UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Epistemic dependence and autonomy in justification: the case for intellectual autonomy in schools Ross, Murray 1994

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1994-953918.pdf [ 5.81MB ]
Metadata
JSON: 831-1.0088052.json
JSON-LD: 831-1.0088052-ld.json
RDF/XML (Pretty): 831-1.0088052-rdf.xml
RDF/JSON: 831-1.0088052-rdf.json
Turtle: 831-1.0088052-turtle.txt
N-Triples: 831-1.0088052-rdf-ntriples.txt
Original Record: 831-1.0088052-source.json
Full Text
831-1.0088052-fulltext.txt
Citation
831-1.0088052.ris

Full Text

EPISTEMIC DEPENDENCE AND AUTONOMY TN mSTIFICATION:THE CASE FOR INTELLECTUAL AUTONOMY IN SCHOOLSbyMURRAY ROSSB.A., The University ofBritish Columbia, 1973M.A., The University ofBritish Columbia, 1987A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUiREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDiESDepartment of Social and Educational StudiesWe accept this thesis as conforming to the required standardTHE UNIVERSITY OF BRITISH COLUMBIAAugust 1994©Murray Ross, 1994In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)Department of ¶&c&&The University of British ColumbiaVancouver, CanadaDate &A.4St 22DE-6 (2/88)11ABSTRACTIn this thesis I outline a conception of intellectual autonomy, and defend it againstarguments that question the value of intellectual autonomy as an educational goal. Beyondoffering a view of intellectual autonomy that is free of the defects found in other conceptions, Iassess the rationality of belief on testimony, and consider the impact that externalist theories ofepistemic justification might have on arguments such as mine that defend autonomy injustification. I argue that belief on testimony is more rational than is typically allowed, and thatintellectual autonomy cannot be justified solely in terms of the conditions necessary forknowledge or justification. I argue instead that intellectual autonomy is best conceived as anepistemic virtue, and best fostered by introducing students to practices of epistemic justification,in classroom environments characterized by a commitment to open dialogue and debate.111TABLE OF CONTENTSAbstract iiTable of Contents iiiAcknowledgments vIntroduction 1Chapter One: Intellectual Autonomy 8Intellectual Autonomy and Freedom 9Epistemic Dependence 11Barriers to Fostering Intellectual Autonomy 16The Core Pedagogy and the Culture of Teaching 16The Traditions of Schooling 18Teacher Talk 20Strategic Dialogue: The Teaching Exchange 20The Pedagogy of Social Control 23The Epistemic Sensibility of Students 25The Socialization of Student Teachers 27Chapter Two: The Neglect of Epistemic Considerations in Research on Teaching 43Strategic and Logical Acts of Teaching 44Process-Product Research 45Skill and the Pedagogy of Practice 48The Neglect of Reasoning in Classroom Management 50Evidential and Non-Evidential Styles of Belief 51Reasons and Motivation 55Thinking Skills 62Dispositions as Virtues 66Knowledge and Thinking 66The Discipline of the Norm 73Chapter Three: The Contested Nature and Value of Autonomy 83Intellectual Autonomy and the Will to Heterarchy 85Intellectual Autonomy as a Goal of Education 86The Austinian Tradition 88Scheffler and the Evidence Condition 89Green and Evidentially-Held Belief 90Hirst and the Forms of Knowledge 91Siegel on Being Appropriately Moved by Reasons 92Epistemic Paternalism 92The Rationality of Epistemic Dependence 96Externalist Theories of Justification 103Internalist and Deontic Conceptions of Justification 103The Externalist Critique of Internalist Justification 105ivEpistemic Patemism. 108Internalism and the Problem of Infinite Regress 110Chapter Four: The Case Against Externalist Theories ofKnowledge and Justification 119The Internalist Response 125The Surplus Value of the Internalist Condition 131Acquiring the Concept ofBeing Justified 136The Virtues ofEvidentially-Held Belief 138The Prevalence of Unreliable Testimony 142Commitment to Belief 145Chapter Five: Autonomy in Justification Defended 152Belief on Testimony 154Rational Constraints on Accepting Testimony 155Epistemic Paternalism 162Indoctrination 172Chapter Six: Classroom Discourse and the Public Use ofReason 180The Rationality of Belief on Testimony Reconsidered 181Public Reason in the Classroom Context 183Habermas: Discourse Ethics 185The Interest Relativity of Knowledge 190Objectivity 200Vindicating Reason by Means of Dialogue 208Concluding Remarks 214Bibliography 221VACKNOWLEDGMENTSI would like to acknowledge the contribution several individuals have made to thisdissertation. I am indebted to LeRoi Daniels and Murray Elliott for their constructive criticismand support. I am especially indebted to Jerrold Coombs for his careful and patient guidance overthe last several years. I wish to thank Mark Selman for his unflagging interest in my work, andfor first posing the questions that this dissertation seeks to answer. Special thanks go also toRobin Van Heck and LeRoi Daniels for their kindness and generosity, and to Tony Varga for hisinvaluable assistance with the final preparation of this manuscript. Finally and especially, I wantto thank Linda Darling for all her patient support and encouragement. My debt to her isincalculable.1INTRODUCTIONVirtually every educational jurisdiction in North America professes a conmiltment tofostering rational, independent thought in the young. This commitment includes, in the language Ihave chosen to employ, a commitment to fostering intellectual autonomy. Given much traditionalschool practice, however, and the influence of misguided bodies of educational theory andresearch, the expectation that this commitment will soon bear fruit is little more than a pioushope. It is not my intention to offer a comprehensive descriptive account of public schooling’sfailure to foster intellectual autonomy. It is rather to clarifj first the nature of intellectualautonomy and the problem that confronts educators who are interested in pursuing this goal.Following this it is also my aim to offer suggestions for the direction future research might takein establishing the empirical relations between schooling practices and the development ofindependent rational thought. My discussion will, in the main, be conceptual and philosophical,though it will rely on empirical studies for the purpose of illustrating problems and problemsituations. The central problem that confronts those committed to fostering autonomy is not onlyone of determining the most fruitful conception of autonomy, but also one of redressing thepernicious influence of acute epistemic dependence. I will argue that one cannot appreciate thevalue of intellectual autonomy without appreciating the undesirability of acute epistemicdependence.In general terms the problem of acute epistemic dependence is as Dewey expressed itnearly eighty years ago.Men still want the crutch of dogma, of beliefs fixed by authority, to relieve them of thetrouble of thinking and the responsibility of directing their activity by thought. They tendto confine their own thinking to a consideration of which among the rival systems ofdogma they will accept. Hence the schools are better adapted, as John Stuart Mill said, tomake disciples than inquirers.’I am not concerned with the question of whether or not in a given case it is rational tobase one’s beliefs on the authoritative testimony of others, for in many cases it would be foolishto do otherwise. I want to consider the more general question of whether the neglect ofevidential reasoning in schools is a wise policy. A habitual reliance on authorities in the fixing of2belief may well leave individuals unfit to decide for themselves what they ought to believe.2Insofar as schools foster such a reliance they are indeed better adapted to developing in studentsthe heteronomous condition of the disciple than the condition of critical autonomy we recognizein those who can think for themselves.In this sentiment Dewey echoed Kant’s 1784 remarks in What is Enlightenment?, butwith an important difference. While Kant also believed that human beings too often depend upon“the crutch of dogma”, and fix their beliefs on the basis of appeals to authority, he did not takethis state of affairs to be due to a wholly ‘natural’ human disposition to be lazy. Rather the latenttendency toward laziness in grounding belief was, in Kant’s view, a rather small part of theoverall phenomenon of heteronomy. Laziness of the sort in which Kant was interested is aproduct of the social relations between intellectual authorities and those they seek to instruct.In What is Enlightenment? Kant sets out the dichotomy between autonomy andheteronomy and argues that it is heteronomy, not the absence of reason, that accounts for asignificant category of intellectual immaturity. This is the basic distinction that I will employ inmy discussion of autonomy. While there can be considerable philosophical disagreement over theexact meaning of autonomy or heteronomy, especially in their relation to rationality, it isuncontroversial, I think, to hold that there is a clear and coherent distinction to be made betweenwhat it is to hold one’s beliefs on the basis of authority and what it is to hold them on the basis ofgood reasons. There is also a clear distinction to be made between persons who habitually deferto authorities in their thinicing, and persons who endeavor to weigh evidence, assess arguments,and justi1y the beliefs they hold. The latter are “appropriately moved by reasons”, while theformer are not. The former are in a state of acute epistemic dependence; the latter in a state thatdisposes its members toward the pursuit of what Kant called enlightenment. My aim in this thesisis to point the way to a conceptualization of both autonomy and immaturity that makes plain thedanger of heteronomous modes of thought, and coercive classroom practices, while illustratingthe desirability of enlightenment in Kant’s sense of the term.Kant locates autonomy within an interlocking network of concepts such asenlightenmen4 immaturity, heteronomy and tutelage. Significantly, the discussion of autonomy3in What is Enlightenment? is framed within pedagogical metaphors that stress several dangersinherent in situations characterized by teaching and learning.Enlightenment is man’s release from his self-incurred tutelage. Tutelage is man’s inabilityto make use of his understanding without direction from another. Self-incurred is thistutelage when its cause lies not in lack of reason but in lack of resolution and courage touse it without direction from another.... Laziness and cowardice are the reasons why sogreat a portion of mankind, after nature has long since discharged them from externaldirection, nevertheless remains under lifelong tutelage, and why it is so easy for others toset themselves up as their guardians. It is so easy not to be of age. If I have a book whichunderstands for me, a pastor who has a conscience for me, a physician who decides mydiet, and so forth, I need not trouble myself3There are a variety of important considerations worth examining here. One is the claimthat a significant portion of mankind remains immature though ‘nature has long since dischargedthem from external direction’. The reason Kant gives for this sort of immaturity is ‘laziness andcowardice’, not as one might expect, the inability of the immature to reason competently. Kanthere is concerned about the suppression of rationality and the formation of heteronomous habitsover pre-existing rational capacities. While it may be granted that this remark of Kant’s is anempirical claim that may or may not have been true in Kant’s day, let alone in our own, theinteresting conceptual point is that the absence of certain intellectual virtues incapacitates thosewho would otherwise be capable of reasoning adequately. It also hints that there may be coerciveor otherwise threatening circumstances that lead persons to fear the consequences of thinking foroneselfImmobilized by either fear or laziness the immature lead lives of self-incurred tutelage andremain dependent on others to do their thinking for them. It is a condition, says Kant with someirony, that is introduced and maintained by self appointed “guardians” who have “so kindlyassumed superintendence” over the immature. This immaturity is self-incurred insofar as it is dueto a failure of the will to resist superintendence, but it is imposed in the sense it is the product ofa misguided paternalism. The way out of this state of immaturity, in Kant’s view, is to begin tothink for oneself a task that is only possible under conditions of freedom. “Indeed if onlyfreedom is granted, enlightenment is almost sure to follow.” The argument is simple: to be ableto think critically one must be free in certain important respects. The links between autonomy and4freedom, heteronomy and authority postulated by Kant raise a question of educationalsignificance. What is ‘natural’ in the immaturity of the young and to what degree is immaturity aproduct of educational practices that give authority too large a place in the life of the school?Michel Foucault and Jurgen Habermas, in their respective ways, also stress the necessityof certain types of freedom in the development and exercise of intellectual autonomy. Both haveattempted to expose the mysti1,iing influence of power and coercion in human affairs, thoughHabermas has done more to outline how we can rid ourselves of coercive communicativepractices.4Foucault holds a view of immaturity consistent with Kant’s: namely that “immaturityis a certain state of our will that makes us accept someone else’s authority to lead us in areaswhere the use of reason is called for.”5 Foucault’s interest in immaturity is very much related tohis work on the coercive effects of what he calls normalization, and the undue influenceintellectual authority may enjoy in a regime of truth. Foucault has elaborated his notion ofdiscipline in an effort to demonstrate the operation of a new modality of power that is exercisedover individuals simply by comparing them against sets of norms. I will use Foucault’s work toargue that disciplinary practices based on research-based norms of effective teaching can, inprinciple, be used to maintain immaturity and docility in both students and teachers. For thisreason I will argue that these practices and the research from which they have been developedought to be viewed with suspicion by educators interested in fostering intellectual autonomy.The imperfections that a Foucaldian analysis may reveal, can, I think, be traced toconceptions of teaching, motivation, and learning, that fail to conceive of students inappropriately teleological terms, and thus simultaneously fail to emphasize the role of reasons inthought and action. In place of an emphasis on reasons, educational research and theory has toofrequently encouraged teachers to adopt quasi-causal models of human behavior that do little tohighlight the moral and epistemological hazards of neglecting the role of reasons in theconstitution of beliefHabermas’ work on communication and rationality makes it possible to distinguishbetween communicative acts that aim at understanding (communicative action) and those thataim at dominance (strategic action). In light of this distinction my purpose is to argue that too5many of the communicative ventures of teachers are characteristically strategic, thoughsuperficially at least they may appear to be communicative or otherwise benign. Insofar as thespeech acts of teachers tend to be more strategic than communicative in the Habermasian sensethey do little to introduce students to the standards of rational discourse, and may wellperpetuate the practices of distorted communication. I will argue further that Habermas’conception of the ideal speech situation, free of coercive elements, is a regulative ideal by whichwe can judge teachers’ communicative ventures and begin to construct an ideal model ofclassrooms and schools as uncoerced communities.In addition to these concerns there is one final consideration. This inquiry is alsomotivated by a concern that the prospects for fostering autonomy will not be improved bycurrently influential conceptions of thinking and its relation to knowledge. I have in mind here theview of thinking that posits a sharp dichotomy between thinking and knowing. The assumption ofa dichotomous relation between the two may encourage teachers to view the acquisition ofknowledge as an educational objective distinct and perhaps even subordinate to the developmentof thinking abilities and skills. This way of conceiving of knowledge and thinking tends in myview to obscure the epistemic requirements that ought to inform teaching practice. Without clearand frequent access to these requirements that guide and define the rational formation of beliefstudents must have fewer opportunities to develop the dispositions and abilities that areconstitutive of intellectual autonomy.The following four questions are to be the main considerations I want to entertain in thisthesis.1) How should we interpret the notion of intellectual autonomy in the context of consideringeducational goals?2) As an educational goal, how is intellectual autonomy similar to and different from criticalthinking?3) What is the relationship between the goal of developing intellectual autonomy and that ofacquiring knowledge?4) What educational arrangements are inconsistent or incompatible with the goal of intellectual6autonomy? What criteria can serve to help us pick out educational arrangements that arecompatible with intellectual autonomy?Briefly, in summary, the answer to these questions is this. The development of intellectualautonomy requires educational arrangements, in some respects free of compulsion, that givestudents access to the epistemic standards necessary for intelligent thought and action. Therichest repository of these standards is to be found in what Popper has termed the third world ofknowledge. Hence there is a direct relationship between developing intellectual autonomy andacquiring knowledge. The term intellectual autonomy stands roughly for the ability to think foroneself; where ability is understood to include the inclination to do so. In this it is nearly identicalto critical thinking. My reason for employing intellectual autonomy rather than the more familiarcritical thinking is that I wish to emphasize the dangers of basing one’s judgments on authority.In particular I want to stress the pernicious influence of pedagogical practices that directly orimplicitly offer authority as the ground for belief. In this sense my argument is an essay in theethics of belief; and afortiori in the ethics of teaching propositional knowledge.The argument will also turn on the question of whether freedom in some form is anecessary condition of the development of rational autonomy. I will say that it is. I will describeseveral trends in education that threaten autonomy, either by imposing heteronomy on teachers,or by maintaining it in children. And I will sketch a model of classroom practice that promises todo more to preserve and extend the autonomy of students and teachers in schools than presentarrangements. Along the way I will need to contend with recent arguments that suggest that thevalue of intellectual autonomy is overrated and misplaced.6On this view the modern tradition ofepistemic individualism that runs from Kant, Mill and Locke through to the present day is littlemore than a romantic dream that bears scant resemblance to what in fact competent thinkersactually do when deliberating about what to believe. Epistemic justification, it would appear, isnot an individual, but a social phenomenon, built largely on a trust of cognitive authorities. It isargued that a more acceptable view ofjustification than the classical and highly individualist oneis one that admits that trust in cognitive authorities is the basis of most well founded belief I willdeal with this argument by admitting the social element of epistemic justification, while denying7that a conception of epistemic justification based on appeals to authority is satisfactory foreducational purposes. My discussion of epistemic justification is also meant to engage the well-known early arguments ofR. F. Dearden that warn us of the dangers of abandoning the ideal of aliberal education in our pursuit of a chimerical autonomy for students. In the view of intellectualautonomy that I am espousing Dearden’s earliest objections to child-centred construals ofautonomy must be beside the point. This is so because on my view a liberal education thatproperly initiates students into the several traditions of inquiry is the primary means by whichautonomy ought to be developed. Dearden’s more recent views,7 while close to those beingadvanced here, do not insist on the social nature of deliberation and inquiry as I will do.‘John Dewey, Democracy and Education (New York: Macmillan, 1916), 339.2Not all reliance on authorities is unreasonable, of course. The point of distinguishingbetween acute epistemic dependence and less troublesome forms of epistemic dependence isprecisely to get at the problems associated with basing one’s form of life on thoughtlessacceptance of the testimony of others.3lmmanuel Kant, “What is Enlightenment?,” in Foundations of the Metaphysic ofMorals,trans. Lewis White Beck (New York: Liberal Arts Press, 1959), 85.4Michel Foucault, “What is Enlightenment,” in The Foucault Reader ed. Paul Rabinow(New York: Pantheon, 1984); Jurgen Habermas, The Theory of Communicative Action, 2volumes (New York: Beacon Press, 1984).5Foucault, “What is Enlightenment?”, p. 34.6See for example John Hardwig, “Epistemic Dependence,” Journal ofPhilosophy 82, no.7 (1985): 335—349; Stephen Stich & Richard Nisbett, “Justification and the Psychology ofHuman Reasoning,” Philosophy ofScience 47 (1980): 188—202; W. H. Walsh, “Knowledge in itsSocial Setting,” Mind 80, 319 (1971): 321—336; Alvin Goldman, “Epistemic Paternalism,” TheJournal ofPhilosophy 88, no. 3 (1991): 113—131; “Foundations of Social Epistemics,” Synthese73 (1987): 109—144.7R. F. Dearden, “Freedom and the Development of Autonomy: A Reply to VictorQuinn,” Journal of Philosophy of Education, 18, no. 2 (1984): 271—273; R. F. Dearden ed.“Autonomy and Intellectual Education,” Theory and Practice in Education (London: Routledgeand Kegan Paul, 1984), 110—122; R. F. Dearden ed. “Education and the Ethics of Belief”Theory and Practice in Education London: Routledge and Kegan Paul, 1984). 97—109.8CHAPTER ONEINTELLECTUAL AUTONOMYIntellectual autonomy and the means by which its development is impeded is the centraltheme of this thesis. In this chapter I intend to outline what is generally taken to be the nature andvalue of intellectual autonomy, and consider the prospects for fostering intellectual autonomy intoday’s schools. There are several accounts of autonomy that capture what I have in mind, but forthe most part it is Kant’s discussion of intellectual autonomy in What is Enlightenment? that willbe the foundation of my account.In the simplest of terms autonomy means self government. In political philosophy self ruleis roughly equivalent to sovereignty and freedom from external interference. Political autonomy isthus largely conceived in terms of negative freedom. Personal autonomy is another matter. Whilethe issue of sovereignty remains a part of the concept of personal autonomy, the notion ofexternal interference is only one aspect of a very complex term. We can see this in the followingtwo definitions. Wolfi’s definition, for example, does emphasize negative freedom frominterference (“The autonomous man insofar as he is autonomous is not subject to the will ofanother”), while Telfer’s emphasizes freedom from emotional dependency (“An autonomousagent must not have to depend on others for being told what he is to think or do.”2) Bothexamples refer to the subjugation of sovereign will, but in the latter the threat of domination isinternal, the product of a weak will. There is also the hint that autonomy consists not only infreedom from one’s weaknesses, but freedom to think for oneself It is not perfectly obvious whatlink there might be between the negative freedom associated with sovereignty, and the positivefreedom to think for oneself One of my tasks in this thesis is to elucidate what I take this relationto be. What sort of freedom is necessary to safeguard the freedom to think for oneself? Anotherof my tasks is to explain why the freedom to think for oneself is a freedom worth having. It is notobvious that thinking for oneself is a wiser policy than depending on others who may know betterthan you what it is reasonable to believe.What we might call epistemic individualism is a common theme in philosophical writing.Scanlon’s remarks are typical of the sentiment that one should think for oneself9The autonomous person cannot accept without independent consideration the judgment ofothers as to what he should do. If he relies on the judgment of others, he must be preparedto advance independent reasons for thinking their judgment likely to be correct.3Scanlon here underlines what he sees as the close relation between independence of mindand the rational formation of belief This context (ofjudging the acceptability of testimony) is veryclose to the context in which students find themselves. Students are told what to believe (as itwere), and, in a special sense, they have a choice in what they will believe.4They may accept whattheir teachers tell them, or they may refuse to believe their teachers. They may suspendjudgement, or they may feign acceptance to avoid trouble, and replace genuine belief with whatScheffler calls “behavioural manifestations of belief’ . When it comes to the question ofintellectual autonomy, it makes a difference on what basis students come to accept what they aretold. To the extent persons simply accept what they are told, they are not autonomous, onScanlon’s view. Indeed on most accounts a person is said to be lacking in intellectual autonomywhen he blindly follows peer pressure, tradition, or fads. And yet one could act in conformity withpeer pressure without acting on the basis of peer pressure, and thus be autonomous still. On theother hand impulsive rebelliousness against tradition or parental authority hardly counts asautonomy either. This is so because there is something about the quality of one’s deliberationsthat distinguishes the autonomous person from the non-autonomous. In most accountsautonomous deliberation is distinguished by its reliance on impersonal, objective standards ofthought or action. The compulsive gambler or the drug addict is governed more by his desiresthan by such standards, and is therefore a paradigm example of someone who lacks autonomy.INTELLECTUAL AUTONOMY AND FREEDOMStanley Benn has made a number of useful distinctions regarding the relationship betweenautonomy, freedom and self-rule.6On his view freedom is an extremely complex concept, whosestructure involves relations between other concepts such as “authority,” “rights,” “will,”“autonomy,” “self-realization,” and “personhood.” The ideally free person, he says, satisfies threeconditions, one of which is autonomy. The other two are the capacity for free agency, and thecapacity for freedom of action. Benn terms the capacity for free agency “autarchy”, and means bythis “the ability to make decisions in light of reasons and to act on them.”7 Typical among the10defects that militate against the development of autarchy are varieties of impulsiveness. One maybe governed by “inner-impulsions” brought on by paranoia, kleptomania or other neuroticdisorders. Or one may be governed by “other-impulsions”, what Benn terms “heterarchy”.Heterarchy is a condition of submissiveness or suggestibility due, in the case of children, to a blindtrust in adults, especially familiar adults. Heterarchy in adults is more often due to a weakness ofthe will, or the domineering force of another’s will. Autarchy is not the same as autonomy,however. While autarehy is a feature of many mature adults, autonomy is an ideal that few, if anypeople, actually instantiate. At best autonomy can be approximated to various degrees at differenttimes in the lives of some, but not necessarily all people. It consists in the use of one’s autarchy,that is, in basing one’s decisions about what to believe or do on a coherent set of reasons that onehas critically examined and made one’s own. There are two ways of falling short of this ideal. Thefirst is to act “on impulse or on whim, not because [one] is inner impelled or compulsive, butbecause [one] acknowledges nothing as a reason for doing otherwise.”8 Persons who act onimpulse due to an utter lack of reasonableness are, in Benn’s terminology, anomic persons. Suchpersons have the capacity “to grasp a more considered way of living but have simply neverlearned how to do it or to value anything which living by nomos could offer in return for thepostponement of gratification.”9The last condition enjoyed by the ideally free person is freedom of action. Benn construesfreedom of action as the absence of unwarranted external interference by another. Examples ofexternal interference include, obviously, physical force and threats of violence, as well as lessobvious forms of interference such as the manipulation of one’s beliefs. To the extent one acts soas to escape harm, or where one’s actions are the product of willful manipulation by another thiscondition of freedom is absent. In the chapters to follow one of my aims will be to show that as aresult of common teaching practices children stand a risk of emerging from school in a state ofheterarchy insufficiently diminished in degree from the state they were in upon first enteringschool. Of greatest concern is that students might emerge from school anomie persons, that is, asstudents who have the capacity for intelligent and critical reflection, but who for reasons much todo with school practice have simply never learned how to think for themselves or to value the sort11of life available to those who are intellectually autonomous. I will argue that autarchy is aworthwhile educational goal because of the epistemic benefits associated with thinking foroneself I won’t employ this term however since ‘autonomy’ is the more familiar term. Thecontext of my discussion should make clear whether the intended sense refers to the ideal ofautonomy or the ability to make decisions in the light of reasons.EPISTEMIC DEPENDENCEWhat remains unclear is how to specif’ the criteria that distinguish rational acceptancefrom naive or unquestioning acceptance. To put the question more precisely what distinguishesepistemic independence from epistemic dependence, and to what extent is the former morerational than the latter? This is the problem taken up in Kant’s What is Enlightenment?, and theproblem that will be the focus of this thesis: What, exactly, is the proper relation betweenauthority and intellectual autonomy in the constitution of belief; and what bearing should theanswer to this question have on pedagogical practice? My thesis, therefore, is simultaneously anessay in the ethics of belief and the ethics of communicating belief It centres on questions ofepistemic justification, and takes epistemic dependence as its main problem.’°Enlightenment, for Kant, was a “way out” or “exit” from the sort of immaturity found inpeople unable to think for themselves. Immaturity, on this view, is characterized by the dispositionto let others do our thinking for us, and to let ourselves be ruled by judgments other than ourown. Kant puts the matter this way:Laziness and cowardice are the reasons why so great a portion of mankind, after naturehas long since discharged them from external direction, nevertheless remains underlifelong tutelage, and why it is so easy for others to set themselves up as their guardians. Itis so easy not to be of age. If I have a book which understands for me, a pastor who has aconscience for me, a physician who decides my diet, and so forth, I need not troublemyself I need not think, if I can only pay—others will readily undertake the irksome workfor me.When a book stands in place of our understanding, or a priest in the place of ourconscience we are not autonomous, but heteronomous and ruled externally. Autonomy, then, is acondition where one has the ability and the inclination to reason for oneself Immaturity, on theother hand, is a condition of self-incurred tutelage, self-incurred because “its cause lies not in lack12of reason but in lack of courage and resolution to use it without direction from another.. “12Foucault interprets immaturity in Kant’s argument to mean “a certain state of our will that makesus accept someone else’s authority to lead us in areas where the use of reason is calledfor.”13(emphasis added) This refinement is an important one since some have taken Kant to beprohibiting any reliance on authorities.’4I take Foucault’s point to be that there are circumstancesin which a reliance on authorities is reasonable and times when it is not. Immaturity and the stateof heteronomy consist in the chronic failure to think for ourselves when circumstances requirethat we do. It would be a mistake however to suppose the responsibility for heteronomy restssolely with the immature. The initial condition of the immature is used as a pretext for others “toset themselves up as their guardians.”15Kant rhetorically adds that once these guardians “have sokindly assumed superintendence over them” the immature are transformed into domestic cattletoo fearful and unaccustomed to autonomy to be able to “take a single step without the harness ofthe cart to which they are tethered.”6Nonetheless, Kant maintains that if people are encouragedto take the first step, even the most domesticated can, after a few stumbling attempts, learn tocope on their own. These remarks suggest that the most significant obstacles to the developmentof maturity and autonomy may be conditions that limit their exercise. Further, they suggest thateducators have a responsibility to avoid inhibiting students’ free use of reason, or creating in theirclassrooms conditions of intellectual servitude. In what seems to be a piece of uncharacteristicallyromantic excess Kant suggests that the main condition necessary to overcome immaturity andgain enlightenment is a condition offreedom. Since immaturity is due to a combination of lazinessand timidity on the one hand and the willingness of authorities to maintain individuals in states ofdependence, the way out of immaturity, then, is not simply a matter of deciding to think foroneself it is also a matter of participating in a milieu characterized by freedom from interferencein the use of one’s capacity for reason.The matter of determining the appropriate relation between authority and intellectualautonomy in the constitution of belief is not as simple as it may at first appear. Appeals toauthority are based on a trusting of authorities to know what they are talking about. This sort oftrust has bothered a great many philosophers. Standard logic texts, for example, identilj13arguments from authority as fallacious, though they vary in their description of the circumstancesunder which appeals to authority are problematic. 17 Locke, perhaps, puts this point stronger thanmost, insisting that no appeal to authority has any logical force. Locke’s term for this fallacy,argumentum ad verecundiam, stresses that it is excessive modesty or shame (verecundia) thatleads individuals to doubt their own opinions and defer to those of authorities.18 Interpreters ofLocke, such as Hamblin, note that Locke is not warning us simply against bogus or irrelevantauthorities.19 He is reminding us that any authority, no matter how competent, may be wrong.Moreover, even if these authorities are right, those who depend upon them cannot be said tounderstand what it is they profess to believe on the basis of authority.20When we base our beliefson authorities, says Locke, our understanding is incomplete and of little use to us. We can’t besaid to know that which we don’t understand, and we can no more “know by other Men’sunderstandings” than we can “rationally hope to see with other Men’s Eyes.”2 When our beliefsare taken on trust, it is “opiniatrety” we possess, not knowledge: “Floating other Men’sOpinions.. .makes us not one jot the more knowing.”22Mill’s view is similar to Locke’s, except that Mill was more worried about the undesirableinfluence authoritarian personalities and strictures might have on public debate.23 Those who have“no knowledge whatever of the grounds of the opinion, and could not make a tenable defense of itagainst the most superficial objections” are, says Mill, a danger to the traditions of rationalinquiry.24 Mill is anxious that authoritarian personalities may come to stifle genuine debate bymeans of their characteristic tendency to reject out of hand any beliefs that are inconsistent withreceived opinion. This tendency makes it difficult for the received view to be rejected wisely andeasier for it to be rejected rashly. In a well known passage from On Liberty Mill outlines the chiefdifficulty with restrictions on debate. Each time an opinion is “compelled to silence” we lose theopportunity to discover a hitherto unknown truth, or portion thereof that might emerge from “thecollision of adverse opinions.” Moreover, if well established beliefs are not allowed to be“vigorously and earnestly contested” they will, says Mill, be “held in the manner of a prejudice,with little comprehension or feeling of [their] rational grounds.” Held in this way beliefs tend tobecome “mere formal professions”, “deprived of [their] vital effect on the character and conduct.”14As a consequence they are in danger of being lost or enfeebled.25The Lockean view of argumentum ad verecundiam, according to Schmitt, amounts to arejection of the possibility that there could be knowledge based on testimony, since the idea ofknowledge on testimony “entails that the subject must forego the source’s reason for the belief;whether that reason be understood as a set of beliefs,. . . or an experience.”26 The upshot ofLocke’s view is that there is no way a layperson can be justified on expert testimony. Oneenormous difficulty with this view is that it appears that the courts’ reliance on the testimony ofeither eye-witnesses or experts is, at best, weakly justified, or at worst, not justified at all. Agrowing number of philosophers have, in the last decade, begun to acknowledge the rationality ofbelief on testimony, and have argued that not only are appeals to authority justified, they arerepresentative of what is ideal in the constitution ofjustified belief According to this perspectivethe philosopher’s traditional distrust of appeals to authority is nothing more than a hopelesslyromantic sort of epistemic individualism whose deficiencies should be obvious to anyone who hasbothered to observe how people actually justify’ their beliefs. It is simply a fact that our mostsuccessful forms ofjustification are based on appeals to authority. Given the complex division ofcognitive labour in the modern world we should not want it any other way. A tolerably highdegree of reliability in the pronouncements furnished by experts is ensured by an elaborate systemof cognitive checks and balances within institutions and professions. Admission requirements tohigher levels of education or to membership in professional associations, the conventionsgoverning the acceptance of books or articles for publication, forms of peer review and sanction,laws with respect to perjury and expert witnesses in courts of law are all devised to ensuresatisfactory levels of reliable testimony. The most advanced scientific experiments today arecarried out by international teams of researchers who have neither the time nor the expertise toconfirm the claims of their colleagues on which they all depend.27 No one person is in anepistemically most favoured position. Most of the time, then, we find ourselves in a position ofepistemic dependence with respect to the testimony of others and, it is claimed, there is no reasonto be worried about this. The alternative is a sort of epistemic autism that no one could seriouslyendorse. To the extent the view outlined above is correct there doesn’t seem to be any compelling15consequentialist reason for promoting intellectual autonomy.28So what exactly is objectionable about the sort of intellectual immaturity identified byKant? To say such immaturity leaves us like children, or like oxen tethered to a yoke is not aparticularly informative objection since it is not perfectly clear why, on purely epistemic grounds,immaturity of this sort is such a terrible thing. I want to concede that in many contexts acceptingthe testimony of others is a sensible policy. It is my contention, however, that a general policy ofdeferring to others, even bonafide cognitive authorities, is not a suitable policy for education, forthe simple reason that educationally relevant epistemic goals are not best served by socializingstudents to accept the testimony of others. I have in mind four epistemic goals that, when takentogether, combine into the single aim of filling out one’s system of beliefs with as many truebeliefs as possible.291) The goal of acquiring true beliefs2) The goal of deleting false beliefs3) The goal of abstaining from acquiring false beliefs4) The goal of abstaining from deleting true beliefsWhile in many contexts deferring to cognitive authorities is the most rational strategy forincreasing one’s stock of true beliefs, it cannot be the sole strategy for acquiring true beliefs.Since it is unlikely that any child’s life will be free of occasions where the ability to think criticallyis necessary, it must fall to schools (and others) to develop in children the capacity and inclinationto think for oneself Hence there is a need to develop epistemic independence, or more precisely,autonomy in justification. It is precisely over this point that those who argue for the rationality ofbelief on testimony part company with me and each other. Some argue that schooling is and oughtto be the socialization of students into deference to cognitive authority.30Others who would likelyagree that schools should teach children how to think do not appear to believe that autonomy injustification is a necessary feature of rational belief3’ So there is a philosophical problem to sortout. There is also a large practical problem concerning the lack of congruence between acceptedschool practice and the sort of pedagogy needed to foster autonomy. I want now to turn tocommon educational practice and the influence of certain bodies of educational research that16make the prospects for fostering epistemic independence in schools seem dim indeed. Insubsequent chapters I shall return to a more detailed discussion of the philosophical issues athand.BARRIERS TO FOSTERING INTELLECTUAL AUTONOMYIn this section a portion of my analysis will be devoted to showing the persistence of anoccupational culture of teaching within schools that serves to perpetuate noxious pedagogicalorthodoxies.32Another portion will be devoted to detailing these orthodoxies and revealing whatis objectionable about them. Following this I will argue that the absence of sufficient attention toepistemic considerations in influential educational research and theory leaves the likelihood ofschools fostering intellectual autonomy small indeed.Educational philosophers have long been calling for classrooms that would do more tofoster the virtues and dispositions of the critical, autonomous thinker. More than thirty years agoIsrael Scheffier objected to teaching which, because it failed to appeal to the reason of pupils,could not in any strict sense count as teaching at all.33 Many philosophers of education have sinceargued that instructional activities that do fail in this regard are bound to be indoctrinatory, or atleast only weakly connected with what we might call education.34 Some have called for open-mindedness35,others for reflective skepticism36, or thoughtfllness37,or the disposition to beappropriately moved by reasons38.By suggesting teachers ought to appeal to the reason of pupilsScheffier meant that teachers are obliged to provide pupils with the reason for believing x to bethe case, and to see that pupils come to appreciate the force of that reason in an evidentiaryargument that serves to justify the belief in question. In Scheffler’s words, often quoted, “Toteach is thus, in the standard use of the term, to acknowledge the ‘reason’ of the pupil, i.e. hisdemand for and judgement of reasons, even though such demands are not uniformly appropriateat every phase of the teaching interval.”39Unfortunately, there is evidence that suggests that manyteachers today are no more appealing to the reason of the child than did their predecessors. Nor isthere ample reason to believe that future prospects will be more promising in this regard. Thispessimism is due, in large part, to the influence of “the culture of teaching” on the maintenance ofteaching orthodoxies.4°17The Core Pedagogy and the Culture of TeachingNearly a century of classroom observation in the United States reveals that despitesignificant changes in teacher rhetoric, a “core pedagogy” has remained surprisingly constant.41What has persisted is teacher-centred, whole class lecture, recitation, and drill in the most basiceducational outcomes—although these at times bear little relation to the fine sounding rhetoricand slogans of periodic educational reform.42 Even in schools that have adopted child-centredcurricula and co-operative small group instruction observers have noted a continuation of theformalist presuppositions that underlie more traditional pedagogies.43 Many researchers accountfor the durability of formalist teaching by reference to the concrete details of classroom life. Thestructures of time, space, and teacher-student ratios have changed very little, and over the yearshave taken on a cultural momentum of their own. Viewed from the perspective of criticaltheory, these structures represent the material basis on which the cultural superstructure rests.According to this view, the conditions of teachers’ work have engendered teacher ideologies thattake the concrete details of classroom life as givens to which teachers must adapt.45 Thus theconditions of work (“the classroom press”), are the foundation of the culture of teaching. It isthis culture that helps to determine “what works” in a classroom, and what will count amongteachers as a sensible approach to teaching. Models of teaching that differ substantially from thecore pedagogy are more likely than not to be rejected as impractical. Thus so long as the basicstructures remain more or less the same the tendency will be for the core pedagogy to remainalso, despite its apparent limitations. Hargreaves notes this culturally reproductive aspect ofteaching:The culture of teaching, then, is a culture in which classroom experience is exalted aboveall else in collective discussions of educational matters. It is a culture whose conditions ofexistence in the pressing and recurring immediacy of classroom work and in the isolatedcontext of classroom performance make sustained and shared reflection of a rigorousnature difficult. And it is a culture whose conditions in the allocation of time to differentareas of the teacher’s task, place the classroom at the centre and all else at the periphery ofthis work. Once in motion, the culture of teaching is reproductive and self-generating, butonly as long as the conditions of its existence—the isolation and the constraints of theclassroom, the limited opportunities for reflection, the minimal allocation of statutory timeto non-classroom work—persist and continue to sustain18Thus, the concrete circumstances of the teacher’s job have, over the years, fostered “anethic of practicality” that rules on the suitability of instructional methods.48In the busy classroom press, where the pre-occupation is largely with means andimmediate outcomes, the orientation of teachers is towards the concrete and proceduralrather than the theoretic or reflective. They have an understandable interest in day-to-dayprocedures for maintaining an orderly classroom, but often without critically examiningconsequences or debating longer term ends. New ideas are judged for their immediate andutilitarian impact.49Historically, the teacher’s day has provided few opportunities for critical thought andreflection, and thus few opportunities for teachers to see the relevance of critical reflection totheir work. In recent years the situation appears to have worsened as more responsibilities havebeen added to an already overcrowded teaching schedule. Despite significant improvements inremuneration and training over the last half century, teachers in Canada, Britain and the UnitedStates continue to report that while their occupational role has been expanded to include extranon-instructional committee work, and additional record-keeping, the amount of time in whichthey are to carry out these duties has not been expanded to keep pace. Teachers complain ofhaving more social work responsibilities, more paper work, more meetings and conferences thanever before.5°The increase in administrative tasks are the result of increased demands for teacherand student accountability, while the social work fi.inetions are a response to mainstreamingpolicies and the changing demographics of school populations. In the view of labour processanalysts teachers’ work is undergoing a process of intensification,51 where teachers’responsibilities increase without a commensurate increase in the time or training necessary tocarry out these responsibilities in a competent manner. One effect of the time crunch is anincreased teacher reliance on external expertise.52 In an effort to reduce preparation and markingtime teachers report feeling compelled to turn to externally produced curricular materials that areeasily administered and graded but which do not engage the critical intelligence of students.53Given that the litmus test for the acceptability of new teaching ideas is whether or not theyfit easily within existing routines and resources this process of intensification will do little to moveteaching in the direction needed to foster epistemic independence.19The Traditions of SchoolingIn this century the earliest ethnographic foray into classrooms, that of Stevens in 1912,revealed that pupils typically spent most of their time engaged in recitation and drill, or in copyingnotes from the blackboard.54 Verbal interaction between teachers and pupils took the form ofteachers prompting pupils to recall information. Since then surprisingly little has changed.55Teachers dominated talk in the classroom then and they still do. Students were discouraged fromasking substantive questions and they still are. The characteristics of classroom interaction notedby Stevens are still reported to be the dominant form of teacher-pupil interaction, and definestoday the nature of modem classroom teaching in the West.56As late as 1956 Benjamin Bloom attempted to reduce the emphasis teachers gave to thememorization of facts and principles.57 His highly influential taxonomy of educational objectivesurged teachers and curriculum planners to consider more intellectually challenging sorts ofobjectives than those of factual recall.58 More than a decade later Hoetker and Ahibrandcondemned the prevalence of the recitation method in elementary schools59, while Bellack’s6°investigation into teachers’ questioning techniques revealed that teachers typically ask more than90 per cent of the questions, even in classrooms where the teachers were committed to promotinginquiry. The ratio of teacher talk to pupil talk was such that even if opportunities for pupil talkhad been distributed fairly each pupil in a class of more than 25 would receive less than 1 per centof the available opportunities to talk, ask questions, express doubts, etc. Indeed Bellack et a!.estimated that in an average year teachers asked up to 50,000 questions while pupils asked 10questions or less, usually concerning details of class routine, the layout of notebooks and so on.Teacher questions typically required only a rote or memory response, rather than a reasonedreply.In 1970 Flanders observed that roughly two thirds of classroom time was spent in talking,and two thirds of that was teacher talk.6’These proportions, subsequently known famously as thetwo thirds rule, have been found to prevail still in North American high schools.62 Delamont in1976 found that 50% of teacher talk was spent lecturing and questioning pupils about the contentof the lesson.63 The other half was devoted to establishing and maintaining control of the class. In201978, sixty five years after the Stevens’ study, Edwards and Furlong found teachers still raisedmost of the questions, and that 80 per cent of these questions were oriented to the recall of factsor simple generalizations.64In the same year McHoul obtained similar findings: students askedonly about 10 per cent of the questions, and of these very few were substantive questionsregarding the validity of the knowledge claims that were presented to them in school.65 Bothstudies showed that the importance of having good reasons or grounds for belief was an issue thatwas virtually ignored by teachers, and students consistently failed to ask for such grounds. Ofinterest is McHoul’s additional finding that teachers actively discouraged students from askingquestions that might require them to justify knowledge claims or classroom procedures.66 JohnGoodlad’ s 1983 study of 1000 American classrooms revealed the same teacher preoccupationwith the transmission of facts, and the same student reluctance to examine critically theknowledge claims put before them.67 These findings from North American classrooms have beenconfirmed by similar studies in Sweden, the United Kingdom, the Federal Republic of Germanyand Australia.68 As a group these studies suggest that classroom life is dominated by teacher talkthat fails to give students any sense of what reason there may be for accepting or questioning theclaims put before them. Let’s consider teacher talk in more detail.Teacher TalkTeacher talk can be broken down into three categories: instructional monologues, socialexchanges and teaching exchanges. Teachers ask questions of students in all but the first. In socialexchanges teacher questions aim at organizing and controlling pupil behaviour. In many cases theinterest in controlling pupil behaviour determines the nature and content of the teachingexchanges as well.69 Let us consider the general features of the teaching exchange then turn to theissue of social control generally.Sfrategic Dialogue: the Teaching ExchangeThe teaching exchange typically conforms to what has been termed the Initiate, Respond,Feedback (IRF) cycle, where teachers initiate the questions, students respond, and teachers offerfeedback. Teaching exchanges, like teacher monologues, are concerned with transmitting the21content of the curriculum, but differ from teacher monologues in the amount of participationallowed students. Indeed this teaching strategy sometimes goes by the name “active participation”and is held up as a superior alternative to monologues on the ground of greater pupil engagementand interest.70 It is sometimes favourably compared with the Socratic method of teaching andseen, for this reason, to be a useful way to foster a critical spirit in students.Its critics, however, view the teaching exchange as a device for indoctrinating students,and consider the practice a sham attempt at engaging children’s interest and thoughtfulness.71Given the authoritative status of truth claims advanced by teachers, and teachers’ manipulationand restriction of dialogue and debate in these exchanges students are neither encouraged toassess knowledge claims nor see such evaluative activities as legitimate for themselves. Theteaching exchange, on this interpretation, has the latent function of socializing children to acceptuncritically what they are told to believe.The fundamental problem is one of the unavoidable asymmetry of the relations betweenteacher and pupil during the teaching exchange. First of all, there is a lack of symmetry in theepistemic status of teacher and pupil utterances. The utterances of teachers carry more epistemicauthority than those advanced by pupils. There is further asymmetry in the opportunities teachersand pupils have for talking in class. Teachers talk more often and control who shall talk, when andabout what. They also control what shall count as sensible, relevant and justified by means ofglossing practices called epistemic formulations. Teachers first ask questions, and then putglosses on the students’ responses that serve to regulate the admissibility of their contribution tothe discussion. Of course in one very important sense this is as it should be. Teachers are, after all,both the cognitive and institutional authorities in the classroom. That the relations between pupiland teacher are asymmetrical in these ways is part of the reason for having such a relationship.Nevertheless, an analysis of transcripts from classroom dialogue reveals that teacher questions areoften not sincere efforts to discover student points ofview or engage in critical dialogue. They arerather part of a disingenuous strategy of leading students to predetermined conclusions byforeclosing debate on certain lines of argument, while maintaining the appearance of dialogue.Students eventually come to recognize that the point of such exercises is to discover what the22teacher is driving at (“Guess what the teacher is thinking?”).72 Seen in this light the teachingexchange is little more than an inefficient and circuitous method for transmitting knowledge thatcould otherwise be communicated by teacher monologues. Its primary advantage seems to be itsusefulness as a strategy for keeping students “on their toes.”The concern is that such activities communicate to students a sense that their role is moreto answer questions to the satisfaction of the teacher rather than one of satisf’ing public standardsof warrant and justification. Students’ answers are to be consistent with the teacher’s beliefs aboutwhat is true, though the teacher’s reasons for taking her beliefs to be true are seldom expressedand evaluated. Teachers give the impression they are the sole arbiters of the truth by publiclyevaluating student responses while at the same time discouraging students from evaluating theirown truth claims. At one level the appearance is not one of the teacher imposing her view on thestudents. The teacher instead appears to have integrated student points of view into the dialogue.Edwards and Furlong describe the manner in which the process is made to appear legitimate.The teaching exchange provides a framework into which pupil talk is fitted, and that talk isassessed according to the closeness of fit. Brief pupil contributions are taken as beingrepresentative of the group, and the interaction then proceeds as though the other pupilseither know already, or shared the same and now corrected inadequacies as those whospoke. In its orderliness, and in the shaping of meaning, the interactions can be seen as themanaged product of one of the participants.73The teaching exchange ironically appears to invite dialogue while at the same time itprevents the emergence of any genuine dialogue involving the free exchange and defense of ideas.It manages to do this by means of glossing practices, steering pupil contributions towards theviewpoint held by the teacher but unstated by the teacher. Pupils’ responses that appear to leadthe discussion away from the teacher’s preferred view are implicitly rejected, while pupils’responses closer to the preferred view are employed by the teacher in a variety of ways. Pupilresponses of this sort are either reiterated in their totality or are reformulated for a closer fit withthe teacher’s view.Reformulations take various forms. They range from reiteration of the question, completebut disapproving reiteration of pupil responses, partial reiteration and paraphrases, or in the caseof longer responses, gists and summaries. In the course of providing summaries, gists or23paraphrases teachers frequently alter the meaning of the pupil’s contribution by adding anddeleting details, or putting an interpretation to the response that departs slightly from the original.In each case teachers may register their approval or disapproval of pupil responses obliquely bytone of voice, facial expression and gesture. In these ways teachers channel and shape thediscussion toward a pre-specified conclusion. Worth noting is that these glossing practices oftenamount to putting words in the mouths of students, even as they appear to invite the pupil’s pointof view. They aim less at clarifying pupils’ points of view than at shaping them in conformity withthe teacher’s point of view. Yet such approaches to classroom instruction are commonly offeredas examples of how teachers can encourage pupil participation and develop points of view. Youngnotes the fundamentally manipulative nature of these glossing practices in this way:The foregoing analysis draws our attention to those formulating practices which arestructurally located in teacher reactions to pupil answers and which purport to formulatethe pupil talk itself In these formulations we have a practice which, while forming asmuch a part of the teacher’s contribution to the dialogue as teacher monologue, passesitself off as a versions of the pupil’s contribution.... It is difficult to avoid the conclusionthat the dominant pattern of classroom communication is indoctrinational. In not one, buta multitude of ways, it is structured so as to exclude, repress and prevent exploration ofquestions concerning the validity of the facts and simple generalizations which make upthe bulk of the information transmitted in classrooms.74The Pedagogy of Social ControlResearchers have found that teacher decisions regarding both the selection andorganization of subject matter are strongly influenced by concerns related to the management ofchildren.75 Some of this emphasis can be traced to the advice of educational authorities. In his1984 summary of research on classroom management Brophy suggests that the pace, depth andsubstance of lessons ought to be determined by a concern for minimizing student opportunities formischief76To the dismay of Beyer too many teachers think as Brophy does:Getting students through the lesson on time and in a quiet and orderly manner, frequentlybecomes the primary basis for accepting or rejecting the use of a particular teachingactivity. If a technique “works”, that is, solves the immediate problem at hand, it is oftenperceived as good or appropriate regardless of possible larger consequences or theexistence of alternative approaches.77Other observers of classroom practice have noted that the content of lessons is frequently24“downsized” to fit the perceived imperatives of order.78 Knowledge is broken down into isolatedbits, then organized around instructional objectives that lend themselves to assessment by meansof worksheets, short answer quizzes and end of chapter comprehension questions. Assessmentinstruments such as these are easily administered and measured, and are useful for the purpose ofcontrolling students by means of grading practices. One unanticipated consequence of such anapproach, noted famously by Freire, is that knowledge transmitted in this manner is trivialized andcommodified.79Students are tempted to view knowledge as bits of information that can be tradedin for accreditation or grades. The goal for students becomes one of completing assignments thatrequire only that they find the correct answer to recall questions. Seldom is there any opportunityfor students to challenge or expand upon the points of view presented in their texts. Topics thathold the potential for disagreement are also often avoided by teachers who fear that order may becompromised by raucous debate, or that their own authority may be undermined if they are drawninto the debate and asked to defend their own points of view.80 In Beyer’s view, “being a studentmeans acquiring. . . knowledge and learning how to use it in a context that does not includecriticism and has little patience with analysis.”8’The preoccupation with order is not entirely misplaced, however. At one time or anothermost students find their personal and social interests are either not satisfied in school, or directlyconflict with the teacher’s instructional plans.82 Yet for most students attendance at school iscompulsory. Students often react to this state of affairs with behaviour ranging from passive toopen resistance, the latter most common in working class secondary schools.83 Willis, forexample, refers to the “aimless air of insubordination” and the manifestations of “cagedresentment” that he encountered in working class British youth.84 By most accounts teachersrespond to this resistance by assigning work that is cognitively undemanding, easy to grade, andtakes little time to explain.85 The very tasks that would contribute to the development of thinkingability—infrequently assigned long term tasks of sufficient complexity to demand criticalthought—are the very sort of assignments teachers avoid giving. In Woods’, as well as otherBritish studies, teachers have been reluctant to employ instructional activities that foster criticalthought because of the opportunities for mayhem such activities are felt to generate.86 In the25United States, Cusick87 as well as Powell, Farrar and Cohen88 have documented the uneasyequilibrium of some schools where students exhibit a perfunctory compliance with school normsin exchange for an academically undemanding course of studies. In both Britain and the UnitedStates the undemanding character of much schoolwork can also be attributed to the imperatives ofgrading assigned work. As noted, school grades and credits are one of the most powerful socialcontrol mechanisms available to teachers. Yet if not handled properly the distribution of gradescan present control problems of its own. Compliance is gained by distributing grades frequentlyfor schoolwork that requires no more than normal effort and perseverance. Schrag concludesfrom his analysis of interviews with “difficult” students:If students feel that a decent grade is dependent primarily on their native aptitudes orintelligence, on extraordinary effort rather than on the modicum of work they are willingto undertake, the less talented and ambitious will lose their incentive to comply.9°McNeil observed a similar phenomenon in Wisconsin high schools. She noted that thesocial studies teachers in her study tended to reduce the content of their lessons to disjointed listsof information. The teachers themselves were quite conscious of downsizing the curriculum in aneffort to control unruly students. In their view more challenging or complex assignmentsantagonized the less able, while the time needed to explain such assignments provided too manyopportunities for trouble to break out. As an activity, furnishing students with undemanding listsof facts and generalizations that they then write down and memorize has the virtue of stabilizingthe classroom and making it more predictable.The Epistemic Sensibility of StudentsOne finding from the 1985 PEEL Project in Australia points to the effect the corepedagogy may have on the epistemic sensibilities of students.9’During the course of the project aclassroom teacher wrote a set of notes on the board, that the students copied down. The notesrelated to weather patterns in Australia, except they contained nonsense terms and patentfalsehoods.92 Prior to his writing on the board the teacher stressed the importance of askingquestions and of thinking critically. The notes would have appeared to the students as comingdirectly from the textbook that the teacher was holding while at the board. This procedure was26repeated in two different classes of high school students. In each class there was no more than onequestion, despite the repeated call for questions. One question asked for the meaning of a term,while the other hesitantly raised doubts about the truth of one of the claims (e.g. soil evaporates inwarm weather.) Once apprised of the situation the chagrined students admitted they were so usedto accepting uncritically the things they were told that they failed to notice anything wrong withthe notes on the board. It is perhaps useful to pause and reflect on the significance of thisexample. It seems clear that in one sense the students in this class did not believe what they werebeing told by their teacher. That is, on the view that belief is any proposition or set of propositionsto which one would assent it is reasonable to expect few students would admit to believing thatsoil evaporates in warm weather. Yet virtually all the students acted as if they did assent to such aclaim. I would suggest the proper interpretation of this incident is not simply that students tooreadily accept what they are told by their teachers,93 but rather that the students were largelyunaware of or ignored, the epistemic import of someone’s asserting this claim or any other claim.In other words the students did not take reflective skepticism to be among their responsibilities asstudents. It is this assumption that critical intelligence is not called for in school that is mostdisturbing, for it suggests how rare it may be for students to consider the truth of statements theyencounter in school. It also suggests how rare must be the opportunities for developing epistemicresponsibility in students. Regardless of the accuracy of what is taught in school, if we wishstudents to become epistemically responsible we must, at a minimum, show them what suchresponsibility consists in.A further consequence of students simply accepting claims in this manner is that suchclaims will likely have little bearing on the beliefs students already hold. A case in point is the1981 study of first year university physics students that revealed that a surprisingly largeproportion of them did not understand Newtonian physics though their exam results indicated thatin a narrow sense they did. When asked which of two balls, one heavier than the other, wouldreach the ground if released simultaneously the majority of respondents chose the heavier of thetwo.94 Despite having spent a good deal of time learning Newtonian physics, these students wereinclined to give Aristotelian answers to questions regarding gravity. The ideas concerning the27Newtonian theory of gravity were, in Whitehead’s sense, inert.95 If these ideas were understood,they were not understood in a way that made them lively elements in the belief systems ofstudents. These consequences should not surprise us. Mill and many other philosophers since havenoted the relationship between critical reflection and understanding.96Karl Popper distinguishedbetween three levels ofunderstanding.97The lowest level of understanding consisted in possessing“a pleasant feeling of having grasped the argument”.98The intermediate level involved having theability to repeat the argument. The highest level of understanding was characterized by having theability to refi.ite the argument. For Popper, criticism is the heart of understanding. Mill, one of thebest known expositors of this view, held that unless a belief is “vigorously and earnestlycontested, it will, by most of those who receive it, be held in the manner of a prejudice, with littlecomprehension or feeling of its rational grounds.”99 In Mill’s view the truth is most efficientlyarrived at through “the collision of adverse opinions”. ‘°°The research just reviewed suggests thatin many schools there has been a notable absence of any such collision.The Socialization of Student TeachersThe episodes from the PEEL project and the physics class are only two illustrations of aphenomenon we have every reason to believe has been a feature of schools since their inception.The persistence of this feature has been traced to a variety of factors. First, teachers aresimultaneously isolated and, while engaged in their classrooms, busy to the point of distraction.The latter factor, we have seen, predisposes teachers to reject teaching methods whose merits arenot immediately obvious, or whose introduction into the classroom is at all complicated, or likelyto pose management problems. On the other hand, the eggcrate architecture of schools cutsteachers off from their colleagues, and has been credited with establishing and maintainingpeculiar professional norms of individualism and non-interference that reduce the likelihoodinnovative teaching methods will be communicated within the school.’°’ One might expect thatnovice teachers, fresh from teacher’s college, would inevitably bring new ideas into the school.No doubt novices do have an impact on their more experienced colleagues. This impact, however,appears to be slight compared to the influence of experienced teachers on the novice teacher. Tostart with, the teaching styles of novice teachers appear to be much more shaped by experienced28teachers than by university courses on how to teach.102 Indeed no other group is as powerful asocializing agent as the experienced teachers with whom the novice comes into contact in the firstfew years of teaching.’°3The substantive character of this influence is to be found in the models ofinstructional technique exhibited by established teachers, and the responses of these teachers tothe questions and concerns of novices. Hammersley noted a variety of tactics employed byexperienced teachers to disabuse new teachers of ideas that threaten established patterns withinthe school: denigrating the relevance of university coursework to teaching; denigrating theidealism of novice teachers, especially with respect to issues of discipline and control of students;asserting and collectively reinforcing their definition of a particular situation over interpretationsoffered by novices. In the majority of cases these “staifroom ideologies” were communicated bymeans of jokes and humorous stories, staffroom “folktales” as it were, told again and again toconfirm the collective view and deride contrary views.’°4 He also notes that under thecircumstances faced by novice teachers it is only too easy to begin to adopt the collective view.“Whatever one thinks of the views expressed in this staifroom there is enormous influence onindividuals to conform to the collective view. There is reassurance, security, restored pride,personal dignity, power in belonging to a team, in a situation where individuals are working underconditions of. . . constant stress.”105The socialization of novice probationary teachers begins much earlier than their firstpracticum or field experience, however. It begins with the childhood experience of the teacher tobe. Every teacher was once a student, and while in school each underwent what has beenvariously called an “apprenticeship of observation”06 or an “apprenticeship of memory.”107Hanson and Herrington, for example, note that “with their entrance into school as young childrenthe teachers of the future begin their apprenticeship of memory. Students entering [teachers’]college already know what teaching j•”108 Even among those whose memories of school havebeen dimmed by time there will be significant numbers of aspiring teachers who will learn thegeneral features of the core pedagogy from family members and other relatives who arethemselves teachers. Teaching runs in the family.’09The characteristics of teachers may also pose problems for those interested in fostering29independent thought in schools. Research in the United States reveals that large numbers ofAmerican teachers share something of a common personality profile—a profile that can be used toexplain patterns of selective recruitment and retention of teachers, as well as the formation ofintellectual propensities and professional norms within the profession. In broad terms the teachingprofession, in the United States at least, is attractive to, and comprised of significant numbers ofmen and women who are authoritarian, conformist, and “other-directed”.”° These generalizationsconcerning the personality of teachers come from measures of central tendency; exceptionsabound. Moreover the studies from which these generalizations are drawn are neither recentenough, nor sufficiently free of methodological problems to accept at face value. Until more andbetter designed studies are conducted these findings must be viewed with a degree of skepticism.Nonetheless, there is a remarkable degree of agreement among teacher educators that part of thetepid intellectual climate found in many schools may be traced to the characteristics of theteachers who work in them. As well, compared to other professional groups teachers havehistorically been shown to possess a higher proportion of individuals with below averageacademic standing,11’who are disinclined to value independent or critical reflection”2,and wholook to others to tell them what to Teachers with above average academic standing doenter teaching, but they are among the first to leave, usually within the first five years. Attritionwithin the first few years is very high. In some studies as much as forty percent of the noviceteacher cohort quit teaching in the first two years.”4Over ten years the attrition rate has reachedas high as seventy percent.”5 In Canada the situation is often as serious. Recently in Alberta, forexample, it was found that less than fifty percent of teachers in training enter and remain in theschool system for more than three years.’16A further problem is the nature of coursework in education faculties and the willingness ofmany teacher educators to communicate sets of narrowly conceived prescriptions, both of whichcontribute to what has been described as the “tell us what to do” mentality of student teachers.’17Floden notes that the prescriptive character of many teacher education programs fails to“acknowledge the rationality of teachers and places researchers in an undeservedly superiorposition in which teachers were not able to assess the worth ofwhat they were being told.”8The30inconclusive nature of much research on teaching, or the basis on which prescriptions for teachinghave been developed are typically not conveyed to student teachers, thus leaving these teachers inan epistemically dependent position, roughly parallel to the position in which experienced teachershave been shown to leave their own students.”9Where teacher education programs do stress theneed for reflection and independent thought they have not always provided students with a clearsense of the relevant epistemic standards by which intelligent reflection is to be carried out.’2°These results should not be surprising in light of the characteristics of many teacher educators,especially those most directly responsible for teacher methodology classes and practicumsupervision. Research on the characteristics of teacher educators in the US points to markeddifferences between education professors and other university faculty.’2’Most notable is thecomparatively low productivity of teacher educators with respect to scholarship and research.Broudy suggests that the reason for this can be found in the fact that education professors have anexcessively practical orientation that devalues abstract thought and decision-making.’22Lanier andLittle suggest the roots of this anti-intellectualism lie in the social background and career paths ofeducation professors:Faculty in institutions of higher education are expected to value intellectual challenge,questioning, criticism, and conceptual analysis. Advancing higher learning requires thatscholars enter uncharted intellectual territory, and as they explore the not yet known, theymust maintain a cognitive flexibility and conmitment to examine alternative, sometimescompeting, beliefs and assumptions. Diverse views and openness to new evidence, novelideas, and controversial opinions are long accepted values of the academy. Conversely, thetendency to ignore or reject competing ideas and evidence, to accept old or new ideasuncritically, or to proselytize unexamined truths are signs of academic weakness. Evidencesuggests that the typical lineage of teacher educators has not prepared them to appreciatethe traditional values of higher education.’23In addition Tabachnick et al. noted in discussions with student teachers that practicumsupervisors are more intent on socializing student teachers into existing classroom practices thanin encouraging them to view such practices critically.By focusing upon how things are to be done in classrooms to the exclusion of why, theuniversity originated discussions which tended to accept the ongoing patterns and beliefsillustrated earlier. Instead of responsibility and reflection, the actions of universitypersonnel encouraged acquiescence and conformity to existing school routines. The latentmeanings of workshops and seminars were established in a variety of ways. For example,students were continually reminded that they needed to get along if they wanted good31recommendations for their job placement folders.. .The content of supervisory conferencesalso gave legitimacy to existing classroom priorities.’24By contrast, professors in arts and science faculties (these, in a special sense, are alsoteacher educators)’25 are much less concerned with the practical use to which their students’knowledge may be put. Nonetheless, undergraduate teaching, especially in the freshman andsophomore years, is often identified with survey courses characterized by a preoccupation withcovering vast amounts of knowledge in a superficial manner. A close examination of the evidenceor arguments in support of knowledge claims, if found at the undergraduate level at all, is mostoften restricted to undergraduate honors programs. Thus it seems reasonable to suppose that theneglect of epistemic considerations by both pre- and in-service teachers has also much to do withtheir experience as university students as well. Teachers themselves tend not to have beenschooled in the strategies of critical thinking and epistemic justification, so little in theirexperience, either as schoolchildren, or as teachers, will have given them much insight into whichamong their preferred teaching strategies are least appropriate for fostering critical thinking. Thusit seems reasonable to suppose that classrooms now and in the past have not been fertile breedinggrounds for intellectual autonomy. Insofar as the development of intellectual autonomy iscontingent upon students being initiated into the practices of critical reflection the univocal natureof classroom discourse, and the systematic neglect of epistemic considerations must surely beobstacles to the flourishing of autonomy.‘Robert Wolfi In Defense ofAnarchism (New York: Harper and Row, 1985).2Elizabeth Telfer, “Autonomy and Education,” in S. C. Brown ed. Philosophers DiscussEducation (London: Macmillan, 1975).3Thomas Scanlon, “A Theory of Freedom of Expression,” Philosophy and Public Affairs1 (1972): 216.4Strictly speaking we can’t choose what to believe. I can’t choose for example to believethat Calgary is the capital of Canada. The sense of choice invoked here is the choice whether todeliberate or not.5lsrael Scheffler, Reason and Teaching (New York: Routledge and Kegan Paul, 1973).6Stanley I. Benn, A Theory ofFreedom (Cambridge: Cambridge University Press, 1988).327lbid., 131.8lbid., 176.9lbid., 176.10Harvey Siegel, “Rationality and Epistemic Dependence,” Educational Philosophy andTheory 20, no. 1 (1988): 1—6.110ne of many translations, from which this adapted, is “What is Enlightenment?,”published with Immanuel Kant’s Foundations of the Metaphysics ofMorals, trans. Lewis WhiteBeck (New York: Liberal Arts Press, 1959), 85.12Ibid., 85.‘3Michel Foucault, “What is Enlightenment?” Paul Rabinow ed. in The Foucault Reader(New York: Pantheon Books, 1984), 34.14Wolff In Defense ofAnarchism.‘5Kant, What is Enlightenment?, 38.‘6Thid., 39.‘7See for instance Irving Copi’s Introduction to Logic, 6th ed. (New York: Macmillan,1982).18”The first is to allege the opinion of men whose parts, learning, eminency, power, orsome other cause has gained a name and settled their reputation in the common esteem with somekind of authority. When men are established in any kind of dignity, it is thought a breach ofmodesty for others to derogate any way from it, and question the authority of men who are inpossession of it. This is apt to be censured as carrying with it too much of pride, when a man doesnot readily yield to the determination of approved authors which is wont to be received withrespect and submission by others; and it is looked upon as insolence for a man to set up andadhere to his own opinion against the current stream of antiquity, or to put it in the balanceagainst that of some learned doctor or otherwise approved writer. Whoever backs his tenets withsuch authorities thinks he ought thereby to carry the cause, and is ready to style it impudence inanyone who shall stand out against them. This I think may be called argumentum adverecundiam.” John Locke, An Essay Concerning Human Understanding, ed. John W. Yolton(London: Dent, 1961), 278.‘SC. Hamblin, Fallacies (London: Methuen, 1970), 68.20Welbourne (1981) suggests an alternative reading to Hamblin’ s. It is possible that Lockemeans only to say that belief based on authority is less reliable, and uiirther that since it lacks fullunderstanding it is less useful.3321Locke, An Essay Concerning Human Understanding, 278.22Jbid., 278.23Jol Stuart Mill, On Liberty (New York: The Liberal Arts Press, 1956).24Jbid., 13.25Ibid., 15.26Frederick F. Schmitt, “Justification, Sociality and Autonomy,” Synthese 73 (1987): 54.27’t)j H. Walsh, “Knowledge in its Social Setting,” Mmci 80, no. 319 (1971): 321—336;John Hardwig, “The Role of Trust in Knowledge,” The Journal ofPhilosophy 88, no. 12, (1991),693-694; R. T. Allen, “Because I Say So Some Limitations Upon the Rationalization ofAuthority,” Journal ofPhilosophy ofEducation 21, no. 1 (1987): 15—24.28There is of course the well known teleological justification for intellectual autonomywhich originated with Kant. I am more interested in providing a consequentialist justification forautonomy in terms of the epistemic benefits autonomy of this sort confers on those who can thinkfor themselves. The teleological account runs as follows. Because human beings have a capacityfor rational choice, they possess a value that is unconditional. Objects of conditional value, on theother hand, are valuable only under certain conditions: that is, for what they may lead to or bring.Thus, objects of conditional value are valuable only as means to ends. Objects of unconditionalvalue, such as human beings, have value then, not as means, but as ends in themselves. Onaccount of their unconditional value persons are said to deserve respect, hence the intimateconnection between the concept of autonomy and what is called the principle of respect forpersons. To respect persons, on this account, is to see them as unconditionally valuable agents,and so to recognize they should not be treated as the means to our own private ends. To treatpersons merely as a means is to treat them as if they were not rational and moral agents. So it thiscapacity for rational and moral choice that confers a special dignity on persons which other moralagents ought to respect. To reject the considered judgments of other persons, or interfere withtheir right to act on those judgments is to demonstrate a lack of respect. Therefore the principle ofrespect for persons requires that, subject to certain side constraints, moral agents have the right tohold their own opinions and act on them, free from our interference. Those who possess thedignity due to persons are morally entitled to determine their own destinies. The fact that freedomfrom interference is contingent upon one’s possessing a capacity for rational judgment sets avariety of limits on an agent’s right to autonomy. In the first place to the extent individuals lackthis capacity paternalistic treatment is seen to be justified. To the extent an individual’s actionsmay bring harm to others she forfeits the right to non-interference. Autonomy as a right is a rightagainst attempts to interfere with or diminish an agent’s ability to act autonomously. So, forexample, a student’s right to autonomy is violated when she is treated as if she has no capacity forautonomy, or if she is treated in a way which fails to respect what autonomy she does have. It ison this ground that paternalism and indoctrination are often objected to. Indoctrination violatesthe student’s right to autonomy because it undercuts the student’s ability to assess critically her34beliefs and desires. Threats, coercion and psychological manipulation similarly violate the agent’sright to autonomy if it can be shown that the justification for paternalistic treatment fails. Amongthe consequentialist objections that can be leveled against paternalism of this sort is the dangerthat continuous subjection to even mild forms of coercion and manipulation will in all likelihooddiminish the agent’s ability to act autonomously. Even though I am more interested in theconsequentialist argument than the teleological, I have included this discussion of the latterbecause it helps to bring out the manner in which moral and epistemic concerns begin toconverge.29This characterization of epistemic goals is taken from Schmitt, “Justification, Sociality,and Autonomy.”30Stich and Nisbett, “Justification and the Psychology of Human Reasoning”; Allen,“Because I Say So!”31Goldman, “Epistemic Paternalism”; Hardwig, “Epistemic Dependence.”32It is perhaps misleading to speak of a single culture of teaching since there are in factmany. The culture I intend to examine most closely is the one most relevant to the failure ofschools to challenge the minds of students.33Israel Scheffler, The Language ofEducation (Springfield, IL: Thomas, 1960).34Thomas Green, “A Topology of the Teaching Concept,” in Concepts of Teaching:Philosophical Essays, ed. C. J. B. MacMillan and Thomas W. Nelson (Chicago: Rand McNally,1968); The Activities of Teaching (New York: McGraw Hill, 1971); John Kleinig, PhilosophicalIssues in Education (London: Croom Helm, 1982); R. E. Young, “Teaching EqualsIndoctrination: The Dominant Epistemic Practices of our Schools,” British Journal ofEducational Studies 32, no. 3 (1984): 220-238.35William Hare, Open-Mindedness and Education (Montreal: McGill-Queens UniversityPress, 1979); In Defence of Open-Mindedness (Montreal: McGill-Queens University Press,1985).36Robert Ennis, “A Conception of Critical Thinking,” in Proceedings of the Philosophy ofEducation Society 1979, ed. Jerrold R. Coombs (Normal, IL: 1979); John McPeck, CriticalThinking and Education (New York: St. Martin’s Press, 1981).37Francis Schrag, Thinking in School and Society (London: Routledge, 1988).38Harvey Siegel, Educating Reason: Rationality, Critical Thinking and Education (NewYork: Routledge, 1988).39Scheffler, The Language ofEducation, 57—8.owe this contrast to Francis Schrag. Thinking in School and Society, 733541Larry Cuban, “Persistent Instruction: Another Look at Constancy in the Classroom,” PhiDelta Kappan. 6 (1986): 711; Cuban, How Teachers Taught: Constancy and Change inAmerican Classrooms; 1900—1980 (New York: Longman, 1984); Cuban, “Persistent Instruction:The High School Classroom 1900—1980,” Phi Delta Kappan. 64 (1982): 113—118.42Neil Sutherland, “The Triumph of ‘Formalism’: Elementary Schooling in Vancouverfrom the 1920s to the 1960s,” B C Studies no. 69—70 (1986): 175—2 10.435ee Sutherland, “The Triumph of Formalism,” for evidence in the Canadian context, andPeter Woods, Sociology and the School: An Interactionist Viewpoint (London: Routledge andKegan Paul, 1983); Andy Hargreaves, Curriculum and Assessment Reform (Toronto: OISEPress, 1989) for evidence from British sources.44Seymour B. Sarason, The Predictable Failure of Educational Reform (San Francisco:Jossey Bass, 1990).45Peter Woods, “Culture of the School—Teachers,” in Sociology and the School: AnInteractionist Viewpoint (London: Routledge and Kegan Paul, 1983); Peter Woods ed., TeacherStrategies: Explorations in the Sociology of the School (London: Croom Helm, 1980).46Sharon Feiman-Nemser and Robert Floden, “The Cultures of Teaching,” in Handbook ofResearch on Teaching, ed. Merlin C. Wittrock (New York: Macmillan, 1986), 505—526.47Hargreaves, Curriculum andAssessment Reform, 56.48Walter Doyle and 0. Ponder, “The Practicality Ethic in Teacher Decision-Making,”Interchange. 8, 3 (1978): 1—12.49Walt Werner, “Curriculum Integration and School Cultures,” Forum on CurricularIntegration: Occasional Paper #6 (Burnaby: Tn-University Integration Project, 1991), 20.50Quoted in Andy Hargreaves and Rouleen Wignall, Time for the Teacher: A Study inCollegial Relations and Preparation Time Use among Elementary School Teachers (Toronto:Ontario Institute for Studies in Education, 1989).51See Michael Apple, Teachers and Texts: A Political Economy of Class and Gender(New York: Routledge and Kegan Paul, 1989); Andy Hargreaves, “Time and Teachers’ Work:An Analysis of the Intensification Thesis,” Teachers College Record 94, no. 1 (1992): 87—107; S.M. Larson, “Proletarianization and Educated Labour,” Theory and Society 9 (1980): 13 1—175.52Michael Apple and Susan Jungck, “You Don’t Have to be a Teacher to Teach This Unit:Teaching, Technology and Control in the Classroom,” in Understanding Teacher Development,eds. Andy Hargreaves and Michael Fullan (New York: Teachers College Press, 1992), 20—42.53Apple, Teachers and Texts; Hargreaves, “Time and Teachers’ Work.”3654Romniett Stevens, The Question as a Measure of Efficiency in Instruction: A CriticalStudy of Classroom Practice (New York: Teachers College Press, 1912).55Larry Cuban, How Teachers Taught (New York: Longmans, 1984).56Robert E. Young, A Critical Theory of Education: Habermas and Our ChildrenFuture: Habermas and Our Children Future (New York: Teachers College Press, 1990);Francis Schrag, Thinking in School and Society.5’7Benjamin Bloom, A Taxonomy ofEducational Objectives Volume 1: Cognitive Domain.58lnterestingly enough Stevens too had years earlier complained that teacher questionsstressed only memory and the recall of simple generalizations. She also noted that thequestion/answer method generated a destructive sort of nervous tension among students, thoughit was teachers and not students who were doing most of the thinking.59J• E. Hoetker and W. Ahibrand Jr., “The Persistence of the Recitation,” AmericanEducation Research Journal, 6 (1969): 145—167.60A. Bellack et al., The Language of the Classroom (New York: Teachers College Press,1965).61See Robert E. Young, “Critical Theory and Classroom Questioning,” Language andEducation 1, no. 2 (1987): 125—134.62Robert E. Young, “Teaching Equals Indoctrination: The Dominant Epistemic Practicesof Our Schools,” British Journal ofEducation Studies 22, no. 3, 115—145.63Sara Delamont, Interaction in the Classroom (London: Methuen, 1976).64A. Edwards and V. Furlong, The Language of Teaching (London: Heinemann, 1978).65A. McHoul, “The Organization of Turns at Formal Talk in the Classroom,” Language inSociety 7 (1978): 183—213.66McHoul, “The Organization of Turns at Formal Talk,” 197.67Jolm Goodlad, A Place Called School: Prospects for the Future (New York: McGrawHill, 1984).68Young, A Critical Theory ofEducation.69See Walter Werner, “Curriculum Integration and School Cultures,”. Occasional PaperSeries: FOCI, Tn-University Project, 1991); Peter Woods, “Culture of the School—Teachers,” inSociology and the School: An Interactionist Viewpoint (London: Routledge and Kegan Paul,1983).3770Carol Cummings, Teaching Makes a Dfference (Edmonds, WA: Teachers, 1980).7tPeter Woods, “Culture of the School—Teachers,” in Sociology and the School: AnInteractionist Viewpoint (London: Routledge and Kegan Paul, 1983).72Robert Young, Critical Theory and Classroom Talk (London: Taylor and Francis,1993).73Edwards and Furlong, The Language of Teaching (London: Heinemann, 1978).74Young, “Teaching Equals Indoctrination,” 126.75F. Schwartz, “Supporting or Subverting Learning: Peer Group Patterns in Four TrackedSchools,” Anthropology and Education Quarterly 12, 2 (1981): 99—121; D. Eder, “AbilityGrouping as a Self-Fulfilling Prophecy: A Micro Analysis of Teacher-Student Interaction,”Sociology ofEducation. 54, 3 (1981): 15 1—161; R. Alington, “The Reading Instruction ProvidedReaders of Differing Reading Ability,” Elementary School Journal 83, 5 (1983): 548—55976Jere Brophy, “Classroom Management and Organization,” Elementary School Journal83, no. 4 (1983): 265—285.77Linda Beyer, “What Knowledge is of Most Worth in Teacher Education? in EducatingTeachers: Changing the Nature ofPedagogical Knowledge, ed. J. Smyth (London: Falmer Press,1987), 21.78Milbrey Wallin McLaughlin, “What Matters Most in Teachers’ Workplace Context?,” inTeachers’ Work: Individuals, Colleague and Contexts, eds. Judith Warren Little and MilbreyWallin McLaughlin (New York: Teachers College Press, 1993); Thomas Popkewitz, “The SocialStructure of Schools and Reform: A Case Study of IGE/S,” in Qualitative Evaluation: Conceptsand Cases in Curriculum Criticism, ed. G. Willis (Berkeley: McCutchan, 1978), 413—441; JohnGoodlad, A Place Called School: Prospects for the Future (New York: McGraw-Hill, 1984);Carol Evertson, “Differences in Instructional Activities in Higher and Lower Achieving JuniorHigh English and Math Classes,” Elementary School Journal 82, 4 (1982): 329—3 50; R. N. Page,“Lower-Track Classes at a College-Prepatory High School: A Caricature of EducationalEncounters,” a paper presented at the annual meeting of the American Educational ResearchAssociation, New Orleans, 1984); R. A. Hanson and R. E. Schutz, “A New Look at SchoolingEffects from Programmatic Research and Development,” in Making Change Happen?, ed. DaleMann (New York: Teachers College Press, 1978), 120—149; J. Oakes, Keeping Track: HowSchools Structure Inequality (New Haven, CT: Yale University Press, 1985).79PaoIo Freire, Pedagogy of the Oppressed (New York: Seabury, 1970).80Woods, Teacher Strategies, 142.81Jbid., 22.3882Martin Hammersley and Peter Woods eds., Lfe in School: The Sociology of PupilCulture (Milton Keynes, England: Open University, 1984).$3Jjd84Paul Willis, Learning to Labour (Westmead: Saxon House, 1977), 36.851t is a commonplace of sociological research in education that these non-epistemicpractices are more pervasive in the schools serving low income neighbourhoods. See F. Schwartz,“Supporting or Subverting Learning: Peer Group Patterns in Four Tracked Schools,”Anthropology and Education Quarterly 12, no. 2 (1981): 99—12 1; Wilbur Brookover, N. V.Brady and M. Warfield, Educational Policies and Equitable Education: A Report of Studies ofTwo Desegregated School Systems (East Lansing, MI: Center for Urban Affairs, Michigan StateUniversity, 1981); A. G. Powell, E. Farrar and David Cohen, The Shopping Mall High School:Winners and Losers in the Educational Marketplace (Boston: Houghton Muffin, 1985); LoisWeis, Issues in Education: Schooling and the Reproduction of Class and Gender Inequalities(Buffalo, NY: State University of New York, Buffalo, Department of Educational Organization:Occasional Paper Number Ten, 1986); R. A. Hanson and R. E. Schutz, “A New Look atSchooling Effects from Programmatic Research and Development,” in Making Change Happen?,ed. Dale Mann (New York: Teachers College Press, 1978), 120—149.86Woods, Sociology and the School.8Whilip Cusick, Inside High School: The Student’s World (New York: Holt, Rinehart andWinston, 1973).88Powell et al., The Shopping Mall High School.89Woods, Sociology and the School; Linda McNeill, “Defensive Teaching and ClassroomControl,” in Ideology and Practice in Schooling, eds. Michael Apple and Lois Weiss (Boston:Routledge and Kegan Paul, 1983); Hammersley and Woods, Life in School.90Schrag, Thinking in School and Society, 94.91Jim MacKenzie, “Authority,” Journal of Philosophy of Education. 22, no. 1 (1988):107—111.92For example, the notes contained terms like “hygration,” and suggested that soilevaporates. See appendix in Jim MacKenzie, “Authority,”.93This is the interpretation offered in Mackenzie, “Authority”.94R. F. Gunstone & R. T. White, “Understanding Gravity,” Science Education 65 (1981):291—299.95Alfred North Whitehead, Aims ofEducation (London: Macmillan, 1929).3996John Stuart Mill, On Liberty (New York: The Liberal Arts Press, 1956).97Karl Popper, “Intellectual Autobiography,” in The Philosophy ofKarl Popper, ed. PaulArthur Schlipp (LaSalle, IL: Open Court Publishing, 1974).98Ibid., 58.99Mil1, On Liberty, 43.‘°°Ibid., 43.‘°1Sarason, 162—163.‘°2According to Zeichner there is virtually no empirical support for the idea that universitybased supervisors have any significant influence on student teachers. See Kenneth M. Zeichner,“The Ecology of Field Experience: Toward an Understanding of the Role of Field Experiences inTeacher Development,” in Advances in Teacher Education, volume 3, eds. Martin Haberman andJulie M. Backus (Norwood, NJ: Ablex Publishing Co., 1987), 124.103W. Hoy, “Organizational Socialization: The Student Teacher and Pupil ControlIdeology,” Journal of Educational Research 61(1967): 153—155; “The Influence of Experienceon the Beginning Teacher,” School Review 76 (1968): 312—323; “Pupil Control Ideology andOrganizational Socialization: A Further Examination of the Influence of Experience on theBeginning Teacher,” School Review 77 (1969): 257—265; W. Roy and R. Rees, “TheBureaucratic Socialization of Student Teachers,” Journal of Teacher Education 28, no. 1, 23—26;D. J. Willower, “The Teacher Subculture,” Samplings 1 (1968): 45—59; Peter Woods and MartinHammersley, School Experience (London: Croom Helm, 1977); David Lortie, Schoolteacher(Chicago: University of Chicago Press, 1975); Philip W. Jackson, Lfe in Classrooms (New York:Holt, Rinehart & Winston, 1968); Andy Hargreaves, Curriculum and Assessment Reform(Toronto: OISE Press, 1989).‘°4Peter Woods, “The Meaning of Staifroom Humour,” in Classrooms and Staffrooms:The Sociology of Teachers and Teaching, ed. Andy Hargreaves and Peter Woods (MiltonKeynes: Open University Press, 1984), 190—202.‘°5Woods and Hammersley, 48.‘°6Lortie, Schoolteacher, 13107David Hanson and David Herrington, From College to Classroom: The ProbationaryYear (London: Routledge and Kegan Paul, 1976).‘°8lbid., 12.109Judith E. Lanier and Judith W. Little, “Research on Teacher Education,” in HandbookofResearch on Teaching, 3d ed., ed. M. C. Wittrock (New York: Macmillan, 1986), 525—569.40110M. L Kohn and C. Schooler, “Job Conditions and Personality: A LongitudinalAssessment of their Reciprocal Effects,” American Journal of Sociology 87, 6 (1982):1257—1286; Lortie, Schoolteacher; M. L. Kohn, Class and Conformity. A Study in Values(Homewood, IL: Dorsey Press, 1969); Robert Floden, Margaret Buchmann and L. R. Schwille,“The Case for the Separation of Home and School,” paper presented at the annual meeting of theAmerican Educational Research Association, New Orleans, 1984).111Lapjer and Little’s 1986 survey of the research on teacher education concluded that“persons with low measures of academic talent dominate the field” (p. 565). See Judith E. Lanierand Judith W. Little, “Research on Teacher Education,” in Handbook of Research on Teaching,3d ed., ed. M. C. Wittrock (New York: Macmillan, 1986), 525—569. See also D. H. Kerr,“Teaching Competence and Teacher Education in the United States,” in Handbook of Teachingand Policy, Lee S. Shulman and G. Sykes eds. (New York: Longman, 1983), 126—149; V. S.Vance and P.C. Schlechty, “The Distribution of Academic Ability in the Teaching Force: PolicyImplications,” Phi Delta Kappan 64, 1, September (1982): 2—27; P.C. Schlechty and V. S.Vance, “Recruitment, Selection and Retention: The Shape of the Teaching Force,” ElementarySchool Journal, 83, (1983) pp. 469—487.“2Dan Lortie, Schoolteacher (Chicago: University of Chicago Press, 1975); MargaretBuchman and 3. Schwille, “Education: The Overcoming of Experience,” American Journal ofEducation. 92, 1 (1983): 30—51; B. R. Tabachnick, T. S. Popkewitz and Kenneth Zeichner,“Teacher Education and the Professional Perspectives of Student Teachers,” Interchange. 10, 4(1979—1980): 12—29; Walter Doyle and G. Ponder, “The Practicality Ethic in Teacher Decision-Making,” Interchange. 8, 3 (1978): 1—12.‘13Walter Doyle and G. Ponder, “The Practicality Ethic in Teacher Decision-Making,”Interchange 8, no. 3 (1978): 1—12; M. B. Bierly and D. C. Berliner, “The Elementary SchoolTeacher as Learner,” Journal of Teacher Education 33, no. 6 (1982): 37—40; Sharon FeinmanNemser and Margaret Buchman, “Pitfalls of Experience in Teacher Education,” Teachers CollegeRecord 85, no. 1 (1985): 53—67; K. K. Zumwalt, “Research on Teaching: Policy Implications forTeacher Education,” In Policy Making in Education: 81st Yearbook of the National Society forthe Study of Education, A. Lieberman and M. W. McLaughlin eds., Part 1. (Chicago: Universityof Chicago Press, 1982), 215—248; Lortie, Schoolteacher.“4Richard Murnane, Judith Singer, John Willett, James Kemple and Randall Olsen, WhoWill Teach? (Cambridge, MA: Harvard University Press, 1991).115Murnane et al., “How Long Do Teachers Stay in Teaching?,” Who Will Teach?“6L. T. Williams, Developmental Patterns of Teaching Careers. Unpublished DoctoralDissertation, University of Alberta, Edmonton, Alberta, 1986), quoted in R. F. J. Jevne and H. W.Zingle, Striving for Health: Living with Broken Dreams (Edmonton, Alberta: Alberta SchoolEmployee Benefit Plan, 1991).41‘17Margaret Buchmann, “The Use of Research Knowledge in Teacher Education andTeaching,” American Journal ofEducation 92, no. 4, (1984): 421—439; Gary D. Fenstermacher,“On Learning to Teach Effectively from Research on Teacher Effectiveness,” in C. Denham andA. Lieberman eds. Time to Learn (Washington, DC: National Institute of Education, 1980);Zumwalt, “Research on Teaching: Policy Implications for Teacher Education.”‘18Robert E. Floden, “The Role of Rhetoric in Changing Teachers’ Beliefs,” Teaching andTeacher Education.119Frederick Erickson, ed. “Qualitative Methods in Research on Teaching,” in M. C.Wittrock Handbook ofResearch on Teaching, 3d ed. (New York: Macmillan, 1986).‘20A recent example of this phenomenon can be seen in James G. Henderson’s ReflectiveTeaching: Becoming an Inquiring Educator where terms like reflection and inquiry areprogrammatically defined in terms more moral than epistemic, and fitted into an interlockingsystem of fine-sounding but vague slogans. For example consider the following from chapter one,“Reflective Teaching and Educational Inquiry,”:“[Reflective teachers] are sincere and thoughtful professionals who constantly learn fromtheir reflective experiences.,”“The caring teacher takes the time to help all students discover their individual inclinationsand capitalize on them.,”“Caring teachers must also be ‘co-operative educators’. Teachers guided by an ethic ofcaring understand that they can’t practice personal confirmation and honest dialogue unless theywork co-operatively with their students, and perhaps with their students’ parents as well.”“Caring teachers think of themselves as facilitators of learning; they act as counsellors andadvisors in their subject fields and not just as imparters of knowledge.”“The reflective teacher takes this constructivist approach and sees learning as a complexinteraction among each student’s past experiences, personal purposes and subject matterrequirements.”“Teachers who become skilled problem solvers by following an ethic of caring and theconstructivist theory of learning provide a special service in their classrooms.”“Reflective teaching is enhanced by an inquiring attitude toward education. The essence ofeducational inquiry involves taking a questioning, pondering, democratic perspective on thepersonal and public virtues of teaching and learning.”121Lanier and Little, “Research on Teacher Education.”‘22Harry S. Broudy, “What Do Professors of Education Profess? Educational Forum 44,4 (1980): 441—451.42‘23Lanier and Little, “Research on Teacher Education,” 533.‘24Tabachnick et al., “Teacher Education and the Professional Perspectives of StudentTeachers”, 22.‘25By one American estimate four-fifths of a secondary teacher’s required program, andtwo-thirds of an elementary teacher’s program, is provided by undergraduate study in facultiesoutside education. As more faculties of education in the United States and Canada phase outundergraduate degrees in education the proportion of teacher education outside educationfaculties will likely increase. See Wittrock, Handbook ofResearch on Teaching, 529.43CHAPTER TWOTHE NEGLECT OF EPISTEMIC CONSIDERATIONS IN RESEARCH ON TEACHINGTo this point I have argued that the culture of teaching and the socialization of noviceteachers into this culture combine to preserve long-standing, but undesirable methods of teaching.My chief worry concerned teaching practices which, by their neglect of reasoning, run the risk offostering intellectual dependence in schools. Let us suppose that the arguments thus far haveexaggerated the extent to which successfhl educational reform is unlikely. We might think that theculture of teaching and the forces of teacher socialization are not the powerful influences I’vemade them out to be, that effective programs of teacher education nullify their effects. Effectiveteacher preparation might then fortify novice teachers against the influence of teacher socializationin schools, and thereby lessen the chances that undesirable teaching practices will persist. In thisdiscussion I plan to cast doubt on such an optimistic thesis. For while some courses of study inteacher education programs challenge established practices that stunt intellectual virtue, there areinfluential bodies of educational research and theory whose dissemination in initial teacherpreparation programs reinforce these practices, not least because they already resemble teachingpractices familiar to pre-service teachers. Of the many bodies of research and theory relevant toteacher training I will consider only two:1) Process-product research on teaching effectiveness.2) A conception of thinking skills that dichotomizes the relation between thinking and knowing.My main objections can be summed up as follows. In process-product research we findrecommended teaching strategies that are premised on conceptions of learning, achievement, andmotivation that are either unduly restrictive or altogether wrong-headed. In matters related toclassroom management and motivation as well as matters more directly related to teaching,process-product research advocates a pedagogy that fails to respect the intelligence of the learnerand overlooks the place of critical thought in educational achievements.Secondly, popular notions of the relation between thinking and knowing posit a sharpdichotomy between the two, and conceive of thinking in terms of discrete and highly generalizableskills. I will argue that this skill talk risks giving rise to a pedagogy of practice that is44inappropriate to the development of critical intelligence. It also fails to make clear thepedagogically significant relations between knowing and thinking. The picture of knowing andthinking to be found in many teacher education programs is more a caricature of the achievementswe seek in education. To the extent teachers aim at ephemera of this sort their efforts to developintellectual autonomy will be off the mark. Or so I will argue.STRATEGIC AND LOGICAL ACTS OF TEACHINGFollowing Green, we can say that considerable emphasis in the literature of teacherpreparation is given to the ‘strategic acts’ of teaching, while teaching’s ‘logical acts’ are virtuallyignored.’ Strategic acts include planning, motivating, disciplining, questioning and the like.Logical acts, on the other hand, include explaining, concluding, giving reasons, amassingevidence, inferring and so on. Logical acts are essentially epistemic in character insofar as theyaim at fostering rational belief while strategic acts lack this epistemic purpose. The neglect oflogical acts in teacher preparation texts is a serious defect of this research, a defect, that, I believe,can be traced to a crudely causal view of teaching and learning. While it is commonly admitted byteacher educators that teaching requires a knowledge of subject matter in addition to pedagogicalknowledge, teachers colleges tend to restrict themselves to training teachers in the methodsrelated to the strategic acts of teaching. Teacher education texts have the character of how-tomanuals, providing prospective teachers with lists of do’s and don’ts, sets of simple prescriptionsfor planning, motivating, questioning, etc. Effective teaching is viewed in terms of the utilizationof discrete skills; the logical character of acts that aim at fostering rational belief are virtuallyignored. E. D. Hirsch has observed of this tendency thatAmerican schools of education are conceived on the principle that pedagogy itself is a skillthat can be applied to all subject matter. Many of the courses taken by prospectiveteachers emphasize techniques of teaching and ways of improving students’ “inferencingskills” and other general abilities as they are defined by theories of educational psychology.Thus the principle that abstractly defined skills are more important than specificinformation cannot be relinquished without compromising the fundamental assumptions ofeducation schools. If educationists did not assert that skill in pedagogy is more importantthan mere information (which can always be looked up) they would not be able to resistthe common sense view that the best teachers. . . would tend to be those who are wellprepared in the subjects that they teach.245While Hirsch may exaggerate the importance of subject preparation to teaching, hisobservation that education faculties have an institutional bias toward inflating the importance ofgeneric skills is an important one. Teachers are not typically taught which questions on a giventopic are important to ask. Instead they are taught the form of questions that require students toevaluate or analyze. There is little guidance given in how to determine which are the importantclaims, actions or events in need of evaluation or analysis. Thus there is the danger that teacherquestions will be trivial or pointless even while they possess the appropriate form. Skill inclassroom management tends to be viewed as a matter of instituting routines, procedures andsanctions that make the classroom a predictable environment. Absent is the idea that studentmisconduct will frequently be due to a lapse in moral judgment. Textbook discussions concernedwith techniques of classroom management seldom describe student misconduct as selfish,inconsiderate, rude, insolent, or cruel—terms that carry moral significance. Nor do these textscommonly employ terms like ‘foolish,’ ‘wrong-headed,’ ‘self-defeating,’ or ‘reckless’. Insteadstudent misconduct is discussed in terms of “off-task behavior” and non-compliance with schoolor classroom rules. The emphasis on derived over fundamental rules encourages anauthoritarianism that, I will argue, is out of place in institutions charged with fostering intelligentbelief and action.Process-Product ResearchProcess-product research has been called the most vigorous and productive of theresearch programs investigating teaching.3 It, more than any of its rivals, has had a significantimpact on teacher training and practice in the field, though the research community has lost faithin its ability to generate robust findings. Despite the misgivings of many empirical researchers andmore than a decade of criticism from educational philosophers the findings from this researchconstitute a staple of initial teacher education programs. Though the bulk of this research isconducted in the United States, its influence has been internationalized by commercial programssuch as Madeline Hunter’s Instructional Theory into Practice, teacher education textbooks, and avariety ofjournals, of which the Elementary School Journal is but one example.Classroom teachers are more likely to know process-product research by a different46name—teacher effectiveness research,—or be familiar with the instructional programs andconstructs it has produced such as direct instruction, active instruction, time on task, andacademic learning time.4Process-product research is so named because of researchers’ interest indefining the relation between what teachers do when teaching (the process) and what studentsachieve as an apparent result of this teaching (the product). The research to date has generatedseveral recommendations regarding teacher effectiveness, especially in the areas of classroommanagement and motivation, and has led to the development of instructional models based onmeta-analyses of the data.With direct instruction, the best known of these models, teacher behaviors are designed tomaximize the length of time students remain engaged in their work, thus in the literature there is aheavy emphasis on teacher directed lessons characterized by high rates of student ‘success’, andcontinuous monitoring of student progress. The price of comparatively high rates of success andengagement time is a disturbing redefinition of what students ought to study in school. Althoughit may appear that direct instructional methods are the most efficient way to use time andresources, in practice direct instruction means that teacher talk dominates the class, while thecontent is intentionally simple and easy to grasp so teachers can ask many questions and get manyresponses. Questions of this sort tend to be exclusively closed; that is, they are questions forwhich there is only one right answer. Open-ended questions are to be discouraged, as isdiscussion, and debate. Topics and issues of any complexity, controversy, or ambiguity (for thestudent) are pushed to the side.Barak Rosenshine, a direct instruction advocate, has admitted that process-productresearch has little to say on how to teach composition, increase reading comprehension or developanalytic skill.5 He does claim, however, that the research results are relevant in:teaching mathematical procedures and computation, reading decoding, explicit readingprocedures (such as distinguishing fact from opinion), science facts and concepts, socialstudies facts and concepts, map skills, grammatical concepts and rules, and foreignlanguage vocabulary and grammar.6Significantly, what distinguishes these two lists is the extent to which students maysucceed simply by means of memorizing facts, rules, concepts, or procedures. Very little47reasoning is required by these outcomes (although a rather barren notion of learning a concept isimplied). Little room exists for developing knowledge or understanding in any serious sense whenteaching is largely given over to the transmission of basic facts and skills. Students are notencouraged to hold or abandon their beliefs on the basis of reasons, evidence and warrant. Littletime in the “effective” lesson is to be allotted to deliberating about the adequacy of evidentiaryarguments, or producing and defending one’s own. In what sense is the neglect of reasoning aproblem? It might be argued that most of elementary schooling, at least, ought to be devoted tobasic outcomes that will form a foundation for more sophisticated understanding at a later date.There is some sense to this view, especially when applied to the teaching of concepts and rulefollowing procedures, but it would be a mistake to see the teaching of concepts as somehowdivorced from teaching people how to think.7 The remainder of Brophy’s list—teachingmathematics, science and social studies—suggests a crudely mechanistic view of what it is toacquire knowledge in these disciplines. A common criticism of mathematics teaching is thatchildren are instructed in rote procedures that they unthinkingly follow to obtain a correct answer.There are good reasons to reject this view of mathematical understanding. In an essay entitled“Basic Mathematical Skills” Scheffler criticizes what he calls “the false public image ofmathematics”.8This is an image of mathematics as something “exact, mechanical, numerical andprecise—yielding for every question a decisive and unique answer in accordance with an effectiveroutine.” The gist of his argument is that mathematics is quite different from this public image.Mathematics is not a collection of procedures for obtaining answers to problems, but rather a fieldof human understanding. To understand mathematics one must comprehend the truth ofmathematical expressions and propositions. Though there may be step-by-step procedures one canfollow in arriving at correct answers, one cannot be said to understand these procedures unlessone understands the reasoning behind each step in the procedure, and why a correct answer iscorrect.Skill and the Pedagogy of PracticeScheffler’ s observation on mathematics instruction stresses two points: first, that skill andcomprehension are quite different things, and second, that the differences between them mean that48methods that are suitable for developing skills are not suitable for developing understanding orcomprehension.I have said much about skill, virtually nothing about comprehension. What sort of skill isthat? Elsewhere I have argued that it is not a skill at all. To approach education as if itwere always a matter of equipping the pupil with skills distorts our thinking. . .. [Onecannot] speak of practice in the realm of comprehension as one does in reference to skills.One cannot develop an understanding of quantum theory by understanding it over andover again, nor can one deepen one’s understanding by faithfully repeated performances ofunderstanding. One can tell a pupil to practice writing out a proof; it makes no sense totell him to practice understanding it.The general point is this, and it applies to all fields of human understanding, not justmathematics. If teachers aim at fostering understanding they must provide the explanations, theanalysis, etc., that will enable students to understand why we take some set of propositions, evenmathematical propositions, to be true. This means giving students the reasons why we takesomething to be true. Much mathematics teaching in the past appears to have mistakenly takenmathematical understanding to be a matter of skill)° Accordingly, a great deal of mathematicsteaching resembled a direct instruction lesson with its demonstration and guided practice. Thesame complaint holds for science education. Numerous studies have shown that science studentstend to approach scientific problem solving as an exercise in the application of scientific formulae.Even very successful students were shown to put most of their effort into memorizing formulaeand identifying the sort of problem to which a given formula would apply.” By contrastprofessional scientists tend to operate within what Larkin and Chabay have called “the mentalspace of scientific reasoning”. Scientists, they say, are more likely to “talk qualitatively of forcemomentums, velocity changes, and the relations between them, without ever writing anequation.”12Students can obtain correct answers by following the steps they practice, but this sort ofpedagogy provides very little understanding of what these problems actually mean, why a correctanswer is correct, what bearing these mathematical operations or scientific formulae have on realworld problems, or how mathematics and science are, indeed, forms of human understanding.Social studies instruction fares no better in this regard than instruction in mathematics andscience. Unlike the knowledge claims in the latter two domains the knowledge claims advanced in49social studies are often contested. It is the nature of the disciplines that make up the social studiesthat the knowledge claims and value positions with respect to particular issues are the subject ofmuch debate among experts. Where authorities fail to agree appeals to epistemic authority areillegitimate. Thus didactic teaching in social studies lacks the justification it might have if broadconsensus among experts prevailed as it does in many areas of mathematics and science. It is well-established in research on social studies instruction that the majority of social studies textbooksfail to acknowledge the range of disagreement among experts on various topics. Discussions ofculture, for example, frequently adopt a stance of cultural relativism without acknowledging thecontentious nature of this view. Topics in history, economics, and politics frequently receive achauvinistic treatment that would not be accepted by all historians, economists, or politicalscientists. For example, sympathetic interpretations of American foreign policy are frequently theonly interpretations made available to American students. The ideal functioning of free markets isfrequently presented in terms that suggest the ideal is the actual, just as the superiority ofAmerican political institutions is argued for on the basis of comparisons between the theoreticallyideal tinctioning of these institutions and the real problems of political institutions in othercountries. The latter example is perhaps more clearly an example of textbook bias than one ofdenaturing topics in social studies. The prevalence of bias in social studies textbooks is one of themost well documented failures in the entire domain.’3 Textbooks in both the United States andCanada have been criticized for gender’4 and political bias,’5 and for misrepresenting the natureand extent of conflict in history and contemporary society.’6 Gender stereotyping in texts ischarged with socializing girls and boys into accepting the inferiority and passivity of girls andwomen. Texts commonly trivialize or ignore women’s role in societies past and present. Womenwho are acknowledged in history texts tend to be those who have distinguished themselves inmale domains. Historically significant topics in which women have played a major role—midwifery and human reproduction, housework, sexuality, marriage, and divorce—have not untilvery recently received sufficient attention. In Canadian textbooks biographical portraits of menoutnumbered those of women by a ratio of 6 to 1.17 Critics have also observed that mosttextbooks present an unrealistically upbeat account of the nation’s past and present. The shameful50chapters in national history receive considerably less treatment. Conflicts that do receive treatmentare more often international conflicts than domestic ones. Labour history is either ignored, orpresented as the history of conciliation between management and workers. The history ofaboriginal people, when not overlooked altogether, is presented in terms that downplay thecatastrophic moral and physical impact of European contact and domination. The mistreatment ofwomen and children, especially crime against women and children, is virtually ignored in theelementary grades where the family and community are major topics of study. Instead students areleft with the impression that in society “a happy consensus reigns,” where members of families,communities, and states have their needs met by existing social/political institutions andtraditions. 18 Students who continue their education to the university level will encounter in theirsociology, history, and political philosophy textbooks arguments and points of view intentionallywithheld from them a few years earlier. Critics of textbooks in both the United States and Canadaobserve that few teacher education programs alert prospective teachers to these deficiencies, orgive them training in the analysis of textbooks. 19The Neglect of Reasoning in Classroom ManagementProcess-product research has been especially influential in the area of classroommanagement. Jere Brophy, one of the most enthusiastic promoters of this research, gives usexamples of how an effective teacher manages her class.20 Significantly, in these examples theteacher does not give students reasons for study and co-operation. There is no effort at moraleducation, nor any direct reasoning with students about the purpose and value of their studies.Misbehavior is not viewed as the product of faulty reasoning, so much as the effect of a poorlystructured set of activities and procedures. Teachers, therefore, are not encouraged to justifylessons to students. Neither are they to engage students in practical reasoning. Instead they are toreduce opportunities for idleness and mischief by closely watching students, and manipulating thepace and difficulty of the lesson. Indeed, the effective manager is primarily an effectivemanipulator of children’s emotional states, particularly the anxiety of children with respect toearning the teacher’s approval or censure. Effective managers, according to Brophy, usepresentation and questioning techniques that keep students alert and accountable. These51techniques, borrowed from Kounin, include withitness, overlapping, signal continuity and groupalerting.2’ In practical terms these techniques amount to teachers “looking around the groupbefore calling on someone to recite, keeping the students in suspense as to who would be calledon next by selecting randomly, getting around to everyone frequently...” (group alerting).22Effective teachers “monitored the classroom regularly, stationing themselves where they could seeall of the students continuously. . . [to] let students know their teachers were ‘with it.”(withitness).23Effective managers also “move near the inattentive students, use eye contact wherepossible, direct a question to them, or cue their attention with a brief comment...” (signalcontinuity).24 What is striking about these suggestions, apart from their banality, is the way inwhich they sidestep the question of whether we ought to give students reasons for what we askthem to do.Since challenging material or subjects of any complexity may present managementdifficulties, Brophy recommends easy seat work for students on the grounds that “confusionabout what to do or lack of even a single important concept or skill will frustrate students’progress and lead to both management and instructional problems for teachers.”25 When studentsare working with the assistance of teachers the research, says Brophy, suggests that success ratesof 75—80% should be expected. When students are working independently success rates of95—100% are necessary.Evidential and Non-Evidential Styles of BeliefSo long as teachers teach propositional knowledge as if it is only a set of disjointed factsto be remembered there is the danger that students will develop a non-evidential style of belief 26If we are committed to developing intellectual autonomy in students then at the least we arecommitted to the idea that students should hold their beliefs on the basis of good reasons, andadequate evidence. We must therefore give students good reasons for what we want them tobelieve, and encourage a reasonable skepticism until such reasons are provided. With respect tomotivation and classroom management the commitment to critical thinking ought to alert us to thedangers of giving logically irrelevant reasons for study, hard work, and reasonable behaviour.Following Foley, such reasons may be termed non-evidential prudential reasons, that class of52reasons one might adopt on practical grounds in the interests of your non-epistemic well being.The motivation for accepting a belief on such practical grounds is essentially Pascalian. Pascalurges us to believe in God because the consequences of non-belief are severely distressing(providing God exists and possesses the punitive inclinations against non-believers attributed toHim in the Bible). The implicit argument for student’s acting in compliance with teacher requestsor with the school’s standard of behavior is one restricted to the prudential considerations ofavoiding sanction, or gaining rewards of a non-epistemic nature. Student attention is directed notto a set offundamental rules that have their ground in the goals of education, but rather to a setof derived rules whose derivation gives little sense of their relation to what is fhndamental. Rulesof behavior in classrooms and schools tend to be stated as imperatives or prohibitions that do notreveal the moral nature of the fundamental rule from which they are derived. Prohibitions againstrunning in the halls or talking in class have a tendency to become enforced or defied onauthoritarian or anti-authoritarian grounds where the fundamental ground is overlooked. A moreappropriate situation would be one where teachers offer reasons for adopting a particular belief orembarking on a particular course of action. Student resistance to these efforts can be lessambiguously viewed as a refusal to believe the propositions advanced in classes, or more generallya refusal to engage in inquiries whose relevance and value is unclear to the students. Appealing tostudent capacities to reason disambiguates at least to some degree student resistance. As thingsnow stand, student resistance tends to be interpreted by teachers and prominent classroommanagement theorists as immature defiance to institutional authority. Resistance of this sort is tobe circumvented by reducing opportunities for off task behavior, and/or instituting a system ofcontingently applied penalties and incentives unrelated to the epistemic merits of the beliefs heldup for consideration. Few of the prominent management theorists stress the need for reasoning,and this oversight lowers the probability that students will become acquainted with thefundamental moral and epistemic standards that define the form of life associated with theeducated person. Instead student purposes are depicted in terms of the pursuit of high grades,teacher approval and compliance with school rules. None of these courses of action will do muchin themselves to acquaint students with the logically relevant reasons for study or considerate and53reasonable behavior.The difficulty here can perhaps be best illuminated by reference to the idea of a practiceand the standards immanent in a practice. Maclntyre has argued there is an important tensionbetween acting in the pursuit of goods that are extrinsic to a practice and acting in the pursuit ofgoods that are internal to it.By a practice I. . . mean any coherent and complex form of socially established cooperativehuman activity through which goods internal to that form of activity are realized in thecourse of trying to achieve those standards of excellence, and human conceptions of theends and goods involved, are systematically extended.27Maclntyre lists as examples of practices: architecture, agriculture and various traditions ofinquiry such as physics, chemistry, biology and history. Each has its own standards and purposes,its own methods and traditions. Unlike goods internal to a practice, external goods arecontingently attached to practices, as are money, status and prestige to the practice of medicine.Unlike external goods that can be pursued in a great variety of practices, internal goods can beacquired only within the practices. Maclntyre’s discussion is especially instructive for those whoare attempting to initiate children into complex human practices such as those found in varioustraditions of inquiry that comprise the bulk of school subjects.Consider the example of a highly intelligent seven-year-old child whom I wish to teach toplay chess, although the child has no particular desire to learn the game. The child doeshowever have a very strong desire for candy and little chance of obtaining it. I thereforetell the child that if the child will play chess with me once a week I will give the child 50cents worth of candy. Thus motivated the child plays and plays to win. Notice howeverthat, so long as it is the candy alone which provides the child with a good reason forplaying chess, the child has no reason not to cheat and every reason to cheat, provided heor she can do so successfully. But, so we may hope, there will come a time when the childwill find in those goods specific to chess, in the achievement of a certain highly particularkind of analytical skill, strategic imagination and competitive intensity, a new set ofreasons, reasons now not just for winning on a particular occasion, but for trying to excelin whatever way the game of chess demands. Now if the child cheats, he or she will bedefeating not me, but himself or herself28Goods internal to a practice can only be specified in terms of the practice, and can only beidentified and acquired within the experience of participating in the practice itself Maclntyredistinguishes between two types of goods internal to a practice: the excellence of the product(which includes excellence in the production of the product and the product itself), and the good54of a certain form of life. A practice, therefore, involves standards of excellence that apply both tothe product and its production, and define excellence in means and ends. To learn these standardsand to come to accept their normative force is to enter into a form of life. Richard Petersanticipated Maclntyre’s view of practices in his discussion of what it is to be educated.A man cannot really understand what it is to think scientifically unless he not only knowsthat evidence must be found for assumptions, but knows also what counts as evidence andcares that it should be found. In forms of thought where proof is possible cogency,simplicity, and elegance must be felt to matter. And what would historical or philosophicalthought amount to if there was no concern about relevance, consistency, or coherence?All forms of thought and awareness have their own internal standards of appraisal. To beon the inside of them is both to understand and to care.29This picture of a practice, its defining standards and goods, and the sort of schoolingnecessary to initiate students into an intellectually complex practice is, unfortunately, quite remotefrom the experience most teachers have had in their own education, and quite remote also fromthe sort of schooling many teachers are capable of providing to their own students. The state ofbeing on the inside of a practice such as science or history is not likely to be available to studentswho do not reach the level of advanced study in university. Is there any special reason why thisshould be so, other than its being due to the traditions of schooling and undergraduate education?I think not. There is little evidence to suggest children cannot develop at least the disposition tobe critical and the inclination to base their beliefs on evidence. Indeed the insistent curiosity andthe demand for reasons among small children are well known features of early childhood. There isreason to believe that these dispositions are more extinguished than fostered in schools as childrenpass from the elementary grades to secondary school and beyond. Nor do schools appear tointroduce students to the critical standards that partially define the traditions of inquiry and markout the domain of epistemic justification. Denied experiences that would put them in touch withthese standards it is little wonder that students fail to develop a critical spirit, and instead developa non-evidential style of belief Schools and universities must take some responsibility for thisstate of affairs to the degree they neglect evidential reasoning and offer non-epistemic motivationsfor belief and action.There are two important points here for teachers to bear in mind. The first is that it makes55a difference what sort of reason we give students for doing what we ask them to believe or to do.The second is that our motivational practices in school, including grading practices, promote adevotion to external goods that is at cross purposes with our fundamental educational goals ofhaving students act on the basis of good reasons, and of their discovering the value to be found inthe world of knowledge and skill. If the reason for belief or action is intimately and logicallyrelated to the activity we want students to engage in and succeed at then that reason will revealsomething of the point of the practice. Without an understanding of the purpose of a practice thereason for thinking certain methods are superior to others will remain obscure. There arestandards that define excellence in the activity as well as standards that define excellence in theproducts of activity. Without an emphasis on justification for action and belief these standardsmust remain hidden from view. Ifjustifications are only contingently related to the activity, as thecandy was in Maclntyre’s example, then such justifications indicate little of what is intrinsicallyvaluable in the practice, and thereby tend to stand as barriers to the pursuit of those goods that areinternal to the practice. Among these are the goods of a particular way of life, in this case the wayof life of an educated person. Students who have no iniding of what these goods may be aresimply not educated. It is difficult to see how school practices as described here will give studentsmuch insight into this way of life or the standards that define it.Reasons and MotivationOne of the most widely used introductions to research on teaching, Arends’ Learning toTeach, devotes a chapter to summarizing research on the variety of ways in which teachers canstructure the classroom environment and manipulate students’ psychological states to ensure cooperation and effort. In this chapter Arends devotes several pages to Madeline Hunter’s methodsfor motivating students. These include the following:1) Level of concern: Level of concern is the level of stress or anxiety students feel during yourlesson. Hunter insists that if students aren’t moderately stressed they won’t be properlymotivated. Hence she suggests teachers do such things as “stand next to a student who is notparticipating to raise concern”; Announce that “This will be probably be on the test”.2) Feeling tone: According to Arends, “students put forth more or less effort according to the56unpleasantness or pleasantness of the learning environment.” An unpleasantfeeling tone is theresult of a teacher intentionally saying negative things (“That story must be finished before youare excused for lunch”) A pleasant feeling tone is the result of the teacher saying somethingencouraging or pleasant (“You write such interesting stories, I’m anxious to read this one.”)3) Success: By this Hunter means to remind us that success at a task can be encouraging, whilefailure may result in a discouraged and unmotivated learner. To this basic insight she adds thatif tasks are too easy we don’t feel successfiil in mastering them, while if tasks are too hard wecan’t be successfiil. Thus, we are motivated by tasks that are moderately challenging. Teachers,on this view, should set challenging tasks that still allow students to be successftul.4) Interest: Teachers ought to make their lessons interesting, novel, and vivid.5) Knowledge of results: Teachers need to give specific and immediate feedback on studentwork.6) Influence and Affiliation Motives: Teachers need to give students some say or influence inthe way things are done in the classroom, as well as giving students opportunity to work insocial groups.On the face of it most of these “strategies” for motivating students seem perfectlyinnocent, sensible even. The problem is that not one of them involves giving students the logicallyrelevant reason(s) for engaging in their schoolwork. Presumably there is a good reason forstudents to do the things we ask them to do. We need to give students these reasons, and expectstudents to demand them. Hunter’s virtual silence on the importance of giving these reasons is amistake in the view of those committed to fostering a reason-for-acting mentality in students.Manipulating the feeling tone and level of concern fails to communicate any sense of theeducational benefits to be derived from academic study. What is being appealed to here, instead ofreason of this sort, are emotions or psychological states like fear, pride, and anxiety. No doubtthere are times when teachers will need to appeal to such things, particularly when appeals toreason have failed. But the impression left by Hunter is that appealing to these states is whatteachers ought to do in the first instance, not as a last resort. Interest, knowledge of results andthe rest are fine, in themselves, but they are not a replacement for giving students the genuine57reasons we have for thinking that what we are asking them to do is worth doing. The argumentagainst Hunter’s approach can be summed up thus. If we, as a matter of course, appeal to thereason of students (by giving them the logically relevant reasons for doing what we ask them todo) students have a much greater chance of getting the message that to be an intelligent person isto do things for a good reason.It might be offered in defense of teacher manipulation that children lack the maturity togovern themselves, or lack the ability to understand any justifications that might be advanced byteachers. After all the effectiveness of rational persuasion with respect to children depends on therational capacity of children. But anyone who wishes to depart from rational persuasion inargumentation must carry the burden of proof for demonstrating rather than merely assuming theinability of children to benefit from rational persuasion. Following Kohlberg, Habermas notesthree phases of cognitive development; only the first of these is compatible with the neglect ofrational persuasion.1) For the pre-school child, who is cognitively still at the stage of pre-operational thought, thesector of his symbolic universe relevant to [social] action consists only of individual, concrete,behavioral expectations and actions, as well as consequences of action that can be understoodas gratifications or sanctions.2) As soon as the child has learned to play social roles his symbolic universe [can now include]actions as the fulfillment of temporally generalized behavioral expectations [norms].3) When finally the youth has learned to question the validity of social rules and norms ofaction.. .there.. . appear principles in accordance with which opposing norms can be judged.3°At the first level the extent to which children can be held morally or epistemicaflyresponsible is quite limited. At the second and third levels, however, children and adolescents arecapable of reasoning along conventionally normative lines at the least. Adolescents at the thirdlevel are capable of a significant degree of epistemic and moral independence. They are able, saysHabermas, “to assert their identities independent of concrete roles and particular systems ofnorms.”31 That is they become capable of the criticism of norms on the basis of principles. MaxMiller, an associate of Habermas, has found that children as young as three years old take note of58circumstances where justification is called for, and enter into basic forms of argumentation toresolve the issue. Miller lays out the formal structure of an argument between three year olds thatshows that there is no formal logical difference between the justificatory argument of a rationaladult and that of a three year old.32 The difference lies in the norms upon which children rely intheir efforts to come to an agreement. Of special interest is Miller’s finding that theappropriateness of the norms employed by young children is limited by the context or problemsituation in which they find themselves.33 Miller concludes that, within the appropriate problemlevel, children can, in principle, engage in justificatory argument much more the equals of adultsthan hitherto allowed. What is crucial, however, to the success of classroom argumentation is thatstudents be allowed to set the problem level by virtue of the questions they raise and the problemsthey encounter. This recommendation must sound very much like the constructivism of currenteducational theory, but it needn’t possess the same solipsistic implications. It is not beingsuggested, as is sometimes argued, that because children ‘make meaning’ out of their existingsystems of belief teachers should stifle their inclination to follow an established course of study,and instead hand over the task of curriculum design to the children in their classes. The claim ismuch more modest. It is that the occasions when children identifS’ what is problematic for themmay be used by teachers to identifj the level of debate most appropriate to the goal of fosteringintellectual autonomy. As things currently stand the evidence suggests teachers conduct theirclasses at a level more appropriate to pre-school students operating in the first of theHabermasfKohlberg stages. To the degree Miller is correct, there appears to be no good reason toteach in the ways suggested by behaviourist educational researchers like Madeline Hunter,providing our goal is the development of rational capacities and passions.Once one has been alerted to the absence of a reason for acting perspective in theresearch on classroom management, it is easy to see that this lacuna is a defining feature of theorthodoxies with respect to motivation and preventative discipline. In textbook after textbook onefinds little or no mention of and certainly no emphasis of the importance of reasoning withchildren about what to do.34 Instead one finds quasi-causal perspectives supplied by behavioral35and clinical psychology36,classroom ecology37,and teacher effectiveness research38.The first two59perspectives emphasize the psychological causes of behavior, while teacher effectiveness researchis more concerned with teaching behaviors that are believed to cause, or otherwise bring about“on-task” student behavior and superior test scores. The list of what causes misbehavior is long,but includes pupil insecurity, a neurotic need for teacher attention, or power, as well as the morefunctional need for affiliation, and self-esteem. The effects of inept parenting, bad peerrelationships or disadvantaged social backgrounds are often cited as well, but these typically areviewed as being beyond the power of teachers to remedy. Their significance to the presentargument is that they exemplify the emphasis on non-epistemic causes of behavior, and are part ofa view of motivation premised on an incomplete account of human action. While there is nodiscounting the causal influence of factors such as the need for affiliation and self esteem, in thecontext of education student motivation is best understood in terms of student thinking. So thecommon reminder to focus on student behavior instead of their reasons is little help in alertingteachers to the view that as human agents students do what they do for a reason. If the reason is apoor one, and reflects more than a momentary lapse of judgement, the appropriate teacherintervention is, ideally, one which attempts to show to the student the inadequacy of her reasons.Instead what is most often prescribed is an intervention that supplies or withholds whateverextrinsic goods are valued by students. This strongly behaviourist orientation fails to encourageteachers to see students in these teleological terms, and thus contributes to the neglect of reasonsas the basis for thought and action.Classroom management texts repeatedly stress the need to minimize opportunities forstudent decision-making about what to do, and treat evidence of emotional dependency not asmaking a case for the need to reduce this dependency so much as underlining the need to modifyone’s teaching in the face of it. Emmer, Evertson and Sanford echo Brophy’s claim that a keydifference between effective and ineffective classroom managers is that effective managerseliminate any ambiguity about what students are to do. Effective managers, they say, establishroutinized procedures for student participation and movement, and where instructions or workrequirements must be communicated they are unequivocally clear.39 Classroom activities that donot lend themselves to straightforward routinization are to be considered less attractive than those60whose simplicity reduce the need for student judgment and interpretation. The advantage ofroutinizing procedures, claims Brophy, is that such classrooms “seem to work automatically”.40Doyle and Carter go so far as to endorse the maintenance of heteronomy in the interest ofcontrol. In one of their studies, they observed students who demanded to be told what to say andwhat to do in an creative writing assignment. It was obvious that the teacher was attempting tofoster creativity and self-direction by providing a range of writing options. It was also clear thatthe class feigned confusion to force the teacher to become more explicit in her directions, and todo more of their thinking for them. Some were merely wasting time to avoid work, while otherswere made anxious by the prospect of thinking for themselves. Recall Kant’s claim that “lazinessand cowardice are the reasons why so great a portion of mankind, after nature has long sincedischarged them from external direction, nevertheless remains under lifelong tutelage, and why itis so easy for others to set themselves up as their guardians.” Doyle and Carter’s advice is thatorder must take priority over fostering autonomy. Rather than conclude that heteronomy ought tobe reduced or eliminated, they take emotional dependency as a given to which teachers mustaccommodate themselves in the interest of maintaining control.Some students became quite adamant in their demands.. .On such occasions, order beganto break down and the normal smoothness and momentum of the classes were reinstatedonly when the teacher provided the prompts and resources the students were requesting.The teacher was pushed, in other words, to choose between conditions for students’ self-direction and preserving order in the classroom.41Doyle and Carter note with some satisfaction that the teacher in this case was experiencedenough to know that “order had to come first or everything else was lost.” This sentiment iswidespread in the research literature as well as in schools. When control breaks down, the cause isseen to lie in the teacher’s failure to anticipate dependency and immaturity. It is as if students aremerely reactive agents in an environment of teacher regulated stimuli.The manner in which students are characterized in this research is indeed puzzling, as isthe fact that few people have objected to this characterization. One of the more revealingobjections, by John Meyer, points to the way in which the prescriptions from this research maydull the moral sensitivity of teachers. “Too much research looks at the technology of teaching as ifit were the mechanical action of a person on an object.” In Meyer’s view the student is seen as “a61mildly intelligent monkey (or occupant of a monkey-like role), constrained by the immediatedistribution of rewards.”42 Students are typically depicted as being the passive recipients of one oranother of various educational “treatments” or “inputs”, or as being subject to the influence of ahost of stimuli that will, under optimal conditions, produce the desired educational “outcomes”.Once students are conceived of as the passive recipients of “treatments” they become littlemore than receptacles of inputs, or objects to be manipulated. This carries a number ofconsequences. The student is to a large extent freed of any responsibility for learning; theresponsibility therefore shifts to the teacher. In addition, manipulating students begins to appearreasonable, even desirable. Aside from the moral dangers associated with this sort of manipulationis the threat to intellectual autonomy: students may not be given a chance to develop the criticalabilities and dispositions that are constitutive of being an educated person. Since “effective”teaching methods of this kind are authoritarian in the strictest sense of the word, “effective”teaching methods carry the risk of impressing on youngsters the values of obedience and industryat the expense of independent critical thought. Students are to accept the truth of statements,insofar as the issue of truth and justification ever arises, on the basis of authority, either theauthority of the teacher or the authority of the text. They are to complete their exercises and dotheir homework, it would appear, because the teacher expects them to and will catch them out ifthey don’t. This is just the sort of teaching practice that is likely to engender the habit of holdingbeliefs dogmatically, on the basis of someone else’s word, or of acting unthinkingly to avoidcensure. Without opportunities to consider the reasons why a belief is widely held to be true, itwill seem to many students that it is not their business to know why a claim is justified so long assomeone does.43So the main argument against the most widely cited prescriptions in educational researchis that students are not expected to think critically about what their teachers tell them to believe ordo. Indeed, teachers are encouraged to reduce the occasions where student thoughtfulness is evennecessary. Therein lies the maintenance of immaturity in students.THINKING SKILLSTalk of skills is ubiquitous in current educational discourse. By one estimate the number of62books and articles that offer lists of educationally important skills runs in the hundreds! Thereare thinking skills, psycho-motor skills, listening skills, reading skills, inter-personal skills,problem solving skills, information processing skills, communication skills, generating andintegrating skills, remembering skills. The list appears to go on forever. This fecund proliferationof skills should not surprise us. According to Daniels the list of skills could indeed go onforever.45 Because of the polymorphous nature of cognitive concepts any and every intentionalaction that eventuates in an achievement, no matter how modest, may count as a skill. Just asthere are an indefinite number of behaviors that would count as obeying, given certainassumptions regarding the agent’s intentions, the list of cognitive skills and processes is virtuallyboundless as well. There is little point, however, in elaborating longer and more finely detailedlists of skills. Such lists point us in the wrong direction, away from the normative requirementsthat define the successfhl act, and toward an empty pedagogy of practice. What is also strikingabout this list is the manner in which distinctions between widely disparate capacities anddispositions are effaced by bringing them together under the category of skill. It should be plainthat the motor skill of dribbling a basketball or executing a forward somersault is categorically adifferent thing from the “skill” of analyzing an argument or treating another person with courtesyand respect. The term “skill” when applied to these various contexts blurs important differences,some of which are crucial to intelligent curriculum planning or teacher education. Danielsremarks:Classifying knowledge, skills, attitudes, interests, strategies, and processes together asspecies of the same sort of thing masks the differences between them. Attitudes, forinstance, are propensities or inclinations to act in certain ways. Skills are capacities thatmay be exercised, but that need not be. Processes are neither propensities nor capacities;they are, in one sense of the word, events. We have attitudes, we have and use skills; butprocesses happen or we learn to make them happen. It may be that as teachers we want toinculcate attitudes, develop skills, and teach students how to carry out certain processes.But it will not do to pretend that all the upshots we seek are simple subdivisions of asingle psychology of instruction—just because we as teachers happen to seek them all.47Not only are there distinctions to be made between these categories, the skills listedabove, if they are that, differ with respect to determinacy and complexity. The skill of snappingone’s fingers is a relatively simple and determinate skill, whereas riding a bicycle or operating a63lathe involve sets of skills that are not easily or sensibly described in isolation from one another.More complex still are the skills of an historian or surgeon which in addition to being morecomplicated require considerable judgement in their execution. This complexity, the need forjudgement, the relative value of surgery over cycling signal important differences. Suchdifferences are significant because they give us a sense, when properly understood, of how weought to go about developing the competencies in which we are most interested. Whenmisunderstood these differences may lead us to emphasize wrong-headed and ineffectivepedagogies.The failure to distinguish important differences among competencies might lead to grief inseveral ways. A major difficulty has to do with the physical and manual connotations thatsurround the word skill. Skill talk seems most at home in the domain of psycho-motor skills.Dribbling basketballs, doing somersaults, planing a block of wood, making an incision, sewing adress—these are all central cases of skill. Dictionary definitions, though not the final word,support this view of skills. Webster’s Dictionary, for example, defines skills as “(1) a great abilityor proficiency, expertness that comes from training or practice, or (2a) an art, craft or science,especially one involving the use of hand or body, and (2b) ability in such an art, craft, orscience.”48Talk of skills, then, is suggestive of physical proficiency in an art, craft or science that isbest developed by training or practice. Forms of competence that are not of this sort will not bebest developed by teaching methods that are appropriate to the more physical and manual skills.Scheffier makes a similar point in his discussion of the difference between propositional andprocedural knowledge. Procedural knowledge—know-how and skills—are categorically differentfrom propositional knowledge that is more closely connected with questions of truth andjustification than with matters of efficacious procedures. Scheffier observes there arepedagogically important differences between skills and understanding:The notion of practice seems clearly relevant to skills and know-how; they are, indeed,typically built up through repeated trials or performances. . . One cannot develop anunderstanding of the quantum theory by understanding it over and over again, nor can onestrengthen or deepen one’s understanding by repeated performances ofunderstanding.4964Dribbling a basketball well, then, is largely a matter of guided practice and drill. Reading,when construed as understanding what one reads, is less a matter of practice, than ofunderstanding. In order to understand what one reads one must have the conceptual resources tomake sense of the words on a page and the pattern they take in a narrative, an argument, adescription or what have you. This may involve many different kinds of things: a knowledge ofliterary conventions and rules, or a knowledge of human behaviour; what is needed is at least apartial understanding of the theme or subject of the text. Successful reading most of all requiresone’s having the appropriate conceptual resources. The requisite resources will vary much morefrom context to context than the talk of reading skills allows. The tale of the Emperor’s NewClothes is unintelligible to those who don’t have the concepts of vanity, greed, kings and theirrelation to courtiers, at least in embryonic form. Reading and understanding in another context,say that of a narrative ofNewton’s influence on science depends on the reader possessing at leasta rudimentary understanding of gravity and velocity. It is undeniable that one cannot learn to readwithout practice, but practice alone will not yield up an understanding of texts. Skills talk runs therisk of promoting a pedagogy of practice where some other approach is called for. It alsoobscures the importance of context in the exercise of intellectual abilities. The skill of dribbling aball is something that can be exercised in a number of situations and for a number of purposes. Itcan easily be applied across contexts in a way that intellectual abilities cannot. Barrow observesthat “the skills of an historian—the ability to weigh evidence, for example, are not things that canbe transferred. This has nothing to do with empirical arguments about transference; it is a matterof logic that weighing evidence in moral philosophy requires understanding that is not given bylearning to weigh evidence in history.”5°The skills of an historian are abilities that depend onunderstanding history more than on some form of mechanical practice. These conceptual pointscan be buttressed by reference to empirical findings. Though Daniels’ objections to skills talkdates back twenty years, critics today make the same points. One of the best known efforts toenumerate the skills of information processing, those of Sternberg,5’have been criticized for theendless proliferation of skills, which despite their number, have little transference.52Nowhere is the significance of context more apparent than in the domain of critical65thinking. In this general category one often finds such skills as decision-making skills, problemsolving skills, inferencing skills, or so called higher order thinking skills. What these have incommon is the supposition that there are highly transferable thinking skills that have theirapplication in a wide variety of circumstances. Proponents of these thinking skills appear tobelieve it is possible to provide practice in the exercise of these skills in a small number ofcontexts then have pupils successfully apply these skills in a much wider range of contexts. Theinformation or subject matter on which these skills are brought to bear is of secondaryimportance. The point I wish to argue is that the ability to think critically cannot be developed inthis way.I say this for two reasons. The first is the conceptual point often raised by McPeck53,andargued for even more persuasively by Ryle54: thinking, whether critical or not, is always thinkingabout something. What should be obvious from even a cursory examination of arguments in eitherpractical or theoretical reasoning is the fundamental role played by factual premises. Those whoare ignorant of the relevant facts will not be able to construct or detect sound arguments, evenwith an excellent grasp of deductive validity. This fact in itself does not establish a need forknowledge in the strong sense conceived by Scheffler, for true beliefs will serve just as well asjustified true beliefs in any case. It does however establish that thinking is not a contentindependent activity. In support of this view there are numerous empirical studies which suggestthat knowledge that is simply given to students to remember will not prove as useful in thinking asknowledge that is examined with a critical eye.”Secondly, while it may be possible to have the ability to think without exercising thisability in a given case, an individual could not sensibly be called a critical thinker without revealinga disposition to think critically. A person’s being disposed to do something depends on thecommitments and values that person holds. So whether a person is disposed to be logical, critical,reasonable, etc., will depend on what he has come to take seriously. What we take seriously, ormore simply, what we value, is also something that we can reason about. If we want children tolearn to think critically we must introduce them to various domains and subject matters, for eachwill contain, in the practices of criticism which lie at their heart, the value standards that motivate66and guide inquiry. If we want children to possess so called social skills we would do well tointroduce them to the arguments in the more explicitly normative domain of moral and valuesreasoning.Dispositions as VirtuesSkills talk, which renders every sort of desirable trait or ability a skill, tends also toobscure the role of intellectual virtues in the exercise of intellectual autonomy. Wallace offers auseful way to distinguish skills from virtues: each overcomes different impediments.56Lacking askill is a technical difficulty, while lacking the requisite virtue leaves one unable to overcome acontrary inclination. Baron’s review of the empirical evidence concerning rational thinkingidentifies two deviations from rationality, two impediments, if you like: individuals typically“overweigh the immediate costs of thinking relative to the long term benefits” and “gather andinterpret evidence in a way that does not challenge possibilities about to be adopted.”57These twotendencies are not attributable to a lack of technical skill in thinking so much as they areinclinations that reduce the likelihood of an individual thinking as carefully as he or she might. Asantidotes to contrary inclinations virtues are not restricted to specific activities in the way skillsare. Von Wright notes that playing a piano or driving a car are matters of skill limited to theactivity in question, while the courage of a pianist filling in for a famous virtuoso has much incommon with the courage of the psychiatrist interviewing a psychopath. The activities themselveshave little in common, yet the virtue of courage has application across contexts.Knowledge and ThinkingOne of the ironies of the current spate of reform efforts in education is that while mostcurriculum documents stress the importance of promoting intellectual development few evince anunderstanding of the role of knowledge in thinking. The view that knowledge and thinking areseparate is well established in education. In Benjamin Bloom’s 1954 Taxonomy of EducationalObjectives in the Cognitive Domain knowledge and what are now called higher order thinkingprocesses are distinguished, with knowing being tagged as the process that has been given toomuch emphasis in schools. The higher order processes—analysis, synthesis, and evaluation were67seen as being neglected since the bulk of teaching in schools appeared to be aimed at thetransmission of facts and principles to be remembered. Indeed Bloom clearly equated knowledgewith information to be remembered. This identification of knowledge with information is afundamental mistake that has caused much mischief in education for it obscures the dimension ofepistemic justification in genuine knowledge claims. Bloom was right to object to the emphasis onrote learning in schools, but mistaken in thinking that the practice of memorizing facts isconsistent with a full bodied conception of knowledge. A full bodied conception of knowledgewould require that what we call knowledge would include the element ofjustification or warrant.One cannot be said to know much of value if one is ignorant of the reasons (the justification) wehave for thinking a particular conclusion is true. It is in the “logical space of reasons”58 where webest develop our ability to reason, and the logical space of reasons is situated in the justificationswe have for our knowledge claims. Rather than de-emphasize the teaching of knowledge, we needto do a much better job of teaching knowledge, one that stresses the strengths and faults of thejustifications we take to be adequate in the knowledge claims we generally accept.Given this construal of knowledge it should not surprise us that there is significantresistance to the view that one of a teacher’s primary responsibilities is the transmission ofknowledge. Instead something referred to as “process” is to be favoured over “content”.Although what is meant by process is not always clear, the term content appears to besynonymous with such things as concepts, facts, principles and the like. In some cases thedistinction employed to derogate knowledge is one between higher and lower order mentalprocesses: knowing is a lower order mental process while analyzing and evaluating are higherorder mental processes. In other schemes knowledge is seen as mere information, and thinking amatter of information processing. These various synonyms for knowledge are even less suggestiveof epistemic considerations. The terms content and information give no hint of the need forassessing truth claims.For example, Involvement in Learning: Realizing the Potential of Higher Education, aninfluential report of the National Institute of Education in the United States, claims that“capacities and skills are the truly enduring effects of higher education.”59Less attention is given68to knowledge which is variously referred to as the “content or raw material of a discipline.” Thereport leaves educational institutions to decide on the “raw material” that will serve as the meansto the development of skills. Among the skills are those of writing and speaking, critical thinkingand analysis, synthesizing, imagining, and creating. One commentator has observed of thisemphasis thatsomething is missing from this formula, and the reason is the separation of knowledgefrom skills and so forth, as if one of these things were the broth and the other the pot.What happens if the institutional decisions about the objects of knowledge areinadequate?. . . Will all the skills in the world then be of any value?. . Can a report that doesnot speak of the what of knowledge be anything more than a series of encouraging (ordiscouraging shouts)?6°What is striking about the characterization of skills in curriculum and policy documents isnot only the importance attached to these skills relative to knowledge and attitudes, but thenumber of cognitive abilities now classified as skills. This emphasis on skills, present in schoolsfor many years, has been given official sanction in British Columbia as a result of a majorcurriculum revision program known as the Year 2000,61 and in Saskatchewan’s on-goingcurriculum revision efforts. The proposed changes to the curriculum found in the Year 2000 arepredicated on an assessment of social and economic trends in British Columbia that have a bearingon education. One of these trends, the knowledge explosion, is reported to be global in its scopeand is used to make the case against education’s traditional emphasis on the transmission ofknowledge. The significance of the knowledge explosion is that it is unrealistic and miseducativeto persist in teaching knowledge since not only is there too much knowledge at present forstudents to acquire, there will be a great deal more in the fUture. It is claimed that the rate atwhich knowledge is produced and then made obsolete is so rapid that the traditional reliance onknowledge is no longer suitable in the age of information. It would be much more sensible, on thisview, to teach children how to access and process information stored in libraries and data banksthan have them commit facts to memory. The Carnegie Task Force on teacher education adoptsthis line in its report, A Nation Prepared: Teachers for the 21st Century. It is not the job ofschools to provide students with the knowledge they will need to know in their adult roles, butrather with a knowledge of how to find out what they need to know: “They [students] will not69come to the workplace knowing what they need to know, but knowing how to figure out whatthey need to know, where to get it, and how to make meaning of jt.”62 The primary sense of“knowing” in this passage is that of knowing how or skills; it is also a commodifled view ofknowledge whose utility lies in its applicability to the world of work, and in its interpretiveusefulness. On this view there appears to be a skill by means of which one “makes meaning” outof the knowledge demanded by one’s work. These are odd distinctions to say the least. There isfirst the split between knowing how and knowing that where the latter does little to supply theformer. Then there is the split between knowledge and meaning where the latter is imposed on theformer.Of the sort of knowledge needed the report has this to say:The skills needed now are not routine. Our economy will be increasingly dependent onpeople who have a good intuitive grasp of the ways in which all kinds of physical andsocial systems work. They must possess a feeling for mathematical concepts, and the waysin which they can be applied to difficult problems, an ability to see patterns of meaningwhere others see only confusion: a cultivated creativity that leads them to new problems,new products, and new services before their competitors get to them; and in many cases,the ability to work with other people in complex organizational environments where workgroups must decide for themselves how to get the job done.63While the wholesale abandonment of knowledge is not being proposed, its pride of placein education has been put into question. It is to be displaced, at least somewhat, by a greateremphasis on skills and attitudes. Prominent among these skills are information processing skills,thinking skills, decision-making skills, problem-solving skills and communication skills. This list ofskills gives some hint of a postulated dichotomy between knowledge and thinking. One hasknowledge, perhaps in a data bank, or in one’s head, and then, depending on the situation, oneprocesses this knowledge in a variety of ways. If confronted with a problem, a problem-solvingprocess is called for. If one needs to make a decision, a decision-making process is called for.Once the requisite process is identified it can then be employed in relation to one’s “knowledgebase”. These various skills are seen to be distinct enough from knowledge and from each otherthat they can be taught and exercised separately. That is, one can be taught how to solve problemsin a general sort of way and then go on to solve specific problems by accessing and processing therelevant information. Furthermore, since there is more information now, and more sophisticated70technologies for processing information than ever before educators ought to give more weight incurriculum planning to these processes and technologies and less weight to the fact stuffing andcramming that has characterized a great deal of schooling to date. In Saskatchewan thebifurcation of knowledge and thinking is written explicitly into curriculum guides. On the adviceof Barry Beyer, the Saskatchewan curriculum guides divide thinking skills into a hierarchy oftwelve skills that are to be taught in sequence, one or two skills per grade level. Each successiveskill is said to depend on the preceding skill. The skill of “data location” begins the sequence ingrade four, with data gathering in grade five, and data organization in grade six. The remainingsequence goes as follows:Grade 12: EvaluatingGrade 11: Synthesizing and AnalyzingGrade 10: Analyzing and HypothesizingGrade 9: Inferencing, Generalizing and ClassifyingGrade 8: ComparingGrade 7: SummarizingSince it seems the absurdity of this arrangement is not transparent to everyone somediscussion of its weaknesses is called for. First of all it is assumed that evaluating, synthesizingand analyzing are higher order thinking skills in this hierarchy. They are the most complex, andare less likely to be mastered if lower order thinking skills have not been mastered first. Indeed itis insisted that “students are not able to learn to the mastery level more than 5 skills per year.”64 Itis also assumed that school age children will lack these skills or will be somehow deficient in theiruse. These skills are to be mastered by means of “frequent, but intermittent (not massed) practiceof the skill.”65 The guide further cites Beyer as claiming that research shows that thinking skillsmust be learned in the context of course content since “skills learned in isolation of content willnot automatically transfer to any content that may be selected later.”66 The implication appears tobe that transfer to other contexts, perhaps even automatic transfer, will be possible if teachersdevelop these skills in relation to some course content.What should be obvious is that skill in the employment of these “skills” is highly context71dependent. In many contexts even pre-schoolers will be seen to possess these skills. No argumentis necessary to justify the claim that pre-schoolers are capable of making sound comparisons(introduced in grade 8), classifying objects into categories (grade 9), and generalizing fromexperience (grade 9). All three “skills” are required to draw the conclusion and make the claimthat “cotton candy is sweeter than apples”. And yet teachers are warned against attempting toteach more than two or three of these skills per year since students are not able to learn more thanfive skills per year. Take evaluation, allegedly the most complex, therefore reserved for thehighest level of secondary education. To the extent that the skill of evaluation is a necessarycondition of developing stable preferences and acting on them this skill is present in infancy.Presumably what skills advocates are after is the ability to make intelligent evaluations across arange of contexts. But if that is what is desired neither this taxonomy nor the prescriptions forteaching that frequently accompany it will suffice. What is not made at all clear in the curriculumdocument is the criteria-driven nature of evaluation. An evaluation of the new cars for 1995 willdraw on evaluative criteria relevant to cars, while the criteria for evaluating Napoleon’s tenure asEmperor of France will require altogether different criteria. Mileage per gallon, frequency ofrepair rates, and the results of low speed collision tests will be of no use in assessing Napoleon’sgreatness. To say evaluative criteria are context sensitive is just to say that attempts to developthese skills in a general way is an idle fantasy. An ability to analyze chemical compounds will be ofno use in tasks requiring the analysis of philosophical arguments or stock market reports. To beable to analyze chemical compounds what is needed is some defining purpose for the analysis anda knowledge of chemical compounds as extensive as is required by the analysis. To analyzephilosophical arguments (which is to evaluate them as well) is to employ critical standards such aslogical coherence and an extensive knowledge of other philosophical arguments that bear on thetopic under discussion. The upshot is that there is little reason to believe this taxonomy of genericskills is a useful way to characterize thinking tasks because it drains thinking of its substantive andfrequently content-dependent character.In order to maintain the view that knowing and thinking are separate one must ignore adistinction brought forward by Plato, between what it is to know something to be true and what it72is to have a true belief67 Having knowledge, on this view, implies that one is able to give asatisfactory account of how one knows.68 That is, one must be able to provide evidence orgrounds for taking a belief to be a true belief Within epistemology this view falls withininternalist theories of epistemic justUlcation. Internalists can be distinguished from externalistsby their emphasis on the internal character of epistemic justification. Justification, for theinternalist, is a matter of the subject being able to offer an argument that justifies a belief inquestion, while for the externalist the subject may be quite unaware of the reason why a belief isjustified and thus would be unable to offer any justification. The neglect of epistemic justificationin schools is more problematic to internalists than externalists, and for this reason the debatebetween these two will be considered in some detail in the chapter to follow.An implication that the internalist view carries for teachers is that children must be able toprovide an evidential argument of at least a rudimentary sort to be justified in claiming to knowanything. That is, children must be able to demonstrate, to some degree at least, that their groundsfor holding a belief are in accord with public standards of relevance and adequacy, the force ofwhich they understand. If teachers were to emphasize the need for students having evidentiallyheld beliefs instead of beliefs held on the basis of convention or authority, as well as give theirstudents access to this evidence, students would be better positioned to learn to think forthemselves.It is also common in curriculum documents to find references to critical thinking thatsuggest a more or less complete failure to understand its normative nature. Teachers areencouraged to view thinking as a matter of following invariant procedures and steps. Forexample, many curriculum guides advocate the teaching of a five step problem solving methodmodeled on Dewey’s scientific problem solving method. Teachers are instructed to providestudents with opportunities to practice solving problems by working through the successive stagesof problem identification, hypothesis formation, testing, etc., on the assumption that anunfamiliarity with these steps explains why children fail to solve problems intelligently. (Indeed thechoice of problems to be solved often seems not to matter in teaching methods texts and courses.Any topic of interest to students will do). Yet what reason is there to believe children are73unfamiliar with these steps, or that the problem domain is unimportant? Imagine a five year oldlooking for his teddy bear. He walks into his room and realizes he doesn’t know where his teddybear is. In other words he identifies the problem. He then generates several hypotheses regardingthe whereabouts of his bear—under the bed, in the closet, in his toy box, etc. He then proceeds totest each hypothesis by looking under the bed, in the closet, in the toy box. Upon discovering histeddy bear in the toy box he not only confirms one of his hypotheses but is able to infer that hismother put the toy there because he knows he wouldn’t have, and because she always insists thetoy box is the place for toys. Scenarios like this one no doubt occur again and again well beforemost children enter the first grade and receive what their teacher may think is their first exposureto the problem solving method. It is not unfamiliarity with solving problems that accounts for agiven failure in problem solving, so mere practice in problem solving will do little to preparestudents for solving difficult problems. As Dewey recognized, what makes problem solvingdifficult is an inability to tell the exact nature of a problem situation, or the range of potentialhypotheses that may be worth pursuing, or the standards by which hypotheses may be tested.Absent in many educational discussions of problem solving is any clear grasp of these difficultiesor how to go about resolving them for children. Problem situations of educational importance,along with competing hypotheses and the standards by which they can be tested tend to revealthemselves in traditions of critical thought. To ignore these traditions or these elements withinthem is to squander an opportunity to familiarize students with some of the most successfularguments and standards to date.The Discipline of the NormIt is worth noting that both the aforementioned research on teaching, and the generic skillsconception of thinking can be viewed as rudimentary attempts to standardize the way teachersteach. They offer pedagogical norms of effectiveness as definitive of good practice and therebysupply the means by which poor teaching can be identified, and corrected. Norms such as theseare institutionally powerful. Once they have been accepted, they have a way of discipliningindividuals into conformity with them. It is Foucault’s special contribution to have elaborated thepolitical power of the norm which, in the context of this discussion, merits further examination.74Let us begin with Foucault’s discussion of the emergence and operation of disciplinary power inthe first modern prisons.69 Foucault saw in the first prisons the emergence of techniques of controlbased on supervision and the establishment of norms that this supervision made possible.Discipline, in Foucault’ s sense, began to operate in prisons and in other social institutionsresponsible for human improvement (schools, asylums, clinics, barracks, etc.) during thenineteenth century. By means of three instruments: hierarchical observation, normalizingjudgement, and the examination—these are Foucault’s terms—it became possible to knowindividuals and thereby transform them.Hierarchical observation refers to the manner in which supervisory arrangements areorganized within a bureaucratic framework. Each supervisor must in turn be supervised in such away that the entire organization is knitted together in a network of inspection. Once it is possibleto observe those within an organization, and once those within it are brought into view, itbecomes possible to know them and thereby alter them. Where direct observation is not possible itis necessary to develop indirect supports or “relays” that over time connect the information ofaccumulated periods of time. Periodic inspections of schools and classrooms are examples of theformer, while standardized norm referenced achievement tests are examples of the latter.Systematic observation makes it possible to gather enough information about people todevelop a sense of what is normal. From there it is a relatively simple matter to employ norms tomaintain normality or to begin to move toward some ideal or superior state of affairs. Normsprovide a standard against which individuals can be judged. In the context of inspectionthroughout an organization, normalizing judgement makes possible the fixing of individuals alonga continuum that has reward and punishment at its poles. Examination and inspection, governedby norms and operating within a bureaucratic hierarchy, combine to produce “a normalizing gaze”through which individuals can be judged and classified.In short, the art of punishing, in the regime of disciplinary power, is aimed neither atexpiation, nor even precisely at repression. It brings five quite distinct operations into play:it refers individual actions to a whole that is at once a field of comparison, a space ofdifferentiation, and the principle of a rule to be followed. It differentiates individuals fromone another, in terms of the following rule: that the rule be made to fI.inction as a minimalthreshold, as an average to be respected, or as an optimum toward which one must move.75It measures in quantitative terms and hierarchizes in terms of value the abilities, the level,the “nature” of individuals. It introduces, through this “value-giving” measure, theconstraint of a conformity that must be achieved. Lastly, it traces the limit that will definedifference in relation to all other differences, the external frontier of the abnormal.... Theperpetual penality that traverses all points and supervises every instant in the disciplinaryinstitutions compares, differentiates, hierarchizes, homogenizes, excludes. In short, itnormalizes.70The ideal classroom of the “effective teacher” is one that is firmly situated with this“regime of disciplinary power”. Norms of effective teaching (that emerged from the examinationof teaching behaviors and student achievement) permit comparisons to be made between teachers.Interpreted in this way the norms derived from process-product research place teachers under“the constraint of a conformity that must be achieved.” They serve as “minimum thresholds” ofcompetence, “averages to be respected”, or optimums toward which [teachers] must move.” Thelink to achievement test scores permits the measuring in quantitative terms of “the ability, thelevel, the nature of individuals”, and makes possible the ranking of both teachers and students.Once ranked, individuals who have been constituted as the “ineffective teacher”, or the “learningdisabled student”, who have been placed in “the external frontier of the abnormal” theseindividuals can be improved, normalized, or excluded. Regardless ofwhich treatment they receive,they have been identified, constituted as individuals, and made into sites for the exercise andproduction of power and knowledge.Teachers are thus pressured to teach in much the same way, with the paradoxical resultthat this similarity between them makes the differences between them (in pedagogy, in success)more readily discernible. What is especially interesting about these disciplinary instruments is thatat the same time during which they control teachers, they control students; both are caught in adisciplinary web.In the hierarchically ordered power structures of schools, supervisors can insist thatbehaviors that the research has identified as being the most efficacious are those that teachersmust employ. Teachers are then pressured to standardize their teaching practices, while studentsare pushed to exhibit a narrow range of competencies. This sort of homogeneity permits themaking of comparisons that would be impossible, but for these conditions of formal equality. Onthe basis of such comparisons, individuals (both students and teachers) can be related to one76another, ranked and classified. Comparative knowledge such as this can then be employed in theallocation of rewards, incentives and sanctions, the pursuit or avoidance of which becomes thereason for acting. In this manner teachers are rnled heteronomously.It is important to note that the pressure to teach along the lines suggested by the researchmay be felt either as external pressure or as a conviction about how one ought to teach. Becausethis research is stamped with the legitimacy of science it may be more readily accepted and maymore easily displace the professional judgment of individual teachers who would prefer to teach insome other way. Because it ‘exudes a penality of the norm’ it may bully teachers into teaching asthey are told. In Foucault’s words it has the potential to make us “accept someone else’s authorityto lead us in areas where the use of reason is called for.” In Erickson’s phrase it places “externallimits on the capacity of a teacher to reflect critically on his or her practice.” Effectivenessresearch may well become for the teacher precisely what Kant cautioned his contemporariesagainst. It may become the book that understands for us. In this sense the research would becomethe analog of Kant’s guardians “who have so kindly assumed superintendence over us”. It serveson the one hand to do our thinking for us, and on the other to prevent us from thinking. Kant’s‘way out’ of immaturity is the free use of reason, which in the disciplinary regime, is the verything most threatened.‘Thomas Green, The Activities of Teaching (New York: McGraw Hill, 1971).2E D. Hirsch, “The Primal Scene ofEducation,” The New York Review ofBooks March 2,1989, 29.3Lee S. Shulman, “Paradigms and Research Programs in the Study of Teaching,” inHandbook ofResearch on Teaching, 3d. ed., ed. M. C. Wittrock (New York: Macmillan, 1986),3—36.774Strictly speaking, teacher effectiveness research is only part of the larger researchprogram known as process-product research. Since it is teacher effectiveness research that is mostoften translated into prescriptions for teachers, and since teacher effectiveness research sharesmost of the important assumptions that frame inquiries in process-product research the two willbe viewed as equivalent here. For a discussion of the ways in which teacher effectiveness researchfigures as only a part of the larger tradition see N. L. Gage and Margaret C. Needels, “ProcessProduct Research on Teaching: A Review of Criticisms,” Elementary School Journal, 89, no. 3,1989); Jere Brophy and Thomas L. Good, “Teacher Behavior and Student Achievement,” inHandbook ofResearch on Teaching, 3d. ed., ed. M. C. Wittrock (New York: Macmillan, 1986),328—3 75; Alan Tom, Teaching as a Moral Craft (New York: Longmans, 1984).5Barak V. Rosenshine, “Explicit Teaching,” in Talks to Teachers: A Festschrfi for N LGage, ed. D. C. Berliner and B. V. Rosenshine (New York: Random House, 1987), 75.6lbid., 75.70n the relation between concept learning and critical thinking see Jerrold Coombs,“Critical Thinking and Problems of Meaning,” in Critical Thinking and Social Studies, ed. IanWright and Carol LaBar (Toronto: Grolier, 1987).8lsrael Scheffler, “Basic Mathematical Skills,” in In Praise of the Cognitive Emotions(New York: Routledge, 1991).9lbid., 75.‘°Recent surveys of student opinion in elementary mathematics classes reveals that “manychildren believe that the goals of mathematics problem solving is to find the single correct answeras determined by the teacher. For these children, problems are seen only as opportunities to findand apply proper computational rules to some arbitrary set of numbers.,” See Rochelle G. Kaplan,Takashi Yamamoto, and Herbert P. Ginsburg, “Teaching Mathematics Concepts,” in Toward theThinking Curriculum: Current Cognitive Research: Yearbookfor the Association of Supervisionand Curriculum Development, ed. Lauren B. Resnick and Leopold E. Klopfer (1989), 63;Rochelle G. Kaplan, Burgess, and Herbert P. Ginsburg, “Children’s Mathematical RepresentationsAre Not Always Mathematical,” Genetic Epistemologist (1990).“M. T. H. Chi, J. Feltovich, and R. Glaser, “Categorization and Representation of PhysicsProblems by Experts and Novices,” Cognitive Science 5 (1981): 121—152; D. and H. A. Simon,“Individual Differences in Solving Physics Problems,” in Children Thinking: What Develops?, ed.R. S. Siegler (Hillsdale, NJ: Lawrence Erlbaum Associates, 1978); J. H. Larkin, “The Role ofProblem Representation in Physics,” in Mental Models, ed. D. Gentner and A. L. Stevens(Hilisdale, NJ: Lawrence Erlbaum Associates, 1983); J. H. Larkin, “Enriching Formal Knowledge:A Model for Learning to Solve Problems in Physics,” in Cognitive Skills and Their Application,ed. J. R. Anderson (Hillsdale, NJ: Lawrence Erlbaum Associates, 1981).7812Ji11 H. Larkin and Ruth W. Chabay, “Research on Teaching Scientific Thinking:Implications for Computer Based Instruction,” in Toward the Thinking Curriculum: CurrentCognitive Research: Yearbookfor the Association of Supervision and Curriculum Development,ed. Lauren B. Resnick and Leopold E. Klopfer (1989), 151.‘3Report of the Royal Commission on the Status of Women in Canada (Ottawa, 1970);Jane Gaskell, “Stereotyping and Discrimination in the Classroom,” in Precepts Policy andProcess: Perspectives on Contemporary Education, ed. J. Donald Wilson and Hugh Stevenson(Calgary: Detselig, 1977); Batcher et al., And Then There Were None: A Report Commissionedby the Status of Women Committee, Federation of Women Teachers of Ontario (Toronto, 1975).British Columbia Teachers Federation, Women in Teaching Textbook Study (Vancouver, 1975).Richard Paul et al., “Some Common Problem with Social Studies Texts,” in Critical ThinkingHandbook. 6th to 9th Grades. 172—176; Isabel L. Beck and Margaret G. McKeown, “Substantialand Methodological Considerations for Productive Textbook Analysis,” in Handbook ofResearchon Social Studies Teaching and Learning, ed. James P. Shaver, 496—512; Donald Fisher, “ThePolitical Nature of Social Studies Knowledge,” History and Social Science Teacher 18, no. 4(1983): 2 19—225; Patricia Baldwin and Douglas Baldwin, “The Portrayal of Women in ClassroomTextbooks,” Canadian Social Studies 26, no. 3 (Spring 1992): 110—114.‘4Baldwin and Baldwin, “The Portrayal of Women in Classroom Texts”.‘5Paul et al., “Some Common Problems with Social Studies Texts”; Fisher, “The PoliticalNature of Social Studies Knowledge”.‘6Fisher.‘7Baldwin and Baldwin.‘8Fisher.‘9Baldwin and Baldwin, McKeown and Beck, “Substantial and MethodlogicalConsiderations for Productive Textbook Analysis.”20Jere Brophy, “Classroom Organization and Management,” in Elementary SchoolJournal, 83, no. 4 (1983): 265—285.21John S. Kounin, Discipline and Group Management in Classrooms (New York: Holt,Rinehart & Winston, 1970).22Brophy, 267.23Ibjd 267.24Jbjd 267.25Jbjd 268.7926J owe this term to Thomas Green’s discussion of the teaching concept in “A Topologyof the Teaching Concept.”27Alasdair Maclntyre, After Virtue (Notre Dame, Indiana: University of Notre DamePress, 1984), 187.28Jbjd29R. S. Peters, Ethics and Education (London: Allen and Unwin, 1966): 31.30Jurgen Habermas, Moral Consciousness and Communicative Action trans. ShierryWeber Nichoolson (Cambridge, MA: MIT Press, 1990).31Ibid., 14.32Max Miller, “On First Learning How to Contradict,” in Children ‘s World andChildren Language, ed. J. Cook-Gumperz, W. Corsaro, & J. Streeck (Berlin: Mouton deGruyter, 1986).33Robert E. Young, A Critical Theory of Education: Habermas and Our ChildrenFuture (New York: Teachers College Press, 1990), 117.34Richard I. Arends, Learning To Teach (New York: McGraw Hill, 1991); Gary D.Borich, Effective Teaching Methods (Toronto: Maxwell Macmillan, 1992); Allan C. Ornstein,Strategiesfor Effective Teaching (New York: Harper and Row, 1990); Thomas L. Good and JereBrophy, Looking Inside Classrooms, 5th ed. (New York: Harper Collins, 1990).35L. Canter and D. M. Canter, Assertive Discipline (Los Angeles: Canter and Associates,1976). It is important to note that despite the enormous popularity of the Canters’ approachamong teachers, the research upon which Assertive Discipline rests only weakly supports theprogram of interventions.36Robert Dreikurs, Psychology in the Classroom: A Manual for Teachers, 2d. ed. (NewYork: Harper and Row, 1968).37Richard A. Schmuck and Patricia Schmuck, Group Processes in the Classroom, 5th ed.(Dubuque, Iowa, W. C. Brown, 1988).38Walter Doyle and K. Carter, “Academic Tasks in the Classroom,” Curriculum Inquiry14 (1984): 124—149; John Kounin, Discipline and Group Management in Classrooms (NewYork: Holt, Rinehart & Winston, 1970); V. Gump, “School Settings and their Keeping,” inHelping Teachers Manage Classrooms, ed. D. L. Duke (Alexandria, VA: Association forSupervision and Curriculum Development, 1982); Walter Doyle, “Classroom Organization andManagement,” in Handbook ofResearch on Teaching, 3d. ed., ed. M. C. Wittrock (New York:Macmillan, 1986).8039E. T. Emmer, Carol Evertson, J. Sanford, B. S. Clements, W. E. Worsham, ClassroomManagement for Elementary Teachers (Englewood Cliffs, NJ: Prentice-Hall, 1984); E. T.Emmer, Carol Evertson, J. Sanford, B. S. Clements, W. E. Worsham, Classroom Managementfor Secondary Teachers (Englewood Cliffs, NJ: Prentice-Hall, 1984).40Brophy, “Classroom Organization,” 266.41Walter Doyle and K. Carter, “Academic Tasks in Classrooms,” Curriculum Inquiry. 14(1984): 146.42J0 W. Meyer, “Levels of the Educational System and Schooling Effects,” in TheAnalysis of Educational Productivity, ed. C. Bidwell and D. Windham (Cambridge: Baflinger,1980), 53.43As we shall see in the chapter following there are educationists who think attitudes ofthis sort are perfectly reasonable, and ought to be fostered in schools.44Robert Marzano, Ronald S. Brandt, Carolyn Sue Hughes, Beau Fly Jones, Barbara Z.Presseisen, Stuart C. Rankin and Charles Suhor, Dimensions of Thinking: A Framework forcurriculum and Instruction (Alexandria, VA: Association for Supervision and CurriculumDevelopment, 1988).45LeRoi B. Daniels, “What is the Language of the Practical? Curriculum Theory Network.1974).‘The following discussion owes a great deal to the exchange between Robin Barrow,Morwenna Griffiths, and Richard Smith in Journal ofPhilosophy ofEducation, 21, no. 2 (1987):187—214.47Daniels, p. 14.48quoted in Barrow, 190.49Israel Scheffler, Conditions of Knowledge. (Chicago: University of Chicago Press,1961), 20.50Barrow, 19251Robert Sternberg, “Sketch of a Componential Subtheory of Human Intelligence,”Behavioral and Brain Sciences 3 (1980): 573—614; “Components of Human Intelligence,”Cognition 15 (1983): 1—48; Beyond IQ: A Triarchic Theory of Human Intelligence (London:Cambridge University Press, 1985).52Jonathan Baron, Rationality and Intelligence (London: Cambridge University Press,1985); U. Neisser, “Components of Intelligence or Steps in Routine Procedures,” Cognition 15(1983): 189—197; Francis Schrag, Thinking in School and Society (New York: Routledge, 1988).8153John McPeck, Critical Thinking and Education (New York: St. Martin’s, 1981).54Gilbert Ryle, On Thinking (Totowa, NJ: Roman and Littlefield, 1979).55Lauren B. Resnick, “Toward the Thinking Curriculum: An Overview,” in Lauren B.Resnick and Leopold E. Klopfer ed. Toward the Thinking Curriculum: Current CognitiveResearch: Yearbook for the Association of Supervision and Curriculum Development(Alexandria, VA: ASCD, 1989), 1—18. See also Isabel L. Beck, “Improving Practice ThroughUnderstanding Reading”; James A. Minstrell, “Teaching Science for Understanding”; Jill H.Larkin and Ruth W. Chabay, “Research on Teaching Scientific Thinking: Implications forComputer Based Instrnction,” all in Toward the Thinking Curriculum: Current CognitiveResearch: Yearbookfor the Association ofSupervision and Curriculum Development, ed. LaurenB. Resnick and Leopold E. Klopfer (Alexandria, VA: ASCD, 1989).56James D. Wallace, Virtues and Vices (New York: Cornell University Press, 1978).57Jonathan Baron, Rationality and Intelligence (London: Cambridge University Press,1985), 8558Wilfred Sellars, Science, Perception and Reality. (London: Routledge and Kegan Paul,1963).59U. S. National Institute of Education, Study Group on the Conditions of Excellence inAmerican Higher Education. Involvement in Learning: Realizing the Potential of AmericanHigher Education (Washington: U. S. Government Printing Office, 1984), 232.60Hazard Adams, Antithetical Essays in Literaty Criticism and Liberal Education(Tallahassee: Florida State University Press, 1990), 233.61An early draft version of the Mission Statement indicates the relative lack of statusknowledge is to have in the new curriculum. Knowledge is not even mentioned. Significantly, noris the idea of enabling learners to develop their potential mentioned. The complete text is asfollows: “The major purpose of the British Columbia school system is to enable our youth todevelop the intellectual skills and attitudes needed to maintain a healthy society and prosperouseconomy.” In the final version of the Year 2000 knowledge is, of course, indicated as one of thethree “learning dimensions,” which make up the “framework for learning,”. One wondersnevertheless how much of the original derogation of knowledge is carried over into the fmaldocument. The extent to which the economy is to be in the driver’s seat with respect to justi1jingeducation and defining its purpose is also worth noting.62Carnegie Forum on Education and the Economy, Task Force on Teaching as aProfession, A Nation Prepared: Teachersfor the 21st Century (Hyattsville, MD: Carnegie Forumon Education and the Economy, 1986), 20.63A Nation Preparec4 20.8264Province of Saskatchewan, Department of Education, Training and Employment, SocialStudies JO: Social Organizations—Curriculum Guide. 1992), 1665Province of Saskatchewan, Department of Education, Training and Employment, SocialStudies JO: Social Organizations—Curriculum Guide. 1992), 16.66Province of Saskatchewan, Department of Education, Training and Employment, SocialStudies 10: Social Organizations—A Curriculum Guide. 1992), 16.67Plato, “Meno,” in Plato, 12 volumes, translated by H. N. Fowler (Cambridge, MA:Harvard University Press, 1921), 97a198c.68This view has recently come under attack from externalist epistemology which denies therelevance of the subject’s own evaluations and perspective. See Frederick F. Schmitt, Knowledgeand Belief (London: Routledge, 1992).69Michel Foucault, Discipline and Punish (New York: Vintage Books, 1979).70Ibid., 182—183.83CHAPTER THREETHE CONTESTED NATURE AND VALUE OF AUTONOMYBefore going any further it is necessary to consider whether promoting intellectualautonomy is worth the effort. As indicated in the first chapter not everyone thinks so highly ofautonomy. Iris Murdoch, for example, likens the solitary, autonomous man to Lucifer, expelledfrom heaven for doubting the wisdom and love of God, and laments that autonomy is accordedsuch prestige in the political and social philosophy of Western democracies.’ Dearden thinks thevalue of autonomy is greatly over-rated since independence of mind in itself does not ensuremoral or critical intelligence. The great criminal and robber baron capitalist may instantiateautonomy more than the educated person. The educated person, he thinks, is more the product ofdisciplined inquiry than of unfettered freedom2 Benn echoes this sentiment in his suspicion thatCesare Borgia might well have been every bit as autonomous as Socrates.3 Neither Benn norDearden see much reason to expect the autonomous man will be especially virtuous or reasonable.Elizabeth Telfer, for her part, worries that many accounts of autonomy permit an arrogantlysolipsistic approach to testimony that rejects the views of others, especially those views it wouldbe rash to ignore.4 Gerald Dworkin wonders why autonomy is held in such high regard since itconflicts, or so he thinks, with a host of other values such as loyalty, objectivity, commitment,benevolence and love. Moreover he believes that autonomy as it is commonly presented rules outvaluable human practices such as promising, worship, obedience to command, even conformity tolaw.5There seems to be a conflict between self-determination and notions of correctness andobjectivity. If we are to make reasonable choices, then we must be governed by canons ofreasoning, norms of conduct, standards of excellence that are not themselves the productsof our choices. We have acquired them at least partly as the result of others’ advice,example, teaching—or, perhaps, by some innate coding. In any case, we cannot havedetermined these for ourselves.6Among educational philosophers Brian Crittenden has raised similar objections. Crittendenbelieves that a commitment to intellectual autonomy is a commitment to the idea that the agent isto be the originator of all her beliefs, as well as of the standards by which certain beliefs are to be84judged more justified than others.Intellectual autonomy would require.. .that a person not accept any of his important beliefsprimarily on the authority of others, but on his own experience, his own reflection onevidence and argument, his own sense of what is true and right. For complete intellectualautonomy it would also seem necessary that a person should determine for himself thesecond order question about what constitutes a true claim, adequate evidence, a justifiablemoral principle, and the like. Even the crucial concepts in which he perceives andunderstands should be of his own design, or at least accepted from others only because heis personally satisfied that such concepts are satisfactory.7Crittenden contends such a view is patently mistaken. The radical subjectivism implied bythis view of autonomy dissolves into solipsism where no distinction can be drawn betweenknowledge and belief where there can be no ground for claiming anyone is mistaken. Converselythose who hold to such a view cannot even intelligibly say the autonomous individual believes thatwhich is true for him. For where it is not possible to identify error, neither is it possible todistinguish truth or rationality. Given that education is “an induction into the standards of truthand rationality... as they have been articulated in the on-going public traditions of humanunderstanding” autonomy cannot sensibly be promoted as an aim of education.8Crittenden haslittle patience even for those like Kant whose arguments for autonomy acknowledge theconstraints of rational criteria. According to Crittenden the basis of Kantian autonomy is “the ideaof the will of every rational being as a will which makes universal law.”9 Crittenden claims thisview of self-legislation is both morally hazardous and ftindamentally incoherent. First, anindividual cannot will a universal law without being prepared to challenge the autonomy ofeveryone else. If we suppose all rational individuals must agree on which principles to adoptautonomy seems to be without much point. If we assume that individuals will frequently disagree,the injunction to will a universal law is an invitation to conflict and totalitarian oppression, just asargued by Berlin.’0 Secondly, as Baier points out, “it is logically impossible to claim eachindividual is subject only to the laws of his own creation. If no member of a society were subjectto the will of any other, then there would simply be no law and so no legislation, including selflegislation.” The problem, on this view, is that the very idea of legislation and hence of selflegislation is inappropriate to the determination of which moral principles should apply. Even ifBaier is mistaken in dismissing the metaphor of legislation from moral theory, there is reason, says85Crittenden, to suppose his criticism applies to the field of education. The reason is the simple onethat an individual can never obtain an education by means of self-legislating acts. We would bebetter advised, on this view, to jettison talk of autonomy in favour of talk that emphasizes theneed of students to understand what is learned, not on the authority of the teacher, but on thebasis of the evidence that would justify their claim to know.INTELLECTUAL AUTONOMY AND THE WILL TO HETERARCHYDworkin, for his part, attempts to argue against our retaining anything but a weak,procedural notion of autonomy, one that stresses second order reflection, but which does notspecify any content to this reflection. He raises the possibility that ‘autonomy’ may not be aparticularly useful concept. On Dworkin’s view a person could rationally give up what isordinarily recognized as autonomy and still be autonomous. Dworkin puts the point this way:Suppose we have a person who has not been subjected to the kinds ofinfluence—whatever they turn out to be—that interfere with procedural independence.Suppose the person wants to conduct his or her life in accordance with the following: Dowhatever my mother or my buddies or my leader or my priest tells me to do. Such aperson counts, in my view, as autonomous. 12It is hard to conceive of a view more distant from Kant’s than this. Dworkin calls hisconception of autonomy weak in contrast to the strong view held by many philosophers. Whatdistinguishes the two views is that while the Kantian view gives substantive content to the notionof autonomy, Dworkin’s does not. Dworkin’s autonomy is a purely formal notion that does notspecify which decisions are consistent with autonomy and which aren’t. The concept of autonomyshould not suggest any content to the decisions an autonomous person makes. One can decidethat the best life is a monastic life governed by vows of obedience, or the best course of action thecourse dictated by one’s political party or guru.Dworkin characterizes his procedural conception of autonomy in the following way:Autonomy is conceived of as a second order capacity of persons to reflect critically upontheir first-order preferences, desires, wishes and so forth and the capacity to accept orattempt to change these in light of higher-order preferences and values. By exercising sucha capacity, persons define their nature, give meaning and coherence to their lives, and takeresponsibility for the kind of person they are. 13With this conception of autonomy we avoid the conflict between the different values86indicated above, and refrain from specifying what form a meaningfiul and coherent life must take.To understand what Dworkin means by a formal procedural notion of autonomy we need to focuson the nature of deliberation prior to the agent’s making a promise or taking a vow of obedience.In the case of one who decides always to do what his mother wants, it is the procedure thateventuates in the decision that is relevant to a determination of autonomy. The temporal pointrelevant to a determination of autonomy is one that precedes the decision. It is the deliberationprior to the decision that is to count, not the content of the decision or the agent’s actions thatfollow. We cannot object that autonomy is lost once the decision is made to defer to someoneelse’s judgment since, on Dworkin’s view, the character of the agent’s deliberations carries overto characterize the actions that follow from his decision. In the case of the man who decides hewill do only what his mother wants, “we must,” says Dworkin, “make reference to his intentionsto do what his mother wants. It is his decision, arrived at freely, backed by reasons that makes hismother’s wishes effective in determining his actions. He is doing what he wants to do. He isleading just the kind of life he thinks is worth leading. How can he not be autonomous?”14INTELLECTUAL AUTONOMY AS A GOAL OF EDUCATIONFor the sake of argument let’s accept that intellectual autonomy is a desirable trait. Mustwe accept also that it is equally a desirable educational goal? The answer given in philosophy ofeducation has been an ambivalent one. Those who wish to see more of a place for intellectualautonomy in schools have tended to stress the importance of evidential reasoning, while thosewith less enthusiasm for autonomy emphasize the need to deny autonomy to the young becausechildren are too immature to benefit from the exercise of whatever partial autonomy they dopossess. The arguments in favour of intellectual autonomy depend first upon a theory of epistemicjustification known as internalism, and second upon faith in the ability of children to benefit fromrational discussion.15 In the pages to follow I intend to consider the challenge to intemalism thathas emerged in the last decade, and the arguments that cast doubt on the desirability of fosteringepistemic independence in schools. I intend to preface this discussion with a survey of internalistcommitments in educational philosophy, in the hope that the contrast between internalism and itscritics will throw some light on the issue of autonomy in justification.87Educational philosopher Kenneth Strike once remarked that an essential task for educationis the distribution of rationality.’6 On Strike’s account one obstacle that stands in the way ofsucceeding at this task is the inability of students to revise their beliefs in light of new andcompelling evidence. In this we find Strike echoing Mill’s worry that the practical hazard ofaccepting beliefs on the testimony of authorities is an impaired capacity for rational belief revision.There is a class of persons.. .who think it enough if a person assents undoubtingly to whatthey think true, though he has no knowledge whatever of the grounds of the opinion andcould not make a tenable defense of it against the most superficial objections. Suchpersons if they can once get their creed taught from authority, naturally think that nogood, and some harm, comes of its being allowed to be questioned. Where their influenceprevails, they make it nearly impossible for the received opinion to be rejected wisely andconsiderately, though it may still be rejected rashly and ignorantly; for to shut outdiscussion entirely is seldom possible, and when it once gets in, beliefs not grounded onconviction are apt to give way before the slightest semblance of an argument. Waiving,however, this possibility—assuming that the true opinion abides in the mind, but abides asa prejudice, a belief independent of and proof against, argument—this is not the way inwhich truth ought to be held by a rational being. This is not knowing the truth. Truth, thusheld, is but one superstition the more, accidentally clinging to the words which enunciate atruth.’7The preceding chapter’s discussion of epistemic practices in schools showed, I think, thatthe practices, if not teachers themselves, “make it nearly impossible for the received opinion to berejected wisely and considerately”. Moreover these practices interfere with the intelligentgrounding of belief If it is true that beliefs not grounded in any genuine understanding will beeither feebly held, or dogmatically adhered to, the consequence of teachers neglecting the elementofjustification in knowledge claims is an impaired capacity for rational belief revision. This pointhas been argued most lucidly in educational philosophy by Green’8 whose distinction betweenevidential and non-evidential styles of belief brings Mill’s concern over unintelligent belief revisioninto educational focus.Over the last decade the evidentialism typified by Green has come to be known as a typeof epistemic internalism, largely as a result of the criticism of externalists who doubt that non-evidential belief is in any way unintelligent. As with evidentialism, internalism holds thatknowledge and justification are based on reasons or evidence that are somehow available to theagent. More precisely, the evidence must be possessed by the agent and appreciated as evidence.88Evidence that is appreciated as such is said to be epistemically within the agent’s grasp, orepistemically internal. It should be plain that the notion of epistemic independence is an internalistnotion, so any criticism of internalism will bear on the acceptability of intellectual autonomy as Ihave conceived it. Some discussion of the debate between internalist and externalists is thereforenecessary. Let’s begin with a survey of what we can now recognize as the internalist commitmentsin philosophy of education.The Austinian TraditionAccording to internalism, epistemic agents, in being justified, claim (or could claim) a rightto know that includes being free of any reproach for claiming to know. This feature of internalismis obvious in ordinary language analysis of the type exemplified by Austin and Urmson, and ispresent to a significant degree in the writing of educational philosophers.Austin claimed that “whenever I say I know, I am always liable to be taken to claim that,in a certain sense appropriate to the kind of statement (and to present intents and purposes), I amable to prove it.”19 [my emphasis] Similarly, Urmson noticed that the verbs know, guess, supposeand estimate belong to the family of verbs that indicate the evidential status of a statement and theepistemic situation of the agent:This is the group [of verbs]. . . which is used to signal what the degree of reliability isclaimed for, and should be accorded to, the statements to which they are conjoined. Thus‘I guess that this is the right road to take’ is a way of saying that this is the right road,while indicating that one is just plumping and has no information, so that the statementwill be received with the right amount of caution; ‘I know’ shows that there is all theevidence that one could need, and so on.20This sort of analysis has led educational philosophers to emphasize the importance ofmaking epistemic justifications available to students.2’ It also conforms with the tripartitedefinition of knowledge that is still widely accepted, the Gettier counter-examplesnotwithstanding.22 For example, prior to Gettier, Ayer concluded that “the necessary andsufficient conditions for knowing something is the case are first what one is said to know be true,and secondly that one be sure of it, and thirdly that one should have the right to be sure.”23Chisholm’ s account, roughly identical to Ayers, states that “S knows that h is true’ means: (i) S89accepts h; (ii) S has adequate evidence for h; and (iii) h is true.”24 Following Gettier25 Chishoimoffers this analysis of knowledge: “S knows at t that h is true, provided (1) S believes h at t; (2) his true; and (3) h is evident at t for S.” As late as 1989 Chisholm was still insisting that knowledgeis justified true belief26 Similarly Roderick Firth insists on the internal character of being justifiedin beliefTo decide whether Watson knows that the coachman [committed a crime] we must decidewhether or not Watson is justified in believing that the coachman did it. Thus if Watsonbelieves that the coachman did it, we must decide whether his conclusion is basedrationally on the evidence.27Scheffler and the Evidence ConditionCommon to most of these analyses is the theme that before it can be said an agent isjustified in holding a belief that agent must possess the justification for this belief This theme ispicked up and explored by Israel Scheffler in his influential Conditions of Knowledge. From thestandard analysis of knowledge Scheffler derived practical guidance for the teacher, and outlinedcriteria by which we can understand what it means for a student to know that p.28 Here Schefflerstresses that the distinction between knowing and having a true belief is made by reference to thesurplus value ofknowledge over true beliefIn every case where evidence is required for the right to be sure, knowing involves notmerely having adequate evidential data but also appreciating their value as data, in thelight of an appropriately patterned argument.Citing Augustine, Scheffler further observes that:the pupil who knows.. . is not just someone who has a belief which is true.... He mustfurther have considered within himself whether what has been said is true. He must haveengaged in a personal process of evaluating the belief in question, by reference to his ownsource of interior truth.29 [my emphasis]In this discussion Scheffier’ s commitment to internalism is explicit. It is the student whomust possess adequate evidence for the beliefs she comes to hold. To this end she must be in aposition to follow and evaluate the evidentiary argument that purports to justify the claims athand. Crucially, she must come to feel the force of this argument. Among the obvious implicationsto be drawn is that teachers must provide students with the evidentiary argument that supports agiven claim. They must also ensure that students understand the argument sufficiently well to see90its force. In so doing they guard against students holding beliefs they do not understand or thatare for them inert. Further, without some sense of the grounds for belief students cannotunderstand the nature of the discipline under study, or its distinctive truth tests and standards.Whether the subject be history, physics or mathematics, each has distinctive methods forascertaining the truth, and distinctive standards for judging success in this regard.3°But it is notsimply that the disciplines are ends in themselves and their study justified solely in terms of theirintrinsic value. Rather the disciplines are arenas in which students can come to grips with“whatever competent investigation has achieved in the way of reliable knowledge and skill” inorder that they might be initiated into “the critical life” characterized by “the quest for and theexchange of reasons.” 31In training our students to reason we train them to be critical. We encourage them to askquestions, to look for evidence, to seek and scrutinize alternatives, to be critical of theirown ideas as well as those of others. This educational course precludes taking schoolingas an instrument for shaping their minds to a preconceived idea. For if they seek reasons, itis their evaluation of such reasons that will determine what ideas they eventually accept.32It is on this basis that Scheffler denies the appropriateness of appeals to authority ineducational contexts. Teachers should not base their teaching on such appeals, and studentsshould not be permitted to hold their beliefs on such grounds.Green and Evidentially-Held BeliefIn a similar vein Green has concluded, from an analysis of the family of concepts related toteaching, that beliefs can be held evidentially or non-evidentially.When beliefs are held without regard to evidence or contrary to evidence, or apart fromgood reasons or the canons for testing reasons and evidence, then we may say they areheld non-evidentially. It follows that beliefs held non-evidentially cannot be modified byintroducing evidence or reasons or by rational criticism. When beliefs, however, are held“on the basis of’ evidence or reasons, they can be rationally criticized, and therefore canbe modified in the light of Iuirther evidence or further reasons, then we shall say they areheld evidentially.33Green adds that students who hold true beliefs non-evidentially cannot know why suchbeliefs ought to be taken as true. For them such beliefs can only be ‘correct’, and as aconsequence can never be known. “We cannot be said to know that a belief is true,” says Green,91“if we cannot give any reasons for it, any explanation of it, or any evidence in support of it.” Hisadvocacy of evidential styles of belief does not rest simply on a definition of knowledge. Hisanalysis points to the practical limitation of ‘correct’ over known beliefs: namely that those whodo not hold their beliefs on the basis of good reasons and evidence are not in a position to changetheir beliefs in light of better reasons and more substantial evidence. Rational belief revision isimpaired by the absence of well grounded belief On this point Green echoes Mill’s concern that abelief “not grounded on conviction is [either] apt to give way before the slightest semblance of anargument,” or is apt to “abide as a prejudice, a belief independent of; and proof against,argument.”Hirst and the Forms of KnowledgeThe very idea of a rational mind has been linked in educational philosophy to the variousforms of knowledge and the truth tests that distinguish one from the other. Most of this thinkingcan be traced to the work of Paul Hirst, but its influence is widespread. On Hirst’s account, tohave a rational mind just is to be in possession of knowledge, robustly conceived.The forms of knowledge are.. .the basic articulations whereby the whole of experience hasbecome intelligible to man, they are the fundamental achievement of mind. Knowledgehowever must never be thought of merely as vast bodies of tested symbolic expressions....To acquire knowledge is to become aware of experience as structured, organized andmade meaningful in some quite specific way, and the varieties of knowledge constitute thehighly developed forms in which man has found this possible. To acquire knowledge is tolearn to see, to experience the world in a way otherwise unknown, and thereby come tohave a mind in a fuller sense.35On this account initiation into the various forms of knowledge will involve one’s becomingfamiliar, not only with the conclusions of a field, but also with the truth tests and critical standardsthat yield those propositions that comprise the justifications for truth claims.More recently McPeck has argued a similar position in his effort to defend a conception ofcritical thinking based on a knowledge of the disciplines. In either case, the development ofrationality, or the capacity for critical thought hinges on students having access to the ‘logicalspace of reasons’ •36This sort of argument has implications for the reasonableness of deference to teacher92authority, as deCastell notes:The aim of developing informed rational autonomous persons entails that what the learneraccepts as true comes to be increasingly grounded in “good reasons” for such acceptanceas true. But as Hirst pointed out, the criteria for the truth of propositions vary with thekind of proposition asserted, so that a progressive understanding of the discipline itself isthe necessary condition of an increasingly rational acceptance of educational authority. Alltoo obviously, then, students who are not taught the discipline (which provides suchcriteria) are denied access to the means of rational autonomous development.37Siegel on Being Appropriately Moved by ReasonsSiegel’s conception of critical thinking, though different from McPeck’s, is nonethelessjust as committed to the notion that rationality requires autonomy in justification. The rationalagent, says Siegel, is appropriately moved by reasons. Thus a critical thinker is one “whoappreciates and accepts the importance and convicting force of reasons. When assessing claims,making judgments evaluating procedures, or contemplating alternative actions, the critical thinkerseeks reasons on which to base her assessments, judgments, and actions.”38 From this descriptionof the critical thinker Siegel is able to derive a commitment to epistemic independence. He claimsthat once we accept critical thinking as an important goal of education we have explicitlyacknowledged the importance of autonomy.If we think it good that a student become a critical thinker, we must approve as well of thestudent’s ability and disposition to consult her own independent judgement concerningmatters of concern to her. The critical thinker must be autonomous—that is, free to actand judge independently of external constraint, on the basis of her own reasoned appraisalof the matter at hand.39Epistemic PaternalismMill, though well known as a champion of freedom of thought and expression, isnotorious for his paternalistic denial of these freedoms for children. According to Mill, childrenought to be denied liberty on the ground they have not attained “the maturity of their faculties”.4°Mill indirectly elaborates this notion of maturity by adding that the immature are those who “arestill in a state to require being taken care of by others.”41 This is a group that includes ‘barbarians’as well as “young persons below the age which the law may fix as that of manhood orwomanhood.”42Barbarians are to be denied liberty for the same reason liberty is to be denied the93young. The immaturity of “those backward states of society in which the race itself may beconsidered as in its nonage” is seen as sufficient justification for denying “barbarians” liberty.43Thus:Despotism is a legitimate mode of government in dealing with barbarians, provided theend be their improvement, and the means justified by actually effecting that end. Liberty,as a principle, has no application to any state of things anterior to the time when mankindhave become capable of being improved by free and equal discussion. Until then there isnothing for them but implicit obedience to an Akbar or Charlemagne, if they are sofortunate as to find one. But as soon as mankind have attained the capacity of beingguided to their own improvement by conviction or persuasion (a period long since reachedin all nations with whom we need here concern ourselves), compulsion, either in the directform or in that of pains and penalties for non-compliance, is no longer admissible as ameans to their own good, and justifiable only for the security of others.”AMill’s argument justifies the granting of liberty on epistemological grounds and denies iton the basis of limited rational capacity (defined in terms of the ability to benefit from rationaldiscussion). Thus, while liberty is necessary to the pursuit of the truth, the development ofindividuality and genius, indeed, to the “mental well-being of mankind”, children and barbariansare not entitled to liberty since they are incapable of “improvement by free and equal discussion.”As a result “there is nothing for them but implicit obedience to an authority.”Richard Peters and Francis Dunlop have both argued, though in different ways, that sincechildren are initially incapable of evaluating the truth claims of their teachers, teachers needn’tdevote much time to convincing their students of the rightness of a particular view. Insteadteachers should expect students to take their accounts on trust, and teach accordingly. Dunlop’scondemnation of premature attempts to foster reasonableness is especially forthright, thoughexpressed in the context of arguing against child-centred curricula.The task of the teacher is not to indulge the pupil’s own naive wants and aspirations; it israther to transform desires and aspirations. Therefore, teachers properly display adownward regard for their pupils who in turn evince a reciprocal upward regard. Moreprecisely, the student should trust his teachers beyond all reasons (since he has no criteriaby which to test them) and submit to their authority.45Peters, too, believes that up to a certain point in the child’s development she lacks thecriteria by which to test teachers’ claims. Children, he says, lack the ability to evaluate truthclaims, perhaps not in any general way, but relative to their familiarity with a practice, whether the94practice be science, philosophy, or what have you. Those sufficiently familiar with a practice are,in Peters’ phrase, on the “inside of a form of life” of which the practice is a part. Those who havelittle understanding of the substantive content of a practice, its standards of appraisal, or thosewho do not possess the rational passions that would commit them to these standards are said tobe “outside the form of life”. On Peters’ view students must assiduously study their way insidethis form of life, and, initially at least, they must recognize that their being outside the practiceleaves them reliant on their teachers. Alasdair Maclntyre takes a similar view. His conception ofpractice, you will recall, is more precisely spelled out than Peters’ notion of a form of life, and lesstied to academic disciplines. Nonetheless, every practice, whether it be architecture, farming, orthe historian’s craft, requires a certain kind of relationship between teacher and student.It belongs to the concept of a practice as I have outlined it. . . whether we are painters orphysicists or quarterbacks or indeed just lovers of good painting or first rate experimentsor a well-thrown pass: that its goods can only be achieved by subordinating ourselveswithin the practice in our relationship to other practitioners. We have to learn to recognizewhat is due to whom; we have to be prepared to take whatever self-endangering risks aredemanded along the way; and we have to listen carefully to what we are told about ourown inadequacies and to reply with the same carefulness for the facts.’Practitioners, for their part, are equally bound by the rules and standards that partiallydefine the practice, and that set a limit on the practitioner’s relation to the uninitiated. For thenovice as well as the master to enter into a practice is “to accept the authority of those standardsand the inadequacy of [their] own performance as judged by them. It is to subject [their] ownattitudes, choices, preferences and tastes to the standards that currently and partially define thepractice.”47Of interest in this passage is the idea that though the learner is epistemically subordinate tothe teacher, the teacher is subordinate to the authority of the standards that partially define thepractice. Two consequences for my argument follow from this observation. Teachers themselvesmust be sufficiently “inside the practice” to know and to be committed to the relevant standards,and they must also teach in such a way that these standards are made manifest to their students.Of course, the novice is, in a special sense, more subordinate to the master as well as thestandards, since the master, by virtue of being a master, is able to understand and appreciate the95rules that delimit the practice. The novice must take many more matters on faith. For Maclntyrethe relevant intellectual virtue for the novice is one of humility.1f on starting to listen to music, I do not accept my own incapacity to judge correctly, Iwill never learn to hear, let alone appreciate, Bartok’s last quartets. 1f on starting to playbaseball, I do not accept that others know better than I when to throw a fast ball and whennot, I will never learn to appreciate good pitching let alone to pitch. In the realm ofpractices the authority of both goods and standards operates in such a way as to rule outall subjectivist and emotivist analyses ofjudgment.48Maclntyre’ s discussion of practices points to a variety of the ways in which the integrity ofpractices is threatened. The main threat comes from what he calls “goods external to thepractice”. Goods external to a practice, such as money, fame and the like, are external in the sensethey belong to some individual as property, and are not essentially an aspect of the practice.External goods can always be acquired outside the practice, unlike goods internal to the practicethat are acquired only in the pursuit of those excellences that partly constitute the practice itself.These two sorts of goods stand in an antagonistic relation to one another since external goodsare, in principle, always capable of being given more priority, especially in the institutional settingswhere many practices are carried out. Against this ever present threat posed by external goods arethe virtues, that help preserve standards and the excellences that constitute achievement in thepractice. Humility is among Maclntyre’s list of virtues, along with courage, justice and honesty.Judging from Maclntyre’ s remarks regarding the epistemic dependence of the uninitiated thesometimes subjectivist and emotivist perspectives characteristic of child-centred educationaltheory would likely be viewed by Maclntyre as a threat to the integrity of those practices intowhich education is meant to be an initiation.A teacher’s authority, on this view, is based on her understanding of the standards andrules imbedded in the practice. She is as much subject to these standards as her students will be ifshe succeeds as their teacher. Obedience to authority is thus called for because without suchobedience teachers cannot teach well, and students will never learn.To an extent children do represent a special case, thus educational philosophers aresomewhat divided on this issue: Scheffler49 and Green5°coming close to advocating for childrenthe epistemic rights that Mill advocates for those who have reached adulthood, while Peters51,96Hirst52 and Dearden53 declare themselves as standing somewhere near the midpoint between thosewho press for less epistemic paternalism and those who push for more. In their own ways each ofthe latter three stress that the immaturity of the child, and the child’s relative ignorance providegrounds for some degree of authoritarian and paternalistic teaching. They do, however, viewautonomy as a valuable trait that under the appropriate circumstances teachers should endeavor todevelop in their students. A few, such as R. T. Allen54 and Francis Dunlop55,object even to thisrather modest liberalism. Since their arguments could be used to justify the epistemic practices ofteachers described thus far they deserve some examination.The Rationality of Epistemic DependenceAllen, for example, indicates that when students demand epistemic justifications they haveamong the best justifications when their teachers respond with “Because I say so”. Allen, inparticular, objects to the rationalism of Peters and Hirst and states that “living by the Rationalistprinciples. . . [of] Hirst and Peters would lead us into radical autism”56.Which principles does hehave in mind? Hirst and Peters claim that there is a prima facie antagonism between reason andauthority.57 Allen accepts Hirst and Peters’ definition of authority (“authority is present whensomething is correct or to be done because an individual or body of men, who has been given theright, says so”), but objects to their demand that authority be ‘rationalized’, brought in line withreason. Hirst and Peters insist that authorities in educational institutions ought to be viewed onlyas provisional authorities because authorities are often mistaken. Hence their pronouncementsought to be viewed with a tinge of skepticism and caution. This skepticism is reasonable only sofar as it arises from a familiarity with the forms of knowledge under study. Once students arefar enough along in their studies to be capable of informed judgment it is sensible to expectfruitful criticism from them, but not before. This point notwithstanding Hirst and Peters areexplicitly in favour of epistemic independence as a long term goal of educators:There must be provisional authorities in the different forms and fields of knowledge withwhich universities and schools are concerned. Their job is to hand on an inheritance insuch a way that others can come to criticize it and eventually dispense with their teachers.They must exercise their authority in such a way that another generation can learn to livewithout them.5897Elsewhere Peters has stated that authority in schools is needed so “another generation canlive without authority.”59Allen takes Hirst and Peters to task on this last point. He claims that asa matter of fact higher and lower education teaches students to accept cognitive authorities, andthis is just as it should be.In Allen’s opinion, governance by authorities optimizes our chances of getting a soundeducation. Authorities exist to rule on the admissibility of arguments, as well as the acceptabilityof candidates for tenure, promotion and office. They decide who shall be certified, who shallreceive grants, who shall referee papers and so on. They are also, says Allen, responsible fordisseminating and protecting reigning orthodoxies, not simply through argument and criticism, butthrough administrative controls such as those just listed. The protection of orthodoxies, while itmay extend to the regulation of specific arguments, will ideally only amount to the maintenance ofgeneral theories, interpretations or paradigms, until such time as the accumulation of evidenceagainst them, along with the emergence of a superior alternative, requires they be abandoned ormodified.In addition, Allen insists there is far too much knowledge, of such great complexity, that itis unrealistic to think any one person could know enough to avoid epistemic dependence.Whatever may have been the case in the past, it is definitely impossible for anyone todayto be a polymath and to know everything about everything. Each academic disciplinebecomes ever more divided and each expert is tied to an ever denser yet narrower field.And so, even they, for what lies beyond their fields, have to rely on the authority of otherexperts in those other fields.60According to this view, there is a division of cognitive labour that fhnctions in adisciplinary network of competence, one sub-field dependent on another, yet informing it as well.Experts within one field are in a relation of epistemic dependence to experts in other fields thatmake up their discipline. Polanyi calls this phenomenon of mutual reliance “overlappingcontrol”.6’This sort of dependence is not only restricted to the relations among experts, but existsin more pronounced forms among those, like schoolteachers, who, more often than not, areinexpert in the forms of knowledge they are called upon to transmit. In Allen’s view Peters ismistaken to say the directive authority of teachers derives primarily from their cognitive authority,since, as a group, teachers are not cognitive authorities in anything but a restricted sense, relative98to the ignorance of their students. They are only de facto cognitive authorities, not legitimatecognitive authorities.62 Teachers are legitimate directive authorities nonetheless. Hirst and Petersstate that the authority of cognitive authorities within educational institutions “derives from theirspecial training and mastery of the relevant sphere of knowledge, on their success in getting thingsright in spheres where what is right or true does not depend on the pronouncements of anyindividual, but on reasons and evidence that anyone can, in principle, grasp.”63 In Allen’s opinionthis view is obviously wrong with respect to teachers, though less obviously wrong in the case ofprofessors (for reasons given above). In any event the matter is an empirical one. It is ironic thatPeters, who is well known for chiding educationists for making unsubstantiated empirical claims,would himself be guilty of this same mistake in connection with the expertise of teachers. Ifteachers do indeed know the reasons and evidence which justify knowledge claims it is puzzlingthat they routinely fail to communicate these reasons to their students. The more plausibleexplanation for their neglect of epistemic justifications is that, like most people, they simply don’tknow them.64 As far as Allen is concerned whatever cognitive authority teachers do enjoy derivesfrom their reliance on authorities whose knowledge, while more complete, is still dependent uponthe testimony of others. Allen also disputes Hirst and Peters’ claim that cognitive authority restsupon reasons and evidence that any rational man would assent to. Rather it is the case thatsatisfactory explanations in mathematics and physics increasingly adduce evidence and groundsthat non-specialists, teachers included, would not be able to understand. Further Allen reminds usof the element of ‘tacit knowledge’ in competent performance that is not readily communicable, atleast not in terms of evidence, reasons, etc. As Polanyi has illustrated, the master, in many cases,will be unable to provide detailed reasons for acting one way rather than another. The apprenticecan do no more than “surrender himself. . . uncritically to the imitation of another.”65 Until ratherlate in their training apprentices must accept the judgments and pronouncements of their teachersthough they do not know the evidence that would justify them.66So, on Allen’s view, Hirst and Peters have overlooked obvious facts about the world ofknowledge when they claim that teachers ought to teach in such a way that their students will oneday become epistemically independent. Such independence is unattainable, and would be99undesirable in any event. Thus reliance on authorities is unavoidable; first because it is rational tobe epistemically dependent, second because schools and universities wisely acknowledge this factin their practices. On Allen’s view, then, one of the primary benefits of an education is thateducation inducts students into the practices of deference to cognitive authorities. Educationsimply is an induction into the acceptance of epistemic authority.From Allen’s perspective Hirst and Peters fail to acknowledge the full range ofcircumstances in which it is rationally acceptable to hold one’s beliefs on the basis of authority.Hirst and Peters would likely agree with the most basic and uncontroversial version of theprinciple of testimony:If A has good reason to believe that B has good reasons to believe p. then A has goodreasons to believe p.But Allen and others are claiming more than A has good reasons to believe B (where thesereasons, presumably, have to do with an assessment of B’s reliability as an authority, and not withan assessment ofp directly). They are claiming that A knows that p on the word of B. A may notpossess the reasons necessary to justif,’ p and A’s belief that p, but nevertheless A knows p. Todefend this counter-intuitive claim John Hardwig67 points out that to deny it would force us tomaintain:1) that there can no longer be knowledge in many scientific disciplines where the complexity ofavailable evidence rules out epistemic independence.2) that one can know p only by ignoring most of the best evidence forp.68Knowledge can be vicarious, says Hardwig, and a community can know something that noone individual knows. In this view Hardwig has considerable company among contemporaryphilosophers who have abandoned internalist conceptions of epistemic justification. For example,the work of D. M. Armstrong69,Goldman70, Stich71, and others72 looks at actual epistemicpractices and concludes that deference to authorities outside one’s limited expertise is rational,perhaps even maximally rational, given the epistemic situation most of us find ourselves in most ofthe time. Epistemic independence, on this view, is simply foolish.Stich and Nisbett forceftully express this viewpoint as follows:100Deference to authority is not merely the habitual practice of educated people, it is,generally, the right thing to do, from a normative viewpoint. The man who persists inbelieving that his theorem is valid, despite the dissent of leading mathematicians, is a fool.The man who acts on his belief that a treatment, disparaged by medical experts, will curehis child’s leukemia, is worse than a fool.73Stich and Nisbett further claim, as Allen does, that “one of the principal effects ofeducation is to socialize people to defer to cognitive authorities”.74Like Allen they see this stateof affairs as salutary.The view taken by Hardwig and others argues that in science epistemic trust is thetouchstone of rational belief “Modern knowers cannot be independent and self-reliant, not evenin their own fields of specialization. In most disciplines, those who do not trust cannot know;those who do not trust cannot have the best evidence for their beliefs.”75 [emphasis added] Here,Hardwig is not simply making the uncontroversial claim that belief on testimony is rational undercertain restricted conditions. He is suggesting that the standard view—that knowledge rests onevidence—is simply mistaken. Knowledge rests on a trust of epistemic authorities. It is not thatappeals to authority are only sometimes justified on grounds of expedience, rather it is that theideal of epistemic individualism that pervades epistemology is a romantic illusion that fails tocapture the character of successful inquiry. Epistemic dependence is the rule, not the exception.Hardwig’s view has been seconded by Webb who claims that in restricting his analysis largely toscience Hardwig fails to take his argument far enough.Trust is necessary if one wishes to have knowledge of anything interesting beyond one’sown immediate experience. . . .It is not only the progress of science that depends on thiskind of interdependence, this division of labour. It is only because we divide up theepistemic work in this way that we can come to know anything from maps, clocks,thermometers, newspapers, telephone directories, and so on. It is a commonplace that wecannot know very much about history without relying on the testimony of those who werecloser to the events than we are; what is overlooked is that we also cannot know a greatdeal about the here and now without similar reliance on our contemporaries. How would Ifind out about Pluto, Antarctica, quarks and differential equations (not to mention suchtrivia as how to get to the grocery store, or when to expect my paycheck), if not fromother people, whom I trust in these matters?76A similar position, but of more consequence for teachers, is the Wittgensteinian view ofknowledge transmission taken by Welbourne. Welbourne argues that the key to understandingknowledge and how it is transmitted is not to be found in scrutinizing sentences in which ‘know’101and its cognates are used, but rather in attending to our knowledge-seeking and knowledge-communicating practices. If we do this we will find that knowledge is most often transmitted bymere say-so, where what is transmitted is accepted on trust. The important distinction here isbetween the language game of transmitting knowledge and that of arguing for a view. In theformer you accept my pronouncements because I have made them, and because you lack anyreason to doubt my truthfulness. In the latter I am successful when I get you to see things my waybecause of cogent arguments and compelling claims. It is sufficient for the transmission of myknowledge to you that you believe me. By contrast, your simply taking my word for it is notsufficient if I am arguing for a view.Believing someone is par excellence the appropriate response when, and only when, the‘game’ is the transmission of knowledge. That is why one can be confident that it issufficient for the transmission of your knowledge to me that I believe you, and why mybelieving you when you have asserted that p even falsely warrants my saying ‘I know’. Forme to believe you is for me to suppose that I have learnt (come to know) that p from you.In this game believing the speaker is the ‘uptake’ condition. In a perfectly simple case,where all goes smoothly, you speaking from knowledge assert that p; I believe you andthus begin to know that p myself77In the game of knowledge transmission the skepticism that demands proof and argument isout of place. If there is some reason to be skeptical, the game changes from one of transmittingknowledge to arguing a point. Such transformations are relatively rare, however, since, as anyanalysis of these practices will reveal, relations of trust are implicit in the circumstances underwhich questions are asked. To express doubt or demand justification under such circumstanceswould strike one’s interlocutor as odd or mildly insulting.78Welbourne’ s discussion raises the possibility that Scheffier, Green et al. are led to theirview by virtue of a category mistake. They have confused one language game for another. Inorder for students to acquire genuine knowledge it is not necessary that they possess the relevantevidentiary arguments, only that they believe their teachers. Welbourne concludes that “our abilityto engage in the game of transmitting knowledge, whether as transmitter or receiver, requires nofurther intellectual capacity beyond what is implied in having a reasonable command of language,sufficient to understand the import of questioning and answering. It involves no special ability toweigh evidence justly or spot inconsistencies”79102Welbourne, predictably, disagrees with Locke’s position that belief on testimony is nottrue knowledge, but mere belief taken on trust. According to Welbourne, Locke does notunderstand that to believe someone means to regard that person as a source of knowledge. Lockemistakenly believes that knowledge is distinguished from belief by the quality of the reasoning thatleads to it, such that a speaker can know on the basis of her reasons, while a hearer can onlybelieve. In Welbourne’s view Locke’s mistake is one of misrepresenting the grounds of belief ontestimony. He supposes (wrongly) that when we believe on testimony we use the testimony asevidence or warrant for our belief80 He then concludes (rightly) that testimony is poor evidence.The problem, for Welboume, with this line of reasoning is that “to receive testimony as evidenceis precisely not to receive it as testimony.”8’Welbourne’ s view, you will recall, is that belief ontestimony is based on a trust of the speaker, not on an appraisal of evidence. Belief of the speakeris the uptake a speaker hopes for when transmitting knowledge, while truthfhlness and epistemicresponsibility are the features of a speaker expected by those who receive. The assumptions thatunite both are assumptions of trust.[Locke] credits the hearer with the kind of uptake which, other things being unequal,would effect a transmission of knowledge. If hearing you assert that p. I take it on trustfrom you that p, I do not treat your assertion as evidence. I take it that you were speakingfrom knowledge and that I now know that p myself through your say-so... .But Lockethinks that the hearer may end up with (mere) belief, and will if he lacks factualunderstanding. This is not a possible outcome. All the conditions for transmission aresatisfied and the knowledge is not denatured.82If Locke is to be interpreted literally there can never be any transmission of knowledgebetween one person and the next. The pursuit of knowledge must be, perforce, a solitaryendeavor. Yet, clearly it is not. Welbourne argues that the pursuit of knowledge is typicallycarried out within a community, a complex web of epistemic dependence relations that makes thetransmission of knowledge possible. On this point there is considerable agreement among socialepistemologists, who as a group emphasize the benefits of a cognitive division of labour, markedout by expertise.EXTERNALIST TifEORIES OF JUSTIFICATIONAs we have seen, a venerable tradition in philosophy maintains that rational agents hold103their beliefs on the basis of reasons and evidence, the epistemic force of which they understandand appreciate. Typically what is to count as evidence or a good reason does not include appealsto authority.83 When warrant is seen to amount to nothing more than arguments from authority,the received view has been that the justification is at best weak, at worst, no justification at all.This is because, on most accounts, what justifies an agent’s belief must be available to the agent insuch a way that the agent’s autonomy is preserved. The stress on evidence or reasons available tothe agent is what identifies this position as internalist.84 Over the last few years, however,internalism has been losing adherents due to the trenchant criticism of those who favourexternalist theories of knowledge and justification. Reduced to its essentials the externalistposition states that a cognitive agent can know that p and be justified in believing that p withoutknowing what justifies the belief that p. This thesis allows that one who has no knowledge of theground for her belief that p still knows that p nonetheless, provided the proper relation existsbetween the belief and the world. We can characterize two sorts of relations with respect toepistemic justification: a causal relation between, let us say, Susan and her environment, and apsychological relation of understanding and appreciating between Susan, her belief and itswarrant. The former relation is epistemically external; for it to contribute to Susan’s justification,it is not necessary that she think it does. This is so because there can be something true of theperson (that she has grounds) without that person knowing those grounds. Oddly enough, writersin educational philosophy, even critics of internalist intuitions, do not seem to have taken note ofexternalist theories of justification and knowledge. This is both surprising and unfortunate sinceexternalism, as a doctrine, has the potential to ground arguments that would strip educationists ofmuch of the justification they claim for wanting to revise teaching practice in light of the demandsof reason.INTERNALIST AND DEONTIC CONCEPTIONS OF JUSTIFICATIONThere is, as well, a deontological cast to internalism, namely that cognitive agents choosewhat to believe, and are therefore responsible for the choices they make. Insofar as our beliefsresult from acts ofjudgment, ethical predicates apply to matters of beliefs. So it is legitimate, onthis view, to conceive of the cognitive agent as being subject to an ethics of belief Agents, if they104are to be rational, have epistemic obligations that require them to form or revise their opinions onthe basis of reasons and evidence. Steup’s sentiments here stress the link between being rationaland choosing one’s beliefs:No matter how grim the circumstances are, if an agent holds a belief contrary to evidence,it is within his power, given that he is a rational agent, to reflect upon his belief andthereby find out that he had better withhold, or even assent to its negation. Being arational agent, I would say, involves the capacity to find out, with respect to any beliefwhether or not it is being held on good grounds.85In these remarks Steup echoes Clifford’s classic expression of Victorian epistemicdeontologism: “It is wrong everywhere and for anyone to believe anything on insufficientevidence.”86 Clifford’s view, which combines a rigorous evidentialism with an unequivocalcommitment to the ethics of belief exists among today’s internalists in a muted form. Latter daydeontologists prefer to talk in terms of epistemic responsibility and obligation, but the moral tenoris still unmistakably present, as in Bonlour’s remarks:We cannot, in most cases at least, bring it about directly that our beliefs are true, but wecan presumably bring it about directly (though perhaps only in the long run) that they areepistemically justified. It follows that one’s cognitive endeavors are epistemically justifiedonly if and to the extent that they are aimed at this goal, which means very roughly thatone accepts all and only those beliefs which one has good reasons to think are true. Toaccept a belief in the absence of such a reason. . . is to neglect the pursuit of truth; suchacceptance is, one might say, epistemically irresponsible. My contention here is that theidea of avoiding such irresponsibility, of being epistemically responsible in one’sbelievings, is the core of the notion of epistemic justification. 87Internalism in itself does not rule out belief on testimony, providing the agent makes somerational assessment of the authority’s reliability. Internalism requires therefore a discriminatingdeference to epistemic authorities. In this respect it differs from externalism. On the externalistaccount the reliability of testimony derives from the relation between the testimony and states ofaffairs which are testified to. It is not derived in any way from the subject’s knowledge of thatrelation or her assessment of reliability. Thus internalism has utility as a normative view ofepistemic justification that externalism does not.TIlE EXTERNALIST CRITIQUE OF INTERNALIST JIJSTII?ICATIONWe have seen that internalists stress the cognitive agent’s active role in deciding what to105believe. Indeed the distinguishing feature of the rational person is not so much what she believesas the manner in which she acquired her beliefs. Bertrand Russell claimed, in terms nearly identicalto those used by Green subsequently, that “it is not what the man of science believes thatdistinguishes him, but how and why he believes it. His beliefs are tentative, not dogmatic; they arebased on evidence, not on authority or ifltUjtiOfl.”88 (original emphasis) The externalist has nosuch doxastic requirement for a person to be justified. Others, such as authorities, may know whatevidence or reasons justifies a belief; or perhaps no one person knows. Beliefs needn’t be justifiedby the agent to be justifiedfor the agent.Justification is also a matter of choice for most internalists. Cognitive agents are seen tohave some degree of control over what they come to believe. Beliefs are not simply caused. It isnot clear, however, that what one believes is a matter of choice.89 Despite the intimations offreedom and responsibility provided by ordinary language in such expressions as—”You shouldn’tjump to conclusions”, or “I had every right to think she was honest”—many beliefs, uponexamination, do not appear to be the sort of thing about which we have a choice. Alston, forexample, argues that in all but a very few cases we cannot believe at will.9°This is obviously truein the case of perceptual beliefs, but just as true, he says, with ordinary beliefs formed byintrospection, memory, or uncontroversial inferences. If only uncertain beliefs leave room forchoice, and these are comparatively small in number, we cannot, on this view, rightly favour adeontological conception of justification. And since ought implies can there should be nonormative requirement in places where voluntary control is absent. Given that externalism doesnot require that what we believe is a voluntary matter, at least with respect to states of affairs, theethics of belief play little or no part in their account of epistemic justification.Internalism is seen to be flawed at an even more fundamental level, however. Alstonargues that the internalist view is fundamentally incoherent since it attempts to unite the notionthat justification is necessary and nearly sufficient for knowledge with the additional notion thatjustification is a matter of doing one’s duty. “I may have done what could reasonably be expectedof me in the management and cultivation of my doxastic life, and still hold a belief onoutrageously inadequate grounds.”9’What exactly is the position of externalists and their critique106of internalism? BonJour characterizes the externalist position as follows:Though there must in a sense be a reason why a basic belief is likely to be true, the personfor whom such a belief is basic need not have any cognitive grasp of this reason. On thisview, the epistemic justification or reasonableness of a basic belief depends on theobtaining of an appropriate relation, generally causal or nomological in character, betweenthe believer and the world. This relation.. . is such as to make it either nomologicallycertain or else highly probable that the belief is true. It would thus providefor anyone whoknew about it, an undeniably excellent reason for accepting such a belief.... The person forwhom the belief is basic need not (and in general will not) have any cognitive grasp of anykind of this reason or of the relation that is the basis for it in order for this basic belief tobe justified; all these matters may be entirely external to the person’s subjectiveconceptions of the situation.92Armstrong claims that for a belief to be justified there must be a law-like connectionbetween two states of affairs: a’s believing that p on the one hand [Bap] and the state of affairswhich makes p true. In short, “given Bap it must be the case that p.”93 Armstrong calls this thethermometer model of non-inferential knowledge because in the same way that the readings of areliable thermometer reflect the temperature so too do one’s beliefs reflect the states of affairs thatmakes them true. On this view, a person whose beliefs are true is a “reliable cognitiveinstrument.” It is in virtue of this reliability that this person’s beliefs are justified.Epistemology has traditionally had two principal preoccupations. On the one hand therehas been an effort to sketch out a sound theoretical account of knowledge and show howknowledge is possible. On the other hand epistemologists have sought to identify the criteria orrules by which individuals might decide what to believe. Both enterprises have been motivated bya desire to find answers to the questions posed by Socrates: What can I know? How can I beconfident my beliefs are justified? How can I improve on the beliefs which I now hold? Thus thetheoretical enterprise is meant to inform the normative one.The normative enterprise, unlike the theoretical one, has deontological presuppositionsbuilt into it. Ought implies can, in other words. This view—that cognitive agents decide what tobelieve and thus need guidance in belief formation—goes by various names, depending onwhether it is the aspect of choice being emphasized (doxastic voluntarism)94or the obligationsand responsibilities that attend the freedom to choose our beliefs (deontologicaljustfication)95The fixing of belief can, in principle, be free, insofar as an epistemic agent decides whether or not107to believe a given proposition at a given time. Where our choices are rational, they are guided bythe right epistemic principles. Where they are irrational they are epistemically irresponsible andopen to criticism. It is Goldman’s claim that internalists have assumed that the terms which aresought for a theoretically sound analysis of knowledge are the same terms that can serve theregulative function of helping the epistemic agent decide what to believe. Put simply,epistemologists have taken the necessary and sufficient conditions of knowledge to be the sameconditions that agents themselves need to satisi,r consciously and deliberately in order to bejustified. Goldman thinks this is a mistake, brought on by conflating an interest in spelling out thenecessary and sufficient conditions of knowledge with an interest in identilying epistemicprinciples that are useful guides in deciding what to believe. In Conditions ofKnowledge Schefflerdoes precisely this. He moves from an analysis of knowledge as justified true belief to aninternalist conception ofjustification whose regulative function he makes explicit in his defense ofevidentially held belief and in his ruling out authority as a suitable ground for one’s beliefs. Sucha move is mistaken, says Goldman, the result of a mistaken picture of what it is to be justified. Inplace of internalism Goldman proposes a causal account ofjustification which he terms “historicalreliablism”. According to this view, “a belief is justified just in case its causal ancestry consists ofreliable belief-forming processes, i.e., processes that generally lead to truth.”96This account has two features that set it apart from internalist accounts. First, as anexternalist account ofjustification, it makes no reference to the deliberations of agents. The agentmay or may not know the causal ancestry of her belief She may also be unaware of the reliabilityof whatever processes led to the belief Second, as such, it cannot offer any rules or prescriptionsfor choosing beliefs. It considers an already formed belief and identifies which features arenecessary and sufficient for that belief to be justified.Why favour externalism over internalism? The answer given by Goldman is thatinternalism is either fundamentally confused or unattainable. Distinctive of internalism is therequirement that justification be carried out by the agent before accepting a belief Therefore, afundamental concern of internalists is to identify the rule or set of rules by which an agent coulddetermine the justifiedness of a belief Goldman’s strategy is to consider what conditions108determine the right principles. The conditions appropriate for externalism, he says, can beexpressed as follows:1) Rule of justification X is right if and only if: X is actually optimal in avoiding error andproducing true beliefs.This is insufficient for the internalist since it does not require that the agent in some senseknow and utilize X in order to arrive at a justified belief The conditions that appear to Goldmanto be appropriate for internalism are:2) Rule of justification X is right if and only if: we are justified in believing that X is optimal inavoiding error and producing true beliefs.3) Rule ofjustification Xis right if and only if: we believe that Xis optimal.4) Rule ofjustification Xis right if and only if: (A) we believe that X is optimal, and (B) this beliefwas caused by reliable cognitive processes.The fatal difficulty with 2 is that it is circular. Our motivation in seeking a regulativetheory ofjustification is to provide rules that do not presuppose the prior existence of such rules,and 2 fails in this regard. 3 fails because it accomplishes too little. It is possible under 3 that anagent believes X is optimal for foolish reasons. 4 fails because it is no longer internalist, given thatthe causes of our beliefs and the reliability of our cognitive processes are not generally available tous. Internalism requires that they be available. Goldman’s opposition to internalism provides thebasis for his defense of epistemic paternalism. If he is correct the implication for teaching that hastraditionally been derived from internalism must give way. It may be salutary after all thateducation socializes students into deference to epistemic authorities.EPISTEMIC PATERNALISMAmong the evidentialist principles upon which internalists rely the “requirement of totalevidence principle” is especially important. This principle (RTE) states that the agent, to berational, must base her beliefs on the available evidence and not neglect any evidence that bearscrucially on the beliefs in question. In light of this principle, those who have some control overevidence relevant to another’s rational decision ought to accept the following corollary of RTE:(C-RTE).109If agent X is going to make a doxastic decision concerning question Q, and agent Y hascontrol over the evidence that is provided to X then, from a purely epistemic point ofview, Y should make available to X all of the evidence relevant to Q which is (at negligiblecost) within Y’s control.97In terms relevant to teaching children, teachers should, on epistemic grounds, makeavailable all evidence relevant to the particular claims they wish their students to adopt. Youmight say the point of this thesis is to have educators see the wisdom of RTE and its corollary inorder that these two principles might transform the way teachers teach. The RTE principle is seento be epistemically useful because those who follow it are in the best position to acquire truebeliefs and avoid false ones. Its corollary is epistemically useful for much the same reason.Goldman wants C-RTE rejected as a universal principle, and epistemic paternalism acknowledgedas a legitimate stance to take toward the epistemically dependent. He argues these points in termsof the veritistic benefits paternalistic treatment confers on the dependent. Despite volumes ofphilosophical writing that claim otherwise, C-RTE does not, he says, maximize the chances thatthe epistemically dependent will have more true beliefs and fewer false beliefs. To see this oneneed only examine the rules that govern the admissibility of evidence in courtrooms, or theFederal Trade Commission’s regulations with respect to false or deceptive advertising, or theconsiderations that justify excluding certain points of view from school curricula. Evidence thatmay reduce the chances of a juror, consumer, or student arriving at the truth or avoiding errorought to be restricted by the relevant authorities, and this restriction, says Goldman, violates bothRTE and C-RTE.The requirement of total evidence and the doctrine of epistemic individualism are part ofthe Millian legacy that Goldman thinks we would do well to reconsider. He doubts the “collisionof adverse opinion” will directly bring about the “clearer perception and livelier impression of thetruth” as Mill claimed it would. Rather a clearer perception of the truth, if it comes about via theclash between opinions, does so indirectly because more true beliefs and fewer erroneous beliefshave emerged. The fundamental principle, then, is to maximize the availability of true beliefs andminimize the number of false ones. Epistemic paternalism is a sensible and therefore justifiablestrategy for doing just that.110INTERNALISM AND THE PROBLEM OF INFINITE REGRESSHetherington has noted additional problems for internalism.98 Internalists insist thatjustification is based on reasons or evidence that are somehow available to the agent. Moreprecisely the evidence must be held and appreciated as evidence. Not only must the evidence bewithin an agent’s grasp, it must be epistemically within her grasp. I may have a piece of evidencebut not see its significance. Such a piece of evidence is internal, but not, on Hetherington’saccount, epistemically internal. For evidence to be epistemically internal, then, a person must notonly have the evidence, she must appreciate it as her justification. This feature of internalism isalso explicitly invoked by Scheffier in his analogy with detective novels. The fatal difficulty withthis distinction is that it leads to an infinite regress. The problem is this:Suppose Wi (a piece of evidence) is epistemically within a person S: e.g., S not only hasthe belief; but appreciates it as his or her justification. However, since appreciation is a doxasticattitude of a kind with belief; it too must be epistemically within S. For this doxastic attitude toalso be epistemically within S it is necessary that S appreciate her appreciating. This appreciatingin turn must be epistemically internal, and so on.Hetherington states that for Wi to be internal to S there must be a chain lookingsomething like this: From Wi there must follow:the appreciating of WI as epistemically internal to 5, the appreciating of that firstappreciating as epistemically internal to 5, the appreciating of that second appreciating asepistemically internal to S.. 100For each member of the sequence to be epistemically internal the next member of thesequence must also be epistemically internal. Thus we have an infinite regress. So internalismleads to skepticism in two ways. Either it collapses under the weight of an infinite regress or itfails to square with what is commonly taken to be knowledge. If the externalist position is correctit will no longer be possible to argue for the educational importance of evidential styles of beliefsolely on the basis of an analysis of knowledge and rational belief Such is the radical nature ofexternalism that it runs counter to the long-standing normative requirement that cognitive agentsadopt the epistemic point of view, and base their beliefs on evidence they themselves possess and111understand. It therefore lends support to those who think education is best conceived in termsquite unlike those commonly expressed in educational philosophy.‘Iris Murdoch, The Sovereignty of the Good (London: Fontana, 1979).2R. F. Dearden, “Autonomy and Education,” in Education and the Development ofReason, ed. R. F. Dearden, Paul Hirst and R. S. Peters (London: Routledge and Kegan Paul,1972), 461; “Autonomy as an Educational Ideal,” in Philosophers Discuss Education, ed. S. C.Brown (London: Routledge and Kegan Paul, 1975), 18.S. I. Benn, “Freedom, Autonomy and the Concept of a Person,” Proceedings of theAristotelian Society 1975—1976, 130.4Elizabeth Telfer, “Autonomy as an Educational Ideal” in S. C. Brown ed. PhilosophersDiscuss Education (London: Macmillan, 1975), 19-28.5Gerald Dworkin, The Theory and Practice of Autonomy (Cambridge: CambridgeUniversity Press, 1988).6lbid., 12.7Brian Crittenden, “Autonomy as an Aim of Education,” in Ethics and EducationalPolicy, ed. Kenneth Strike and Kieran Egan (London: Routledge and Kegan Paul, 1978), 107.8lbid., 117.9lmmanuel Kant, Groundwork of the Metaphysic ofMorals, trans. by H. J. Paton (NewYork: Harper and Row, 1964), 98.‘°Isaiah Berlin, Four Essays on Liberty (London: Oxford University Press, 1969),152—153.‘1Kurt Baier, “Moral Autonomy as Aim of Moral Education,” in New Essays in thePhilosophy of Education, ed. Glenn Langford and D. J. O’Connor (London: Routledge andKegan Paul, 1973), 102.‘2Dworkin, 21.‘3lbid., 20.‘4lbid., 23.‘5This is not to suggest that those who advocate intellectual autonomy have internalisttheories of justification in mind, or even that they are aware of these theories. It is to suggest,rather, that various forms of evidentialism, some of which date back to Plato, are in the endinternalist in their basic commitments.11216Kenneth A. Strike, Educational Policy and the Just Society (Urbana, IL: University ofIllinois Press, 1982).17Jol’i Stuart Mill, On Liberty (New York: The Liberal Arts Press, 1956), 43.‘8Thomas Green, “A Topology of the Teaching Concept,” in Concepts of Teaching:Philosophical Essays, ed. C. J. B. MacMillan and Thomas W. Nelson (Chicago: Rand McNally,1968).19J L. Austin, “Other Minds,” Philosophical Papers (Oxford: Oxford University Press,1961), 53.20Johi 0. Urmson, “Parenthetical Verbs,” in Essays in Conceptual Analysis, ed. AntonyFlew (London: MacMillan and Company, 1956), 198—199.21Israel Scheffler, Conditions ofKnowledge; The Language ofEducation (Springfield, IL:Charles Thomas Publishers, 1960); Paul H. Hirst, Knowledge and the Curriculum (London:Routledge and Kegan Paul, 1974)22For defense of the view that knowledge entails justified true belief see Marshall Swain,Reasons and Knowledge (Ithaca: Cornell University Press, 1981); Keith Lehrer, Theory ofKnowledge (London: Routledge, 1990); Lawrence BonJour, The Structure of EmpiricalKnowledge (Cambridge, Massachusetts: Harvard University Press, 1985).23A. J. Ayer, The Problem ofKnowledge (Harmondsworth: Pelican Books, 1956), 35.24Roderick Chishoim, Perceiving (Ithaca, NY: Cornell University Press, 1957), 16.25For the most part epistemologists have continued to favour the tripartite definition,though they no longer think the three conditions are sufficient, only nearly sufficient.26Roderick Chishoim, Theory ofKnowledge (New York: Prentice Hall, 1989), 90.27Roderick Firth, “Are Epistemic Concepts Reducible to Ethical Concepts?,” in Valuesand Morals, ed. Alvin Goldman and Jaegwon Kim (Dordrecht: D. Reidel Publishing Co., 1978),219.28Scheffler makes an important qualification regarding the internal character ofjustification. In schools the criteria are to be more stringent than in everyday life since the purposein schools is an educational one. In everyday life where typically expediency is reasonablyafforded more weight appeals to authority are justified. In schools they are less justified. For hisdiscussion see Conditions ofKnowledge, p. 67—68.29Ibid., 56.11330This aspect of the world of knowledge is central to the conception of liberal educationproffered by Paul Hirst, as well as to John McPeck’s conception of education for critical thinking.Both argue that a knowledge of epistemic justification across the forms of knowledge is aconstitutive feature of rational minds.31Israel Scheffler, “Responses,” in Synthese, ed. Catherine Z. Elgin, 94, no. 1 (1993): 130.32Israel Scheffler, Reason and Teaching (Indianapolis: Hackett, 1973), 143.33Green, “A Topology of the Teaching Concept,” 42.34Ibid., 43.35Paul Hirst, “Liberal Education and the Nature of Knowledge,” Knowledge and theCurriculum: A Collection of Philosophical Papers (London: Routledge and Kegan Paul, 1974),40.36This expression is from Wilfred Sellars who claimed that “in characterizing anepisode...as that of knowing...we are placing it in the logical space of reasons, of justifying andbeing able to justify what one says.” See his Science, Perception and Reality (London: Routledgeand Kegan Paul, 1963), 169.37Suzanne deCastell, “Epistemic Authority, Institutional Power, and CurricularKnowledge,” Journal ofEducational Thought. 16, no. 1 (1982): 27.38Harvey Siegel, Educating Reason: Rationality, Critical Thinking and Education (NewYork: Routledge, 1988), 33.39Ibid., 54.40John Stuart Mill, On Liberty, 1641Ibid., 1642thid., 1743Ibid., 1544Ibid., 16.45Eamonn Callan’s paraphrase of Dunlop’s argument (from Francis Dunlop, “On theDemocratic Organization of Schools,” Cambridge Journal ofEducation 9, no. 1 (1979), 43—54)in Eamonn Callan, ed., Autonomy and Schooling (Montreal: McGill-Queens University Press,1988), n. 26, 160—161.“Ibid., 188.11447Ibid., 188.48Ibid., 190.49Scheffler, Conditions ofKnowledge; In Praise of the Cognitive Emotions.50Thomas Green, The Activities of Teaching; “A Topology of the Teaching Concept.”51R. S. Peters, Authority, Responsibility and Education (New York: Atherton Press,1965); Ethics andEducation (London: Allen and Unwin, 1966); “Reason and Habit: The Paradoxof Moral Education,” in Moral Education in a Changing Society, ed. W. Niblett (London: Faberand Faber, 1963).52Paul H. Hirst and R. S. Peters, The Logic ofEducation (London: Routledge and KeganPaul, 1970), 114—123.53R. F. Dearden, The Philosophy of Primary Education (London: Routledge and KeganPaul, 1968), see esp. ch. 2 and 3.54R. T. Allen, “Because I Say So!: Some Limitations Upon the Rationalisation ofAuthority,” Journal of Philosophy of Education 21, no. 1 (1987): 15—24; “I’ll Say It Again: ARejoinder to Jim Mackenzie,” Journal ofPhilosophy ofEducation 22, no. 1 (1988): 113—114.55Francis Dunlop, “On the Democratic Organization of Schools,” Cambridge Journal ofEducation 9, no. 1, 1979).56Allen, “I’ll Say It Again,” 113.57Hirst and Peters, 116.58Paul H. Hirst and R. S. Peters, The Logic ofEducation (London: Routledge and KeganPaul, 1970), 117.59R. S. Peters, Ethics and Education (London: Allen and Unwin, 1966), 265.60Allen, “Because I Say So!, 17.61Michael Polanyi, Knowing and Being (London: Routledge and Kegan Paul, 1969), 55ft’Science, Faith and Society (Chicago: University of Chicago Press, 1964), 48ff62J owe this distinction to Richard T. DeGeorge, The Nature and Limits of Authority(Lawrence, KN: University of Kansas Press, 1985).63Hirst and Peters, 116.11564I is highly improbable that elementary school teachers would have such depth ofunderstanding since they are called upon to teach a wide range of courses (mathematics, science,history, etc.) where the demand for epistemic justification is appropriate. Little in the formalpreparation of these teachers would prepare them to communicate the relevant evidentiaryarguments.65Michael Polanyi, Personal Knowledge (London: Routledge and Kegan Paul, 1968), 53.Alasdair Maclntyre makes an identical point in After Virtue.67John Hardwig, “Epistemic Dependence,” The Journal of Philosophy LXXXII, no. 7(1985): 335—349.68See Ibid., especially pp. 346—349. Hardwig cites the research on charm particles inphysics carried out by a team of 99 researchers as an example of what he calls “vicariousknowledge,”. Vicarious knowledge is knowledge that can be possessed without direct evidenceand with only partial or no understanding of its grounds. It can be expressed as follows:A knows that m.B knows that n.C knows (1) that A knows that m, and (2) that if m, then o.D knows (1) that B knows that n (2) that C knows that o, and (3) that if n and o, thenp.E knows that D knows that p.thereforeE knows that p.69D. M. Armstrong, Belief Truth, Knowledge (London: Cambridge University Press,1973).70Alvin I. Goldman, “What is Justified Belief?,” in Just/Ication and Knowledge, ed.George S. Pappas (Dordrecht, Holland: D. Reidel Publishing Co., 1979), 1—23; “The InternalistConception of Justification,” in Midwest Studies in Philosophy: Studies in Epistemology, VolumeV, ed. Peter A. French, Theodore E. Uehling, Jr. and Howard K. Wettstein (Minneapolis:University of Minnesota Press, 1980), 27—52; “Epistemic Paternalism,” The Journal ofPhilosophy; 88, no. 3, (1991): 111-131.71Stephen P. Stich and Richard E. Nisbett, “Justification and the Psychology of HumanReasoning,” Philosophy ofScience 47 (1980): 188—202.72Frederick F. Schmitt, “Justification, Sociality and Autonomy,” Synthese 73 (1987):43—85; W. H. Walsh, “Knowledge in its Social Setting,” Mind 80, no. 319 (1971): 321—336;Michael Welbourne, “The Community of Knowledge,” Philosophical Quarterly 29 (1981): 302-314; “The Transmission of Knowledge,” Philosophical Quarterly (1979): 29, no. 114, 1—9;Hilary Kornblith, “Some Social Features of Cognition,” Synthese 73 (1987): 27—41.73Stich and Nisbett, 199.11674Ibid., 199.“John Hardwig, “The Role of Trust in Knowledge,” The Journal of Philosophy. 88, no.12 (1991): 693-694.76Mk Owen Webb, “Why I Know About As Much As You: A Reply to Hardwig,” TheJournal ofPhilosophy 7, no. 3 (1993): 260—261.77Michael Welbourne, “The Transmission of Knowledge,” Philosophical Quarterly. 29,no. 114 (1979): 5.78By way of contrast, it is worth noting that where demands for argument and justificationare called for, uncritical acceptance is out of place. For example, Welbourne observes thatneophyte philosophy students tend to view their lectures as occasions for the dissemination ofknowledge. To his dismay they accept what he says without question. Under such circumstances,says Welbourne, “to believe me would be to misunderstand me in a very important way, it wouldbe to misunderstand the game I was playing. Believing me is not the right response in this game.”The proper response would be one of assessing the merits of the argument, raising objections,posing questions and counter-examples, or engaging in other, truth testing maneuvers. SeeWelbourne, 9.79Welbourne, “The Transmission of Knowledge,” 9.80Ii this Locke is not alone. It is commonly asserted that testimony is useful primarily asevidence for one’s beliefs. This remark of Price’s, for example, is typical: “Each of us would liketo know what happened before he was born.... His own first hand observations.. .will not enablehim to answer [this question]. If he cannot know.. .he would still like to be able to hold the mostreasonable beliefs that he can, on the best evidence he can get. And very often indeed the onlyevidence he can get is the evidence of testimony.” See H. H. Price, Belief (New York: HumanitiesPress, 1969), 125.81Welbourne, “The Community ofKnowledge,” 312.82thid., 312.83Walton, for example, states that “generally speaking we only appeal to experts, if in fact,it may be too expensive or otherwise difficult for us to have direct evidence. That is why we maylegitimately appeal to experts as a secondary source of subjective knowledge when we have tomake a decision.” It is not exactly clear what Walton means by subjective knowledge but it isreasonable to suppose that he means to draw a distinction between objective and subjectiveknowledge where the former is justified on purely epistemic grounds, and the latter on thepragmatic grounds of expedience. See D. Walton, Informal Fallacies: Toward a Theory ofArgument Criticisms (Amsterdam: 3. Benjamins Publishing Company, 1987). For similar viewssee Trudy Govier, A Practical Study ofArgument (California: Wadsworth, 1988); Ralph Johnsonand T. Blair, Logical Se(f-Defense (Toronto: McGraw-Hill Ryerson, 1983).11784lnternalist and deontological conceptions of justification are not identical though theyboth emphasize the agent’s active role in weighing reasons and therefore stand together incontrast to externalism.85Mathias Steup, “The Deontic Conception of Epistemic Justification,” PhilosophicalStudies, 53, no. 1 (1988): 78.86W. K. Clifford, “The Ethics of Belief;” Lectures and Essay Volume II (London:MacMillan, 1879, 186.8Thawrence BonJour, The Structure of Empirical Knowledge (Cambridge, MA: HarvardUniversity Press, 1985), 8.88Bertrand Russell, A History of Western Philosophy (New York: Simon and Schuster,1945), 372.See Bernard Williams, “Deciding to Believe,” in Problems of the S4f (Cambridge:Cambridge University Press, 1972); William P. Alston, “The Deontological Conception ofEpistemic Justification,” in his Epistemic Just/Ication: Essays in the Theory of Knowledge(Ithaca, NY: Cornell University Press, 1989), 115—152.90Alston, “The Deontological Conception of Epistemic Justification”.91William Alston, “Concepts of Epistemic Justification,” Epistemic Justfication: Essays inthe Theory ofKnowledge (Ithaca, NY: Cornell University Press, 1989), 67.92Lawrence BonJour, “Externalist Theories of Empirical Knowledge,” Midwest Studies inPhilosophy, 5, 1980), 56.93D. M. Armstrong, Belief Truth and Knowledge (London: Cambridge University Press,1973), 166.94Alvin I. Goldman, “Epistemic Paternalism,” The Journal of Philosophy. 88, no. 3(1991): 113—131.95Alston, “The Deontological Conception of Epistemic Justification”.96Alvin I. Goldman, “The Internalist Conception of Justification,” in Midwest Studies inPhilosophy: Studies in Epistemology, Volume V, ed. Peter A. French, Theodore E. Uehling, Jr.and Howard K. Wettstein (Minneapolis: University of Minnesota Press, 1980), 28.97Alvin I. Goldman, “Epistemic Paternalism,” 114.98Stephen Cade Hetherington, “On Being Epistemically Internal,” Philosophy andPhenomenological Research 51, no. 4 (1991): 855—871.11899The first sense of internal given here is the sense in which externalists view evidence asinternal, and this is all they require.100Hetherington, “On Being Epistemically Internal,” 860.119CHAPTER FOURTILE CASE AGAINST EXTERNALIST THEORIES OF KNOWLEDGE ANDJUSTIFICATIONWe have to this point considered a variety of positions regarding epistemic justification.The most familiar present a particular picture ofjustification, where the solitary agent deliberatesabout what to believe on the basis of the available evidence and then, on this basis, determineswhether to accept a proposition, reject it, or suspend belief It was demonstrated in the lastchapter how this picture has been sharply challenged by those who doubt justification is typically amatter of individual deliberation, or deny that agents have much choice in the beliefs they come toadopt. The most decisive objection has come from those who insist the traditional picture ofjustification sets an unrealistically high standard. For some of these critics being justified does notdepend on epistemically responsible deliberation about what to believe. There may be occasionswhere deliberation is called for, but since there are many where it is not, the insistence that aninternalist condition is necessary for knowledge and justification is too strong. For other critics,being justified is more a matter of believing what you are told than it is a matter of weighingevidence and assessing the reliability of claims. If this picture is correct then it is not on the basisof a need to be justified that one can demand, as I have, that teachers foster evidential styles ofbelief in school children.Externalism and its supplementary arguments for epistemic paternalism thus pose aformidable challenge to my thesis. In essence, externalism denies that knowledge and justificationare best understood by reference to the cognitive agent’s deliberations about what to believe. Itdenies that an appraisal of evidence is a necessary feature of knowing, or even morefundamentally that in the majority of cases we can choose what to believe. The argument ofexternalists relies heavily on a critique of internalism, which by their lights collapses under theweight of an infinite regress, or sets such stringent conditions on knowing that skepticism is theresult. The regress problem bears on the internalist claim that in order to know one mustappreciate the force of one’s evidence, for there will always be some further act of appreciatingrequired of the cognitive agent before she can be said to know. Perceptual beliefs pose a more120serious problem for a deontic conception of justification since it is obviously not the case thatcognitive agents follow an evidential argument, weigh the evidence or assess the adequacy andrelevance of reasons before they determine it is indeed a hand they see before them.’The externalists’ reliance on the twin epistemic goals of acquiring true beliefs and avoidingfalse ones helps to sustain counter-intuitive conclusions regarding epistemic paternalism andappeals to authority. If we are to allow that one can know without possessing and evaluating thegrounds for belief then the sheer efficiency of trusting epistemic authorities is hard to deny.Moreover if we do deny the rationality of believing what we are told what shall be made of theubiquity of belief on testimony in courts, schools, in life generally. Internalists appear to haveignored Wittgenstein’s injunction to think less and observe more.2 They have, on this account,been held captive by a faulty picture of knowledge and justification that has blinded them to whathuman beings actually do in connection with beliefWhat are we to make of the externalist challenge? Some of the issue between externalistsand internalists can be set aside since not every aspect of the debate bears on the thesis underdiscussion. Even if we assume that externalism is the best account of justified belief we can stillmake a case for partial autonomy in terms of the development of expertise. This is a move thatought to be attractive to educators since it is consistent with much current thinking concerning theinstrumental purpose of education. In the course of the last quarter century a human capital viewof education’s purpose has become firmly entrenched in policy. On this view, education ought toprepare students for their future occupational roles within a modern diversified economycharacterized by an division of skilled labour. The development of expertise is therefore one of thelegitimate goals of education.The limitation of this view is that it casts a rather narrow net, since it fails to take intoaccount the variety of other educational goals that might require a degree of intellectualautonomy. Nonetheless the educational implications of the view that epistemic virtue is necessaryto expertise are worth pursuing, for even in cases where agents possess reliable testimony theyfrequently fail to utilize it sensibly. There are as well other valuable attainments, such as the abilityto thinlç that have a point beyond satisi4ng the requirements of specific occupations. Kornblith,121for example, has argued that intellectual autonomy is an attribute of the epistemically responsibleagent, so we might explore the ways in which epistemic virtue guards against error.3There are numerous examples of widespread epistemic irresponsibility upon which tomake a case for epistemic virtue.4 Let us consider three common sorts of error that have beenidentffied in the empirical study of human inference by Nisbett and Ross.5 The first is due to theinfluence of “vividness”. Information is vivid, and therefore likely to attract and hold ourattention, if it is especially interesting, concrete or proximate in either a sensory, temporal orspatial way. Numerous empirical studies have revealed that for the majority of people vivid datahave more impact and influence on belief formation, independent of their objective relevance orreliability, than data which are not vivid.Nisbett et al. offer this illustration of the undue influence of vividness. Suppose you intendto buy a new car, but have not yet decided between a Volvo and a Saab. You consult an issue ofConsumer Reports from which you learn the consensus of expert opinion is that the Volvo ismechanically superior to the Saab. From a reader’s poll of Volvo owners you learn that the Volvohas a superior repair record. On the basis of this information you resolve to buy a Volvo. Butbefore you can make your purchase you meet an acquaintance who upon hearing of your intentionreacts with disbelief and alarm. He relates that his brother-in-law bought a Volvo and had nothingbut trouble with it. The electronic fuel injection failed, then the brakes. He had trouble with therear end, then the transmission went. In frustration he sold the car for junk after three years. Thetale of woe was so dramatic, and came in familiar surroundings from someone less distant than theeditors of Consumer Reports, so you change your mind. According to Nisbett Ct al. most peoplewould attach such disproportionate weight to this sort of testimony from an acquaintance:The logical status of this information is that the N of several hundred Volvo-owningConsumer Reports readers has been increased by one, and the mean frequency of repairrecord shifted by an iota on three or four dimensions. However anyone who maintains thathe would reduce the encounter to such a net informational effect is either disingenuous orlacking in the most elemental self-knowledge.6A second tendency that interferes with the acquisition of true belief is the propensity ofagents to ignore base rate information. In one study7 subjects claimed that the personalitycharacteristics of individuals are poor predictors of the academic field graduate students will122ultimately choose. These same subjects were then asked to predict the field of study for anindividual on the basis of various sorts of information. Included in this information was base rateinformation on the percentage of graduate student selecting various fields, as well as informationon the personality characteristics of the individuals in question. Despite having claimed thatinformation on personality was a poor predictor for such purposes, subjects based theirpredictions entirely on personality characteristics and ignored the base rate information on thepopularity of selected fields.The third example concerns people’s intuitions about random sampling. It is widelyappreciated that the larger the sample drawn from a population the more likely that sample is toreflect the characteristics of the population as a whole. Yet people who would readily assent tothis principle of inductive inference draw their own inferences from remarkably small samples.The tendency to over generalize on the basis of limited data is ubiquitous.What unites these examples is a particular type of irrationality. In each case people shouldknow better than to make the mistakes they do. Their errors are not errors of ignorance, butfailures of critical reflectiveness that an education aimed at fostering epistemic virtue would domuch to prevent.Epistemic virtue is also necessary to offset the influence of non-epistemic reasons for theacceptance of belief. While there are doubtless many non-epistemic, though justified, reasons foraccepting or rejecting the testimony of others (e.g., simplicity, fertility, problem solving capacity)some influential sorts of reasons are clearly without much relevance. In particular there isconsiderable evidence to indicate that receivers of information trust sources of information to theextent these sources look and sound like them.8 This phenomenon is commonplace in secondaryschools where young, inexperienced teachers often have more in common with their students thanthey do with their older, experienced colleagues. This common ground between novice teacherand adolescent frequently simplifies the young teacher’s tasks of persuasion, with the consequencethat young teachers are both more popular and successful, despite the inevitable blunders thataccompany inexperience. Given the prevalence of testimony as the basis for belief in and out ofschools this result should not surprise us. When people base their beliefs on testimony the123characteristics of the source must play a larger role than the epistemic merits of an evidentialargument. Thus it should be no surprise that epistemically irrelevant features of sources areinfluential in the formation of belief Nor should it surprise us that student discussions withinclassrooms frequently degenerate into ad hominem attacks between participants. When the truthof arguments is so frequently to be determined in light of the characteristics of persons, thetemptation must be great to view the perceived fault of positions as being the fault of the person.The prevalence of epistemic dependence in schools makes epistemic virtue all the more necessary.We can also turn Wittgensteinian arguments against externalists. If we accept that themeaning of a term lies in its use, several of the problems for my thesis posed by externalismappear less menacing. For my purposes knowledge and justification are best understood incontexts relevant to the activities of teaching and learning. Thus we can jettison perceptualknowledge from my account. In the main much of what is to be transmitted in schools byinstruction and study is not perceptual knowledge, though clearly students gain what they come toknow through their senses. In terms of teaching acts the teacher begins to offer explanations justat that point where there is a need to argue for a view. Following Grice,9 Fuller provides a rangeof thresholds that differentiate between the degrees of epistemic support required for various sortsof claims. Fuller terms these “thresholds of decidability” for a claim. These thresholds representthe amount of information a speaker must provide to her audience in order to enable that audienceto make a decision. In effect these thresholds determine how explicit a speaker must be in layingout the grounds for belief1) The mere assertion of a claim is sufficient for acceptance or rejection. Quintessentially analyticclaims are the sort where no more than this first threshold must be passed. Other sorts includeself-evident truths and synthetic truisms.2) The assertion must be supplemented with further argument or explication, though the claimcould be easily inferred from other claims already held to be true by the audience.3) The assertion must be accompanied by a “loose” statement of evidence. That is, the claim mustbe supported with evidence, though this evidence may not satisfS’ the most rigorous sort ofstandard (i.e., circumstantial evidence, or evidence found in analogous cases).1244) The assertion must be supported with a more rigorous statement of evidence that satisfieshigher standards. This might include perceptual evidence as well as more technically logicaltruth preserving steps in argument.5) The assertion must be supported as in #4, but in addition must be shown to be counter-intuitivewithout that support. That is, without such evidential support the audience would think theclaim had the opposite truth value from the one demonstrated.Here the thresholds of decidability are determined largely but not exclusively by the natureof the claim. Whether a claim is self-evident or easily inferred from other beliefs is a function ofthe experience of the audience, and so in the case of children there may be a higher threshold ofdecidability for claims that among adults would need less support. It should be plain that many ofthe knowledge claims advanced in schools can only be understood (in Austin’s sense of the term)when teachers go beyond the lower thresholds of decidability and provide more generousepistemic support. The discussion in chapter 1 makes clear that a great many of claims presentedto students are presented as if there is no reason to rise above the lowest thresholds. Yet it isequally clear that many claims require more justification than this. In addition if we take one of thegoals of education to be introducing children to the range of justificatory practices, there shouldbe some effort to familiarize students with these various justificatory moves, even if for ordinarypurposes, there is little need to be more generous in supplying epistemic support.The response of internalists has been to insist that externalism fails to square withcommonsense intuitions regarding knowledge, or that it sidesteps altogether the importantphilosophical questions that have driven epistemological inquiry throughout its history. I will turnto a more detailed examination of their views in a moment. But first I wish to make clear that myintention is not to refute externalism. It isn’t necessary to resolve this dispute amongepistemologists in order to determine whether or not teaching the justificatory arguments thatsupport knowledge claims will help foster the capacity for independent rational thought amongchildren. If internalists are correct in thinking one cannot know without making some subjectiveappraisal of evidence then I can argue my case as others have in the past. That is, I can argue thatif teachers are to transmit knowledge they must transmit as well the evidentiary argument that125provides the warrant for belief 1f however, externalists are correct in supposing justification isexternal to the agent I needn’t abandon my position, so much as modify it. Even among thosewho think that externalism is substantially correct there are several who say the internalistcondition is epistemically worthwhile, though not necessary for knowledge or justification. 10 Thisconcession may provide me sufficient justification for identifying evidentially held belief as aneducational goal of fundamental importance. Kornblith and Schmitt, for example, do not believethat internalism offers an adequate account of knowledge, though they both agree that internalistnotions of justifications pick out worthwhile cognitive traits and virtues. Schmitt’s argument tothis effect features the idea of epistemic virtue. Kornblith’s suggestion is that the disposition toseek reasons for belief is a constitutive feature of “epistemically responsible agency”. Let us firstconsider the internalist reply to the externalists before turning to the arguments that suggest thevalue of the internalist condition is best explicated in terms of epistemic virtue.THE INTERNALIST RESPONSEThe most basic objection against externalism it that it has redefined fundamental epistemicterms in a way that makes them irrelevant to the inquiries concerning the nature of knowledge andjustification. Externalism, on this view, does not only fail to square with our intuitions aboutjustification; it bypasses them altogether by substituting non-epistemic concepts for the mostfundamental epistemic concepts.’1 This strategy, it is said, cannot succeed since the most basicepistemological concepts are sui generis, and therefore defy further analysis or reduction. Someepistemic notions—such as the idea of one proposition being more reasonable to believe thananother—are simply primitive. As well, for every nomological analysis of epistemologicalconcepts offered by externalists, internalists can demand that externalists give us an account ofhow they know externalism is the correct view. As Fumerton remarks externalists will not be ableto answer without utilizing an internalist framework.It is the nomological analyses of epistemic concepts that leads us to keep moving up alevel to ask the externalist how he knows that he knows, or knows that he knows that heknows. The externalist might be able to give correct answers within the framework of hisview, but we, as internalists, will keep asking the questions until his answer invokes a126concept of knowledge or justified belief not captured in terms of nomologicalconnection.12In a similar vein, Chishoim has charged that externalist explications of justification areempty (since they reduce justified belief to true belief) or they are viable only when sosupplemented by internalist concepts they cease to be externalist. Externalism is empty because ittells us nothing that would advance our understanding of knowledge and justification. To see howthis is so consider the central premise of externalism—that S is justified in believing p on thecondition that p is true and S is a thinking subject. Chishoim calls this premise the essential “non-theory” of externalism. If it fails to explain knowledge and justification (as it does) so too mustthe existing varieties of externalism that take it as their main premise. Since this “non-theory”makes no distinction between the true beliefs of an agent and those beliefs he would be justified inhaving its contribution to a theory of knowledge is negligible. If the more finely articulatedversions of externalism also reduce justification to truth they too are non-theories.Take Goldman’s reliability theory of justification, for example. It suggests that one isjustified if one’s belief was formed by a reliable belief forming process.’3 If “reliable process” istaken to mean activities that result in one’s acquiring true belief reliability theory is a non-theoryin Chisholm’s sense. On the assumption that some processes are more reliable than others, indeedthat some of them are unreliable, there remains the inescapable question of which processes aresuch that an agent would be justified in believing them to be reliable. If we restrict our account ofjustification to only those processes that yield true beliefs (as externalists do) we have anuninformative non-theory. If we enlarge our account of justification to include an agent’sdecisions about which processes are reliable we have introduced internalist notions that suggestagents make a subjective appraisal of reliability based on evidence.Since externalists have obliterated the distinction between knowledge and mere true beliefexternalists must allow then that any true belief counts as knowledge. They thereby turn everylucky guess into knowledge.14 The conviction that such a view is wrong is precisely what givesthe Gettier counterexamples their force.’5 These counterexamples (in which some form of luckplays a part in the cognitive agent’s holding a true belief) spoke to the deep seated convictionamong epistemologists that epistemic luck is incompatible with knowledge. That the idea of127epistemic justification has a subjective component can be plainly seen when it is compared tomoral justification. There is a clear difference between genuinely moral action and action with afortuitously moral result. The former is morally justified while the latter is not. Moral luck doesnot make an action morally justified. The person who sets out to murder another person butbungles the job in a way that saves that person’s life from some other source of harm is hardlyacting morally. Vicious and immoral intentions that have been thwarted by circumstances are notthereby redeemed by luck and justified by chance. By analogy epistemic luck does not make anirrational and irresponsible agent epistemically justified.The internalist argument is an argument for evidentially held belief: without beliefs beingheld on this basis there can be no genuine knowledge. BonJour, for example, asserts that forbeliefs to count as knowledge they must be epistemically, and not just causally justified. That is tosay the acceptance of a belief must be epistemically rational, and carried out in an epistemicallyresponsible maimer. Rather than concentrating solely on the conditions in light of which a belief isjustified, as externalists do, we need instead to consider a variety of epistemic practices in whichepistemic agents ordinarily engage. Among these practices are public acts of justification, anddemands that such acts be satisfactorily carried out. If we are to show a belief is justified, saysBonJour, we typically will need to produce “a justificatory argument in which the belief to bejustified is shown to follow inferentially from some other belief which is thus offered as a reasonfor accepting it.”16 Our practices reveal that our interest in acts of justification is a practical one.We want to be justified in believing what we do, so we engage in justificatory argument toconvince ourselves and others that we are correct in believing as we do. Externalism ignores thispractical motivation by adopting a God’s eye point of view and claiming that being justified is aproperty of beliefs or persons as seen from that omniscient perspective.To this basic objection BonJour adds two refinements. First, the distinguishingcharacteristic of epistemic justification is its “internal relationship to the cognitive goal of truth. Acognitive act is epistemically justified. . . only if and to the extent that it is aimed at thisgoal—which means at a minimum that one accepts only beliefs that there is adequate reason tothink are true.”7 Second, the notion of epistemic justification is a normative one that suggests128agents have epistemic duties or obligations. One’s intellectual duty, therefore, is to accept beliefsthat are true and reject beliefs that are false. To do otherwise is to be epistemically irresponsible.It is to this conviction that BonJour appeals in a set of cases that attempt to show that sinceexternalism allows agents to be irrational and epistemically irresponsible its account ofjustification must be wrong. For, externalists have no way of criticizing the irrationality of anirrational person’s belief; save pointing to some sort of dysfunction in the cognitive apparatus.Externalists, on this reading, have no way of denying that a belief may be justified even though theagent either has no reason to think it is justified (i.e. reliably formed) or does have reason to thinkit is unjust/Ied. BonJour’s strategy is to make a prima facie case for internalism by citingexamples that violate our internalist intuitions, and thereby shift the burden of proof toexternalists.Each case involves an agent who has clairvoyant powers and who on the basis of thesepowers correctly believes that the president is in New York. If such powers exist they wouldsatisfy the externalist requirement of a reliable cognitive process. In each case offered by BonJourthe agent cannot be said to know or be justified in claiming to know even though his or her beliefis true and reliably formed. This is because:1. The agent has no reason to believe in his or her clairvoyant powers, or has good reasonto believe his or her powers are unreliable.2. The agent disregards the available evidence or has no evidence.The fact that in four different cases the agent holds a true belief based on a reliable processdoes little, says BonJour, to dispel the intuition that someone who ignores evidence and lacks anyreason to trust in clairvoyance is not justified. Our idea of being justified simply is one thatrequires rational deliberation and epistemically responsible action.One of the motivations for the internalist response to externalism is the desire to showhow knowledge is possible despite the skeptical conclusions forced on us by the infinite regressproblem. If there is to be a foundation, and thus no vicious infinite regress, there must be someempirical beliefs that are genuinely justified without being inferentially dependent on any otherempirical beliefs. Beliefs such as these are to be the foundations upon which inferential beliefs may129rest. Thus, it is the justification of non-inferential, largely perceptual beliefs that is frequently thefocus of externalism. From the perspective of internalists who hold to a coherence theory ofknowledge the externalists’ tendency to consider justification solely in terms of the justification ofnon-inferential beliefs is a question-begging strategy. This is so because such a strategy ignoresthe very thing at issue between foundationalists and non-foundationalists, namely the role ofsurrounding beliefs in the justification of a single belief An objection of this sort is prominent inFeldman’s discussion of Alvin Plantinga’ s theory of proper functionalism, according to whichwarranted belief is the result of the proper fi.inctioning of the cognitive system. According toPlantinga the crucial problem for internalists, and one they have no answer to, is that cognitiveagents could do everything necessary in order to be epistemically responsible and still be badlymistaken. On this view neither having evidence nor assuring oneself of its relevance and adequacyis sufficient for warrant. This is because malfunctioning perceptual faculties might providesubjectively unequivocal evidence for a belief that is objectively false. What is required forwarrant, says Plantinga, is proper function, or more precisely, the absence of cognitive pathology.To make this case Plantinga introduces the naturalist’s version of Descartes’ evil demon: braindisorders that cause the agent to hallucinate. One example concerns the appearance of an alfalfapatch at ten minute intervals that inclines the agent to believe there is alfalfa before him.Feldman’s objection is that Plantinga’s example omits the sort of background beliefs that whenconsidered would prevent a rational agent from believing that what he saw was really alfalfa andnot a hallucination. These background beliefs ordinarily would count as part of the overallevidence for a belief so the fact that such beliefs are missing in Plantinga’ s example counts heavilyagainst his argument. Feldman observes that Plantinga’s examples seem to support his positiononly because they omit the very details that would force an epistemically responsible agent toreconsider beliefs that otherwise seems true and self-evident.Consider again the alfalfa stand example. It is intended to show that a true belief based ongood evidence need not be knowledge.... Now what else are we supposed to think aboutthis case? Here’s one way things might go. I’m sitting in my study looking at my computerscreen, my keyboard, my notes, and Plantinga’s manuscript while I write this paper. Everyten minutes I get a vivid appearance of an alfalfa stand and I find myself inclined to believethat I do indeed see a fine stand of alfalfa. Right. Do I think that it is odd that there’s130suddenly an alfalfa stand in my study, or do I think that they sprout up indoors all thetime? Do I realize that I didn’t think there was one there a moment ago? Do I rememberthat I did think there was one there ten minutes previously? If my background beliefs andevidence are anything like normal, then it is far from clear that my overall evidence doessupport the belief that I see an alfalfa field. The mere fact that I have an alfalfa standappearance and an inclination to believe that I see an alfalfa stand does not guarantee thatthat belief has evidential support. So, it is far from clear that evidentialism goes wronghere.’8Feldman claims further that in cases where there is a coincidence between the hallucinationand what actually obtains the situation is analogous to Gettier type examples that only show thatsome further condition is necessary in addition to the traditional three. Plantinga’s argument, then,either depends on examples that don’t in fact make his case or it depends on Gettier typeexamples that pose no challenge to the idea that evidence is necessary for justified true beliefPlantinga’s efforts at refuting deontologism fail for the same reasons. His specific target isCbisholm’ s claim that epistemic justification requires that cognitive agents form and hold theirbeliefs in accordance with their epistemic duties, especially the obligation to seek true beliefs andshun false ones. Plantinga’s strategy is to provide a case where you fulfill your epistemic dutiesbut nonetheless end up holding an outrageously incorrect belief Due to a brain lesion you arecertain you will be the next president of the United States, though you lack any evidence insupport of this view. In the absence of any genuine doubt you needn’t mull over the possibilitythat you are mistaken so you accept this belief as true when it is not. You have been epistemicallydutiful yet seriously mistaken. Feldman’s response to this scenario is that “even if the obviousnessof the belief counts in its favor, everything else about the situation counts against it.”19 Otherbeliefs would alert the epistemically responsible agent that something was amiss. You wouldknow, for example, that you had never run for political office, that you had made no plans to raisethe enormous sums necessary for a successful campaign, that this certain feeling of imminentsuccess was not based on anything more than a peculiar and groundless confidence.The dispute between internalists and externalists is at an impasse since both camps havebeen able to identify serious difficulties with the opposing theories. Nevertheless, doubts remainamong some externalists that internalist intuitions can be abandoned altogether. On this viewextemalism’ s value has been to stimulate refinements to the received view.131The Surplus Value of the Internalist ConditionAmong externalists there are some, like Schmitt, who accept the idea of an ethics of belief;and do see a limited role for intellectual autonomy.20According to Schmitt, internalists have beenguided by the view that intellectual autonomy does more to ensure the agent will hold true beliefsand avoid error than does a reliance on authorities. In this, says Schmitt, they are mistaken. Notonly, he says, do we obtain more true beliefs and fewer erroneous beliefs through the testimony ofothers, we frequently and rightly do so without evaluating the reliability of the source. Manywould dispute this last point, granting only that belief on testimony is rational provided the agentassesses the source’s reliability. This was Hume’s position. But Schmitt argues that subjects are inno position to make informed judgements regarding a source’s reliability, unless they themselvesare sufficiently expert. This is the defect of what Schmitt calls Humean testimony. In place ofHumean testimony Schmitt argues for nonHumean testimony where justification of belief ontestimony takes the form of a process that maps the beliefs of a collection of authorities (who aremostly unknown to the subject) onto the beliefs of the epistemically dependent agent.Since justification on nonHumean testimony evidently requires that the source’s belief bejustified, and most justification is based on nonHumean testimony, the present view entailsa regress ofjustification: the subject’s justification traces to the source’s, but the source’smust trace to further sources. But on the present view, such a regress is not necessarilyharmful and is indeed consistent with a collective empiricism. While an individual’sjustification regresses, individuals each contribute some observations to the justification.No individual’s observations are enough for justification, but the right combination ofobservations and configuration of processes connecting these observations suffices forjustification. Justification is in this sense collective.21What makes nonHumean testimony justified, says Schmitt, is the reliability made possiblethrough the cognitive division of labour. By virtue of its role in the production and disseminationof knowledge this division of labour helps to ensure that those invested with authority areepistemically reliable. In this Schmitt’s position is sympathetic to a point argued for by Allen,Hardwig, Welbourne and Webb: there exists a cognitive division of labour that is to everyone’sadvantage since individually we can acquire very few justified beliefs. Collectively we can acquiremany. It is a view that is both descriptive of how we are typically justified in our beliefs, andprescriptive of how we ought to guide our efforts at belief formation. It represents an attempt to132reach some sort of broad reflective equilibrium between our practices and our epistemologicaltheories.22These theories have led many to suppose that only epistemic independence can serve asthe ideal in belief formation, but have left us with the consequence that very few people knowanything, and the majority of one’s beliefs are without justification. Some reconciliation isnecessary therefore between our practices and our theories.NonHumean testimony is justified in a way analogous to the justification of beliefs frommemory. On the strictest empirical grounds it might be tempting to say we lack justification forbeliefs based upon anything less than immediate perception or inferences from perception. Dretskeand others have noted, however, that memory fi.inctions as a carrier or surrogate for the vast arrayof one’s observations that go into one’s beliefs being justified.23 Even though my presentobservations are not focused on picking up my mail, I distinctly recall picking up my mail earlierthis morning, and have no doubt I am justified in believing this. Similarly, Harman’s discussion ofthe principle of clutter avoidance in belief formation also suggests it would be unreasonable tosuppose that justified beliefs cease to be justified once we forget the reasons we had for holdingthe belief in the first place.24 On the account by Dretske, even though we lack direct perceptualaccess to these observations, we are nonetheless justified in claiming knowledge from memory.This is because memory is a more or less reliable carrier of or surrogate for those observations.On the account by Harmon, forgetting a beliefs justification is seldom reason enough to doubtone is justified, again for the reason that the practice of trusting one’s memory is generally morereliable than not, while keeping track of one’s justifications is simply not feasible. In sum, then,observations and evidentiary arguments that were once available to the agent and which justifiedbelief have over time become inaccessible. Yet justification remains. So too does the justifiabilityof our relying on memory in this way, despite the fact that our memories are periodicallymistaken. By analogy we can view the process of nonHumean testimony as one where manyobservations or evidentiary arguments, though never accessible to us, justi our beliefsnonetheless. This despite the fact that the chain of observation reports may contain errors andfalsehoods. Although it is the individual who comes to hold true beliefs, in most cases it is not theindividual, but the relevant social group that is “the bearer of epistemic epithets.”25 The133observations on which the individual must rely are the observations of the group.This account will be disconcerting to many, especially among those who conclude from itthat the majority of our beliefs must be based on little more than an uncritical aggregation ofopinions. This worry overlooks the critical element in the array of epistemically distantobservations and reports upon which nonHumean testimony depends. In Schmitt’s account thiselement is present in the role played by partial autonomy in the cognitive division of labour;present first, in expertise, and then second, in the indirect assessment of source reliability.Schmitt insists that in the majority of cases beliefs are justified on nonHumean testimony.Such testimony appears to require no autonomous judgment whatsoever since the subject takes asource’s pronouncement on faith (insofar as the subject lacks any observations of her ownregarding either the truth of the belief or the reliability of the source). This faith would bemisplaced only if the processes that constitute the practice of nonHumean testimony wereunreliable. Schmitt, of course, claims these processes are reliable, and depend in some measure onintellectual autonomy. What makes them reliable? In Schmitt’s phrase these processes are“underwritten” by the partial autonomy of individual agents.The essence of Schmitt’s position regarding justification is that, with respect to a givenbelieJ one can be justified on the testimony of others, even when the reliability of these others isnot known. Autonomy, even partial autonomy, is not generally necessary for justification. Thecircumstances under which it is necessary is a contingent matter. In the absence of genuine doubtregarding the truth of a claim or the reliability of a source, simple trust in authority is justified.26But even in cases where trust is misplaced justification doesn’t require that a given belief be theproduct of autonomous deliberation, only that surrounding beliefs be, and then only to the extentrequired by the circumstances.Though Schmitt dismisses Hume’s claim that belief on testimony is justified on the basis ofobserved regularities between source pronouncements and the truth, he does believe that in somecircumstances agents can determine the reliability of sources. The agent is to evaluate only thosepronouncements she is qualified to assess by virtue of her own knowledge of surrounding topics.She can thereby seek topical justification for beliefs regarding reliability, or topical justification134for the original belief From an assessment of these surrounding beliefs she can infer some degreeof source reliability, even though the assessment of surrounding beliefs will likely rely on beliefsbased on testimony. The process can be expressed thus: Source S is judged to be reliable withrespect to p in domain D (or on topic T) because she has been judged to be reliable in domain E(or on topic U) in light of beliefs that may be only testimonially justified.Empirical evidence suggests this is precisely what people do. Research on persuasionreveals that people are persuaded on the basis of reasons.27 That is, they actively seek reasons andarguments even under circumstances where the proffered basis for belief is testimony. That someindividuals are more demanding than others in this regard points not to the requirements ofjustification, but of epistemic virtue. Schmitt distinguishes between the requirements ofjustification and of virtue in terms of the contribution each makes to the realization of epistemicgoals. Being justified on nonHumean testimony adequately serves the epistemic goals of acquiringtrue beliefs and avoiding erroneous ones. The standard of justification fi.irnishes us with anadequate number of true beliefs, while the standard of virtue furnishes us with a generous numberof such beliefs. Adequate service is a minimal standard, however, since generous service ispossible and desirable. The distinction between adequate and generous service of epistemic goalsis roughly analogous to the distinction between morally obligatory and supererogatory action: theformer is required because the latter is scarce and hard to motivate. Considered within aframework of the cognitive division of labour epistemic virtue exists in degrees and constitutesone of the identifying features of expertise. A doctor may be justified in believing what she doeson the testimony of experts (in textbooks or from lectures) though she has no understanding ofthe non-testimonial warrant. Though justified, she is not, on Schmitt’s account, virtuous, sincevirtue requires much more independent assessment and understanding of the grounds for beliefA system for the transmission of knowledge will permit significant acquisition of truebeliefs only on the condition that sources transmit true information or subjects who receiveinformation can distinguish truths from falsehoods. Unfortunately not all sources are reliable, andsubjects often fail to distinguish the true from the false. The extent to which subjects mustdistinguish true statements from false depends on the reliability of their sources, that varies across135domains of knowledge. In domains characterized by complete agreement, where authorities areequally and highly reliable sources of information there is little or no need for autonomy. But suchdomains are exceedingly rare. So in domains where experts disagree subjects need a modicum ofepistemic acuity. In domains characterized by some unreliable sources and controversy subjectsneed at least partial autonomy since inherited attributions of reliability are always prima faciesuspect in such cases. Non-experts can evaluate source reliability by relying on:1) inherited topical beliefs from outside the controversy or problematic domain.2) inherited correlations between reliability and evidence of recognition within a field (rate ofpublication, awards, research grants, etc.)3) inherited attributions of reliability from reliable judges of reliability.In domains where few sources are reliable or controversy is high subjects need to be ableto consider topical evidence for the beliefs under consideration. This will likely require expertise.Expertise is also required in every domain to underwrite the reliability of the cognitive division oflabour within that domain. At various points in the chain of observations and inherited beliefsexpertise is needed to act as a check on the inherited error from belief on testimony.So it would appear that epistemic virtue is necessary insofar as a defining feature of thecognitive division of labour is expertise. The non-expert can and does legitimately trust epistemicauthorities. What consequences does this view have for my argument and for education? If theteacher’s sole goal is to transmit knowledge, then on the externalist view, we can ignore evidentialarguments. But if we are interested in developing epistemically virtuous intellectual autonomy, anemphasis on evidential argument is justified.Acquiring The Concept ofBeing JustUiedOne of internalism’s most trenchant critics, Wiffiam Alston, argues that while there are nosuccessful internalist arguments extant, some internalist intuitions ought not to be abandoned. Theintuition that Alston says cannot be jettisoned in cases like those brought forward by BonJour andFeldman is the intuition that our grounds for belief must be accessible to us. Inexplicably reliablecognitive processes whose reliability I do not assess when forming my beliefs simply do not satisfyall the requirements of being justified. What is missing, says Alston, is136any basis or ground that S has, or possesses for his belief anything he can point to orspecify as that which gives him something to go on in believing this, any sign or indicationthat the belief he has is true.28Aiston calls this necessary feature of being justified “the requirement of epistemicaccessibility ofground for the belief’. This feature is best understood by reference to the practicesof critical reflection that form the background against which the concept of justification hasdeveloped. These practices include that of the epistemic assessment of beliefs, the challenging ofbeliefs, and the response to these challenges. A successful response will, of necessity, specify anadequate ground for belief one that provides a sufficient indication of the truth of the belief It isnot necessary for justification, of course, that every belief must be put to and survive such a test,only that, in principle, it could.29An implication of this view is that beingjustJied is distinct from the activity ofjustzfyingone belief On Aiston’s view the activity of justifying one’s belief is fundamental to beingjustified insofar as it is practically and historically prior to the concept of being justified.Though the activity of responding to challenges is not the whole story, I do believe that ina way it is fundamental to the concept of being justified. Why is it that we have thisconcept of beingjustfied in holding a belief and why is it important to us? I suggest thatthe concept was developed and got its hold on us, because of the practice of criticalreflection on our beliefs, of challenging their credentials and responding to suchchallenges—in short the practice of attempting to justfy beliefs. Suppose there were nosuch practice; suppose that no one ever challenges the credentials of anyone’s beliefs;suppose that no one ever critically reflects on the grounds or basis of one’s own beliefs. Inthat case would we be interested in determining whether one or another belief is justified?I think not. It is only because we participate in such activities, only because we are alive totheir importance, that the question of whether someone is in a state of being just/Ied inholding a belief is of live interest to us.3°It does not follow from this that being justified is a matter of engaging in activities ofjustification. Nor does it follow that one must be successful in these activities before one can besaid to be justified in holding a belief The fact of being justified is not dependent on any particularactivity ofjustifying since there are many justified beliefs a person might hold without ever havinghad the opportunity or inclination to justify them to anyone.It is Aiston’s contention that the development of the concept of being justified has beenstrongly influenced by our social practices of justification. What has emerged from the history of137this development is a sense of what would have to be specified in order to succeed in justifying abelief; in meeting challenges, etc.. In other words our sense of what would serve as an adequateground for belief has emerged from these practices ofjustification. So it is understandable that theconcept of being justified would historically have contained the internalist requirement that whatjustifies be accessible to the subject. The point where Alston departs from internalists is theirclaim that the accessibility of the ground must be available to the agent in consciousness. Hispreference is to say that the ground must be the sort of thing that is typically accessible to normalhuman beings. “To be a justifier an item must be the sort of thing that, in general, the subject canexplicitly note the presence of just by sufficient reflection on his situation.”31 This view is to bepreferred because it avoids the too stringent requirement that agents can only be justified whenthey do, in fact, offer a justification that succeeds. By “sufficient reflection” Aiston meanssomething more than immediately obvious and something less than that which entails lengthyresearch or experimentation.Alston’s externalism is to be found in his discussion of the adequacy of grounds. That thegrounds must be accessible is an internalist condition. That the adequacy of the grounds need notbe accessible to the agent allows Aiston to term his position “an internalist externalism.” There aregood reasons for not insisting that the adequacy of the grounds be appreciated by the agent:1) the infinite regress problem that leads to skepticism2) the skeptical consequence of denying that one can know on the basis of authority3) the skeptical consequence of denying that one can know without being able to show that one’sgrounds are adequateSo on this view the grounds need only be adequate, viz, they need only be indicative of thetruth. If one does see how the grounds are indicative of the truth of one’s belief; so much thebetter. But it is not necessary for justification that a cognitive agent see this truth indicativerelationship between ground and beliefIn order for my belief that p which is based on ground G, to be justified, it is quitesufficient as well as necessary, that G be sufficiently indicative of the truth ofp. It is noway required that I know anything, or be justified in believing anything, about thisrelationship. No doubt, we sometimes do have justified beliefs about the adequacy of ourgrounds, and that is certainly a good thing. But that is icing on the cake.32138This position then acknowledges the contribution of internalist intuitions to our sense ofbeing justified, but it does not view any intemalist condition as necessary or sufficient forjustification. It is the practice of attempting to justify beliefs that gives children the concept ofbeing justified in holding a belief Learning this concept deserves to be an integral part of theeducation of children, for reasons that I will now make plain.The Virtues ofEvidentially-Held BeliefWe have seen that the internalist perspective is the perspective of the epistemicallyresponsible agent. Internalism may not be satisfactory as an account of knowledge or justification,but it serves a usefbl purpose in laying down criteria for rational and responsible belief formation.In a phrase the epistemically responsible agent is guided by the maxim that “an agent should arriveat his beliefs, internally viewed, in whatever manner would result from his beliefs being regulatedby a desire for true belief”33 Beliefs may be regulated in two ways:1) the desire for true belief may inhibit other desires from playing a deleterious role in beliefacquisition and retention.2) the desire for true beliefs may serve to regulate the agent’s action.Consider these cases of epistemic irresponsibility where the agent is motivated by somedesire other than the desire for true beliefs. One’s wish to be admired may be so strong that itcomes to interfere with the acquisition of true beliefs about oneself and others. Or one’s wish thatthe fhture will be bright may lead you to ignore indications that prospects for the future are notgood. In either case the agent’s beliefs are not being regulated by the desire for true beliefs. Onthis view the beliefs we come to possess cannot be a matter of our direct voluntary control, wecan, however, control the actions we undertake in pursuit of justified belief given, that is, wehave the desire to do so. We can, for example, seek and gather evidence responsibly and therebyput ourselves in a position to acquire true beliefs. Or, motivated by our desire for true beliefs, wecan resist the tendency to follow habitual and unreliable practices in the acquisition of beliefThese include those mentioned above: attaching great importance to vivid information, thoughthis information runs counter to reliable information; holding contradictory views; and drawinginductive inferences from extremely small samples.34 This view of epistemic responsibility meets139the externalist objection that beliefs are more upshots than choices. Justification here is tied to theidea of action, and responsibility for that action. So questions regarding whether an individual isjustified in belief are asking whether she has done all she should have in order to acquire truebeliefs. Questions ofjustification are still linked to an ethics of belief but not to an ethics based onrules of acceptance, as several internalists have insisted. Instead what is being proposed is anethics of belief based on rules of conduct. On the view being advanced here the manner in whichone goes about gathering evidence is relevant to the justificatory status of one’s beliefs.The notion of desire is central to this account because, unlike justification, desire can onlybe construed as internal to agent. Clearly there is a change internal to the agent when her desireschange. Thus the desire for true beliefs is something internal to the agent, as are the regulatoryeffects generated by this desire. The epistemically responsible agent’s perspective is therebyinternally generated and rightly termed internalist.This account’s restriction to the desire for true belief may seem to be uninformative,showing only how epistemically responsible action is internal, for it appears to give little practicalguidance in how one might obtain true beliefs. Other candidates for the touchstone of epistemicresponsibility—the desire for coherent belief for example, or freedom from doubt—at least havethe virtue of being informative. But ultimately if alternative views are to avoid being compatiblewith error they must first be grounded in a desire for true beliefs. In sum the desire for true beliefsmust do more than simply inhibit wishful thinking, it must affect the actions of the agent whichpertain to forming and retaining true beliefs. Among these actions are seeking and gatheringevidence, as well as actions that modify or regulate the way we conduct our search for, and use ofevidence.In addition to the benefits associated with epistemically responsible action a knowledge ofjustifications carries several other benefits. Moore, for example, has argued that the surplus valueof knowledge over mere true belief is not simply the truth conducive character of justificationssuggested by internalist accounts.35 Truth, he claims, is only one of the important epistemic goalsserved by justification, others include what he calls the extra-ver/Ic utilities ofjustfication. Itseems to Moore that while epistemologists have devoted enormous amounts of energy to140determining the nature of knowledge they have not thought very deeply about why knowledge isimportant or why we should engage in epistemic endeavors. He notes in passing how odd thislacuna appears, given that the absence of a satisfactory answer to these questions leaves us eitherwithout a justification for education, or with a justification that is seriously weakened in light ofthe central place knowledge is usually accorded among educational goals. Most epistemologistswho have bothered to consider the practical value of knowledge have emphasized theindispensability of truth in human affairs. For this reason Moore terms the standard, exalted viewof knowledge vericentrism. Vericentrism is the view that holds that all epistemic or cognitiveendeavors have true belief as their only goal. Several leading epistemologists appear to subscribeto vericentrism, though they do not so much argue for this position as assert it without argument.For example, Goldman claims that the principal utility of our cognitive faculties is the promotionof true belief36 Alston considers the basic aim of cognition to be one of believing truths andavoiding falsehoods.37BonJour claims that it is only as a means to the truth that justification isnecessary, for if the truth were transparent justification would be beside the point.38Moore contends that there is more utility to justification and knowledge than as a meansto truth. Put simply the surplus value of knowledge lies in the extra-verfic utility ofjustfication.Moore distinguishes five extra-verific benefits ofjustiflcation, internally conceived:1) Since justification provides the agent with the reassurance that his belief is true justificationdecreases the probability he will disregard or abandon justified beliefs.2) Justification for a belief can make the agent aware of certain methods for acquiring justifiedbeliefs. Such an awareness is beneficial because it enables the agent to acquire additionaljustified beliefs in the future.3) Justification increases the “infectiousness” of the belief That is, because the justificationprovides a compelling argument in support of a belief it increases the agent’s ability todisseminate the belief to others. Infectiousness of this sort is beneficial in three inter-relatedways. i) It is beneficial to the community since the belief will spread more easily throughout thecommunity if reasons can be offered in support of it. ii) It gives the agent greater power toselectively influence the beliefs of others. And finally, iii) the infectiousness of a belief will tend141to increase unanimity of opinion within the community.39Obviously, these benefits are not a part of the externalist picture since, on the externalistaccount, justification needn’t be available to the agent. This, in Moore’s view, is a deficiency ofexternalism, one that leads him to conclude that “because externalism makes knowledge a kind ofbelief that lacks any significant extra-verific utility, externalist knowledge is not as useful a thingas we might expect knowledge to be.”4° Knowing the justification for beliefs is useful both forindividuals and communities. A knowledge ofjustification gives the agent power to influence thebeliefs of others. It also enhances the ability of a community to engage in and succeed at cooperative ventures by virtue of the fact that practices of justification permit the growth ofunanimous correct opinion.Catherine Elgin offers an ironic view that goes several steps beyond Moore’s in claimingthat epistemologists of either stripe have attached too much importance to the goals of acquiringtrue beliefs and avoiding erroneous ones. This is because knowledge (construed internally orexternally) is more easily attainable to those who are the least epistemically virtuous. Indeed it isElgin’s contention that those who are stupid are the most epistemically advantaged.4’Unfortunatethough it may be, the sad fact is “that since qualities of mind like sensitivity, breadth, and logicalacumen often interfere with the satisfaction of the requirements for knowledge, individualsdeficient in such qualities have an epistemic edge.”42 The more sophisticated and numerous anagent’s conceptual schemes the greater are the chances she will err. This is because theemployment of crude categories where distinctions are obvious offers a higher probability ofacquiring true beliefs than the employment of numerous and subtly differentiated categories. Aswell, the more searching and responsible an agent the more likely she is to be distracted bynuances. Those who are most epistemically responsible will tend to suspend belief more often, andfor longer periods of time since the number and complexity of their discriminations introduces andebilitating level of uncertainty. In the meanwhile the credulous and obtuse, less troubled bycompeting hypotheses, will obtain more true beliefs. Oddly enough, “stupidity can enhance, andintelligence diminish, one’s prospects for knowledge.”43 The upshot of her argument is thatknowledge is not a particularly valuable cognitive achievement. We would do well, she says, to142attach more importance to other sorts of cognitive excellence, since those who know more are notfor that reason cognitively better off Among those excellences worth pursuing are logicalacumen, breadth and depth of understanding, conceptual sophistication and the capacity todistinguish important from trivial truths.The Prevalence of Unreliable TestimonyA problem for Elgin’s argument is the challenge to epistemic independence offered byStich, Allen et al. On their view, which depends on minimalist conceptions of knowledge andjustification, cognitive excellence of this sort has little point. If one will acquire more true beliefsby deferring to cognitive authorities, it would be inefficient and irrational to think for oneself Thisview seems hard to credit. Let us consider why. The arguments from Stich et al. succeed as muchas they do on the basis of a limited range of examples taken from academic and professional life.The reliability of beliefs on testimony comes from institutional and legal sanctions againstdeception and error. In universities and laboratories the claims of our colleagues will tend to betrustworthy because there are mechanisms to ensure erroneous beliefs are detected or preventedfrom reaching their intended audience. Newspapers and magazines have fact checkingdepartments, academic journals have blind reviews of submitted manuscripts, witnesses in courtare cross-examined, office-holders must possess the relevant credentials. These mechanisms aresaid to guard against incompetence as well as deception. The sanctions against lying, in courts aswell as in professional life, are sufficiently severe, one would suppose, to deter deception andensure an acceptable level of trustworthiness in the testimony of others. And yet it is obvious thata great deal of the information made available to us is false or misleading. Bookshelves sag underthe weight of patently unreliable books authored by quacks, New Age mystics, and ideologues ofvarious sorts. Faith-healing television evangelists purport to heal afflictions ranging from arthritisto AIDS. Holocaust deniers have today an avid readership.44Newspapers and the television newsmedia mislead the public either by failing to offer comprehensive accounts of the events theycover, or by ignoring some events altogether.45 The tabloid presses routinely print outlandishstories regarding supernatural phenomena and the sighting of celebrities who by all accounts havebeen dead for years. One might reply to these examples that few people are persuaded by these143sources since the bulk of our sources are reliable and thus fail to square with those marginalaccounts that form the minority. In the main those who do read the National Inquirer do so to beamused, not to be informed. But this reply concedes that individuals do more than trust what theyread and hear. It also concedes that some indefinite number of people are credulous in theextreme. That tabloid accounts are more often the objects of scorn and derision attests to theepistemic independence of readers, not simply their conformity to majority opinion.But it might be argued that examples of superstition and foolish belief make my case tooeasy. Let us, then, take the preferred domain of Stich, Nisbett, Polanyi and Allan: that of science.On their view the cognitive division of labour is so finely articulated that a great deal of scientificresearch must be based on trust in the findings of one’s colleagues. The reliability of scientificresearch comes from the institutional mechanisms discussed above, and the shared norms ofethical conduct that informally govern the lives of scientists. This sanguine view of thetrustworthiness of scientists can be easily challenged. There is, first of all, the probability thatresearch findings will, at times, be compromised by careless or sloppy research practices. Evendiscounting this possibility, there is growing evidence of fraudulent work within the scientificresearch community, and evidence that this state of affairs is ftuily appreciated by scientists.47 Thebio-medical research community, for example, has witnessed several highly publicized cases ofresearch fraud.48 Researchers have been exposed for plagiarizing, and falsifying or fabricatingdata. That these researchers were caught out is both surprising and revelatory of the fragile basisof trust. Much of this trust is based on faith in the efficacy of peer review and replication to rootout flawed research. But neither of these mechanisms is as reliable as supposed.49 Several factorsare likely to make effective peer review and replication rare:1) There is a shortage of qualified referees in relation to the enormous number of articlessubmitted for publication.2) The complexity and multiplicity of research techniques further limits the availability of qualifiedor sufficiently zealous referees.3) Internally consistent and plausible falsification cannot be detected by referees since a) they donot examine the original data, and b) they do not witness the gathering of that data.1444) There are few incentives for replicating, others’ work since a) funding is seldom available forreplicative studies, and b) academic credit tends to be awarded only for original research. Evenin cases where attempts are made to replicate others’ work anomalous results are seldominterpreted as evidence of fraud or incompetence.One might expect reliability in co-operative research ventures to be maintained by thevarious members of the team, each eager to protect themselves from a reputation for sioppy orfraudulent work. Yet the reason for co-operation—the cognitive division of labour amongexperts—is the very thing that prevents team members from being able to detect errors in eachothers’ work. As well, collaborators historically have not been held responsible for the errors oftheir partners, so there is no strong prudential reason to be vigilant. In any event, many scientistsdo not believe serious consequences would attend revelations of fraud in their work.5° Indeedthere are several prudential reasonsfor being lax in reporting colleagues for misconduct.1) The confidentiality of informants is difficult to ensure since charges of scientific fraud will in alllikelihood lead to an investigation requiring their testimony.2) When scientific misconduct is found within a research facility or university the institution’sreputation suffers, along with the prestige of the scientists working at that institution. Theseconsequences may in turn lead to a diminution of funding for future research at that institution,or, in extreme cases, a crisis in public confidence regarding the entire field which leads to anoverall reduction in funding.Whistleblowing is not likely to be worth the trouble. One may wonder therefore whetherthis discussion shows we rightly tolerate a significant level of unreliability in information we havelittle choice but to accept, or whether the standards on which we accept information as reliableare intolerably low. Hardwig’s answer is neither of these, but rather that knowledge must veryoften be based on trust in the character of others. One effect of this view is to turn on its head thephilosophical orthodoxy that epistemology is more basic than ethics, that ethics must first meetepistemological tests to be credible. Yet if significant categories of scientific knowledge dependon the moral character of individual scientists working together, then in these categories anecessary condition for knowledge is morality. In short, if epistemic claims are to qualif’ as145knowledge they must meet ethical standards as well as epistemic ones. Experts must not only beepistemically virtuous, but morally virtuous as well. Indeed the two types of virtue appear toconverge.Commitment to BeliefIt is not clear that Hardwig’s insisting on the moral basis of knowledge claims addssufficient reliability to these claims to make blind trust in cognitive authorities rational. It does,however, undercut the claims of those who insist that institutional checks and balances renderbelief on testimony reliable. It may well be that for many or most of the knowledge claims weencounter, but for which we have nothing more than testimonial support, the appropriate courseof action is the suspension of belief Another possibility is that when the testimony is plausible,that is, when it coheres with our other beliefs, what is called for is something less than belief butmore than its suspension. Richard Foley calls this intermediate ground “commitment to belief”5’There are various ways, says Foley, of committing yourself to a belief You can presuppose it,postulate or hypothesize it. You can assume it. Each of these is a doxastic attitude distinct frombelief They are also distinct from merely acting as if one believes. This latter possibility is one wehave reason to suspect characterizes the doxastic attitude of students in school. Such an attitude isnot a matter of intellectual commitment to the belief so much as it is a matter of public display. Ofsignificance to my thesis is the fact that acting as if one believes does share some features ofcommitment. Both are context dependent in a way that genuine belief is not. When you commityourself to a proposition, as when you merely act as if it true, you ordinarily do so only in anarrow range of contexts. A scientist may commit herself to a hypothesis for the purpose ofseeing where this commitment might take her in an experiment. Her reason for acting as if thebelief is true is a non-epistemic practical one. If asked outside the lab whether she believed thishypothesis to be true, she could say without fear of self-contradiction that she did not. Similarly astudent could act as if she believed that leaden balls of different weights would fall at the samerate because she has a practical reason to do so (such as maintaining her grade point average).She might act this way even though outside the context of a test that counted for grades, shemight state that the heavier of the two balls would fall at a faster rate. Thus we can understand146this example from chapter 2 as an illustration of the non-portability across contexts of “beliefs”based on non-epistemic considerations. For educators this lack of portability ought to be seen as aserious deficiency. “Beliefs” grounded in practical considerations may cease to have force outsidean unduly narrow range of situations. Genuine beliefs are different. Beliefs aren’t held relative tocontext, so they carry across contexts. You believe them or you don’t. It follows that belief isneither necessary nor sufficient for commitment. I would submit that a great deal of what is takento be evidence of students’ beliefs in school, may be evidence only of their commitments. It maywell be that, on account of common teaching practices, students are not only barred access toknowledge, but to belief as well. This distinction also lends credence to Mill’s claim that “beliefsnot grounded on conviction are apt to give way before the slightest semblance of an argument”,for the simple reason that with non-epistemic reasons there is so very little to give way.52If we are to view a significant category of student “beliefs” as commitments, then it wouldbe appropriate to clarify for students the epistemological status of their commitments. It is afeature of commitments like hypotheses that agents hold them for the purpose of acquiring truebeliefs by means of further inquiry. But if the proper status of these commitments is not madeplain there is less motivation to carry out further inquiry, or for that matter to keep an open mind,be alert to disconfirming data and so on. The most straightforward way to apprise students of theepistemological status of beliefs they may hold tentatively is to introduce them to the strengthsand weaknesses of evidential arguments in their favour, to the arguments and counter-argumentsthat bear on the question of their being justified. The question isn’t whether it is permissible tohave students hold commitments rather than beliefs since from an educational standpoint both areacceptable. The point is that regardless of whether or not students genuinely believe gases expandwhen heated, teachers have an obligation to make clear what sort of epistemic entity they haveput in the way of their students. Students ought to understand what grounds there are for beliefsthey are expected to adopt in school, and have some sense of the strength of these grounds. Sucharguments, counter-arguments and problematical situations that would convey a sense of thesegrounds and their strength are to be found in what Popper referred to as the third world ofknowledge. In selected circumstances it seems perfectly reasonable to introduce students to this147world to a much greater degree than is done at present.‘Oddly enough, this externalist objection is in keeping with Wittgenstein’s reminder toMoore that knowledge as a term picks out that class of beliefs for which it is necessary to offersome argument. One’s perceptions are not ordinarily in need ofjustification since they are not thesort of thing about which one has any choice but to believe. This is not to suggest, of course, thatWittgenstein would endorse externalism.2Ludwig Wittgenstein, Philosophical Investigations (Oxford: Basil Blackwell, 1986).3Hilary Komblith, “Some Social Features of Cognition,” Synthese, 73 (1987): 27—41.4These examples are drawn from Kornblith’s discussion of epistemic responsibility.Komblith does not take these examples to be indications of epistemic irresponsibility so much asevidence of the restrictions under which cognitive agents must operate. They are for all theselimitations still rational, though their reasoning at times falls far short of ideal reasoning. This isnot my view, for agents do have other choices than to jump to conclusions on the basis of limiteddata, or let themselves be swayed by vivid examples when less dramatic information amplysupports a contrary view.5Richard Nisbett and Lee Ross, Human Inference: Strategies and Shortcomings of SocialJudgment (New York: Prentice-Hall, 1980).6Richard Nisbett et al. “Popular Induction: Information is Not Necessarily Informative,” inJudgment Under Uncertainty: Heuristics and Biases, ed. Daniel Kahneman, C. Slovik, and AmosTversky (Cambridge: Cambridge University Press, 1982), 112—113.7Daniel Kahneman and Amos Tversky, “On the Psychology of Prediction,” in JudgmentUnder Uncertainty: Heuristics and Biases, ed. Daniel Kahneman, Slovik, and Amos Tversky(Cambridge: Cambridge University Press, 1982).8M. I. Alpert and W. T. Anderson, “Optimal Heterophily and CommunicationEffectiveness—Some Empirical Findings,” Journal of Communication 23 (1973): 328—343; E.W. Rogers, Dffusion ofInnovation (New York: Free Press, 1983).9Paul Once, “Logic and Conversation,” in Syntax and Semantics, voL 3: Speech Acts, ed.P. Cole and J. L. Morgan (New York: Academic Press, 1975).‘°William P. Alston, “An Internalist Externalism,” Synthese, 74 (1988): 269; Frederick F.Schmitt, “Justification, Sociality and Autonomy,” Synthese 73 (1987): 43—85;“Richard Fumerton, “The Internalism/Extemalism Controversy,” PhilosophicalPerspectives: Epistemology, 2 (1988), 455.‘2lbid., 454.‘31n “The Internalist Conception of Justification,” Goldman states that “beliefs are justifiedif and only if they are produced by (relatively) reliable belief-forming processes,” 47.148‘4BonJour, 57.‘5Edmund Gettier, “Is Knowledge Justified Belief?,” Analysis 23 (1963): 121—123.16Lawrence BonJour, “Externalist Conceptions of Justification,”“7BonJour, 57.‘8chard Feldman, “Book Symposium: Proper Functionalism,” Nous. 27, no. 1 (1993):3 5—36.19Jbid 36.20Frederick F. Schmitt, “Justification, Sociality and Autonomy,” Synthese 73 (1987):43—85.21Jbjd., 60.22Stich and Nisbett are not in this company, since they expressly repudiate what they call aDavidsonian belief in our epistemic practices. On their view the empirical evidence regarding theirrationality of common practice disallows such optimism. Most of their pessimism rests onresearch which shows the majority of people fail to appreciate what is fallacious in the gambler’sfallacy.23Fred Dretske, “A Cognitive Cul-de-Sac” Mind 91(1982): 109-111..24Gilbert Harman, Change in View: Principles ofReasoning (Cambridge, MA: MIT Press,1986).25Scl-itt 62.261n stressing genuine doubt I mean to distinguish between it and philosophical doubt. Thesentiment here is best expressed in Wittgenstein’s remark that we are not in doubt simply becauseit is possible for us to imagine a doubt. See Philosophical Investigations, paragraph 84.27J. Saltiel and J. Woelfel, “Inertia in Cognitive Processes: The Role of AccumulatedInformation in Attitude Change,” Human Communication Research 1 (1975): 333—344; B. J.Calder, C. A. Insko and B. Yandell, “The Relation of Cognitive and Memorial Processes toPersuasion in a Simulated Jury Trial,” Journal ofApplied Social Psychology 4 (1974): 62—93; 3.E. Danes, J. Hunter and J. Woelfel, “Mass Communication and Belief Change: A Test of ThreeMathematical Models,” Human Communication Research. 4 (1978): 242—253; W. B. Lashbrook,W. B. Snavely and D. L. Sullivan, “The Effects of Source Credibility and Message InformationQuality onto the Attitude Change of Apathetics,” Communication Monographs 44 (1977):242—262.14928William P. Aiston, “Internalism and Externalism in Epistemology,” in EpistemicJustification: Essays in the Theory of Knowledge (Ithaca, NY: Cornell University Press, 1989),224.29That we see no point in asking if a dog is justified in thinking that his master is at thedoor lends indirect support to this view. Beings that are incapable of critical reflection, thatcannot cite reasons which might reassure us that the belief in question was true are not the sort ofbeings for which the question of their being justified arises. Philosophers might wonder if dogshave beliefs which are justified, but they do not, so far as I know, actually ask dogs to justify theirbeliefs, then go about determining the adequacy of the proffered justification. This is so, despitethe fact that dogs have the cognitive apparatus that enables them to have true beliefs. They maybe “reliable cognitive instruments,” as Armstrong puts it, but they are not part of the communityof those who give and demand reasons.30William P. Alston, “An Internalist Externalism,” in Epistemic Just/1cation: Essays in theTheory ofKnowledge (Ithaca, NY: Cornell University Press, 1989), 227—245.31Ibid., 238.32Thid., 243—244.33Hilary Komblith, “Naturalizing Rationality,” 126.“I should point out that Kornblith’s own view is that one can over generalize on the basisof