributed. Because structural inequality can prevent social group members from bringing problems with social systems’ orthodoxy to the level of explicit statement, inequality blocks deliberation and so undermines the universal capacity for self- and collective-rule, contributing to domination. 151 Chapter 6: Inequality, Oppression, and Domination Structural social and material inequalities engender asymmetrical empowerments that act as disabling constraints, depriving disempowered social group members of the opportunities to participate in, and influence political practices that contribute to self-development and self- and collective-rule. Furthermore, these objective inequalities are inscribed in subjectivities, in habitus’ enduring, social cognitive and motivational structures. Because habitus’ tendencies are imitatively learned by each new generation, subjective, interactional inequalities can outlast the objective conditions that created them. Just as political and material inequalities produce disabling constraints that entail external exclusions from social and political practices, these subjectively internalised inequalities act as disabling constraints that entail internal exclusions. In Chapter 3, I described how inequality generates asymmetrical empowerments that prevent generic tools such as talk from functioning democratically. In this chapter, I expand my discussion by illustrating how, under conditions of inequality, talk can contribute to harms of oppression (by undermining self-development), and to harms of domination, by blocking deliberation from occurring at all (thus preventing talk from contributing to self- and collective-rule). In the first section of this chapter I review the relationship between inequality and exclusion (on the one hand) and oppression and domination (on the other). In the second section of this chapter, I describe in more detail what I mean when I say that inequality can become embedded in thought and perception, describing how features of the social and material world practically structure subjectivity to engender interactional inequalities, or asymmetries in social, economic, or political practices. Because of my focus on using talk to achieve democratic 152 collective opinion and agenda formation functions, I focus on a form of interactional inequality that I call discursive inequality, which refers to when subtle and often subconscious cognitive and motivational dispositions generate asymmetries in opportunities to participate in, or influence communication, undermining self-development and self- and collective-rule. In the third section, I use the case of patriarchy as a concrete example to illustrate my claims about how objective and interactional inequalities (and external and internal exclusions) contribute to oppression and domination. I describe how, prior to the women’s liberation movement, explicit means – laws limiting women’s political participation, and openly recognised normative prescriptions limiting women’s economic participation – were used to maintain political and material inequalities between men and women, engendering asymmetrical power relations that entailed women’s external exclusion from political and economic life. I then show how, even after external exclusions are removed, habitus’ motivating dispositions can continue to engender asymmetrical power relations, and women’s internal exclusion from political and economic life. I then distinguish between two forms of discursive inequality – discursive inequity, and norm-conditional communication – and elucidate how, under conditions of discursive inequality, talking contributes to different harms of oppression and domination. In the fourth section I focus on how talking can contribute to different harms of oppression. I first consider discursive inequity, and describe how discursive inequity generates asymmetries in communicative influence by structuring how people perceive themselves, others, and social interactions, contributing to the harm of oppression Iris Young (2009, 2011) refers to as powerlessness. I then turn my attention to norm-conditional communication. I describe how norm-conditional communication presents people with an unenviable choice: participate in norm-conditional 153 communication that is harmful to their self-identities and risk the harm of oppression Young (2009, 2011) refers to as cultural imperialism (and risk discursive inequity and powerlessness), or exclude themselves from talk, and risk the harm of marginalisation. Finally, in the fifth section, I explain how structural inequalities – including discursive inequalities (both discursive inequity and norm-conditional communication) – can contribute to domination (prevent self- and collective-rule) by blocking deliberation from occurring at all. 6.1 The Relationship Between Inequality and the Harms of Oppression and Domination Social systems include the shared intentionality inscribed in language games, spaces of public appearance, habitus, and other institutions – including the government institutions comprising political fields designed to preserve, organise, and direct the political community’s shared intentionality. Within social systems, agents – the embodied knowledge of accumulated cultural and personal life experiences – respond to social systems’ structures through their practices, including their political practices that serve democratic functions. Agents modify and maintain social systems’ structures through practices in everyday activity, available procedures, and through confrontation. Social systems and the government institutions designed to direct the shared intentionality inscribed in social systems’ language games, spaces of public appearance, habitus, and other institutions, range in the degree to which they are democratic. As I explained in Chapter 2, democracy is about empowering those who are affected by social systems’ structures to participate in practices that contribute to self-development and self- and collective-rule. We can say that social systems (and political fields within social systems) are democratic to the 154 extent that they achieve three interrelated functions. They must engender symmetrical empowerments that entail inclusions in practices that contribute to self- and collective-rule, so that those affected by social systems’ structures – including the government institutions designed to direct the shared intentionality inscribed in social systems – can form their individual preferences into collective agendas. And they must empower collectives (e.g., political communities) to execute collective agendas as collective decisions. Democracy is characterised by two interrelated, universal values: (1) the ability to develop and exercise one’s capacities and express one’s experiences; and (2) the ability to participate in determining one’s actions and the conditions of one’s actions. As I explained in Chapter 1, these universal values are connected to the notion of justice derived from Habermas’s communicative ethics which stipulates that – for norms or social conditions to be just – everyone affect by them must, in principle, have “an effective voice” in their creation, and be able to agree to them without coercion (Young 2011, 38). Democracy is linked to this conception of justice because democratic conditions – conditions that promote the universalist values of self-development and collective-rule – can be expected to produce opinions, decisions, norms, and social conditions that are more attentive to the needs and preferences of those affected by them. In short, democratic conditions can be expected to produce functionally better (e.g., more responsive), and normatively better (i.e., more just) opinions, decisions, norms, and social conditions. These two universalistic democratic principles map onto the problems Young (2011) identifies as oppression and domination. Recall that oppression refers to institutional constraints on self-development that prevent social group members from participating in language games and social play. By contrast, domination refers to institutional constraints on self- and collective-155 rule that prevent social group members from participating in determining their actions or the conditions of their actions. Oppression and domination are structural concepts, in that they refer to institutional constraints that prevent certain social group members from participating in self-determination and self- and collective-rule. 6.1.1 Five Faces of Oppression Oppression refers to institutionalised social processes that inhibit social group members’ self-development; their ability to use skills in socially recognised settings, and to participate in social identifications, fields of social play, and language games – notably the everyday communicative processes necessary for developing democracy-supporting social ties, personalities, and facilitative attitudes. Young (2009, 2011) distinguishes between five distinct “faces” of oppression: exploitation, marginalisation, powerlessness, cultural imperialism, and violence. Exploitation is when “some people exercise their capacities under the control, according to the purposes, and for the benefit of other people” (Young 2011, 49). For instance, when owners of production expropriate surplus value from worker’s labour for their own economic gain, owners of production are exploiting workers. Another example is men’s gender exploitation of women, such as when men expropriate women’s material labour (i.e., unpaid household and childcare work), and women’s “nurturing and sexual energies” (Young 2011, 50). As Young (2011) notes, racial oppression in Western democracies is increasingly taking the form of marginalisation, rather than exploitation. Marginalisation refers to when “a whole category of people is expelled from useful participation in social life” (Young 2011, 53). Marginalisation typically occurs when the labour system cannot or will not use social group 156 members (such as the elderly or single mothers in liberal democracies, black people in the Americas, or migrants in Europe), forcing them into welfare dependencies that can undermine participation in social and communicative practices, such as through the suspension of “basic rights to privacy, respect, and individual choice” (Young 2011, 54). While power in modern societies is widely dispersed, and most people do not (and in democratic societies, should not) exert significant personal power, powerlessness is when social group members “lack authority or power even in this mediated sense” (Young 2011, 56). The social division inherent in all differentiated, industrial societies “between those who plan, and those who execute” entails some degree of powerlessness (Young 2011, 58). However, powerlessness constitutes structural oppression when the probability of falling in one category (those who are more likely to give orders) or the other (those who rarely have the right to give them) varies by social group membership. Powerlessness, when social group members “are situated so that they must take orders and rarely have the right to give them,” occurs in other fields besides the economy, including in public sphere, domestic or private sphere, political, and cultural fields, to name a few. Cultural imperialism refers to the “universalization of a dominant group’s experience and culture, and its establishment as the norm” (Young 2011, 59). As my discussion of cultural capital in Chapter 5 began to illustrate, often, without realising it, dominant group members project their own experiences as representative of humanity, and “their cultural expressions become normal, or the universal, and thereby unremarkable,” while disempowered social group members become marked as “Other” (Young 2011, 59; see also Bourdieu 1984). Finally, violence refers to the systemic violence that members of disempowered groups members face. This includes random, unprovoked attacks against their persons or property, such 157 as black Americans’ fear of unprovoked violence from the police (or fear of unprovoked violence from white Americans “standing their ground”), or women’s and LGBTQ people’s fear of physical and sexual violence. Violence includes both physical attacks, as well as “less severe incidents of harassment, intimidation, or ridicule simply for the purpose of degrading, humiliating, or stigmatizing group members” (Young 2011, 61). 6.1.2 Domination Domination refers to institutional conditions that prevent social group members from participating in self- and collective-rule; that prevent social group members from determining their actions, or the conditions of their actions (Young 2011). In extreme cases of domination (for instance, in the case of totalitarian governments) people might have no say over their social practices: where they study or work, who they can marry, whether they can pray, and what they can say or who they can speak to. People will also have little or no choice over the conditions of their actions: not only are they conscribed in terms of what they can say and who they can speak to, there are no available procedures to change the conditions conscribing their speech (for instance, to legal recourse to challenge restrictions on free speech). Democratic systems may also be characterised by a degree of domination. For instance, even in welfare democracies where agents may enjoy universal liberties and participation rights, and where a social safety net guards against material deprivation, “the expansion of bureaucratic administration over increasing areas of work and life” can bring about “new experiences of domination” (Young 2011, 78). As I noted in Chapter 1, oppression does not always entail domination: institutional conditions may prevent agents from using skills in socially recognised settings, and prevent 158 agents from participating in language games with others, while leaving agents some opportunities to determine their actions and the conditions of their actions. But domination almost certainly entails oppression. When institutional conditions prevent agents from determining their actions or the conditions of their actions, their capacity for self-development is circumscribed. Social systems characterised by oppression and domination can be expected to produce functionally worse (e.g., less responsive), and normatively worse (i.e., less just) opinions, decisions, norms, and social conditions. Objective inequalities (e.g., structural gaps in political rights or material resources between social group members) manifest as asymmetrical empowerments that entail external exclusions and asymmetries of participation and influence, and contribute to different harms of oppression and domination. Furthermore, objective inequalities have a long “memory” (M. S. Williams 2000). Even after political institutions designed to openly and explicitly maintain political and material inequalities between social groups (such as systems of apartheid, or patriarchal laws and institutions) and external exclusions (such as formal prohibitions of non-whites’ or women’s political participation) are dismantled, objective inequalities continue to operate through habitus, to subconsciously and implicitly maintain interactional inequalities and internal exclusions that help maintain systems of oppression and domination. So, how do features of the objective (external) world – such as political rights and material inequalities – structure habitus to engender asymmetrical empowerments that entail internal exclusions, contributing to harms of oppression and domination? 159 6.2 When Habitus Engenders Interactional Inequality and Internal Exclusions As I explained in Chapter 4, the self is “habitus,” an embodied comprehension of the world, or a system of dispositions that practically comprehends physical and social space. The social order inscribes itself in bodies through confrontation – in both everyday pedagogic action (a parent might tell a child to “sit like a lady”), and rites of institution – which is often marked by affectivity, particularly embarrassment (Bourdieu 2000; see also Goffman 1967). An example are gender differences, which are taught and learned through clothing and appropriate dress, and even through appropriate walking, talking, sitting, and speaking. Habitus are “isomorphic” to the external (social and material) world in the sense that objective factors – including political rights and material inequalities – shape schemes of thought and expression, ensuring that agents’ habitus are roughly adjusted to their world’s structures and tendencies, guaranteeing agents are spontaneously attuned to their world’s structures and tendencies (Bourdieu 2000, 139). As such, habitus’ dispositions become “the accomplices of the processes that tend to make the probable a reality” (Bourdieu 1990, 64). For instance, when working-class youth internalise a sense of belonging in the world that precludes higher education, and tastes and preferences that disparage higher education, these dispositions and motivations in habitus engender asymmetrical empowerments and internal exclusions: working class youth are less likely to apply to university when they feel “out of place,” foolish, or embarrassed applying. And the dispositions and motivations in habitus that engender asymmetrical empowerments and internal exclusions help maintain systems of oppression analogously to the way objective inequalities and external exclusions help maintain systems of oppression. In this example, no formal prohibitions prevent working class youth from 160 applying to prestigious universities. But just as legal prohibitions can be used to externally exclude certain social group members from university applications, social group members’ internalised tastes and preferences can preclude their application to university (and preclude pursuing the “mental” labour of middle class jobs), and motivate them to follow in the footsteps of their parents and peers, motivating them to pursue the “manual” labour of working class jobs (see Young 2011, 56–57). Interactional inequalities that entail internal exclusions contribute to systems of oppression, and particularly the harms of powerlessness (not only because the working class often lack autonomy at work, but also because internal exclusions undermine autonomous decision-making, such as whether one would like to apply to university at all) and marginalisation (for instance, from the “respectability” of middle class life) (see Young 2011). Interactional inequalities that entail internal exclusions contribute to domination when they prevent agents from participating in self- and collective-rule. As I explained in Chapter 4, no two agents have identical life experiences, and so no two agents have identical habitus. However, the convergent experiences of the same social group (including social class) members unite group members in a relationship of homology. One of the primary consequences of this are that habitus come to define the boundaries of what agents with homologous habitus perceive to be thinkable or unthinkable (their sense of reality, including their perceptions of prospects and closed doors), and whether agents with homologous habitus perceive utterances to be acceptable or unacceptable (for instance, whether an utterance is rude or forthright). This can be problematic given that, as I have explained, habitus are isomorphic to social structure. Recall how working class agents’ homologous minds will be more likely to incline them “to make a virtue of necessity” by, as in my earlier example, dissuading them from pursuing higher education or higher-paying jobs (Bourdieu 1977, 16:77). And, to return to my 161 earlier point, because habitus are inscribed in bodies, and often experienced in affect, habitus can survive long after the social conditions that produced them disappear (Bourdieu 2000, 180). For instance, racism and sexism persist in part because they are bound to minds and bodies (in dispositions and corporeal responses to social and physical stimuli), and so can perpetuate ongoing systems of oppression through unthinking habits. The practical knowledge granted by habitus often takes the form of emotion, of feeling one’s place in the world, and recall the system of dispositions defining one’s feeling of belonging in the world is isomorphic to material structures. In my earlier example, working class youth internalise a feeling of belonging that is isomorphic to their life chances (they are less likely to be admitted to a prestigious university), and they feel embarrassed applying. Another example is how women, socialised to embody femininity defined positively in terms of caregiving and emotional labour, and negatively in terms of not being authoritative or confrontational, might feel more at ease and seek out jobs related to caregiving (such as nursing or teaching), and might feel “out of place” and avoid jobs that involve argumentation or bargaining (such as law or finance). The practical knowledge granted by habitus are motivating structures that help spontaneously reproduce sex segregation in the labour market, including the overrepresentation of women in low-paid caregiving roles and their underrepresentation in high-paid positions of authority. When habitus entail dispositions that engender asymmetrical empowerments and internal inclusions and exclusions (such as feeling in one’s place doing caregiving work, or out of place doing authoritative work), they help to maintain systems of oppression even in the absence of objective – such as legal – prohibitions and external exclusions. The examples I have given thus far illustrate how habitus produces interactional inequalities that entail disempowered social group members’ internal exclusion from 162 participating in, or influencing social and economic practices: when a working-class youth’s embarrassment dissuades them from applying to university, or when a woman’s sense of “feeling out of place” dissuades her from pursuing a position that commands authority. But more importantly for the focus of my dissertation, habitus produce interactional inequalities that entail disempowered social group members’ internal exclusions from participating in, or influencing political practices, including talking, joining, voting, representing, exiting, recognising, and resisting. Even more specifically, habitus produce a specific genus of interactional inequalities I am calling discursive inequalities. Discursive inequalities entail disempowered social group members’ internal exclusion from participating in and influencing talk. In the next section I use the example of patriarchy to illustrate how (on the one hand) language games, spaces of public appearance, and habitus, and (other the other hand) agents’ reactions to social systems’ structures, can contribute to harms of oppression. The shared intentionality inscribed in social systems can openly and deliberately generate asymmetrical empowerments that entail certain social group members’ external exclusion from political practices. On the other hand, the shared intentionality inscribed in social systems can implicitly and unintentionally generate asymmetrical empowerments that entail certain social group members’ internal exclusion from political practices. More specifically, habitus’ cognitive and motivational dispositions can generate discursive inequalities that entail disempowered social group members’ internal exclusion from participating in and influencing talk. 163 6.3 The Example of Patriarchy Patriarchal oppression can function explicitly through language games, institutions, and spaces of public appearance, and implicitly through habitus’ cognitive and motivational structures. Consider Western societies prior to the successes of the women’s liberation movements, when patriarchal power relations were openly and explicitly inscribed in social identities and all fields, including the overlapping fields of the public sphere, law, politics, and the economic production, as well as the segregated field of the private sphere. At that time, patriarchal power relations were overtly and explicitly recognised in norms, and consecrated in laws designed to disempower women relative to men. Under conditions of inequality and exclusion, political practices such as talking or voting can reinforce oppressive political systems. Consider how, until the 20th century, legal prohibitions precluded women from voting or running for office, and clear normative expectations regarding women’s place as caregivers – typically (at least among white women), in the private sphere as homemakers – contributed to their external exclusion from political and economic participation and influence. These legal and normative prohibitions on women’s participation in politics and in the paid (or for women of colour, well-paid) labour market were designed to maintain asymmetrical empowerments (men’s domination of women), that entailed external exclusions (the exclusion of women from political practices and paid, or at least well-paid labour). Men and women’s participation/non-participation and influence/non-influence in politics and the economy contributed to a system of racialized patriarchal oppression, binding women in relations of dependence and powerlessness, marginalising them from public life, and 164 enabling the exploitation of their unpaid sexual and emotional labour (Pateman 1988; Young 2011). These racialized patriarchal power relations also overtly impacted language games. Consider how many jobs were denoted with the suffix “-man,” such as fireman, policeman, or congressman. While this may not seem important, power relations and language games have consequences for the habitus’ embodied schemes of thought and expression. When a woman who was subjected to these power relations and language games applied for a job with the local police force, or ran for a seat in congress, she was seeking a “-man’s” job. Women who took -men’s jobs, or who even advocated for the right to take -men’s jobs (such as suffragettes) were likely treated as if they were less feminine, and even perhaps as if were a threat to gendered identities. This contributed to a system of patriarchal oppression where dominant perspectives of masculinity (associated with authority and the public sphere) and femininity (associated with subordination and, often, the private sphere) were universalised as the norm, and subaltern identities become marked as “Other.” This face of oppression, cultural imperialism, has consequences for social group members’ sense of self, and can produce misrecognised or damaged identities. The practical, embodied comprehension of the world afforded by habitus also likely engendered interactional inequalities and internal exclusions. Women who strongly identified as feminine may have felt greater aversion seeking -men’s job at all. And men with -men’s jobs likely felt that their gender-identity was threatened when women tried to join their ranks. These identity threats, and the way people respond to these threats – such as when men avoid or are prohibited from hiring women, and when women do not apply or do not even advocate for the right to apply for certain jobs – contribute to disempowered social group members’ oppression 165 when they block self-development. When they block self- and collective-rule, they contribute to disempowered group members’ domination. In contemporary, post-civil rights era Western democracies, formal prohibitions to women’s participation in the paid labour market and politics have been removed: employers are not allowed to openly discriminate against women when hiring, and men and women enjoy the same political rights (to vote, associate, run for office, and so on). Even formal language games tend to reflect equalising efforts: for instance, formerly designated -men’s jobs are now designated by gender neutral titles, as in the case of firefighters, police officers, and members of congress. But recall my discussion of humans’ culturally-mediated learning techniques from Chapter 4. As I explained, when children imitate their elders’ actions they are not simply trying to achieve certain ends; children are also trying to discover the underlying conceptual space that makes actions meaningful by participating in action the “right” way. A child imitates their parents because they perceive their parents’ actions – the way their parents walk, speak, sit, or engage in any other intentional activity – as “constitutive of the social being of the accomplished adult” (Bourdieu 2000, 154). People imitatively learn to signify who or what they are through their actions, including how to signify that they are a “woman” or “man” by working, speaking, walking and participating in other practices in a certain way. Girls may imitate their parents – and find themselves attracted to their mothers’ (and other women role models) more “feminine” work, habits, and practices, and avoid their fathers’ (and other male role models) more “masculine” work, habits, and practices – and over time, develop an embodied knowledge of (and comfort doing) traditionally women’s work, habits, and practices. 166 As I noted, in Western democracies today, employers are not allowed to openly discriminate against women when hiring, and men and women enjoy the same formal political rights. But the historical memory of patriarchal power relations – the formally legal and explicit distinction between women’s roles as unpaid or menially paid caretakers, particularly in the private sphere, and men’s (particularly white men’s) roles as providers and decision-makers, particularly in the public sphere – remain as implicit associations in habitus’ schemes of thought and expression. As I discussed in Chapter 5, the need for cognitive economy contributes to categorical thinking, and inter-nodal links that vary with perceivers’ experiences in the social and material world structure mental representations. Implicit, semantic associations between gender and concepts relating women to nurturing and emotional care, and white men to authority and power, remain as stereotypes in habitus’ cognitive and motivational structures. That is, implicit semantic associations, or “stereotypes,” relating women to nurturing and emotional care, and white men to authority and power, remain. Today, women are formally invited to participate in politics and paid labour, but they are still punished when they violate implicit, stereotypic expectations (e.g., of the nurturing-caregiving woman) through norm-violating behaviours (e.g., acting authoritatively) (Lakoff 1990). Women who engage in norm-violating behaviours, such as by violating the expectation that women are warm and caring, or by behaving in an agentic or authoritative fashion experience a “backlash” effect, eliciting negative emotional responses (Cuddy, Fiske, and Glick 2007, 2008; Fiske, Cuddy, and Glick 2007; Fiske et al. 2002, 1999, Rudman and Glick 1999, 2001). For instance, women university instructors who behave authoritatively, or who are perceived as not being nurturing or caring, elicit negative affect and receive lower teaching evaluations (Andersen and Miller 1997; Baldwin and Blattner 2003; Bennett 1982; Statham, 167 Richardson, and Cook 1991). Relatedly, research shows that when a women instructor gives students negative feedback, students rate women instructors as less competent than women instructors who give positive feedback; but this pattern does not hold for men instructors (men instructors can give negative feedback and will not be viewed as less competent than man instructors who give positive feedback) (Sinclair and Kunda 2000). This is because students can use available stereotypes (i.e., that women are incompetent) to justify women instructors’ negative feedback, but cannot use the same stereotypes to justify men instructors’ negative feedback (since men are stereotyped as competent). Reliance on stereotypes helps students cope with women instructors’ norm-violating (i.e., critical and therefore non-nurturing) behaviour. Another problematic discursive inequity is structural (group-based) differences in global self-worth, or self-esteem. Self-esteem refers to the general “attitudes individuals hold about themselves, embracing what they believe to be their desirable (and undesirable) qualities and whether they like (or dislike) themselves” (Sniderman 1975, 44). Historically disempowered groups are more likely to internalise debased or devalued self-identities (Allport 1954; Goffman 2009; Honneth 1995). Discourses about the laziness and intrinsic inferiority of disempowered social group members “can inflict harm, can be a form of oppression, imprisoning someone in a false, distorted, and reduced mode of being” (Taylor 1997, 25). As I started to explain, habitus’ practical, embodied comprehension of the world impacts social action and interaction. For instance, women who identify strongly as feminine may avoid “authoritative” jobs (associated with masculinity) to avoid damaging their gender-identities. And women who do take these jobs may experience misrecognition or non-recognition. As my examples demonstrate, women who take these jobs may also experience other harms, such as 168 negative affective responses and, consequently, negative reviews and feedback on their work, potentially contributing to higher female attrition rates from authority positions. Habitus’ dispositions can also impact men’s behaviour. Habitus’ implicit dispositions may include an aversion to hiring or promoting women to such positions, further subjecting women to the harms of powerlessness and, potentially, marginalisation. And the subconscious association between men’s own masculinity and power or dominance may motivate men to ask women colleagues to perform menial tasks related to cleaning or tidying, demand emotional labour, or engage in sexual harassment. As my discussion shows, habitus can engender interactional inequalities, or asymmetrical participation and influence in social life. More specifically, habitus can engender asymmetrical participation and influence in generic political practices (contributing to domination). For instance, the historical legacy of women’s external exclusion from politics continues as a “gender gap” in politics today, where women seem to know less about politics than men, are less interested in running for office, and are underrepresented in legislatures (Fox and Lawless 2014; Gidengil et al. 2003; Stolle and Gidengil 2010; Wängnerud 2009). I return to this point in Chapter 7, where I analyse gendered variation in communicative political participation. The cognitive schemes of perception and appreciation that give rise to these asymmetries in political participation are “largely the product of a non-conscious, unwilled avoidance,” that often originate “from adults themselves formed in the same conditions” (Bourdieu 1990, 61). As I explained, girls learn imitatively from their parents, not just copying their behaviour to achieve an identical outcome, but to learn to signify who they are – the accomplished adult, or more specifically, the accomplished woman – by doing. Over time, they develop an embodied knowledge of (and comfort doing) traditionally women’s work, habits, and practices. 169 One of the most important predictors of whether girls become interested in politics is whether they have politically active parents, particularly politically active mothers (Fox and Lawless 2014; Gidengil, O’Neill, and Young 2010). Given that women are on average less interested and knowledgeable about politics, their daughters tend to be too – and girls’ on average greater disinterest in politics translates into on average less politically engaged women (Fox and Lawless 2014). The long memory of women’s external exclusion from political practices of voting, representing, and so on lives on in habitus’ schemes of perception and appreciation which engender internal exclusions and a tendency for girls’ and women’s disproportionate non-conscious, unwilled avoidance of political practices. Although habitus can engender interactional inequalities with consequences for participation and influence in all practices – whether social, economic, or political – I am particularly interested in when habitus engenders a genus of interactional inequality I refer to as discursive inequality. Discursive inequality refers specifically to unjustifiable asymmetries in communicative participation and influence. Within the class of effects, I further distinguish between two species of discursive inequality: discursive inequities and norm-conditional communication. Discursive inequities refer to when principles of vision and division embedded in habitus – such as social stereotypes – generate unjustifiable asymmetries of communicative influence between social group members. Norm-conditional communication is when the terms of successful communication (norms regulating what makes an utterance successful, a claim persuasive, and so on) are harmful to certain social group members’ identities, which discourages those certain social group members from participating in communication to produce unjustifiable asymmetries of communicative participation. Norm-conditional communication presents speakers with an unenviable choice: 170 abstain from identity-threatening talk (and face the harm of marginalisation), or participate in norm-conditional communication and contribute to cultural imperialism (and face discursive inequities, and the associated harm of powerlessness). As I explained in Chapter 3, discursive inequalities prevent talk from functioning democratically. In the next two sections, I elucidate how discursive inequities and norm-conditional communication contribute to harms of oppression and domination. 6.4 Discursive Inequality Contributes to Harms of Oppression As I explained, discursive inequalities can prevent social group members from using skills in socially recognised ways, and circumscribe social group members’ capacity to participate in social identities, fields of social play, and language games. In this section, I describe how different forms of discursive inequality contribute to different harms of oppression. I explain that discursive inequity tends to cause powerlessness, while norm-conditional communication prevents agents with the choice of participating in norm-conditional communication and face the harm of cultural imperialism (and often discursive inequities and powerlessness), or abstain from talk and face the harm of marginalisation. 6.4.1 Discursive Inequity and the Harm of Powerlessness The first form of discursive inequality I discuss are discursive inequities. As I mentioned in Chapter 2, the term ‘equity’ comes from the Latin aequitas, the Latin concept for justice, fairness, and symmetry (Skinner 2002). Discursive equity refers to the symmetrical, fair, and just 171 (or justifiable16) influence of speakers in communicative processes. Discursive inequities refer to when agents can (and do) formally participate communicative processes, but social systems’ language games, fields and social identities, and habitus generate unjust or unjustifiable asymmetries in social group members’ ability to influence communication, contributing internal exclusion. The motivations and predispositions of habitus can engender asymmetries in speaker status when semantic associations form between cognitive categories denoting social group members and human attributes. These semantic associations, or stereotypes, are not formed in a vacuum. Rather, as I have been suggesting, past inequalities and external exclusions become “memorised” in habitus in the form of stereotypes. When members of social groups (for instance, White men) fill roles characterised by leadership and authority over long periods of time, an implicit association between the cognitive categories white man and leader (or, white man and authoritative) develops. Like all other cognitive categories, stereotypes can prevent cognitive overload by simplifying large quantities of information, organising our perceptions of the world (Hinton 2000). But evidence from social psychology shows that stereotypes also set implicit normative conditions for communicative success, by disproportionately enhancing the communicative influence of positively stereotyped groups (typically, empowered social group members), and 16 The requirement of just or justifiable influence keeps open the possibility that there are instances where it is just, or justifiable by recourse to good reasons, for certain speakers to have greater influence (that is, where discursive equity requires asymmetrical influence). For instance, at an inquiry on missing and murdered Indigenous women, it may be deemed justifiable that Indigenous women be allotted more speaking time, or that their testimony be given the greatest weight. This is because they are most affected by the outcomes of the inquiry, and may have greater practical knowledge of the subject. However, asymmetric influence must always be defended through explicit recourse to publicly defendable reasons. 172 undermining the communicative influence of negatively stereotyped groups (typically, disempowered social group members). Asymmetries of communicative influence work by structuring how social group members perceive their own influence, by structuring how they are perceived by others, and ultimately by structuring how agents perceive social interactions. Discursive inequities shape agents’ practical identities, and shape how agents perceive their own influence. It is well known that negative stereotypes and stigmatisation have harmful psychological and emotional consequences for those who are subject to them (Allport 1954; Goffman 2009; Honneth 1995).17 Taylor (1997, 25) notes that non-recognition or misrecognition, such as discourses about the laziness and intrinsic inferiority of less powerful group members, “can inflict harm, can be a form of oppression, imprisoning someone in a false, distorted, and reduced mode of being.” The internalisation of negative stereotypes can have behavioural consequences, such as “stereotype threat,” which refers to when stereotype awareness subconsciously induces the agent to behave in a stereotypical manner (R. P. Brown and Pinel 2003). Consider the well-known finding that women perform worse on math tests after being primed with the stereotype that women are bad at math (S. J. Spencer, Steele, and Quinn 1999), a finding that has been replicated with low socioeconomic status students (B. Spencer and Castano 2007), and ethnic or racial minorities (Aronson, Fried, and Good 2002; Gonzales, Blanton, and Williams 2002). Stereotypes also generate unjustifiable asymmetries in discursive influence by structuring affective responses to disparaged social group members. Stereotypes shape affect in at least two 17 Allport (1954, 42) begins his classic analysis of prejudice by asking, ‘What would happen to your personality if you heard it said over and over that you are lazy and had inferior blood?’ The answer, of course, is that you might begin to internalize a debased sense of self-worth. 173 ways. Stereotypes directly shape affect, by eliciting automatic feelings, such as pity, contempt, and aversion (Cuddy, Fiske, and Glick 2008; Fiske, Cuddy, and Glick 2007). And as I stared to explain, stereotypes can also shape affect indirectly, by generating expectations of category members, which – if the expectations are violated – can induce negative feelings directed toward the transgressing category members. Stereotypes generate “normal” expectations, such as the expectation that a woman is nurturing (Lakoff 1990). When a person violates the expectations generated by stereotypes, they often arouse negative feelings. Continuing with my earlier example, when a female professor acts in an authoritative, rather than in a nurturing manner, she often elicits negative responses (Andersen and Miller 1997; Baldwin and Blattner 2003; Bennett 1982). As I already explained, stereotypic expectations can have consequences for people’s identities: women who identify strongly as feminine may avoid authoritative roles, and women in authoritative roles may experience non- or misrecognition. The result may be fewer women selecting into authoritative roles, or higher attrition rates among women in positions of power. This contributes to the disproportionate number of men in positions of authority, replicating patriarchal power relations and the implicit associations between concepts of “man” and “authority.” Discursive inequities also colour interpretations of social dynamics. For instance, in one experiment research participants were either primed with a series of words that cued black American stereotypes or not, and then were asked to read a vignette describing a neutral social interaction (Devine 1989). Research participants who were primed with stereotype-related words perceived greater hostility in the vignette, as compared to the control group, even though the vignettes displayed the same interaction. Besides colouring perceptions of other peoples’ 174 interactions, stereotypes also directly trigger corresponding non-conscious behaviours. For instance, in another study white research participants who were primed with black American images interacted with other (also white) participants in a more hostile manner (as rated by a double blind, outside judge) as compared to white research participants who were primed with white American images (Chen and Bargh 1997). Simply being primed with the stereotype of a negatively stereotyped outgroup can automatically induce hostile behaviour (even toward those who are not the subject of the stereotype). Reminding research participants of implicit associations between a racial minority that has been stereotypically associated with features of social marginalisation (social disintegration and violence) colours how people perceive and interpret neutral social interactions, and how they treat others. Agents have less discursive influence if they have lower self-worth and self-confidence, if they arouse negative affective responses when they speak authoritatively, and if their neutral social interactions are perceived as being aggressive. This lack of discursive influence contributes to oppression as powerlessness. Those who have internalised debased self-identities or lower self-worth contribute to their powerlessness when they come to believe in their inferiority. Others are disempowered by the undeservedly negative, automatic affective and behavioural responses to their utterances and interactions. This undermines agents’ capacities to use skills in socially recognised settings, and to participate in language games and fields of social play, and to enact their social identities. Compounding this problem is the fact that speakers with less discursive influence will have a harder time challenging the discursive inequities that oppress them (because of their reduced discursive influence). In addition to contributing to harms of oppression, this blocks reflexive communication – preventing misunderstanding or disagreement from being solved 175 through deliberation. When discursive inequities prevent social group members from raising discursive challenges, or from influencing deliberation they are undermining social group members’ capacities for self- and collective-rule. I will return to this point after my discussion of norm-conditional communication. 6.4.2 Norm-Conditional Communication and the Choice Between Cultural Imperialism or Marginalisation The second species of discursive inequality I consider is norm-conditional communication. Norm-conditional communication refers to when participating in talk is conditional on a normative framework that is harmful to a social group member’s identity, and causes the social group member to take issue with the very procedures required for reaching mutual understanding, recognition, and social coordination. Norm-conditional communication presents agents with the unenviable choice. On the one hand, there is the Scylla of participation in the identity-threatening procedures of communication. This exposes disempowered group members to the harm of cultural imperialism, and often to discursive inequities and the associated harm of powerlessness. On the other hand, there is the Charybdis of non-participation or abstaining from talk, which risks the harm of marginalisation. Social systems’ language games, and other social and political institutions comprising fields, available social identities, and the homology of the majority or dominant habitus that defines the common-sense world comprise the context and set the terms for communicative success. Social systems’ structures are more likely to damage or threaten a social group member’s practical identity, and to disempower a social group member in asymmetrical power 176 relations, when members of that group (and their ancestors) were excluded from constituting or influencing the social systems’ structures scaffolding mutual comprehension. When social group members (or their ancestors) played little role in constructing social systems, then it is more likely that language games, spaces of public appearances, and the common-sense world will define practical subjectivities that are foreign or harmful to them. Consider the European conquest and colonisation of Indigenous peoples in Canada. Indigenous peoples were excluded from participating in processes of identity- and agenda formation unless they formally accepted and participated in the backdrop of mutual intelligibility of Anglo-Protestant (or, to a lesser extant, Franco-Catholic) Canadians. Indigenous peoples were only given official Canadian citizenship, and invited to participate in political practices if they legally forfeited their Indigenous title, thus rejecting, at least in a legal-bureaucratic sense, their cultural identities. This historical example is a clear instance of the choice between norm-conditional a political practice18 that harms one’s self-identity (in this case, by renouncing one’s identity), or abstention (and exclusion from the political community). Like with norm-conditional communication, choosing the Scylla of norm-conditional participation entails the harm of oppression Young (2009) calls cultural imperialism. In my previous example, Indigenous persons would have to accept the culture of Canada’s settlers as the norm (and legally reject their Indigenous title, which, by its very existence, threatens the norms of settler society). This includes not only the threat of damaged identities (through non- or misrecognition), but also often involves facing discursive inequities and the associated harm of 18 Recall from Chapter 5 that norm-conditional practices refer to social and political practices that take place within a dominant orthodoxy, according to the terms of social play determined by the majority or empowered social group. Norm-conditional communication refers more specifically to norm-conditional practices of talk. Norm-conditional practices (including norm-conditional communication) contribute to cultural imperialism. 177 powerlessness. Cultural imperialism often involves marking members of empowered groups (such as white Christians) with positive stereotypes, and members of disempowered groups (Indigenous peoples, and other ethnic or religious minorities) with negative stereotypes. As I explained in the previous section, given the ways in which stereotypes structure cognition, affect, and interpersonal interactions, norm-conditional participation (and the experience of cultural imperialism) often also brings the risk of discursive inequities and the associated harm of powerlessness. But choosing the Charybdis of abstention – and rejecting, or refusing to participate in norm-conditional participation – also has consequences. Most obviously, the choice of abstention carries the risk of marginalisation. Marginalisation results in external exclusion, where participants are formally shut out from participating in communicative processes of opinion and agenda formation. Marginalisation can expel agents, and entire groups of agents, from participation in large swaths of social or political life. Most Indigenous people refused to renounce their title, and until the 1960s were legally excluded from electoral politics. But as I have explained, historical marginalisation leaves an enduring memory in habitus. Even after receiving unconditional citizenship rights, the habit of political abstention continues, with Indigenous citizens voting at far lower rates than non-Indigenous citizens (Fournier and Loewen 2011). Norm-conditional communication is not restricted to instances where social group members find themselves suddenly thrust into (or ensconced by) dominant systems of government that are foreign to them. Norm-conditional communication also occurs when social group members (or their forbearers) were prevented from participating or influencing social systems language games and other institutions, spaces of public appearance, and common-sense 178 world. Consider again the example of patriarchal societies, where women have historically been denied full participation and influence in self- and collective-rule. In patriarchal societies, the shared, implicit background assumptions are shot through by what Freud (1925) describes as a “Madonna-whore complex,” which dichotomises women into the category of saintly mothers (the Madonna) or objects for male sexual gratification (the whore). In such societies, women may find that complaints about sexual violence only have persuasive force when the women complainants fit the Madonna characterisation. That is, the women were behaving in a pure, motherly manner when the violence took place. Claims made by women behaving in a non-saintly manner, thus fitting the whore characterization – a scantily-clad woman walking alone at night, a college student drinking at a party, or a sex worker – may find that the persuasive force of their claim is undermined. Background assumptions may induce people, including friends, family, police officers, and judges and juries, to conclude that the woman who drank at a party, wore revealing cloths, or sold sex, was “asking for” (and thus consenting to) sex. Here it is not explicit, acknowledged norms guiding listeners’ practices of downplaying or ignoring the agents’ complaints about sexual violence; rather, it is the implicit normative conditions of communicative success embedded in social systems’ language games, fields, social identities, and the homology of the dominant or majority habitus (the common-sense world). This example of norm-conditional communication also presents agents with the ugly choice of participation and cultural imperialism (and the risk of discursive inequity and powerlessness) or abstention and marginalisation. Cultural imperialism, combined with the fear of being powerless, may induce women to participate in certain behaviours or discourses conscribed by the common-sense world. To guarantee the security of knowing they can raise 179 complaints about sexual assault (and to avoid facing the powerlessness of a raped “whore”), women may feel obliged to behave and dress in ways befitting the Madonna, which may have consequences for their identities (for instance, by supressing their sexual identities). When women who behave or dress in ways unbefitting the Madonna are sexually assaulted and raise a complaint, they often face discursive inequity and powerlessness. Being aware of, and fearing of the powerlessness of raped “whores” who raise complaints can deter women from reporting, or speaking about the violence they endured when behaving unlike the Madonna – exposing them to marginalisation. The threats of cultural imperialism, powerlessness, and marginalisation, as well as the ways women respond to these threats, help affirm the very backdrop of implicit, normative assumptions that undermine women’s ability to participate in raising validity claims, as well their discursive influence when they do participate in raising validity claims. Discursive inequalities – both discursive inequities and norm-conditional communication – generate inequalities in the ability to raise, challenge, and defend validity claims. As I explained in Chapter 3, discursive inequity and norm-conditional communication can inhibit discourse from achieving its democratic functions. As my discussion in this chapter shows, discursive inequalities can also contribute to harms of oppression. Finally, as these examples are starting to make clear, discursive inequalities can also block reflexive communication – that is, prevent misunderstandings and disagreements from being solved through rational communication, or deliberation. 180 6.5 Discursive Inequalities Contribute to Domination when they Block Deliberation Normally, when speakers fail to sufficiently connect the component parts of a speech act, communication fails – it falls apart, or results in a misunderstanding. Because misunderstandings and disagreements typically involve a breakdown of communication, conversation partners can identify them and make the misunderstanding or disagreement the theme of communication (they can explicitly talk about it) (Bohman 2000a, 2000b). Ideally, when conversation partners talk through misunderstanding or disagreement, they deliberate – they engage in epoché and temporarily suspend their naïve adherence to the world and explain, justify, and challenge one another’s presuppositions – until they can establish (or re-establish) some kind of mutual understanding or agreement, even if it is an agreement to disagree. This ability to raise and defend validity claims is an essential aspect of self- and collective-rule and – as I discussed in Chapter 1 and Chapter 2 – links democracy to justice (defined as norms, decisions, and social structures that could, in principle, be agreed to by those affected by them). Struggles to challenge a dominant vision of social space involve struggles to bring the “self-evidences” of orthodoxy to the level of explicit statement “to break out of the silent self-evidences” of orthodoxy (Bourdieu 2000, 188). In other words, the implicit, shared understandings and intentions inscribed in social systems’ language games, social identities, fields, and habitus – and, by definition, in the common-sense world created by the homology of habitus – can be problematized (brought to explicit statement, and rationally challenged and defended) through deliberation. However, as I have explained, inequality generates asymmetrical power relations that entail the political exclusion of the disempowered. When the disempowered are excluded from 181 participating in or influencing communicative processes, they cannot effectively highlight misunderstandings or raise points of disagreement. Power asymmetries distort communication and violate implicit norms of every day practice when they block reflexive communication, “so that communicative failure cannot become the theme of public communication” (Bohman 2000, p. 385). The symbolic capital required to bring orthodoxy to the level of explicit statement (and to expand the space of heterodox public opinion) are unevenly distributed. Asymmetrical empowerments and exclusions mean that the disempowered are less likely to be able to push back the self-evidences of orthodoxy imposed by dominant social groups. Inequalities prevent social group members from bringing problems with social systems’ orthodoxy to the level of explicit statement. Discursive inequalities prevent communicative failures from becoming the theme of public communication when, for instance, the descendants of historically disempowered groups, (such as women) internalise the feeling of inferiority prescribed by social systems’ language games, fields, available social identities, and habitus, and come to feel out of place raising claims authoritatively or participating in political practices. Discursive inequalities block reflexive communication when Indigenous agents refuse to raise validity claims regarding colonial relations, because doing so requires participating in norm-conditional practices, including norm-conditional communication, that threaten their collective identities. Discursive inequalities also block reflexive communication when women fear their conduct prior to sexual violence will discredit and shame them, and so refrain from reporting the violence. Discursive inequalities not only block deliberation because it dissuades the disempowered from speaking up, they also block deliberation by undermining the discursive influence of their utterances, preventing their claims from receiving uptake. For instance, when 182 negatively stereotyped groups reject social systems’ prescribed roles and engage in norm-violating behaviour, such as speaking authoritatively in political or public sphere fields, they may still be viewed as incompetent or “out of place” by listeners, and their words and deeds may have less impact. When Indigenous agents engage in political acts outside a social systems’ normal procedures, by engaging in illegal or unsanctioned protests, demonstrations, or sit-ins, or through artistic or other expressive acts, their words and actions may be unintelligible to non-participants, and so may not have the influence they otherwise should. And as I discussed, when women who behave or dress in ways unbefitting the Madonna are sexually assaulted and raise a complaint, their utterances often have less influence. Inequalities generate power asymmetries that entail exclusions and blocks misunderstandings and disagreements from being brought up at all, and even when they are brought up, it can prevent them from being solved through deliberation. 6.6 Conclusion As I explained in Chapter 3, inequality prevents talk from achieving democratic social integration, reciprocity, identity, learning, and legitimation functions. Furthermore, as my discussion in this chapter shows, under conditions of inequality talk can contribute to harms of oppression, and – because they can block deliberation from occurring at all – contribute to domination. Social systems are oppressive to the extent that institutional constraints on self-development prevent social group members from participating in language games, participating in fields of social play, or from enacting their social identities. Social systems are characterised by domination to the extent that institutional constraints on self- and collective-rule prevent 183 social group members from participating in determining their actions or the conditions of their actions. Using the example of patriarchal social systems, I described how, prior to the women’s liberation movement, explicit means – laws limiting women’s political participation, and openly recognised normative prescriptions limiting women’s economic participation – maintained political and material inequalities between men and women, engendering asymmetrical power relations that entailed women’s external exclusion from political and economic life. And I argued that even after external exclusions were removed, habitus’ motivating dispositions continue to engender interactional inequalities (including the genus of interactional inequality I call discursive inequalities) that entail women’s internal exclusion from social, economic, and political practices (including women’s internal exclusion from talk). In Chapter 7, I use data from the 2015 Canadian Election Study (CES) to show that, although there are no legal barriers preventing Canadian women from participating in talk, there is an ongoing gender gap in political talk in Canada. I this chapter I also described how social and material inequalities become embedded in habitus to engender discursive inequalities, or asymmetries of communicative participation and influence. Discursive inequities, including semantic associations between social group members and human attributes, or systematic variation in self-esteem, contribute to unjustifiable asymmetries in discursive influence and the harm of powerlessness. In Chapter 8, I use evidence from a survey experiment I designed, to show how gendered variation in self-esteem undermines women’s discursive influence (relative to men’s) even when women do participate in talk. Today, democratic systems are characterised by formal equality between citizens, and, in most if not all democratic systems, at least some steps have been taking to achieve equity (such 184 as social welfare programs, affirmative hiring, and so on) to mitigate external exclusions from political practices that serve democratic functions. Unfortunately, discursive inequalities continue to generate asymmetries in communicative participation and influence. However, there are practices that can address – neutralise, or even eliminate – discursive inequalities. In Chapter 9, I return to this question to discuss how we might achieve discursive equality and internal inclusion, so that practices such as talk can achieve their democratic functions, and so that misunderstandings and disagreements can be resolved through rational discourse. 185 Chapter 7: An Analysis of Gendered Asymmetries in Political Participation As I explained in Chapter 6, the battle for formal, legal political participation rights is mostly won: in today’s democracies, women can legally vote, run for office, and participate in other political acts. However, I have been claiming that – although formal barriers to women’s political participation have been removed – ongoing asymmetries in women’s discursive participation and influence undermine the universal democratic values of self-development and self- and collective-rule. But do women really participate in political practices – particularly communicative political practices – at lower rates than men? In this chapter, I review the literature on the gender gap in political participation and offer a novel analysis of the extent of, and determinants of women’s communicative political participation, paying attention to the way that gender intersects with other important social attributes, including ethnicity and poverty. In the first section, I review existing literature on the gender gap in political participation (more generally) as well as the existing empirical literature on both everyday political talk and deliberation. My review of the literature shows that the nature of the participatory gender gap varies across different forms of participation – including more traditional forms of participation, such as electoral participation or contacting elected officials, and newer forms of activism, such as “buycotting” (Coffé and Bolzendahl 2010; Dalton 2008; Stolle and Micheletti 2013). But gendered variation in communicative political participation – deliberation or everyday talk about matters of collective concern – remains largely unexamined. This raises the questions: do women participate in talking politics at the same rate as men? And does gender intersect with other identities and ascribed attributes – such as ethnicity or poverty – to impact political talk in important ways? In the second section of this chapter, I 186 explain how I will answer these questions using a cross-sectional research design with data from the Canadian Election Studies (2015) survey. In the third section of this chapter, I review my findings. My results reveal that women participate significantly less than men in the most demanding form of political talk: group deliberation. I find that ethnicity interacts with gender in some important ways, although other economic and resource-based factors – such as poverty and working for pay – do not disproportionately suppress women’s political talk. Social and political capital are more important for explaining variation in political talk, although they do not completely close the gender gap. In the fourth section, I discuss the implications of my findings. I conclude by discussing the limitations of my research design and data, and propose avenues for future work. 7.1 Literature 7.1.1 A Gender Gap in Political Participation An engaged, participatory public is an intrinsic, democratic good, instrumental to producing just norms, policies, and social structures, and to promoting democratic responsiveness. Ongoing patterns of unequal participation along social divides such as gender or other social attributes are “threats to both political equality and democratic performance” (Coffé and Bolzendahl 2010, 319). With the exception of voting, there is a small but persistent gender gap in formal political participatory acts – including working for a candidate, joining a political party, contacting political officials, and donating money – across democracies worldwide (Inglehart and Norris 2003; Dalton 2008; Coffé and Bolzendahl 2010; Paxton, Kunovich, and 187 Hughes 2007). With respect to voting, in most countries women have closed the gender gap (Paxton, Kunovich, and Hughes 2007), and, in Canada, women have “reversed” the gender gap, and vote in higher proportions than men (Elections Canada 2016). The importance of distinguishing between forms of participation is evinced by the fact that gender does not always predict newer, or less conventional forms of political activism. For instance, Dalton (2008) finds that although men are more likely to participate in conventional politics (with the exception of voting), gender does not significantly predict different forms of protest (including boycotting and buycotting), nor does gender predict internet activism (visiting a website, forwarding a political e-mail, or other Internet political activity). Coffé and Bolzendahl (2010) further distinguish between types of activism, and find that women prefer what they call “private activism” such as signing a petition, boycotting, or raising funds (which are low-conflict forms of activism). By contrast, the authors find that women are less likely to participate in what they call “collective activism,” such as demonstrating, or joining a meeting or rally (which are more confrontational forms of activism), or “contacting activism,” which includes contacting politicians, the media, or joining internet forums. Finally, a growing body of literature reveals that the gender gap is reversed with respect to “buycotting” – purchasing products for ethical reasons – with women participating at higher rates than men (Stolle, Hooghe, and Micheletti 2005; Stolle and Micheletti 2006, 2013). Clearly, research on the participatory gender gap must distinguish between different forms of political participation. Gidengil (2007) identifies two other important “gender gap” pitfalls to avoid. First, most gender gap research considers men as the baseline category, implicitly taking men’s behaviour as the norm. This can result in overlooking men’s contributions to gendered phenomena, such as men’s declining support for the New Democratic 188 Party (NDP) in the 1970s (and men’s growing support for Reform in the 1990s). Second, gender gap research risks what Gidengil refers to as “categorical thinking,” or essentializing gender. By contrasting women’s and men’s behaviour, gender gap research often ignores differences among women or among men. According to Gidengil (2007), gender gap research must not only attend to both women’s and men’s contributions to gendered patterns of political behaviour, it must attend to the ways in which gender intersects with other identities, particularly ethnicity and class. With respect to explaining the participatory gender gap, there are two general approaches: resource-based explanations (on the one hand), and explanations that focus on gendered patterns of socialisation and social and political capital (on the other). Resource explanations hypothesise that material-based gender inequalities – in finances, free time, and occupational status – suppress women’s aptitude and time for politics (Schlozman et al. 1995; Schlozman, Burns, and Verba 1999; Verba, Burns, and Schlozman 1997). Because donating money is largely explained by having money, the “feminisation of poverty” is assumed to suppress women’s ability to donate to campaigns, or engage in other resource-intensive activities such as volunteering for a campaign (Brady, Verba, and Schlozman 1995). And because childrearing and housework duties disproportionately fall on women’s shoulders, the presence of children in the home is assumed to reduce women’s free time and thus participatory engagement (Coffé and Bolzendahl 2010). The “double day hypothesis” posits that working women with children will have particularly low political engagement (Gidengil, Giles, and Thomas 2008; Thomas 2012). But existing empirical findings offer mixed evidence for resource-based expectations. The existing research finds that resource-based factors – education, income, job skills, and 189 having pre-school aged children in the home – impact the political engagement of both women and men (Schlozman, Burns, and Verba 1994; Verba, Burns, and Schlozman 1997; Coffé and Bolzendahl 2010; see also Gidengil, Giles, and Thomas 2008; Thomas 2012). There is little evidence that the difference between women’s and men’s participation is due to gendered differences in socio-economic resources. There is also little empirical evidence for the “double-day” hypothesis (Gidengil, Giles, and Thomas 2008; Thomas 2012). Existing research also reveals the effects of children on political engagement is ambiguous. Some findings suggest that although young children reduce both women’s and men’s participatory engagement, children have a slightly bigger effect of suppressing women’s participation (Coffé and Bolzendahl 2010). However, research must be attentive to context, since local factors – such as access to parental leave and affordable childcare – shape children’s effect on parents’ political engagement. A comparative study of women’s and men’s political self-confidence19 reveals that the presence of young children in the home suppresses women’s political self-confidence in the United States (where there is no guaranteed parental leave), but not in Canada (Gidengil, Giles, and Thomas 2008). In Canada, young children may even motivate women’s psychological political engagement and confidence. For instance, Melanee Thomas (2012) finds that working women with children report significantly higher political self-confidence than working women with no children. The second set of explanations, that gendered patterns of socialization and resulting social and political capital disproportionately benefit men, and thus asymmetrically boost men’s 19 Measured by respondents’ response to the statement: “sometimes politics is too complicated for a person like me to understand.” 190 (or suppress women’s) political participation, goes a bit further in explaining the persisting gender gap in political participation. Socialisation is considered essential for boosting political capital, defined as political knowledge, efficacy, and interest. For instance, as I explained in my earlier chapters, of the most important predictors of whether girls become interested in politics is whether they have politically active parents, particularly politically active mothers (Fox and Lawless 2014; Gidengil, O’Neill, and Young 2010). Given that women are on-average less interested and knowledgeable about politics, their daughters tend to be too – and girls’ on-average greater disinterest in politics translates into on average less politically engaged women (Fox and Lawless 2014). Women’s on-average lower political capital – for instance, their on-average lower political interest and efficacy – is doubly problematic because political capital seems to matter more for women’s political participation than for men’s. The gendered effect of political knowledge on conventional participation led Lazarsfeld, Berelson, and Gaudet (1948, 48–49) to conclude that: Men are better citizens but women are more reasoned… If a woman is not interested [in an election], she just feels that there is no reason why she should vote. A man, however, is under more social pressure and will therefore go to the polls even if he is not ‘interested’ in the events of the campaign. This nearly 70-year-old observation – that politics is perceived to be a man’s game, and so men feel greater social obligation to play, regardless of interest in, or knowledge of the sport – may unfortunately continue to hold weight today. Although political interest boosts both women’s and men’s participation in conventional politics, it has a bigger effect on women’s propensity to 191 vote (Coffé and Bolzendahl 2010). And political efficacy has a slightly bigger effect on women’s propensity to participate in private activism (signing a petition, boycotting, or raising funds). Like socialisation, social capital boosts political capital and political participation, but social capital’s contribution to the political gender gap is more ambiguous. For instance, Schlozman, Burns, and Verba (1994) find that organisational engagement significantly helps explain women’s – but not men’s – conventional political participation. This leads Schlozman, Burns, and Verba (1994) to suggest that civic engagement is especially important for reducing (albeit not eliminating) the participatory gender gap. 7.1.2 Discursive Participation (Or Political Talk) I have considered different forms of political participation (conventional and non-conventional), and reviewed different purported explanations for variation in different forms of political engagement. But what about communicative political participation? So far, only one U.S. study has looked at the extent and determinants of communicative political participation. In their nationally-representative survey study of United States’ citizens, Jacobs, Cook, and Delli Carpini (2009, 36) find that despite low expectations regarding the frequency of public talk, “a sizable portion of Americans are engaging in various forms of discursive participation.” By “discursive participation” (or what I refer to as political talk) the authors mean conversations – formal and informal, in person or online – about matters of collective concern. The authors distinguish between six types of political talk. The first two, “face-to-face deliberation” and “Internet deliberation” refer to more intensive, formal groups organised for the specific purpose of discussing local, national, or international issues (either face-to-face 192 deliberation, or deliberation taking place in chatrooms, message boards, or other online message groups). The authors also identify two forms of informal public talk, “traditional talk” (or informal face-to-face or telephone conversations about public issues), and “Internet talking” (e-mail or instant messaging about issues of public issues). Finally, Jacobs and colleagues distinguish between two forms of political persuasion: persuading someone to adopt one’s view, and persuading someone about whom to vote for. Only controlling for socio-demographic characteristics (gender, ethnicity, age, and income) and political characteristics (partisanship and ideology), the authors find significant evidence women are significantly less likely to have participated in Internet deliberation, or to have tried to persuade others on issues and on vote choice. Only controlling for socio-demographic characteristics, the authors also find some evidence women are less likely to participate in internet talking (p=0.10). Jacobs et al. (2009) also find that ethnic minorities are also less likely than white Americans to participate on most measures of political talk, but income only significantly explains vote and issue persuasion. Considering “discursive intensity” (the number of discursive acts participants engaged in) as the dependent variable, it appears that women, ethnic minorities, and those with lower incomes participate in significantly fewer discursive acts than men, white Americans, or rich Americans, when controlling for other socio-demographic and political characteristics. However, when Jacobs et al. (2009) add controls for social capital (organisational membership, church attendance, and length of residency) and political capital (political efficacy, trust, and knowledge, social trust, political attention, and political tolerance) to their model, they find that the effects of socio-demographic characteristics wane. After controlling for social and political capital, gender only significantly explains Internet deliberating. The effect of gender on 193 vote persuasion and discursive intensity drops to marginal significance, and there are no differences between women and men on other measures of political talk after controlling for social and political capital. Controlling for social and political capital also explains most of the apparent effect of race on political talk, although African Americans still participate significantly less than their white counterparts on certain measures of political talk. And controlling for social and political capital explains the apparent effect of income on each discursive participatory act, as well as on discursive intensity. Jacobs et al.’s (2009) work offers hope that, at least with respect to political talk, social and political capital explain the gender gap. That is, closing economic-gaps as well as gaps in social and political capital should largely eliminate the gender gap (and most gaps) in political talk. However, Jacobs’ et al.’s (2009) study does not give sufficient insight into gendered patterns of political talk. For one thing, although the authors control for socio-demographic facts, they treat women (and African Americans, and Latinos, and low-income Americans) like monolithic groups, and do not consider the ways in which social group memberships interact. Second, because the authors are not specifically interested in a gender gap in participation, they do not consider other variables that might be important in explaining gendered variation in political talk, such as the presence of young children in the home, or whether the respondents work for pay. Finally, Jacobs et al.’s (2009) survey was fielded over a decade ago, in 2003 – and Internet usage access has changed dramatically in that time.20 20 To put this in context, the first iPhone was released a full four years later, in 2007. Social media was also barely nascent: Facebook was invented a year later, in 2004 (for Harvard students only), and was only released to the public three years after Jacob et al.’s survey, in 2006. Twitter was also released in 2006. 194 The authors found only four percent of their sample (n=61) had ever participated in Internet deliberation (Internet chatroom or message board conversations about matters of collective concern), and only 24 percent of their sample (n=360) had ever participated in Internet talking (emailing or instance messaging about issues of public concern). The marginal or non-findings related to Internet political talk could be an artifact of low power. And furthermore, growing Internet usage could mean changing patterns of online political talk, and raises the question: has increasingly ubiquitous Internet access democratised online political talk, or does online political engagement continue to be a “weapon of the strong” (Schlozman, Verba, and Brady 2010)? Drawing from these literatures, I identify expectations about gendered variation in Canadians’ political talk. First, I expect to find a gender gap in participation, even accounting for the main effects of resource-related factors, including poverty, employment status, education, and the presence of young children in the home. I expect that adding measures of social and political capital (civic engagement, social trust, political interest, and political efficacy) will do a better job of accounting for variation in men’s and women’s discursive engagement. I consider will also consider how gender intersects with other ascribed attributes (ethnicity), structural features (poverty and working a double-day), and social and political capital. Considering how ethnicity intersects with other social identities and attributes helps avoid categorical thinking, and will offer better insight into the causes of the gender gap in discursive participation. I hypothesise that gender and ethnicity intersect to produce a “double disadvantage” for visible minority women. I also test the “feminisation of poverty” hypothesis to see if poverty particularly suppresses women’s discursive engagement. And while the existing literature is ambiguous with respect to the effects of children on women’s political engagement, I 195 hypothesise that – because young children make disproportionate demands on women’s time – the presence of young children in the household will particularly suppress women’s political talk. More specifically, I test the “double-day hypothesis,” to see whether working women with young children participate less than other social group members in talking politics. Finally, I expect that controlling for asymmetries in social and political capital will lend more insight into the gender gap in political talk. I hypothesise that social capital (civic engagement and social trust) increase political talk among both women and men, and that it is especially important for women’s political talk. I also expect that political capital (political efficacy and political interest) is especially important for women’s political talk. Table 7-1 Gendered variation in communicative political participation hypothesis list Resource Model Predictions H1 Resources (social privilege associated with ascribed status of gender and ethnicity, including: income, employment status, education, and marital status) are related to discursive participation, but accounting for resources will not close the gender gap in political talk. Full Model: Accounting for Resources and Social and Political Capital H2 Double disadvantage hypothesis: There is a “double disadvantage” for visible minority women, even controlling for social and political capital. H3 Feminisation of poverty hypothesis: Poverty particularly suppresses women’s political talk, even controlling for social and political capital. H4 Double-day hypothesis: Working a “double day” (working for pay and caring for young children) particularly suppresses women’s political talk, even controlling for social and political capital. H6A Importance of social capital (civic engagement) hypothesis: Organisational membership is especially important for women’s participation in political talk. H6B Importance of social capital (social trust) hypothesis: Social trust is especially important for women’s participation in political talk. H7A Importance of political capital (political interest) hypothesis: Political interest is especially important for women’s political talk. H7A Importance of political capital (political efficacy) hypothesis: Political efficacy is especially important for women’s political talk. 196 7.2 Methods and Data I use data from the 2015 Canadian Election Studies (CES 2015) mixed-mode (telephone and Web) survey to test my hypotheses about gendered variation in Canadians’ political talk. The telephone survey was conducted by the Institute of Social Research at York University, whereas the online survey was conducted by Survey Sampling International’s panel of Canadian respondents. I consider three forms of communicative political participation: group deliberation, everyday political talk, and Internet political talk. Following my discussion in Chapter 3, I distinguish between deliberation and everyday talk about matters of collective concern. The first two dependent variables, group deliberation and everyday political talk, refer to more conventional conversations happening in real life (not online), while Internet political talk refers to online communication. Group deliberation is the only variable in my analysis measuring deliberation, while the other two outcome variables are measures of everyday communication about matters of collective concern, or everyday political talk. With respect to more conventional conversations in real life, group deliberation is an indicator variable measuring whether respondents have attended a meeting to discuss a local, national, or international issue in the past 12 months (1=have attended a meeting) (see Table 7-2). This item was only asked on the CES 2015 web survey (n=4,146). These kinds of face-to-face groups organised to discuss matters of public concern are the “gold standard” of deliberation, but place the highest demands on participants in terms of time and political capital (Jacobs, Cook, and Delli Carpini 2009, 35). Although I expect to find the same patterns of results for all outcome variables in my analysis, I expect that the gender gap will be largest for group deliberation. 197 Everyday political talk refers to more informal conversations about politics and matters of collective concern. It is measured by asking respondents whether they “discussed politics and news” with friends or family in the past week (1=have discussed news with friends or family) (see Table 7-2). This item was asked on both the CES 2015 web and phone surveys (n=6,790). Political talk should be far more common than group deliberation, since it places the lowest demands on participations. As such, I expect that the effect of gender on political talk will be smaller than on group deliberation. Internet political talk is also measured with an indicator variable measuring whether respondents “exchanged political news and ideas on the Internet” in the past week (1=have exchanged news and ideas on the Internet). This item was asked on both the CES 2016 web and phone surveys. Although internet political sharing might not always constitute a conventional “conversation” (Tweeting or posting can be fairly unidirectional), it does constitute a form of everyday political talk. People express themselves – their beliefs, opinions, and preferences – by posting and sharing links to articles, blogs, or other media online. As such, political sharing is important for self-development, even if it is further removed from self- and collective-rule. Because Internet political talk requires lower investments in time and political capital than group deliberation, I expect the impact of gender to be smaller on Internet political talk than on group deliberation. Finally, it should be noted that my measures of online political talk differ considerably from Jacobs et al.’s (2009) measures. This is necessary to capture the ways online communication have changed since Jacobs and colleagues fielded their 2003 survey. For instance, the release of the first iPhone and popularization of smartphones, which blurred the 198 distinction between “talking on the phone” and “talking online,” and the invention of Facebook and Twitter with the advent of “posting” or “Tweeting” as a form of political expression. Table 7-2 Distributions of dependent variables Variable Obs Mean SD Min Max Group deliberation 4,146 0.22 0.41 0 1 Everyday political talk 6,790 0.81 0.39 0 1 Internet political talk 6,979 0.29 0.45 0 1 To identify the determinants of political talk, I first test a baseline model, then I consider the resource model of political talk, and finally I test the full social and political capital model of political talk. The baseline model includes my only primary independent variables of interest, and basic controls. My primary explanatory variables are an indicator variable measuring gender (coded female=1), and an indicator variable measuring visible minority status (coded visible minority=1)21 (see Appendix A, Table A-1 Distributions of independent variables). I also control 21 Note that on both the web and phone surveys, the modal response to the ethnicity question was “Canadian.” The phone survey (but not the web survey) included a follow-up prompt which asked about the ethnic identity of the respondents’ parents or grandparents. The web survey allowed respondents to select more than one response, and many respondents additionally selected “Other” (and typed in their response). Even using the follow-up responses from the phone survey and open-ended responses from both surveys to recode the ethnicity variable, a sizable number of respondents fell into the category “Canadian.” As such, the visible minority measure is comparing visible minorities (visible minority=1) to non-visible minorities (people of European ancestry) and those who identify as just “Canadian” (non visible minority/Canadian=0). Comparing the mean responses of visible minorities, non-visible minorities, and “Canadians” on thermometers measuring feelings toward immigrants and feelings toward Aboriginal Canadians reveals “Canadians’” mean attitudes toward these social group members are almost identical non-visible minorities’ attitudes, and significantly different from visible minorities’ mean attitudes toward these social group members. Still, the visible minority category (13 percent of the distribution) underestimates the actual number of visible minorities in Canada (19 percent of the population). 199 for age category (18-29, 30-39, 40-49, 50-64, and 65 years of age and older), and region (an indicator variable, coded Quebec=1).22 In the resource model, I add an indicator of poverty (income below $25,000=1), an indicator of employment status (working for pay=1),23 and a five-category measure of education. The education measure ranges from less than secondary, secondary completion, some college or university, college or university completion, to graduate or professional degree. I also add an indicator variable capturing marital status (currently married or living with a partner=1), and an indicator variable capturing the effect of young children in the home (one child or more under the age of six years=1). In the full model, I add two measures of social capital: an indicator of civic engagement (organisational membership=1), and an indicator of social trust (agreeing that “most people can be trusted”=1). I also add two measures of political capital: an indicator of political efficacy (disagreeing that “politicians don’t care what people like me think”=1),24 and an indicator of high political interest (1=high political interest).25 I use the CES (2015) main weighting variable (CombWgt) which combines the provincial weights (available for the Web survey) and national weights (available for the phone survey) to provide post-stratification 22 As a robustness check, I ran the models with a categorical variable measuring region (Maritime, Quebec, Ontario, Prairies, British Columbia, and Northern). Interestingly, although Quebec residents are significantly different from residents in all other regions, none of the regions outside Quebec are significantly different each other. 23 Note that this is a recoded version of the CES 2015 employment status variable, coded so that respondents who indicated they are “working for pay”=1, and everyone else is=0. “Everyone else” includes respondents who indicated they are “students and working for pay,” “caring for family and working for pay,” and “retired and working for pay.” This was an oversight on my part that I realised in retrospect – these three categories should be included in “working for pay.” Including these three groups in “working for pay” makes almost no difference to the results, but future work should include a properly recoded employment variable. 24 I initially included political efficacy (responses to the statement “Politicians don’t care what people like me think” on a four-category scale from strongly disagree to strongly agree). However, there is no difference between any of the categories except those who indicated the highest political efficacy (strongly disagree). I recoded the variable to simplify interpretation. 25 Again, simplifying this variable does not change the substantive findings, but it makes estimating (and plotting) margins (particularly of interactions) must simpler. 200 weight to province, age, gender, and household size. Because the outcome variables in my analysis are dichotomous,26 I proceed with logistic regression. Of course, considering only the main effects of gender and other variables risks treating social groups (such as “women”) as coherent wholes, and potentially misses important differences within groups (such as between visible minority and non-visible minority women, or poor and non-poor women). To test my remaining hypotheses, I predict seven separate interaction models that control for demographics, resources, and social and political capital. The first interaction model includes an interaction between gender and visible minority status (to test the “double disadvantage” hypothesis), the second includes an interaction between gender and poverty (to test the “feminisation of poverty” hypothesis), and the third model includes a three-way interaction between gender, working for pay, and the presence of young children (to test the “double-day” hypothesis). The final four models test the importance of social and political capital hypotheses, and include interactions between gender and civic engagement, gender and social trust, gender and political interest, and gender and political efficacy, respectively. Finally, a note on quantitative research and intersectionality – while I recognise that quantitative research requires the use of categories (such as gender and ethnicity), and that categories themselves can be harmful, I follow the school of intersectional research that recognises there is a trade-off between scale (i.e., the complexity of social identities) and 26 The original variables were categorical, but because I could not assume the distances between the cutpoints were symmetrical, and because the outcome variables are highly skewed (the large majority of respondents indicate they have “never” engaged in most forms of discursive politics), I would have to either proceed with ordered logistic regression, or recode the dependent variables as dummies and proceed with logistic regression. The results (categorical versus indicator dependent variables) are almost indistinguishable (contact author for a table with the coefficients and odds rations of categorical DVs). As such, I recoded the dependent variables into indicatory dummy variables to simplify interpreting results. 201 coherence (i.e., operationalising concepts as measurable variables) (K. Davis 2008; McCall 2005). I believe that there are more effective strategies for resistance than rejecting categories out of hand, or than rejecting methods appropriate for the task at hand – quantitative analysis – because these methods are sometimes conflated with a philosophy of science (positivism). I believe that the provisional use of categories in quantitative research can respect the demand for complexity inherent in intersectional research. I also recognise that society and social inequality are “dynamic, complex, and contingent,” but that – despite this complexity and contingency – the social world is still “amenable to explanation” (McCall 2005, 1794). 7.3 Findings 7.3.1 Main effects Comparing the results for all four measures of political talk reveals some interesting findings (see Appendix A for results tables). Looking at the baseline model (see Table A-2 for baseline regression results), it appears that women participate in all measures of political talk – group deliberation, everyday political talk, and Internet political talk – at lower rates than men (although gender is only marginally significantly related to everyday political talk). And another important main effect in the baseline model, visible minority status, significantly reduces participation in everyday political talk. The resource model, which adds marital status, education, poverty, employment status, and the presence of children under the age of six in the home, improves upon the baseline model (see Table A-3 for resource model regression results). After accounting for material factors, 202 gender is no longer even related to everyday political talk, and the size coefficient measuring the effect of gender on group deliberation is reduced. As expected (Hypothesis 1), controlling for resource-related variables does not fully account for the gender gap in discursive participation. The full main effects model, which includes controls for social capital (civic engagement and social trust) and political capital (political interest and political efficacy) further reveals that some of the effects that appeared to be due to gender are in fact due to gendered differences in social and political capital (see Figure 7-1 for regression results predicting group deliberation, see Figure 7-2 for regression results predicting everyday political talk, see Figure 7-3 for regression results predicting Internet political talk, or see Table A-4 for the results tables for all three outcome variables). After controlling for resources and social and political capital, gender is no longer significantly related to Internet political talk (Figure 7-3). However, a small but persistent gap in the likelihood men and women participate in the most taxing form of political talk – group deliberation – remains (Figure 7-1). 203 Figure 7-1 Full model (main effects) logistic regression results, predicting group deliberation (95% confidence intervals) 204 Figure 7-2 Full model (main effects) logistic regression results, predicting everyday political talk (95% confidence intervals) 205 Figure 7-3 Full model (main effects) logistic regression results, predicting Internet political talk (95% confidence intervals) 206 Controlling for the main effects of demographics, resources, and social and political capital, and calculating the predictive margins (average adjusted predictions) for gender reveals that there is a 25 percent predicted probability a man participates in group deliberation, and a little less than a 22 percent predicted probability that a woman participates in group deliberation (Figure 7-4, or see Table A-5 for predictive margins). The 3 percent difference in the predicted probabilities men and women participate in group deliberation is statistically significant (see Table A-6 for the discreet change in gender). Figure 7-4 Predicted probabilities men and women participate in group deliberation (main effects) 207 7.3.2 Interaction effects Modelling interactions helps clarify how ethnicity, resources, and social and political capital help moderate the relationship between gender and different forms of discursive participation. First, including an interaction between gender and ethnicity in the full model (controlling for demographics, resources, and social and political capital) seems to suggest that gender only matters for group deliberation among white Canadians. Comparing white men and women suggests that, at least among white respondents, gender impacts the likelihood of participating in group deliberation. There is just over a 24 percent predicted probability a white man participates in group deliberation, and a 21 percent predicted probability a white woman does the same (see Table A-8 for predictive margins). The three percent difference in the predicted probability white men and white women participate in group deliberation is statistically significant (see Table A-9 for the discreet effect of gender by ethnicity). The difference between visible minority men and visible minority women is not significant. Second, ethnicity matters, but for men’s discursive engagement. Visible minority status only significantly reduces the predicted probability that men participate in everyday political talk (Figure 7-5, see Table A-7 for regression results). While there is an 84 percent predicted probability that a white man participates in everyday political talk, there is only a 79 percent predicted probability that a visible minority man does the same (see Table A-8 for predictive margins). Men of colour are significantly less likely than white men to talk with friends and family about news and politics (Figure 7-5, or see Table A-10 for discreet effect of ethnicity by gender). 208 Figure 7-1 Average marginal effects of ethnicity (the effect of being a visible minority on the predicted probability of participating in discursive politics) by gender, for three forms of discursive politics (with 95% confidence intervals) Contrary to expectations, there is no evidence that visible minority status particularly impacts women’s participation in everyday political talk. While there is an 84 percent predicted probability a white man or a white woman participates in everyday political talk, there is an 82 percent predicted probability a visible minority woman does the same – and the small difference between white women and visible minority women does not reach conventional levels of statistical significance (see Table A-10 for discreet effect of ethnicity by gender). Contrary to 209 expectations, there is also no evidence that visible minority status particularly impacts women’s (or men’s) participation group deliberation. However, the results for group deliberation should be interpreted with caution, because my measure of “group deliberation” was only included on the Canada Election Study’s online survey (it was omitted from the telephone survey). Since a small number of visible minority respondents (n=457, about equal numbers of women and men) completed the Canadian Election Study’s online survey questionnaire, these non-findings could be due to low power. As Figure 7-6 illustrates, poverty also moderates the effect of gender on group deliberation and everyday political talk (see A-11 for regression results). There is a 26 percent predicted probability that a man who is not suffering from poverty participates in group deliberation, and only a 21 percent predicted probability a poor man does the same (see Table A-12 for predictive margins), a significant difference of five percentage points (Figure 7-6, or see Table A-14 for discreet effect of poverty by gender). And although poor women appear to participate less than women who do not suffer from poverty in group deliberation, this difference is not statistically significant. Similarly, there is a significant, seven percentage point difference in the predicted probability poor and not poor men participate in everyday talk, whereas the difference among women is negligible and does not reach conventional levels of statistical significance. Poverty is not significantly related to Internet political discussion. 210 Figure 7-2 Average marginal effects of poverty (the effect of being poor on the predicted probability of participating in discursive politics) by gender, for three forms of discursive politics (with 95% confidence intervals) Interestingly, I also found no evidence for the “double day” hypothesis. Modelling the interaction between gender, working for pay, and the presence of young children in the home reveals these variables do little to moderate the effect of gender on group deliberation, everyday political talk, or Internet political expression (see Table A-15 for regression results, and see 211 Table A-16 for predictive margins). Having children and working for pay does not suppress either women’s or men’s political talk. Looking now at the interaction between gender and civic engagement, it is clear civic engagement – measured as whether respondents volunteer for a group or organisation, such as a school, religious organisation, or sports or community associations – is centrally important for predicting discursive political participation (see Table A-17 for regression results). And, at least with respect to group deliberation, the effect of participating in civil society differs across genders. There is only an eight percent predicted probability a woman who does not volunteer for a civic association participates in group deliberation, and only a 10 percent predicted probability a non-engaged man does the same – this small difference between non-civically engaged women and men is not significant (see Table A-18 for predictive margins). However, there is a whopping 38 percent predicted probability that a civically engaged woman participates in group deliberation, and about a 43 percent predicted probability that a civically engaged man does the same. The significant difference of just over five percentage points between civically engaged men and women suggests that, contrary to expectations, men gain greater rewards than women from participating in civil society with respect to participating in group deliberation (Figure 7-7, and see Table A-19 for the discreet effect of gender by civic engagement). Civic engagement also boosts the predicted probability a person participates in everyday political talk by about nine percentage points, but the effect of civic engagement on everyday political talk is similar for men and women. Finally, civic engagement boosts the predicted probability a person participates in everyday Internet political talk by the same amount (just 212 under nine percentage points), but again the effect of civic engagement on Internet political talk is similar for both women and men. Figure 7-3 Average marginal effects of civic engagement (the effect of civic engagement on the predicted probability of participating in discursive politics) by gender, for three forms of discursive politics (with 95% confidence intervals) 213 Like with civic engagement, the effect of my other measure of social capital – social trust – on group deliberation also differs for men and women (see Table A-21 for regression results). While there is no significant difference in the predicted probability trusting men and non-trusting men participate in group deliberation, the probability a trusting woman participates in group deliberation is four percentage points higher than a non-socially trusting woman (Figure 7-8, or see Table A-22 for predictive margins, and Table A-23 for the discreet effect of gender by social trust). This suggests that, as hypothesised, social trust matters more for encouraging women’s participation in the most taxing form of discursive politics. 214 Figure 7-4 Average marginal effects of social trust (the effect of having high social trust on the predicted probability of participating in discursive politics) by gender, for three forms of discursive politics (with 95% confidence intervals) Social trust also significantly increases the likelihood people participate in everyday political talk, but the effect of social trust on everyday political talk is similar for both women and men. Interestingly, social trust is not significantly related to the likelihood people participation in Internet political talk, and this is true for both men and women. Analysing the way civic engagement and social trust moderate gender’s effect on different forms of political 215 talk offers mixed results. Contrary to expectations, there is no evidence that civic engagement disproportionately boosts women’s participation in political talk to close the gender gap. In fact, men seem to gain greater rewards from civic engagement with respect to participating in group deliberation. However, there is evidence that – as I hypothesised – social trust is particularly important for boosting women’s engagement in group deliberation. Political capital – defined here as political interest and political efficacy – also interacts with gender to impact discursive political participation. People who are interested in politics are significantly more likely than those with less interest in politics to participate in group deliberation (Figure 7-9), but this effect varies by gender – high political interest gives men a greater boost with respect to participating in group deliberation (see Table A-25 for regression results, see Table A-26 for predictive margins, and see Table A-27 for the discreet effect of gender by political interest). 216 Figure 7-5 Average marginal effects of political interest (the effect of having high political interest on participating in discursive politics) by gender, for three forms of discursive politics (with 95% confidence intervals) High political interest also significantly increases the predicted probability people participate in everyday political talk (by about 15 percentage points), but the effect of political interest on everyday political talk is similar for women and men. Finally, high political interest is quite important for Internet political talk. Increasing the predicted probability of participating in 217 Internet political talk by about 24 percentage points. But again, the effect of political interest is similar for both women and men. Political efficacy – measured as whether respondents disagree that “politicians don’t care what people like me think” – is less important for discursive politics (see Table A-29 for regression results, for predictive margins see Table A-30). Political efficacy has essentially no effect on participating in group deliberation or everyday political talk, among both men and women (Figure 7-10, or see Table A-39 for the discreet change in gender by political efficacy). This is interesting, because it suggests that regardless of whether people feel disempowered by an important feature of electoral democracy – politicians’ responsiveness – citizens still attend meetings to talk about local, national, or international issues, and still talk about politics and the news with their friends and family. 218 Figure 7-6 Average marginal effects of political efficacy (the effect of having high political efficacy on participating in discursive politics) by gender, for three forms of discursive politics (with 95% confidence intervals) However, political efficacy does interact with gender to impact the predicted probability men and women participate in Internet political talk. Although there is no evidence that higher political efficacy impacts women’s propensity to participate in Internet political talk, the predicted probability a man participates in Internet political talk is nine percentage points higher among high efficacy men, as compared to low efficacy men. 219 7.4 Discussion My analysis offers mixed hope for the ideal of achieving equal discursive political participation among men and women, and across the social divides of ethnicity and poverty. The bad news is that, even accounting for demographics (age and region), the effects of resource-related factors (poverty, education, marital status, employment status, and having young children), and the effects of social and political capital (civic engagement, social trust, political interest, and political efficacy), women are still significantly less likely than men to participate in group deliberation, the most taxing form of discursive participation. That women seem as interested as men in everyday political talk and Internet political discussions (controlling for social and political capital), but are less likely to participate in more formal group deliberation is congruent with existing research that suggests women prefer informal political acts and organisations in civil society, as well as more “private” political acts (Siim 2000; Coffé and Bolzendahl 2010). The good news is that controlling for social and political capital there are no gendered differences in everyday political talk, and Internet political talk. In other mostly good news, visible minority women do not appear to suffer from a “double disadvantage” – visible minority women do not differ significantly from white women in terms of the predicted probabilities they participate in group deliberation, everyday political talk, or Internet political expression. This is only a “mostly good” news because when it comes to talking politics women are still at a disadvantage. However, it appears there is a single, gender disadvantage for women: both white women and women of colour are less likely to participate in group deliberation than men. 220 Unfortunately, visible minority status still matters for talking politics. Among men, visible minority status does significantly reduce the predicted probability of participating in everyday political talk. Further examining the suppressing effects of visible minority status on men’s (but not women’s) everyday political participation would be another interesting avenue for future research. Another somewhat happy finding is that, contrary to expectations drawn from the literature, there is no evidence that poverty disproportionately suppresses women’s political engagement with respect to talking politics. However, poverty still suppresses discursive engagement – although interestingly, it appears poverty disproportionately impacts men. Poor men are significantly less likely than men who are not suffering from poverty to participate in group deliberation and everyday political talk. Further examining the suppressing effects of inequality on men’s (but not women’s) discursive engagement would be another interesting avenue for future research. Contrary to my “benefits of social and political capital” hypotheses, I also found little evidence that social and political capital give women a special boost relative to men. Although social trust is positively related to the likelihood women (but not men) participate in group deliberation, both civic engagement and political interest boost men’s participation in group deliberation more than they boost women’s participation in group deliberation – which helps explain why there is an ongoing gender gap in group deliberation, even after controlling for social and political capital. And political efficacy is positively related to the likelihood men (but not women) participate in Internet political talk. 221 Unfortunately, this means that boosting women’s social and political capital so that they are equally civically engaged as men, and equally interested in politics will not close the gender gap in group deliberation. Closing the gender gap in group deliberation will require even greater efforts by activists and policy-makers to encourage women’s participation in this form of political talk. I return to this question in Chapter 9, where I consider ways to increase women’s participation in group deliberation. 7.5 Conclusion Although women enjoy formally equal political participation rights in contemporary democracies, the ongoing gender gap in many measures of participatory engagement is concerning insofar as gendered asymmetries in participation entail collective outcomes that are less attentive to women’s interests or preferences. As the existing literature shows, the nature of the gender gap in political participation varies across political acts. My work contributes to the gender gap in political participation literature by illustrating the extent of asymmetries in communicative participation, including group deliberation, everyday political talk, and Internet political talk. My work also illustrates how gender intersects with other salient identities to impact participation in discursive politics, and sheds like into the determinants of discursive engagement for both women and men. However, my present work suffers from certain limitations, particularly with respect to operationalising visible minority status (or “race”). Although I found small effects for ethnicity, and little evidence of a double disadvantage for women of colour, I am hesitant to draw the 222 conclusion Canadians have achieved equality between historically empowered (non-visible minority) and historically disempowered (racialised) ethnic groups. Given the underrepresentation of visible minorities in the CES 2015 sample (13 percent) relative to the Canadian population (19 percent), and difficulties operationalising the ethnicity variable (due to the large proportion of Canadians who only identified themselves as “Canadian”), my non-findings may be a product of measurement error. Future studies should endeavour to better measure minority status, and would benefit from an oversample of Canadians who identify as visible minorities. An oversample of Indigenous (Aboriginal) Canadians would be especially interesting – the small proportion of the CES 2015 who identify as Aboriginal (149 respondents) means making statistical inferences about Indigenous Canadians’ discursive engagement is impossible with the current dataset. But given that Indigenous Canadians have been excluded from formal political rights even longer than women have, and participate in formal politics at much lower rates than other ethno-national groups (Fournier and Loewen 2011), it would be interesting to gain insight into Indigenous engagement in more informal politics. It would be particularly interesting to gain insight into how the intersection of gender and Indigeneity impacts different forms of political engagement. 223 Chapter 8: An Analysis of Gendered Asymmetries in Communicative Influence I have been claiming that – although formal barriers to women’s political participation have been removed – ongoing asymmetries in women’s discursive participation and influence undermine the universal democratic values of self-development and self- and collective-rule. In Chapter 7, I reviewed the literature on the gender gap in political participation and offered a novel analysis that reveals there is an ongoing gender gap in communicative political participation. In this chapter, I consider how social cognitive and motivational structures can create unjustifiable asymmetries in discursive influence, by influencing how speakers respond to counterarguments. I use evidence from a vignette experiment to show how discursive inequities, or unjustifiable asymmetries of communicative influence, can contribute to the internal exclusion of women speakers from communicative processes of opinion formation. Discounting women’s and enhancing men’s influence in communicative process of opinion formation can reinforce existing patriarchal relations of domination, since women exert less influence in discursive political practices required to modify their social and political conditions. In the first section, I discuss the problem of gendered “discursive inequities” that I introduced in Chapter 6 in more detail. In the second section, I describe the survey experiment and research design and methods that I use to test my expectations about how gender inequality can structure cognition to contribute to the internal exclusion of women speakers. In the third section, I discuss my findings. I find that all respondents are more open to receiving a man’s 224 counterargument, but that female respondents seemed especially influenced by a man’s counterargument. Gendered differences in self-esteem partially explain why females are especially sensitive to a man’s counterargument. I conclude by describing discursive inequity’s “system justifying” (Jost, Banaji, and Nosek 2004) effects, as discounting women’s and enhancing men’s influence in communicative process of opinion formation can reinforce gender-based hierarchies even in the absence of formal patriarchal structures. 8.1 Literature: Discursive Inequity and Women’s Internal Exclusion Democratic processes always begin with inclusion. The aims of forming collective opinions and agendas, and making collective decisions are undermined if those affected by collective endeavours are excluded from participating in them (Fung 2013; Goodin 2007; Young 2000). Most democratic theories consider the problem of external exclusion, and “call for limiting the influence of wealth or position on the ability to participate in a democratic process” (Young 2000, 55). However, as I have explained, inequality can also structure social cognition to engender internal exclusions. This refers to when, even in the absence of formal political or economic asymmetries, historical asymmetries of social and political influence remain in the long history of habitus, shaping how people view themselves and other social group members to engender asymmetries in political participation and influence. When inequality structures cognition, habitus’ cognitive schemes of perception and appreciation can entail internal inequalities that prevent disempowered social group members from taking advantage of formally equal opportunities to participate in everyday talk about 225 matters of collective concern, and can generate unjustifiable asymmetries in disempowered social group member’s communicative influence. A review of the literature on small-group deliberation suggests that members of historically disadvantaged groups are less likely to speak in discussions, and feel less influential (Mendelberg 2002). Continuing with my example of women’s political participation, the legacy of women’s political exclusion embedded in cognitive schemes of perception and appreciation can lead women to perceive themselves (and lead others to perceive them) as less knowledgeable and influential in political conversations. For instance, research shows that when women and men talk politics, both women and men are more likely to perceive men conversation partners as more knowledgeable about politics than women, regardless of the conversation partners’ objective political knowledge (Mendez and Osborn 2010, 270). As Mendez and Osborn (2010, 270) note, “these differences have the potential to discount women’s contributions to political conversation.” I refer to cognitive schema that engender unjustifiable, systematic asymmetries of communicative participation and influence as discursive inequities. Discursive inequities are disabling constraints that entail internal exclusion, undermining disempowered social group members’ influence in conversations about matters of collective concern. This can have the teleological effect of reinforcing existing systems of oppression, since the disempowered cannot participate in (and may not even be motivated to participate in) political practices to modify their social and political conditions. Discursive inequities contribute to the internal exclusion of disempowered group members by influencing how people perceive speakers from social outgroups, and how they perceive themselves and other speakers from social ingroups. As I already mentioned, social 226 stereotypes are a particularly problematic discursive inequity that undermine disempowered social group members, including women’s, communicative influence. Since there are stronger stereotypical associations between men and attributes related to good decision-making (authoritative, decisive, and so on), then both men and women may give more weight to men’s utterances. Recall from Chapter 6 that another problematic discursive inequity is structural (group-based) differences in global self-worth, or self-esteem. Women report lower levels of self-esteem than men on average (Kling et al. 1999). This finding is robust across cultures and over time (Bleidorn et al. 2015). And gendered variation in self-esteem likely matters for discursive political participation. A study of small-group discussions of political science students shows that female students report lower levels of self-esteem, and that self-esteem is related to frequency of verbal participation (Beauvais and Yaylaci 2017b). Other researcher suggests that when avoidance is not an option in politics, low self-esteem individuals are more likely to conform to dominant views and behaviours (McGuire 1968; Gibson 1981). Currently, there are no studies on whether features of social cognition (such as stereotypes, or self-confidence) that engender unjustifiable, systematic asymmetries of communicative participation and influence – or what I call discursive inequities – impact the uptake of arguments in discourse. However, studies of small group deliberations and political conversations offer reason to suspect they might. One of the strongest predictors of a person’s influence in a discursive interaction is whether other participants in the conversation perceive them as more expert or competent in the task at hand (Bottger 1984; Kirchler and Davis 1986; MacRae Jr 1993; see Mendelberg 2002 for a review of this literature). Studies of jury 227 deliberations show that higher-status jurors tend to be perceived as more accurate, even though social status does not correlate with accuracy (Hastie, Penrod, and Pennington 1983). And as I mentioned, women and men talking politics are more likely to perceive men conversation partners as being more knowledgeable, regardless of their actual political knowledge (Mendez and Osborn 2010, 270). I designed a survey experiment to see whether discursive inequities impact the uptake of arguments in political discourse. I expect that, ceteris paribus, people would be more likely to change their minds after hearing a man’s counterargument, as compared to hearing a woman’s identical counterargument. I also consider what might be causing these asymmetries of communicative influence. I expect to find an interaction between the gender of the respondent and the gender of the person presenting the counterargument: Women will be particularly willing to change their minds after hearing a man’s counterclaim. And I expect this effect to be moderated by self-esteem: Lower self-esteem women will be the most sensitive to hearing counterarguments (particularly from men), whereas higher self-esteem women will be more resistant to counterarguments (and more resistant to men’s excess discursive influence). Gendered differences in self-esteem,27 and the differential ways lower self-esteem impacts 27 I also created a unique Implicit Association Test (IAT) to identify whether stereotypes contribute to gendered asymmetries of communicative influence as discursive inequities (see Table B-4 in Appendix). While most IATs measuring gender stereotypes measure the association between gender and high/low status jobs, or between science and arts, my novel IAT specifically tested implicit associations between gender and attributes related to good/bad decision-making (to increase the construct validity of the test). The IAT does effectively measure variation in implicit associations: Male and female respondents have significantly different mean D-scores, in the predicted direction (Table B-5 in Appendix B), and the D-scores correlate with explicit gender attitudes (Table B-6 in Appendix B). Unfortunately, I could not include this variable in my analysis. Very few respondents (n=67) completed the IAT component of the experiment. This is likely because students completed the survey experiment remotely. Most of the survey experiment was hosted on Qualtrics, and the students would have to click on a 228 women and men, contribute to gendered asymmetries of communicative influence as discursive inequities, which entail the internal exclusion of women speakers from processes of opinion formation. 8.2 Methods 8.2.1 Research Design I designed an online vignette experiment to test my hypotheses. Respondents were asked their opinion on a non-salient policy question, and then were randomly assigned to receive identical counterarguments from either a “woman” or a “man.” I chose a non-salient, non-partisan policy issue to try and ensure that respondents be somewhat willing to revise their initial stance, after hearing counterargument. The survey first asked respondents if they thought car insurance should be mandatory (required to everyone to drive on the road), or optional (those who show they can cover the costs of accidents do not have to pay for car insurance). Car insurance is mandatory in every province in Canada, and the province of British Columbia has operated a universal compulsory automobile insurance scheme since 1974 (Kent 1995), well before any of the undergraduate student respondents were born. As far as I am aware, there has hyperlink to open the IAT in a new web browser page. It seems most students did not complete this last step. As such, I do not include the D-Score in my dissertation, and instead focus solely on gendered variation in self-esteem. 229 never been any public debate or media coverage regarding eliminating the mandatory insurance requirement, and so this issue is not politicized, or partisan. Regardless of whether respondents indicated that insurance should be mandatory or optional, they were presented with a vignette counterargument (see Appendix B.1 for question wording). The treatment was the gender – denoted by a name, no picture – of a fictional person giving the counterargument. Half the sample was randomly assigned to receive the counterargument from “Jessica,” and the other half was randomly assigned to receive the counterargument from “Michael.”28 The survey then asked respondents follow-up questions, including the Rosenberg Self-Esteem (RSE) Questionnaire to test my hypotheses about how gendered variation in self-esteem contributes to the internal exclusion of women (see Data and Analysis). There are benefits and drawbacks to using an online survey experiment to study discursive processes. The benefit is that by indicating a respondent’s gender by only their name, with no visual or audible clues to signify race, education or social class, or other potentially relevant factors that might be related to persuasion, I control for these alternative causal variables and isolate the independent effect of speaker gender. The cost, of course, is external validity. One of the primary concern with survey experiments is that “the artificially clean environment of the survey question makes treatment easier to receive” (Gaines, Kuklinski, and Quirk 2007, 16; see also Barabas and Jerit 2010). However, many of the factors the experimental design controls 28 Two of the most popular names in the mid-1990s (when most of my respondents were born). Respondents should be familiar with many men named Michael and women named Jessica. 230 for may further advantage men. In real life, men possess additional attributes that are associated with authoritativeness: men tend to be taller (Hamstra 2014; Blaker et al. 2013; Egolf and Corder 1991), and men have louder and deeper voices (Strand 1999; Carli 2001). In this case, the “pristine experimental setting” (Barabas and Jerit 2010, 238) likely under exaggerates the effect of gender, making this a harder test of my hypotheses. Another drawback related to external validity is that in my experiment, respondents merely articulate a policy preference, are presented with one counterargument, and are asked to indicate the likelihood they will reconsider their initial preference. According to deliberative democratic theorists, discussants should have more time to exchange arguments, and should be able to hear from anyone affected by deliberative outcomes (Chambers 1996; Habermas 1990, 1998). My findings are less relevant for high quality deliberation, which might take place in small-scale, institutionally-bounded civic forums, where trained moderators and rules of discourse ensure higher quality communicative problem-solving (Beauvais and Bächtiger 2016; Beauvais 2017). Instead, my findings are relevant for the kind of discursive participation taking place in the unbounded, mass public sphere, when neighbours, co-workers, or strangers chat about matters of collective concern in informal settings. In such settings people may make and respond to off-hand comments about matters of collective concern, but – in most casual, informal conversations – are unlikely to stop and engage in a full-fledged debate. 231 8.2.2 Data and Analysis The data presented in this paper was collected in April 2016, from undergraduate students recruited through the Department of Political Science’s Subject Pool at the University of British Columbia. Respondents completed the survey experiment on their personal computers, remotely. The dependent variable is respondents’ response to the question “Given [Jessica’s/ Michael’s] counterargument, what is the likelihood you will change your mind about car insurance?” The dependent variable is a five-point scale that ranges from one, “I will definitely NOT change my mind,” to five, “It is very likely, or I will definitely change my mind”29 (Table 8-1). Since the dependent variable is ordinal, and the distance between the categories unequal, I proceeded with ordered logistic regression. Table 8-1 Distribution of the dependent variable Will Change Mind Freq. Percent Cum. Definitely NOT (1) 44 11.58 1.31 Very unlikely (2) 115 30.26 13.75 Somewhat unlikely (3) 119 31.32 54.17 Somewhat likely (4) 79 20.79 96.56 Very likely/Definitely (5) 23 6.05 100 Total 380 100 The primary independent variables in my analysis are respondent gender, counterargument gender condition, and self-esteem (distributions of independent variables in 29 The scale was originally a balanced, six-point scale. However, very few respondents selected the highest value (“I will definitely change my mind”), so I collapsed categories five (“very likely to change”) and six (“definitely change”) together. See Table B-1 for question wording. 232 Table 8-2). Because respondents were asked to self-report if they identify as “female” or “male” I refer to respondents by their gender (rather than as women or men). Respondent gender is a dummy variable indicating whether the respondent is female (coded female=1).30 For clarity, I refer to the “Jessica” counterargument as the woman counterargument, and the “Michael” counterargument as the man counterargument (rather than female or male counterargument). The counterargument condition is a dummy variable indicating whether the respondent received a man’s counterargument (coded Michael counterargument=1). Self-esteem is measured using the Rosenberg Self-Esteem (RSE) Scale (M. Rosenberg 1965). The RSE Scale is the most widely used self-esteem scale in the social sciences (Gray-Little, Williams, and Hancock 1997). It is a highly reliable, and internally consistent unidimensional (single-factor), 10-item scale measuring global self-worth (see Table B-2 in Appendix B for item wording and coding). The average self-esteem score for the sample is 18.76 (on a scale that ranges from zero to 30). Consistent with existing research, I find that male respondents have significantly (p<0.01) higher self-esteem than females. The mean self-esteem score among males is 19.78, whereas among females it is 17.98 (see Table B-3 in Appendix B). Because I am using ordered logistic regression, I recoded the continuous RSE Scale variable into three roughly equal categories, creating a categorical self-esteem variable measuring low self-esteem, average self-esteem, and high self-esteem. 30 The binary gender variable (male or female) is always collected from the Subject Pool participants. In my survey, I also included a 7-point gender variable (completely male, somewhat male, a little male, neither male or female, a little female, somewhat female, and completely female). I ran my analysis with this continuous indicator out of curiosity, but (other than reducing the p-values marginally) it did not affect the findings. Although, of course, interpreting the marginal predicted probabilities for a seven-category independent variable (and a five-category dependent variable) becomes unwieldy. 233 Table 8-2 Distributions of independent variables Variable Obs Mean SD Min Max Female 380 0.57 0.50 0 1 Male counterargument 380 0.49 0.50 0 1 Self-esteem 380 2.00 0.83 1 3 Original position 380 0.68 0.47 0 1 I also control for respondents’ initial position on the policy question, since this is predicted to correlate with how persuasive respondents find the counterargument. Most respondents (68%) indicated that car insurance should be mandatory (the more easily justified position) and so faced a relatively weaker counterargument. Those who indicated that car insurance should be optional (the less easily justified position) faced a relatively stronger counterargument. Ideally, I could have picked a debate where both sides have similar merit, but I wanted to choose a topic that was not salient or politicized. Perhaps unsurprisingly, it is difficult to think of a policy represented by two sides of a debate, where both sides have roughly equal merit, characterized little or no debate. That this debate is one-sided should not matter for testing my theory, if I find that there is a difference in how subjects respond to a man’s or woman’s counterargument, controlling for the effect of subjects’ initial position. 8.3 Findings I first look for evidence of gendered asymmetries a communicative influence. A difference of means test (t-test) comparing the average responses to the two counterarguments reveals that, ceteris paribus, participants who received Michael’s counterargument (the man 234 counterargument) are significantly more willing to change their minds than participants who received Jessica’s counterargument (the woman counterargument), as indicated by the higher mean score on the dependent variable which ranges from one, ‘I will definitely NOT change my mind,’ to five, ‘I will likely/definitely change my mind’ (Table B-3). Table 8-3 Mean likelihood of mind change, by counterargument gender condition Counterargument Condition Likelihood of Mind Change (Std. Err.) Frequency Jessica Counterargument 2.663 (0.076) 193 Michael Counterargument 2.931 (0.080) 187 Combined 2.795 (0.056) 380 Difference -0.267** (0.111) Standard errors in parentheses. Note: No assumption of equal variances. *** p<0.01, ** p<0.05, * p<0.1 The modest, but significant different of 0.27 points on the five-point scale offers evidence of gendered asymmetries of communicative influence: All else being equal, respondents are more willing to change their minds after hearing a man’s counterargument. I use ordered logistic regression to get insight into what is causing this gendered asymmetry of communicative influence. My regression results offer additional evidence that gendered asymmetries of communicative influence structure political argumentation: even controlling for 235 original position,31 respondents are significantly more resistant to a woman’s counterargument. Setting the other variables at their means, there is only an eight percent predicted probability a respondent indicates they will “Definitely NOT” change their mind after learning about a man’s counterargument. But there is an 11 percent predicted probability a respondent indicates they will “Definitely NOT” change their mind after learning about a woman’s identical counterargument. Phrased another way, respondents are more open to a man’s counterargument. There is a full 22 percent predicted probability a respondent is “Somewhat likely” to change their mind after hearing a man’s counterargument, but only a 17 percent probability a respondent is “Somewhat likely” to change their mind after hearing a woman’s identical counterargument. The Main Model’s marginal predicted probabilities (for all five outcomes of the dependent variable) are available in Appendix B (Table B-7). 31 The independent variables should not be correlated with the treatment (since this is an experiment), but I want to control for the fact that respondents who took the weaker initial position (can insurance should be optional) should be – and are – more open to changing their minds in general. That taking a weaker position is significantly correlated with willingness to change one’s mind in all the models is evidence that respondents were reading the survey experiment items. 236 Table 8-4 Ordered logistic regression coefficients and odds ratios for likelihood of changing one’s mind after receiving a counterargument (three models) Model 1 Baseline Model 2 Respondent Gender and Counterargument Gender Interaction Model 3 Respondent Gender, Counterargument, and Self-Esteem Interaction VARIABLES Logit coef Odds Logit coef Odds Logit coef Odds Insurance Mandatory -1.504*** 0.222*** -1.519*** 0.219*** -1.612*** 0.200*** (0.213) (0.0474) (0.214) (0.0468) (0.221) (0.0442) Self-Esteem -0.0961 0.908 -0.0937 0.911 (0.115) (0.104) (0.115) (0.105) Man Counterargument 0.367* 1.443* (0.189) (0.273) Respondent Gender (Female=1) 0.642*** 1.901*** (0.196) (0.372) Respondent Gender & Man Counterargument Interaction Male respondent, man counterargument 0.156 1.169 (0.294) (0.344) Female respondent, woman counterargument 0.470* 1.599* (0.269) (0.430) Female respondent, man counterargument 0.981*** 2.667*** (0.282) (0.752) Respondent Gender, Man Counterargument, and Self-Esteem Three-way Interaction Male R, woman counterarg., ave. self-esteem 0.363 1.438 (0.542) (0.780) Male R, woman counterarg., high self-esteem -0.284 0.753 237 Model 1 Baseline Model 2 Respondent Gender and Counterargument Gender Interaction Model 3 Respondent Gender, Counterargument, and Self-Esteem Interaction VARIABLES Logit coef Odds Logit coef Logit coef Odds Logit coef (0.517) (0.389) Male R, man counterarg., low self-esteem -0.191 0.826 (0.582) (0.480) Male R, man counterarg., ave. self-esteem 0.0419 1.043 (0.515) (0.537) Male R, man counterarg., high self-esteem 0.462 1.588 (0.513) (0.814) Female R, woman counterarg., low self-esteem 0.645 1.906 (0.469) (0.894) Female R, woman counterarg., ave. self-esteem 0.502 1.652 (0.489) (0.807) Female R, woman counterarg., high self-esteem 0.284 1.328 (0.503) (0.668) Female R, man counterarg., low self-esteem 1.122** 3.072** (0.475) (1.460) Female R, man counterarg., ave. self-esteem 1.280** 3.598** (0.524) (1.884) Female R, man counterarg., high self-esteem 0.573 1.773 (0.525) (0.932) Observations 380 380 380 380 380 380 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 238 The differences (or “contrasts”) between the predictive margins comparing respondents who received a man’s counterargument and respondents who received a woman’s counterargument for all five outcomes of the dependent variable (with confidence intervals for each difference) are displayed in Figure 8-1. Recall the counterargument variable is a dummy variable (coded man counterargument=1), so a score of -0.03 in Figure 8-1 means that on average respondents are three percentage points less likely to indicate they will “Definitely NOT” accept a man’s counterargument, as compared to a woman’s. Figure 8-7 Contrasts of predictive margins of man and woman counterargument 239 To discern whether female respondents are particularly sensitive to hearing a man’s counterargument, I included an interaction between respondent gender and the gender of the speaker giving the counterargument (Table 8-3, Model 2). The coefficients reveal that the odds male respondents who receive a man’s counterargument changes their minds is not significantly different from the odds male respondents who received a woman’s identical counterargument changes their minds. The results in Table 8-3 (Model 2) also suggest that, compared to male respondents, female respondents respond more strongly to receiving a man’s counterclaim than a woman’s identical counterclaim. Calculating the marginal predicted probabilities more clearly illustrates this effect. Holding the variable constant at their means, there is a 19 percent predicted probability a female respondent is “Somewhat likely” to change their mind after hearing a woman’s counterargument, but there is a full 27 percent predicted probability a female respondent is “Somewhat likely” to change their mind after hearing a man’s identical counterargument – a difference of nearly seven percentage points (Figure 8-2). By contrast, there is a 15 percent predicted probability that a male respondent is “Somewhat likely” to change their mind after hearing a woman’s counterargument, and only a 17 percent predicted probability after hearing a man’s –difference of only two percentage points. Marginal predicted probabilities for the interaction model, for male and female respondents, for all five outcomes of the dependent variable, are available in Appendix B (Table B-8). 240 Figure 8-8 Difference between the man and woman counterarguments in the predicted probability of being “somewhat likely” to change mind, by respondent gender Finally, I hypothesize that gendered effects of self-esteem may act as a discursive inequity, influencing how people perceive men and women to produce gendered asymmetries in communicative influence that entail the internal exclusion of women speakers. Specifically, I expect to find while lower self-esteem women are more likely to accept a man’s counterargument, higher self-esteem women will be more impervious to men’s excess authority (among higher self-esteem women, there will be a smaller difference between counterargument 241 conditions). To test this, I ran a model that includes a three-way interaction between respondent gender, self-esteem, and the counterargument condition (Table 8-3, Model 3). The results from Table 8-3 (Model 3) help clarify what is driving women’s responses to men’s counterarguments. Female respondents with low and even average self-esteem are significantly more likely to change their minds after hearing a man’s counterargument (as compared to the baseline category of male respondents with low self-esteem who received a woman’s counterargument). Only female respondents who have higher than average levels of self-esteem do not significantly differ from male respondents in terms of how they react to a man’s or woman’s counterclaims. Estimating the marginal effects helps clarify that self-esteem moderates the way a man’s counterargument impacts the likelihood a woman changes her mind. There is a 22 percent predicted probability that a female respondent with low self-esteem is “Somewhat likely” to change their mind, after hearing a woman’s counterargument. But there is a full 28 percent predicted probability that a female respondent with low self-esteem is “Somewhat likely” to change their mind, after hearing a man’s identical counterargument – a difference of 6 percentage points (Figure 7-4). For female respondents with an average self-esteem score, the difference between hearing a woman’s, versus a man’s counterclaim is a full 10 percentage points (in the same direction). Finally, for female respondents with high self-esteem, the difference all but disappears: the difference between those who received a man’s counterargument as compared to a woman’s counterargument is only two percentage points. And although the coefficients (and odd ratios) for male respondents with different levels of self-confidence who received difference counterarguments in Table 8-3 (Model 3) do not 242 reach statistical significance, estimating the marginal effects offers some evidence of cross-gender effects among male respondents as well. Interestingly, both low self-esteem and average self-esteem male respondents responded similarly to a man’s and a woman’s counterargument, with a negligible preference for women speakers’ counterclaims. There is a 17 percent predicted probability a low self-esteem male is “Somewhat likely” to change their mind after hearing a woman’s counterargument, and a 15 percent predicted probability a low self-esteem male is “Somewhat likely” to change their mind after hearing a man’s identical counterargument, a difference of two percentage points in favour of the woman’s counterclaim (Figure 8-4). The difference among average self-esteem men is similar: a three percent point difference in the predicted probability they are “Somewhat likely” to change their minds, in the same direction (favouring the woman’s counterclaim). However, high self-esteem male respondents seem more resistant to a woman’s counterargument. There is only a 20 percent predicted probability a high self-esteem male is “Somewhat likely” to change their mind after hearing a woman’s counterargument, and a full 30 percent predicted probability a high self-esteem male is “Somewhat likely” to change their mind after hearing a man’s identical counterargument, a difference of 10 percentage points in favour of a man’s counterargument (Figure 8-4). The marginal predicted effects for the three-way interaction model, by low, average, and high self-esteem, for females and for males, for all five outcomes of the dependent variable are available in Appendix B (Table B-9, Table B-10, Table B-11). 243 Figure 8-9 Predicted probabilities for outcome “somewhat likely” to change mind, for man and woman counterargument, by respondent gender and self-esteem 8.4 Discussion My experiment reveals that, all else being equal, people who receive a man’s counterargument are more willing to change their minds than people who receive a woman’s identical counterargument. This effect varies by gender, as females seem especially sensitive to men’s discursive authority. This means that a man and woman can make the same claim, and, all 244 else being equal, the man will have greater discursive influence – particularly when he is speaking to women. This effect is partly explained by the way self-esteem’s gendered effects distribute asymmetries of communicative influence as discursive inequities. Female respondents with low and average self-esteem are significantly more likely to change their minds when they receive a man’s counterargument, as compared to similar males who received a man’s counterargument (but this gap does not exist among low and average self-esteem female and male respondents who received a woman’s identical counterargument). Comparing women, the predicted probability an average self-esteem female is “Somewhat likely” to change their mind is 10 percentage points higher after hearing a man’s counterclaim than after hearing a woman’s identical counterclaim; but this difference drops to only three percentage points among high self-esteem women. Comparing men, it appears that low and average self-esteem men are marginally more likely to change their minds after hearing a woman’s counterclaim; but the predicted probability that high self-esteem men are “Somewhat likely” to change their minds is 10 percentage points greater after hearing a man’s counterclaim. Despite these cross-gender effects, most males are still more resistant to a woman’s counterargument, and so are contributing to the internal exclusion of women speakers. Looking at the distribution of the self-esteem measure by gender illustrates that male respondents’ modal response (42 percent of males) is “high self-esteem,” whereas female respondents’ modal response (41 percent of females) is “low self-esteem” (Table 8-4). This means that at the most commonly occurring categories of self-esteem – high self-esteem among male respondents, and low self-esteem among female respondents – people are more open to men’s counterclaims, and 245 more resistant to women’s identical counterclaims. Gendered variation in self-esteem (and the gendered effects of this variation) distribute asymmetries of communicative influence as discursive inequities, contributing to the women’s internal exclusion, even when women are nominally included in political talk. Table 8-5 Self-esteem scores, by respondent gender Respondent Gender Male Female Total Low Self-Esteem 24% 41% 34% (40) (89) (129) Average Self-Esteem 33% 30.56% 32% (55) (66) (121) High Self-Esteem 42% 28% 35% (70) (61) (131) Total 100% 100% 100% (165) (216) (381) 8.5 Conclusion Clearly, group-based differences legal political participation rights – such as a restricted franchise – create external barriers that prevent certain groups from participating in politics, and engender asymmetries in political participation and influence. Group-based economic or resource gaps also constitute external barriers to political participation and influence (for instance, by excluding the poor from accessing candidates or influencing elections through donations). But even after formal prohibitions to political participation are removed, and efforts to address gross economic and resource gaps have been made through social welfare programs and legislation regulating campaign donations, the legacy of political exclusion can become 246 embedded in cognitive schemes of perception and appreciation as discursive inequities, which engender internal exclusion from communicative processes of opinion formation. Discursive inequities contribute to the internal exclusion of disempowered group members by influencing how people perceive speakers from social outgroups, and how they perceive themselves and other speakers from social ingroups. For instance, gender stereotypes characterized by an association between men and authoritativeness, and women and agreeability likely generate asymmetries in communicative influence. And as my research here indicates, gendered effects of self-esteem on a willingness to change one’s mind in light of men’s and women’s counterarguments also seems to generate asymmetries in communicative influence. Most people – particularly women with lower and average self-esteem, and men with higher self-esteem – are more likely to indicate they are willing to change their minds when a man explains a counterargument, as compared to when a woman explains an identical counterargument. Discursive inequities are disabling constraints that entail internal exclusion, undermining disempowered social group members’ influence in conversations about matters of collective concern. Women’s on-average lower self-esteem makes them more prone to self-doubt, and more susceptible to a man’s excessive discursive authority. Discursive inequities contribute to reinforcing existing patriarchal hierarchies, since women exert less influence in the discursive political practices required for changing oppressive social and political conditions. Because I deliberately chose the non-salient, non-partisan policy issue of car insurance, it may be less obvious why this is the case. But consider everyday discursive exchanges about matters of collective concern related to women’s empowerment, such as paid parental leave, public childcare subsidies, pay equity, women’s reproductive health, and other issues. Men and 247 women may have different opinions on these issues, particularly since redistributing empowerments to women mean men lose some of their privilege. For instance, women’s equal participation in the labour force and equal earnings would mean men in heterosexual partnerships would have to participate more equally in unpaid home and childcare, which means men would have to sacrifice some of their free time, or workplace earnings. Or at least, taxpayers would have to subsidize public childcare through taxation, and men – who currently earn more money on average than women, and so disproportionately feel a higher income tax burden – would lose disproportionately more of their wealth. In other words, many men have an interest in social policies that maintain the status quo (and their relative privilege), whereas many women have an interest in adopting social practices and policies to promote greater gender equality. If discursive inequities mean both men and women give less weight to women’s arguments, and women are especially quick to reconsider their opinions after hearing a man’s counterargument, then discursive inequities act as a conservative counterweight preventing discursive challenges to the status quo from receiving uptake. Discursive inequities contribute to system justification, where both the disempowered and empowered social group members contribute – often inadvertently – to maintaining existing patterns of structural inequality and oppression (Jost, Banaji, and Nosek 2004). My findings about how discursive inequities structure argumentation do not mean we give-up on discursive political practices any more than studies showing how formal inequalities structure women’s likelihood of running for office mean we give up on representative democracy. Power is always unavoidably at play when we use language, but as Jane Mansbridge (2015) argues, the closer we get to an absence of power, the better discursive political practices 248 are on normative grounds. My hope is that by identifying the discursive inequities that threaten discursive political practices, we might be able to think of new practices and institutional arrangements to address them. In Chapter 8 I describe practices and institutional arrangements to help address – to eliminate or at least neutralise – discursive inequities and norm-conditional communication. 249 Chapter 9: Achieving Discursive Equality The purpose of this chapter is to identify institutions and practices that promote equality – particularly discursive equality – to enable the inclusion of all social group members in self- and collective-rule. Because of my interest in achieving the democratic collective opinion and agenda formation functions, and the centrality of talk for achieving these functions, I pay special attention to discursive equality. Recall that by “discursive equality” I mean equal opportunities for participating in communicative processes, and equitable communicative influence. Talking is essential for self-development and self- and collective-rule. Talking is the tool for thinking about and bringing up or bringing to light unfreedoms, and the harms of oppression and domination. And talking is the best tool for sharing these thoughts, and bringing others on board so agents can form collective agendas oriented to achieving freer and more equal social systems. In the first section, I draw on James Tully’s (2002, 540) description of how social systems’ structures open “a diverse field of potential ways of thinking and acting in response” to those structures. These responses include practices in regular activity (playing by the “rules of the game”) inscribed in social systems’ structures, and challenging or questioning social systems’ structures using “the available procedures of negotiation, deliberation, problem solving, and reform with the aim of modifying” them (Tully 2002, 540). In this section I consider both macro-systemic and micro-institutional approaches for achieving or safeguarding discursive equality. Both regular activity and using available procedures of negotiation and reform require that agents have a degree of control over their actions and the conditions of their actions, and so 250 are not useful in social systems characterised by extreme domination (such as totalitarian systems). In the second section, I consider social systems characterised by more domination, where there are insufficient or no procedures for changing social systems’ structures. In closed systems of domination, agents’ only options are to go outside the available procedures, and change social systems’ structures through confrontation or exit (Tully 2002, see also 2008a, 2008b). 9.1 Structures and Practices that Engender Symmetrical Empowerments In this section, I suggest structures and practices that should engender more symmetrical empowerments in systems claiming to be democratic, where procedures – the rules of the game – create at least some opportunities for agents to determine their actions and the conditions of their actions. I focus on discursive equality, and engendering symmetrical opportunities for communicative participation and equitable communicative influence. My normative reflection is, of course, historically and socially contextualised. It reflects my experiences living in a liberal democracy in the late 20th and early 21st centuries. Like Young (2011), I recognise that I cannot speak for every social group members’ experiences. As a young, able-bodied, Canadian-born white woman, I cannot speak to the experiences of older people, physically disabled people, colonised peoples, immigrants and migrants, or people of colour, “but the political commitment to social justice which motivates my philosophical reflection tells me that I also cannot speak without them” (Young 2011, 14). My suggestions are oriented to achieving more egalitarian structural relations that engender 251 symmetrical empowerments across social groups. However, this is not meant to be a complete list, nor is this meant to be the final word. My suggestions are only a starting point in the ongoing conversation about what kinds of institutions and practices help engender or protect symmetrical empowerments, so agents can participate in practices that serve democratic functions. Achieving discursive equality requires attending to social systems’ more “objective” structures (laws and institutions) to promote symmetrical empowerments and external inclusions. Universal moral equality must be recognised in the legal framework defining the context of political and deliberative practices, so that those affected by discourses cannot be barred from participating in them. This constitutional or legal framework must enable empowerments by recognising both “the universal right to equal liberties,” as well as “participatory rights,” which include not only voting rights, but also the right to speak and associate (Habermas 1998, 548). Limiting the degree to which governments and groups can interfere in citizens’ lives should include not only constitutional guarantees against state intrusion into the lives of citizens (such limiting the degree to which governments can inhibit speech, political participation, or association), but should also include non-discrimination legislation, such as civil human rights legislation. And, of course, this should include procedures (such as human rights tribunals), for pursuing civil actors who engage in discriminatory practices. This guards against coercion and distributes symmetrical empowerments as formally equal opportunities for participation. This is the first step to promoting inclusive communicative practices that contribute to democratic public opinion and agenda formation. 252 Furthermore, within this “framework guaranteed by constitutional rights,” the mass media must effectively channel public communication through a pluralistic, open network of subcultural publics that develop “more or less spontaneously” within the public sphere (Habermas 1998, 307). In addition to legally recognising universal political liberties (such as the right to free speech and association) and democratic participatory rights (such as voting and protesting), democratic systems require a self-regulating media that is independent of market forces (Chambers 2009). Protecting mass media from market forces also requires positive efforts to make media more equitable, such as such as campaign finance reforms, or regulating political advertisements during and between elections (Bohman 1996). Other positive efforts to make media more equitable might include public subsidies for local and regional public television and radio, local or minority language programs, and internet and phone infrastructure in remote areas. As I started to explain, even when a legal framework protecting political freedoms and participation rights ensures social group members have equal opportunities to participate in political practices, and a self-regulating media channels communication through different publics, inequalities can produce asymmetries in social group members’ abilities to use these universal empowerments. For instance, structural inequalities engender external exclusions when the poor and other historically disempowered groups do not have the time or resources to participate in political practices at all. Inequalities also engender “internal exclusions” when disempowered groups are formally present in deliberation, but their utterances are given less weight, or are ignored (Young 2000). As I explained in Chapter 2, people do not arrive at life’s starting lines with the same capacities and resources. People vary in their facultative and 253 rhetorical abilities (Knight and Johnson 1997; S. W. Rosenberg 2007; Young 2000), and stereotypes, loaded metaphorical language, and aggressive conversational behaviours do disproportionate harm to disempowered social group members (Beauvais 2015; Mendelberg, Karpowitz, and Oliphant 2014; Young 2011). As I explained in Chapter 6, when structural inequalities entail external and internal exclusions that prevent historically disempowered group members from using their political rights or from influencing deliberative processes, communication can contribute to maintaining social hierarchies and harms of oppression. Distributing symmetrical empowerments that engender deliberative inclusion also requires pursuing the value of equity, and positive efforts to develop individual and collective capacities. This means that deliberation must be underwritten by enabling rights and practices that attend to social circumstances, and aim to ensure the participation and equitable influence of all social group members in the variegated, interlocking deliberative sites in democratic systems. Enabling rights include social programs designed to increase individual and collective capacities, including the provision of public services such public education, healthcare, childcare provisions, and redistributive efforts such as unemployment insurance or social security. Other enablements include legislation that actively promotes equity, such as affirmative hiring in the civil service, and guaranteed parental leave for both men and women. Enablements might also include some public funding for low-income complainants pursuing anti-discrimination cases through human rights tribunals, as well as for constitutional challenges through court of law. Social services, social welfare rights, positive legal efforts, and subsidies for anti-discrimination or constitutional challenges help mitigate the exclusionary consequences of socioeconomic 254 inequalities, which otherwise might deprive disadvantaged social group members from having the skills, resources, or time to participate in communicative practices, or keep them from problematizing discriminatory or unjust practices and social systems’ structures. As I mentioned, positive efforts can also increase the accessibility and responsiveness of public media to ensure more genuinely universal and equitable access to the means of communication (Bohman 1996). The state can also work with civil society actors to help ensure all social group members make use of their formal opportunities for participation, and enjoy equitable communicative influence by encouraging disempowered group members’ participation in spaces of public appearance. For instance, in the 1960s, the Canadian federal government began publicly funding minority language protection, multicultural, and women’s advocacy groups to try and achieve the policy goal of a more unified, harmonious country (Pal 1995). Public support for these organisations helped increased historically disempowered group members’ participation in public spaces: in their ability to participate in social identities (by engaging with other minority language speakers, minority group members or allies, and women), and fields of public spheres. This enriched Canada’s democratic communicative practices, and collective opinion and agenda formation. Furthermore, by maintaining links to minority language protection, multicultural, and women’s advocacy groups, the government created a communication channel between communicative agenda formation among groups typically underrepresented in electoral politics, and Canada’s federal legislative, decision-making body. Civil society actors also independently play an important role in encouraging social group members to use their formally equal participation rights, and in promoting equitable communicative influence. For instance, organisers of civic associations or public meetings may 255 use selective recruitment techniques to ensure disempowered social group members are singularly, or over-represented in deliberative venues in civil society, to help boost the discursive authority of disempowered group members in deliberative interactions. Consider Nancy Fraser’s (1990, 67) description of the “feminist subaltern counterpublic,” with its diverse array of “lecture series, research centers, academic programs, conferences, conventions, festivals, and local meeting places.” Precisely because these “parallel discursive arenas” are disproportionately comprised of women, and almost exclusively comprised of feminists, they empower feminist women to participate in and influence counterdiscourses, and to “formulate oppositional interpretations of their identities, interests, and needs,” which feed back into broader public discourses about women’s interests and roles in society. Most communicative interactions take place in the mass public. For instance, through mass media’s highly mediated and unidirectional communication flows, or civic associations’ and religious or cultural groups’ fields that create space for conversations about matters of collective concern and deliberation. And, most ubiquitously, in everyday conversations in day-to-day life, such as everyday chitchat with friends, neighbours, and colleagues. Many important deliberative interactions also take place in state institutions, including legislatures, the courts, the bureaucracy, and through the dialogue between state institutions and civil society. And, as such, most reforms – including the suggestions I have listed – are macro-structural efforts to promote equal opportunities to participate, and equitable influence in communicative processes in the mass public. The hope is that “once the established rules constitute the right game – one that promises the generation of considered public opinion – then even the powerful actors will only contribute to the mobilization of relevant issues, facts, and arguments” (Habermas 2001, 420). 256 9.1.1 Engineering Micro-Institutional Communicative Processes There is also a growing body of research considering how to design small-scale deliberative forums that are “symmetrical, face-to-face, and equal” (Chambers 2009, 339; see also M. Warren 2007), or what I refer to as “micro-institutional” processes. The drawback of these micro-institutional processes is that they typically focus on small-scale, face-to-face forums and institutions such as deliberative minipublics. Their main benefit, however, is that because they are designed, organisers can structure them to mitigate or neutralise existing inequalities – including discursive inequalities – that otherwise plague communication in the mass public. As such, micro-institutional processes may be especially useful for neutralising inequalities when known inequalities engender asymmetrical empowerments and the empowered/ disempowered are expected to have different preferences. For instance, consider the case of neighbourhood planning. The socioeconomic inequalities that allow some social group members to buy property and prevent others from doing the same tend to engender empowerments that benefit property owners. We can reasonably expect that homeowners will have more resources (time, money, and social status) to participate in and influence conversations about neighbourhood planning. This is particularly problematic considering homeowners and renters often have different interests with respect to planning for density and the availability of rental units, particularly with respect to planning for low-income, subsidised rental units. In this kind of situation, self-selected participation into conversations in the unstructured mass public may come to be dominated by empowered speakers, or may lead to 257 greater polarisation (Beauvais and Warren 2017; see also Mendelberg and Oleske 2000 for a related discussion of racially segregated neighbourhoods). In these cases, micro-institutional forums can help ensure renters are fairly represented, that their participation is subsidised (e.g., through childcare), and that they have fair speaking time. A growing body of empirical research on micro-institutional deliberation lends insight into different institutional design choices for promoting the twin values of equality and inclusion in micro-institutional forums. In this section, I review the empirical literature on institutional design choices related to participant selection methods, facilitation styles, communication formats, and decision rules. The requirement for face-to-face interaction imposes practical limitations on participation, and those designing deliberative events must decide who gets to participate and how they are selected. One of the most popular selection methods is self-selection, where deliberating groups are open to any interested participants (at least, until the deliberating group is at capacity). Self-selection promotes a kind of universal opportunity for participation that ostensibly ignores social circumstances.32 The primary benefits of self-selection include ease of implementation, and the kind of legitimacy that comes with allowing people to use their formal equalities to influence things they care about. Of course, self-selection is not always blind to social circumstances. Since historically empowered social group member are often more likely to 32 Of course, institutional design choices can and should be made with an awareness of context in which the deliberation takes place. Allowing participants to self-select into a deliberating body can result in enclave deliberation, which may be the intention of the organizers. For instance, if organizers wish to encourage deliberation among a disempowered minority group, they could invite residents from disproportionately minority neighbourhoods to self-select into a deliberative process. The result should be enclave deliberation among the disempowered. 258 participate in political practices, self-selection can produce more homogeneous groups that reflect these inequalities (Urbinati and Warren 2008; M. E. Warren 2001). Another popular selection technique for promoting universal equality is random selection. However, rather than creating an open opportunity for participation, random selection promotes universal equality through principle of justice referred as “isegoria,” or the equal chance for every member of a population to have their voice heard (Dworkin 2000, 194–98). Furthermore, because random selection prevents empowered social group members and powerful organised interests from over-selecting into deliberations, random selection helps prevent asymmetrical empowerments from engendering exclusions in deliberative bodies. As compared to self-selection, random selection also tends to do a better job of ensuring a diverse range of voices are included, which – in addition to promoting inclusion – has epistemic benefits related to learning (Landemore 2013). One drawback of a purely random sample is that it still does not guarantee the representation of disempowered social group members, particularly when disempowered group members comprise a numerically small proportion of the population. This problem is exacerbated when populating small deliberating groups, since smaller random draws are, of course, less likely to be representative. Random stratified sampling – a random sample designed to ensure the proportionate (or over-representation) of certain group members – can be used to overcome these problems. Random stratified sampling retains the fairness embodied in the principle of isegoria, but is more attentive to social contexts. Because of these attributes, random stratified sampling is one of the preferred techniques for selecting participants for minipublics, including deliberative polls, 259 citizens’ assemblies, and citizen juries (Fung 2003; Grönlund, Bächtiger, and Setälä 2014; Smith 2009). For instance, random stratified sampling was used to ensure the representation of Indigenous participants in the Australian Citizens’ Parliament (J. Dryzek 2009), as well as in citizens’ assemblies in Canada (Beauvais and Warren 2017; M. E. Warren and Pearse 2008). Because Indigenous citizens comprise a small proportion of the population in both Australia and Canada, they likely would have been excluded in a purely random draw. By ensuring their representation in the citizens’ parliament and assemblies, organizers helped ensure that the resulting deliberations and collective agenda formation were attentive to these historically marginalised group members. However, as I discussed, even when disempowered social group members are formally represented in conversations, internal exclusions can undermine their discursive influence. Purposive sampling (or targeted recruitment), can be used to specifically recruit participants based on some social or sociodemographic criteria. Purposive sampling can be used to achieve a “threshold presence” of disempowered group members in mixed groups (Kymlicka 1995; J. J. Mansbridge 1981), which refers to when a minimum number of disempowered group members are included. Purposive sampling can also be used to promote enclave deliberation among the disempowered, which refers to when groups are mostly or entirely populated by disempowered social group members. And while enclave deliberation is sometimes treated with skepticism because of the concern it contributes to groupthink and polarisation (Sunstein 2000), this problem is mitigated if the enclave meets under deliberative conditions (Grönlund, Herne, and Setälä 2015). And although greater diversity in deliberation is generally associated with positive epistemic outcomes such as learning (Bohman 2006; Landemore 2013), enclave deliberation 260 among the disempowered can produce many of the same benefits as heterogeneous deliberation (Karpowitz, Raphael, and Hammond 2009). Another important institutional design choice for promoting equality and inclusion in micro-institutional settings is facilitation. Afsoun Afsahi’s (2017) research shows that in unstructured (non-facilitated) talk about difficult issues, empowered social group members (i.e., white participants and men) are much more likely to “divest” deliberative capital, where deliberative capital is defined in terms of the willingness to participate in reason-giving, respect, reflection on or incorporation of others’ views, sincerity, empathy, and productive dialogue. That is, in unstructured (non-facilitated) talk, empowered social group members are less likely to offer justifications, are more likely to show disrespect, and so on. Facilitators can have a large and significant effect on empowering social group members, reducing their divestments and increasing their investments in deliberative capital, which opens space for disempowered group members to talk and be heard. In unstructured communication, interactional inequalities engender symmetrical empowerments that contribute to disempowered group members’ internal exclusion (and empowered group members’ excessive participation and influence). But as Afsahi’s research shows, facilitators use techniques that can neutralise this inequality to engender symmetrical empowerments, and the fair inclusion of all speakers in micro-institutional processes. Because facilitation is useful for ensuring both that everyone can use the formal opportunities to speak, and promotes equity by ensuring that different sides of the debate are heard, facilitation is one of the most important techniques for ensuring participants’ internal inclusion (Landwehr 2014). Facilitation style also matters for promoting universal equality and 261 equity in deliberation. It is useful to think of facilitator style in terms of a three-category distinction of passive facilitators, who play a “turn-taking enforcer” role, moderate facilitators, who play a “designated driver” role (moving the conversation along, without adding new interpretations), and involved facilitators, who play a “quasi-participant” role (editorializing or interpreting the conversation) (Dillard 2013, 220). Passive facilitators may be most effective in simply ensuring everyone can use their formal opportunities to speak, while moderate and active facilitators may more effectively achieve equitable discourses that are attentive to disempowered group members, including those who may not be present. Moderate and active facilitators can engage in a kind of “discursive representation” (J. S. Dryzek and Niemeyer 2008) to speak for those who are not present. For instance, they might offer hypotheticals (“What would you reply if someone argued that because of p, q?”) to increase awareness for diverse views, and engender empathetic concern for those who are affected by the outcomes of collective opinions and decisions, even if they are not present in homogeneous deliberating groups (Landwehr 2014). Relatedly, communication format impacts equality and inclusion in micro-institutional settings. Specifically, the question of whether deliberation should involve more combative, argumentative discussion formats (dialectical inquiry and devil’s advocacy), or more consensual, supportive communication. There is a concern that debate-style, argumentative forms of deliberation may suppress marginalised voices and contribute to the internal exclusion of the disempowered. For instance, research shows there is a gender bias when assigning roles such as the devil’s advocate, as women’s reputations can be harmed when they challenge men (Sinclair and Kunda 2000). Summarizing their findings on gender inequality in deliberation, Karpowitz et 262 al. (Mendelberg, Karpowitz, and Oliphant 2014, 35) conclude “that egalitarian discussion rests not on adversarial but on supportive communication, which lifts women’s authority.” However, while argumentative communication formats carry the risk of marginalizing the disempowered, debate can promote learning. Debate can unravel inconsistencies, bring to light unconsidered facts, and push participants to provide more, and better, reasons for their positions (Manin 2005). Experimental research suggests both dialectical inquiry and devil’s advocacy can lead to a higher level of critical evaluation of assumptions and better quality recommendations than consensus decision-making (David M. Schweiger, William R. Sandberg, and Ragan 1986). When making institutional design decisions, practitioners should be clear on what goals they want to achieve, as well as be attentive to the context within which deliberation takes place. For instance, where disempowered group members’ status is lower (for instance, if they are the numerical minority in a mixed group), consensual communication that achieves the goal of equal influence may be more appropriate. But where disempowered group members’ status is higher (such as in enclave deliberation), more argumentative formats that achieve deliberation’s learning aims might be preferable (Beauvais and Bächtiger 2016). Finally, if deliberating groups are expected to reach a collective decision, the decision rule affects how participants interact with one another, with consequences for equality and inclusion. For instance, experimental research reveals that the wrong institutional rules can exacerbate women’s internal exclusion in deliberation (Karpowitz and Mendelberg 2014; Mendelberg, Karpowitz, and Oliphant 2014). For instance, under majority rule when there are few women – and women’s status is lowest – the balance of male interruptions can contribute to women’s internal exclusion, silencing women’s voices. This problem appears to be mitigated 263 when women’s authority is highest, such as when they comprise the majority under majority rule. 9.2 Circumventing Normal Procedures: Agonism and Escape In this section, I consider social systems characterised by a high degree of domination, where agents have little control over their actions or the conditions of their actions (Young 2011, 38). In such social systems, which Tully (Tully 2008a, 38) refers to as “closed systems of domination,” agents’ only options are to go outside the available procedures for negotiation and reform. In systems characterised by a high degree of domination, the primary way to change social systems’ features is through confrontation or escape (Tully 2002, see also 2008a, 2008b). In loosely democratic systems, where procedures for changing social systems exist but may be insufficient, confrontation and escape are typically not violent, but instead involve forms of “agonism” (Mouffe 2000). Agonistic practices involve making the implicit assumptions in the backdrop of the common-sense world explicit through discomforting practices to thrust them into the light of rational critique. Bringing the always implicit, taken-for-granted backdrop of shared intentionality into the foreground and forcing subjects to confront it does not require violence, and is probably most effective without violence. But these acts still require a disruption of everyday life, a form of agonism analogous to a politically motivated version of Garfinkel’s (1967) “breaching experiments.” Consider again the normative assumption that women who do not fit the “Madonna” characterisation – women who wear skimpy cloths or walk alone at night – are therefore objects 264 of male sexual gratification. Women face an unenviable choice: Participate in normative expectations and follow the impossible example set by the Mother of God (and suffer cultural imperialism, with consequences for their sexual identities), or flaunt the norms regulating their behaviour and dress, express or explore their sexualities, and face the risk that if they are sexually assaulted, they will likely face discursive inequities if they complain (and suffer powerlessness) or face marginalisation if they remain silence. A police officer’s comments that women “should avoid dressing like sluts in order not to be victimized” sparked a transnational movement called “SlutWalk,” which challenges rape culture (Rush 2011). And while this movement uses available procedures (the legal right to protest and demonstrate), it always involves confrontation – simply using the term “slut” is to elicit discomfort, and to bring the implicit Madonna/whore dichotomy to the foreground of rational critique, where it can be challenged. Confrontation can also effectively take place through passive resistance. Consider how Martin Luther King’s “passive resistance” signalled not only a refusal to participate in norms, institutions, and discourses that kept black Americans separate and unequal, but furthermore the act of not engaging in violence carried an illocutionary rejection of the association of black Americans with violence and social disintegration. As Hannah Arendt (2013, 201) says, passive resistance is “one of the most active and efficient ways of action ever devised, because it cannot be countered by fighting.” Passive resistance cheats the victor of their prize in the sense that by refusing to participate in oppressive institutions, practices, or discourses, participants are refusing to re-constitute the normative backdrop of shared intentionality that oppresses them. Although it is a form of “non-participation,” passive resistance is unlikely to bring the risk of marginalisation 265 because it is a deliberate political act, an active response to social systems’ oppressive structures, rather than a silent consequence of them. Escape is another response to social systems’ structures. Escape can mean physically fleeing, as in the case of refugees fleeing oppression and domination. But just as confrontation should not be conflated with violence, escape should not be conflated with physically absconding. Escape can involve constructing and participating in subaltern counterpublics in the midst of a majority culture’s normative framework. Examples of this include the commune of Christiania in the centre of Copenhagen, Denmark, where agents experiment with a communal, non-capitalist mode of production and exchange. In North America, Indigenous peoples’ living in capitalist societies still characterised by the vestiges of colonialism work to escape capitalist and colonial contexts by constructing and participating in “normative lifeways and resurgent practices” (Coulthard 2014, 179). And in democracies where public spheres have historically been dominated by white men, African Americans participate in the “black public sphere” (The Black Public Sphere Collective 1995; Squires 2002), and women participate in feminist subaltern counterpublics (Fraser 1990). And, although escaping to subaltern fields of social play and social identities can be interpreted as a form of non-participation (at least, non-participation in dominant, norm-conditional communication), it can help counter the risk of marginalisation (as well as counter the risk of cultural imperialism). This is because subaltern spaces of social play create alternative spaces for mutual recognition and understanding, so agents can engage in communicative processes of opinion and agenda formation that in egalitarian contexts characterised by symmetrical power relations. 266 Consider the example of how various sites of subaltern discourses linked Indigenous peoples across Canada, helping them launch the Idle No More movement. Idle No More is an instance of what Coulthard (214, p. 179) would describe as a set of “resurgent practices” – series of teach-ins, rallies, and protests – that aim not only to protect Indigenous rights and reduce inequality between Indigenous and settler subjects, but also to “provide an analysis of the interconnections of race, gender, sexuality, class and other identity constructions in ongoing oppression” (Idle No More, n.d.). In other words, to act as a vehicle challenging and modifying social systems’ language games, institutions, and spaces of public appearance. 9.3 Conclusion Both universal moral equality, which requires abstracting from social circumstances, and equity, which requires attending to social circumstances, are essential for distributing symmetrical empowerments that entail inclusion in political practices, including talk. Institutional arrangements that promote the twin values of equality include, firstly, a legal framework guaranteeing liberties and political rights to engender formally equal opportunities for participation. Democratic systems require a self-regulating media that is independent of market forces. Furthermore, positive efforts – including social welfare programs, and actively engaging disempowered group members in civil society – must be made to ensure disempowered social group members can use their opportunities for political and deliberative participation. A growing body of research also offers insight into the different ways equality’s twin values can be achieved when designing “micro-institutional” forums, such as deliberative mini-267 publics. Organisers designing these kinds of micro-institutional forums have different choices related to participant recruitment, facilitation style, communication style, and decision-making, which have different consequences for empowering participation and both external and internal inclusion. Practitioners organising micro-institutional forums should consider the goals of deliberation and the context within which the deliberation is taking place, with an eye to potential consequences or trade-offs of institutional design choice (Beauvais and Bächtiger 2016). Of course, we cannot ignore deliberation in the mass public, since it comprises the context within which discreet deliberative events take place, and can determine the success or failure of small-scale deliberative experiments. For instance, small-scale deliberative minipublics may produce excellent policy proposals, only to be ignored by the benighted mass public (Chambers 2009; M. E. Warren and Pearse 2008). Alternatively, the outcomes of micro-institutional deliberations could be ignored by decision-makers, or worse, used as legitimating devices for decision-makers’ pre-determined policies (Johnson 2015; Pateman 2012). Micro-institutional processes also risk falling capture to organised interests in civil society (Beauvais & Warren, 2016). This is particularly true if organised interests anticipate they will not like the outcomes of deliberation (Hendriks 2011). No matter how effectively we learn to design micro-institutional deliberative forums, talk in the mass public is the primary means by which individuals engage in self-development, and communicative link their preferences to collective agendas through processes of self- and collective-rule. 268 Chapter 10: Conclusion Among crowded artisans’ stalls at a local cultural festival, I noticed a familiar face: one of the local artists had participated in the Grandview-Woodland Citizens’ Assembly, a municipal engagement process for neighbourhood planning I had been involved with the year prior. When I asked her for her retrospective take on the process, the assembly member told me that, overall, she was happy with the citizens’ assembly. She said she enjoyed meeting new people, had learned a lot, and it felt good to have a say in the process. My dissertation starts from the normative precept that more democracy is always possible, and always better. When people like the Grandview-Woodland Citizens’ Assembly members participate in communicative practices that contribute to self- and collective-rule under egalitarian conditions, they build bonds of reciprocal interdependence by developing mutual respect and trust for one another. These kinds of communicative practices are essential for developing democracy-supporting social integration, by fostering cross-cutting, non-hierarchical social ties. The Grandview-Woodland process helped integrate assembly members into a somewhat artificial community – the Grandview-Woodland neighbourhood, a mostly bureaucratic designation comprised of several smaller subareas that residents normally think of as their neighbourhoods – by creating common understandings and shared intentions with other Grandview-Woodland residents. Participating in communicative practices under egalitarian conditions can also help people develop democracy-supporting personalities by strengthening speakers’ self-efficacy and self-esteem. Participants also learn new information, including about others’ experiences, 269 interests, and opinions. Not least of all, when participants develop and refine their opinions with others in these ways they can have more confidence in resulting collective opinions and wills. Continuing with my example, assembly members can have confidence in the citizens’ assembly’s proposals because they largely reflect the opinions they developed while talking things through with each other, and with other community members, stakeholders, and experts. 10.1 Limitations Much of my discussion assumes some effective equalities; it assumes some social systems’ features (institutions, fields of play, social identities) empower agents in some way to participate in practices that serve democratic functions. As such, my work may be most applicable to democratic, or at least ostensibly democratic contexts. Although I considered some practical responses to closed systems of domination, agonism and escape still presume a degree of empowerment. In tightly closed systems of domination such as North Korea, it is not clear whether or to what extent these suggestions can help change the oppressive system’s structures. My empirical study of gendered asymmetries in women’s communicative participation (Chapter 7) suffers from certain limitations, particularly with respect to operationalising visible minority status (or “race”). Although I found some small effects of ethnicity on measures of discursive politics – for instance, that visible minority men are less likely than non-visible minority men to participate in everyday political talk – I found little evidence of a double disadvantage for women of colour. However, I am hesitant to draw any strong conclusions about Canadians achieving equality between the ethnic majority groups (non-visible minorities) and 270 racialized minorities. Given the underrepresentation of visible minorities in the CES 2015 sample (13 percent) relative to the Canadian population (19 percent), and difficulties operationalising the ethnicity variable (due to the large proportion of Canadians who only identified themselves as “Canadian”), my non-findings may be a product of measurement error. Future studies should endeavour to better measure minority status, and would benefit from an oversample of Canadians who identify as visible minorities. Given that Indigenous Canadians have been excluded from formal political rights even longer than women have, and participate in formal politics at much lower rates than other ethno-national groups (Fournier and Loewen 2011), an oversample of Indigenous respondents would be especially interesting to gain insight into Indigenous engagement in more informal politics. Finally, my empirical study of asymmetries in women’s communicative influence (Chapter 8) suffers from two main methodological drawbacks. The first is sample size. Given the small-n, my analysis does not have much power. In future work, I will endeavour to increase my sample size. Second, there is the question of external validity. As I discussed in Chapter 8, one of the main concerns with survey experiments is that “the artificially clean environment of the survey question makes treatment easier to receive” (Gaines, Kuklinski, and Quirk 2007, 16; see also Barabas and Jerit 2010). However, as I explained, many of the factors my experimental design controls for may further advantage men. In real life, men possess additional attributes that are associated with authoritativeness: men tend to be taller (Hamstra 2014; Blaker et al. 2013; Egolf and Corder 1991), and men have louder and deeper voices (Strand 1999; Carli 2001). As such, the “pristine experimental setting” (Barabas and Jerit 2010, 238) of my experiment likely under exaggerates the effect of gender, making this a harder test of my hypotheses. 271 Another drawback related to external validity is that in my experiment, respondents merely articulate a policy preference, are presented with one counterargument, and are asked to indicate the likelihood they will reconsider their initial preference. Ideally, discussants should have more time to exchange arguments, and should be able to hear from anyone affected by deliberative outcomes (Chambers 1996; Habermas 1990, 1998). But many communicative exchanges in everyday talk are characterised by this kind of brief exchange. People hear unidirectional counterpoints to their opinions through the television or radio. People hear brief counterpoints to their opinions in passing conversations with neighbours, co-workers, or strangers in informal settings. However, my findings are less relevant for high quality deliberation, which might take place in small-scale, institutionally-bounded civic forums, where trained moderators and rules of discourse ensure higher quality communicative problem-solving (Beauvais and Bächtiger 2016; Beauvais 2017). Instead, my findings are relevant for the kind of everyday talk that takes place in the unbounded mass public. 10.2 Contributions Despite these relatively minor shortcomings, my dissertation makes several important contributions to democratic theory and political science. First, my work contributes to the problem-based, systems’ approach to democratic theory. As Warren (2017) argues, democratic systems must achieve three normatively necessary functions to count as democratic: empowered inclusion, collective opinion and agenda formation, and collective decision-making. Empowered inclusion is functionally and normatively prior to the other two functions, because the aims of 272 forming collective opinions and wills, and making collective decisions are undermined if those affected by collective endeavours are excluded from participating in them (Fung 2013; Goodin 2007; Young 2000). I elucidated this point by showing that a degree of equality helps distribute empowerments symmetrically, so that those affected by collected endeavours can communicatively form individual preferences into collective opinions and agendas. I also contributed to democratic theory by clarifying what kind of equality is needed for practices to achieve their democratic functions. “Equality,” broadly understood, can be thought of as symmetrical empowerments that enable people to participate in forming individual and collective judgments and decisions through political liberties (i.e., equality entails the absence of constraints), as well as through positive supports that develop individual and collective capacities, and ensure that all social group members can effectively make use of their formal, universal equalities. It requires both the values of universal equality and equity. Once a baseline of equality engenders symmetrical empowerments that entail inclusions, individual preferences need to be communicatively linked to collective judgements to link self-government with collective governance. I also expanded the system’s approach by considering how the practice of “talk” – including, but not limited to deliberation – can achieve democratic functions. Focusing on using talk to achieve democratic functions, I built-out the collective opinion and agenda formation function, identifying five interrelated, but distinguishable components comprising this general function – the social integration, reciprocity, personality, learning, and legitimation component functions – and described how talking helps achieve each component function. I also expanded 273 Warren’s (2017) discussion of generic tools that achieve democratic functions to include “talk” broadly understood as not only deliberation and everyday talk about matters of collective concern, but also mundane talk. I make an important contribution to democratic theory by showing that everyday communication – particularly mundane talk – is essential for building the kinds of relational ties and facilitative attitudes that allow speakers to withstand the strain of disagreement endemic to deliberation. My work also contributes to the study of democracy – and particularly to deliberative democracy, and empirical research on communicative political participation – by clarifying three problems that inequality poses for talk’s democratic potentials. First, inequality can simply prevent talking from achieving the collective opinion and agenda formation functions. Second, under conditions of inequality, talking can contribute to harms of oppression. And third, inequality can block deliberation. With respect to this last point, when asymmetrical empowerments block the “forceless force” of the better argument from solving communicative breakdowns (i.e., misunderstandings or disagreements), or impede efforts to problematize normative assumptions and bring them to the level of explicit disagreement in the first place, asymmetrical empowerments contribute to domination. Most democratic theories consider the problem of external exclusion, and “call for limiting the influence of wealth or position on the ability to participate in a democratic process” (Young 2000, 55). My work builds on Young’s (2000) description of “internal exclusions,” to showing how inequalities inscribed in social systems’ language games, institutions, fields of public appearance, and other cultural artefacts can structure agents’ socio-cognitive and motivational predispositions. Those who have historically been excluded from the franchise – 274 women, the poor, or non-whites – may internalise a sense of “feeling out of place” in politics, and so may be less likely to participate in practices such as voting or talking about matters of collective concern, even after formal prohibitions to their participation have been lifted. I introduce the concept of “discursive inequalities” to refer to cognitive schema that engender unjustifiable, systematic asymmetries of communicative participation or influence. Discursive inequalities are particularly problematic for talk. When inequality structures cognition, cognitive schemes of perception and appreciation can entail internal inequalities that prevent disempowered social group members from taking advantage of formally equal opportunities to participate in talk, and can generate unjustifiable asymmetries in disempowered social group member’s communicative influence. I identify two distinct species of discursive inequality: discursive inequities and norm-conditional communication. Discursive inequities are disabling constraints that entail internal exclusion, undermining disempowered social group members’ communicative influence when they do participate in talk. Social stereotypes and structural variation in self-esteem are two particularly problematic discursive inequities that undermine disempowered social group members communicative influence. Norm-conditional communication refers to when the terms of successful communication (norms regulating what makes an utterance successful, a claim persuasive, and so on) are harmful to certain social group members’ identities. Norm-conditional communication discourages disempowered social group members from participating in communication at all (produce unjustifiable asymmetries of communicative participation), and entails discursive inequities (and unjustifiable asymmetries of communicative influence) when disempowered social group members do participate in norm-conditional communication. Discursive inequalities 275 can reinforce existing systems of oppression, since the disempowered are less likely to influence or participate in communicative practices. My dissertation also offers insight into where discursive inequalities come from. Like many other animals, humans have the cognitive hardware enabling a neural experience of intersubjectivity (sharing automatic, sensorimotor understandings, and emotional contagion) (e.g., Gallese 2001; Gallese, Eagle, and Migone 2007; Gallese and Goldman 1998; Iacoboni 2009; Iacoboni et al. 2005; Kohler et al. 2002; Rizzolatti and Craighero 2004). Humans are both capable and motivated to go beyond this temporary, corporeal experience of intersubjectivity: people expand and give permanence to intersubjective understandings by inscribing them into language games and other social institutions, a process that enables more cognitively-oriented mind reading skills (attributing the mental states of others) and expansive collaborative activity (e.g., Ardila 2008; Burkart, Hrdy, and Van Schaik 2009; Carpenter 2006; Hrdy 2011; Tomasello 2010; Tomasello and Carpenter 2007). Not only do agents inscribe mutual understandings into language games and other institutions, they inscribe them in minds, in cognitive and motivational structures, internalising a practical comprehension of the world in which they have been inserted (Bourdieu 1977, 1990, 2000). Human plurality – that humans are simultaneous similar to one another, and yet distinct from every other person – means that humans are both capable of understanding one another, but also need to make themselves understood (Arendt 2013). Human plurality motivates two fundamental psychological needs: belonging and similarity (on the one hand) and distinctiveness and differentiation (on the other) (Brewer 2003, 2011; Brewer and Pickett 1999; Leonardelli, Pickett, and Brewer 2010). The simultaneous needs for belonging and distinctiveness motivates 276 agents to create and participate in spaces of public appearance, including collective social identities and what Bourdieu (1977, 1984, 1990, 2000) describes as fields of social play. However, the desire for mutual understanding can motivate agents to forcibly co-opt others into participating in their social systems, to expand and give greater permanence to the shared intentionality inscribed in their social systems’ language games, institutions, and spaces of public appearance. This is particularly true under conditions of inequality and asymmetrical empowerments, when forcibly co-opting new participants reinforces or increases the co-opting group’s power. Forcible co-optation entails norm-conditional communication; it entails the co-opting group inducing another group to participate in the co-opting group’s language games, according to the terms inscribed in the co-opting groups’ social systems. Furthermore, because experiences in the social and material world structure semantic associations and personalities, structural inequalities often entail the association of disempowered group members with negative attributes that reflect their disempowerments (subordination, stupidity, weakness, impurity, etc.) (Cuddy, Fiske, and Glick 2008; Fiske et al. 1999; Macrae, Stangor, and Hewstone 1996; Operario and Fiske 2001). In other words, social and material inequalities become embedded in cognitive structures as discursive inequities, including harmful stereotypes and distorted personalities. These discursive inequalities act as disabling constraints preventing disempowered social group members from participating in language games and fields of social play, and from enacting social identities; that is, discursive inequalities block capacities for self-development (and contribute to oppression). Discursive inequalities also prevent disempowered social group members from having a say over social systems’ language games, spaces of public appearance, and other institutions that affect them; 277 that is, discursive inequalities block capacities for self- and collective-rule (and contribute to domination) My dissertation also offers empirical evidence that – although formal barriers to women’s political participation have been removed – there are ongoing asymmetries in women’s discursive participation and influence. In Chapter 7, I reviewed the literature on the gender gap in political participation and offered a novel analysis revealing an ongoing gender gap in group deliberation. In Chapter 8, I offered evidence from a novel vignette experiment to show that even when women do participate in political talk, discursive inequities (i.e., the gendered effects of women’s on-average lower self-esteem) can undermine women’s discursive influence. That women are less likely to participate in political talk, and have less influence when they do participate in political talk threatens the democratic values of self-development and self- and collective-rule. By offering evidence of ongoing gaps in communicative participation and influence, my empirical research does important political work by convincing skeptics that discursive inequality is an ongoing problem that needs to be addressed. Both universal equality and equity can – and must – be expanded and preserved to better achieve the self-development and self- and collective-rule. For instance, democratic systems must enable empowerments through a baseline of universal equality consecrate in law as constitutional or human rights. And in political systems where historical inequalities result in continuing asymmetries in social group members’ abilities to use these universal empowerments, it is necessary to pursue the value of equity to promote the empowered inclusion of historically disadvantaged group members, such as through needs-based social welfare redistributions. 278 These efforts to achieve or safeguard symmetrical empowerments help create a context in which talk can achieve its democratic opinion and agenda formation functions in everyday contexts, or what James Tully (2002) calls “normal” activity. This backdrop of empowerments even helps when ongoing structural inequalities (including discursive inequalities) prevent talk or other political practices from functioning democratically, because agents can use regular procedures of reform and deliberation to address disabling constraints and injustices (Tully 2002). For instance, agents can address discrimination in hiring practices by filing suit in a court of law or human rights tribunal, or can address problematic normative assumptions inscribed in language games or other cultural artifacts in social systems by making them explicit and challenging them through deliberation. When this backdrop of empowerments is absent – or when it does not go far enough to empower agents to address disabling constraints or ongoing injustices – agents can work outside the normal procedures, such as through agonistic practices of confrontation and protest, or through escape. 10.3 Future Research My future work will extend my dissertation in two main ways. First, I will consider the relationship between talking’s democratic reciprocity and social integration functions. Second, I want to consider gender dynamics in Indigenous agents’ agonistic responses to norm-conditional inclusion in colonial and post-colonial contexts. With respect to the first extension, I will draw from my dissertation work pertaining to talk’s democracy-supporting reciprocity functions to consider whether facilitative attitudes (particularly mutual respect and tolerance) must be met 279 both within and across salient ethnic group boundaries, so that everyday talk can contribute to democracy-supporting social integration in diversifying populations. It is not sufficient if, for instance, French citizens of North African descent feel that conversations with citizens of North African descent are characterised by norms of reciprocity, if they feel that conversations with white French citizens (of French descent) are not. Whether democratic reciprocity functions – what Dietlind Stolle and Allison Harrell (2014) call “facilitative values” – are achieved impacts the kinds of social ties people form. When democratic reciprocity functions are satisfied during intra-group (within group) communication, but are not satisfied during cross-cutting (across group) communication, this discourages cross-cutting social interactions, which is the relational measure of democratic social integration. The absence of cross-cutting social ties can contribute to societal fragmentation, which is problematic because societal fragmentation can prevent citizens from solving collective problems in democratically legitimate ways. The purpose of this proposed extension of my dissertation research is twofold. The first aim of this proposed project is to identify the extent to which democratic reciprocity functions are satisfied both within, and across group communication in Canada and France, and so diagnose a primary driver of societal fragmentation. The second aim of this proposed project is to prescribe practical solutions for improving inter-group communication (and facilitating cross-cutting, non-hierarchical social ties), to achieve the kind of social cohesion needed for increasingly diverse societies to function democratically. My second future research project draws on my dissertation’s discussion of norm-conditional communication, and will explore communicative practices in post-colonial contexts. Research suggests that Canada’s Indigenous peoples participate in conventional politics at much 280 lower rates (Fournier and Loewen 2011). But do existing conceptions of what constitutes “political practices” miss alternative or subaltern forms of engagement and contestation? With respect to Indigenous political practices in post-colonial contexts, I am particularly interested in looking at gender dynamics and agonistic practices. Since “argumentative” discussion formats seem to undermine women’s communicative influence, how does this impact Indigenous women’s opportunities to participate in and influence agonistic communicative and expressive practices in post-colonial contexts? Understanding how structural inequalities shape women’s communicative participation and influence, whether members of different cultural groups living in the same society talk to one another and develop social ties, and whether (and what) opportunities Indigenous persons (particularly Indigenous women) have for participating in social and political practices is important because democracy is defined by the degree to which people are empowered to participate in self-development and self- and collective-rule. These universal values are connected to the notion of justice derived from Habermas’s communicative ethics which stipulates that – for norms or social conditions to be just – everyone affect by them must, in principle, have “an effective voice” in their creation, and be able to agree to them without coercion (Young 2011, 38). Democracy is linked to this conception of justice because democratic conditions – conditions that promote self-development and collective-rule – can be expected to produce opinions, decisions, norms, and social conditions that are more attentive to the needs and preferences of those affected by them. Structural inequalities act as disabling constraints that prevent social group members from enacting social identities, and from participating in language games and spaces of social 281 play – in other words, structural inequalities contribute to harms of oppression. Structural inequalities also act as disabling constraints that prevent social group members from participating in self- and collective-rule – in other words, structural inequalities contribute to harms of domination. As a political practice, talking is especially important because it is the tool for thinking about and bringing to light inequality and unfreedoms, and the harms of oppression and domination that constitute injustices. Talking helps women in patriarchal societies think through and articulate “oppositional interpretations of their identities, interests, and needs,” which feed back into broader public discourses about women’s interests and roles in society (Fraser 1990, 67). Cross-cutting communication is how members of different cultural groups get to know one another, allowing them to form non-hierarchical, cross-cutting bonds of mutual trust and friendship – the source of democracy-supporting social integration in culturally diverse societies (Harell and Stolle 2010). And talking is the means by which Indigenous peoples create and preserve “normative lifeways and resurgent practices” in colonial/postcolonial contexts (Coulthard 2014, 179). The long-term importance of studying how talk enhances self-development and contributes to self- and collective-rule is important, because talk is the best tool for sharing thoughts about ongoing inequalities and unfreedoms, and for forming collective agendas oriented to achieving a more equal, free, autonomous, and just future. 282 References Afsahi, Afsoun. 2017. “Can We Talk?: Examining Willingness and Facilitating Deliberative Capital.” Doctoral thesis, University of British Columbia. Alexander, Garrett E., and Michael D. Crutcher. 1990. “Neural Representations of the Target (Goal) of Visually Guided Arm Movements in Three Motor Areas of the Monkey.” J Neurophysiol 64 (1): 164–178. Allport, Gordon. 1954. The Nature of Prejudice. Cambridge: Cambridge, MA: Addison-Wesley. Andersen, Kristi, and Elizabeth D. Miller. 1997. “Gender and Student Evaluations of Teaching.” PS: Political Science & Politics 30 (02): 216–219. Ardila, Alfredo. 2008. “On the Evolutionary Origins of Executive Functions.” Brain and Cognition 68 (1): 92–99. Arendt, Hannah. 2013. The Human Condition: Second Edition. Chicago: University of Chicago Press. Arneil, Barbara. 2006. Diverse Communities: The Problem with Social Capital. Cambridge: Cambridge University Press. Aronson, Joshua, Carrie B. Fried, and Catherine Good. 2002. “Reducing the Effects of Stereotype Threat on African American College Students by Shaping Theories of Intelligence.” Journal of Experimental Social Psychology 38 (2): 113–125. Axelrod, Robert. 1984. The Evolution of Cooperation. New York: New York: Basic Books. Baldwin, Tamara, and Nancy Blattner. 2003. “Guarding against Potential Bias in Student Evaluations: What Every Faculty Member Needs to Know.” College Teaching 51 (1): 27–32. 283 Barabas, Jason, and Jennifer Jerit. 2010. “Are Survey Experiments Externally Valid?” American Political Science Review 102 (2): 226–42. Bartels, Larry M. 2005. “Homer Gets a Tax Cut: Inequality and Public Policy in the American Mind.” Perspectives on Politics 3 (01): 15–31. Baumeister, Roy F., and Mark R. Leary. 1995. “The Need to Belong: Desire for Interpersonal Attachments as a Fundamental Human Motivation.” Psychological Bulletin 117 (3): 497. Beauvais, Edana. 2015. “Political Communication and the Use of Political Language.” Journal of Parliamentary and Political Law, no. Special Issue: The Informed Citizens Guide to Elections: Electioneering Based on Rule of Law. ———. 2017. “Deliberation and Equality.” In The Oxford Handbook of Deliberative Democracy, edited by Jane Mansbridge, Mark Warren, André Bächtiger, and John Dryzek. Oxford University Press. Beauvais, Edana, and André Bächtiger. 2016. “Taking the Goals of Deliberation Seriously: A Differentiated View on Equality and Equity in Deliberative Designs and Processes.” Journal of Public Deliberation, Special edition: Equality, Equity, and Deliberation, 12 (2). Beauvais, Edana, and Mark E. Warren. 2017. “Can Citizens’ Assemblies Deepen Urban Democracy?” Working Paper. Beauvais, Edana, and Sule Yaylaci. 2017a. “The Role of Conversational Dynamics and Facilitators in Promoting Empathy in Small-Group Deliberation.” Working Paper. ———. 2017b. “The Role of Social Group Membership on Classroom Participation.” PS: Political Science & Politics 50 (2). 284 Benhabib, Seyla. 1992. Situating the Self: Gender, Community, and Postmodernism in Contemporary Ethics. New York: Routledge. Bennett, Sheila K. 1982. “Student Perceptions of and Expectations for Male and Female Instructors: Evidence Relating to the Question of Gender Bias in Teaching Evaluation.” Journal of Educational Psychology 74 (2): 170. Blackwell, Debra L. 1998. “Marital Homogamy in the United States: The Influence of Individual and Paternal Education.” Social Science Research 27 (2): 159–188. Blackwell, Debra L., and Daniel T. Lichter. 2004. “Homogamy among Dating, Cohabiting, and Married Couples.” The Sociological Quarterly 45 (4): 719–737. Blaker, Nancy M., Irene Rompa, Inge H. Dessing, Anne F. Vriend, Channah Herschberg, and Mark Van Vugt. 2013. “The Height Leadership Advantage in Men and Women: Testing Evolutionary Psychology Predictions about the Perceptions of Tall Leaders.” Group Processes & Intergroup Relations 16 (1): 17–27. Bleidorn, Wiebke, Ruben C. Arslan, Jaap JA Denissen, Peter J. Rentfrow, Jochen E. Gebauer, Jeff Potter, and Samuel D. Gosling. 2015. “Age and Gender Differences in Self-Esteem—A Cross-Cultural Window.” Journal of Personality and Social Psychology 111 (3): 396–410. Bohman, James. 1996. “Deliberative Inequalities.” In Public Deliberation: Pluralism, Complexity, and Democracy. Cambridge: The MIT Press. ———. 2000a. “Distorted Communication: Formal Pragmatics as a Critical Theory.” In Perspectives on Habermas, edited by Lewis Edwin Hahn, 3–20. Chicago: Open Court. 285 ———. 2000b. “‘When Water Chokes’: Ideology, Communication, and Practical Rationality.” Constellations 7 (3): 382–392. ———. 2006. “Deliberative Democracy and the Epistemic Benefits of Diversity.” Episteme 3 (03): 175–191. Bottger, Preston C. 1984. “Expertise and Air Time as Bases of Actual and Perceived Influence in Problem-Solving Groups.” Journal of Applied Psychology 69 (2): 214. Bourdieu, Pierre. 1977. Outline of a Theory of Practice. Vol. 16. Cambridge: Cambridge university press. ———. 1984. Distinction: A Social Critique of the Judgement of Taste. Boston: Harvard University Press. ———. 1990. The Logic of Practice. Stanford University Press. ———. 2000. Pascalian Meditations. Stanford: Stanford University Press. Bourdieu, Pierre, and Jean-Claude Passeron. 1970. “Foundations of a Theory of Symbolic Violence.” In Reproduction in Education, Society and Culture, 1–69. London, UK: Sage Publishing. Brady, Henry E., Sidney Verba, and Kay Lehman Schlozman. 1995. “Beyond SES: A Resource Model of Political Participation.” American Political Science Review 89 (02): 271–294. Brewer, Marilynn B. 1991. “The Social Self: On Being the Same and Different at the Same Time.” Personality and Social Psychology Bulletin 17 (5): 475–482. ———. 2003. “Optimal Distinctiveness, Social Identity, and the Self.” In Handbook of Self and Identity, edited by Mark R. Leary and June P Tangney, 480–91. New York: Guilford Press. 286 ———. 2011. “Optimal Distinctiveness Theory: Its History and Development.” In Handbook of Theories of Social Psychology: Volume Two, edited by Paul A M Van Lange, Arie W. Kruglanski, and E. Tory Higgins, 2:81–98. London: Sage Publications. Brewer, Marilynn B., and Samuel L. Gaertner. 2001. “Toward Reduction of Prejudice: Intergroup Contact and Social Categorization.” In Blackwell Handbook of Social Psychology: Intergroup Processes, edited by Rupert Brown and Samuel L. Gaertner, 451–472. Malden: Blackwell. Brewer, Marilynn B., Jorge M. Manzi, and John S. Shaw. 1993. “In-Group Identification as a Function of Depersonalization, Distinctiveness, and Status.” Psychological Science 4 (2): 88–92. Brewer, Marilynn B., and Cynthia L. Pickett. 1999. “Distinctiveness Motives as a Source of the Social Self.” In The Psychology of the Social Self, edited by R. Tyler Tom, Roderick M Kramer, and P. John Oliver, 71–87. New York: Psychology Press. Brown, Rupert, and Miles Hewstone. 2005. “An Integrative Theory of Intergroup Contact.” Advances in Experimental Social Psychology 37 (37): 255–343. Brown, Ryan P., and Elizabeth C. Pinel. 2003. “Stigma on My Mind: Individual Differences in the Experience of Stereotype Threat.” Journal of Experimental Social Psychology 39 (6): 626–633. Browne, Lewis. 1961. The World’s Great Scriptures: An Anthology of the Sacred Books of the Ten Principal Religions. London: Macmillan. 287 Burkart, Judith M., Sarah B. Hrdy, and Carel P. Van Schaik. 2009. “Cooperative Breeding and Human Cognitive Evolution.” Evolutionary Anthropology: Issues, News, and Reviews 18 (5): 175–186. Carli, Linda L. 2001. “Gender and Social Influence.” Journal of Social Issues 57 (4): 725–741. Carlston, Donal E., and Robert S Wyder, eds. 1994. Associated Systems Theory: A Systematic Approach to Cognitive Representations of Persons. Hillsdale: Lawrence Erlbaum Associates, Inc. Carpenter, Malinda. 2006. “Instrumental, Social, and Shared Goals and Intentions in Imitation.” In Imitation and the Development of the Social Mind: Lessons from Typical Development and Autism, edited by Sally J Rogers and Justin H G Williams, 48–70. New York: Guilford Press. Chambers, Simone. 1996. Reasonable Democracy: Jurgen Habermas and the Politics of Discourse. Cambridge: Cambridge University Press. ———. 2009. “Rhetoric and the Public Sphere: Has Deliberative Democracy Abandoned Mass Democracy?” Political Theory 37 (3): 323–50. Chen, Mark, and John A. Bargh. 1997. “Nonconscious Behavioral Confirmation Processes: The Self-Fulfilling Consequences of Automatic Stereotype Activation.” Journal of Experimental Social Psychology 33 (5): 541–560. Chong, Dennis, and James N. Druckman. 2007. “A Theory of Framing and Opinion Formation in Competitive Elite Environments.” Journal of Communication 57 (1): 99–118. Christ, Oliver, Katharina Schmid, Simon Lolliot, Hermann Swart, Dietlind Stolle, Nicole Tausch, Ananthi Al Ramiah, Ulrich Wagner, Steven Vertovec, and Miles Hewstone. 288 2014. “Contextual Effect of Positive Intergroup Contact on Outgroup Prejudice.” Proceedings of the National Academy of Sciences 111 (11): 3996–4000. Codol, J-P. 1984. “Social Differentiation and Nondifferentiation.” In The Social Dimension, edited by Henri Tajfel. Cambridge: Cambridge University Press. Coffé, Hilde, and Catherine Bolzendahl. 2010. “Same Game, Different Rules? Gender Differences in Political Participation.” Sex Roles 62 (5–6): 318–333. Coulthard, G. 2014. Red Skin, White Masks. Rejecting the colonial politics of recognition. Minneapolis: University of Minnesota Press. Crutcher, Michael D., and Garrett E. Alexander. 1990. “Movement-Related Neuronal Activity Selectively Coding Either Direction or Muscle Pattern in Three Motor Areas of the Monkey.” J Neurophysiol 64 (1): 151–163. Cuddy, Amy JC, Susan T. Fiske, and Peter Glick. 2007. “The BIAS Map: Behaviors from Intergroup Affect and Stereotypes.” Journal of Personality and Social Psychology 92 (4): 631. ———. 2008. “Warmth and Competence as Universal Dimensions of Social Perception: The Stereotype Content Model and the BIAS Map.” Advances in Experimental Social Psychology 40: 61–149. Dalton, Russell J. 2008. “Citizenship Norms and the Expansion of Political Participation.” Political Studies 56 (1): 76–98. David Jary, and Julia Jary. 1995. “Interaction.” Collins Dictionary of Sociology, Second Edition. Glasgow: Harper Collins. 289 David L. Hamilton, and Tina K. Trolier. 1986. “Stereotypes and Stereotyping: An Overview of the Cognitive Approach.” In Prejudice, Discrimination, and Racism: Historical Trends and Contemporary Approaches., edited by JF Dovidio and Samuel L. Gaertner, 127–63. Academic Press. David M. Schweiger, William R. Sandberg, and James W. Ragan. 1986. “Group Approaches for Improving Strategic Decision Making: A Comparative Analysis of Dialectical Inquiry, Devil’s Advocacy, and Consensus.” The Academy of Management Journal 29 (1): 51–71. Davis, Mark H. 1994. Empathy: A Social Psychological Approach. Boulder, CO: Westview Press. De Waal, Frans B. M. 2008. “Putting the Altruism Back into Altruism: The Evolution of Empathy.” Annual Review of Psychology 59 (1): 279–300. De Waal, Franz B. 2008. “Putting the Altruism Back into Altruism: The Evolution of Empathy.” Annual Review of Psychology 59 (1): 279–300. Delli Carpini, Michael X. Delli, Fay Lomax Cook, and Lawrence R. Jacobs. 2004. “Public Deliberation, Discursive Participation, and Citizen Engagement: A Review of the Empirical Literature.” Annu. Rev. Polit. Sci. 7: 315–344. Delli Carpini, Michael X. Delli, and Scott Keeter. 1997. What Americans Know about Politics and Why It Matters. New Haven: Yale University Press. Derrida, Jacques. 1998. Monolingualism of the Other, Or, The Prosthesis of Origin. Stanford: Stanford University Press. Devine, Patricia G. 1989. “Stereotypes and Prejudice: Their Automatic and Controlled Components.” Journal of Personality and Social Psychology 56 (1): 5. 290 Di Pellegrino, Giuseppe, Luciano Fadiga, Leonardo Fogassi, Vittorio Gallese, and Giacomo Rizzolatti. 1992. “Understanding Motor Events: A Neurophysiological Study.” Experimental Brain Research 91 (1): 176–180. Dillard, Kara N. 2013. “Envisioning the Role of Facilitation in Public Deliberation.” Journal of Applied Communication Research 41 (3): 217–235. Driver, Janice L., and John M. Gottman. 2004. “Daily Marital Interactions and Positive Affect during Marital Conflict among Newlywed Couples.” Family Process 43 (3): 301–314. Druckman, James N., and Kjersten R. Nelson. 2003. “Framing and Deliberation: How Citizens’ Conversations Limit Elite Influence.” American Journal of Political Science 47 (4): 729–745. Dryzek, John. 2009. “The Australian Citizens’ Parliament: A World First.” Journal of Public Deliberation 5 (1): 1–9. Dryzek, John S., and Simon Niemeyer. 2008. “Discursive Representation.” American Political Science Review 102 (04): 481–493. Dworkin, Ronald. 2000. Sovereign Virtue: The Theory and Practice of Equality. Cambridge: Harvard University Press. Egolf, Donald B., and Lloyd E. Corder. 1991. “Height Differences of Low and High Job Status, Female and Male Corporate Employees.” Sex Roles 24 (5–6): 365–373. Elections Canada. 2016. “Estimation of Voter Turnout by Age Group and Gender at the 2015 Federal General Election.” Online. Fanon, Frantz. 2007. The Wretched of the Earth. New York: Grove Press. 291 Finlay, Krystina A., and Walter G. Stephan. 2000. “Improving Intergroup Relations: The Effects of Empathy on Racial Attitudes.” Journal of Applied Social Psychology 30 (8): 1720–37. Fishkin, James S. 1997. The Voice of the People: Public Opinion and Democracy. New Haven: Yale University Press. Fiske, Susan T., Amy JC Cuddy, and Peter Glick. 2007. “Universal Dimensions of Social Cognition: Warmth and Competence.” Trends in Cognitive Sciences 11 (2): 77–83. Fiske, Susan T., Amy JC Cuddy, Peter Glick, and Jun Xu. 2002. “A Model of (Often Mixed) Stereotype Content: Competence and Warmth Respectively Follow from Perceived Status and Competition.” Journal of Personality and Social Psychology 82 (6): 878. Fiske, Susan T., Juan Xu, Amy C. Cuddy, and Peter Glick. 1999. “(Dis) Respecting versus (Dis) Liking: Status and Interdependence Predict Ambivalent Stereotypes of Competence and Warmth.” Journal of Social Issues 55 (3): 473–489. Fogassi, Leonardo, Pier Francesco Ferrari, Benno Gesierich, Stefano Rozzi, Fabian Chersi, and Giacomo Rizzolatti. 2005. “Parietal Lobe: From Action Organization to Intention Understanding.” Science 308 (5722): 662–667. Fournier, Patrick, and Peter Loewen. 2011. “Aboriginal Electoral Participation in Canada.” Elections Canada Report. Fox, Richard L., and Jennifer L. Lawless. 2011. “Gendered Perceptions and Political Candidacies: A Central Barrier to Women’s Equality in Electoral Politics.” American Journal of Political Science 55 (1): 59–73. ———. 2014. “Uncovering the Origins of the Gender Gap in Political Ambition.” American Political Science Review 108 (03): 499–519. 292 Fraser, Nancy. 1990. “Rethinking the Public Sphere: A Contribution to the Critique of Actually Existing Democracy.” Social Text, no. 25/26: 56–80. Fung, Archon. 2003. “Survey Article: Recipes for Public Spheres: Eight Institutional Design Choices and Their Consequences.” Journal of Political Philosophy 11 (3): 338–367. ———. 2013. “The Principle of All-Affected Interests: An Interpretation and Defense.” In Representations: Elections and Beyond, edited by Jack H. Nagel and Rogers M. Smith, 236–68. University Park, PA: Pensylvania State University Press. Gaertner, Samuel L., John F. Dovidio, Phyllis A. Anastasio, Betty A. Bachman, and Mary C. Rust. 1993. “The Common Ingroup Identity Model: Recategorization and the Reduction of Intergroup Bias.” European Review of Social Psychology 4 (1): 1–26. Gaertner, Samuel L., John F. Dovidio, and Betty A. Bachman. 1996. “Revisiting the Contact Hypothesis: The Induction of a Common Ingroup Identity.” International Journal of Intercultural Relations 20 (3): 271–290. Gaines, Brian J., James H. Kuklinski, and Paul J. Quirk. 2007. “The Logic of the Survey Experiment Reexamined.” Political Analysis 15 (1): 1–20. Galinsky, Adam D., and Gordon B. Moskowitz. 2000. “Perspective-Taking: Decreasing Stereotype Expression, Stereotype Accessibility, and in-Group Favoritism.” Journal of Personality and Social Psychology 78 (4): 708. Gallese, Vittorio. 2001. “The Shared Manifold Hypothesis: From Mirror Neurons to Empathy.” Journal of Consciousness Studies 8 (5–7): 5–7. 293 ———. 2003. “The Manifold Nature of Interpersonal Relations: The Quest for a Common Mechanism.” Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences 358 (1431): 517–528. Gallese, Vittorio, Morris N. Eagle, and Paolo Migone. 2007. “Intentional Attunement: Mirror Neurons and the Neural Underpinnings of Interpersonal Relations.” Journal of the American Psychoanalytic Association 55 (1): 131–175. Gallese, Vittorio, Luciano Fadiga, Leonardo Fogassi, and Giacomo Rizzolatti. 1996. “Action Recognition in the Premotor Cortex.” Brain 119 (2): 593–609. Gallese, Vittorio, Pier Francesco Ferrari, and Maria Alessandra Umiltà. 2002. “The Mirror Matching System: A Shared Manifold for Intersubjectivity.” Behavioral and Brain Sciences 25 (01): 35–36. Gallese, Vittorio, and Alvin Goldman. 1998. “Mirror Neurons and the Simulation Theory of Mind-Reading.” Trends in Cognitive Sciences 2 (12): 493–501. Garfinkel, Harold. 1967. Studies in Ethnomethodology. New York: Wiley. Gibson, James L. 1981. “Personality and Elite Political Behavior: The Influence of Self Esteem on Judicial Decision Making.” The Journal of Politics 43 (01): 104–125. Giddens, Anthony. 1984. The Constitution of Society: Outline of the Theory of Structure. Berkeley: University of California Press. Gidengil, Elisabeth. 2007. “Beyond the Gender Gap: Presidential Address to the Canadian Political Science Association, Saskatoon, 2007.” Canadian Journal of Political Science 40 (04): 815–831. 294 Gidengil, Elisabeth, André Blais, Richard Nadeau, and Neil Nevitte. 2003. “Women to the Left? Gender Differences in Political Beliefs and Policy Preferences.” In Women and Electoral Politics in Canada, edited by Manon Tremblay and Linda Trimble, 140–159. Don Mills, ON: Oxford University Press Canada. Gidengil, Elisabeth, Janine Giles, and Melanee Thomas. 2008. “The Gender Gap in Self-Perceived Understanding of Politics in Canada and the United States.” Politics & Gender 4 (04): 535–561. Gidengil, Elisabeth, Elizabeth Goodyear-Grant, Neil Nevitte, and André Blais. 2006. “Gender, Knowledge, and Social Capital.” In Gender and Social Capital, edited by Brenda O’Neill and Elisabeth Gidengil, 241–72. New York, USA: Routledge. Gidengil, Elisabeth, Brenda O’Neill, and Lisa Young. 2010. “Her Mother’s Daughter? The Influence of Childhood Socialization on Women’s Political Engagement.” Journal of Women, Politics & Policy 31 (4): 334–355. Gidengil, Elisabeth, and Dietlind Stolle. 2009. “The Role of Social Networks in Immigrant Women’s Political Incorporation.” International Migration Review 43 (4): 727–763. Goffman, Erving. 1967. Interaction Ritual: Essays in Face to Face Behavior. New York: Pantheon Books. ———. 2009. Stigma: Notes on the Management of Spoiled Identity. New York: Simon and Schuster. Gonzales, Patricia M., Hart Blanton, and Kevin J. Williams. 2002. “The Effects of Stereotype Threat and Double-Minority Status on the Test Performance of Latino Women.” Personality and Social Psychology Bulletin 28 (5): 659–670. 295 Goodin, Robert E. 2007. “Enfranchising All Affected Interests, and Its Alternatives.” Philosophy & Public Affairs 40: 40–68. ———. 2008. Innovating Democracy: Democratic Theory and Practice After the Deliberative Turn: Democratic Theory and Practice After the Deliberative Turn. Oxford: Oxford University Press. Gottman, John M., and Janice L. Driver. 2005. “Dysfunctional Marital Conflict and Everyday Marital Interaction.” Journal of Divorce & Remarriage 43 (3–4): 63–77. Government of Canada. 1995. Employment Equity Act. Vol. c. 44. Gray-Little, Bernadette, Valerie SL Williams, and Timothy D. Hancock. 1997. “An Item Response Theory Analysis of the Rosenberg Self-Esteem Scale.” Personality and Social Psychology Bulletin 23 (5): 443–451. Grice, H. P. 1975. “Logic and Conversation.” In Syntax and Semantics, Vol. III, edited by John P. Kimball, 41–58. New York: Academic Press. Grönlund, Kimmo, André Bächtiger, and Maija Setälä. 2014. Deliberative Mini-Publics: Involving Citizens in the Democratic Process. Wivenhoe Park, UK: ECPR Press. Grönlund, Kimmo, Kaisa Herne, and Maija Setälä. 2015. “Does Enclave Deliberation Polarize Opinions?” Political Behavior 37 (4): 995–1020. Gutmann, Amy, and Dennis Thompson. 1996. Democracy and Disagreement. Cambridge, MA: Cambridge Univ Press. ———. 2004. Why Deliberative Democracy? Princeton: Princeton University Press. Habermas, Jürgen. 1984. The Theory of Communicative Action: Vol. 1. Reason and the Rationalization of Society. Translated by Thomas McCarthy. Boston: Beacon Press. 296 ———. 1990. Moral Consciousness and Communicative Action. Cambridge: MIT press. ———. 1998. Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy. Cambridge: MIT Press. ———. 2001. On the Pragmatics of Social Interaction: Preliminary Studies in the Theory of Communicative Action. Cambridge: MIT Press. Hamilton, W. D. 1964. “The Genetical Evolution of Social Behaviour.” Journal of Theoretical Biology 7 (1): 1–16. Hamstra, Melvyn RW. 2014. “‘Big’ Men: Male Leaders’ Height Positively Relates to Followers’ Perception of Charisma.” Personality and Individual Differences 56: 190–192. Harell, Allison, and Dietlind Stolle. 2010. “Reconciling Diversity and Community? Defining Social Cohesion in Developed Democracies.” In Social Cohesion : Contemporary Theoretical Perspectives on the Study of Social Cohesion and Social Capital, edited by Marc Hooghe, 8–43. Brussels: Royal Flemish Academy of Belgium for Science and the Arts. ———. 2015. “Diversity and Social Cohesion.” In Routledge International Handbook of Diversity Studies, edited by Steven Vertovec, 294–301. New York, NY: Routledge. Harper, Stephen. 2008. “Statement of Apology.” Presentation, November 3. https://www.aadnc-aandc.gc.ca/eng/1100100015644/1100100015649. Hastie, Reid, Steven Penrod, and Nancy Pennington. 1983. Inside the Jury. Union: The Lawbook Exchange, Ltd. Hendriks, Carolyn M. 2011. The Politics of Public Deliberation: Citizen Engagement and Interest Advocacy. London: Palgrave Macmillan. 297 Hinton, Perry Roy. 2000. Stereotypes, Cognition and Culture. New York: Psychology Press. Hogg, Michael A. 2006. “Social Identity Theory.” Contemporary Social Psychological Theories 13: 111–1369. Honneth, Axel. 1995. The Struggle for Recognition: The Moral Grammar of Social Conflicts. Cambridge: Polity. Hooghe, Marc, and Dietlind Stolle. 2003. Generating Social Capital: Civil Society and Institutions in Comparative Perspective. London: Palgrave Macmillan. Hornsey, Matthew J. 2008. “Social Identity Theory and Self-Categorization Theory: A Historical Review.” Social and Personality Psychology Compass 2 (1): 204–222. Hrdy, Sarah Blaffer. 2011. Mothers and Others. Boston: Harvard University Press. Huckfeldt, R. Robert, Paul E. Johnson, and John D. Sprague. 2004. Political Disagreement: The Survival of Diverse Opinions within Communication Networks. Cambridge: Cambridge University Press Cambridge. Huckfeldt, Robert, Paul Allen Beck, Russell J. Dalton, and Jeffrey Levine. 1995. “Political Environments, Cohesive Social Groups, and the Communication of Public Opinion.” American Journal of Political Science 39 (4): 1025–1054. Iacoboni, Marco. 2009. “Imitation, Empathy, and Mirror Neurons.” Annual Review of Psychology 60: 653–670. Iacoboni, Marco, Istvan Molnar-Szakacs, Vittorio Gallese, Giovanni Buccino, John C. Mazziotta, and Giacomo Rizzolatti. 2005. “Grasping the Intentions of Others with One’s Own Mirror Neuron System.” PLoS Biology 3 (3): e79. Idle No More. n.d. “Idle No More: The Story.” http://www.idlenomore.ca/story. 298 Inglehart, Ronald, and Pippa Norris. 2003. Rising Tide: Gender Equality and Cultural Change around the World. Cambridge: Cambridge University Press. Jacobs, Lawrence R., Fay Lomax Cook, and Michael X. Delli Carpini. 2009. Talking Together: Public Deliberation and Political Participation in America. Chicago: University of Chicago Press. Johnson, Genevieve Fuji. 2015. Democratic Illusion: Deliberative Democracy in Canadian Public Policy. Toronto: University of Toronto Press. Jost, John T., Mahzarin R. Banaji, and Brian A. Nosek. 2004. “A Decade of System Justification Theory: Accumulated Evidence of Conscious and Unconscious Bolstering of the Status Quo.” Political Psychology, 881–919. Kakei, Shinji, Donna S. Hoffman, and Peter L. Strick. 1999. “Muscle and Movement Representations in the Primary Motor Cortex.” Science 285 (5436): 2136–2139. ———. 2001. “Direction of Action Is Represented in the Ventral Premotor Cortex.” Nature Neuroscience 4 (10): 1020–1025. Kant, Immanuel. 2008. Groundwork for the Metaphysics of Morals. New Haven: Yale University Press. Karpowitz, Christopher F., and Tali Mendelberg. 2014. The Silent Sex: Gender, Deliberation, and Institutions. Princeton: Princeton University Press. Karpowitz, Christopher F., Chad Raphael, and Allen S. Hammond. 2009. “Deliberative Democracy and Inequality: Two Cheers for Enclave Deliberation among the Disempowered.” Politics & Society 37 (4): 576–615. 299 Kent, Nigel. 1995. “No Fault Benefits in British Columbia: Must Out-of-Province Auto Insurers Pay the Same as ICBC?” https://www.cwilson.com/services/18-resource-centre/361-no-fault-benefits-in-british-columbia-must-out-of-province-auto-insurers-pay-the-same-as-icbc.html. Kiefer, Amy K., and Denise Sekaquaptewa. 2007a. “Implicit Stereotypes and Women’s Math Performance: How Implicit Gender-Math Stereotypes Influence Women’s Susceptibility to Stereotype Threat.” Journal of Experimental Social Psychology 43 (5): 825–832. ———. 2007b. “Implicit Stereotypes, Gender Identification, and Math-Related Outcomes a Prospective Study of Female College Students.” Psychological Science 18 (1): 13–18. Kirchler, Erich, and James H. Davis. 1986. “The Influence of Member Status Differences and Task Type on Group Consensus and Member Position Change.” Journal of Personality and Social Psychology 51 (1): 83. Kling, Kristen C., Janet Shibley Hyde, Carolin J. Showers, and Brenda N. Buswell. 1999. “Gender Differences in Self-Esteem: A Meta-Analysis.” Psychological Bulletin 125 (4): 470. Knight, Jack, and James Johnson. 1997. “What Sort of Political Equality Does Deliberative Democracy Require?” In Deliberative Democracy: Essays on Reason and Politics, 279–319. Cambridge: The MIT Press. Knobloch, Katherine R., and John Gastil. 2015. “Civic (Re) Socialisation: The Educative Effects of Deliberative Participation.” Politics 35 (2): 183–200. 300 Kohler, Evelyne, Christian Keysers, M. Alessandra Umilta, Leonardo Fogassi, Vittorio Gallese, and Giacomo Rizzolatti. 2002. “Hearing Sounds, Understanding Actions: Action Representation in Mirror Neurons.” Science 297 (5582): 846–848. Kuklinski, James H., Paul J. Quirk, Jennifer Jerit, David Schwieder, and Robert F. Rich. 2000. “Misinformation and the Currency of Democratic Citizenship.” Journal of Politics 62 (3): 790–816. Kymlicka, Will. 1995. Multicultural Citizenship: A Liberal Theory of Minority Rights. Cambridge: Cambridge Univ Press. Lakoff, George. 1990. Women, Fire, and Dangerous Things: What Categories Reveal about the Mind. Cambridge: Cambridge Univ Press. Landemore, Hélène. 2013. “Deliberation, Cognitive Diversity, and Democratic Inclusiveness: An Epistemic Argument for the Random Selection of Representatives.” Synthese 190 (7): 1209–1231. Landwehr, Claudia. 2014. “Facilitating Deliberation: The Role of Impartial Intermediaries in Deliberative Mini-Publics.” In Deliberative Mini-Publics: Involving Citizens in the Democratic Process, edited by Kimmo Grönlund, André Bächtiger, and Maija Setälä, 77–92. Wivenhoe Park, UK: ECPR Press. Lazarsfeld, Paul Felix, Bernard Berelson, and Hazel Gaudet. 1948. The People’s Choice: How the Voter Makes Up His Mind in a Presidential Campaign. New York: Columbia University Press. Lemaine, Gerard. 1974. “Social Differentiation and Social Originality.” European Journal of Social Psychology 4 (1): 17–52. 301 Leonardelli, Geoffrey J., Cynthia L. Pickett, and Marilynn B. Brewer. 2010. “Optimal Distinctiveness Theory: A Framework for Social Identity, Social Cognition, and Intergroup Relations.” Advances in Experimental Social Psychology 43: 63–113. Lowndes, Vivien. 2000. “Women and Social Capital: A Comment on Hall’s ‘Social Capital in Britain.’” British Journal of Political Science 30 (03): 533–537. Luskin, Robert C., James S. Fishkin, and Roger Jowell. 2002. “Considered Opinions: Deliberative Polling in Britain.” British Journal of Political Science 32 (03): 455–487. Macrae, C. Neil, and Galen V. Bodenhausen. 2000. “Social Cognition: Thinking Categorically about Others.” Annual Review of Psychology 51 (1): 93–120. Macrae, C. Neil, Charles Stangor, and Miles Hewstone. 1996. Stereotypes and Stereotyping. New York: Guilford Press. MacRae Jr, Duncan. 1993. “Guidelines for Policy Discourse: Consensual versus Adversarial.” In The Argumentative Turn in Policy Analysis and Planning, edited by Frank Fischer and J Forester, 1–17. Durham: Duke University Press. Manin, Bernard. 2005. “Democratic Deliberation: Why We Should Promote Debate rather than Discussion.” In Paper Delivered at the Program in Ethics and Public Affairs Seminar, Princeton University. Vol. 13. Mansbridge, Jane. 1999. “Everyday Talk in the Deliberative System.” In Deliberative Politics: Essays on Democracy and Disagreement, edited by Stephen Macedo, 211–39. Oxford: Oxford University Press. 302 ———. 2015. “A Minimalist Definition of Deliberation.” In Deliberation and Development: Rethinking the Role of Voice and Collective Action in Unequal Societies, edited by Patrick Heller, 27–50. Washington, DC: World Bank Group. Mansbridge, Jane, James Bohman, Simone Chambers, Thomas Christiano, Archon Fung, John Parkinson, Dennis F. Thompson, and Mark E. Warren. 2012. “A Systemic Approach to Deliberative Democracy.” In Deliberative Systems: Deliberative Democracy at the Large Scale, edited by John Parkinson and Jane Mansbridge, 1–26. New York: Cambridge University Press. Mansbridge, Jane J. 1981. “Living with Conflict: Representation in the Theory of Adversary Democracy.” Ethics 91 (3): 466–476. Marschall, Melissa J., and Dietlind Stolle. 2004. “Race and the City: Neighborhood Context and the Development of Generalized Trust.” Political Behavior 26 (2): 125–153. Maslach, Christina. 1974. “Social and Personal Bases of Individuation.” Journal of Personality and Social Psychology 29 (3): 411. McCarthy, Thomas. 1992. Practical Discourse: On the Relation of Morality to Politics. McGuire, William J. 1968. “Personality and Susceptibility to Social Influence.” Handbook of Personality Theory and Research 2: 1130–1187. McPherson, Miller, Lynn Smith-Lovin, and James M. Cook. 2001. “Birds of a Feather: Homophily in Social Networks.” Annual Review of Sociology, 415–444. Mead, George Herbert. 1934. Mind, Self and Society. Chicago: University of Chicago Press. Meltzoff, Andrew N. 1988. “Infant Imitation after a 1-Week Delay: Long-Term Memory for Novel Acts and Multiple Stimuli.” Developmental Psychology 24 (4): 470. 303 Mendelberg, Tali. 2002. “The Deliberative Citizen: Theory and Evidence.” Political Decision Making, Deliberation and Participation 6 (1): 151–93. Mendelberg, Tali, Christopher F. Karpowitz, and J. Baxter Oliphant. 2014. “Gender Inequality in Deliberation: Unpacking the Black Box of Interaction.” Perspectives on Politics 12 (01): 18–44. Mendelberg, Tali, and John Oleske. 2000. “Race and Public Deliberation.” Political Communication 17 (2): 169–191. Mendez, Jeanette Morehouse, and Tracy Osborn. 2010. “Gender and the Perception of Knowledge in Political Discussion.” Political Research Quarterly 63 (2): 269–279. Morrell, Michael E. 2005. “Deliberation, Democratic Decision-Making and Internal Political Efficacy.” Political Behavior 27 (1): 49–69. ———. 2007. “Empathy and Democratic Education.” Public Affairs Quarterly 21 (4): 381–403. ———. 2010. Empathy and Democracy: Feeling, Thinking, and Deliberation. University Park, PA: The Pennsylvania State University Press. Mouffe, Chantal. 2000. The Democratic Paradox. London: Verso. Mutz, Diana C. 2002a. “Cross-Cutting Social Networks: Testing Democratic Theory in Practice.” American Political Science Review 96 (01): 111–26. ———. 2002b. “Cross-Cutting Social Networks: Testing Democratic Theory in Practice.” American Political Science Review 96 (01): 111–126. ———. 2006. Hearing the Other Side: Deliberative versus Participatory Democracy. Cambridge: Cambridge University Press. 304 Nagell, Katherine, Raquel S. Olguin, and Michael Tomasello. 1993. “Processes of Social Learning in the Tool Use of Chimpanzees (Pan Troglodyte) and Human Children (Homo Sapiens).” Journal of Comparative Psychology 107 (2): 174. Nawyn, Stephanie J., Linda Gjokaj, DeBrenna LaFa Agbényiga, and Breanne Grace. 2012. “Linguistic Isolation, Social Capital, and Immigrant Belonging.” Journal of Contemporary Ethnography 41 (3): 255–282. Oluo, Ijeoma. 2017. “The Heart of Whiteness: Ijeoma Oluo Interviews Rachel Dolezal, the White Woman Who Identifies as Black.” The Stranger. http://www.thestranger.com/features/2017/04/19/25082450/the-heart-of-whiteness-ijeoma-oluo-interviews-rachel-dolezal-the-white-woman-who-identifies-as-black. Operario, Don, and Susan T. Fiske. 2001. “Stereotypes: Content, Structures, Processes, and Context.” Blackwell Handbook of Social Psychology: Intergroup Processes 1: 22–44. Page, Benjamin I. 1996. Who Deliberates?: Mass Media in Modern Democracy. Cambridge Univ Press. Pal, Leslie A. 1995. Interests of State: The Politics of Language, Multiculturalism, and Feminism in Canada. Montreal: McGill-Queen’s Press - MQUP. Pateman, Carole. 1988. The Sexual Contract. Stanford: Stanford University Press. ———. 2012. “Participatory Democracy Revisited.” Perspectives on Politics 10 (01): 7–19. Paxton, Pamela, Sheri Kunovich, and Melanie M. Hughes. 2007. “Gender in Politics.” Annu. Rev. Sociol. 33: 263–284. 305 Perdue, Charles W., John F. Dovidio, Michael B. Gurtman, and Richard B. Tyler. 1990. “Us and Them: Social Categorization and the Process of Intergroup Bias.” Journal of Personality and Social Psychology 59 (3): 475. Pettigrew, Thomas F. 1979. “The Ultimate Attribution Error: Extending Allport’s Cognitive Analysis of Prejudice.” Personality and Social Psychology Bulletin 5 (4): 461–476. ———. 1998. “Intergroup Contact Theory.” Annual Review of Psychology 49 (1): 65–85. Preston, Stephanie D., and Frans De Waal. 2002. “Empathy: Its Ultimate and Proximate Bases.” Behavioral and Brain Sciences 25 (01): 1–20. Putnam, Robert D. 1993. Making Democracy Work. Princeton, NJ: Princeton University Press. ———. 1995a. “Bowling Alone: America’s Declining Social Capital.” Journal of Democracy 6 (1): 65–78. ———. 1995b. “Bowling Alone: Democracy at the End of the Twentieth Century.” presented at the Nobel Symposium, Uppsala, Sweden. ———. 2000. Bowling Alone: The Collapse and Revival of American Community. New York: Simon and Schuster. ———. 2002. Democracies in Flux: The Evolution of Social Capital in Contemporary Society. Oxford: Oxford University Press. Rakoczy, Hannes, Felix Warneken, and Michael Tomasello. 2008. “The Sources of Normativity: Young Children’s Awareness of the Normative Structure of Games.” Developmental Psychology 44 (3): 875. Richerson, Peter J., and Robert Boyd. 2008. Not by Genes Alone: How Culture Transformed Human Evolution. Chicago: University of Chicago Press. 306 Rizzolatti, Giacomo, Rosolino Camarda, Leonardo Fogassi, Maurizio Gentilucci, Giuseppe Luppino, and Massimo Matelli. 1988. “Functional Organization of Inferior Area 6 in the Macaque Monkey.” Experimental Brain Research 71 (3): 491–507. Rizzolatti, Giacomo, and Laila Craighero. 2004. “The Mirror-Neuron System.” Annu. Rev. Neurosci. 27: 169–192. Rochat, Magali J., Fausto Caruana, Ahmad Jezzini, Irakli Intskirveli, Franck Grammont, Vittorio Gallese, Giacomo Rizzolatti, Maria Alessandra Umiltà, and others. 2010. “Responses of Mirror Neurons in Area F5 to Hand and Tool Grasping Observation.” Experimental Brain Research 204 (4): 605–616. Rosch, Eleanor. 1999. “Principles of Categorization.” In Readings in Cognitive Science, a Perspective From Psychology and Artificial Intelligence, edited by Allen Collins and Edward Smith, 189–206. Burlington: Morgan Kaufmann Publishers. Rosch, Eleanor, and Carolyn B. Mervis. 1975. “Family Resemblances: Studies in the Internal Structure of Categories.” Cognitive Psychology 7 (4): 573–605. Rosenberg, Morris. 1965. Society and the Adolescent Self-Image. Princeton: Princeton University Press. Rosenberg, Shawn W. 2007. “Rethinking Democratic Deliberation: The Limits and Potential of Citizen Participation.” Polity 39 (3): 335–60. Rudman, Laurie A., and Peter Glick. 1999. “Feminized Management and Backlash toward Agentic Women: The Hidden Costs to Women of a Kinder, Gentler Image of Middle Managers.” Journal of Personality and Social Psychology 77 (5): 1004. 307 ———. 2001. “Prescriptive Gender Stereotypes and Backlash toward Agentic Women.” Journal of Social Issues 57 (4): 743–762. Rudman, Laurie A., Anthony G. Greenwald, and Debbie E. McGhee. 2001. “Implicit Self-Concept and Evaluative Implicit Gender Stereotypes: Self and Ingroup Share Desirable Traits.” Personality and Social Psychology Bulletin 27 (9): 1164–1178. Rush, Curtis. 2011. “Cop Apologizes for ‘sluts’ Remark at Law School | Toronto Star.” Thestar.com. February 18. https://www.thestar.com/news/gta/2011/02/18/cop_apologizes_for_sluts_remark_at_law_school.html. Schlozman, Kay Lehman, Nancy Burns, and Sidney Verba. 1994. “Gender and the Pathways to Participation: The Role of Resources.” The Journal of Politics 56 (4): 963–990. ———. 1999. “‘What Happened at Work Today?’: A Multistage Model of Gender, Employment, and Political Participation.” The Journal of Politics 61 (1): 29–53. Schlozman, Kay Lehman, Nancy Burns, Sidney Verba, and Jesse Donahue. 1995. “Gender and Citizen Participation: Is There a Different Voice?” American Journal of Political Science, 267–293. Schlozman, Kay Lehman, Sidney Verba, and Henry E. Brady. 2010. “Weapon of the Strong? Participatory Inequality and the Internet.” Perspectives on Politics 8 (02): 487–509. Schunk, Dale H. 1984. “Sequential Attributional Feedback and Children’s Achievement Behaviors.” Journal of Educational Psychology 76 (6): 1159. Searle, John. 2010. Making the Social World: The Structure of Human Civilization. Oxford: Oxford University Press. 308 Searle, John R. 1995. The Construction of Social Reality. New York: Simon and Schuster. ———. 2013. Freedom and Neurobiology: Reflections on Free Will, Language, and Political Power. New York: Columbia University Press. Siim, Birte. 2000. Gender and Citizenship: Politics and Agency in France, Britain and Denmark. Cambridge: Cambridge University Press. Sinclair, Lisa, and Ziva Kunda. 2000. “Motivated Stereotyping of Women: She’s Fine If She Praised Me but Incompetent If She Criticized Me.” Personality and Social Psychology Bulletin 26 (11): 1329–1342. Singer, Tania. 2006a. “The Neuronal Basis and Ontogeny of Empathy and Mind Reading: Review of Literature and Implications for Future Research.” Neuroscience & Biobehavioral Reviews 30 (6): 855–63. ———. 2006b. “The Neuronal Basis and Ontogeny of Empathy and Mind Reading: Review of Literature and Implications for Future Research.” Neuroscience & Biobehavioral Reviews 30 (6): 855–863. Skinner, Quentin. 2002. Visions of Politics. Cambridge: Cambridge University Press. Smith, Graham. 2009. Democratic Innovations: Designing Institutions for Citizen Participation. Cambridge: Cambridge University Press. Sniderman, Paul M. 1975. Personality and Democratic Politics. University of California Press. Snyder, C. R., and Howard L. Fromkin. 1980. Uniqueness: The Pursuit of Human Difference. New York: New York: Plenum Press. 309 Spencer, Bettina, and Emanuele Castano. 2007. “Social Class Is Dead. Long Live Social Class! Stereotype Threat among Low Socioeconomic Status Individuals.” Social Justice Research 20 (4): 418–432. Spencer, Steven J., Claude M. Steele, and Diane M. Quinn. 1999. “Stereotype Threat and Women’s Math Performance.” Journal of Experimental Social Psychology 35 (1): 4–28. Squires, Catherine R. 2002. “Rethinking the Black Public Sphere: An Alternative Vocabulary for Multiple Public Spheres.” Communication Theory 12 (4): 446–468. Statham, Anne, Laurel Richardson, and Judith A. Cook. 1991. Gender and University Teaching: A Negotiated Difference. New York: SUNY Press. Stepien, Kathy A., and Amy Baernstein. 2006. “Educating for Empathy.” Journal of General Internal Medicine 21 (5): 524–30. Stolle, Dietlind. 2002. “Trusting Strangers–the Concept of Generalized Trust in Perspective.” Austrian Journal of Political Science 31 (4): 397–412. ———. 2003. “The Sources of Social Capital.” In Generating Social Capital: Civil Society and Institutions in Comparative Perspective. New York, NY: Palgrave Macmillan. Stolle, Dietlind, and Elisabeth Gidengil. 2010. “What Do Women Really Know? A Gendered Analysis of Varieties of Political Knowledge.” Perspectives on Politics 8 (01): 93–109. Stolle, Dietlind, and Allison Harell. 2013. “Social Capital and Ethno-Racial Diversity: Learning to Trust in an Immigrant Society.” Political Studies 61 (1): 42–66. Stolle, Dietlind, Marc Hooghe, and Michele Micheletti. 2005. “Politics in the Supermarket: Political Consumerism as a Form of Political Participation.” International Political Science Review 26 (3): 245–269. 310 Stolle, Dietlind, and Michele Micheletti. 2006. “The Gender Gap Reversed: Political Consumerism as a Women-Friendly Form of Civic and Political Engagement.” In Gender and Social Capital, Chapter: The Gender Gap Reversed: Political Consumerism as a Women-Friendly Form of Civic and Political Engagement, edited by Elisabeth Gidengil and Brenda O’Neill, 45–72. London: Routledge. ———. 2013. Political Consumerism: Global Responsibility in Action. Cambridge: Cambridge University Press. Stolle, Dietlind, and Thomas R. Rochon. 1998. “Are All Associations Alike? Member Diversity, Associational Type, and the Creation of Social Capital.” American Behavioral Scientist 42 (1): 47–65. Stolle, Dietlind, Stuart Soroka, and Richard Johnston. 2008. “When Does Diversity Erode Trust? Neighborhood Diversity, Interpersonal Trust and the Mediating Effect of Social Interactions.” Political Studies 56 (1): 57–75. Strand, Elizabeth A. 1999. “Uncovering the Role of Gender Stereotypes in Speech Perception.” Journal of Language and Social Psychology 18 (1): 86–100. Sunstein, Cass R. 2000. “Deliberative Trouble? Why Groups Go to Extremes.” The Yale Law Journal 110 (1): 71–119. Tajfel, Henri. 1981. Human Groups and Social Categories: Studies in Social Psychology. Cambridge: Cambridge University Press. Taylor, Charles. 1997. “Multiculturalism.” In The Politics of Recognition, edited by Amy Gutmann, 25–74. Princeton: Princeton University Press. 311 The Black Public Sphere Collective, ed. 1995. The Black Public Sphere: A Public Culture Book. Chicago: University of Chicago Press. Thomas, Melanee. 2012. “The Complexity Conundrum: Why Hasn’t the Gender Gap in Subjective Political Competence Closed?” Canadian Journal of Political Science 45 (02): 337–358. Thompson, Debra. 2009. “Racial Ideas and Gendered Intimacies: The Regulation of Interracial Relationships in North America.” Social & Legal Studies 18 (3): 353–371. Tomasello, Michael. 2000. “Culture and Cognitive Development.” Current Directions in Psychological Science 9 (2): 37–40. ———. 2004. “Learning through Others.” Daedalus 133 (1): 51–58. ———. 2009a. The Cultural Origins of Human Cognition. Boston: Harvard University Press. ———. 2009b. Why We Cooperate. Cambridge: MIT Press. ———. 2010. Origins of Human Communication. Cambridge: MIT Press. ———. 2015. “Joint Attention as Social Cognition.” In Joint Attention: Its Origins and Role in Development, edited by Chris Moore and Philip J Dunham, 103–130. New York: Psychology Press. Tomasello, Michael, and Malinda Carpenter. 2007. “Shared Intentionality.” Developmental Science 10 (1): 121–125. Tomasello, Michael, Malinda Carpenter, Josep Call, Tanya Behne, and Henrike Moll. 2005. “Understanding and Sharing Intentions: The Origins of Cultural Cognition.” Behavioral and Brain Sciences 28 (05): 675–691. 312 Tomasello, Michael, Ann Cale Kruger, and Hilary Horn Ratner. 1993. “Cultural Learning.” Behavioral and Brain Sciences 16 (03): 495–511. Trivers, Robert L. 1971. “The Evolution of Reciprocal Altruism.” Quarterly Review of Biology 46 (1): 35–57. “Truth and Reconciliation.” n.d. Tully, James. 2002. “Political Philosophy as a Critical Activity.” Political Theory 30 (4): 533–555. ———. 2008a. Political Philosophy in a New Key, Volume 1: Democracy and Civic Freedom. Cambridge: Cambridge University Press. ———. 2008b. Public Philosophy in a New Key Volume 2, Imperialism and Civic Freedom. Cambridge: Cambridge University Press. Turner, John C. 1987. Rediscovering the Social Group: A Self-Categorization Theory. Hoboken: Blackwell. Twenge, Jean M., and W. Keith Campbell. 2002. “Self-Esteem and Socioeconomic Status: A Meta-Analytic Review.” Personality and Social Psychology Review 6 (1): 59–71. Umiltà, M. A., L. Escola, I. Intskirveli, F. Grammont, M. Rochat, F. Caruana, A. Jezzini, V. Gallese, and G. Rizzolatti. 2008. “When Pliers Become Fingers in the Monkey Motor System.” Proceedings of the National Academy of Sciences 105 (6): 2209–13. doi:10.1073/pnas.0705985105. Urbinati, Nadia, and Mark E. Warren. 2008. “The Concept of Representation in Contemporary Democratic Theory.” Annual Review of Political Science 11: 387–412. 313 Verba, Sidney, Nancy Burns, and Kay Lehman Schlozman. 1997. “Knowing and Caring about Politics: Gender and Political Engagement.” The Journal of Politics 59 (4): 1051–72. Wängnerud, Lena. 2009. “Women in Parliaments: Descriptive and Substantive Representation.” Annual Review of Political Science 12: 51–69. Warneken, Felix, Frances Chen, and Michael Tomasello. 2006. “Cooperative Activities in Young Children and Chimpanzees.” Child Development 77 (3): 640–663. Warren, Mark. 2007. “Institutionalizing Deliberative Democracy.” In Deliberation, Participation, and Democracy: Can the People Govern?, 272–88. New York, USA: Palgrave Macmillan. Warren, Mark E. 1993. “Can Participatory Democracy Produce Better Selves? Psychological Dimensions of Habermas’s Discursive Model of Democracy.” Political Psychology, 209–234. ———. 1999. Democracy and Trust. Cambridge: Cambridge University Press. ———. 2001. Democracy and Association. Princeton: Princeton University Press. ———. 2017. “A Problem-Based Approach to Democratic Theory.” American Political Science Review 111 (1): 39–53. Warren, Mark E., and Hilary Pearse. 2008. Designing Deliberative Democracy: The British Columbia Citizens’ Assembly. Cambridge: Cambridge University Press. Weinstein, Barbara E., and Ira M. Ventry. 1982. “Hearing Impairment and Social Isolation in the Elderly.” Journal of Speech, Language, and Hearing Research 25 (4): 593–599. Williams, Bernard. 1972. “The Idea of Equality.” In Problems of the Self: Philosophical Papers 1956-1972, 230–49. Cambridge: University of Cambridge Press. 314 Williams, Melissa S. 2000. Voice, Trust, and Memory: Marginalized Groups and the Failings of Liberal Representation. Princeton: Princeton University Press. Wittgenstein, Ludwig. 1953. Philosophical Investigations. Hoboken: Wiley-Blackwell. Wuthnow, Robert. 1998. Loose Connections: Joining Together in America’s Fragmented Communities. Boston, MA: Harvard University Press. Young, Iris Marion. 2000. Inclusion and Democracy. New York: Oxford University Press. ———. 2009. “Five Faces of Oppression.” Geographic Thought: A Praxis Perspective, 55–71. ———. 2011. Justice and the Politics of Difference. Princeton: Princeton University Press. 315 Appendices Appendix A Extra Material for Chapter 7 A.1 Distributions of Independent Variables Table A-1 Distributions of independent variables Variable Frequency Mean Standard Dev. Min. Max Gender (Female=1) 11,548 0.5259785 0.4993463 0 1 Visible Minority (Minority=1) 10,835 0.1307799 0.337175 0 1 Age Category 10,957 3.316693 1.39521 1 5 Region (Quebec=1) 11,556 0.3085843 0.4619291 0 1 Marital status (married=1) 7,221 0.6202742 0.4853522 0 1 Education Secondary 11,064 0.195047 0.3962548 0 1 Some college/university 11,064 0.189624 0.3920212 0 1 College/university 11,064 0.4135936 0.4924996 0 1 Grad/professional degree 11,064 0.1103579 0.3133495 0 1 Poverty 9,976 0.2080994 0.4059687 0 1 Working for pay 10,811 0.42984 0.495076 0 1 Young children 11,040 0.0976449 0.2968474 0 1 Civic engagement (Volunteer=1) 7,184 0.4640869 0.4987433 0 1 Trust (Most people can be trusted=1) 6,940 0.4090778 0.4916991 0 1 Political efficacy (High efficacy=1) 6,805 0.0662748 0.2487801 0 1 Political interest (High=1) 7,227 0.3403902 0.4738732 0 1 316 A.2 Main Effects Table A-2 Baseline model (main effects) logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK VARIABLES Logit coef Odds Logit coef Odds Logit coef Odds Gender (female=1) -0.306*** 0.737*** -0.136* 0.873* -0.204*** 0.815*** (0.0821) (0.0605) (0.0711) (0.0621) (0.0626) (0.0510) Ethnicity (visible minority=1) 0.163 1.177 -0.251** 0.778** 0.0928 1.097 (0.119) (0.140) (0.107) (0.0833) (0.0930) (0.102) Age -0.116*** 0.890*** -0.106*** 0.899*** -0.302*** 0.739*** (0.0321) (0.0286) (0.0278) (0.0250) (0.0243) (0.0180) Region (Quebec=1) -0.528*** 0.590*** -0.395*** 0.673*** -0.243*** 0.784*** (0.0890) (0.0525) (0.0728) (0.0490) (0.0685) (0.0537) Constant -0.618*** 0.539*** 2.095*** 8.129*** 0.281*** 1.324*** (0.129) (0.0697) (0.127) (1.030) (0.104) (0.138) Observations 3,983 3,983 6,528 6,528 6,705 6,705 317 Table A-3 Resource model (main effects) logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK VARIABLES Logit coef Odds Logit coef Odds Logit coef Odds Gender (female=1) -0.230*** 0.794*** -0.0785 0.925 -0.201*** 0.818*** (0.0871) (0.0692) (0.0763) (0.0705) (0.0657) (0.0538) Ethnicity (visible minority=1) 0.136 1.146 -0.346*** 0.707*** 0.0779 1.081 (0.127) (0.145) (0.115) (0.0814) (0.0971) (0.105) Age -0.0848** 0.919** -0.107*** 0.899*** -0.309*** 0.734*** (0.0376) (0.0346) (0.0323) (0.0291) (0.0289) (0.0212) Region (Quebec=1) -0.506*** 0.603*** -0.382*** 0.682*** -0.260*** 0.771*** (0.0943) (0.0569) (0.0787) (0.0537) (0.0722) (0.0557) Marital status (married=1) -0.100 0.904 0.170** 1.186** -0.0128 0.987 (0.0945) (0.0855) (0.0853) (0.101) (0.0727) (0.0717) Education (baseline=no high school) Highschool degree -0.0841 0.919 0.277** 1.319** 0.262* 1.300* (0.202) (0.186) (0.140) (0.184) (0.153) (0.198) Some college 0.139 1.149 0.722*** 2.059*** 0.504*** 1.656*** (0.194) (0.223) (0.146) (0.300) (0.148) (0.246) College degree 0.340* 1.405* 0.799*** 2.224*** 0.501*** 1.651*** (0.183) (0.258) (0.133) (0.296) (0.141) (0.232) Post-grad degree 0.786*** 2.194*** 1.500*** 4.483*** 0.962*** 2.618*** (0.211) (0.462) (0.188) (0.842) (0.159) (0.417) Poverty -0.413*** 0.661*** -0.472*** 0.623*** 0.0932 1.098 (0.123) (0.0811) (0.102) (0.0634) (0.0941) (0.103) Employment (working=1) -0.0537 0.948 -0.191** 0.826** -0.143* 0.867* (0.0943) (0.0894) (0.0829) (0.0685) (0.0737) (0.0639) Kids (children under six years=1) 0.312** 1.366** 0.364** 1.439** -0.0645 0.938 (0.156) (0.213) (0.157) (0.225) (0.122) (0.114) 318 GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK VARIABLES Logit coef Odds Logit coef Logit coef Odds Constant -0.863*** 0.422*** 1.495*** 4.458*** -0.0676 0.935 (0.247) (0.104) (0.194) (0.863) (0.192) (0.179) Observations 3,576 3,576 5,834 5,834 5,987 5,987 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 319 Table A-4 Full model (main effects) logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK VARIABLES Logit coef Odds Logit coef Odds Logit coef Odds Gender (female=1) -0.240** 0.787** 0.0495 1.051 -0.107 0.898 (0.102) (0.0805) (0.0846) (0.0889) (0.0721) (0.0647) Ethnicity (visible minority=1) 0.237 1.268 -0.292** 0.747** 0.0828 1.086 (0.148) (0.188) (0.127) (0.0949) (0.105) (0.115) Age -0.0924** 0.912** -0.170*** 0.844*** -0.385*** 0.681*** (0.0434) (0.0396) (0.0370) (0.0312) (0.0326) (0.0222) Region (Quebec=1) -0.335*** 0.715*** -0.271*** 0.762*** -0.216*** 0.806*** (0.108) (0.0775) (0.0883) (0.0673) (0.0794) (0.0640) Marital status (married=1) -0.207* 0.813* 0.0891 1.093 -0.0596 0.942 (0.109) (0.0887) (0.0943) (0.103) (0.0794) (0.0748) Education (baseline=no high school) 0.0353 1.036 0.232 1.261 0.270* 1.310* Highschool degree (0.229) (0.238) (0.157) (0.197) (0.163) (0.213) 0.00104 1.001 0.492*** 1.635*** 0.325** 1.384** Some college (0.221) (0.221) (0.163) (0.267) (0.161) (0.223) 0.200 1.222 0.589*** 1.803*** 0.295* 1.344* College degree (0.208) (0.254) (0.152) (0.274) (0.151) (0.203) 0.335 1.399 0.966*** 2.627*** 0.569*** 1.767*** Post-grad degree (0.240) (0.335) (0.206) (0.541) (0.173) (0.305) -0.308** 0.735** -0.330*** 0.719*** 0.201* 1.222* Poverty (0.139) (0.102) (0.116) (0.0837) (0.103) (0.126) 0.154 1.166 -0.0386 0.962 -0.0505 0.951 Employment (working=1) (0.112) (0.130) (0.0905) (0.0871) (0.0804) (0.0764) 0.119 1.127 0.167 1.182 -0.202 0.817 (0.182) (0.206) (0.169) (0.199) (0.133) (0.109) Civic engagement (volunteers=1) 2.017*** 7.517*** 0.784*** 2.190*** 0.448*** 1.565*** (0.110) (0.830) (0.0914) (0.200) (0.0734) (0.115) 320 GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK VARIABLES Logit coef Odds Logit coef Logit coef Odds Logit coef Social trust (=1) 0.242** 1.273** 0.313*** 1.367*** -0.0542 0.947 (0.104) (0.133) (0.0915) (0.125) (0.0749) (0.0710) Political efficacy (politicians care=1) 0.418** 1.518** 0.343* 1.409* 0.197 1.218 (0.192) (0.292) (0.196) (0.277) (0.140) (0.171) Political interest (high interest=1) 0.908*** 2.479*** 1.547*** 4.696*** 1.195*** 3.302*** (0.105) (0.261) (0.117) (0.549) (0.0753) (0.249) Constant -2.343*** 0.0961*** 1.029*** 2.799*** -0.377* 0.686* (0.285) (0.0274) (0.224) (0.626) (0.208) (0.143) Observations 3,197 3,197 5,346 5,346 5,472 5,472 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Table A-5 Predictive margins for gender (main effect) (1) (2) (3) VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET DELIBERATION Men 0.251*** 0.835*** 0.323*** (0.0102) (0.00748) (0.00974) Women 0.218*** 0.841*** 0.302*** (0.00983) (0.00687) (0.00939) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 321 Table A-6 Discreet change in gender (effect of being a woman) VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET DELIBERATION Female -0.0337** 0.00599 -0.0204 (0.0143) (0.0102) (0.0137) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 322 A.3 Interactions Interactions between gender and ethnicity Table A-7 Full model with gender and ethnicity interaction logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk Deliberation Everyday talk Internet expression VARIABLES Logit coef Odds Logit coef Odds Logit coef Odds Visible minority man 0.271 1.312 -0.428** 0.652** 0.241* 1.273* (0.202) (0.265) (0.180) (0.117) (0.141) (0.180) Non-visible minority woman -0.230** 0.795** 0.0135 1.014 -0.0633 0.939 (0.110) (0.0872) (0.0902) (0.0914) (0.0772) (0.0725) Visible minority woman -0.0283 0.972 -0.145 0.865 -0.148 0.862 (0.213) (0.207) (0.177) (0.153) (0.156) (0.134) Age -0.0926** 0.912** -0.170*** 0.843*** -0.386*** 0.680*** (0.0434) (0.0396) (0.0369) (0.0311) (0.0326) (0.0222) Region (Quebec=1) -0.336*** 0.715*** -0.272*** 0.762*** -0.215*** 0.806*** (0.108) (0.0775) (0.0883) (0.0673) (0.0795) (0.0641) Marital status (married=1) -0.208* 0.813* 0.0887 1.093 -0.0598 0.942 (0.109) (0.0887) (0.0943) (0.103) (0.0794) (0.0748) 0.0344 1.035 0.234 1.264 0.266 1.305 Highschool degree (0.229) (0.237) (0.156) (0.197) (0.163) (0.213) 0.000304 1.000 0.493*** 1.638*** 0.324** 1.382** Some college (0.221) (0.221) (0.163) (0.267) (0.161) (0.223) 0.198 1.219 0.594*** 1.811*** 0.290* 1.337* 323 Deliberation Everyday talk Internet expression Logit coef Odds Logit coef Logit coef Odds College degree (0.208) (0.254) (0.152) (0.275) (0.151) (0.202) 0.333 1.396 0.973*** 2.647*** 0.562*** 1.754*** Post-grad degree (0.240) (0.335) (0.206) (0.546) (0.173) (0.303) -0.308** 0.735** -0.334*** 0.716*** 0.204** 1.226** Poverty (0.138) (0.102) (0.116) (0.0833) (0.103) (0.127) 0.153 1.165 -0.0384 0.962 -0.0542 0.947 Employment (working=1) (0.112) (0.131) (0.0905) (0.0871) (0.0805) (0.0763) 0.119 1.127 0.164 1.179 -0.199 0.820 (0.182) (0.205) (0.168) (0.198) (0.133) (0.109) Civic engagement (volunteers=1) 2.018*** 7.524*** 0.781*** 2.184*** 0.452*** 1.572*** (0.111) (0.832) (0.0914) (0.200) (0.0734) (0.115) Social trust (=1) 0.242** 1.273** 0.313*** 1.367*** -0.0529 0.948 (0.104) (0.133) (0.0915) (0.125) (0.0749) (0.0710) Political efficacy (=1) 0.414** 1.513** 0.343* 1.409* 0.193 1.213 (0.193) (0.292) (0.197) (0.277) (0.140) (0.170) Political interest (high interest=1) 0.907*** 2.477*** 1.549*** 4.706*** 1.194*** 3.300*** (0.105) (0.261) (0.117) (0.550) (0.0753) (0.249) Constant -2.345*** 0.0959*** 1.048*** 2.852*** -0.393* 0.675* (0.285) (0.0273) (0.223) (0.635) (0.209) (0.141) Observations 3,197 3,197 5,346 5,346 5,472 5,472 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 324 Table A-8 Predictive margins for the interaction between gender and ethnicity GROUP DELIBERATION EVERYDAY POLITICAL TALK EVERYDAY INTERNET TALK VARIABLES Predicted probabilities Predicted probabilities Predicted probabilities White man 0.245*** 0.842*** 0.317*** (0.0110) (0.00781) (0.0106) White woman 0.214*** 0.844*** 0.305*** (0.0105) (0.00730) (0.0101) Visible minority man 0.286*** 0.785*** 0.364*** (0.0290) (0.0243) (0.0262) Visible minority woman 0.241*** 0.824*** 0.289*** (0.0280) (0.0208) (0.0259) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 325 Table A-9 Discreet change in gender (effect of being a woman), by ethnicity GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK VARIABLES Non-visible minorities -0.0320** 0.00160 -0.0120 (0.0153) (0.0107) (0.0146) Visible minorities -0.0448 0.0389 -0.0752** (0.0398) (0.0320) (0.0367) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Table A-10 Discreet change in ethnicity (effect of being a visible minority), by gender VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK Men 0.0407 -0.0568** 0.0476* (0.0312) (0.0256) (0.0285) Women 0.0279 -0.0195 -0.0156 (0.0298) (0.0222) (0.0280) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 326 Interaction between gender and poverty Table A-11 Full model with gender and ethnicity interaction logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK VARIABLES Logit coef Odds Logit coef Odds Logit coef Odds Poor man -0.354* 0.702* -0.502*** 0.605*** 0.211 1.235 (0.188) (0.132) (0.169) (0.102) (0.146) (0.180) Not poor woman -0.254** 0.776** -0.0171 0.983 -0.104 0.901 (0.113) (0.0879) (0.0944) (0.0928) (0.0793) (0.0715) Poor woman -0.525*** 0.591*** -0.212 0.809 0.0876 1.092 (0.189) (0.112) (0.146) (0.118) (0.133) (0.145) Ethnicity (visible minority=1) 0.235 1.265 -0.301** 0.740** 0.0832 1.087 (0.148) (0.188) (0.127) (0.0940) (0.106) (0.115) Age -0.0929** 0.911** -0.173*** 0.841*** -0.385*** 0.681*** (0.0435) (0.0396) (0.0371) (0.0312) (0.0326) (0.0222) Region (Quebec=1) -0.336*** 0.715*** -0.273*** 0.761*** -0.216*** 0.806*** (0.109) (0.0776) (0.0883) (0.0672) (0.0794) (0.0640) Marital status (married=1) -0.207* 0.813* 0.0899 1.094 -0.0597 0.942 (0.109) (0.0888) (0.0943) (0.103) (0.0794) (0.0748) 0.0354 1.036 0.232 1.261 0.270* 1.310* Highschool degree (0.230) (0.238) (0.157) (0.198) (0.163) (0.213) 0.000376 1.000 0.492*** 1.636*** 0.325** 1.384** Some college (0.221) (0.221) (0.164) (0.268) (0.161) (0.223) 0.201 1.223 0.594*** 1.812*** 0.295* 1.343* College degree (0.209) (0.255) (0.152) (0.275) (0.151) (0.203) 0.335 1.398 0.970*** 2.637*** 0.569*** 1.767*** 327 Deliberation Everyday talk Internet expression VARIABLES Logit coef Odds Logit coef Logit coef Odds (0.240) (0.336) (0.206) (0.543) (0.173) (0.305) Employment 0.154 1.166 -0.0431 0.958 -0.0503 0.951 (0.112) (0.131) (0.0905) (0.0867) (0.0805) (0.0765) Young kids 0.119 1.126 0.160 1.174 -0.202 0.818 (0.183) (0.206) (0.168) (0.198) (0.133) (0.109) Civic engagement (volunteers=1) 2.016*** 7.510*** 0.785*** 2.192*** 0.448*** 1.565*** (0.111) (0.831) (0.0914) (0.200) (0.0734) (0.115) Social trust (=1) 0.241** 1.273** 0.313*** 1.367*** -0.0542 0.947 (0.104) (0.133) (0.0915) (0.125) (0.0749) (0.0710) Political efficacy (=1) 0.418** 1.520** 0.342* 1.408* 0.197 1.218 (0.193) (0.293) (0.196) (0.276) (0.140) (0.171) Political interest (high interest=1) 0.908*** 2.479*** 1.548*** 4.701*** 1.195*** 3.303*** (0.105) (0.261) (0.117) (0.549) (0.0754) (0.249) Constant -2.334*** 0.0969*** 1.072*** 2.921*** -0.379* 0.684* (0.286) (0.0277) (0.227) (0.664) (0.209) (0.143) Observations 3,197 3,197 5,346 5,346 5,472 5,472 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 328 Table A-12 Predictive margins for the interaction between gender and poverty VARIABLES GROUP DELIBERATION INTERNET POLITICAL TALK INTERNET POLITICAL TALK Not poor man 0.260*** 0.847*** 0.317*** (0.0116) (0.00798) (0.0105) Poor man 0.210*** 0.780*** 0.358*** (0.0226) (0.0228) (0.0269) Not poor woman 0.224*** 0.845*** 0.297*** (0.0112) (0.00778) (0.0104) Poor woman 0.188*** 0.821*** 0.333*** (0.0210) (0.0163) (0.0231) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Table A-13 Discreet change in gender (effect of being a woman), by poverty VARIABLES GROUP DELIBERATION INTERNET POLITICAL TALK INTERNET POLITICAL TALK Not poor -0.0364** -0.00201 -0.0196 (0.0163) (0.0111) (0.0149) Poor -0.0221 0.0408 -0.0245 (0.0299) (0.0264) (0.0337) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 329 Table A-14 Discreet change in poverty (effect of being poor), by gender VARIABLES GROUP DELIBERATION INTERNET POLITICAL TALK INTERNET POLITICAL TALK Men -0.0499* -0.0671*** 0.0414 (0.0257) (0.0245) (0.0291) Women -0.0356 -0.0243 0.0365 (0.0240) (0.0184) (0.0256) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 330 Interaction between gender, working for pay, and young kids Table A-15 Full model with gender, working for pay, and young kids interaction logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK VARIABLES Logit coef Odds Logit coef Odds Logit coef Odds Man, not working, has kids -0.384 0.681 -0.428 0.652 -0.0737 0.929 (0.520) (0.354) (0.372) (0.243) (0.361) (0.335) Man, working, no kids 0.254 1.289 -0.00998 0.990 0.0987 1.104 (0.155) (0.200) (0.131) (0.129) (0.109) (0.120) Man, working, has kids 0.381 1.463 0.461 1.586 -0.260 0.771 (0.319) (0.467) (0.287) (0.456) (0.215) (0.166) Woman, not working, no kids -0.118 0.889 0.100 1.105 0.0118 1.012 (0.142) (0.126) (0.112) (0.124) (0.0973) (0.0984) Woman, not working, has kids -0.339 0.712 0.467 1.596 -0.231 0.793 (0.351) (0.250) (0.354) (0.566) (0.265) (0.210) Woman, working, no kids -0.187 0.830 -0.00505 0.995 -0.181 0.834 (0.170) (0.141) (0.128) (0.127) (0.115) (0.0958) Woman, working, has kids 0.315 1.371 -0.0294 0.971 -0.241 0.786 (0.327) (0.448) (0.296) (0.288) (0.250) (0.196) Ethnicity (visible minority=1) 0.236 1.266 -0.292** 0.747** 0.0766 1.080 (0.149) (0.188) (0.127) (0.0947) (0.106) (0.114) Age -0.0954** 0.909** -0.172*** 0.842*** -0.384*** 0.681*** (0.0438) (0.0398) (0.0372) (0.0313) (0.0327) (0.0223) Region (Quebec=1) -0.335*** 0.716*** -0.267*** 0.766*** -0.217*** 0.805*** (0.109) (0.0777) (0.0885) (0.0677) (0.0795) (0.0640) Marital status (married=1) -0.196* 0.822* 0.0864 1.090 -0.0549 0.947 (0.109) (0.0900) (0.0942) (0.103) (0.0794) (0.0752) Highschool degree 0.0198 1.020 0.230 1.259 0.272* 1.312* 331 GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK VARIABLES Logit coef Odds Logit coef Logit coef Odds Logit coef (0.231) (0.236) (0.156) (0.196) (0.163) (0.213) Some college 0.00363 1.004 0.484*** 1.622*** 0.332** 1.394** (0.222) (0.223) (0.163) (0.264) (0.161) (0.224) College degree 0.196 1.216 0.582*** 1.790*** 0.300** 1.350** (0.210) (0.255) (0.151) (0.271) (0.151) (0.204) Post-grad degree 0.338 1.402 0.959*** 2.609*** 0.580*** 1.786*** (0.241) (0.338) (0.206) (0.537) (0.173) (0.309) Poverty -0.319** 0.727** -0.338*** 0.714*** 0.200* 1.222* (0.139) (0.101) (0.116) (0.0830) (0.103) (0.126) Civic engagement (volunteers=1) 2.010*** 7.467*** 0.791*** 2.207*** 0.445*** 1.560*** (0.111) (0.827) (0.0917) (0.202) (0.0734) (0.115) Social trust (=1) 0.241** 1.273** 0.322*** 1.380*** -0.0558 0.946 (0.105) (0.133) (0.0916) (0.126) (0.0748) (0.0708) Political efficacy (=1) 0.415** 1.515** 0.331* 1.393* 0.196 1.216 (0.193) (0.292) (0.197) (0.274) (0.141) (0.172) Political interest (=1) 0.906*** 2.474*** 1.548*** 4.700*** 1.195*** 3.304*** (0.106) (0.261) (0.117) (0.550) (0.0755) (0.249) Constant -2.366*** 0.0938*** 1.021*** 2.777*** -0.445** 0.641** (0.296) (0.0278) (0.226) (0.628) (0.211) (0.135) Observations 3,197 3,197 5,346 5,346 5,472 5,472 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 332 Table A-16 Predictive margins for the interaction between gender, employment status, and the presence of young children in the home VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK Man, not working for pay, no kids 0.234*** 0.833*** 0.321*** (0.0147) (0.0103) (0.0143) Man, not working for pay, has kids 0.184*** 0.775*** 0.307*** (0.0619) (0.0546) (0.0657) Man, working for pay, no kids 0.272*** 0.832*** 0.340*** (0.0165) (0.0125) (0.0152) Man, working for pay, has kids 0.291*** 0.884*** 0.273*** (0.0464) (0.0257) (0.0345) Woman, not working for pay, no kids 0.218*** 0.845*** 0.323*** (0.0136) (0.00940) (0.0139) Woman, not working for pay, has kids 0.189*** 0.884*** 0.278*** (0.0412) (0.0321) (0.0443) Woman, working for pay, no kids 0.209*** 0.833*** 0.287*** (0.0166) (0.0117) (0.0147) Woman, working for pay, has kids 0.281*** 0.830*** 0.277*** (0.0468) (0.0352) (0.0411) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.10 333 Interaction between gender and civic engagement Table A-17 Full model with gender and civic engagement interaction logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK VARIABLES Logit coef Odds Logit coef Odds Logit coef Odds Civically engaged man 2.027*** 7.588*** 0.777*** 2.176*** 0.439*** 1.550*** (0.149) (1.128) (0.134) (0.291) (0.101) (0.156) Non-civically engaged woman -0.225 0.798 0.0455 1.047 -0.118 0.889 (0.183) (0.146) (0.102) (0.107) (0.100) (0.0889) Civically engaged woman 1.780*** 5.931*** 0.835*** 2.305*** 0.340*** 1.405*** (0.149) (0.886) (0.122) (0.282) (0.0987) (0.139) Ethnicity (visible minority=1) 0.237 1.268 -0.292** 0.747** 0.0825 1.086 (0.148) (0.188) (0.127) (0.0949) (0.105) (0.115) Age -0.0925** 0.912** -0.170*** 0.844*** -0.385*** 0.681*** (0.0434) (0.0396) (0.0370) (0.0312) (0.0326) (0.0222) Region (Quebec=1) -0.336*** 0.715*** -0.271*** 0.762*** -0.216*** 0.806*** (0.108) (0.0775) (0.0884) (0.0674) (0.0794) (0.0640) Marital status (married=1) -0.207* 0.813* 0.0893 1.093 -0.0594 0.942 (0.109) (0.0889) (0.0945) (0.103) (0.0794) (0.0748) Highschool degree 0.0352 1.036 0.232 1.261 0.270* 1.311* (0.229) (0.238) (0.157) (0.197) (0.162) (0.213) Some college 0.00157 1.002 0.492*** 1.635*** 0.325** 1.384** (0.221) (0.222) (0.163) (0.267) (0.161) (0.223) College degree 0.201 1.222 0.589*** 1.803*** 0.295* 1.343* (0.208) (0.255) (0.152) (0.274) (0.151) (0.203) Post-grad degree 0.335 1.399 0.966*** 2.627*** 0.569*** 1.767*** (0.240) (0.335) (0.206) (0.541) (0.173) (0.305) 334 VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK Logit coef Odds Logit coef Logit coef Odds Logit coef Poverty -0.308** 0.735** -0.329*** 0.719*** 0.201* 1.223* (0.139) (0.102) (0.116) (0.0837) (0.103) (0.126) Employment (working=1) 0.153 1.166 -0.0384 0.962 -0.0501 0.951 (0.112) (0.130) (0.0906) (0.0872) (0.0805) (0.0765) Young kids 0.119 1.127 0.167 1.182 -0.202 0.817 (0.182) (0.206) (0.168) (0.199) (0.133) (0.109) Social trust 0.242** 1.273** 0.313*** 1.368*** -0.0543 0.947 (0.104) (0.133) (0.0917) (0.125) (0.0749) (0.0710) Political efficacy 0.418** 1.519** 0.343* 1.408* 0.197 1.218 (0.192) (0.292) (0.196) (0.276) (0.140) (0.171) Political interest 0.908*** 2.480*** 1.547*** 4.696*** 1.195*** 3.303*** (0.105) (0.261) (0.117) (0.548) (0.0754) (0.249) Constant -2.348*** 0.0955*** 1.031*** 2.804*** -0.373* 0.689* (0.294) (0.0281) (0.225) (0.630) (0.210) (0.145) Observations 3,197 3,197 5,346 5,346 5,472 5,472 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 335 Table A-18 Predictive margins for the interaction between gender and civic engagement (volunteering for a civic association) VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK Non-civically engaged man 0.101*** 0.794*** 0.280*** (0.0106) (0.0109) (0.0126) Civically engaged man 0.433*** 0.888*** 0.365*** (0.0198) (0.0105) (0.0152) Non-civically engaged woman 0.0832*** 0.801*** 0.259*** (0.0103) (0.0106) (0.0131) Civically engaged woman 0.378*** 0.893*** 0.345*** (0.0187) (0.00881) (0.0140) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Table A-19 Discreet change in gender (effect of being a woman), by civic engagement VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK Non-civically engaged -0.0182 0.00668 -0.0210 (0.0147) (0.0150) (0.0177) Civically engaged -0.0552** 0.00533 -0.0201 (0.0273) (0.0137) (0.0208) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 336 Table A-20 Discreet change in civic engagement (effect of volunteering for a civic association), by gender (2) (4) (6) VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK Men 0.332*** 0.0932*** 0.0859*** (0.0226) (0.0105) (0.0141) Women 0.295*** 0.0932*** 0.0859*** (0.0214) (0.0105) (0.0141) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 337 Interaction between gender and social trust Table A-21 Full model with gender and social trust interaction logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK VARIABLES Logit coef Odds Logit coef Odds Logit coef Odds Man, high trust 0.146 1.157 0.352*** 1.422*** -0.0231 0.977 (0.142) (0.164) (0.128) (0.182) (0.102) (0.1000) Woman, low trust -0.319** 0.727** 0.0749 1.078 -0.0795 0.924 (0.129) (0.0936) (0.103) (0.111) (0.0928) (0.0857) Woman, high trust 0.0304 1.031 0.349*** 1.418*** -0.167 0.846 (0.152) (0.156) (0.130) (0.184) (0.103) (0.0872) Ethnicity (visible minority=1) 0.237 1.268 -0.292** 0.747** 0.0833 1.087 (0.148) (0.188) (0.127) (0.0949) (0.105) (0.115) Age -0.0929** 0.911** -0.170*** 0.844*** -0.384*** 0.681*** (0.0435) (0.0396) (0.0370) (0.0312) (0.0326) (0.0222) Region (Quebec=1) -0.332*** 0.717*** -0.272*** 0.762*** -0.217*** 0.805*** (0.109) (0.0779) (0.0883) (0.0673) (0.0794) (0.0640) Marital status (married=1) -0.207* 0.813* 0.0885 1.093 -0.0601 0.942 (0.109) (0.0887) (0.0945) (0.103) (0.0794) (0.0747) Highschool degree 0.0250 1.025 0.232 1.261 0.270* 1.310* (0.230) (0.236) (0.156) (0.197) (0.163) (0.213) Some college -0.0102 0.990 0.493*** 1.636*** 0.326** 1.385** (0.221) (0.219) (0.163) (0.267) (0.161) (0.223) College degree 0.190 1.209 0.590*** 1.805*** 0.296* 1.345* (0.209) (0.252) (0.152) (0.274) (0.151) (0.203) Post-grad degree 0.320 1.377 0.968*** 2.633*** 0.571*** 1.770*** (0.240) (0.331) (0.206) (0.543) (0.173) (0.306) 338 VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK Logit coef Odds Logit coef Logit coef Odds Logit coef Poverty -0.310** 0.734** -0.330*** 0.719*** 0.200* 1.221* (0.139) (0.102) (0.116) (0.0837) (0.103) (0.126) Employment (working=1) 0.154 1.167 -0.0385 0.962 -0.0502 0.951 (0.112) (0.131) (0.0906) (0.0871) (0.0804) (0.0764) Young kids 0.119 1.127 0.167 1.182 -0.202 0.817 (0.183) (0.206) (0.169) (0.199) (0.133) (0.109) Civic engagement 2.020*** 7.539*** 0.783*** 2.189*** 0.448*** 1.565*** (0.110) (0.832) (0.0915) (0.200) (0.0734) (0.115) Political efficacy 0.417** 1.517** 0.339* 1.403* 0.195 1.216 (0.192) (0.292) (0.196) (0.275) (0.140) (0.170) Political interest 0.907*** 2.477*** 1.547*** 4.695*** 1.195*** 3.304*** (0.105) (0.261) (0.117) (0.549) (0.0753) (0.249) Constant -2.294*** 0.101*** 1.016*** 2.761*** -0.393* 0.675* (0.288) (0.0291) (0.227) (0.626) (0.211) (0.143) Observations 3,197 3,197 5,346 5,346 5,472 5,472 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 339 Table A-22 Predictive margins for the interaction between gender and social trust (most people can be trusted, or you can’t be too careful) VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK Man, low social trust 0.244*** 0.820*** 0.325*** (0.0130) (0.00995) (0.0125) Man, high social trust 0.265*** 0.862*** 0.320*** (0.0165) (0.0113) (0.0152) Woman, low social trust 0.200*** 0.829*** 0.310*** (0.0119) (0.00884) (0.0126) Woman, high social trust 0.248*** 0.862*** 0.293*** (0.0176) (0.0114) (0.0145) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Table A-23 Discreet change in gender (effect of being a woman), by social trust VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK Not socially trusting -0.0436** 0.00969 -0.0152 (0.0176) (0.0133) (0.0177) Socially trusting -0.0171 -0.000273 -0.0271 (0.0242) (0.0159) (0.0209) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 340 Table A-24 Discreet change in trust (effect of agreeing that most people can be trusted), by gender VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK Man 0.0215 0.0423*** -0.00445 (0.0210) (0.0151) (0.0197) Woman 0.0480** 0.0323** -0.0163 (0.0212) (0.0146) (0.0195) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 341 Interaction between gender and political interest Table A-25 Full model with gender and political interest interaction logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL DISCUSSION VARIABLES Logit coef Odds Logit coef Odds Logit coef Odds Man, high political interest 0.947*** 2.579*** 1.445*** 4.244*** 1.212*** 3.359*** (0.140) (0.361) (0.152) (0.643) (0.101) (0.339) Woman, low political interest -0.207 0.813 0.0164 1.017 -0.0925 0.912 (0.130) (0.106) (0.0917) (0.0932) (0.0951) (0.0867) Woman, high political interest 0.653*** 1.921*** 1.712*** 5.540*** 1.082*** 2.951*** (0.158) (0.304) (0.185) (1.023) (0.110) (0.325) Ethnicity (visible minority=1) 0.235 1.265 -0.291** 0.748** 0.0824 1.086 (0.148) (0.187) (0.127) (0.0950) (0.106) (0.115) Age -0.0921** 0.912** -0.171*** 0.843*** -0.384*** 0.681*** (0.0435) (0.0396) (0.0370) (0.0312) (0.0326) (0.0222) Region (Quebec=1) -0.336*** 0.714*** -0.271*** 0.762*** -0.216*** 0.806*** (0.109) (0.0775) (0.0882) (0.0673) (0.0794) (0.0640) Marital status (married=1) -0.207* 0.813* 0.0913 1.096 -0.0605 0.941 (0.109) (0.0887) (0.0943) (0.103) (0.0794) (0.0748) Highschool degree 0.0341 1.035 0.232 1.261 0.270* 1.310* (0.230) (0.238) (0.157) (0.197) (0.163) (0.213) Some college -0.00323 0.997 0.496*** 1.643*** 0.324** 1.382** (0.221) (0.220) (0.163) (0.268) (0.161) (0.223) College degree 0.200 1.221 0.591*** 1.805*** 0.295* 1.343* (0.208) (0.254) (0.152) (0.274) (0.151) (0.203) Post-grad degree 0.333 1.396 0.969*** 2.636*** 0.568*** 1.765*** 342 VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL DISCUSSION Odds Logit coef Odds Odds Logit coef Odds (0.240) (0.335) (0.206) (0.543) (0.173) (0.305) Poverty -0.307** 0.736** -0.329*** 0.720*** 0.200* 1.222* (0.138) (0.102) (0.116) (0.0837) (0.103) (0.126) Employment (working=1) 0.153 1.166 -0.0390 0.962 -0.0505 0.951 (0.112) (0.130) (0.0905) (0.0871) (0.0804) (0.0764) Young kids 0.120 1.128 0.165 1.180 -0.202 0.817 (0.182) (0.206) (0.169) (0.199) (0.133) (0.109) Civic engagement 2.017*** 7.517*** 0.785*** 2.192*** 0.448*** 1.565*** (0.111) (0.831) (0.0914) (0.200) (0.0735) (0.115) Social trust 0.242** 1.274** 0.312*** 1.367*** -0.0538 0.948 (0.104) (0.133) (0.0915) (0.125) (0.0749) (0.0710) Political efficacy 0.413** 1.512** 0.345* 1.411* 0.197 1.217 (0.193) (0.292) (0.196) (0.277) (0.140) (0.171) Constant -2.360*** 0.0944*** 1.047*** 2.849*** -0.385* 0.681* (0.290) (0.0274) (0.225) (0.640) (0.211) (0.143) Observations 3,197 3,197 5,346 5,346 5,472 5,472 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 343 Table A-26 Predictive margins for the interaction between gender and political interest VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK Male, low political interest 0.203*** 0.789*** 0.239*** (0.0127) (0.0104) (0.0120) Male, high political interest 0.351*** 0.936*** 0.491*** (0.0180) (0.00789) (0.0166) Female, low political interest 0.176*** 0.791*** 0.224*** (0.0110) (0.00946) (0.0103) Female, high political interest 0.301*** 0.950*** 0.461*** (0.0209) (0.00793) (0.0192) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Table A-27 Discreet change in gender (effect of being female), by political interest VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK Less political interest -0.0267 0.00253 -0.0154 (0.0168) (0.0141) (0.0158) High political interest -0.0500* 0.0138 -0.0298 (0.0275) (0.0111) (0.0251) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 344 Table A-28 Discreet change in political interest (effect of high interest), by gender VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK Man 0.148*** 0.148*** 0.252*** (0.0221) (0.0130) (0.0204) Woman 0.125*** 0.159*** 0.237*** (0.0237) (0.0124) (0.0220) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 345 Interaction between gender and political efficacy Table A-29 Full model with gender and political efficacy interaction logit coefficients and odds ratios for group deliberation, everyday political talk, and Internet political talk GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK VARIABLES Logit coef Odds Logit coef Odds Logit coef Odds Man, high efficacy 0.376 1.457 0.325 1.384 0.441** 1.554** (0.253) (0.369) (0.277) (0.383) (0.183) (0.285) Woman, low efficacy -0.246** 0.782** 0.0478 1.049 -0.0662 0.936 (0.106) (0.0829) (0.0865) (0.0908) (0.0744) (0.0697) Woman, high efficacy 0.221 1.247 0.412 1.509 -0.220 0.802 (0.291) (0.363) (0.274) (0.414) (0.219) (0.176) Ethnicity (visible minority=1) 0.239 1.270 -0.292** 0.747** 0.0799 1.083 (0.149) (0.189) (0.127) (0.0949) (0.105) (0.114) Age -0.0927** 0.911** -0.170*** 0.844*** -0.383*** 0.682*** (0.0435) (0.0396) (0.0370) (0.0312) (0.0324) (0.0221) Region (Quebec=1) -0.335*** 0.715*** -0.271*** 0.762*** -0.215*** 0.806*** (0.108) (0.0775) (0.0883) (0.0673) (0.0795) (0.0641) Marital status (married=1) -0.207* 0.813* 0.0891 1.093 -0.0587 0.943 (0.109) (0.0887) (0.0943) (0.103) (0.0794) (0.0748) Highschool degree 0.0360 1.037 0.232 1.261 0.268* 1.307* (0.229) (0.238) (0.156) (0.197) (0.163) (0.213) Some college 0.00204 1.002 0.492*** 1.635*** 0.328** 1.389** (0.221) (0.221) (0.163) (0.267) (0.161) (0.224) College degree 0.201 1.222 0.589*** 1.802*** 0.299** 1.349** (0.208) (0.254) (0.152) (0.274) (0.151) (0.204) Post-grad degree 0.337 1.401 0.966*** 2.627*** 0.567*** 1.764*** 346 VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK Odds Logit coef Odds Odds Logit coef Odds (0.240) (0.336) (0.206) (0.541) (0.173) (0.305) Poverty -0.308** 0.735** -0.330*** 0.719*** 0.201* 1.222* (0.139) (0.102) (0.116) (0.0837) (0.103) (0.126) Employment (working=1) 0.153 1.165 -0.0385 0.962 -0.0519 0.949 (0.112) (0.130) (0.0906) (0.0872) (0.0802) (0.0761) Young kids 0.119 1.126 0.167 1.182 -0.194 0.824 (0.183) (0.206) (0.169) (0.199) (0.133) (0.110) Civic engagement 2.017*** 7.515*** 0.784*** 2.190*** 0.448*** 1.566*** (0.110) (0.830) (0.0914) (0.200) (0.0735) (0.115) Social trust 0.241** 1.273** 0.313*** 1.368*** -0.0576 0.944 (0.104) (0.133) (0.0914) (0.125) (0.0750) (0.0708) Political interest 0.909*** 2.483*** 1.547*** 4.696*** 1.193*** 3.298*** (0.105) (0.262) (0.117) (0.549) (0.0754) (0.249) Constant -2.340*** 0.0963*** 1.030*** 2.802*** -0.403* 0.668* (0.285) (0.0274) (0.224) (0.627) (0.208) (0.139) Observations 3,197 3,197 5,346 5,346 5,472 5,472 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 347 Table A-30 Predictive margins for the interaction between gender and political efficacy (1) (2) (3) VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK Man, lower political efficacy 0.248*** 0.833*** 0.316*** (0.0106) (0.00774) (0.0101) Man, higher political efficacy 0.306*** 0.870*** 0.405*** (0.0390) (0.0278) (0.0373) Woman, lower political efficacy 0.214*** 0.839*** 0.304*** (0.0101) (0.00712) (0.00970) Woman, higher political efficacy 0.281*** 0.879*** 0.276*** (0.0434) (0.0260) (0.0377) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Table A-31 Discreet change in gender (effect of being female), by political efficacy VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK Lower political efficacy -0.0343** 0.00585 -0.0125 (0.0148) (0.0106) (0.0141) Higher political efficacy -0.0244 0.00870 -0.130** (0.0586) (0.0379) (0.0529) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 348 Table A-32 Discreet change in political efficacy (effect of high efficacy), by gender VARIABLES GROUP DELIBERATION EVERYDAY POLITICAL TALK INTERNET POLITICAL TALK Man 0.0576 0.0368 0.0891** (0.0403) (0.0288) (0.0385) Woman 0.0674 0.0396 -0.0281 (0.0446) (0.0270) (0.0390) Observations 3,197 5,346 5,472 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 349 Appendix B Extra Material for Chapter 8 B.1 Experimental Treatment Table B-1 Initial policy question and counterarguments Initial Policy Question: Some people think that car insurance should be mandatory. That is, required for all drivers who want to use the road. Others think that car insurance should be optional, with a ‘personal responsibility requirement’. This means that a person is not required to buy insurance, if they prove they could pay for damages in the event of a car accident. What is your opinion on this issue? Car insurance should be mandatory. Car insurance should NOT be mandatory (should be optional). Counter argument for “Car insurance should be mandatory” Counter argument for “Car insurance should NOT be mandatory (should be optional)” Another student, [Jessica/Michael], has a different opinion. Please carefully read [Jessica’s/Michael’s] counter argument, and then answer a couple of questions. [Jessica’s/Michael’s] response: “Car insurance definitely shouldn’t be mandatory! People shouldn’t be forced to buy things they don’t need or want. The potential cost of accidents can be taken care of by enforcing a personal responsibility requirement. If a person can prove that they would be able to pay for the damage, if they ever got into an accident, then they shouldn’t be forced to buy something they don’t need and don’t want.” Another student, [Jessica/Michael], has a different opinion. Please carefully read [Jessica’s/Michael’s] counter argument, and then answer a couple of questions. [Jessica’s/Michael’s] “Car insurance should definitely be mandatory! People have to buy insurance before they drive on public roads because there’s always a chance that at-fault drivers won’t pay for the damage they cause, either because they can’t afford the costs or because they don’t admit to the damage, like in hit-and-run accidents. Laws asking drivers to prove that they would be able to pay for damages without insurance just aren’t good enough to take care of this problem.” 350 B.2 Self-Esteem Measure Table B-2 Rosenberg self-Esteem scale wording and coding Below is a list of statements dealing with your general feelings about yourself. Please indicate whether you strongly agree, agree, disagree, or strongly disagree with the statements: Strongly Agree Agree Disagree Strongly Disagree 1. On the whole, I am satisfied with myself 2.* At times, I think I am no good at all. 3. I feel that I have a number of good qualities. 4. I am able to do things as well as most other people 5.* I feel I do not have much to be proud of. 6.* I certainly feel useless at times. 7. I feel that I’m a person of worth, at least on an equal plane with others. 8.* I wish I could have more respect for myself. 9.* All in all, I am inclined to feel that I am a failure. 10. I take a positive attitude toward myself. Scoring: SA=3, A=2, D=1, SD=0. Items with an asterisk are reverse scored, that is, SA=0, A=1, D=2, SD=3. Sum the scores for the 10 items. The higher the score, the higher the self-esteem. See Rosenberg (1965). Table B-3 Mean difference on Rosenberg Self-Esteem (RSE) scale, by respondent gender Gender Mean RSE Scale Score (Std. Err.) Frequency Male 19.776 (0.383) 165 Female 17.977 (0.349) 215 Combined 18.758 (0.262) 380 Difference 1.799*** (0.519) Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 351 B.3 Implicit Association Test Table B-4 Implicit association test stimulus materials Concepts Related to Good Decision-Making Concepts Related to Bad Decision-Making Authoritative, Strong, Rational, Intelligent, Important, Confident, Decisive Subordinate, Weak, Confused, Foolish, Useless, and Uncertain Male Faces Female Faces Table B-5 Mean differences on d-scores between male and female respondents Gender Mean D-Score (Std. Err.) Frequency Male 0.283 (0.080) 26 Female -0.077 (0.064) 41 Combined 0.063 (0.054) 67 Difference 0.360*** (0.105) Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Note: Recall the D-score ranges from -2 to 2. The highest score (2) indicate stronger implicit associations between MALE faces and traits related to GOOD decision-making and FEMALE faces and traits related to BAD decision-making. The lowest value (-2) indicates the opposite (stronger implicit associations between FEMALE-GOOD and MALE-BAD). Zero indicates no implicit association between gender and decision-making traits. 352 Table B-6 Mean D-scores, by respondent gender and measure of explicit gender attitude (mean responses to the statement: “society would be better off if more women stayed home.”) Respondent Gender Male Female Total AGREE (or neutral) that society would be better off if more women stayed home. 0.348 (0.302) n=4 0.139 (0.462) n=6 0.222 (0.401) n=10 DISAGREE that society would be better off if more women stayed home. 0.271 (0.430) n=22 -0.114 (0.394) n=35 0.034 (0.447) n=57 Total 0.283 (0.409) n=26 -0.077 (0.409) n=41 0.062 (0.442) n=67 Note: Male respondents who AGREE society would be better off if more women stayed home have the highest scores (implicit associations between males and attributes related to good decision-making), with a mean D-Score of 0.348. Female respondents who DISAGREE society would be better off have the lowest scores, with a mean D-Score of -0.114. 353 B.4 Predicted Marginal Effects Table B-7 Marginal predicted probabilities for main model: five categories of the dependent variable, by counter argument condition (1) (2) (3) (4) (5) VARIABLES Def. NOT Very Unlikely Somewhat Unlikely Somewhat Likely Very Likely/Definitely Woman Counterargument 0.114*** 0.336*** 0.340*** 0.171*** 0.0380*** (0.0187) (0.0292) (0.0269) (0.0226) (0.00912) Man Counterargument 0.0821*** 0.280*** 0.361*** 0.222*** 0.0540*** (0.0149) (0.0279) (0.0275) (0.0268) (0.0123) Observations 380 380 380 380 380 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 NOTE: Variables at Means Table B-8 Marginal predicted probabilities for model 2: five categories of the dependent variable, by gender and counterargument interaction (1) (2) (3) (4) (5) VARIABLES Def. NOT Very Unlikely Somewhat Unlikely Somewhat Likely Very Likely/Definitely Male R.*Woman Counterargument 0.135*** 0.363*** 0.323*** 0.148*** 0.0312*** (0.0285) (0.0376) (0.0317) (0.0283) (0.00917) Male R.*Man Counterargument 0.118*** 0.341*** 0.338*** 0.167*** 0.0363*** (0.0230) (0.0327) (0.0291) (0.0256) (0.00935) Female R.*Woman Counterarg. 0.0985*** 0.311*** 0.353*** 0.194*** 0.0441*** (0.0217) (0.0377) (0.0284) (0.0333) (0.0128) Female R.*Man Counterargument 0.0615*** 0.232*** 0.364*** 0.271*** 0.0715*** (0.0137) (0.0313) (0.0277) (0.0355) (0.0172) Observations 380 380 380 380 380 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 354 Table B-9 Marginal predicted probabilities for model 3: five categories of the dependent variable, by gender, counterargument, and self-esteem interaction (for low self-esteem) (1) (2) (3) (4) (5) VARIABLES Def. NOT Very Unlikely Somewhat Unlikely Somewhat Likely Very Likely/Definitely Male R.*Woman Counterargument 0.156*** 0.344*** 0.294*** 0.166*** 0.0404** (0.0505) (0.0441) (0.0337) (0.0462) (0.0171) Male R.*Man Counterargument 0.181*** 0.360*** 0.279*** 0.146*** 0.0337** (0.0557) (0.0376) (0.0365) (0.0401) (0.0137) Female R.*Woman Counterarg. 0.0994*** 0.293*** 0.325*** 0.220*** 0.0628*** (0.0311) (0.0487) (0.0272) (0.0453) (0.0226) Female R.*Man Counterargument 0.0646** 0.229*** 0.331*** 0.279*** 0.0954** (0.0267) (0.0595) (0.0253) (0.0575) (0.0381) Observations 380 380 380 380 380 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 355 Table B-10 Marginal predicted probabilities for model 3: five categories of the dependent variable, by gender, counterargument, and self-esteem interaction (for average self-esteem) (1) (2) (3) (4) (5) VARIABLES Def. NOT Very Unlikely Somewhat Unlikely Somewhat Likely Very Likely/Definitely Male R.*Woman Counterargument 0.115*** 0.306*** 0.314*** 0.208*** 0.0565*** (0.0351) (0.0427) (0.0279) (0.0419) (0.0200) Male R.*Man Counterargument 0.151*** 0.340*** 0.296*** 0.171*** 0.0420*** (0.0435) (0.0392) (0.0310) (0.0399) (0.0153) Female R.*Woman Counterarg. 0.113*** 0.311*** 0.318*** 0.203*** 0.0552*** (0.0287) (0.0371) (0.0271) (0.0345) (0.0165) Female R.*Man Counterargument 0.0558*** 0.208*** 0.328*** 0.299*** 0.109*** (0.0167) (0.0407) (0.0266) (0.0430) (0.0327) Observations 380 380 380 380 380 Table B-11 Marginal predicted probabilities for model 3: five categories of the dependent variable, by gender, counterargument, and self-esteem interaction (for high self-esteem) (1) (2) (3) (4) (5) VARIABLES Def. NOT Very Unlikely Somewhat Unlikely Somewhat Likely Very Likely/Definitely Male R.*Woman Counterargument 0.195*** 0.366*** 0.272*** 0.137*** 0.0309*** (0.0500) (0.0332) (0.0331) (0.0329) (0.0111) Male R.*Man Counterargument 0.106*** 0.294*** 0.318*** 0.220*** 0.0619*** (0.0281) (0.0373) (0.0264) (0.0368) (0.0181) Female R.*Woman Counterarg. 0.136*** 0.336*** 0.305*** 0.178*** 0.0452*** (0.0431) (0.0432) (0.0336) (0.0413) (0.0163) Female R.*Man Counterargument 0.106*** 0.302*** 0.322*** 0.211*** 0.0588*** (0.0347) (0.0482) (0.0293) (0.0449) (0.0209) Observations 380 380 380 380 380 "@en ; edm:hasType "Thesis/Dissertation"@en ; vivo:dateIssued "2017-09"@en ; edm:isShownAt "10.14288/1.0354400"@en ; dcterms:language "eng"@en ; ns0:degreeDiscipline "Political Science"@en ; edm:provider "Vancouver : University of British Columbia Library"@en ; dcterms:publisher "University of British Columbia"@en ; dcterms:rights "Attribution-NonCommercial-NoDerivatives 4.0 International"@* ; ns0:rightsURI "http://creativecommons.org/licenses/by-nc-nd/4.0/"@* ; ns0:scholarLevel "Graduate"@en ; dcterms:title "Discursive equality"@en ; dcterms:type "Text"@en ; ns0:identifierURI "http://hdl.handle.net/2429/62658"@en .