Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The psychology of democratic deliberation : from practice to system Moscrop, David 2017

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

Item Metadata

Download

Media
24-ubc_2017_may_moscrop_david.pdf [ 3.49MB ]
Metadata
JSON: 24-1.0343290.json
JSON-LD: 24-1.0343290-ld.json
RDF/XML (Pretty): 24-1.0343290-rdf.xml
RDF/JSON: 24-1.0343290-rdf.json
Turtle: 24-1.0343290-turtle.txt
N-Triples: 24-1.0343290-rdf-ntriples.txt
Original Record: 24-1.0343290-source.json
Full Text
24-1.0343290-fulltext.txt
Citation
24-1.0343290.ris

Full Text

					THE	PSYCHOLOGY	OF	DEMOCRATIC	DELIBERATION:	FROM	PRACTICE	TO	SYSTEM			by		David	Moscrop		B.Soc.Sci	(Honours),	University	of	Ottawa,	2007	M.A.,	University	of	Ottawa,	2008			A	THESIS	SUBMITTED	IN	PARTIAL	FULFILLMENT	OF	THE	REQUIREMENTS	FOR	THE	DEGREE	OF		DOCTOR	OF	PHILOSOPHY			in		THE	FACULTY	OF	GRADUATE	STUDIES	AND	POSTDOCTORAL	STUDIES		(Political	Science)				THE	UNIVERSITY	OF	BRITISH	COLUMBIA		(Vancouver)			March	2017		©David	Moscrop,	2017		 	ii	Abstract 	Accounts	of	democratic	deliberation	assume	and	require	citizens	who	are	capable	of	rational	and	autonomous	cognition.	Such	individuals	are	expected	to	be	able	to	gather,	process,	and	communicate	information	in	such	a	way	that	allows	them	to	accurately	account	for	their	preferences,	including	providing	reasons	for	those	preferences.	The	epistemic	defense	of	deliberative	democracy	suggests	that	this	is	possible	and	that	citizens	who	deliberate	can	generate	good	judgments	and	decisions.	In	this	dissertation,	I	bring	findings	from	social	and	political	psychology	to	bear	on	the	question	of	whether	citizens	can	make	good	judgments	and	decisions	through	democratic	deliberation.	Data	collected	over	the	last	five	decades	casts	some	doubts	over	whether	they	can.	However,	as	I	argue,	there	is	good	reason	to	believe	that	deliberation,	despite	these	challenges,	is	often	superior	to	alternative	approaches	to	decision	making	and	that,	moreover,	there	are	individual	practice	and	institutional	design	responses	that	can	mitigate	the	deleterious	effects	of	phenomena	that	bring	about	cognitive	distortion,	bias,	and	error	when	citizens	deliberate.	In	the	first	section	of	this	dissertation,	I	argue	that	the	epistemic	defense	of	deliberation—including	the	need	for	rational,	autonomous	citizens—is	challenged	by	findings	from	social	and	political	psychology,	but	that	democratic	deliberation	remains	a	possible	and	superior	form	of	public	judgment	and	decision	making.	In	the	second	section,	I	use	institutional	theory,	deliberative	systems	literature,	and	findings	from	psychology	to	discuss	ways	of	thinking	about	autonomy	and	deliberation,	and	I	develop	approaches	to	limiting	or	overcoming	the			 	iii	challenges	mentioned	in	section	one.	These	approaches	are	rooted	in	both	broader	institutional	design	and	deliberative	system	design	and	in	specific	deliberative	practices.																											  		 	iv	Preface 	This	dissertation	is	original,	unpublished	work	by	the	author,	David	Moscrop.																																							  		 	v	Table of Contents 	Abstract	...............................................................................................................................	ii	Preface	.................................................................................................................................	iv	Table of Contents	...............................................................................................................	v	List of Figures	.....................................................................................................................	x	Acknowledgements	...........................................................................................................	xi	Dedication	..........................................................................................................................	xii	Chapter 1: Can we deliberate? The problem of rationality and autonomy in democratic societies	.........................................................................................................	1	Judgments and decisions	.................................................................................................................	4	Reason giving, good/correct judgments and decisions, and deliberative democracy	.....	6	Rationality and autonomy	...............................................................................................................	10	Epistemic democracy	......................................................................................................................	15	Nature, culture, and deliberation	..................................................................................................	17	Affect and autonomy in political philosophy	.............................................................................	20	Cognitive limitations, distortions, and bias in political contexts	.........................................	27	Models from social and political psychology	............................................................................	33	Chapter outline	..................................................................................................................................	36	Chapter 2: The Epistemic argument for democratic deliberation and the need for autonomy	...........................................................................................................................	46	A note on truth	...................................................................................................................................	51	Deliberative democracy as epistemology	...................................................................................	54	Politics and expertise in deliberations	........................................................................................	57	Two more reasons why deliberation is a more rational way to generate judgments and decisions	.............................................................................................................................................	60	Epistemic proceduralism: the best of two worlds?	.................................................................	63			 	vi	Elstub’s critique of the epistemic defense of deliberation	.....................................................	67	Autonomy in deliberative democracy: the keystone capacity	..............................................	69	Autonomy and the principle of non-self-deception	.................................................................	77	Self-determination and democracy	..............................................................................................	83	Relational autonomy	........................................................................................................................	86	A note on the philosophical debate between free will and determinism	............................	88	Conclusion and summary	...............................................................................................................	90	Chapter 3: Challenges to autonomy in democratic deliberation	..............................	91	A note on epistemology, the brain, and our environment	......................................................	94	What is expected and required of deliberative agents?	.........................................................	96	The expression of rationality and self-determination/autonomy	..........................................	97	Self-determination and autonomy	..............................................................................................	100	Validity of reasons	..........................................................................................................................	102	More legitimate outcomes	............................................................................................................	104	From evolutionary and social psychology to deliberation: dual process cognition in the system 1 and system 2 modes	....................................................................................................	105	The systems approach: how we generate judgments and beliefs	.....................................	110	What is cognitive distortion and bias?	......................................................................................	112	Models of cognition	........................................................................................................................	116	The elaboration likelihood model (ELM): persuasion and opinion/attitude change	.............	116	System justification	.......................................................................................................................	120	Social intuitionism	.........................................................................................................................	128	Automaticity	...................................................................................................................................	137	Now what? Options for responding to challenges to autonomous deliberation	...........	145	Conclusion and Summary	............................................................................................................	146	Chapter 4: Motivated reasoning and democratic deliberation	.................................	149			 	vii	Motivated reasoning	.......................................................................................................................	151	Overview	........................................................................................................................................	151	Hot cognition	..................................................................................................................................	155	Directional goals, accuracy goals, and reason-giving	.............................................................	156	The model of motivated reasoning in summary	.....................................................................	166	Cognitive models, motivated reasoning, and democratic deliberation	............................	169	Conclusion and Summary	............................................................................................................	175	Chapter 5: Institutional deliberative autonomy and practice-based responses to the psychological challenges of democratic deliberation	........................................	176	How institutional design can be used to respond to challenges to autonomy and rationality	..........................................................................................................................................	177	Tradeoffs and limits: Who can and who cannot deliberate	..................................................	180	Habermas and autonomy	..............................................................................................................	183	Institutional deliberative autonomy	............................................................................................	187	The problem	...................................................................................................................................	187	The solution	...................................................................................................................................	191	How to evaluate IDA	.......................................................................................................................	197	Assessment one: Discourse Quality Index	................................................................................	201	Assessment two: expert qualitative reports	..............................................................................	203	Assessment three: Self-reporting	...............................................................................................	205	Summary of institutional deliberative autonomy	....................................................................	207	Concepts for deliberative design within the context of institutional deliberative autonomy	..........................................................................................................................................	209	A note on facilitation	.....................................................................................................................	209	Iteration	...........................................................................................................................................	211	A-rational receptivity	.....................................................................................................................	215	Cognitive diversity	.........................................................................................................................	218	Targeted motivation	......................................................................................................................	221	Next steps for deliberative institutional design: hypotheses and tests	...........................	223	Conclusion and Summary	............................................................................................................	224	Chapter 6: Democratic deliberation and institutions	................................................	226			 	viii	What is an institution and why do institutions matter?	........................................................	227	The rules of the game: Denzau and North	................................................................................	228	The ontology of institutions: Searle	............................................................................................	230	How do institutions and cognition interact?	............................................................................	234	Ontological security: What is there in the world?	.....................................................................	236	Meaning maintenance: Start making sense?	............................................................................	238	How do institutions encourage poor reasoning, judgments, and decisions?	................	242	Campaigns and elections	............................................................................................................	243	The media and attitude formation	...............................................................................................	249	Political parties and partisanship	................................................................................................	251	Electoral systems	..........................................................................................................................	255	Ideology and hegemony: a special case	...................................................................................	258	Conclusion and summary	.............................................................................................................	264	Chapter 7: A deliberative systems approach to generating epistemically better democratic deliberation	............................................................................................................	266	Understanding the deliberative system: Concerns, functions, and roles	........................	267	What are the concerns of a deliberative system?	....................................................................	267	What are the functions of a deliberative system?	....................................................................	268	What are the roles of experts, policy-makers, and citizens in a deliberative system?	........	272	Deliberative systems and cognitive error	.................................................................................	276	More deliberations, integrated deliberations	...........................................................................	278	Situated and balanced technocracy	...........................................................................................	282	Stability and disruption	.................................................................................................................	287	Experiments in deliberation	.........................................................................................................	289	Conclusion and summary	.............................................................................................................	291	Chapter 8: Practicing democratic deliberation	...........................................................	293	Why bother with epistemically good democratic deliberation?	.........................................	294	Implications and further research	..............................................................................................	295	Next steps	.........................................................................................................................................	299	The place of deliberation in our political lives	........................................................................	301	Deliberation, decency, and human capacity	............................................................................	303			 	ix	Bibliography	.....................................................................................................................	304			  		 	x	List of Figures 	Figure	1.	Institutional	deliberative	autonomy	and	its	components……………….............………….195																																  		 	xi	Acknowledgements 	Writing	a	dissertation	is	at	times	a	lonely	endeavour,	but	it	is	nonetheless	a	collective	one.	In	some	ways,	writing	this	has	been	the	most	enjoyable	part	of	my	graduate	school	experience—of	the	official	bits	of	my	program,	anyway.	I	owe	this	fact	to	several	people—far	more	than	I	can	name	here.	However,	a	few	stand	out.	First,	I	wish	to	acknowledge	and	thank	my	supervisor,	Mark	E.	Warren.	I	came	to	the	University	of	British	Columbia	mostly	because	of	him,	and	I'm	glad	I	did.	His	help	in	the	process	has	been	invaluable.	Mark	has	a	singularly	remarkable	way	of	synthesizing,	simplifying,	and	communicating	complex	ideas.	I've	been	the	beneficiary	of	these	skills,	and	I'm	awfully	thankful	for	it.			 Next,	I	thank	my	committee	members.	Andrew	Owen	and	Steven	Heine	have	been	at	once	rigorous	and	encouraging	in	supporting	a	dissertation	that	includes	research	from	philosophy,	neuroscience,	political	theory,	social	and	political	psychology,	and	other	fields.	Their	patience	and	insights	have,	along	with	Mark's,	made	this	dissertation	much,	much	better	than	I	could	have	made	it	on	my	own.	I	chose	my	committee	to	keep	me	honest;	I	figured	that	you	could	perhaps	fool	one	expert	in	one	field	about	research	from	another	area	of	study—but	it	would	be	hard	to	fool	three	remarkably	talented	experts	from	three	different	areas	of	study.	I	chose…wisely.			 Finally,	I	thank	my	family	and	friends;	there	far	too	many	folks	to	name	here—other	than,	of	course,	my	mom,	whose	support	has	been	instrumental	in	me	making	it	this	far	in	the	first	place.	They	know	who	they	are.	My	deepest	thanks	to	them.		  		 	xii	Dedication 	To	the	family	I	have	lost	during	these	years	of	study.	I	was	so	lucky	to	have	you	with	me	at	the	beginning;	I	wish	you	were	here	at	the	end.															 											 			 	1	Chapter 1: Can we deliberate? The problem of rationality and autonomy in democratic societies 	Can	individuals	make	rational,	autonomous	political	judgments	in	democratic	deliberative	settings?	That	is	the	question	I	ask	and	answer	in	this	dissertation.	To	ask	this	question	is	to	ask	two	related,	but	independent	questions.	First:	Are	there	available,	sufficient,	and	meaningful	opportunities	for	citizens	to	become	educated	about	political	issues;	to	come	together	in	a	setting	free	from	coercion—emotional,	psychological,	or	physical—to	discuss	their	preferences,	desires,	and	goals;	and	to	give	reasons	for	and	against	propositions	in	an	attempt	to	reach	considered,	just,	and	legitimate	decisions?	And	second:	Are	human	beings	equipped	with	or	able	to	cultivate	the	necessary	cognitive	capacities	for	rationally	and	autonomously	engaging	in	democratic	deliberation	with	one	another?	While	the	first	question	raises	significant	ethical	and	practical	concerns	about	democratic	systems	and	the	role	of	individual	citizens	in	collective	self-determination	and	self-government,	it	is	the	second	question	that	I	am	interested	in	here.	Why?	Because	the	second	question	points	to	concerns	about	the	extent	to	which	human	beings	are	capable	of	reaching	the	rational	and	autonomous	political	judgments	and	decisions	that	proponents	of	deliberative	democracy	argue	that	they	can,	and	will,	reach	when	deliberating.	So,	the	viability	of	theories	of	deliberative	democracy	thus	depend,	in	part,	on	the	answer	to	this	second	question.	Concerns	about	the	capacity	of	deliberators	to	make	good	judgments	and	decisions	emerge	because	there	is	a	gap	between	what	we	normatively	expect	or	desire	from	those	individuals—critical,	factual,	and	relevant	judgments	and	decisions	whose	origins	are	rooted	primarily	in	the	considered	and	expressed	considerations	of	the	agent,	which	are	themselves			 	2	drawn	from	known	internal	or	external	motivations—and	what	they	tend	to	be	able	to	deliver	day-to-day.	This	dissertation	is	about	what	creates	that	gap	and	how	we	might	go	about	bridging	it.	Given	this	concern	with	autonomy	in	democratic	deliberation,	in	this	dissertation	I	spend	lots	of	time	discussing	human	cognition—its	individual	and	social	manifestations—and	how	it	relates	to	both	the	normative	(i.e.	desired)	and	practical	(i.e.	necessary)	requirements	of	deliberative	citizenship.	In	the	following	pages,	while	I	address	both	the	capacities	of	autonomy	and	rationality	I	pay	particularly	close	attention	to	autonomy	and	argue	that	it	is	the	keystone	capacity	for	good	deliberation.	Autonomy	is	central	to	deliberation	because	while	both	it	and	rationality	are	important	deliberative	capacities,	autonomy	is	necessary	to	deliberation	in	a	special	way.	Democratic	deliberation	requires	that	individuals	put	reasons	on	the	table	for	discussion	and	debate.	While	an	individual	might	have	irrational	reasons	or	justifications	for	their	preferences	(which	are	not	ideal	from	a	deliberative	point	of	view),	if	those	irrational	considerations	can	be	put	on	the	table,	they	can	at	least	be	interrogated	by	the	group.	The	ability	to	deliberate	in	this	way	requires,	as	I	will	argue	below	and	more	extensively	in	chapter	two,	a	capacity	for	autonomy.	So,	autonomy	allows	rational/irrational	reasons	to	be	put	on	the	public	register	and	taken	up	by	those	who	deliberate.	That	is	why	I	argue	it	is	a	keystone	capacity,	and	why	I	spend	extra	time	analyzing	it	vis-à-vis	deliberation.		Again,	the	primary	argument	I	make	in	this	dissertation	is:	When	it	comes	to	democratic	deliberation,	there	is	a	gulf	between	what	we	expect	and	required	of	citizens,	and	what	our	cognitive	capacities	and	the	institutions	that	they	shape/that	shape	them	can	deliver.	In	short:	we,	as	individuals	who	deliberate	in	democratic	settings,	are	expected	to	make	better	judgments	and	decisions	than	we	tend	to	make	outside	of	them,	on	our	own	or	with	others.	We			 	3	are	supposed	to	be	thoughtful	and	critical;	we	are	expected	to	adopt	a	higher	standard	of	critical	thinking	than	if	we	were	just	popping	into	the	grocery	store	to	choose	some	vegetables	and	meat	for	dinner,	or	picking	which	film	to	watch	for	an	evening,	or	choosing	between	the	red	shirt	and	the	blue	one.	But	often	we	fail	to	live	up	to	those	higher	standards.	Often,	something	gets	in	the	way	of	us	making	good	judgments	and	decisions.	Now,	this	argument	could	—no	surprise—also	apply	to	all	kinds	of	human	behaviour	if	there	were	good	reason	to	apply	it,	including	those	just	mentioned:	how	we	eat,	how	we	purchase	goods	and	services,	what	we	watch,	or	even	how	we	choose	a	partner.	But	in	this	dissertation,	I	am	specifically	concerned	with	how	certain	institutional	and	cognitive	structures	and	phenomena	interact	with	individuals	as	potential	participants	in	public,	political	deliberation	as	outlined	by	contemporary	deliberative	democracy	theory.	Proponents	of	democratic	deliberation	argue	that	deliberation	is	a	good	way	to	generate	political	outcomes;	I	want	to	put	that	claim	to	the	test	in	light	of	data	from	political	and	social	psychology	that	(often	implicitly,	when	it	comes	to	deliberation)	suggests	otherwise.		My	focus	in	this	dissertation	is	thus	on	the	relationship	between	individuals,	institutions,	and	deliberative	democracy	vis-à-vis	the	epistemic	function	deliberation	serves	as	a	democratic	practice.	And	while,	as	I	argue,	there	are	reasons	to	be	sceptical	about	the	extent	to	which	individuals	tend	to	make	“good”	judgments	and	decisions—a	term	I	define	below—there	is	also	reason	to	be	optimistic	about	improving	those	judgments	and	decisions	through	careful	personal	practice	and	institutional	reform.	To	get	to	that	point,	I	first	call	into	question	the	extent	to	which	individuals	can	engage	in	epistemically	good	democratic	deliberation.	I	explore	individual	cognitive	tendencies—especially	those	that	distort	or	bias	our	thinking—drawing			 	4	from	empirical	evidence	found	in	social	and	political	psychology.	I	explore	how	these	tendencies	interact	with	the	social	and	political	institutions	found	in	contemporary	democratic	societies	and	argue	that	to	maximize	the	quality	of	judgments	and	decisions	made	in	deliberative	settings,	we	require	changes	in	our	personal	practice,	institutional	design,	and	deliberative	systems.	Nonetheless,	despite	significant	challenges	to	rationality	and	autonomy	posed	by	certain	cognitive	tendencies,	I	argue	that	with	good	practice	and	design,	we	can	achieve	and	maintain	a	high	standard	of	epistemically	valuable	democratic	deliberation.		Judgments and decisions 	In	this	dissertation,	when	I	discuss	a	“judgment,”	I	mean	to	refer	to	a	single,	considered	conclusion	reached	by	an	individual.	For	instance,	I	might	reach	the	judgment	that	raising	taxes	on	the	richest	two	percent	is	good	policy	since	we	can	leverage	the	relatively	minimal	marginal	cost	to	those	individuals	for	proportionately	greater	goods	for	social	programs	to	help	poorer	people.	That	is	a	considered	judgment.	It	is	personal,	discrete,	and	internal.	Related	to	a	judgment,	but	distinct	from	it,	is	a	“decision”—by	which	I	mean	a	deliberate,	collective	choice	to	undertake	(or	to	reject)	some	action	or	series	of	actions.	So,	a	legislative	body	might	take	up	the	judgment(s)	of	one	or	more	individuals	on,	say,	this	matter	of	raising	taxes	and	indeed	make	changes	to	legislation	that	does	just	that.	Decisions,	as	I	mean	them	in	this	dissertation,	are	collective	and	external,	though	they	are	enabled	by	judgments.	Individual	judgments	are	separate	and	distinct	from	collective	decisions,	even	though	they	likely	contribute	to	bringing	about	a	decision.	In	a	debate	or	deliberation,	individual	judgments	will	affect	group	outcomes,	though	they	may	be	of	different	sorts:	moral,	technical,			 	5	strategic,	and	so	forth.	So,	for	instance,	a	group	might	be	made	up	of	individuals	who	make	different	and	incompatible	judgments,	but	who	take	a	collective	decision	for	any	number	of	reasons.	As	an	example,	imagine	a	deliberation	around	what	to	do	about	climate	change.	Jennifer	reaches	the	judgment	that	a	carbon	tax	is	needed	to	address	climate	change;	William	prefers	a	cap-and-trade	system.	Each	explains	their	judgment—their	preference	for	one	system	over	the	other—and	gives	reasons	for	it.	The	body	of	which	they	are	a	part	decides	to	choose	cap-and-trade.	While	Jennifer	and	William	reached	separate	judgments,	the	body	made	a	single	decision.	If	the	body	is	required	to	vote,	perhaps	William	and	Jennifer	will	make	a	deal	in	which	Jennifer	supports	William,	which	is	a	different	sort	of	judgment	leading	to	a	different	decision	(e.g.	to	compromise	or	bargain	in	exchange	for	some	other	good).	The	examples	and	variations	of	examples	of	this	distinction	are	endless,	but	the	important	takeaway	is	that	judgments	and	decisions	are	distinct,	but	related	in	complex	and	variable	ways.	Typically,	however,	and	especially	when	it	comes	to	democratic	deliberation,	a	judgment	is	in	part	the	result	of	an	epistemic	process	of	collecting,	interpreting,	and	sharing	information,	and	then	reaching	a	considered	conclusion	that	made	lead	to	a	decision—which,	again,	for	the	purposes	of	this	dissertation,	is	a	shared,	single	act	of	a	deliberative	or	other	decision-making	body.	This	distinction	is	important	to	remember	because	throughout	this	dissertation	I	will	argue	that	by	improving	individual	judgments—that	is,	by	improving	the	epistemic	value	of	deliberation	by	enhancing	individual	capacities	for	rationality	and	autonomy—we	can	improve	collective	decision	making.	This	argument	rests	on	the	claim	that	if	we	can	improve	the	quality	of	information,	preferences,	types	of	reasoning	and	the	sorts	of	reasons	that	emerge	from	those	combined	considerations,	we	can	better	facilitate	shared			 	6	understanding,	mutual	respect,	legitimacy,	and	the	representativeness	of	decisions.	(Note	that	the	qualities	of	a	good	decision	include	non-cognitive	goods.)		Now,	to	the	heart	of	the	matter:	to	answer	the	question	of	whether	our	cognitive	capacities	of	rationality	and	autonomy	are	sufficient	for	the	type	of	deliberative	democracy	advocated	by	leading	theorists	of	democracy,	I	ask	the	following	research	question:	What	is	the	impact	of	a-rational	cognitive	processes—cognitive	processes	that	precede	or	circumvent	autonomous,	rational	reflection,	but	which	may	have	an	impact	on	conscious	thought—on	the	possibility	of	producing	epistemically	good	judgments	and	decisions	in	deliberative	settings?	I	also	ask,	in	response	to	challenges	to	autonomous	and	rational	deliberation	posed	by	such	processes:	How	can	political	theory	and	political	science	respond	to	the	effects	of	such	phenomena	in	ways	that	underwrite	capacities	of	autonomy	and	rationality	in	relation	to	deliberation?	By	now	you	might	think	that	I	assume	the	answer	to	the	question	“Can	individuals	make	rational,	autonomous,	political	judgments	in	democratic	deliberative	settings”	is	“No.”	Close,	but	that	is	not	quite	true.	I	believe	the	answer	is	“Not	always	and	not	particularly	well;	but	there	are	ways	we	can	do	better.”	Indeed,	the	latter	half	of	this	dissertation	will	be	specifically	about	how	we	can	do	better	through	changes	to	our	personal	practices	and	our	institutions.		Reason giving, good/correct judgments and decisions, and deliberative democracy 	At	this	point,	a	bit	more	definitional	clarity	and	precision	are	required.	What	is	a	rational,	autonomous	judgment	or	decision?	And	what	is	an	epistemically	“good”	judgment	or	decision?			 	7	Answering	each	of	these	questions	is	essential	since	a	great	deal	depends	upon	how	one	understands	each	of	these	concepts;	accordingly,	I	will	carefully	address	them	in	detail	below.	First,	though,	it	is	important	to	broadly	define	what	it	is	I	mean	by	“deliberative	democracy”—since	it	is	within	the	context	of	this	family	of	theories	of	democracy	that	I	will	ask	these	questions—and	why	deliberation	is	critical	to	democratic	theory	in	the	context	of	contemporary	liberal	democracies	(and	elsewhere	for	that	matter).	While	deliberative	democracy	is	a	contested	concept,	one	of	its	core	elements	is	common	among	the	many	theories	of	deliberation—enough	that	we	can	say	it	is	accepted	and	held	in	common:	reason	giving.	Reason	giving	is	a	central	requirement	to	deliberative	democracy;	indeed,	the	practice	is	at	the	very	heart	of	deliberation	as-such	(Bohman	1998;	Gutmann	and	Thompson	2004;	Schneiderhan	and	Khan	2008;	Warren	2002).	Reason	giving	refers	to	the	exchange	of	reasons	among	participants	in	a	deliberation	for	or	against	a	claim	or	proposal	such	that	each	is	accountable	to	the	others	concerning	why	they	want	what	they	want.	Reasons	are	given	for	the	purposes	of	justification,	but	also	so	that	individual	preferences	might	be	produce,		transformed,	and/or	justified	to	others.	The	deliberative	reason-giving	approach	stands	in	contrast	to	the	mere	aggregation	of	preferences	(e.g.	by	a	tally),	which	only	requires	that	a	participant	in	a	political	exercise	states	what	they	want	(typically	before	a	vote	is	taken	to	decide	the	matter).	The	concept	and	practice	of	reason	giving	are	important	for	this	dissertation.	As	I	will	discuss	in	more	detail	in	chapters	three	and	four,	reasons	are	essential	to	deliberation	because	they	are	the	means	by	which	individuals	in	a	deliberation	communicate.	Indeed,	in	whatever			 	8	way	they	are	delivered,	which	may	be	in	a	better	or	worse	way	depending	on	the	individual	and	how	a	given	deliberation	is	structured	and	carried	out,	reasons	enable	deliberation.	Reasons	allow	for	“deontic	scorekeeping”:	essentially	a	tally	kept	of	what	people	say	to	one	another,	which	each	takes	as	binding	the	other	to	their	utterance	and	its	implications	(Brandom	1994).	Reasons	also	generate	democratic	goods	including	trust,	coordinated	actions,	motivation	to	engage	in	future	exchanges,	and	compliance.	Reasons,	when	exchanged	honestly	and	in	a	constructive	way,	create	the	foundations	on	which	democratic	deliberation	rests—or,	if	you	prefer,	serve	as	a	currency	exchanged	between	participants.	But,	just	as	importantly,	as	I	will	argue	throughout	this	dissertation,	when	we	connect	reasons	to	autonomous	judgments	and	decisions,	participants	in	deliberation	become	more	likely	to	produce	better	outcomes.	What	kind	of	better	outcomes?	Warren	(2002)	argues	that	democratic	deliberation—through	“the	give	and	take	of	reasons”—is	concerned	with	bringing	about	three	types	of	outcomes:	political,	ethical,	and	epistemic.	First,	deliberative	democracy’s	political	aims	are	related	to	generating	better	governance	institutions	and	forming	citizen	preferences	with	an	emphasis	on	reaching	acceptable	decisions	based	on	reason	giving;	this	approach	stands	in	contrast	to	alternative	(possibly	complementary)	approaches	to	generating	decisions,	including	aggregating	existing	preferences	through	voting	or	bargaining,	or	by	coercion.	Next,	deliberative	democracy’s	ethical	function	and	related	outcomes	are	bound	up	in	the	deontological	claim	that	individuals	ought	to	be	treated	as	ends	rather	than	means,	and	thus	should	be	a	part	of	the	decision-making	process	when	it	comes	to	decisions	that	will	affect	them—whether	the	questions	at	hand	are	political	or	moral.	This	ethical	commitment	is	part	of	the	“all	affected	principle,”	by	which	all	of	those	affected	by	a	decision	are	accepted	as	having	a			 	9	moral	claim	to	be	a	part	of	the	process	by	which	that	decision	is	made.	Finally,	deliberation’s	epistemic	function	is	based	on	the	argument	that	deliberative	democracy	provides	a	more	rational	and	thus	better	way	of	making	decisions	than	expert/elite,	power-based,	or	strategic	approaches	to	decision	making,	such	as	technocracy,	coercion	or	preference-aggregation.	This	claim	is	grounded	in	the	argument	that	deliberation	is	better	at	producing	collective	judgments	based	on	shared	knowledge	and	understanding	(hence	the	“better”	outcomes).	Knowledge	generated	in	this	way	also	tends	to	enjoy	the	benefit	of	being	valid	(what	it	takes	for	a	statement	to	be	considered	valid	in	the	context	of	deliberation	is	something	I	will	examine	below).	In	this	dissertation,	this	epistemic	function	of	deliberative	democracy	is	most	important	to	me	since	I	am	interested	how	good	or	correct	judgments	and	decisions	can	be	brought	about	by	rational,	autonomous	deliberators—which	requires	that	deliberations	generate	epistemically	reliable	information	flows	and	uptake.		Accordingly,	I	will	focus	primarily	on	the	epistemic	dimension	of	deliberation	and	the	epistemic	goods	that	emerge	from	good	deliberation.	While	the	three	functions	of	deliberative	democracy	may	be	related	and	mutually	reinforcing,	I	do	not	spend	much	time	exploring	the	ways	they	are	related,	except	for	a	few	interludes	where	warranted	(and	explicitly	noted).	It	is	within	this	context	of	the	epistemic	dimensions	of	democratic	deliberation	that	I	define	a	“good”	judgment	and	a	“good”	decision.	A	“good”	judgment	in	deliberation	is	transparent,	valid,	and	reliable.	It	is	transparent	when	the	motivations	of	the	individual	making	it	are	accessible	and	known	to	her	and,	if	required,	to	others;	it	is	valid	such	that	the	premises	and/or	reasons	upon	which	it	is	based	logically	lead	to	the	conclusion	in	a	way	that	can	be			 	10	understood	and	accepted	by	others;	and	it	is	reliable	in	such	a	way	that	one	should	reasonably	expect	an	individual	to	reaffirm	it	in	the	future	given	the	same	facts	and	context.	A	good	decision	is	one	that	is	taken	collectively	based	on	the	compatible	principles	of	good	individual	judgments,	and	which	is	collectively	scaled	so	that	the	outcome	is	based	on,	as	much	as	possible,	accepted	epistemic	foundations	(even	though	individual	judgments	may	vary).1	A	good	decision	should	meet	the	criteria	set	for	good	judgment.	If	in	translating	many	judgments	into	a	decision	something	is	lost,	and	the	criteria	cannot	be	met,	the	decision	is	not	a	good	one.2	It	is	important	to	note	that	I	do	not	intend	the	usage	of	the	concept	of	“good”	in	any	way	that	relies	on	the	correspondence	theory	of	truth	in	which	an	outcome	“matches”	the	world	“as	it	is.”		Rationality and autonomy 	When	it	comes	to	reasoning—which	is	an	important	part,	but	not	the	only	part,	of	how	we	generate	preferences	reasons	to	support	them—I	do	not	assume	that	the	method	of	reasoning	must	be	based	on	either	ends-means	rationality	or	utility	maximization.	By	this	I	mean	that	I	do	not	consider	instrumental	rationality	to	be	the	only	sort	of	rationality	that	matters	when	making	political	judgments	and	decisions	in	a	deliberative	context.	The	key	issue	regarding	deliberation	                                                1	The	end	goal	of	deliberation	is	not	consensus,	but	mutual	understanding	and	acceptance	of	whatever	outcomes	are	reached	by	those	who	deliberate.	The	same	focus	applies	here	since	we	can	base	a	good	decision	on	many	different,	but	nonetheless	good,	individual	judgments.	Any	collective	decision	that	deliberators	reach	may	or	may	not	be	compatible	with	the	substance	of	each	judgment,	but	should	be	compatible	with	the	principles	of	that	judgment.		2	There	are	other	criteria	for	a	good	decision	relevant	to	democratic	norms	and	deliberative	democracy	goods	(e.g.	publicity,	legitimacy,	etc.).	They	are	in	some	ways	to	related	to	my	concerns	here,	but	I	have	bracketed	any	discussion	of	them	since	I	am	primarily	concerned	with	the	epistemic	function	of	deliberation.			 	11	is	whether	there	is	a	defensible	logic	at	work	during	an	exchange	that	reflects	a)	both	a	basic	understanding	of	the	factual	and	normative	reality	of	the	world	and	b)	which	can	be	shared,	understood,	and	in	principle	accepted	by	others	under	reasonable	circumstances.	Within	this	approach,	there	is	also	such	a	thing	as	“more	or	less”	good	(i.e.	better	or	worse)	in	a	given	context.	As	we	will	see	in	chapter	five,	there	is	a	“sliding	scale”	when	it	comes	to	one’s	capacities	for	rationality	and	autonomy,	and	there	is	also	such	a	scale	for	judgments	and	decisions	(within	a	given	system).	This	is	to	say	that,	on	balance,	given	different	approaches	to	reasoning	and	different	interpretations	of	“facts”	about	the	natural	world	and	about	the	normative	world	in	the	context	of	politics,	there	will	be	space	for	dispute	about	what	counts	as	good;	accordingly,	there	is	no	scientific	test	for,	or	objective	measure,	that	signifies	the	“right”	or	“true”	way	of	judging	or	deciding—especially	when	it	comes	to	ethical	or	moral	matters	in	a	society	marked	by	deep	and	persistent	disagreement.3		However,	within	the	context	of	a	given	and	generally-accepted	system,	some	judgments	and	decisions	will	be:	i)	closer	to	reasonable	interpretations	about	what	motivates	the	agent	who	decides;	ii)	based	on	more	or	less	defensible	accounts	about	what	exists	in	the	world	and	how	it	exists	alongside	and	interacts	with	other	things	and	phenomena;	iii)	backed	up	by	more	or	less	honest	and	accurate	accounts	about	the	motivation(s)	one	has	for	holding	the	preferences	that	generate	a	judgment	or	decision;	iv)	more	or	less	likely	to	be	reproduced	in	the	future	given	the	same	or	very	similar	contexts;	and	v)	more	or	less	accessible	to	other	agents	who	have	a	political	and	moral	right	to	participate	in	collective	decision	making	about	the	political	and	social	world	they	share	and	                                                3	Although	science	as	a	way	of	accumulating	knowledge	is	characterized	by	procedures	for	establishing	validity	intersubjectively.			 	12	whose	rights	are	better	respected	by	being	a	part	of	a	political	practice	to	which	they	have	access	to	a	shared	epistemic	foundation	for	decision	making.	Of	course,	there	are	contexts	in	which	conceiving	of	a	capacity	for	rationality	in	terms	of	ends-means	or	utility-maximization	makes	sense—especially	if	that	happens	to	be	how	some	individuals	think	in	some	circumstances	or	how	groups	elect	to	work	through	an	issue	or	some	set	of	issues.	For	the	purposes	of	this	dissertation	I	am	most	often,	unless	otherwise	stated,	specifically	interested	in	how	ordinary	(i.e.	non-elite)	individuals	cognitively	navigate	and	make	sense	of	the	complex	political	world	day-to-day;	and	I	am	especially	interested	in	how	they	think,	judge,	and	decide	when	they	are	asked	to	engage	in	democratic	deliberation,	however	they	may	tend	to	(cognitively)	do	so.	But	cognition	tends	to	rely	on	a	human	capacity	for	rationality	(and	autonomy)	that	is	constrained	or	“bounded”	(March	and	Simon	1958).	Accordingly,	my	argument	is	grounded	in	the	belief—and	supported	by	the	literature	that	I	will	explore	throughout	these	pages—that	rationality	is	inherently	bounded	in	the	sense	of	being	constrained	by	real-world	limitations	that	limit	ends-means/utility-maximizing	rationality.	My	argument	is	also	based	on	the	idea	that	rationality	is	also	inherently	and	inextricably	bound	up	with	a-rationality	to	some	extent.	And	while	rationality	may	be	context-dependent	and	shaped	by	structures	of	authority	and	power	(Flyvbjerg	1998),	there	must	be	nonetheless	a	basic	shared	contextual	rationality	in	specific	contexts	if	we	are	to	achieve	stable	and	widely	understood—though	not	necessarily	permanent	or	universally-agreed-upon—democratic	decisions,	though	navigating	how	we	conceive	of	rationality	in	different	contexts	will	require	some	flexibility	and	openness	to	adaptation.				 	13	Accordingly,	the	conception	of	rationality	that	I	rely	on	for	this	dissertation	is	broad;	by	rationality,	I	refer	to	the	human	capacity	to	make	stable	sense	of	the	world	alongside	others	with	whom	they	live	in	a	social	and	political	community.	This	conception	of	rationality	requires	that	individuals	be	able	to	draw	facts	from	the	world	in	a	more-or-less	consistent,	reliable,	and	objectively	accurate	way	(within	a	given	system	of	meaning),	and	that	they	be	able	to	communicate	them	to	others	on	basic	shared	grounds	(determined	epistemically);	by	implication,	this	conception	of	rationality	also	involves	that	individuals	are	more	or	less	able	to	agree	on	some	normative	facts—which	can	be	decidedly	trickier	in	a	political	context.	My	conception	of	rationality	borrows	a	bit	from	the	instrumental	sense	of	the	term	mentioned	above,	but	is	balanced	by	a	need	for	also	including	a	communicative	rationality	perspective,	which	involves	“processes	of	discussion	and	persuasion”	to	“[help]	form	bonds	of	understanding”	between	individuals	with	different	histories,	beliefs,	and	political	and	scientific	understandings	of	the	world	(Parkinson	2006:	127;	see	also	Dryzek	1990;	Habermas	1996).	Rationality,	for	my	purposes,	is	thus	individually	held	as	a	capacity,	but	publicly	shaped	and	shared	through	interpersonal	communication.	Rationality	in	this	context	is	a	personal	capacity	but	it	is	other-dependent.	Why?	Because	as	I	conceive	of	it	for	the	purposes	of	this	dissertation,	rationality	depends	on	an	intersubjective	process	of	validation—we	come	to	know	that	a	judgment	is	rational	through	exchange	with	others	in	a	given	system,	which	is	essential	to	democratic	(and	other	sorts	of)	life.	Another,	related,	approach	to	conceiving	of	rationality	is	practice-based	and	aimed	at	navigating	the	world	through	practical	understanding.	This	conception	is	often	known	as	the	“practical	wisdom”	approach	to	rationality.	Grounded	in	Aristotle	(1999	[350	B.C.E.])	and	his			 	14	concept	of	“phronesis”	(i.e.	“prudence”	or	“practical	wisdom”),	this	approach	to	conceiving	of	rationality	is	based	upon	reason	as	moral	sentiment	and	it	imagines	human	agency	as	being	best	realized	in	terms	of	a	practice-based	conception	of	judgment	and	decision	making	that	aims	to	produce	outcomes	based	on	knowing	‘what	to	do,	when	to	do	it,	and	how	to	do	it’	(Cameron	2014).	In	this	dissertation,	I	consider	practical	wisdom	as	a	form	of—or	approach	to—rationality	similar	to	those	mentioned	above,	but	suited	to	its	particular	domains	and	contexts.	If	I	am	talking	about	a	specific	conception	of	rationality	in	a	given	context,	I	will	note	this;	otherwise	I	am	referring	to	a	suite	of	approaches	of,	each	with	its	strengths	and	weaknesses,	each	susceptible	to	the	sorts	of	challenges	to	cognition	that	I	will	outline	in	chapters	to	come,	and	each	most	concerned	with	the	requirement	that	individuals	be	able	to	make	sense	of	the	world	in	a	consistent,	valid	way	that	they	are	capable	of	sharing	with	and	communicating	to	others.		Before	proceeding,	I	want	to	make	clear	that	in	this	dissertation	I	am	not	concerned	with	asking	whether	citizens	are	“smart	enough”	for	democracy	in	the	sense	of	them	having	the	raw	intelligence,	I.Q.,	or	so	on,	to	think,	organize,	and	decide	individually	or	collectively.	Obviously,	to	some	extent,	we	have	managed	to	do	relatively	well	at	all	of	this	despite	certain	challenges	to	autonomy	and	rationality	(Heath	2014).4	I	am	far	less	interested	in	the	issue	of	raw	intelligence	than	in,	on	the	one	hand,	how	we	conceive	of	what	we	are	capable	of	and	the	expectations	that	emerge	from	such	a	conception,	and,	on	the	other	hand,	the	ways	in	which	institutions	support,	enable,	interrupt,	or	distort	our	judgments	and	decisions	through	exploiting	cognitive	architectures	or	tendencies	that	initially	evolved	in	humans	as	adaptations	                                                4	The	many	bizarre,	disturbing,	and	disconcerting	events	of	2016	notwithstanding.				 	15	aimed	at	other	purposes.	So,	while	such	structures	or	tendencies	may	be	related	to,	or	a	component	of,	intelligence	as	processing	power,	in	this	dissertation,	I	am	interested	in	how	individuals	reason	and	how	they	explain	and	justify	their	reasoning	and	the	conclusions	they	reach—whatever	their	“level	of	intelligence”	may	be	on	traditional	measures.	This	approach	is	defensible	primarily	because	the	cognitive	distortions	and	biases,	as	I	will	show	in	chapters	three	and	four,	are	pervasive	across	and	throughout	populations—though,	as	I	note	in	chapters	three	and	four,	there	are	some	distinct	intersections	that	tend	to	vary	with	education	and	socio-economic	status	and	are	worth	paying	special	attention	to.			Epistemic democracy 	In	the	following	pages,	I	will	examine	autonomy,	rationality,	(individual)	judgment,	and	(collective)	decision	making	as	they	relate	to	deliberative	democracy	as	a	political	process	aimed	at	generating	judgments	that	are	epistemically	valid	and	authoritative	in	relation	to	the	norms,	facts,	preferences,	and	goals	of	those	who	participate	in	deliberation—what	I	mean	by	deliberation	as	producing	epistemically	good	judgments	and	decisions.	As	Warren	(2002)	argues,	deliberation	makes	public	the	information	and	otherwise	private	reasoning	required	for	reaching	collective	decisions.	This	process	requires	that	processes	of	“challenge,	reason-giving,	and	verification”	(192)	be	undertaken	so	we	can	have	confidence	in	the	epistemic	validity	of	judgments.	As	he	notes	“…the	rational	validity	of	a	statement—its	authority—cannot	be	separated	from	the	processes	that	establish	this	authority	in	the	absence	of	privileged	or	objective	or	independent	knowledge”	(192).	However,	as	he	adds,	this	requires	that	a	pragmatic	consensus	is	reached	(i.e.	public	understanding)	among	participants	on	the			 	16	epistemological	authority,	and,	accordingly,	the	validity,	of	claims	and	judgments.	As	Warren	concludes,	“validity	is	a	product	of	procedures,	suggesting	that	institutionalized	deliberation	can	establish	the	epistemic	validity	of	claims	and	assertions”	(193,	emphasis	mine).	This	stands	in	contrast	to	a	correspondence	theory	of	meaning	or	truth	in	which	validity	is	a	product	of	“linking”	or	“matching”	the	world	“as	it	is”	to	the	world	in	your	head	(and	by	implication	in	the	heads	of	others).	If	deliberative	democracy	is	a	theory	of	democratic	decision	making	concerned	with,	among	other	things,	producing	desirable	outcomes	based	on	judgments	that	are	generated	by	and	which	reinforce	epistemic	validity	and	authority,	then	rationality	and	autonomy	are	essential	capacities	for	those	who	render	judgments.	After	all,	those	judgments	are	required	to	produce	decisions,	and	both	are	underwritten	by	the	normative	basis	of	democratic	deliberation	and	the	requirement	that	facts,	norms,	preferences,	motivations,	and	goals	be	apprehended,	affirmed,	publicized,	scrutinized,	and	publicly	known	to	be	logical	and	relatively	stable.	Consequently,	to	the	extent	that	individual	rationality	or	autonomy	is	undermined	by	cognitive	limitations,	distortions,	or	biases,	epistemic	validity	and	authority	are	also	compromised—and	so	are	the	judgments	and	subsequent	decisions	generated	by	deliberative	processes.	This	undermining	occurs	because	the	epistemic	validity	and	authority	that	these	judgments	and	decisions	require	can	only	emerge	from	a	pragmatic	consensus	emerging	from	exchanges	between	rational,	autonomous	agents	under	(more	or	less)	egalitarian	conditions,	and	within	the	context	of	certain	personal	practices	and	institutional	arrangements	that	enable	good	judgment.	This	reasoning	chain	is	essential	to	my	argument:	the	undermining	of			 	17	rationality	and	autonomy	compromise	deliberative	outputs	by	short-circuiting	validity	and	authority,	and	thus	producing	epistemically	suspect	decisions	(and	judgments).	Some	judgments	in	deliberative	contexts	are	reached	under	conditions	in	which	the	rationality	and	autonomy	of	participants,	and	consequently	any	emergent	epistemic	validity	and	authority,	are	compromised—for	instance	in	cases	of	misunderstood	motivations	or	distorted	interpretations	of	relevant	data/arguments	or	intentions	(a	key	subject	that	I	explore	in	chapters	three	and	four).	It	cannot	be	said	that	such	judgments	reflect	the	actual	(considered)	will	of	the	participants	since	it	is	possible	that	agents	would	reach	different	judgments	and	outcomes	under	conditions	of	complete,	or	near-complete,	autonomy	or	rationality.	As	I	will	argue	in	chapter	two,	this	concern	is	particularly	acute	when	inequalities	in	information	processing	capacities	(e.g.	the	capacity	to	critically	judge	between	several	alternatives)	reflect	structural	cognitive	inequalities	(e.g.	increased	susceptibility	to	framing	effects),	both	in	terms	of	the	extent	to	which	cognitive	distortion	affects	particular	populations	and	in	which	specific	populations	are	affected.		Nature, culture, and deliberation		Of	particular	importance	before	proceeding	is	the	scope	of	this	dissertation.	As	I	will	argue	further	in	chapter	three,	the	root	of	the	fundamental	problem	addressed	herein	emerges	from	a	disconnect	between,	on	the	one	hand,	what	is	required	and	expected	of	deliberative	citizens	(normatively)	and	institutional	arrangements	that	have	emerged	from	decades	of	social	and	political	activity	(what	I	am	calling,	for	my	own,	specific	purposes,	"culture"),	and,	on	the	other	hand,	the	reality	of	how	citizens	tend	to	engage	in	cognition	in	deliberative	(and	other)	settings			 	18	(what	I	am	calling,	again,	for	my	own,	specific	purposes,	"nature")	based	on	our	capacities	for,	among	other	things,	autonomy	and	rationality.	In	short:	our	expectations	and	our	capacities	often	fail	to	link	up,	rendering	both	our	theories	and	our	practice	open	to	critiques	about	just	what	we	are	producing,	how,	and	for	whom.	To	maximize	the	potential	effect	of	deliberative	democracy	in	generating	democratic	goods	through	epistemically	good	deliberation,	the	gap	between	nature	and	culture5	as	it	relates	to	rational,	autonomous	deliberation	must	be	bridged.	Of	course,	the	cultural	milieu	in	which	I	am	writing	this	dissertation,	and	the	one	it	interrogates,	is	a	historical	particularity:	it	comes	from	a	specific	time	and	place,	and	much	of	what	I	address	here	is	far	from	universal.	Maybe	none	of	it	is.	Even	the	way	we	think,	in	a	general	sense,	is	particular.	As	Henrich	et	al.	(2010)	have	convincingly	demonstrated,	styles	of	cognition	in	some	instances	vary	significantly	from	region	to	region	and	it	is	a	mistake	to	state	that	“we	all	think”	in	way	X,	Y,	or	Z.	Moreover,	I	do	not	think	it	would	be	such	a	leap	to	suggest	that	ways	of	thinking	vary	from	historical	period	to	historical	period	given	the	variety	of	concepts,	institutions,	norms,	and	imperatives	that	we	see	throughout	history;	however,	by	and	large,	enough	of	the	general	mechanisms	of	cognition	remain	the	same	despite	particular	adaptive	tendencies	in	response	to	upbringing	and	environment,	including	our	habits	and	the	ways	we	design	the	world	in	which	we	live	that	I	am	able	to	say	something	important	about	how	we	think	and	deliberate	in	contemporary	democratic	societies.	This	suggestion	is	reminiscent	of	the	claim	by	Xenophanes	that	while	humans	tend	to	anthropomorphize	their	deities,	if	horses	could	draw,	they	would	draw	their	gods	as	horses;	the	expression,	in	such	a	                                                5	My	use	of	“culture”	includes	institutions	in	both	a	formal	sense	and	in	the	sense	of	normative	behavioural	expectations.			 	19	case,	would	be	different,	but	the	act	would	be	the	same.	In	the	pages	that	follow,	I	draw	our	horses.	Accordingly,	this	dissertation	examines	deliberative	democracy,	autonomy,	rationality,	epistemology,	and	cognition	within	the	context	of	contemporary	Western	liberal	democracies	marked	by	the	following	characteristics:	entrenched,	multi-level,	and	complex	governance	institutions;	the	rapid	speed	and	extensive	reach	of	mass	communication	technologies;	persistent	disagreement	on	many	issues,	both	superficial	and	substantive;	and	increasingly	urban-dwelling	populations	marked	by	social,	political,	cultural,	religious,	and	ethnic	diversity.	These	characteristics	are	not	chosen	arbitrarily;	rather,	they	reflect	the	types	of	democratic	states	that	deliberative	democracy	has	typically	been	concerned	with,	though	not	exclusively.	These	phenomena	also	influence	cognition,	shaping	and	directing	it	through	external	environmental	cues.	Of	course,	none	of	this	necessarily	precludes	the	possible	portability	of	much	of	what	I	will	argue	in	this	dissertation;	it	does,	however,	constrain	the	claims	that	I	am	making,	situating	them	in	a	particular	time	and	place,	about	particular	political	systems	and	those	who	inhabit	them	in	the	early	twenty-first	century.	Also,	as	I	noted	above,	not	all	deliberation	is	democratic.	I	am	careful	to	use	the	term	"democratic	deliberation"	whenever	I	refer	to	the	democratic	uses	of	deliberation—i.e.	accessible	(to	public	participants),	fair,	open	(to	public	scrutiny	and	review),	and	regarding	questions	of	a	public	nature.	As	He	and	Warren	(2011)	find,	authoritarian	regimes	(e.g.	China)	use	quasi-public	deliberation	for	political	purposes.	And	John	Rawls	(1971)	considered	the	United	States	Supreme	Court	to	be	an	exemplary	deliberative	body.	After	all,	you	can	have	deliberation	without	democracy,	and	you	can	have	democracy	without	deliberation.	I	am			 	20	concerned	here	with	the	cases	in	which	there	are	both.	I	am	interested	in	such	cases	because	I	am	concerned	with	the	(public)	epistemic	defense	of	democratic	deliberation	as	it	relates	to	generating	good	decisions	made	by	citizens	for	citizens.	In	such	vastly	different	settings	(e.g.	carefully-managed	authoritarian	political	exercises,	or	in	camera	court	deliberations)	there	will	be	various	incentives/disincentives,	standards,	institutions,	levels	of	expertise,	resources,	and	so	on	would	factor	into	outcomes.	Such	variance	does	not	mean	that	my	approach	to	analyzing	deliberation	cannot	be	carried	out	for	such	deliberations;	it	could.	But	that	is	not	my	focus	in	this	dissertation.		Affect and autonomy in political philosophy 	To	fully	understand	how	and	why	citizens	deliberate	(or	fail	to	deliberate)	well	(i.e.	rationally	and	autonomously),	we	must	consider	human	psychology.	Traditionally,	however,	political	theorists	in	general	and	democratic	theorists	in	particular	have	paid	only	limited	attention	to	the	impact	of	a-rational	cognition	on	rational,	autonomous	political	judgments	and	decisions,	despite	the	fact	that	these	phenomena	can	contribute	significantly	to	shaping	each.	More	specifically,	to	the	extent	that	a-rational	cognitive	processes	condition,	alter,	or	direct	autonomous	judgment,	they	may	significantly	impact	real-world	political	outcomes,	both	directly	in	the	moment	of	judgment	or	decision	and	in	the	future	by	forging	a	familiar	path	that	subsequent	judgments	and	decisions	may	traverse	(i.e.	via	path	dependency).	Understanding	how	such	processes	and	stimuli	affect	autonomous,	rational	judgment	in	deliberative	settings	will	help	reveal	a	full	portrait	of	the	citizen	as	a	deliberative	agent.	It	may	also	yield	some			 	21	effective	personal	and	institutional	tactics	and	strategies	for	improving	judgment	in	the	future—which	are	the	focus	of	the	second	half	of	this	dissertation.	While	theorists	of	politics	have	yet	to	provide	a	full	account	of	how	a-rational	cognitive	processes	affect	rational,	autonomous	judgment	in	deliberative	settings,	several	political	theorists	and	philosophers	have	provided	accounts	of	the	impact	of	affect6	on	political	judgments	and	decisions.	In	fact,	stretching	back	at	least	as	far	as	Plato,	political	theory	and	philosophy	have	folded	affect	into	the	question	of	what,	if	anything,	makes	a	human	capable	of	rational,	autonomous	judgment	and	under	which	conditions	the	exercises	of	each	capacity	is	possible.	And	while	such	enquiries	were	not	always	expressly	political,	for	these	thinkers,	understanding	the	origins	of	the	causes	of	one’s	judgment	has	been	essential	to	account	for	agency	and	freedom,	since	such	origins	may	have	implications	for	how	politics	should	or	should	not	be	undertaken	and	under	which	conditions.	Ultimately,	in	this	dissertation	I	aim	to	build	upon	these	past	discussions	of	rationality	and	autonomy,	updating	previous	accounts	with	new	findings	from	social	and	political	psychology,	and	placing	the	matter	within	the	context	of	deliberative	democracy	as	a	theory	of	how	we	can	generate	epistemically	good	judgments	and	decisions.	But	first,	we	return	to	the	beginning.	As	noted,	Plato	was	one	of	the	first	philosophers—perhaps	the	first—to	examine	how	affect	affects	judgment.	In	The	Republic,	Plato	divides	and	orders	the	human	soul	(or	mind)	into	three	elemental	components,	placing	the	logical	or	reasoning	element	at	the	top	of	the	hierarchy,	and	the	appetitive	and	spirited	(including	                                                6	In	this	dissertation	I	assume	that	affect	includes	emotion	and	feeling.	Traditionally,	political	theorists	have	often	used	emotion/feeling/affect	interchangeably,	and	in	these	pages	I	will	do	the	same.	In	the	few	cases	where	I	need	to	distinguish	between	the	terms,	I	do	so	expressly.			 	22	emotional)	elements	below	the	rational	element,	since	they	are	in	need	of	regulation	and	control	in	order	for	a	human	to	be	well-ordered	and	capable	of	rational,	autonomous,	enlightened	judgment	(2004	[c.380	BCE]).	Aristotle,	a	student	of	Plato’s	as	well	as	his	critical	interlocutor,	was	also	wary	of	the	potentially	deleterious	impact	of	affect	run	amuck,	arguing	in	The	Nicomachean	Ethics	that	the	incontinent	man—one	who	is	without	self-control,	such	as	one	ruled	by	a-rational	forces—“acts	from	desire	but	not	from	choice”	and	therefore	does	not	act	entirely	autonomously	(2004	[c.350	BCE]:	1111b14-16).	He	even	goes	so	far	as	to	determine	that	the	best	(i.e.	happiest)	activity	of	human	life	is	the	rational	contemplation	of	the	universe	and	its	eternal	truths	(2004:	Book	X).	Nonetheless,	Aristotle	makes	space	for	affect	in	human	life,	arguing	that	it	is	an	essential	element	of	well-being,	and	even	useful	in	judgment	insofar	as	it	may	guide	judgment,	but	only	if	its	role	is	understood,	moderated,	and	its	use	is	kept	in	check	(within	the	mean	between	extremes).	Plato	and	Aristotle	initiated	the	tradition	in	political	theory	and	philosophy	of	asserting	reason	as	the	prime	and	most	effective	capacity	that		makes	a	human	being	capable,	in	a	cognitive	sense,	of	undertaking	political	activity	autonomously,	thus	linking	the	two	capacities.	Whatever	the	place	or	necessity	of	affect	to	the	process	of	thought,	it	was	seen,	at	best,	as	suspect—and	likely	to	be	a	distorting	phenomenon.	The	subsequent	history	of	the	study	of	politics	in	the	West	stayed	the	course	set	by	Plato	and	Aristotle,	remaining	skeptical	of	affect	and	its	effects	on	thought	and	behaviour,	if	not	quite	universally	denigrating	it.	One	of	the	most	pronounced	and	sustained	critiques	against	affect	and	its	place	in	cognition	is	offered	by	Spinoza,	who,	in	The	Ethics	(2000	[1677]),	argues	that	to	the	extent	that	human	beings	act	from	emotion	(a	term	Spinoza	uses	explicitly	and	carefully)	they	act	passively	and	their	autonomy	is			 	23	significantly	diminished.	Agents	who	are	influenced	by	emotions,	especially	those	that	are	not	fully	understood,	are	acting	passively.	Passive	action	occurs	when	an	agent	is	unaware	of	the	origins	of	her	actions,	and	since	freedom	for	Spinoza	is	active	judgment,	then	to	the	extent	that	one	is	unaware	of	the	origins	of	her	actions,	she	is	unfree.		Roger	Scruton	(2002)	summarizes	Spinoza’s	argument	nicely,	noting	that	for	him	“freedom	is	not	freedom	from	necessity,	but	the	consciousness	of	necessity”	(64).	Thus,	Spinoza	locates	an	agent’s	autonomy	in	her	capacity	to	know	the	origins	of	her	judgment.	In	so	doing	he	anticipates—perhaps	initiates—what	would	become	the	apotheosis	of	reason	and	rationality	in	modernity:	the	rise	of	Kantian	epistemology.	In	this	dissertation,	the	conceptions	of	autonomy	offered	by	Spinoza	and	Kant	will	play	an	important	role	in	how	I	define	autonomy	and	conceive	of	its	function	vis-à-vis	cognition	and	political	judgment;	these	conceptions	of	autonomy	underwrite	my	argument	that	good	political	judgments	require	extensive	and	accurate	knowledge	of	their	origins	in	order	to	guarantee	that	a	judgment	is	good,	so	that	it	can	be	communicated	to—and	in	some	way	held	in	common	with—others,	even	if	that	is	only	for	the	purposes	of	disagreeing	with	one	another.	Recall	that	Kant	(2012	[1785]),	for	all	his	revolutionary	reordering	of	our	understanding	of	the	nature	and	expression	of	human	epistemological	capacity,	argues	a	point	similar	to	Spinoza’s—a	point	that	is	at	the	centre	of	my	core	argument	in	this	dissertation.	Kant	makes	the	distinction	between	autarchy,	which	is	the	ability	to	make	choices	for	oneself,	whatever	the	motivation	one	may	have	(known	or	unknown),	and	autonomy,	which	requires	that	the	agent	can	rationally	reflect	upon	and	give	reasons	for	his	judgment.	In	conceiving	of	this	distinction,	Kant	implicitly	set	a	standard	for	judgment	that	entrenched	rationality	as	the	ultimate			 	24	guarantor	of	sound,	autonomous	judgment	(including	autonomy	as	necessary	for	communicating	reasons	for	that	judgment).	Judgment	for	Kant	requires	that	the	agent	engage	his	capacity	for	practical	reasoning	(cf.	Aristotle):	only	rational	reflection	within	the	constraints	of	the	macro-structures	of	human	cognition	and	voluntary	subjection	to	rational	laws	can	yield	autonomous	judgment	and	action;	if	such	reflection	and	subjection	are	absent,	an	individual	is	considered	to	be	acting	heteronomously	(e.g.	acting	from	desire	or	fear,	or	some	other	a-rational	determining	force).	Again,	as	in	Spinoza’s	conceptions	of	autonomy	and	judgment,	to	the	extent	that	an	individual	is	unaware	of	the	reasons	for	their	judgment,	or	unable	control	their	emotion	when	making	a	judgment,	their	autonomy	is	diminished,	if	not	eliminated	altogether.		Returning	to	deliberative	democracy,	the	tradition	of	theories	of	democratic	deliberation	have	remained	thoroughly,	if	not	universally,	Kantian.	Reasons	given	by	an	agent	are	expected	to	reflect	rational	considerations	(even	if	they	contain	some	emotional	element	or	are	presented	emotionally)	and	are	supposed	to	be	anchored	in	goals	and	preferences	about	which	the	agent	can	reason	and	communicate	in	the	hopes	of	generated	shared	understanding	(i.e.	validity)	through	communication	with	others.	Thus,	within	the	cognitive	processes	of	an	agent,	a	rational,	autonomous	link	is	presupposed:	stretching	and	connecting	from	rational,	autonomous	(personal)	reasoning	to	shared	(collective)	understanding.The	universal	presence	and	proper	functioning	of	this	link	in	agents	are,	however,	doubtful.	In	fact,	it	may	not	even	be	the	case	that	this	relationship	is	commonly	present	when	agents	deliberate,	as	I	will	argue	throughout	this	dissertation.				 	25	Perhaps	the	first	great	doubter	in	this	regard	was	Hume.	This	is	because	Hume	de-emphasizes	reason	as	a	capacity	with	significant	causal	force	and	denies	that	it	can	be	the	primary	source	of	either	morality	or	judgment;	instead,	he	introduces	affect	(“passions”)	into	the	equation.	He	defines	the	will	as	“…nothing	but	the	internal	impression	we	feel	and	are	conscious	of,	when	we	knowingly	give	rise	to	any	new	motion	of	our	body,	or	a	new	perception	of	our	mind”	(Hume	1975	[1738]:	399).	He	adds	that	“Reason	is,	and	ought	only	to	be	the	slave	of	the	passions,	and	can	never	pretend	to	any	other	office	than	to	serve	and	obey	them”	(Hume	1975	[1738]:	415).	As	Lara	Denis	(2012)	notes	regarding	the	motivational	roots	of	our	actions	and	judgments	(which	is	worth	quoting	at	length),	according	to	Hume		[a]bstract	(or	demonstrative)	reasoning,	which	involves	a	priori	inferences	and	judgments	about	relations	of	ideas,	cannot	influence	the	will,	but	only	assist	us	in	our	pursuit	of	an	end	we	already	have	(e.g.,	if	mathematical	calculations	would	facilitate	our	achievement	of	our	end).	Probable	(or	causal)	reasoning	helps	us	discover	cause	and	effect	relations	among	objects	of	experience	conducive	to	the	realization	of	pre-selected	ends,	but	such	information	about	cause	and	effect	can	never	motivate	action	on	its	own…In	order	to	be	motivated	to	act,	we	must	first	anticipate	pleasure	or	pain	from	something.	That	anticipated	pleasure	or	pain	gives	rise	to	feelings	of	desire	or	aversion	for	the	object	in	question.	Probable	reasoning	allows	us	to	discern	the	causes	of	this	object;	our	positive	or	negative	feelings	about	the	object	then	spread	to	the	causes	of	it;	and	we	are	then	motivated	to	pursue	or	to	avoid	them…	(no	page	reference)			 	26		This	conception	of	the	process	of	reasoning	includes	the	act	judgment	and,	accordingly,	if	it	is	accurate	as	an	account	of	human	cognition,	renders	elusive	the	kind	of	autonomous	judgment	required	for	deliberative	democracy	to	function	as	a	theory	of	how	epistemically	good	decisions	might	be	made.	Why?	Because	if	Hume	is	right,	the	requirement	that	public	reasons	be	given	in	concert	with	others	in	an	attempt	to	generate	preferences	and	reach	decisions	in	a	deliberation	is	potentially	reduced	to	the	mere	aggregation	of	affect-driven	desires	coated	with	the	gloss	of	ex	post	facto	(non-conscious)	rationalization.		 This	outcome	is,	of	course,	the	epistemically	worst-case	scenario	for	proponents	of	deliberation.	The	concerns	raised	about	our	theoretical	goals	as	deliberative	democrats	as	they	relate	to	epistemic	validity,	emerging	from	autonomous	judgment,	are	thoroughly	Kantian	and	are	based	upon	a	high	standard	of	judgment	and	decision	making.	Nonetheless,	doubts	about	the	extent	to	which	such	validity	is	possible,	emerging	from	work	in	social	and	political	psychology,	which	are	closer	to	the	Humean	tradition,	must	be	taken	seriously—though	they	should	not	be	overstated—if	we	are	to	produce	the	best	arguments	we	can	for	the	importance	of	deliberative	democracy.	To	put	this	dissertation	in	its	historical	place,	then,	it	may	be	fair	to	say,	that	it	is	a	contemporary	attempt	to	continue	the	debate	between	these	grand	ideas	about	autonomy,	rationality,	judgment,	and	decision-making	in	the	history	of	philosophy,	with	an	eye	to	both	grounding	and	extending	contemporary	debates	on	these	subjects.						 	27	Cognitive limitations, distortions, and bias in political contexts 	Notwithstanding	the	divergent	theoretical	positions	on	the	need	for	and	nature	of	autonomy	and	rationality,	observational	and	experimental	work	in	social	and	political	psychology	has	revealed	the	extent	to	which	human	judgment	and	the	capacities	of	autonomy	and	rationality	that	underwrite	them	can	be	compromised	in	political	(and	so	many	other)	contexts.	Kuklinski	and	Quirk	(2000)	summarize	the	problem	well:	cognitive	biases	and	distortions	have	been	hard-wired	into	the	human	brain	(e.g.	the	tendency	to	rapidly	categorize	stimuli	as	friendly	or	threatening	without	deeper	interrogation,	or	stereotyping)	through	a	long	history	of	evolution;	and	while,	as	the	authors	suggest,	such	distortions	are	likely	to	have	served	a	function	at	some	point	in	our	history	(and	perhaps	still	do	in	some	ways),	they	present	serious	challenges	to	contemporary	democratic	citizenship	and	the	normative	expectations	this	citizenship	includes.	As	they	conclude:	“researchers	should	not	presume	that	any	feature	of	human	cognition	is	well	adapted	to	the	tasks	of	citizen”	(Kuklinski	and	Quirk	2000:	165).	In	sum:	we	have	evolved	for	life	in	a	prior	evolutionary	age	and,	to	once	again	quote	Kuklinski	and	Quirk	(2000:	166)	in	their	eloquent	skepticism:	“…we	cannot	assume	that	the	cognitive	processes	people	use	in	making	a	particular	political	judgment	are	well	adapted	to	that	use.	To	the	contrary,	if	close	observers	see	such	processes	as	irrational	or	misleading,	they	probably	are.”	This	conclusion	points	to	the	argument	I	made	earlier—and	will	make	more	extensively	in	the	next	two	chapters—that	there	is	a	divergence	between	ordinary	individual	capacities	as	they	currently	tend	to	be	exercised	(nature)	and	what	is	nonetheless	expected	from	those	individuals.		 There	is	a	myriad	of	experimental	and	observational	data	that	support	this	scepticism.	As	early	as	the	1950s,	systematic	research	emerged	in	political	psychology	that	demonstrate			 	28	that	political	judgments	and	decisions—such	as	vote	choice	or	opinion/preference	formation	and	expression—were	at	least	partially	determined	by	factors	that	were	not	themselves	expressly	political	or	even	rational.	Berelson	et	al.	(1954)	provide	one	of	the	earliest	and	best	iterations	of	the	choice	model,	arguing	that	political	socialization—learned	primarily	from	the	family—helps	to	explain	the	continuity	of	voting	patterns	and	attitudes	throughout	time,	casting	vote	preference	as	a	learned	behaviour	that	is	transferred	between	social	groups	and	then	repeated	over	time,	like	a	habit.	By	the	1960s,	Converse	(1964)	had	emerged	to	present	an	even	more	dire	evaluation	of	(non-elite)	voters,	suggesting	that	they	tend	not	to	have	coherent	belief	systems	and	often	fail	to	evaluate	candidates	or	parties	on	the	left/right	political	spectrum,	though	this	claim	is	disputed	(see	Jost	et	al.	2006).	Voters,	Converse	concludes,	are	not	typically	ideological,	and	many	tend	to	change	their	mind	seemingly	at	random	on	certain	issues.	Reminiscent	of	this	position,	though	somewhat	more	sanguine	about	voter	competence,	Zaller	(1992)	argues	that	citizens	draw	on	competing	and	sometimes	conflicting	considerations	when	making	political	decisions	or	forming	opinions;	the	decision	that	is	ultimately	taken	by	an	individual	comprises	a	sampling	of	the	most	pertinent	(and	often	the	most	recent)	information	found	“at	the	top	of	their	head”	at	the	time	when	they	are	required	to	make	a	decision	or	pass	judgment.	In	Zaller’s	research,	what	emerges	is	an	understanding	of	the	voter	as	having	something	like	a	political	attitude,	though	it	remains	variable	and	unstable.		 Regarding	political	information	processing,	researchers	have	discovered	several	phenomena	that	call	into	question	the	extent	to	which	individuals	are	capable	of	rational,	autonomous	judgment.	For	instance,	framing	and	priming	effects	have	been	found	to	occur	in	subjects	who	process	political	information,	causing	them	to	change	the	weight	they	place	on			 	29	the	significance	of	information	based	on	the	frequency	and	order	of	mentions	of	the	subject	(Chong	and	Druckman	2007;	Iyengar	et	al.	1982).	However,	Lenz	(2009),	argues	that	the	phenomenon	is	in	fact	evidence	of	learning	and	not	manipulation	in	real-world	observational	findings.7	Regarding	elections,	Ballew	and	Todorov	(2007)	find	that	rapid	judgments	of	the	perceived	competence	of	candidates	in	gubernatorial	elections	predicted	outcomes	of	the	race	based	on	100ms	of	exposure	to	the	faces	of	the	candidates	(who	were	previously	unknown	to	the	subjects);	while	researchers	have	not	isolated	the	mechanism	at	work	in	these	predictions,	Ballew	and	Todorov	suggest	that	the	facial	expressions	of	the	candidates—demonstrating	(implied)	competence	or	not—may	play	a	role	in	determining	who	the	subject	predicts	as	the	winner.	Each	of	these	instances	is	a	challenge	to	contemporary	accounts	of	advanced	democratic	citizenship,	given	that	they	undermine	the	normative	expectation	of	sustained,	rational	reflection	and	autonomy	in	judgment	and	decision	making,	and	they	suggest	common	cases	in	which	a	more	rapid,	intuitive	judgment	is	at	work	(which	is	prone	to	error,	though	not	universally	so).8		Psychologists	have	pressed	matters	further	on	this	front,	examining	how	unsettling	stimuli,	including	being	read	a	surrealist	story,	being	subtly	reminded	of	death,	or	playing	cards	                                                7	Lab	findings	from	priming	experiments,	however,	cannot	be	explained	as	a	product	of	learning.	8	Priming	effects	have	been	shown	to,	generally,	have	a	temporally	limited	and	transient	effect.	However,	this	does	not	mean	that	priming	effects	are	insignificant.	For	one,	repeated	priming	that	leads	subjects	in	a	particular	direction	may	have	long	term	effects.	Also,	decisions	tend	to	be	made	by	individuals	under	conditions	in	which	they	are	primed	in	or	around	the	moment—e.g.	at	the	polls,	during	discussions,	while	consuming	news	and	forming	opinion.	The	accumulation	of	decision	outcomes	based	on	priming	in	some	particular	direction—e.g.	framing	a	war	as	a	battle	for	civilization	or	making	a	news	story	seem	more	salient	by	continued	coverage—may	produce	broader	outcomes	of	significance.	Thus	priming	may	be	significance	via	repetition	(at	the	site	of	the	individual)	or	accumulation	(across	individuals).				 	30	with	the	suit	colours	swapped,	affect	subjects’	behaviour.	In	laboratory	studies,	subjects	exposed	to	unsettling	stimuli	tended	to	non-consciously	employ	the	coping	mechanism	of	affirming	alternative	belief	systems	(and	overcompensating	by	behaving	in	more	extreme	ways	when	they	did).	Subjects	presented	with	discordant	stimuli	tended	to	be	more	likely	to	assert	a	“law	and	order”	belief	system,	becoming	more	likely	to	assign	more	significant	punishments	to	a	hypothetical	offender	than	subjects	who	were	not	presented	with	such	stimuli	(Heine	et	al.	2006;	Proulx	and	Heine	2006;	Proulx	et	al.	2010).	The	force	of	these	effects	seems	to	rely	on	non-conscious	motivation	via	implicit	messaging—whether	or	not	those	messages	are	expressly	intended	or	not.	Mendelberg	(2001)	has	shown	the	significant	effect	of	implicit	messaging	surrounding	race	politics	in	the	United	States,	finding	that	race-baiting	is	generally	effective	when	it	circumvents	established	race-norms	of	equality	among	white	voters	by	subconsciously	activating	feelings	of	resentment	and	negative	predispositions	towards	African	Americans	by	pairing	certain	stimuli	(e.g.	the	image	of	a	“frightening”	black	man	and	foreboding	music).		Not	only	do	implicit	messages	impact	judgments	and	decisions,	undermining	the	exercise	of	autonomy	and	rationality	in	the	process,	attempts	at	defending	such	judgments	may	be	compromised	by	further	cognitive	limitations	and	distortions,	thus	even	further	casting	doubt	on	the	possibility	of	autonomous	judgment.	As	Sniderman	et	al.	(1986)	demonstrate	in	their	study	of	Americans’	support	for	welfare	for	African	Americans,	the	reasoning	chains	of	some	voters	may	not	proceed	rationally	and	serially;	rather,	conclusions	justified	ex	post	facto	with	(rational)	reasons	may,	in	fact,	have	their	origins	in	affective	presuppositions,	prejudicial	or	otherwise	(e.g.	masquerading	racism	as	concerns	about	moral	hazard).	Such	rationalizations			 	31	deviate	from	the	standard	of	autonomous	judgment	expected	and	required	by,	among	others,	proponents	of	deliberative	democracy.	Many	of	these	findings	suggest	that	we	cannot	always	assume	the	existence	of	a	common,	sustained,	and	refined	capacity	for	citizens	to	express	rational,	autonomous	judgment	when	it	comes	to	political	matters—especially	when	that	requires	agents	to	know	the	sources	of	their	judgment	(or	behaviour).	Whether	those	sources	of	judgment	are	conscious	or	not,	political	or	not,	or	rational	or	not,	they	must	be	accounted	for	if	judgment	is	to	be	considered	autonomous	and	if	an	individual	is	to	arrive	at	a	generally	rational	judgment	based	on	reasons	they	can	share	with	others	to	generate	validity	in	the	context	of	democratic	deliberation.	While	past	standards	for	understanding	the	place	of	a-rational	cognition	and	the	impact	of	certain	types	of	stimuli	on	judgment	were	crude,	today	we	have	new	data,	methodologies,	and	a	much	more	sophisticated	and	empirically	supported	sense	of	many	(though	probably	not	all)	of	the	factors	that	impact	an	individual’s	judgment	and	a	group’s	decision.	Now,	if	we	wish	to	come	to	a	theory	of	deliberative	democracy,	defended	on	epistemic	grounds—grounded	in	the	capacity	for	autonomous	and	rational	exchanges	of	reasons	and	arguments	and	the	possibility	of	shared	understanding—then	we	must	take	up	the	ancient	question	of	what	are	the	sources	of	autonomy	and	rationality	when	it	comes	to	judgment	and	decision	making;	moreover,	we	must	adapt	this	question	and	its	implications	in	light	of	new	findings	about	the	nature	of	human	cognition	as	it	relates	to	democratic	deliberation—which	is,	in	part,	what	I	set	out	to	do	in	this	dissertation.		Some	scholars	of	political	psychology	have	already	begun	to	do	this.	In	their	exhaustive	study	of	how	voters	think,	Lodge	and	Taber	(2013)	argue	that	the	expectation	that	voters	(and			 	32	by	implication,	citizens	more	generally)	typically	turn	to	intentional	rational	evaluation	to	generate	deliberative	political	judgments	is	grossly	overstated.	As	they	claim,		[f]ar	more	common	[than	intentional	rational	evaluation],	we	believe,	will	be	the	reverse	causal	pathway	from	evaluation	to	deliberation.	This	rationalization	hypothesis…asserts	that	the	causal	pathways…that	travel	through	unconscious	affect,	and	in	particular	the	affect-driven	evaluation	processes,	cause	most	of	our	deliberation	about	politics….	Evidence	is	accumulating…that	attitudes	and	behavioral	intentions—even	behavior	itself—arise	from	automatic,	uncontrolled	processes	and	are	often	set	before	we	begin	seriously	“thinking”	about	them.	This	the	case,	deliberation	serves	to	rationalize	rather	than	cause	(21,	emphasis	in	original).		What	we	are	left	with,	then,	is	what	Lodge	and	Taber	refer	to	as	rationalizing,	rather	than	rational	citizens.	These	citizens	are	guided	by	affect-driven,	dual-process	modes	(I	address	dual-process	modes	in	depth	in	chapter	three)	of	thinking	and	reasoning.	They	tend	to	rely	on	the	second	stream	of	cognition	in	the	dual-process	model:	the	rapid,	automatic,	non-conscious,9	                                                9	Throughout	this	dissertation	I	will	use	the	word	“non-conscious”	rather	than	“unconscious”	wherever	possible—that	is	to	say,	whenever	I	am	not	directly	quoting	an	author	who	has	chosen	the	latter.	This	is	for	the	simple	reason	that	I	wish	to	avoid	any	association	with	common	conceptions	of	the	Freudian	unconscious	(hidden	but	vague	“drives,”	sexualized	motives,	the	tripartite	division	of	personality),	which,	while	broadly	relevant	to	the	subject,	is	merely	adjacent	to	the	project	at	hand.	While	I	will	discuss	drives	that	are	outside	of	awareness,	the	data	and	theories	I	draw	on	are	very	different,	and	more	recent	and	sophisticated	than	those	available	to	or	used	by	Freud.			 	33	affective	stream,	rather	than	the	slower,	deliberate,	and	rational	first	stream	(see	also	Kahneman	2011).	The	authors	summarize	nicely	the	emergent	doubts	about	citizens’	capacity	for	rational,	autonomous	judgment	in	their	conclusion	that	“affect	precedes	and	contextualizes	cognition”	(27).	I	will	revisit	this	conception	of	political	cognition	in	chapter	four.		Models from social and political psychology 	Past	and	recent	accounts	as	well	as	models	of	cognition	offered	by	students	of	social	and	political	psychology	tend	to	support	the	claims	made	by	scholars	such	as	Lodge,	Taber,	Kuklinski,	Quirk,	Sears,	and	Marcus,	that	non-conscious,	affect-driven	thinking	often	drives	citizen	judgment.	In	this	dissertation,	I	will	draw	primarily	on	four	models	of	cognition	developed	by	top	researchers	in	the	field	of	social	and	political	psychology	and	I	will	contextualize	with	reference	to	system	1	and	system	2	cognition	(Kahneman	2011).	The	system	1/system	2	division	is,	more	broadly,	a	dual-process	model	of	cognition	in	which	system	1	represents	fast,	automatic,	emotional,	non-conscious	cognition,	while	system	2	is	characterized	by	slow,	rational,	conscious	thought.	In	chapter	three	I	will	explain	which	elements	or	modes	of	each	model	I	use	falls	into	system	1	and	which	into	system	2,	and	under	what	circumstances;	and	I	will	link	each	back	to	rationality,	autonomy,	and	the	generation	of	epistemically	good	judgments	and	decisions	within	the	context	of	democratic	deliberation	by	highlighting	the	challenges	(and	potential	opportunities)	presented	by	each.	While	several	scholars	have	contributed	to	the	development	of	each	of	these	models,	I	will	refer	to	them	primarily	with	reference	to	their	originators	and	best-known	proponents,	though	I	will	also	engage	with	their	critics	where	it	is	useful	to	do	so.	The	models	are	Richard	E.			 	34	Petty	and	John	T.	Cacioppo’s	elaboration	likelihood	model,	John	T.	Jost’s	model	of	system	justification,	Jonathan	Haidt’s	social	intuitionism,	and	John	Bargh	and	Tanya	Chartrand’s	automaticity.	While	each	of	these	models,	broadly	stated,	offers	an	independent	account	of	the	nature	and	processes	of	particular	modes	of	cognition	as	they	relate	to	human	behaviour	and	judgment,	they	are	not	mutually	exclusive.	Indeed,	part	of	the	appeal	of	these	models	is	that	they	are	consistent	with	one	another	and	also	compatible	with	the	organizing	systems/dual-process	model	mentioned	above.	I	will	also	discuss	motivated	reasoning	on	its	own,	in	chapter	four—which	is	the	kind	of	model	used	by	Lodge	and	Taber,	noted	above.	Motivated	reasoning	has	been	given	its	own	chapter	because	it	stands	apart	from	the	other	models.	Why?	Because	motivated	reasoning	draws	more	on	explicit	motivational	considerations	than	the	other	models,	which	rely	more	heavily	on	cognitive	considerations.	Still,	there	is	overlap	between	motivated	reasoning	and	the	other	models,	which	I	will	note	and	discuss	when	appropriate.	I	have	chosen	to	explore	these	four	theories	for	five	reasons.	First,	they	enjoy	theoretical	compatibility	with	one	another;	second,	they	enjoy	the	status	of	leading	theories	in	the	fields	of	social	and	political	psychology;	third,	they	are	plausible,	cogent,	and	rely	on	rigorously	collected	and	theorized	data;	four,	they	are	intuitively	appealing	(the	irony	is	not	lost	on	me	here);	and	fifth,	they	correspond	with	emerging	theories	of	behaviour	and	judgment	within	political	psychology	and	the	data	supporting	those	theories.	Accordingly,	I	have	chosen	these	to	include	these	models	together	as	elements	of	a	broader	approach	to	understanding	judgment	and	behaviour	through	the	lens	of	non-conscious	political	cognition.	On	balance,	these	are	plausible	and	representative	models—and	the	particular	findings	I	will	use	to	explore	and	support	them	are	just	as	plausible	and	representative.	Moreover,	these	models	are	capable			 	35	of	coherently	absorbing	and	accounting	for	data	from	not	only	social	psychology	but	also	political	psychology	and	cognitive	neuroscience.	The	macro-level	coherence,	representativeness,	and	cogency	of	these	models	is	what,	ultimately,	renders	them	useful	and	very	well	fitted	for	this	study	of	rationality,	autonomy,	and	cognition	as	they	related	to	epistemically	good	(or	not)	democratic	deliberation	in	liberal	democracies.		 I	will	examine	these	models	in	chapters	three	and	four,	so	I	will	only	briefly	sketch	each	of	them	out	here.	First,	Petty	and	Cacioppo’s	elaboration	likelihood	model	suggests	that	attitude	change	occurs	through	one	of	two	cognitive	processing	routes—central	(i.e.	system	2)	or	peripheral	(i.e.	system	1).	Second,	Jost’s	model	of	system	justification	asserts	that	certain	psychological	motives	and	processes—for	instance,	the	desire	for	stability	or	the	fear	of	death—are	causally	linked	to	ideological	predispositions	(see	Jost	2006;	Jost	and	Amodio	2012;	Jost	et	al.	2003).	Third,	Haidt’s	social	intuitionist	model	suggests	that	personal	moral	reasoning	does	not	cause	moral	judgment.	Instead,	such	judgments	emerge	from	interpersonal	processes	and	are	perceived,	rather	than	rationally	discovered.	The	moral	reasoning	that	gives	ideational	foundation	to	these	judgments	appears	only	after	the	fact	(Haidt	2001).	Finally,	Bargh	and	Chartrand’s	model	of	automaticity	suggests	that	our	capacity	for	intentional	and	rational	judgment	is	limited	by	automatic	perceptual	direction	taken	from	the	external	environment,	which	directs	behaviour	and	judgment	(Bargh	and	Chartrand	1999).	Moreover,	action	precedes	reflection;	both	external	stimuli	and	internal	mental	processes	contributing	to	behaviour	and	judgment	precede	rational	reflection	upon—or	even	apprehension	of—behavioural	outcomes	and	judgments;	tendencies	towards	certain	sorts	of	actions	in	these	areas	develop	over	time,	creating	general,	automatic,	and	non-conscious	patterns	of	behaviour	and	judgment	in	citizens.			 	36	Conscious	apprehension	of	such	activity	emerges	later,	acting	as	a	“gatekeeper	and	sense	maker”	(Bargh	and	Morsella	2008)—and	what,	as	I	will	argue,	can	be	characterized	as	rationalizations	or	pre-deliberative	motivations	that	become	expressed	as	preferences,	reasons,	and	judgments	when	individuals	deliberate.		Chapter outline 	As	I	have	noted,	the	primary	question	I	address	in	this	dissertation	is:	How	do	a-rational	cognitive	processes	affect	of	the	possibility	of	rational,	autonomous	judgment	and	the	production	of	epistemic	validity	in	deliberative	settings?	The	secondary	question	that	I	address	is:	How	can	political	theory	and	political	science	respond	to	the	effects	of	such	processes	and	stimuli	in	a	way	that	underwrites	capacities	of	rationality	and	autonomy,	and	the	production	of	good	judgments	and	decisions	from	democratic	deliberation?	These	questions	emerge	in	response	to	an	attempt	to	bridge	the	gap	between	what	is	normatively	expected	of	those	who	deliberate	in	a	liberal	democracy,	and	what	they	tend	to	be	able	to	deliver.	To	address	these	questions	in	a	way	that	relates	them	to	one	another,	this	dissertation	is	separated	into	two	parts.	The	first	part	(chapters	two,	three,	and	four)	makes	the	argument	that	democratic	deliberation	requires	a	relatively	high-level	capacity	rationality	and	autonomy,	and	that	certain	a-rational	cognitive	processes	currently	pose	a	serious	and	persistent	threat	to	this	requirement	by	undermining	the	human	capacity	to	give	reasons	for	judgments	that	reflect	real	motivations,	eroding	the	capacity	for	deliberation	to	generate	epistemically	good	judgments	and	decisions.			 	37	To	this	end,	in	chapter	two	I	argue	that	autonomy	is	the	keystone	capacity	required	for	deliberative	democracy	to	function	as	an	approach	to	generating	epistemically	valid	and	authoritative	judgments	and	decisions	among	citizens.	Recall	that	the	focus	on	autonomy	over	rationality	is	that	when	the	former	is	present	to	a	high	degree,	the	products	of	the	latter	can	be	better	interrogated	and	understood.	Deliberative	democracy	is	founded	on	reason-giving	and	the	force	of	argument	as	a	way	of	reaching	legitimate	and	just	political	outcomes;	for	these	to	be	achieved,	the	sources	of	judgments	must	be	known	and	agents	must	have	confidence	that	the	facts,	norms,	motivations,	and	desires	that	they	adhere	to	or	express	are	not	the	products	of	cognitive	distortion	or	bias.	Otherwise,	the	moral	grounding	of	the	theory	of	deliberation	as	one	based	on	good	faith	attempts	to	approach	decisions	based	on	the	force	of	reasons	and	the	better	argument	is	eroded	by	the	influential	presence	of	possibly	irrelevant	determinants	and	important	factors	that	might	remain	outside	the	realm	of	interrogation	and	the	democratic	goods	produced	by	epistemically	valid	reasons	may	be	undermined	or	lost.	Moreover,	I	argue,	building	on	Kant’s	distinction	between	autarchy	and	autonomy,	that	to	satisfy	the	requirements	for	a	judgment	to	be	autonomous,	not	only	must	an	individual	be	able	to	give	reasons	for	that	judgment,	those	reasons	must	closely	reflect	their	actual	motivations	for	holding	them.	That	requirement	is	meant	to	ensure	that	reasons	are	not	generated	heteronomously—or,	to	be	blunt:	their	reasons	must	be	accurate	and	their	own.	This	requirement	aligns	closely	with	Christman’s	conception	of	autonomy,	the	definition	of	that	concept	that	I	will	use	in	this	dissertation	and	will	explore	in	depth	in	chapter	two.	According	to	Christman,	autonomy	is	a	capacity	in	which	“…the	influences	and	conditions	that	give	rise	to	the	desire	[or	preference	or	intention]	were	factors	that	the	agent	approved	of	or	did	not	resist,	or	would	not	have	resisted			 	38	had	she	attended	to	them,	and	that	this	judgment	was	or	would	have	been	made	in	a	minimally	rational,	non-self-deceived	manner”	(22,	emphasis	mine).		In	chapter	three	I	argue	that	the	capacity	of	autonomy	required	for	deliberative	democracy,	defended	in	chapter	two,	is	undermined	in	some	cases	by	certain	a-rational	cognitive	processes,	and	that	these	processes	also	undermine	capacities	for	rational	judgment.	I	frame	the	discussion	as	being	rooted	in	a	disconnect	between	what	we	expect	and	require	in	our	culture	and	what	we	tend	to	be	able	to	do	to	given	our	evolutionary	cognitive	limitations.	By	culture	I	mean	the	social,	political,	and	technological	institutions	that	we	have	developed	(including	their	normative	content,	such	as	ideas	about	how	we	ought	to	organize	our	institutions	and	how	we	ought	to	behave	towards	one	another);	by	nature,	I	mean	the	central	tendencies	and	limits	of	our	biological	and	psychological	capacities	as	they	have	developed	over	the	course	of	our	time	as	a	species.	I	then	present	findings	from	social	and	political	psychology	and	cognitive	science	that	demonstrate	the	ways	in	which	rational,	autonomous	judgment	is	susceptible	to	influence—or	even	determination—by	apolitical	forces	or	phenomena	that	have	traditionally	been	considered	irrelevant	to	arriving	at	political	judgments	in	deliberative	settings.	Specifically,	as	noted	above,	I	use	four	generally	accepted	models	of	human	cognition	organized	under	system	1/system	2	distinctions:	Petty	and	Cacioppo’s	elaboration	likelihood	model,	Jost’s	system	justification,	Haidt’s	social	intuitionism,	and	Bargh	and	Chartrand’s	automaticity.	I	explain	why	I	have	chosen	these	models	and	how	the	threats	to	autonomous	judgment	they	raise	undermine	normative	arguments	in	favour	of	deliberative	democracy	as	both	a	moral	theory	founded	on	rational,	autonomous	reason	giving	and	the	force	of	the	better			 	39	argument,	and	as	an	approach	to	generating	epistemically	good	or	correct	judgments	and	decisions.	I	also	argue	that	given	the	strength	of	these	threats	to	deliberative	democracy	as	a	political	process	for	reaching	legitimate	and	just	decisions	based	on	such	epistemically	valid	and	authoritative	facts,	norms,	and	preferences,	and	given	our	current	conception	of	rationality	and	autonomy	as	they	relate	to	judgment	in	deliberative	settings,	theorists	of	deliberative	democracy	who	wish	to	address	this	challenge	are	presented	with	four	general	approaches	in	response:	ignore,	revise,	reject,	or	respond.	Scholars	of	deliberation	can	ignore	the	problem	altogether	and	accept	the	potentially	flawed	nature	of	the	judgments	and	decisions	that	follow.	They	can	revise	deliberative	democracy	as	a	moral	theory,	altering	its	moral	foundation	(i.e.	honest	reason	giving	and	the	force	of	the	better	argument)	or	the	requirements	for	what	counts	as	a	good	reason	or	a	good	argument.	They	can	reject	deliberative	democracy	entirely	and	adopt	a	more	explicitly,	traditionally	“political”	(i.e.	realist,	aggregative)	approach	to	arriving	at	political	decisions;	this	substitute	approach	would	need	to	be	one	in	which	concerns	about	the	particular	motivations	and	determinants	of	judgment	were	made	secondary	to	(or	perhaps	eliminated	entirely	from)	outcomes.	Or	they	can,	as	I	argue	they	should,	seek	to	develop	approaches	towards	designing	institutional	set-ups	and	personal	practices	to	be	deployed	within	a	deliberative	system	aimed	at	leveraging	our	cognitive	capacities	in	such	a	way	that	we	can	reach	rational,	autonomous	judgments	through	deliberation.	This	approach	requires	an	account	of	what	counts	as	rational,	autonomous	judgment	in	such	settings	so	that	we	may	know	if	the	requisite	standards	are	met;	it	also	requires	a	statement	of	the	conceptual	and	practical	apparatuses	needed	to	meet	such	a	standard—each			 	40	of	which	I	cover	in	part	two	of	this	dissertation.	While	this	may	seem	like	a	revision	to	the	theory	of	what	counts	as	a	good	decision	or	judgment,	it	is	not;	the	requirement	of	autonomy	offered	in	accounts	of	deliberative	democracy	is	maintained,	as	are	the	guiding	ideal	requirements	that	reasons	are	given	and	that	the	force	of	the	better	argument	carries	the	day.	Instead,	this	approach	adds	a	new	evaluative	criterion	(i.e.	that	those	who	arrive	at	judgments	are	aware	of	all	the	relevant	factors	that	determine	those	judgments)	and	suggests	ways	that	this	standard	might	be	met	by	those	who	deliberate.		In	chapter	four	I	examine	the	cognitive	phenomenon	of	motivated	reasoning—primarily,	though	not	exclusively,	through	Lodge	and	Taber’s	model—as	a	specific	motivational	threat	to	political	judgments	and	autonomous	deliberation,	linking	that	model	back	to	the	four	general	models	explored	in	chapter	three	and	examining	how	it	relates	to	deliberative	democracy.	Motivated	reasoning	is	a	phenomenon	consistent	with	each	of	the	four	models,	especially	when	such	reasoning	occurs	under	the	system	1	mode	of	cognition	(though	it	is	also	compatible	with	system	2).	Additionally,	in	this	chapter	I	define	and	examine	the	related	concepts	of	hot	cognition	and	reasoning	chains	and	link	them	to	democratic	deliberation.	I	explore	critiques	of	the	models	I	discuss.	In	the	second	part	of	this	dissertation	(chapters	five	to	seven),	I	present	a	general	approach	to	rethinking	autonomy	in	deliberative	settings	as	well	as	specific	responses	to	the	challenge	of	a-rational	cognition	to	rationality	and	autonomy.	I	suggest	that	to	address	the	challenges	I	raise,	we	need	to	both	add	new	tools	to	the	deliberative	democracy	toolkit	and	mobilize	and	adapt	existing	tools.	I	also	discuss	the	specific	steps	required	to	move	from			 	41	democratic	deliberation	as	it	currently	is	practiced	to	a	sort	of	democratic	deliberation	that	can	specifically	respond	to	the	challenges	outlined	in	this	dissertation.	In	chapter	five	I	ask	whether	we	can	design	deliberative	practices	and	institutional	arrangements	in	such	a	way	that	underwrites	and	enhances	our	capacities	for	rationality	autonomy.	I	introduce	and	develop	the	concept	of	institutional	deliberative	autonomy	(IDA),	into	which	I	fold	concerns	for	both	capacities	of	rationality	and	autonomy.	IDA	is	a	concept	developed	for	thinking	through	whether	rational,	autonomous	judgments	are	reached	in	deliberative	settings.	Accordingly,	IDA	includes	an	account	of	what	counts	as	rational,	autonomous	judgment	in	deliberative	settings—an	account	that	is	consistent	with	a	robust	theory	of	autonomy.	This	concept	is	not	merely	an	evaluative	concept,	but	also	a	regulative	ideal	meant	to	guide	deliberative	democratic	institutional	designs	and	evaluations	of	participants	and	outcomes.	IDA	does	not	track	the	status	of	a	binary	state	(rational/autonomous	or	not),	but	rather	within	it	I	imagine	a	scale	for	autonomy	that	ranges	from	maximal	autonomy	(i.e.	the	absence	of	any	determining	factors	save	for	rational	judgment—a	theoretical	high-point	that	is	neither	achievable,	nor	necessarily	desirable)	to	total	automaticity	(i.e.	the	full	determination	of	one’s	judgment	by	factors	entirely	outside	of	one’s	awareness).		In	this	chapter	I	also	introduce	and	interrogate	four	concepts	for	thinking	about	autonomous,	rational	deliberation	might	be	brought	about.	The	first	concept	is	iteration:	repeated	discussions	or	exercises,	similarly	structured	over	the	course	of	a	period,	and	designed	to	develop	a	stable	central	tendency	and	to	wash	out	individual	instances	of	random	cognitive			 	42	distortions.	This	approach,	however,	does	not	apply	to	structural	distortions;	for	that,	other	concepts	are	required.	To	that	end,	the	second	concept	I	introduce	is	a-rational	receptivity,	which	refers	to	personal	practices	aimed	at	attempting	to,	as	best	as	possible,	bring	to	the	surface	hidden,	partial,	or	manifest	moods,	feelings,	emotions,	motivations,	fears,	anxieties,	and	so	forth	during	deliberation.	More	specifically,	a-rational	receptivity	refers	to	a	group-wide	disposition	aimed	at	encouraging	the	exploration	of	an	individual’s	a-rational	state.	Such	an	element	is	also	an	institutional	feature	because	it	is	built	into	the	structure	of	the	deliberation	and	employed	as	necessary;	however,	its	operating	mechanisms	are	individual	practices	of	self-interrogation	combined	with	group-wide	receptivity	to	such	interrogations.	Returning	to	Kant:	autonomy	requires	that	individuals	have	both	judgments	and	reasons	for	them;	I	hasten	to	add	that	autonomy	requires	that	individuals	have	reasons	for	judgments,	which	would	include	the	emotions,	feelings,	and	moods	that	play	a	role	in	generating	and	underwriting	those	judgments.10		The	third	concept	I	look	at	is	cognitive	diversity.11	This	concept	reflects	an	attempt	to	prevent	individuals	from	coasting	through	deliberations	by	relying	merely	or	mostly	on	automatic,	low-level	or	non-conscious	heuristics	or	gut-level	guidance.	While	these	automatic	                                                10	Though	it	is	presently	outside	of	the	scope	of	this	dissertation,	a-rational	receptivity	may	be	a	concept	useful	for	bridging	the	gap	between	theorists	of	deliberation	who	rely	on	traditional	understandings	of	autonomy	and	reasoning,	and	theorists	of	difference	who	are	critical	of	the	potential	exclusivity	and	limits	of	this	approach	(see	Benhabib	1992,	1996;	Bohman	and	Rehg	1997;	Young	1990).	11	Cognitive	diversity	does	not	necessarily	require	a	diversity	of	identities,	only	a	diversity	of	approaches	towards	cognition	and	its	related	practices.	That	said,	to	the	extent	that	differences	in	identity	correlate	with	cognitive	diversity,	the	two	can	be	thought	of	as	bound	up	together,	though	this	is	not	logically	necessary.			 	43	processes	may	often	be	useful	for	navigating	a	complex	and	variable	world,	the	kind	of	attention	and	reflection	that	deliberation	requires	is	usually	poorly	served	by	this	approach.	Cognitive	diversity	may	hold	part	of	the	answer	to	getting	subjects	out	of	this	state.	As	defined	by	Landemore	(2013),	building	from	Hong	and	Page’s	(2004)	conception,	cognitive	diversity	“refers	to	a	diversity	of	ways	of	seeing	the	world,	interpreting	problems	in	it,	and	working	out	solutions	to	these	problems.	The	concept	denotes	more	specifically	a	diversity	of	perspectives…interpretations…heuristics…and	predictive	models”	(see	also	Page	2007).	In	deliberative	contexts,	the	presence	of	a	variety	of	individuals	with	different	cognitive	styles	might	not	only	improve	decision	making	(Page	2007)	and	the	quality	of	deliberation	(Landemore	2013),	but	that	presence	of	diverse	ways	of	thinking	may	also	offer	a	cognitive	jolt	to	individuals	who	might	otherwise	rely	heavily	on	the	low-resource	flow	of	automaticity	when	processing	information	and	reaching	judgments.	My	fourth	and	final	concept	is	targeted	motivation.	This	practice	indicates	an	attempt	to	engage	participants	in	central-processing	(i.e.	deliberate,	rational	reflection	rather	than	low-level,	automatic	processing)	by	highlighting	the	relevance	and	importance	of	the	issue	or	issues	at	hand.	Targeted	motivation	may	assist	in	shifting	subjects’	attention	towards	the	subject	matter,	thus	maximizing	the	likelihood	that	they	will	scrutinize	the	data	and	arguments	presented	to	them.	Such	scrutiny,	in	theory,	would	open	individuals	up	to	iteration	and	a-rational	receptivity,	as	well	as	traditional	approaches	reasoning,	by	raising	the	cognitive	stakes	of	the	deliberation.	Ultimately,	targeted	motivation	alone	may	not	work	on	all	deeply	embedded	processes	linked	to	cognitive	distortion;	it	may,	however,	minimize	the	impact	of			 	44	some	of	the	superficial	effects	of	automaticity	and	may	enable	deeper	tactics,	such	as	a-rational	receptivity,	to	work.	In	chapter	six	I	move	away	slightly	from	the	individual	in	the	context	of	democratic	deliberation	and	look	at	the	broader	social	and	political	institutions	that	frame	democratic	deliberation.	I	compare	different	understandings	of	what	an	institution	is	and	examine	how	both	such	conceptions	of	institutions	and	the	institutions	themselves	interact	with	those	who	deliberate.	I	also	ask	how	these	institutions	might	be	understood	and	reformed	in	a	way	that	enhances	democratic	deliberation.	I	look	specifically	at	the	media	and	attitude	formation,	political	parties	and	partisanship,	electoral	systems	and	elections,	and	ideology	and	hegemony.	My	primary	focus	in	that	chapter	is	how	institutions	often	come	to	encourage	poor	thinking,	judgment,	and	decision	making.	This	focus	sets	up	chapter	seven—which	is	an	examination	of	deliberative	systems.	In	chapter	seven	I	discuss	how	an	understanding	of	deliberative	systems	might	help	us	to	respond	to	the	challenges	of	both	individually	limited	and	distorted	cognition	and	institutional	arrangements	that	undermine	our	best	efforts	at	producing	epistemically	good	democratic	deliberation.	I	argue	that	good	democratic	deliberative	systems	design—along	with	the	aforementioned	institutional	reform	and	changes	to	individual	practices—can	help	generate	better	deliberation	despite	challenges	from	the	sorts	of	cognitive	distortion	explained	by	the	four	models	I	examine	in	chapter	three	and	motivated	reasoning.	This	chapter	pulls	together	several	threads:	from	the	individual	and	social	and	political	institutions,	up	to	the	level	of	a	democratic	deliberative	system	that	goes	beyond	mere	one-off	deliberative	events	and			 	45	entrenches	deliberative	activities	as	complementary	approaches	to	democratic	judgment	and	decision	making	in	a	political	system.	In	chapter	eight	I	conclude	the	dissertation	by	recapitulating	how	each	of	the	arguments	raised	in	the	dissertation	connects	to	one	another	and	examining	the	problem	of	cognitive	distortion	as	it	relates	to	undermining	the	rationality	and	autonomy	required	for	reaching	epistemic	validity	in	deliberative	contexts.	I	also	revisit	the	conceptual	approaches	I	have	offered	in	response	to	these	challenges,	further	arguing	that	these	approaches	require	testing	in	real-world	deliberative	contexts	to	determine	whether	they	will	be	able	to	address	the	challenges	to	rationality	and	autonomy	discussed	throughout	the	dissertation.	I	end	with	a	discussion	of	how	the	cognitive	sciences	might	be	further	integrated	with	political	science	and	political	theory	in	the	future.	   		 	46	Chapter 2: The Epistemic argument for democratic deliberation and the need for autonomy 	Liberal	democracy	refers	to	a	political	system	predicated	on,	among	other	foundational	elements,	the	related	premises	that	individuals	are	the	best	judges	of	their	own	interests	and	their	own	preferences	(Dahl	1989;	Raz	1986).	Consequently,	it	follows	that	those	within	such	an	association	ought	to	be	able	to	choose	freely,	at	least	in	a	general	sense,	how	that	association	operates	(Habermas	1996;	Held	1995).	Given	these	premises,	it	is	important	to	carefully	determine	by	what	means	members	of	such	a	body	should	make	decisions	that	will	have	an	impact	on	their	lives	and	the	lives	of	others.	If	each	citizen	were	an	island	unto	herself,	then	it	would	scarcely	matter	to	anyone	else	what	that	individual	wanted	or	why	she	wanted	it;	nor	would	it	matter	what	she	did.	However,	liberal	democracies	are	populated	by	individuals	whose	interests	and	preferences	diverge	from—and	often	conflict	with—one	another	and	within	which	the	interests	and	preferences	often	affect	the	lives	of	others.	These	divergences	must	be	addressed	if	legitimacy,	order,	good	government,	and	just	policies	and	laws	are	to	be	established	and	maintained.	Accordingly,	it	is	necessary	to	ask	precisely	how	this	is	to	be	done	and	to	develop	a	procedure	(or	procedures)12	For	decision	making	that	is	both	fair	(i.e.	acceptable	to	reasonable	participants	and	impartial	towards	all	who	deliberate)	and	effective	at	                                                12	To	say	a	bit	more,	for	this	dissertation	I	conceive	of	fairness	in	limited	terms	that	are	specifically	related	to	democratic	deliberation.	For	my	purposes,	deliberation	is	fair	if	the	terms	of	the	deliberation	are	i)	generally	acceptable	to	all	those	included	or	affected;	ii)	those	terms	are	followed	during	deliberations;	and	iii)	the	deliberation—notwithstanding	whatever	other	terms	have	been	set—is	impartial	among	those	who	participate	in	such	a	way	that	no	single	individual	or	their	perspective	is	a	priori	privileged	over	and	above	that	of	another.	Of	course,	during	the	deliberation	an	individual	or	the	collected	body	may	choose	to	favour	a	perspective,	but	this	should	not	be	taken	as	granted	before	deliberation	has	begun	nor	should	it	be	chosen	for	individuals	or	the	collected	body	externally.			 	47	producing	outcomes	that	link	individual	and	collective	interests	and	preferences	with	outcomes	in	the	formal	political	sphere.		 Deliberative	democrats	advocate	a	participatory	system	in	which	citizens	(or	their	representatives)	exchange	reasons	with	one	another	to	generate	political	judgments	or	decisions13	that	may	be	used	to	generate	recommendations,	laws,	or	policies	that	will	be	enacted	by	decision	makers	or	put	to	the	broader	population	for	approval.	Rather	than	suggesting	that	decisions	are	best	made	by	merely	aggregating	pre-existing	preferences—a	blunt,	if	standard,	majoritarian	approach—deliberative	democrats	argue	that	the	deliberative	approach	to	democracy	yields	more	legitimate,	just,	and	valid	outcomes	(Warren	2002).	Theories	of	deliberative	democracy	assume	that	those	who	participate	in	judgment	and	decision	making	are	autonomous	(i.e.	they	are	the	originator	of	their	actions	and	preferences,	and	can	give	an	account	of	why	they	act	in	such	a	way	or	hold	such	preferences)	and	rational	(i.e.	they	can	more-or-less	accurately	discern	facts	about	the	world	in	a	consistent	way,	that	they	can	communicate	them	to	others	effectively,	and	that	they	are	open	to	being	persuaded	by	reasons).	However,	these	standards	for	deliberation,	especially	the	standard	of	autonomy,	leave	a	gap	between	expectations	about	one’s	ability	to	establish	natural	and	normative	facts	about	the	world,	and	the	extent	to	which,	in	practice,	individuals	tend	to	be	fully	aware	of	why	and	how	they	arrive	at	a	judgment	(and	how	those	collective	judgments	generate	decisions).	My	interest	in	this	dissertation	is	in	the	assumption	underlying	the	requirements	of	autonomy	in	reason	giving,	which	should	also	be	taken	to	imply	an	extended	sense	of	self-                                                13	Recall	that	for	the	purposes	of	this	dissertation,	"judgment"	refers	to	individual	conclusions	arrived	at	after	a	period	of	(perhaps	very	brief)	consideration,	while	"decision"	refers	to	a	binding	or	semi-binding	conclusion	reached	by	a	particular	group	(e.g.	a	deliberative	body).			 	48	ownership	and	self-direction:	namely,	that	one	is	mostly	aware	of	her	motivations	for	holding	or	reaching	her	interests	and	preferences,	that	she	is	also	conscious	of	the	reasons	that	underwrite	them,	and	that	she	can	give	a	more-or-less	accurate	account	of	them	(Christman	1991:	11).	These	requirements	build	on	the	assumption	that	with	a	rational	agent,	there	is	a	connection	between	one’s	motivations,	her	reasons,	and,	finally,	her	judgments	and	the	subsequent	(collective)	decisions	to	which	they	contribute.	The	requirement	of	having	reasons	and,	in	an	extended	sense,	reasons	for	one’s	reasons,	builds	off	and	extends	Kant’s	distinction	between	autarchy,	which	is	the	ability	of	an	individual	to	make	her	own	choices,	and	autonomy,	which	requires	that	a	chooser	is	able	to	state	reasons	for	these	choices	that	reflect	their	motivations	(Elstub	2008).	It	is	the	addition	of	this	latter	requirement—that	one	is	aware	of	and	able	to	state	their	motivations	for	their	reasons—that	complicates	matters	since,	as	we	will	see	briefly	in	this	chapter	and	more	fully	in	chapters	three	and	four,	motivations	can	quite	quickly	become	complicated	and	obscured	from	an	individual’s	own	apprehension	(making	it	complicated	to	communicate	their	reasons	for	having	their	reasons	to	another).		 The	primary	goal	of	this	chapter	is	to	defend	deliberative	democracy	as	both	a	fair	democratic	procedure	(again,	in	the	thin	sense	of	it	being	a	procedure	that	impartial	among	participants,	and	a	procedure	that	is	acceptable	to	them)	and	as	an	epistemic	theory	of	good	and	valid	democratic	judgment	and	decision	making	(including	the	requirement	that	its	participants	be	autonomous	in	the	extended	sense	just	mentioned).	To	this	end,	I	will	argue	that	a	capacity	for	autonomy,	including	the	ability	to	state	one’s	motivations	for	one’s	reasons,	is	the	keystone	requirement	for	a	theory	of	deliberative	democracy	to	hold	sway	as	a	cogent	theory	of	better	democratic	decision	making.	This	claim	relates	specifically	to	the	epistemic			 	49	claim	made	by	Warren	(2002)	that	deliberative	democracy	is	more	epistemically	valid	in	the	context	of	a	liberal	democracy	(i.e.	its	decision-making	procedures	produce	better	decisions	than	alternative	approaches)	than	aggregative	democracy	or	other	forms	of	decision	making	(e.g.	elite-driven,	technocratic	procedures)	and	also	Estlund’s	(2008)	argument	that	democracy,	in	general,	has	more	epistemic	value	as	a	procedure	than	other	forms	of	decision	making.			 Recall	that	I	am	focusing	on	autonomy	over	rationality	here	since	to	the	extent	that	autonomy	is	present,	an	individual	is	theoretically	capable	of	giving	an	accurate	and	reliable	account	of	why	they	have	come	to	some	judgment	(i.e.	a	consistent	account	that	reflects	the	reasons	that	actually	motivate	an	individual).	If	that	judgment	is	irrational,	it	can	be	interrogated,	countered,	and	corrected.	But	autonomy	is	an	essential	requirement	for	deliberation	since	it	enables	an	individual	to	present	accurate	reasons	for	his	judgment,	which	then	allows	those	assembled	to	engage	with	it.	One	could,	in	theory,	give	rational	explanations	that	were	not	accurate	accounts	of	their	actual	motivations,	reasons,	etc.	Autonomy,	when	exercised,	implies	that	the	account	given	is	a	more-or-less	accurate	one.14			 The	epistemic	dimension	of	deliberative	democracy	requires	a	bit	more	elaboration	here,	specifically	when	it	comes	to	the	statement	that	it	generates	“better”	decisions.	By	“better”	decisions,	I	refer	to	Warren’s	claim	that	deliberation	tends	to	generate	decisions	based	on	statements	that	are	“authoritative	in	relation	to	[their]	referent:	facts,	norms	or	subjective	desires	and	experiences”	(Warren	2002:	192).	This	point	is	essential.	In	a	liberal	democracy,	asserting	an	ontological	truth	is	not	enough	to	justify	an	action	because	the	foundation	of	                                                14	In	this	context	I	am	referring	to	“accuracy”	being	related	to	the	specific	causes	of	something,	rather	than	some	(conscious	or	not)	rationalization.			 	50	democracy	is	(broadly)	built	upon	the	requirement	that	all	those	affected	by	decisions	should	have	some	(general)	control	over	how	those	decisions	are	made	and	by	whom.	Accordingly,	Warren’s	point	that	deliberation	generates	better	knowledge	of	a	statement’s	validity	(and,	subsequently,	the	reasons	for	and	justifications	of	a	decision)	rests	on	the	(accurate)	claim	that	“We	gain	such	knowledge,	or	confidence	in	the	validity	of	the	statement,	through	processes	of	challenge,	reason-giving,	and	verification”	(Warren	2002:	192).	These	processes—undertaken	in	a	deliberation	as	an	intersubjective	effort	aimed	at	generating,	among	other	things,	shared	understanding—are	inherent	in	deliberative	democracy.	Moreover,	they	are	better	achieved	and	exhibited	through	deliberation	than	other	means	of	judgment	and	decision	making,	especially	in	the	context	of	liberal	democracy.	To	put	it	simply:	deliberation	fulfills	at	least	three	broad	essential	public	goods	simultaneously:	an	epistemic	good	(i.e.	establishing	the	validity	of	statements),	a	political	(or	democratic)	good	(i.e.	providing	grounds	for	the	generation	of	public	and	legitimate	decisions	and	including	more	individuals	in	decision	making),	and	an	ethical	good	(i.e.	recognizing	the	moral	status	of	those	affected	by	a	decision	through	reason	giving	and	reciprocity).	With	this	established,	in	the	following	pages,	drawing	on	Warren,	I	explore	each	of	the	specific	ways	deliberative	democracy	is	epistemically	superior	(since	this	is	the	focus	of	my	project)	to	aggregative	approaches	to	democracy	to	establish	deliberation	as	an	approach	to	decision	making	more	consistent	with	autonomy.	I	then	present	and	defend	Estlund’s	(2008)	theory	of	epistemic	proceduralism	as	the	strongest	and	most	defensible	framework	of	deliberative	democracy	as	an	epistemic	theory	of	collective	decision	making	and	judgment,	since	it	provides	the	scaffolding	for	understanding	how	deliberation,	autonomy,	and			 	51	epistemology	intersect.15	I	also	address	Elstub’s	(2008)	critique	of	epistemic	defenses	of	deliberative	democracy:	a	critique	that	is	grounded	in	the	claim	that	deliberative	democracy	cannot	generate	true	beliefs	about	the	common	good,	since	if	this	is	true,	then	the	value	of	deliberation	as	an	approach	to	autonomously	address	moral	issues	in	politics	would	be	undermined.	I	then	examine	how	autonomy,	as	it	is	conceived	by	Christman	(1991),	links	to	deliberative	democracy	as	an	essential,	keystone	capacity	for	epistemically-valid	decision	making.	I	then	explain	why	in	order	to	be	autonomous	individuals	must	not	be	self-deceived.	I	finish	with	a	short	note	on	why	the	epistemological	issue	of	autonomy	and	heteronomy	in	decision	making	is	separate	from	the	broader	ontological	issue	of	free	will	and	determinism,	since	the	two	are	likely	to	be	conflated.	 A note on truth 	One	of	the	most	common	and	famous	theories	of	truth	is	the	“correspondence	theory	of	truth.”	In	general,	approaches	to	truth	that	fit	under	the	umbrella	of	correspondence	theories	hold	that	a	proposition	is	“true”	only	insofar	as	it	corresponds	to	“reality”	(whatever	that	is—that	is	another	debate,	and	dissertation,	altogether).	The	purpose	of	this	dissertation	is	not	to	evaluate	the	correspondence	theory	of	truth,	but	rather	to	make	an	argument	about	the	capacity	for	individuals	to	come	together	to	establish	the	validity	of	statements	through	a	deliberative	procedure.	So,	my	concern	here	is	not	whether	statements,	judgments,	and	                                                15	While	Estlund’s	framework	is	concerned	with	justifying	democracy	in	general	on	epistemic	grounds,	I	adapt	it	slightly	where	needed	and	specify	its	strengths	within	the	context	of	deliberative	democracy	in	particular.	This	adaptation	is	entirely	consistent	with	Estlund's	approach,	given	that	the	epistemic	arguments	in	favour	of	democracy	in	general	and	democratic	deliberation	in	particular	are	very	similar.				 	52	decisions	arising	from	either	or	both	are	“true”—that	is	yet	another	important	question	for	another	project.	My	concern	here	is	whether	or	not	we	can	establish	a	democratic	deliberative	procedure,	series	of	procedures,	or	system	that	can	facilitate	the	establishment	of	validity	through	the	exchange,	by	assembled	individuals,	of	preferences	and	reasons	for	those	preferences.	Moreover,	I	am	concerned	about	how	we	can	generate	the	necessary	good	judgments	and	decisions	expected	in	a	liberal	democratic	society.	This	is	to	say	that	my	project	is	a	specific	kind	of	epistemological	endeavor	and	not	an	ontological	one;	while	I	presuppose	that	there	is	indeed	both	a	natural	and	normative16	“world	out	there”	that	can	be	apprehended,	understood,	and	shared,	my	interest	is	in	how	it	is	apprehended,	understood,	and	shared,	and	not	in	the	content	of	what	is	included	in	such	acts	or	the	physical	or	metaphysical	status	of	that	content.	This	interest	in	and	focus	on	epistemology	comes	from	a	more	specific	interest:	how	we	can	collectively	make	good	judgments	and	decisions.	For	my	purposes,	there	are	two	categories	of	requirements	that	need	to	be	met	for	a	judgment	to	be	good,	each	of	which	is	important.	First,	there	is	what	I	am	calling	the	personal	category	in	which	a	“good”	judgment	is	one	for	which	the	motivations	of	the	individual	making	that	judgment	are	accessible	and	known;	from	which	the	conclusion	follows	from	the	premises	(i.e.	logically	valid);	and	which	we	might	reasonably	expect	someone	to	choose	again,	affirm,	or	reaffirm	given	a	similar	context	It	is	“personal”	since	the	focus	is	on	how	a	given	individual	produces	a	judgment.	A	“good”	decision	for	my	purposes	is	a	decision	taken	collectively—and	based	on	the	good	judgments	of	one	or	                                                16	By	a	normative	“world	out	there”	I	mean	a	world	full	of	intersubjectively-established	moral	facts	(though	not	necessarily	universal	facts)	that	can	be	apprehended	and	discussed.			 	53	more	of	the	assembled	individuals.	Second,	there	is	an	impersonal	category	that	corresponds	to	whether	a	judgment	is	“correct,”	by	which	I	mean	valid	in	regards	to	the	sort	of	criteria	required	to	evaluate	a	judgment	given	the	kind	of	reality	it	is	meant	to	address:	natural,	normative,	or	subjective	(corresponding,	respectively,	to	Habermas’	categories	of	truth,	rightness,	and	truthfulness)	(Habermas	1984).	It	is	“impersonal”	insofar	as	its	evaluation	depends	on	factors	primarily	(though	not	entirely)	outside	of	the	individual.17	My	focus	here	is	mostly	on	good	judgments	and	decisions—I	will	only	address	the	correctness	dimension,	as	I	have	defined	it,	occasionally.	It	is	most	important	to	note	that	when	I	do	use	the	term	“correct,”	it	is	in	this	epistemic	sense	that	I	mean	it,	and	not	some	ontological	sense.	So,	again,	my	concern	in	this	dissertation	is	not	about	how	individuals	generate	“true	statements”	so	much	as	it	is	about	the	pragmatic	need	to	come	up	with	shared	epistemic	foundations	and	claims	that	can	be	used	to	reach	legitimate,	effective,	and	just	democratic	outcomes	through	deliberation.18	It	is	also	about	the	ways	this	process	is	undermined	by	common	limits	to	our	cognition,	and	how	those	are	brought	about	or	exacerbated	by	institutions	or	common	ways	of	thinking	and	behaving;	just	as	much,	it	is	about	how	we	might	respond	to	the	challenges	this	presents	to	democratic	life	by	generating	better	processes	for	the	ongoing	testing	of	the	reasons,	judgments,	and	decisions	that	come	out	of	democratic	deliberation.		                                                17	Specifically,	this	mode	includes	whether	a	statement	is	valid	in	relation	to	the	natural	world,	whether	it	is	right	in	relation	to	the	intersubjective	normative	world,	and	whether	it	is	truthful	in	relation	to	others	who	perceive	the	speaker	as	sincere	and	trustworthy	(see	Habermas	1984).	18	My	approach	here	draws	on	Habermas’	pragmatism	(see	1996;	1998);	cf.	with	correspondence	theories	from	Moore	(1993	[1901-1902])	and	Russell	(1994	[1905]).			 	54	Deliberative democracy as epistemology 	Warren	(2002)	argues	that	deliberative	democracy	offers	a	rational	way	of	generating	judgments	and	democratic	decisions.	He	bases	this	claim	on	the	argument	that	deliberative	processes	provide	better	outcomes	than	aggregative	or	other	approaches	to	decision	making	(rational	in	this	sense	refers	to	a	more	effective	and	efficient	means	to	a	given	ends).	In	the	following	section	I	will	present	Warren’s	epistemic	argument	for	the	superiority	of	deliberative	democracy	over	aggregative	democracy,	and	will	add	two	further	arguments	in	support	of	this	claim:	First,	drawing	on	Estlund	(2008)—that	deliberation	solves	the	problem	of	epistocracy;	and	second,	drawing	on	Page	(2007)	and	Landemore	(2012)—that	collective	deliberation	can	address	some	of	the	shortcomings	of	bounded	rationality	that	are	often	present	in	individual	judgment	by	adjusting	deliberative	processes	to	ensure	the	presence	of	cognitive	diversity	(i.e.,	diversity	in	perspectives)	among	participants.19	Taken	together,	these	arguments	help	bolster	the	case	for	democratic	deliberation,	as	currently	conceived,	as	a	way	of	addressing	some	challenges	to	autonomous	judgment	and	decision	making.	I	will	revisit	the	cognitive	diversity	point	again,	too,	in	chapter	five.	First,	though,	let	us	look	at	Warren’s	claim	that	deliberative	democracy	generates	more	rational	and	valid	judgments	and	decisions.	Warren	grounds	this	claim	in	the	supposition	that	deliberation,	insofar	as	it	requires	specific	procedures,	begins	as	a	more	epistemically	sound	approach	to	decision	making	and	judgment.	To	fully	explore	this	point,	let	us	return	to	and	expand	a	point	made	above	regarding	epistemic	validity.	Warren	argues	                                                19	This	argument	will	be	revisited	in	part	two	of	this	dissertation	when	I	discuss	approaches	to	designing	deliberation	and	deliberative	systems.			 	55		Decisions	resulting	from	properly-constituted	procedures	have	more	validity—that	is,	they	are	likely	to	be	truer	or	more	right	or	more	truthfully	related	to	needs….Validity	is,	in	other	words,	given	by	the	knowledge	that	a	statement	is	authoritative	in	relation	to	its	referent:	facts,	norms,	or	subjective	desires	and	experiences.	So	while	a	statement	may	be	true	(or	right	or	truthful)	in	some	ontological	sense,	epistemological	questions	have	to	do	with	how	we	can	know	it	to	be	so.	We	gain	such	knowledge,	or	confidence	in	the	validity	of	the	statement,	through	processes	of	challenge,	reason-giving,	and	verification	(192).			Thus,	the	primary	function	of	deliberative	democracy	as	a	rational,	epistemic	theory	of	democratic	judgment	and	decision	making,	is	the	transmission	of	information	from	a	private,	individual	state	to	a	public,	shared	state	in	which	participants	can	evaluate	claims	and	counter-claims	through	an	intersubjective	process	that	maximizes	validity	(and	good	judgments	and	decisions)	through	understanding	that	results	from	shared	reasons.	This	function	implies	that	there	are	epistemic	norms	outside	of	(and	perhaps	potentially	independent	of)	ontological	“truths”	that	govern	and	assist	in	determining	what	ends	up	counting	and	knowledge.	Deliberative	democracy	as	an	epistemic	theory	of	democracy	draws	its	rational	value	from	its	reliability	as	an	epistemic	procedure	for	generating—or	tending	to	generate—shared	knowledge	in	a	given	context;	it	generates	knowledge	through	a	procedure	rather	than	revealing,	stipulating,	or	insisting	upon	some	absolute	and	non-negotiable	ontological	reality	which	may	exist	in	private,	but	the	validity	of	which,	to	be	shared	publicly	in	a	democratic,	must			 	56	be	defended	rather	than	asserted.	This	validity-based	approach	is	why	deliberation	is	a	more	rational	way	of	generating	judgments	and	decisions.	As	Warren	puts	it,	the	legitimacy	that	follows	from	deliberation	“may	become	more	‘rational’—not	in	Max	Weber’s	sense,	which	refers	rational	legitimacy	only	to	the	outcomes	of	positive	procedures—but	in	the	sense	that	legitimacy	is	generated	out	of	public	deliberations	which	produce	reasons	with	motivational	force”	(Warren	2002:	187).	It	is	important	to	reiterate—as	this	is	the	central	point	of	this	chapter	and,	indeed,	the	animating	force	behind	much	of	this	dissertation—that	for	deliberative	democracy	to	yield	good	judgments	and	decisions,	participants	must	be	autonomous	in	the	extended	sense	of	being	able	to	state	their	motivations	for	holding	their	interests,	preferences,	desires,	and,	moreover,	the	reasons	for	their	reasons	(i.e.	their	“actual	reasons”).	Otherwise,	attempts	to	generate	understanding	through	shared	reasons	may	be	undermined.	This	undermining	will	occur	both	within	the	individual	and	the	group—including	attempts	at	challenge	and	verification—by	either	the	introduction	of	false	or	misleading	antecedents	or	reasons	that,	quite	simply,	are	not	justified	given	that	they	lack	valid	support	and	are	thus	suspect.	Again,	as	we	will	see	in	chapter	three,	there	are	plausible	and,	indeed,	strongly	persuasive	reasons	to	question	the	extent	to	which	agents	are	capable	of	acting	autonomously	in	the	extended	sense,	either	alone	or	in	groups.	Cognitive	distortions	and	limitations	are	pervasive	and	are	not	easily	eliminated,	even	at	the	group	level	(see	Mendelberg	2002,	for	instance,	for	a	comprehensive	review	of	the	ways	in	which	group	deliberation	can	amplify	these	effects).	So,	there	is	good	reason	to	ask	ourselves	whether	human	beings	are	well	adapted	to	the	epistemic	expectations	of	liberal-democratic	citizenship	in	the	deliberative	mold	(though	this	is	not	to	say	that			 	57	improvements	are	impossible	or	even	unlikely	to	work,	but	rather	that	more	empirical	work	needs	to	be	done	evaluating	deliberation	against	its	normative	claims).20		 Politics and expertise in deliberations 	The	special	concern	about	autonomously	(and	rationally)	expressing	one’s	reasoning	comes	from	the	fact	that	the	epistemic	force	of	deliberative	democracy	is,	in	part,	rooted	in	its	purported	capacity	to	act	as	a	theory	capable	of	responding	to	the	twin	fundamental	requirements	of	deliberative	democratic	decision	making:	(1)	that	the	practice	or	system	establishes	or	sustains	public	awareness	that	there	is	an	issue	that	we,	as	a	public,	need	to	address,	since	the	outcome	will	affect	us;	and	(2)	that	it	provides	a	procedure	or	series	of	procedures	by	which	the	“we”	of	the	first	requirement	can	collectively	decide	in	a	way	that	is	legitimate,	procedurally	fair,	and	more	likely	to	be	respected	by	those	involved,	including	those	who	disagree	with	the	outcome).	Each	of	these	imperatives	requires	shared	and	common	reasons	offered	by	more-or-less	equal	participants	who	can	express	their	preferences	and	the	reasons	(and	reasoning)	behind	them.	Deliberative	democracy	responds	to	this	need	by	providing	a	procedure	which	makes	public	the	otherwise	private	reasons,	preferences,	and	motivations	of	participants,	thus	generating	shared,	reliable	knowledge	than	can	be	used	to	generate	judgments	and	decisions,	provided	they	are	rationally	and	autonomously	generated.	Let	me	be	more	precise	about	this	point.	Politics	emerges	at	the	point	at	which	a	decision—broadly	conceived—must	be	taken	to	respond	to	or	adjudicate	among	competing	                                                20	This	issue	of	norms	and	capacities	is,	ultimately,	an	empirical	question	and	related	to	the	second	half	of	this	dissertation.	For	a	full	discussion	on	this	point,	see	chapters	5-7.				 	58	preferences,	wills,	claims,	needs,	or	desires.	This	adjudication	requires	common	grounding	in	basic	facts	and	norms	from	which	the	agents	or	collectives	engaged	might	address	the	issue	at	hand	through	a	procedure	that	is	knowledge	producing.	Ideally,	this	is	established	through	the	force	of	rational,	autonomously-generated	statements	exchanged	between	participants.	As	Warren	(2002)	argues,	in	reference	to	scientific	deliberations		[Scientific	deliberations]	are	institutionalized	in	a	positive	sense	through	peer	review,	discussions	among	colleagues,	scientific	meetings,	and	publications…when	others	recognize	statements	as	true,	they	gain	in	epistemological	authority	and	hence	validity.	It	is	important	that	validity	does	not	come	in	any	direct	sense	from	the	ontological	'objectivity'	a	statement	might	claim.	There	is	nothing	that	could	establish	such	objectivity	outside	of	scrutiny	of	the	statement	by	others…	(192-193)		Thus	expert	claims	achieve	standing	through	scrutiny,	review,	testing	(if	possible)	and,	then,	if	the	claim	is	taken	up,	public	acceptance.	The	a	priori	prohibition	on	merely	asserting	ontological	objectivity	(of	natural	facts,	not	norms)	as	a	basis	for	a	claim’s	validity	is	the	condition	for	that	claim	to	be	collectively	endorsed.	To	simplify:	it	is	never	enough	for	an	expert	judgment	to	rest	on	take	my	word	for	it.	The	same	can	be	said	about	public	deliberations	concerning	political	matters,	whether	carried	out	by	ordinary	citizens	or	their	elected	representatives.	There	are	strong	parallels	between	scientific	deliberation	and	public,	democratic	decision	making	and	judgment	in			 	59	deliberative	contexts.	Warren	(2002)	highlights	two	similarities	and	one	major	difference	between	expert	scientific	deliberations	and	public	political	deliberations,	and	each	reveals	something	about	the	epistemic	value	of	deliberation	as	a	political	response	to	divided	societies	(i.e.	most,	likely	all,	societies)		First,	validity	is	a	product	of	procedure,	suggesting	that	institutionalized	deliberation	can	establish	the	epistemic	validity	of	claims	and	assertions.	Second,	like	science,	politics	works	at	the	frontiers	of	validity,	although	in	a	different	sense.	Political	issues	emerge	precisely	when	epistemic	authority	is	questioned	or	has	yet	to	be	established.	But	in	other	ways,	politics	is	distinct	from	science.	In	politics,	factual	issues	are	intermingled	with	normative	and	expressive	issues,	so	that	the	authority	deriving	from	knowledge	of	facts	is	not	as	easily	achieved	within	political	contexts	as	within	the	relatively	insulated	institutions	of	science…In	politics,	[experts]	must	argue	and	convince.	They	must	enter	into	deliberations	in	order	to	educate,	persuade—or	be	persuaded.	Here,	expertise	has	no	pre-political	rights	(193).		Not	only	should	a	speaker's	claims	not	receive	a	priori	standing	in	the	context	of	political	judgment	and	decision	making,	those	claims,	in	such	a	political	context,	are	a	part	of	a	greater	endeavour	and	must	be	evaluated	alongside	other	considerations,	including	the	normative	concerns	of	all	those	present.	In	essence,	democratic	decisions,	by	their	nature,	require	reasons	based	on	a	different,	higher	standard	of	admissibility.	The	process	of	deliberation,	insofar	as	it			 	60	is	an	epistemic	process,	is	one	of	generating	publicly	valid	statements	by	communicating	private	reasons	from	the	individual	speaker	to	the	collective,	thus	subjecting	those	reasons,	much	like	in	the	case	of	scientific	deliberations,	to	scrutiny,	review,	and,	ultimately,	to	adjudication	alongside	complementary	or	competing	claims—thus	meeting	the	higher	standard	required	for	democratic	decisions.		Regardless	of	the	ontological	status	of	any	relevant	private	interest,	desire,	preference,	claim,	or	experience	that	an	individual	may	assert	as	their	own,	each	participant	in	a	deliberation	must	take	part	in	an	epistemic	process	by	which	each	is	given	an	opportunity—and	indeed	a	responsibility—to	establish	the	validity	of	that	information	alongside	others	through	reason	giving,	verification,	and,	if	necessary,	counter-challenge	(Warren	2002).	Thus,	deliberative	democracy,	insofar	as	it	seeks	to	produce	epistemically	valid	judgments,	offers	a	reliable,	sophisticated,	and	more	rational	(in	the	sense	noted	above	by	Warren)	way	to	make	decisions	collectively	by	requiring	that	potential	outcomes	are	subjected	to	scrutiny	and	are	the	result	of	an	exchange	of	reasons	and	argument.		Two more reasons why deliberation is a more rational way to generate judgments and decisions 	Building	on	Warren	and	taking	individuals	in	diverse	liberal	democracies	as	the	focus	of	this	study,	it	is	relatively	easy	to	further	support	the	claim	that	deliberation	is	a	more	rational	approach	to	decision	making	by	looking	at	two	further	benefits	it	offers,	each	related	to	managing	political	life	in	diverse	communities	in	different	ways.	The	first	relates	to	making	space	for	and	protecting	the	judgments	of	the	minority	and	avoiding	the	problem	of	epistocracy			 	61	(defined	below).	The	second	refers	to	leveraging	cognitive	diversity	for	the	purpose	of	generating	better	decisions	and	judgments,	and	enhancing	group	problem	solving.	Each	of	these	purported	goods	gives	us	reason	to	support	deliberation	as	a	more	rational	way	to	make	decisions	in	diverse	liberal	democracies.	Together,	if	true,	they	contribute	significantly	to	the	appeal	of	democratic	deliberation.	The	first	additional	reason	why	deliberation	is	a	better	way	to	produce	good	judgments	and	decisions	(compared	to	other	forms	of	democratic	decision	making)	is	that	it,	properly	designed,	solves	the	challenge	of	epistocracy.	Epistocracy,	as	defined	by	Estlund	(2008),	refers	to	the	rule	of	experts:	it	suggests	that	instead	of	letting	the	public	(i.e.	non-experts)	rule,	those	who	‘know	best’	ought	to	decide.	The	problem	of	epistocracy	is	grounded	in	the	following	logic:	If	good	outcomes	are	vital	to	democracy—an	issue	we	will	examine	more	carefully	below—and	if	experts	participating	in	decision	making	improve	the	likelihood	that	a	good	decision	(as	defined	above)	will	be	reached,	then	why	not	let	the	experts	rule?	Estlund	deftly	rejects	this	proposition	on	the	grounds	that	a)	authority	does	not	emerge	directly	from	expertise	(2008:	4);	and,	even	if	we	were	to	stipulate	that	it	did,	b)	not	all	citizens	in	a	given	polity	will	agree	on	who	the	experts	are	in	any	given	situation	and,	moreover,	not	all	would	necessarily	agree	on	what	such	experts	think	should	be	done	(2008:	3-4,	103-105).		Democratic	deliberation	serves	as	a	rational	alternative	to	rule	by	experts—though	this	claim	is	subject	to	evaluation	depending	on	the	type	of	deliberative	framework	advocated	(discussed	in	the	next	section).	Deliberative	theories	of	democracy	avoid	the	problem	of	epistocracy	by	taking	seriously	the	claim	made	by	Warren,	noted	above,	that	expertise	has	no	pre-political	authority,	and	by	grounding	the	legitimacy	of	outcomes	in	a	fair	and	open	decision			 	62	making	process.	These	theories,	when	properly	constituted,	also	address	the	problem	of	epistocracy	by	providing	a	way	for	those	holding	minority	positions	to	accept	the	majority	outcome	as	legitimate	without	surrendering	their	judgment	(as	we	will	see	below).		 The	second	additional	way	that	a	deliberative	democratic	approach	to	judgment	and	decision	making	is	more	rational	than	the	alternatives	is	found	in	a	feature	of	deliberation	that	enables	it	to	leverage	the	cognitive	diversity	of	participants	to	solve	problems.21	Recall	that	cognitive	diversity	as	defined	by	Landemore	(2013)	and	Hong	and	Page	(2004)	“refers	to	a	diversity	of	ways	of	seeing	the	world,	interpreting	problems	in	it,	and	working	out	solutions	to	these	problems.	It	denotes	more	specifically	a	diversity	of	perspectives…interpretations…heuristics…and	predictive	models”	(see	also	Mill	1999[1859],	Page	2007).	In	deliberative	contexts,	the	presence	of	a	variety	of	individuals	with	different	ways	of	thinking	has	been	shown	to	have	the	potential	to	improve	reasoning	and	decision	making	(Page	2007).	Accordingly,	it	is	thus	expected	to	increase	the	overall	quality	of	deliberation	by	offering	more	and	different	sorts	of	reasons	from	which	to	draw	a	judgment	or	decision,	and	by	providing	more	robust	tests	for	outcomes	(Bohman	2006;	Druckman	2004;	Landemore	2012,	2013;	Mercier	and	Landemore	2012).	In	this	case,	the	increased	rationality	of	decisions	is	a	product	of	a	practical	concern:	deliberative	democratic	procedures	marked	by	a	cognitively	diverse	group	of	individuals	make	better,	more	creative	decisions	and	judgments.	Moreover,	                                                21	Recall	that	cognitive	diversity	does	not	necessarily	require	a	diversity	of	individual	identities,	only	a	diversity	of	approaches	towards	cognition	and	its	related	practices	(i.e.	different	ways	of	thinking).	That	said,	to	the	extent	that	differences	in	identity	correlate	with	cognitive	diversity,	the	two	can	be	thought	of	as	bound	up	together,	though	this	is	not	necessarily	the	case,	nor	is	it	logically	required.				 	63	those	outputs	are	put	through	more	robust	tests	than	they	would	be	if	merely	left	to	groups	of	experts	(Page	2007)	or,	by	implication,	just	put	to	a	vote.		Epistemic proceduralism: the best of two worlds? 	Estlund	(2008)	offers	the	strongest	and	most	systematic	defense	of	democratic	authority	as	an	epistemic	theory	of	decision	making	and	judgment.	He	calls	his	approach	epistemic	proceduralism:	a	kind	of	compatibilist	theory	of	democracy,	that	can	be	fitted	to	democratic	deliberation,	that	sits	between	full	proceduralist	approaches—generally,	those	approaches	under	which	the	legitimacy	of	decisions	emerges	from	an	internal	procedural	standard,	such	as	fairness	in	the	strict	procedural	sense	of	procedural	equality—and	correctness	approaches—under	which	decisions	are	legitimate	because	they	are,	strictly	speaking,	correct	(e.g.	Plato	and	the	rule	of	the	Philosopher	Kings;	Rousseau	and	the	General	Will).	Estlund	rejects	the	former	on	the	grounds	that	democracy	is	concerned	with	more	than	just	fairness,	and,	to	some	extent,	is	after	the	‘right	decision’—otherwise,	a	coin-toss	would	be,	strictly	speaking,	the	most	‘fair’	way	to	make	a	decision	(since	each	would	have	an	equal	chance	of	affecting	the	outcome,	which	is	to	say	zero	chance)	and	would	be	sufficient,	regardless	of	outcome	(2008:	5-7).	He	rejects	the	latter	by	rejecting	epistocracy	on	the	grounds	that,	as	we	have	just	seen,	not	all	in	society	will	agree	on	who	the	experts	are	and	what	counts	as	expert	judgment	(2008:	3-4,	99,	105).	As	noted,	in	so	doing,	epistemic	proceduralism	solves	the	problem	of	epistocracy	by	requiring	that	the	minority	in	a	given	decision	recognize	and	accept	the	outcome	without	deferring	their	judgment	to	the	majority,	as	correctness	theories	would	have	them	do	(2008:	102-103).	This	practice	makes	space	for	expert	judgment	and	evaluation,	among	other	perspectives,	without			 	64	privileging	experts	over	citizens	or	substituting	their	judgment	for	the	judgment	of	the	participants	in	any	given	deliberation.	The	epistemic	force	of	deliberation,	in	Estlund’s	conception,	remains	in	place.	Instead	of	being	rooted	in	a	pure	proceduralist	or	correctness	theory,	Estlund	roots	epistemic	proceduralism	in	the	claim	that	properly	constituted	(deliberative)	decision-making	procedures	are	preferable	to	nondemocratic	arrangements,	fair	(in	the	sense	of	being	impartial	between	participants	and	generally	acceptable	to	a	reasonable	person),	and	have	a	better	than	random	chance	of	producing	the	correct	decision	(i.e.	correct	by	some	independent	standard,	whatever	it	may	be)	(Estlund		2008:	17,	98,	102-103,	167-168).22	According	to	Estlund,	epistemic	proceduralism	offers	“Procedural	impartiality	among	individuals’	opinions,	but	with	a	tendency	to	be	correct;	[it	is]	the	impartial	application	of	intelligence	to	the	moral	question	at	hand”	(107-108).	So,	epistemic	proceduralism	grounds	the	legitimacy	of	its	outcomes	in	a	procedure	that	is	both	democratic	(fair	and	acceptable)	and	that	has	epistemic	value—and	thus	legitimacy—through	its	tendency	to	get	things	right	(2008:	105-106).	Estlund	compares	epistemic	proceduralism	to	juries—though	not	in	the	sense	of	Condorcet’s	Jury	Theorem.23	                                                22	Such	a	standard	could—indeed	should—be	an	item	for	deliberation,	though,	as	with	constitutional	matters,	one	would	expect	it	to	be	difficult	to	amend	and	for	such	a	deliberation	to	occur	rarely.	23	Estlund	rejects	Condorcet's	Jury	Theorem—the	idea	that	a	sufficiently	large	assemblage	of	individuals,	each	with	a	more	than	fifty	percent	chance	of	being	correct,	will	reach	the	correct	answer	the	vast	majority	of	the	time	since	incorrect	positions	will	be	washed	out	by	the	overall	higher	probability	of	correctness.	Estlund	rejects	this	on	the	grounds	that:	1)	There	is	no	reason	to	assume	that	voters	are	better	than	random	and,	more	importantly,	2)	No	one	should	be	forced	to	substitute	the	judgment	of	the	majority	for	their	own	judgment—especially	since	one's	opposition	to	the	majority	judgment	gives	them	further	reason	to	doubt	that	outcome	on	top	of	the	dubious	likelihood	that	any	given	majority	is	more	likely	than	chance	to	be	correct	(Estlund	1997:	185-186).	See,	related,	Page	and	Shapiro's	(1992)	use	of	Condorcet's	Theorem	to	argue	that	the	public	as	a	public	has	stable,	rational	opinions	and	refutations	by	Althaus	(1998),			 	65	Instead,	he	argues	that	juries	derive	their	legitimacy	from	a	fair	procedure—an	adversarial	trial	in	which	both	the	prosecution	and	defendant	are	given	resources	and	a	chance	to	make	their	case—and	their	tendency	to	reach	the	correct	outcome	on	a	better-than-random-chance	basis	(2008:	8-9,	106-107,	168).	Estlund’s	epistemic	proceduralism	also	maintains	respect	for	an	ontological	reality	separate	from	deliberation	without	asserting	a	single,	mandatory	‘truth’	to	guide	all	deliberation—all	the	while	making	it	possible	for	participants	in	the	system	to	make	‘correct’	decisions	and	epistemically	valid	decisions	and	judgments.	Estlund	achieves	this	by	first	specifying	that	epistemic	proceduralism	is	not	“an	ideal	epistemic	situation”	nor	is	it	a	“constituting	truth”	(2008:	19).	Instead,	Estlund’s	framework	“imagines	deliberators	for	whom	there	are	independent	facts	about	what	ought	to	be	done”	(2008:	19).	Any	slip	towards	relativism	or	groundlessness	is	avoided	by	claiming	that	truth,	which	is	“generally	unspecific,”	is	related	to	some	particular	framework	and	thus	true	or	false	within	a	given	moral	system	(2008:	5).	Thus	deliberators	can	have	it	both	ways:	Estlund’s	framework	recognizes	that	there	is	an	ontological	reality	for	its	participants	that	they	bring	with	them	into	deliberations	(and	perhaps	maintained	throughout	and	after	deliberations,	despite	what	might	be	decided	within	the	deliberation).	It	also	accepts	that	such	a	reality—with	the	rights	and	wrongs	and	whatever	imperatives	it	might	hold—will	render	specific	outcomes	within	the	system	right	or	wrong	for	some	participants	and	not	others,	but	will	nonetheless	be	preferable	to	alternative	decision-making	procedures	(for	the	reasons	noted	above).	However,	again,	the	rights	of	the	minority	in	                                                Kuklinski	and	Quirk	(2000),	and	Zaller	(1992)	that	are,	primarily,	grounded	in	the	arguments	that	voters	are	unlikely	to	be	correct	more	often	than	not	and	that	many	of	them	have	a	tendency	towards	systematic	bias,	as	I	will	discuss	in	chapters	three	and	four.			 	66	a	given	outcome	are	respected	and	protected,	since	minorities	are	given	the	option	to	reject	the	correctness	of	the	outcome	(they	need	not	defer	their	judgment),	even	though	they	are	required	to	recognize	and	obey	it	as	the	product	of	a	legitimate	procedure	that	is	fair,	and	that	tends	to	produce	a	correct	outcome.		Bohman	(1998)	summarizes	the	approach	this	way	“…epistemic	proceduralism	seems	to	require	only	that	citizens	share	the	assumption	that	there	is	some	way	to	judge	better	or	worse	reasons	and	not	that	a	specified	theory	of	public	justification	provides	specific	criteria	to	settle	disputes	about	such	norms”	(407).	This	epistemic	proceduralism	is	what	Estlund	calls	a	“formal	epistemic	account”—an	account	of	a	decision	making	that	has	a	“tendency	to	get	things	right	from	the	standpoint	of	justice	or	common	good	whatever	the	best	conception	of	those	might	be”	(Estlund	2008:	169,	emphasis	in	original).	Democratic	deliberation,	in	this	conception	of	it,	thus	allows	for	a	political,	intersubjective	way	of	working	out	what	ought	to	be	done	without	compromising	the	quality	of	the	sorts	of	decisions	reached.	However,	as	I	argue	below,	this	requires	that	individuals	exercise	their	capacity	for	autonomy.		 Going	forward,	it	is	important	to	keep	in	mind	that	my	approach	diverges	from	Estund’s	in	two	important	ways.	First,	I	rely	on	a	very	thin	conception	of	correctness	for	evaluating	democratic	judgments	or	decisions.	This	conception	is	based	in	part	on	the	requirement	that	validity	is	included	as	part	of	a	measure	of	a	good	outcome	(i.e.	validity	in	regards	to	the	sort	of	criteria	required	for	evaluating	that	kind	of	judgment,	depending	on	whether	it	is	about	natural,	normative,	or	subjective	concerns).	Second,	while	Estlund	discusses	“independent	facts”	about	“what	ought	to	be	done,”	I	am	concerned	with	dependent	facts	about	what	we	would	like	to	do;	I	suggest	that	(normative)	facts	are	dependent	upon	an	intersubjective	process	of	acceptance,			 	67	just	as	any	decision	about	what	is	to	can	comes	from	the	authority	of	a	group	established	to	come	up	with	such	a	decision	through	the	exchange	of	reasons.		Elstub’s critique of the epistemic defense of deliberation 	Before	turning	to	a	look	at	how	autonomy	relates	to	deliberative	democracy	as	an	epistemic	approach	to	democratic	decision	making,	I	want	to	address	a	critique	of	the	theory’s	epistemic	justification	offered	by	Elstub	(2006:	304,	2008:	61-62).		Elstub	characterizes	the	epistemic	justification	of	deliberative	democracy	as	being	grounded	in	its	claim	to	offer	“the	best	method	of	providing	good	decisions….	because,	by	generating	public	reason,	it	can	lead	to	decisions	that	are	true,	well	justified	or	commensurate	with	justice,	needs	or	the	common	good”	(61).	Against	this,	he	claims	that	it	cannot	be	known	whether	deliberative	approaches	to	democratic	decision	making	do	indeed	provide	the	best	method	for	reaching	such	decisions	since	we	cannot	test	this	claim;	and	if	we	could	test	it,	there	could	be	another	approach	to	decision	making	that	is	not	known,	so	deliberation	cannot	be	assumed	as	the	best	approach.	Moreover,	he	notes,	along	with	Cohen	(1989),	that	there	is	reason	to	doubt	the	existence	of	a	“’real	truth’	about	the	common	good.”	Accordingly,	he	concludes	deliberative	democracy	cannot	be	justified	by	reference	to	its	epistemological	superiority	to	other	ways	of	making	decisions.		 While	there	is	very	good	reason	to	doubt	that	a	“real	truth	about	the	common	good”	exists—at	least	in	the	absence	of	a	self-contained	system	in	which	ethical,	social,	and	political	parameters	are	strictly	established	and	controlled—this	fact	does	not	gainsay	the	epistemic	justification	of	deliberative	democracy	offered	by	Estlund	(whose	full	statement	of	epistemic	proceduralism	came	after	Elstub’s	critique).	That	is	because	the	particular	epistemological	claim			 	68	made	by	many	who	justify	deliberation	on	epistemic	grounds,	and	especially	the	framework	offered	by	Estlund,	is	not	that	deliberative	democracy	reaches	or	reveals	some	ontological	truth	about	the	content	of	the	common	good;	rather,	the	claim	proponents	make	is	that	deliberation	offers	a	(fair)	procedure	by	which	a	given	collective	can	establish	validity	through	reason	giving	and,	if	needed,	verification	and	challenge	(Warren	2002).	They	also	argue	that	deliberation	tends	to	generate	decisions	that	are	correct	(according	to	Estlund)	or	good	(as	I	mean	it)	more	often	than	random	within	a	given	moral,	ethical,	social,	and	political	system,	while	also	being	based	on	a	democratic	(i.e.	fair	and	generally	acceptable)	procedure	(Estlund	2008).	This	same	point	applies	to	Elstub’s	critique	that	there	is	no	test	we	can	apply	to	deliberative	democracy	to	determine	if	it	does	indeed	lead	to	“decisions	that	are	true,	justified	or	commensurate	with	justice,	needs	or	the	common	good.”	Again,	deliberation	is	not	about	the	discovery	of	ontological	truths,	but	the	establishment	of	epistemic	agreement	about	the	validity	of	statements,	and	the	generation	of	judgments	and	decisions	about	what	we	want	to	do	in	a	democracy.	Validity	in	this	context	is	a	product	of	a	fair,	generally	acceptable,	and	properly	constituted	procedure.	Deliberative	democracy,	properly	structured,	offers	an	approach	to	establishing	valid	statements	about	political	issues	in	the	hopes	that	we	might	reach	some	decision	in	a	way	that	is	more	rational	and	thus,	hopefully,	better	than	the	alternative	options	and	more	acceptable	to	those	who	will	be	affected	by	any	decisions	that	are	be	made.	Deliberative	democracy,	then,	is	first	and	foremost	a	theory	of	how	a	properly-structured	process	can	lead	to	more	ethical,	rational,	and	legitimate	democratic	decisions;	it	does	not	presuppose	any	particular	ethical	content	or	require	reaching	common	good	as	an	outcome	of	its	proceedings	outside	of	the	core	requirements	and	values	of	liberal	democracy			 	69	(Warren	2002).	Thus,	critiques	of	the	epistemic	justification	of	deliberative	democracy	that	draw	on	doubt	about	the	existence	or	verifiability	of	a	common	good,	in	the	thick	ethical	sense	rather	than	the	thin	functional	sense,	mischaracterize	how	deliberation	is	commonly—though	perhaps	not	universally—justified	on	epistemic	grounds.	Such	a	justification	is	based	on	democratic	deliberation	as	an	inter-subjective	process	aimed	at	securing	common	terms	of	discussion	and	evaluation,	and	facilitating	an	exchange	of	reasons	to	reach	an	internal	understanding	of	the	desires,	interests,	and	preferences	of	participants—and,	sometimes,	though	not	always,	a	specific	political	outcome.		Autonomy in deliberative democracy: the keystone capacity 	If	epistemic	proceduralism	functions	as	a	framework	for	the	theory	and	practice	of	deliberative	democracy	aimed	at	generating	a	foundation	for	evaluating	claims	and	counter-claims	on	an	ongoing	basis,	it	requires	a	cognitive	underwriter.	It	is	in	this	sense	that	I	think	of	autonomy:	the	capacity	that	serves	to	connect	the	will,	through	rationality	and	self-awareness,	to	the	reasons	one	has	for	their	preferences	and	judgments	in	such	a	way	that	they	can	both	reach	and	understand	such	preferences	and	judgments	and	their	origins;	moreover,	autonomy	requires	that	individuals	be	also	able	to	communicate	that	information	to	others.	Autonomy,	in	the	Kantian	sense,	noted	in	chapter	one,	is	present	when	one’s	thinking	is	grounded	in	a	sound	internal	cognitive	procedure.	Building	on	this	core	requirement,	I	take	my	conception	of	autonomy	from	Christman	(1991),	who	defines	it	as	a	state	in	which	“…the	influences	and	conditions	that	give	rise	to	the	desire	[or	preference	or	intention]	were	factors	that	the	agent	approved	of	or	did	not	resist,	or	would	not	have	resisted	had	she	attended	to	them,	and	that			 	70	this	judgment	was	or	would	have	been	made	in	a	minimally	rational,	non-self-deceived	manner”	(22,	emphasis	mine).	Thus,	in	Christman’s	formulation,	an	agent	can	only	be	said	to	be	autonomous	if	she	is	aware	“of	the	changes	and	development	of	her	character	and	of	why	they	came	about”	(11),	since	only	in	this	way	can	she	encourage	or	resist	these	changes	through	implied	or	actual	deliberation	with	others.			 There	are	three	reasons	why	Christman’s	understanding	of	autonomy	is	appropriate	for	the	study	of	deliberative	democracy	that	I	am	undertaking	in	this	dissertation,	and	I	will	briefly	address	each	of	them	here.	First,	Christman	logically	separates	the	possibility	of	a	concept	of	autonomy	from	that	of	any	concept	of	freedom,	but	maintains	a	link	between	the	two,	noting	that	“…the	nature	and	value	of	political	freedom	is	intimately	connected	with	the	presupposition	that	actions	one	is	free	to	do	flow	from	desires	and	values	that	are	truly	an	express	of	the	‘self-government’	of	the	agent”	(Christman	1991:	1).	So,	at	a	conceptual	level,	autonomy,	though	connected	in	practice	to	freedom,	is	a	separate	capacity	that	can	be	interrogated,	understood,	and	even	cultivated,	on	its	own.	The	concept	of	autonomy	advanced	by	Christman	thus	avoids	being	folded	into	the	concept	of	freedom	as	some	variation	on	an	absence	of	restraint.	This	allows	us	to	discuss	autonomy	as	the	internal	process	of	a	particular	agent—or	collection	of	agents—and	provides	us	with	the	conceptual	opportunity	to	qualify	conditions	and	expressions	of	freedom	within	deliberation	as	being,	in	part,	linked	to	the	extent	and	quality	of	the	capacities	for	autonomy	held	by	those	participating	in	the	deliberative	process.		 Second,	Christman	avoids—deliberately—the	tricky	problem	of	infinite	regress	that	is	associated	with	some	definitions	of	autonomy;	indeed,	he	specifically	addresses	the			 	71	problematic	definition	of	autonomy	offered	by	Dworkin	(1988)	(Christman	1991:	4-10).	Dworkin’s	formulation	of	autonomy	rests	on	the	claim	that	a	person	is	autonomous	when	he	is	in	a	position	to	affirm,	reject,	or	alter	his	preferences,	desires,	or	goals	(Dworkin	1988).	However,	as	Christman	notes,	drawing	and	building	on	a	critique	offered	by	Irving	Thalberg,	not	only	does	this	definition	of	autonomy	open	up	the	possibility	that	an	agent	finds	himself	affirming	heteronomous	positions	(Christman	1991:	5),	it	invites	the	problem	of	infinite	regress,	since	any	affirmation	or	rejection	of	a	preference,	desire,	or	goal	will	also	require	its	own	justification	(and	affirmation)	if	it	is	also	to	be	an	autonomous	choice.	Indeed,	any	condition	of	“self-appraisal,”	as	Christman	calls	it,	vis-à-vis	autonomy,	will	face	this	threat.		Christman’s	definition	of	autonomy	avoids	the	problem	of	infinite	regress	by	fixing	the	condition	of	autonomy	at	the	first	level	of	evaluation:	the	process	by	which	a	desire,	preference,	or	interest	is	developed	(Christman	1991:	18-19).	Nothing	would	stop	an	individual	from	running	through	this	process	multiple	times—say,	to	reflect	upon	a	desire	that	is	motivated	by	another	desire;	what	is	essential	to	the	process	is	the	individual’s	awareness	of	what	contributes	to	the	formation	of	the	preference,	interest,	desire,	and	so	on.24	Thus	the	process	of	evaluation,	if	undertaken	in	conditions	of	minimal	rationality	and	self-awareness,	serves	as	both	the	necessary	and	sufficient	condition	of	autonomy	without	the	need	to	evaluate	any	particular	outcome.	The	parallel	with	deliberation	rests	in	the	fact	the	epistemic	force	of	deliberation	is	itself	found	in	a	process,	rather	than	in	a	moment	or	an	outcome;	theories	of	                                                24	The	standard	of	what	counts	as	“being	aware”	is	difficult	to	specify	outside	of	a	particular	agent	and	context;	however,	a	minimal,	reasonable	standard	is	easy	enough	to	approximate:	any	motivation	that	a	reasonable	individual	might	realistically	be	expected	to	uncover	under	conditions	of	self-reflection	and	having	most	of	the	information	immediately	required	in	that	particular	context	is	enough.				 	72	deliberative	democracy	thus	require	a	concept	of	autonomy	suited	to	their	procedural	nature,	and	Christman’s	definition	fits	nicely	while	also	maintaining	a	critical	standard	of	evaluation.	Finally,	and	related	to	the	previous	point,	Christman’s	conception	of	autonomy	avoids	the	“time	slice”	problem	of	theories	of	autonomy	that	ground	the	capacity	for	autonomy	in	the	ability	to	judge	a	particular	outcome	at	a	given	moment.	Again,	when	evaluating	autonomy	in	the	moment,	such	as	is	required	by	a	theory	of	autonomy	such	as	Dworkin’s,	there	is	no	guarantee	that	the	agent	is	not	affirming	or	rejecting	or	altering	a	desire,	interest,	or	preference	heteronomously.	This	problem	is,	once	again,	avoided	by	locating	autonomy	in	an	ongoing	process	of	preference/interest/desire	formation	and	the	conditions25	under	which	this	process	unfolds,	rather	than	in	a	single	moment	of	judgment.	The	parallel	with	deliberative	democracy	in	this	instance	is	the	same	as	noted	above	during	the	discussion	of	infinite	regress:	procedural	autonomy	is	well	fitted	to	deliberative	democracy	because	the	efficacy	of	each	as	a	system	is	predicated	on	a	process	rather	than	a	moment,	and,	indeed,	is	agnostic	when	it	comes	to	the	content	of	a	judgment	or	decision.	This	principle	aligns	nicely	with	the	model	of	deliberation	with	which	I	am	working	here,	a	model	that	accepts	that	an	ontological	reality	for	each	participant	(i.e.	beliefs	about	what	is	true)	will	exist	in	any	given	deliberation	without	assuming	that	everyone	shares	such	a	reality	and	is	willing	to	accept	whatever	suggested	decisions	or	judgments	it	implies.		Before	continuing	to	the	matter	of	non-self-deception	in	autonomy,	it	will	be	useful	to	briefly	revisit,	summarize,	and	comment	a	bit	more	on	Christman’s	formulation	of	autonomy,	                                                25	It	is	worth	mentioning	that	as	Christman	notes	“…these	conditions	may	have	little	to	do	with	how	the	agent	evaluates	the	desire	[or	interest	or	preference]	itself…”	(1991:	11).			 	73	since	it	runs	counter	to	most	approaches	to	defining	the	concept.	According	to	him,	the	conditions	for	autonomy	are		(i) A	person	P	is	autonomous	relative	to	some	desire	D	if	it	is	the	case	that	P	did	not	resist	the	development	of	D	when	attending	to	this	process	of	development,	or	P	would	not	have	resisted	that	development	had	P	attended	to	the	process;		(ii) The	lack	of	resistance	to	the	development	of	D	did	not	take	place	(or	would	not	have)	under	the	influence	of	factors	that	inhibit	self-reflection;	and	(iii) The	self-reflection	involved	in	condition	(i)	is	(minimally)	rational	and	involves	no	self-deception	(Christman	1991:	11,	emphasis	in	original).	 The	meanings	of	(i)	and	(ii)	are	each	reasonably	self-evident	and	have,	anyway,	been	discussed	above.	However,	there	are	two	elements	of	Christman’s	definition	of	autonomy	upon	which	I	wish	to	dwell,	since	they	are	directly	relevant,	though	perhaps	not	obviously	so,	to	an	understanding	of	how	and	why	autonomy	is	necessary	for	deliberative	democracy	to	be	plausibly	considered	as	a	theory	of	democratic	engagement	aimed	at	generating	epistemically	valid	outcomes.			 First,	autonomy	requires	what	Christman	calls	“minimal	‘internal’	conditions	for	rationality”	(14).	He	cites	a	basic	consistency	of	beliefs	and	desires	as	requirements	but	stops	short	of	demanding	that	there	be	an	absolute	and	clear	link	between	the	epistemic	process	of			 	74	developing	internal	consistency	and	the	ontological	objectivity	of	the	external	world.	Thus	autonomy	requires	only	internal	consistency,	and	not	an	objectively	verifiable	connection	to	some	pre-established	ontological	reality.	The	link	to	the	model	of	deliberative	democracy	that	I	am	working	with,	and	to	its	epistemic	defense,	is	clear	enough:	participants	in	deliberation	who	are	in	search	of	epistemic	validity	must	be	capable	of	maintaining	at	least	a	basic	internal	consistency,	otherwise	the	grounding	upon	which	the	deliberative	enterprise	rests	is	subject	to	its	own	inconsistency	(though	this	does	not	necessarily	require	universal	agreement	on	any	given	ontological	reality).	Epistemically,	deliberation	is	primarily	about	process	and	not	outcome,	though	proponents	make	claims	about	the	sorts	of	outcomes	that	are	likely	to	come	about	from	well-designed	and	executed	deliberations;	it	is	not	enough	to	come	to	‘the	right’	decision,	since	it	is	not	clear	on	what	bases	we	would	evaluate	this	absent	some	justifying	procedure—which	is	what	deliberative	democracy	is:	a	way	of	doing	things.	That	decision	must	also	reflect	a	logical	consistency	that	is	unlikely	to	emerge	if	it	is	drawn	from	a	collection	of	illogical	internal	processes;	and	even	if	it	did,	it	could	not	be	said	to	be	the	product	of	an	epistemically-valid	process.	Second,	autonomy	requires	that	“the	influences	and	conditions”	surrounding	a	judgment,	through	the	interests,	preferences,	motivations,	and	desires	that	support	such	a	judgment,	were	approved	of	by	the	agent—or	would	have	been—under	what	we	might	call	conditions	of	sufficient	awareness	(a	minimal	level	of	knowledge	about	factors	relevant	to	the	judgment	at	hand,	which	I	will	discuss	below).	In	other	words,	the	agent	must	be	in	a	position	to	assent	to	all	of	the	immediate	factors	that	contribute	to	their	judgment.	This	assent	requires,	as	Christman	notes	(11),	self-reflection	and,	I	add,	returning	to	the	extended	conception	of	Kant’s			 	75	autonomy	over	autarchy,	it	also	requires	that	the	factors	considered	be	the	actual	mobilizing	agents	related	to	the	judgment	(or	desire,	preference,	etc.).	This	is	to	say	that	an	agent	is	only	autonomous	to	the	extent	that	the	process	of	self-reflection	he	undertakes	in	the	course	of	approving	of	a	judgment	accurately	links	“influences	and	conditions”	to	outcomes	and	is	not	interrupted	by	some	internal	or	external	force.		This	requirement	of	autonomy	is	what	I	am	referring	to	as	the	principle	of	non-self-deception	(examined	further	below).		 Violations	of	this	principle	are	common.	In	chapters	three	and	four	I	will	deal	with	these	violations	in	much	greater	detail	by	looking	at	four	cognitive	models	that	explain	how	certain	modes	of	cognition	explain	why	cognitive	distortion	is	so	common;	but,	for	now,	one	key	study	from	political	behaviour	research	will	serve	to	illustrate	the	form	of	the	problem.	Sniderman	et	al.	(1986)	set	out	to	understand	how	citizens	come	to	decide	what	they	think	about	some	particular	complex	political	issue.	They	found	that	among	both	more	and	less	educated	voters,	broken	reasoning	chains	were	common:	rather	than	building	from	a	premise	or	a	series	of	premises	to	a	conclusion	through	a	series	of	inductive	steps,	many	respondents	began	with	a	preference	and	worked	backward	to	rationalize	it	outside	of	their	awareness.	Low-information	respondents	drew	on	affect-driven	considerations,	while	high-information	voters	drew	on	ideology—but	each	rationalized	their	judgments.	While	neither	of	these	processes	of	chain	reasoning	may	be	particularly	devastating	to	autonomy	in	opinion	formation	on	their	own,	Sniderman	and	his	colleagues	found	that,	in	general,	individuals	in	mass	publics	maintain	consistency	of	belief	by	adjusting	their	policy	preferences	to	their	likes	and	dislikes,	and	then	by	reasoning	backward	from	their	conclusion.	Thus,	for	some	citizens—at	least	on	some	issues—preference	formation	is	about	the	ex	post	facto	rationalization	of	prior	affective	or	ideological			 	76	relationships	(for	more	expressly	political	examples	of	this,	see	Lodge	et	al.	1995;	Lodge	and	Taber	2013).	This	finding	implies	that	whatever	the	conscious	approach	that	some	take	to	explain	their	preferences	or	beliefs,	knowledge	of	how	they	were	formed	and	what	drives	them	may	remain	elusive.	Given	some	respondents’	rationalizations,	they,	strictly	speaking,	cannot	be	said	to	be	behaving	autonomously,	since	the	link	between	influences	and	conditions	and	outcomes	is	broken	or	distorted.	This	phenomenon	is	an	example	of	an	agent	expressing	autarchy	but	not	autonomy	since	the	agent	might	not	have	approved	of	the	outcome	if	they	had	the	full	appreciation	of	their	motivations	and	better	knowledge	of	the	process	by	which	they	came	to	their	judgment.26	In	the	case	of	deliberative	democracy	and	the	procedural	generation	of	epistemic	validity,	the	same	standard	can	be	upheld:	only	reasons	given	that	accurately	reflect	influences	and	conditions	can	count	as	being	part	of	an	autonomous	judgment.	Establishing	epistemic	validity	requires	not	only	an	exchange	of	reasons	directed	at	establishing	common	terms	of	discourse,	but	also	a	common	working	understanding	of	the	interests,	desires,	motivations,	preferences	(and	so	on),	and	the	subsequent	reasons	that	support	them.	Moreover,	it	requires,	at	least,	a	working	agreement	on	basic	facts	and	norms,	and	that	each	reason	is	logically	and	accurately	attached	to	its	motivational	source.	If	parties	in	a	deliberation	are	making	judgments	that	are	disconnected	or	hidden	from	their	ultimate	motivations,	they	cannot	be	said	to	be	behaving	autonomously—which	undermines	any	claim	to	be	producing	epistemically	valid	                                                26	See	also	the	related	phenomenon	of	“motivated	reasoning,”	which	refers	a	process	by	which	issue	partisans	process	a	variety	of	information,	some	of	it	challenging	to	their	position,	in	a	way	that	confirms	or	even	enhances	their	pre-existing	position	(Ditto	and	Lopez	1992;	Taber	et	al.	2001).	The	phenomenon	of	motivated	reasoning	will	be	thoroughly	explored	in	chapter	four.			 	77	outcomes,	and	thus	calls	into	question	what	deliberative	democracy	really	requires	as	sufficient	justification	for	claims	made	to	one	another	by	participants.	Why?	Because	reasons	that	are	partial,	distorted,	or	which	rely	on	hidden	motivations	prevent	potentially	relevant	information	from	being	put	on	the	table	during	a	deliberation,	thus	constraining	the	ability	of	those	assembled	to	establish	the	sort	of	validity	required	to	make	good	judgments	and	decisions.		 So,	for	it	to	be	said	that	an	agent	is	behaving	autonomously	during	deliberation,	she	must	be	able	to	maintain	basic	and	minimal	internal	logical	consistency	and,	upon	reflection,	be	able	to	more-or-less	accurately	link	her	motivations	to	the	reasons	she	is	offering	towards	some	particular	outcome:	if	too	much	is	lost	in	translation	between	one’s	motivation	and	one’s	stated	preferences	and	the	reasons	given	for	those	preferences,	the	agent	cannot	be	acting	autonomously.	Moreover,	for	the	reasons	noted	above,	for	it	to	be	said	that	a	deliberation	is	working	towards	establishing	epistemic	validity,	its	participants,	at	the	very	least	those	who	are	contributing	significantly	to	the	outcome	of	the	deliberations,	must	meet	the	basic	requirements	of	autonomy.		Autonomy and the principle of non-self-deception  	I	have	already	discussed,	via	Christman’s	definition	of	the	concept,	what	I	mean	by	autonomy	and	I	have	laid	out	its	core	requirement	of	non-self-deception:	the	requirement	that	one’s	motivations	are	reasonably	aware	to	them	and	that	the	reasons	one	gives	for	a	judgment	are	those	which	actually	motivate	them.	However,	it	remains	to	be	seen	why	this	principle	is	so	essential	to	autonomy.	I	have	already	addressed	why	Dworkin’s	formulation	of	autonomy	as	a	capacity	to	affirm	or	reject	outcomes	after	the	fact	invites	the	problem	of	infinite	regress.	Also,			 	78	and	again,	as	this	point	is	essential,	his	definition	invites	the	possibility	that	one	is	affirming	heteronomous	sources	without	awareness	of	what	these	sources	are	or	how	they	impact	one’s	decisions	or	judgments	(a	corollary	of	the	infinite	regress	problem).	Autonomy	requires,	for	the	reasons	noted	above,	non-self-deception	to	avoid	a	state	in	which	one’s	motivations	and	consequent	preferences	or	judgments	are	determined	outside	of	their	awareness.	Nonetheless,	neither	of	these	counter-arguments	explains	why	non-self-deception	is	important	to	judgments	and	decisions	in	general	or	in	instances	of	deliberation	in	particular.	If,	for	example,	a	deliberative	procedure	is	fair	and	acceptable,	and	has	the	tendency	to	produce	good	outcomes,	why,	then,	is	it	also	important	that	participants	in	the	system	not	be	self-deceived?	Why	is	it	important	that	they	not	only	produce	often-enough	good	(i.e.	more	often	than	random)	and	legitimate	decisions	but	also	act	autonomously?	We	might	call	this	the	problem	of	a	rose	by	any	other	name,	since	it	poses	the	challenge	of	whether	some	ersatz	copy	of	a	deliberative	system	and	its	participants,	though	of	different	fundamental	composition	(i.e.	without	autonomous	agents),	is	a	priori	undesirable	and	less	of	an	appropriate	standard	for	judgment	and	decision	making	in	liberal	democracies.		 There	are	at	least	two	plausible	lines	of	argument	against	the	potential	banality	and	irrelevance	of	self-deception	and	heteronomy	to	democratic	deliberation—arguments	against	the	ersatz	copy.	The	first	is	empirical:	If	self-deception	and	heteronomy	and	their	effects	were	randomly	distributed	both	among	and	within	participants	in	a	deliberation,	then	the	concern	about	those	effects	might	be	somewhat	mitigated	or	perhaps	even	eliminated	altogether.	However,	those	effects	are	not	always	randomly	distributed.	Part	of	the	problem	of	self-		 	79	deception	and	heteronomy	is	that	their	presence	and	effects	can	become	structurally	distributed	in	at	least	two	ways.		The	first	way	concerns	among	whom	their	effects	are	most	pronounced:	ample	research	shows	that	low-information	individuals	(i.e.	those	who	may	be	engaged	with	politics	but	who	are	poorly	or	partially	informed)		are	more	affected	by	heuristics	whose	true	effect	they	are	unlikely	to	be	aware	of—and	so	the	system,	as-is,	tends	towards	structural	bias	in	the	outcomes	it	produces	(see,	for	instance,	Althaus	1998,	Chong	and	Druckman	2007,	Converse	1964,	Cutler	2002,	Iyengar	et	al.	1982,	Kahneman	2011,	Kuklinski	and	Quirk	2000,	Sniderman	et	al.	1986,	Zaller	1992;	cf.	Lenz	2009).	The	second	concerns	who	tends	to	bear	the	burden	of	the	adverse	outcomes	of	heteronymous	judgments	or	preferences	(however	they	might	be	generated):	for	example,	people	of	colour	(Mendelberg	2001,	Sniderman	et	al.	1986),	the	poor	and	undereducated	(Althaus	1998,	Frank	2004,	Zaller	1992),	and	other	groups	who	already	suffer	the	effects	of	negative	stereotyping	(Kuklinski	and	Quirk	2000).		Whatever	the	many	the	complex	reasons	why	the	presence	and	effects	of	self-deception	and	heteronomy	are	distributed,	the	deleterious	effects	of	their	presence	on	deliberations	and	the	outcomes	of	deliberations	are	significant,	especially	among	certain	populations	of	(often)	already	disadvantaged	groups,	while	the	benefits	of	potential	positive	effects	tend	to	be	accumulated	by	those	who	are	already	advantaged	by	the	status	quo	or	who	begin	from	positions	of	relative	power	or	influence.	For	instance,	the	threat	and	effects	of	climate	change—e.g.	heat	waves	and	extreme	weather	events—are	disproportionately	borne	by	the	poor	(United	States	Environmental	Protection	Agency	2015).	In	an	experiment	on	motivated	reasoning	(discussed	further	in	chapter	4)	and	preferences	on	climate	change	policy	in	the			 	80	United	States,	Hart	and	Nisbet	(2012)	find	that	political	partisanship	influences	support	for	climate	change	and	that	new	information—shared	equally	and	presented	identically	to	each	participant—further	polarizes	opinions	on	climate	change	between	Republicans	and	Democrats	(see	also	Taber,	Cann,	and	Kucsova	2009).	And	if	such	effects	are	combined	with	a	group	in	which	a	minority-type	is	outnumbered,	polarization	can	become	worse	through	increased	(non-cognitive)	bias	and	decreased	cooperation	(Bettencourt	and	Dorr	1998).	So,	in	the	case	of	climate	change,	to	the	extent	that	polarization	leads	to	inaction	or	inadequate	action—which,	to	date,	globally,	it	largely	has—then	cognitive	distortion	(which	undermines	autonomy)	contributes	to	the	structural	and	persistent	disadvantage	of	certain	groups	over	others.		 Even	if	the	effects	of	self-deception	and	heteronomy	were	randomly	distributed,	there	is	no	reason	to	believe	that	they	would	contribute	to	good	decisions	on	a	better-than-random	basis,	while	there	are	many	reasons	to	believe	that	autonomy	and	non-self-deception	would.	Imagine	trying	to	assemble	a	piece	of	furniture.	The	heteronomous/self-deception	approach	is	akin	to	winging	it:	grabbing	the	hex	key	and	slapping	pieces	together	under	the	impression	that	“Hey,	I	know	what	I’m	doing	here.”	The	autonomous	approach	would	be	that	which	requires	that	one	had	a	reason	to	connect	particular	bits	and	pieces—this	leg	to	that	frame—and,	moreover,	reasons	for	those	reasons—to	stabilize	the	piece,	to	keep	it	from	falling	over,	etc.	The	specific	analogy	here,	of	course,	related	to	approaches	that	link	epistemically	valid	claims	and	reasons	to	judgments	or	decisions	and	those	that	do	not.	Being	autonomous,	having	reasons	and	knowing	why	one	holds	them,	should	increase	the	probability	of	good	(or	correct)	decisions	by	increasing	the	reliability	of	both	the	information	one	has	and	the	process	by	which			 	81	one	internally	evaluates	that	information,	connects	it	to	reasons,	and	reaches	purposive	and	considered	judgments.		The	second	argument	against	heteronomy	and	self-deception	is	theoretical	and	aims	to	respond	to	the	hardest	test	of	the	value	of	autonomy	and	epistemic	validity.	Imagine	that	the	presence	and	effects	of	self-deception	and	heteronomy	were	both	randomly	distributed	among	participants	in	a	deliberation,	that	there	was	no	structural	bias	against	any	group,	and	still,	the	outcome	of	the	process	was	guaranteed	to	have	a	better-than-random	probability	of	reaching	a	good	result.	What,	then,	would	be	the	problem	with	self-deception	and	heteronomy?	The	empirical	concerns	raised	above	would	no	longer	apply	since	the	process	would	be	guaranteed	to	be,	at	least,	equal	to	a	process	that	was	fair,	supportive	of	autonomy,	and	included	a	tendency	to	produce	correct,	just,	and	epistemically	valid	outcomes.		 In	this	(extremely	unlikely)	instance,	the	value	of	autonomy	becomes	separated	from	any	particular	outcome	and,	indeed,	leaves	the	realm	of	empirical	concerns	(almost)	all	together	since	we	have	stipulated	that	agents	in	this	hypothetical	(and,	again,	highly	unlikely)	scenario	marked	by	heteronomy	and	self-deception	would	be	functionally	as	sound	as	those	who	were	autonomous	and	fully	aware	of	their	reasons	and	their	motivations	for	having	those	preferences.	Here	we	need	to	consider	the	inherent	value	of	self-determination.	A	significant	part	of	the	value	of	autonomy	and	non-self-deception	is	derived	from	the	tendency	we	have	as	humans	to	desire	self-determination	as	an	organizing	principle	for	our	lives,	both	individually	and	collectively,	separate	from	whatever	other	functions	it	serves.	To	the	extent	that	one	is	heteronomously	directed	and	self-deceived,	she	cannot	be	said	to	be	self-determining,	since,	by	definition,	heteronomy	implies	the	determining	work	of	an	outside	force;	therefore,	self-		 	82	deception	is	logically	inconsistent	with	self-determination,	which,	by	definition,	requires	self-awareness	as	well	as	the	opportunity	to	put	into	practices	one’s	desires,	goals,	and	so	forth.	Of	course,	complete	autonomy—what	is	known	as	maximal	autonomy	and	is	defined	as	radical,	boundless,	and	fully-independent	self-creation	(Berofsky	1995)—is	an	impossible	and	even	counter-productive	standard:	the	capacity	to	define	and	develop	oneself	outside	of	any	external	determining	forces	is	an	illusion,	and	it	is	not	clear	that	it	would	be	desirable	if	it	were	even	possible.	It	does	not	take	much	to	demonstrate	how	such	an	approach	to	understanding	autonomy	quickly	becomes	incoherent:	after	all,	we	are	born	into	a	time	and	place,	each	of	which	has	its	own	social,	political,	and	technological	context	that	begin	to	shape	us	from	the	moment	we	are	born—and	in	some	ways,	even	before	that.	We	are	also	born	with	unique	bodies	prone	to	different	affective	states	and	imbued	with	diverse	mental	and	physical	abilities.	These	realities	conspire	with	others—local	and	global,	cultural	and	biological—to	generate	broad	paths	for	us	to	trek	as	we	develop.	We	cannot	choose	to	live	in	a	neutral	environment,	nor	can	we	choose	neutral	minds	or	bodies,	and	so	the	concept	of	maximal	autonomy	fails	to	allow	for	a	definition	of	autonomy	that	preserves	any	hope	of	reaching	a	functional	level	of	behavioural	correspondence	to	the	definition.		 However,	returning	to	Christman’s	more	constrained	definition	of	autonomy	as	the	rational,	non-self-deceptive,	acceptance	of	the	conditions	of	self-development	and	choice,	we	can	see	that	this	functional	understanding	autonomy	allows	for	self-determination	while	also	respecting	that	some	external	determination	is	necessary.	It	is	the	process	of	autonomous	thought,	the	capacity	to	rationally	explain	and	reflexively	accept	or	deny	our	preferences	or	judgments	that	makes	us	autonomous	and,	thus,	self-determining.	Without	this	capacity,	one	of			 	83	the	important	(though	neither	necessary	nor	sufficient)	elements	of	what	makes	us	human—and	deserving	of/capable	of	engaging	with	and	maintaining	liberal	democratic	governments—is	fundamentally	undermined.		So,	an	absence	of	self-deception	and	heteronomy,	and	by	implication,	the	presence	of	autonomy,	is	necessary	for	deliberation—and	elsewhere—for	(at	least)	two	sets	of	reasons.	First,	because	we	have	good	reason	to	believe	that	autonomous	decisions	are	more	likely	to	be	epistemically	valid	and	well-reasoned	than	those	based	on	self-deception	and	heteronomy,	and	because	they	are	also	less	likely	to	include	structural	bias	both	among	specific	groups	of	deliberators	(who	suffers	from	structural	bias)	and	among	those	who	are	affected	by	decisions	conditioned	by	such	bias	(whom	such	suffering	affects	through	decisions	that	are	made	by	the	group	in	question).	And,	second,	self-deception	and	heteronomy	threaten	self-determination	and	what	individuals	tend	to	want	for	themselves:	the	ongoing	opportunity	to	remain	appraised	of	the	relevant	information	that	has	some	a	bearing	on	that	process	of	self-determination.	However,	as	we	will	see	in	chapters	three	and	four,	structural	cognitive	distortions—which	are	common	and	pervasive—are	a	threat	to	autonomy	and	the	power	of	deliberation	to	generate	correct,	epistemically	valid	judgments	and	decisions.		Self-determination and democracy 	Having	discussed	self-determination	as	it	relates	to	the	individual	and	how	he	directs	his	life,	it	is	worth	saying	a	little	bit	about	why	theorists	of	democracy	in	general	and	deliberative	democracy	in	particular	should	care	about	self-determination	insofar	as	it	related	to	democratic	self-determination	collectively.	While	the	principle	is	often	taken	for	granted	as	important	and			 	84	desirable	for	a	democracy,	that	does	not	exempt	us	from	the	need	to	understand	why	and	to	what	end.	In	the	context	of	democratic	theory,	self-determination	as	a	guiding	principle	recognizes	the	inherent	tendency	for	individuals	who	must	live	together	in	a	bounded	space	to	be	different	from	one	another	and	to	disagree.	Even	individuals	raised	in	homogenous	communities	show	variation	and	unique	identities,	and	they	come	into	conflict	with	one	another.	Self-determination	thus	scales	nicely	from	the	level	of	the	individual	who	can	and	does—and	usually	seems	to	want	to—be	the	author	of	their	life	to	that	of	the	community	(or	city,	province,	or	country)	that	wants	to	do	the	same.	Historically,	with	few,	if	any,	enduring	exceptions,	individuals	and	communities	have	tended	towards	self-determination—though	to	varying	degrees	of	success	and	often	only	after	much	(indeed,	ongoing)	struggle.	As	sceptical	as	we	should	be	about	claims	about	“human	nature,”	one	thing	is	certain—when	given	a	chance,	human	beings	tend	to	strive	for	individual	or	collective	(or	both)	self-determination,	however	inter-subjectively	bounded	that	self-determination	may	be.	Moreover,	at	least	in	the	modern	age,	democracy	has	often	been	the	superior	mode	of	organization	for	enabling	broad	self-determination	among	groups	of	people	with	persistent	and	divergent	interests	and	preferences	(though	not	universally	so);	and	within	the	context	of	contemporary	democracy,	the	(individual)	capacity	for	self-determination	(enabled	in	part	by	the	capacity	for	autonomy)	enables	democratic	citizenship	by	giving	individuals	a	chance	to	be	active	participants	in	self-government	rather	than	passive	objects	of	governance.	Deliberative	democracy	as	a	form	of	democratic	practice	in	decision	making	is	particularly	well-suited	to	self-determination	because	it	gives	individuals—whether	citizens	or	their	elected	representatives—a	chance	to	substantively	participate	in	exchanging	reasons	for	and	against	a			 	85	range	of	preferences,	and	to	use	their	capacities	for	autonomy	to	generate	judgments	in	the	pursuit	of	self-determination.	Also,	while	allowing	individuals	to	practice	self-determination	through	reason-giving,	deliberation	also	requires	that	participants	recognize	one	another	as	participants,	thus	setting	up	a	relationship	in	which	they	are	both	recognizers	of	and	recognized	as	active	citizens	in	the	pursuit	of	self-determination	at	both	the	individual	and	collective	levels.	Through	this	process,	individuals	are	not	only	recognized	as	citizens,	but	also,	as	Arendt	or	Kant	might	suggest,	as	human	beings.	The	reach	of	the	principle	of	self-determination	is	potentially	quite	far.	As	Nedelsky	and	Beiner	put	it	“If	it	can	be	shown…that	the	quality	of	our	experience	atrophies	in	proportion	as	we	passively	yield	to	the	judgments	of	others	and	cede	greater	and	greater	dimensions	of	political	responsibility,	then	we	would	have	powerful	reasons	to	believe	that	active	citizenship	is	a	major	component	of	the	human	good”	(2001:	ix).	While	it	is	outside	of	the	scope	of	this	dissertation	to	fully	pursue	that	“if,”	it	seems	intuitively	true	that	this	is	indeed	the	case.	Moreover,	as	noted	above,	human	history	seems	to	bear	out	this	point	in	many,	if	perhaps	not	all,	cases.	To	Nedelsky	and	Beiner’s	point	I	would	add	that	active	citizenship	is	not	just	a	component	of	citizenship	or	the	human	good,	but	an	important—though	not	necessary—component	of	human	life	as	such,	at	least	insofar	as	it	creates	space	for	pursuing	and	realizing	self-determination.27	                                                27	In	chapter	five	in	a	section	entitled	“Tradeoffs	and	limits:	Who	can	and	who	cannot	deliberate”	I	note	that	a	focus	on	democratic	deliberation	in	general	and	on	autonomous	deliberation	as	I	have	conceived	of	it	in	particular	means	that	certain	individuals	may	be	excluded	from	my	analysis	(e.g.	due	to	severe	cognitive	disability).	In	that	section	I	argue	why			 	86	Relational autonomy 	Now,	returning	to	autonomy,	it	is,	I	have	argued,	a	capacity	exercised	through	a	process	(Nedelsky	2001,	2012).	A	person	who	is	exercising	their	capacity	for	autonomy	can	account	for	their	motivations	and	can	give	valid	reasons	for	their	choices	when	reaching	a	judgment	(or	taking	action).	To	the	extent	that	they	are	unable	to	do	so—either	because	they	cannot	give	any	reason	or	because	the	reasons	they	give	are	not	the	reflections	of	actual	motivations	or	reasons—they	are	less	autonomous	and	less	self-determining.	But,	so	far	my	discussion	of	autonomy	has	been	rooted	in	individual	behaviour	and	cognition;	and	yet,	as	I	have	begun	to	show	in	this	chapter,	and	will	elaborate	upon	in	chapters	three	and	four,	so	much	of	human	behaviour	and	cognition	is	socially	determined.	How,	then,	can	an	understanding	of	individual	autonomy	as	I	conceive	of	it	be	reconciled	with	the	reality	of	human	life	as	the	product	of	a	myriad	of	day-to-day	relationships	carried	out	over	a	lifetime?		 The	work	of	Jennifer	Nedelsky	on	relational	autonomy	helps	bridge	this	gap.	As	she	argues,	“What	makes	autonomy	possible	is	not	being	independent	of	all	others,	but	constructive	relationships—with	parents,	teachings,	friends,	colleagues	and	officials	of	the	state.	Autonomy	is	thus	also	not	a	characteristic	that	we	simply	achieve…its	flourishing	depends	on	the	kinds	of	relationships…of	which	we	are	a	part”	(Nedelsky	2001:	111;	2012).	She	lists	“biases,	fears,	emotions	that	cloud	rather	than	facilitate	judgment”	(111)	as	challenges	that	emerge	from	a	focus	on	“private	considerations.”	Indeed,	she	argues,	echoing	Kant,	that	to	the	extent	that	these	or	similar	factors	drive	our	“judgments,”	we	are	not	making	judgments	at	all,	                                                this	fact,	while	being	far	from	ideal,	does	not	present	a	significant	challenge	to	what	I	am	attempting	to	argue	in	this	dissertation.			 	87	but	have,	rather,	a	certain	“emotion	or	conception”	(111)	masquerading	as	a	judgment.	She	prescribes	taking	multiple	perspectives	when	forming	judgments	to	combat	the	deleterious	and	autonomy-reducing	effects	(including	the	various	dimensions	of	the	self).			 Nedelsky’s	conception	of	autonomy	as	a	relationally-bound	capacity	is	complementary	to	and	consistent	with	the	conception	of	autonomy	that	I	use	in	this	dissertation.	While	my	conception	of	autonomy,	taken	from	Christman,	is	an	individually-executed	process	of	judgment	formation	and	affirmation/review,	Nedelsky’s	relational	understanding	of	autonomy	both	complements	and	contextualizes	the	one	I	use.	Her	understanding	of	relational	autonomy	is	based	on	autonomy	being	enabled	by	social	relations;	still,	it	complements	Christman’s	conception	of	autonomy	by	providing	a	perspective	that	can	be	taken	within	the	process	of	reviewing	one’s	preferences.	For	instance,	when	I	review	whether	some	particular	judgment	is	(truly)	autonomously	reached,	I	can	employ	multiple	perspectives	as	tests	designed	to	interrogate	my	motivations.	Indeed,	as	I	will	argue	in	chapter	five,	there	are	specific	tactics	that	can	be	used	to	do	just	this	within	a	deliberative	setting.	Moreover,	Nedelsky’s	understanding	of	autonomy	as	relational	also	contextualizes	my	understanding	of	autonomy	as	an	individual	process	of	self-reflection	and/or	procedural	checks	and	balances	during	the	process	of	forming	a	judgment.	Specifically,	Nedelsky	characterizes	autonomy	as	inherently	bound	up	in	relationships	with	others—so	that	the	individual	is	socially-bound	and	constituted	(2012).	Accordingly,	her	perspective	provides	another	argument	against	the	maximal-autonomy	perspective	of	radical,	unbounded	self-determination	while	revealing	tactics	for	enhancing	an	individual	capacity	for	autonomy.			 	88	That	said,	there	is	a	difference	between	these	two	conceptions	of	autonomy—they	are	not	wholly	interchangeable.	Ultimately,	for	this	dissertation,	I	think	of	autonomy	as	a	process	conditioned	collectively	but	expressed	individually.	Its	parameters	are	set	externally	and	inter-subjectively,	but	it	is	the	individual	who	must	act	autonomously	or	not,	which	may	not	always	require	others.	Granted,	by	taking	the	perspectives	of	others,	it	may	help	the	individual	to	uncover	their	motivations;	but,	in	the	end,	autonomy	requires	that	it	be	the	individual	who	is	practicing	their	capacity	for	autonomy	(or	reaching	it	more	or	less	relative	to	a	continuum—which	is	based	on	an	approach	for	measuring	autonomy	that	I	will	outline	in	chapter	five)	or	not.	So,	while	autonomy	may	be	relational	in	terms	of	its	broad	constitution,	the	capacity	itself	is	reached	and	expressed	individually—which	is	my	concern	in	this	dissertation.	I	do	not	think	this	limit	is	inconsistent	with	Nedelsky’s	conception	of	autonomy;	indeed,	if	anything,	each	of	the	two	conceptions	makes	the	other	more	complete	and	useful	for	understanding	how	human	beings	reach	better	or	worse	judgments.		A note on the philosophical debate between free will and determinism 	In	closing	this	chapter,	I	want	to	note	that	a	discussion	of	epistemic	autonomy	and	heteronomy	in	deliberative	contexts,	specifically	as	each	relates	to	judgments	and	decisions,	is	adjacent	to	the	broader	discussion	of	the	causal	properties	of	free	will	versus	determinism	(not	self-determination)	in	an	ontological	sense.	As	cognitive	neuroscientist	Michael	Gazzaniga	defines	it,	“…determinism	is	the	philosophical	belief	that	all	current	and	future	events,	actions,	including	human	cognition,	decisions,	and	behavior	are	causally	necessitated	by	preceding	events	combined	with	the	laws	of	nature.	The	corollary,	then,	is	that	every	event,	action,	et	cetera,	is			 	89	predetermined	and	can	in	principle	be	predicted	in	advance,	if	all	parameters	are	known”	(Gazzaniga	2011:	111).		 Whatever	the	merits	or	demerits	of	either	camp’s	position	concerning	free	will	and	determinism,	the	question	of	what	sorts	of	immediate	cognitive	factors	go	into	producing	an	individual	judgment	or	collective	decision	can	be	sectioned	off	as	a	separate	issue,	perfectly	consistent	with	either	perspective,	and	interrogated.	That	is	because,	as	I	have	briefly	touched	on,	and	as	I	will	discuss	in	greater	depth	in	chapters	five	to	seven,	decisions	can	be	made	to	come	from	more	or	from	less	autonomous	agents,	or	from	more	or	less	self-deceived	agents,	depending	on	factors	such	as	education,	self-awareness,	attention	to	detail,	motivation	to	reflect	and	engage	in	thought,	the	structure	and	composition	of	deliberations,	and	so	forth.	Whatever	the	macro-antecedent	causes	of	the	states	in	which	judgments	or	decisions	are	made,	those	within	those	states,	those	making	decisions	and	judgments,	can	be	reliably	evaluated	as	being	more	or	less	autonomous	according	to	a	prescribed	standard;	and,	moreover,	they	can	be	made	more	or	less	autonomous	(as	I	will	argue	in	chapter	five)	through	changes	to	personal	practices	and	institutional	arrangements.	Thus,	the	level	of	analysis	with	which	this	dissertation	is	concerned—the	individual	as	a	knowing	and	thinking	agent	in	a	deliberative	context—is	of	interest	independently	of	any	broader	ontological	questions	about	free	will	and	determinism	as	discussed	above.	To	the	extent	that	I	am	interested	in	freedom,	it	is	of	a	variety	very	similar	to	that	conceived	by	Kant—the	replacement	of	non-rational	causes	of	behaviour	(e.g.	external	forces,	subterranean	desires)	with	causal	reasons,	though	in	my	conception	I	make	more	room	for	“desires”	and	other	affective	considerations	as	legitimate			 	90	grounds	for	reasons,	so	long	as	an	individual	can	communicate	their	actual	reasons	for	reaching	a	judgment	or	for	taking	an	action.		Conclusion and summary 	In	this	chapter,	I	argued	that	deliberative	democracy	is	a	theory	of	how	ordinary	citizens	can,	among	other	things,	make	epistemically	good	decisions	in	a	fair	way	(i.e.	a	way	that	is	impartial	among	participants	and	generally	acceptable	to	all	who	engage	in	democratic	deliberation)	on	a	better-than-chance	basis.	I	also	argued	that	most	of	theories	of	deliberative	democracy	require	autonomous	agents	to	cash	out	these	promises,	but	that	autonomy	stands	out	as	the	“keystone	capacity”	required	for	good	democratic	deliberation.	I	have	used	Christman’s	definition	of	autonomy	as	the	capacity	of	an	agent	to	generate	judgments	that	he	or	she	did	not	or	would	not	have	resisted,	in	a	theoretically	self-reflective,	non-self-deceptive,	and	rational	way,	which	I	think	is	the	most	appropriate	conception	of	autonomy	for	this	dissertation.	Furthermore,	I	explored	Estlund’s	argument	that	deliberative	democracy	is	defensible	on	the	grounds	that	it	does	indeed	fulfill	this	epistemic	process	through	“epistemic	proceduralism”;	in	the	process,	I	have	also	distinguished,	using	the	work	of	Estlund	and	Warren,	between	expertise	in	deliberation	and	ordinary	citizen	deliberation.	I	also	compared	my	conception	of	autonomy	with	the	(complementary)	conception	offered	by	Nedelsky—relational	autonomy.	Finally,	I	distinguished	autonomy	(and	self-determination)	from	free-will	and	determinism,	noting	that	I	am	interested	here	in	the	former	and	not	the	latter.		  		 	91	Chapter 3: Challenges to autonomy in democratic deliberation  	The	problem	of	autonomy,	as	it	relates	to	deliberative	democracy	as	an	epistemic	theory	of	better	judgment	and	democratic	decision	making,	arises	from	a	more	fundamental	problem	that	runs	through	many,	if	not	most,	of	our	social,	political,	and	economic	systems:	there	is	a	significant	gap	between	our	nature	and	our	culture—between	what	our	capacities	tend	to	allow	us	to	accomplish,	and	what	is	normatively	expected	of	us.	This	claim,	when	made	in	a	constrained	sense,	is	both	intuitively	appealing	and	empirically	demonstrable,	but	is	often	treated	as	more	controversial	than	it	ought	to	be.	Such	controversy	is	mostly	misguided,	at	least	when	the	nature/culture	gap	argument	appears	in	the	more	constrained	sense	by	which	I	mean	it	in	this	dissertation.	Much	of	the	controversy	surrounding	this	gap,	I	believe,	stems	from	poorly	defined	or	underdefined	terminology	used	by	those	who	make	this	argument,	or	because	definitions	of	either	nature	or	culture	that	tend	to	either	overreach	or	over-specify	what	they	can	reliably	and	empirically	demonstrate.	I	will	try	to	avoid	this	common	problem	while	I	characterize	this	gap	before	moving	on	to	discuss	precisely	how	it	impacts	the	possibility	of	autonomous	deliberation	in	contemporary	liberal	democracies.	When	I	say	that	there	is	a	gap	between	nature	and	culture,	I	simply	mean	that	when	it	comes	to	certain	behaviour,	practices,	norms,	and	imperatives,	such	as	democratic	deliberation,	what	is	generally	expected	of	us	and	what	we	can	reasonably	be	expected	to	deliver	tend	to	regularly	and	systematically	diverge.	This	gap	emerges	because	the	expectations	and	standards	we	adopt	often	outpace	our	ability	to	meet	them.	On	balance,	this	is	an	effect	of	our	bounded	rationality	and	the	significant	impact	of	a-rational	modes	of	cognition	on	our	day-		 	92	to-day	lives	(Heath	2014).	By	“nature”	I	am	referring	simply	to	the	constrained	evolutionary	cognitive	capacities	that	most	human	beings	have	developed	and	the	behavioural	tendencies—extremely	well	documented	by	social	and	cognitive	psychology,	sociology,	political	psychology,	economics,	and	other	fields—that	tend	to	be	concomitant	with	those	capacities.	I	do	not	mean	to	suggest	that	human	nature	is	universal	and	absolute	or	eternally	fixed	or	fully-determining;	instead,	I	mean	to	suggest	that	our	nature	is,	fundamentally,	a	condition	that	limits	the	range	of	behaviours	that	are	possible	for	us,	sets	certain	cognitive	limitations	that	may	exist	prior	to	our	expectations	of	behaviour,	and	that	tends	to	yield,	on	balance,	certain	predictable	behavioural	patterns	or	responses	in	the	presence	of	specific	stimuli	or	contexts,	regardless	of	what	our	cultural	expectations	might	be.		By	“culture”	and	institutional	arrangements	I	am	referring	broadly	to	the	malleable	(though	they	tend	to	change	only	with	time	and	effort)	time	and	space	specific	sets	of	practices,	rules,	norms,	expectations,	and	standards	that	emerge	from,	among	other	things,	evolutionary	imperatives,	biological	necessities,	historical	particularities,	ideational	supporting	structures,	and	material	circumstances.	Whatever	the	origin	of	expectations	that	emerge	from	these	structures,	there	is	no	a	priori	guarantee	that	such	expectations	will	align	with	our	natural	abilities	in	every	case	or	instance	since	what	we	imagine	or	expect	human	beings	to	be	capable	is	not	necessarily	the	same	as	what	we	are	actually	capable	of.	This	is	the	nature/culture	divide	that	frames	and	informs	this	chapter,	and	indeed	much	of	this	dissertation.	Regardless	of	whatever	causal	mechanisms	might	be	at	work	in	producing	these	circumstances,	regardless	of	the	balance	of	causal	force	between	idealism	and	materialism,	free	will	and	determinism,	and	other	macro-causal	theories,	it	is	indisputable,	and	unremarkable			 	93	that	the	possibilities	and	limitations	of	our	minds	have	an	impact	on	our	behaviour,	including	our	expectations.	The	specific	case	with	which	I	am	concerned	in	this	dissertation	is	the	potential	normative	overreach	by	theorists	of	deliberation	who	implicitly	or	explicitly	rely	on	an	account	of	human	agency	and	cognition	that	tends	not	to	align	with	agents	typically	deliver.	I	am	equally	concerned	with	the	question	of	what	we	might	do	about	this	in	order	to	contribute	to	generating	better	political	judgments	and	decisions—that	is,	judgments	and	decisions	that	come	closer	to	meeting	the	high	standards	and	expectations	we	set	for	ourselves.	In	this	chapter,	to	demonstrate	how	this	overreach	affects	deliberative	democracy,	I	will	do	four	things.	First,	I	will	outline	what	theorists	of	deliberative	democracy	expect	from	deliberative	agents,	specifically	focusing	on	the	normative	content	of	those	expectations—i.e.	what	is	desired	of	deliberation	and	those	who	deliberate.	Second,	I	will	summarize	and	analyze	a	broad	theory	of	cognition	drawn	from	social	and	evolutionary	psychology—systems	theory—which	divides	cognition	into	two	streams	(known	as	a	dual-process	theory).	The	fields	of	social	and	evolutionary	psychology	are	concerned	with,	among	other	things,	showing	that	certain	cognitive	limitations	are	built	into	the	structures	of	human	brains	and	are	expressed	as	thought	patterns	while	tending	to	present	themselves	in	certain	behavioural	patterns.	This	phenomenon	is	perhaps,	though	not	always,	influenced	by	culture	in	certain	expressions,	but	generally	independent	of	it.	Looking	at	dual	process	theories	in	general,	it	becomes	apparent	that	certain	challenges	are	implicit	in	many	cognitive	undertakings,	including	democratic	participation	in	general	and	deliberation	in	particular.	Third,	I	will	lay	out	four	leading	models	of	cognition	from	social	psychology	that	demonstrate	some	of	the	specific	behavioural	implications	of	social	and	evolutionary	psychology	in	the	context	of	contemporary	societies	in			 	94	general	and,	more	specifically,	in	deliberative	contexts	(I	will	tie	them	back	to	the	systems	theory	when	necessary).	The	first	model	is	Richard	E.	Petty	and	John	T.	Cacioppo’s	elaboration	likelihood	model	(ELM);	the	second	is	John	Jost’s	system	justification	model;	the	third	is	Jonathan	Haidt’s	account	of	moral	intuitionism,	and	the	final	is	John	Bargh	and	Tanya	Chartrand’s	automaticity	model.	Finally,	I	will	conclude	by	summarizing	the	implications	for	deliberative	democracy	that	are	generated	by,	in	general,	our	evolutionary	cognitive	capacities	and,	specifically,	by	the	four	models	examined	in	this	chapter	and	suggest	how	theorists	and	practitioners	of	deliberation	might	respond	to	the	challenges	that	emerge	from	those	implications.	While	these	challenges	to	autonomous	deliberation	and	good	judgment	and	decision	making	are	significant	and	entrenched,	I	think	they	can	be	addressed	in	productive	ways.	Ultimately,	this	dissertation	is	about	highlighting	cognitive	challenges	to	deliberation	revealed	by	research	in	political	and	social	psychology;	it	is	also	about	developing,	conceptually,	personal	practices	and	approaches	to	institutional	design	to	address	those	challenges	head	on.	In	that	way,	I	think	the	gap	between	our	nature	and	our	culture—at	least	when	it	comes	to	democratic	deliberation—can	be	bridged	in	a	way	that	will	produce	better	judgments	and	decisions.		A note on epistemology, the brain, and our environment	This	chapter	is	rooted	in	an	epistemological	understanding	of	the	human	being	and	is	not	concerned	with	ontological	explorations.	I	am	interested	in	how	we	come	to	know	what	we	know,	both	individually	and	collectively,	and	how	that	affects	collective	knowledge,	judgment,	and	decision	making.	Because	this	epistemological	approach	is	so	central	to	this	dissertation,	it			 	95	is	worth	saying	a	bit	more	about	it.	To	do	so,	in	this	section,	I	will	draw	on	the	work	of	biologist	Gerald	Edelman	to	discuss	this	approach	that	underlines	this	project.	More	precisely,	I	take	as	my	point	of	departure	his	concept	of	“brain-based	epistemology”	(Edelman	2006).	According	to	Edelman,	this	approach	“…refers	to	efforts	to	ground	the	theory	of	knowledge	in	an	understanding	of	how	the	brain	works”	(Edelman	2006:	2).	His	interest	in	epistemology	and	cognition	is	largely,	unlike	my	own,	concerned	with	consciousness.	However,	insofar	as	he	is	concerned	with	how	epistemology	relates	to	cognition	within	a	triad	that	includes	brain,	body,	and	environment,	his	approach	to	understanding	how	we	come	to	know	things,	given	the	types	of	brains	we	have	evolved,	is	useful	for	the	matter	at	hand.	Indeed,	such	an	approach	both	allows	for	a	broader	enquiry	into	knowing	that	allows	space	for	examining	ideational,	biological,	and	environmental	factors,	and	we	can	place	it	squarely	and	soundly	within	the	traditions	of	both	evolutionary	psychology	and	evolutionary	biology,	which	are	useful	for	understanding	how	(and	why)	our	nature	and	culture	might	have	split	in	certain	areas.	This	is	important	to	keep	in	mind	in	reading	this	chapter	given	that	the	underlying	theory	of	epistemology	that	supports	my	arguments	is	one	concerned	with	an	interaction	between	brain/psychology,	bodily	processes,	and	the	environment:	the	kind	of	general	context	that	deliberators	will	tend	to	find	themselves	in	when	faced	with	generating	political	judgments	and	decisions.	If	we	can	understand	our	thinking	and	reasoning	as	guided	by	several	sorts	of	conditions,	then	we	can	begin	to	unravel	how	to	best	generate	knowledge,	share	it,	reach	our	own	judgments,	and	come	to	share	those	judgments	with	others	to	make	decisions—which	is	what	democratic	deliberation	is	all	about.					 	96	What is expected and required of deliberative agents? 	Theories	and	theorists	of	deliberative	democracy	tend	to	both	assume	and	require	autonomous	and	rational	agents	capable	of	exchanging	reasons	for	their	political	preferences.	Participants	in	a	deliberation	are	also	expected	to	be	open	to	being	swayed	by	the	force	of	the	better	argument	(as	Habermas	puts	it).	In	general,	when	we	consider	deliberation	as	a	better—or	the	best—way	of	making	decisions,	some	combination	of	democratic	goods	and	normative	preferences	are	included	as	justification	for	this	claim—for	instance,	a	better-than-random	chance	at	producing	good	decisions,	the	increased	political	legitimacy	of	those	decisions,	respect	for	the	agency	of	individuals,	the	minimization	of	the	potential	for	violence	in	politics,	and	so	on.	These	goods	and	desires	are	what	I	am	calling	the	normative	expectations	of	deliberation,	since	these	are	outcomes	or	justifications	that	are	tied	to	a	particular	place,	time,	and	theoretical	program:	namely,	a	philosophical	tradition	with	its	origins	in	the	European	Enlightenment	and	developed	in	the	context	of	twenty-first-century	liberal	democracy.	As	noted	above,	the	very	fact	that	such	expectations	are	normative	in	a	broad	sense	means	that	they	are	artifacts	subject	to	limits,	and	there	is	no	a	priori	guarantee	that	their	realization	is	possible,	probable,	or	even,	ultimately	desirable	given	what	I	am	calling	natural	constraints—constraints	based	on	cognitive	or	other	biological	particularities.	In	the	case	of	this	dissertation	and	the	argument	I	am	making,	the	context	is	twenty-first-century	liberal	democracies	marked	by	diversity,	pervasive	disagreement,	speed	of	communication,	the	complexity	of	our	social	and	political	systems,	and	the	status	of	certain	capacities	(e.g.	rationality	and	autonomy).	Regardless	of	whatever	challenges	are	posed	to	theories	of	deliberation	by	cognitive	limitations,	the	desired	goods	themselves	can	be	evaluated	amongst	one	another	against	a			 	97	standard	of	internal	coherence—whether	the	requirement	of	each	is	consistent	with	the	presence	of	the	others.	For	many—likely	most—theorists	of	deliberation,	at	least	four	general	categories	of	democratic	goods	and	normative	standard	can	be	discerned	from	their	theories	of	deliberative	democracy:	the	expression	of	rational	judgments	and	decisions,	the	provision	of	space	for	the	practice	of	self-determination	and	autonomy,	the	generation	of	valid	reasons	for	political	judgments	and	decisions,	and	the	production	of	more	legitimate	political	outcomes.	These	can	be	further	broken	down	into	sub-categories,	with	the	former	two—rationality	and	self	determination/autonomy—relating	to	specific	capacities	of	deliberators,	and	the	latter	two	relating	to	functions	of	the	system	of	deliberation.	All	four	taken	together	produce	an	argument	about	the	value	of	deliberation	as	such.		The expression of rationality and self-determination/autonomy 	Let	us	look	first	at	each	of	the	two	individual-level	assumptions	and	requirements	of	theories	of	deliberation:	rationality	and	self-determination/autonomy.	Each	requirement	implies	assumptions	about	the	cognitive	capacities	of	individuals	who	deliberate.	Theories	of	deliberation	assume	rational	agents,	though	this	assumption	unless	very	narrowly	and	unfairly	construed,	does	not	imply	that	agents	follow	a	strict	rational	choice	framework	of	optimal	ends-means	rationality—an	approach	that	is	better	suited	to	aggregative	theories	of	democracy	(Warren	2002).	Instead,	theorists	of	deliberative	democracy	generally	stipulate	that	deliberators	ought	to	give	publicly	accessible	reasons	for	their	preferences	that	they	are	able	and	willing	to	defend,	drawn	from	a	more-or-less	shared	world	(Benhabib	1996;	Chambers			 	98	1996;	Dryzek	2000;	Elster	1997;	Gutmann	and	Thompson	1996,	2004;	Warren	2002).	Gutmann	and	Thompson	(2004)	further	specify	this	requirement	by	insisting			…any	premises	in	the	argument	[made	within	deliberation]	that	depend	upon	empirical	evidence	or	logical	inference	should	in	principle	be	open	to	challenge	by	generally	accepted	methods	of	inquiry….	[and]	premises	for	which	empirical	evidence	or	logical	inferences	is	not	appropriate	should	not	be	radically	implausible	(72).		Thus,	the	rational	citizen	who	deliberates	is	not	the	ends-means	calculative	machine	that	some	might	imagine.	Instead,	she	is	a	citizen	who	has	the	capacity	to	form,	present,	revise,	defend,	or	reject	preferences	and	judgments	based	on	publicly	justifiable	reasons	that	are,	at	least	in	principle,	subject	to	interrogation	and	evaluation	by	her	peers,	and	that	are	not	a	priori	implausible	or	irrelevant.	These	preferences	and	judgments	should	be	more-or-less	coherent	and	drawn	from	the	world	in	such	a	way	that	aims	to	make	sense	of	that	world.	As	mentioned	earlier,	this	definition	of	rationality	is	not	a	particularly	onerous	one,	but	it	is	functional	and	sufficient	for	my	purposes	in	this	dissertation.	It	is	important	to	keep	in	mind	that	for	this	dissertation,	I	am	referring	primarily	to	deliberative	rationality,	which	is	bound	by	the	requirements	that	individuals	in	a	deliberative	process	draw	on	genuine	reasons	when	arguing	and	defend	their	reasons,	preferences,	judgments,	and	so	on	to	others.	This	approach	contrasts,	for	example,	with	ends-means	rationality	in	which	it	would	be	rational	to	engage	in	strategic	behaviour,	to	lie,	to	withhold	key			 	99	information,	and	so	forth	to	achieve	a	particular	outcome.	In	a	deliberation,	this	behaviour	would	be	irrational	since	the	goals	of	the	deliberation	are	founded	on	an	intersubjective	and	normative	considerations	aimed	at	establishing	better	judgments	and	decisions	through,	among	other	things,	exchanges	based	on	epistemically	valid	statements.			 Let	us	expand	on	the	point	of	deliberation	prohibiting	a	priori	implausible,	irrational,	or	irrelevant	reasons.	A	rational	deliberator	is	concerned	with	rational	considerations.	In	his	entry	on	deliberation	in	the	Oxford	Handbook	of	Political	Philosophy,	Robert	B.	Talisse	gives	the	example	of	Abby,	who	is	deciding	between	attending	a	film	or	visiting	the	library	(Talisse	2012:	204).	She	can	flip	a	coin	to	decide	or	she	can	weigh	the	pros	and	cons	of	each	option.	If	she	chooses	the	coin	toss,	she	is	choosing	an	internally	irrational	process	(which	may	be	perfectly	externally	rational,	if	she	really	cannot	decide).	That	is	because	the	coin	toss	is	blind	to	reasons;	if,	however,	she	deliberates	over	the	decision,	she	can	generate	and	offer	reasons	for	and	against	each	option:	an	internally	rational	process.	Deliberative	democracy	is	the	public	extension	of	this	logic	of	rational	deliberation;	it	does	not	presuppose	specific	ethical	content	for	decisions—at	least	outside	of	very	general	principles	related	to	liberal	democracy	more	broadly.	However,	it	does	presuppose	a	rational	and	deliberate	process	aimed	at	generating	coherent	judgments	and	decisions	in	a	formal	political	context.	My	dissertation	is	concerned	with	precisely	this:	how	can	we	generate	coherent,	valid	judgments	and	decisions	through	democratic	deliberation.						 	100	Self-determination and autonomy  	Next,	let	us	look	at	self-determination	and	autonomy,	each	of	which	I	examined	in	greater	detail	in	chapter	two.	One	of	the	core	justifications	for	the	reason-giving	requirement	of	deliberation	is	that	deliberators	ought	to	be	treated	as	autonomous	agents	capable	of	interacting	with	one	another	towards	some	end	of	their	choosing,	rather	than	as	passive	objects	of	governance	mobilized	for	the	purposes	of	reaching	another’s	end	(Gutmann	and	Thompson	2004:	3-4).	For	this	to	be	possible,	citizens	must	be	autonomous	in	the	broader,	non-cognitive	sense	of	the	term	(recall	that	throughout	the	dissertation	I	discuss	autonomy	in	the	cognitive	sense,	not	the	liberty	sense,	unless	otherwise	noted)—non-cognitive	autonomy	refers	to	the	capacity	to	freely	act	in	such	a	way	that	allows	an	individual	to	be	personally	self-determining	and	a	participant	in	the	determination	of	the	fate	of	their	community.28	This	requires	a	number	of	stipulations	in	the	way	that	deliberations	are	designed,	carefully	set	up	to	ensure	to	the	greatest	degree	possible	that	the	autonomy	of	each	participant	is	respected	to	the	greatest	degree	possible;	as	Estlund	(2008)	puts	it,	deliberations	should	be	set	up	as	to	prevent	“power’s	interference	with	reason”	(193).	Specifically,	this	end	requires	that	both	the	deliberative	design	and	the	behaviour	of	deliberators	be	more	or	less	in	line	with	the	goal	of	respecting	the	autonomy	and	agency	of	each	participant.	This	end	also	requires	institutional	setups	respecting	the	freedom	and	equality	of	each	participant	within	the	context	of	the	                                                28	For	instance,	this	latter	understanding	of	autonomy	is	grounded	in	an	account	of	human	intersubjectivity	discussed	in	chapter	two	in	the	section	on	relational	autonomy.	The	relational	understanding	of	autonomy—consistent	with	but	distinct	from	the	one	I	use	in	this	dissertation—focuses	on	human	beings	as	inherently	bound	up	and	partly	determined	by	communities,	while	my	interest	is	in	the	cognitive	process	that	either	precedes	or	succeeds	specific	instances	of	reflection,	either	alone	or	in	a	group.			 	101	deliberation	and	the	principle	of	non-domination	so	that	each	affected	by	an	outcome	has	a	meaningful	chance	to	engage	in	the	process	that	generates	a	decision.	These	requirements	open	up	space	for	the	possibility	of	self-determination—which	is	defined	here	as	the	possibility	and	practice	of	exercising	(for	now,	non-cognitive)	autonomy	in	the	pursuit	or	realization	of	one’s	preferences,	mirroring,	in	a	slightly	confusing	way,	Rostbøll’s	(2008)	definition	of	autonomy	as	a	process	of	“[living]	under	conditions	where	one	can	engage	with	others	in	deliberative	practices	that	enable	one	continually	to	modify	one’s	preferences	and	opinions	in	light	of	arguments”	(87).	I	have	been	speaking	of	autonomy	in	this	section	in	a	non-cognitive,	traditional	sense.	However,	as	noted	in	chapters	one	and	two,	autonomous	deliberation	requires	the	capacity	for	sustained	autonomy	in	the	cognitive	sense.	To	review,	cognitive	autonomy,	as	opposed	to	autonomy	as	freedom,	requires	that	“…the	influences	and	conditions	that	give	rise	to	the	desire	[or	preference	or	intention]	were	factors	that	the	agent	approved	of	or	did	not	resist,	or	would	not	have	resisted	had	she	attended	to	them,	and	that	this	judgment	was	or	would	have	been	made	in	a	minimally	rational,	non-self-deceived	manner”	(Christman	1991:	22,	emphasis	mine).	The	relationship	between	the	two	senses	of	the	term	is	interesting	but	irrelevant	to	the	argument	I	make	in	this	dissertation.	For	now,	it	is	enough	to	note	that	there	are	at	least	two	distinct	meanings	of	the	term	“autonomy,”	and	to	distinguish	one	(i.e.	autonomy	as	freedom-directed	in	the	non-cognitive	sense)	and	from	another	(i.e.	the	cognitive	sense).	The	latter	adds	a	level	of	analysis	to	the	traditional	understanding	of	autonomy	by	pushing	the	boundaries	of	autonomy	past	the	absence	of	constraints	and	into	to	the	realm	of	cognitive	capacity.	 			 	102	Before	proceeding,	I	must	address	the	question	of	whether	cognitive	autonomy,	as	I	conceive	of	it	here,	is	ever	fully	possible,	or	whether	it	is	an	ideal.	As	I	will	discuss	in	chapter	six,	full	cognitive	autonomy	is	an	ideal,	though	I	think	of	individual	autonomy	as	a	capacity	that	is	exercised	in	the	moment	in	a	better	or	worse	way.	So,	full	autonomy	is	an	ideal,	but	the	exercise	of	one’s	autonomy	falls	along	a	continuum	that	can	be	evaluated	as	more	or	less	autonomous	to	the	extent	that	individuals	are	aware	of	what	drives	them.	My	primary	concern	in	this	dissertation	is	understanding	how	autonomy	can	become	compromised	during	deliberation,	since	insofar	as	it	is	compromised,	the	validity	of	judgments	and	decisions	is	undermined;	my	secondary	concern	is	how	we	can	address	the	challenges	generated	by	compromised	autonomy	and	encourage	better	exercise	of	that	capacity,	whether	or	not	it	ever	reaches	the	full	ideal	of	(theoretical)	total	cognitive	autonomy.		Validity of reasons 	Let	us	now	look	at	the	system-level	assumption	about	the	effects	of	deliberation—or	the	goods	that	theories	of	deliberation	claim	will	be	produced	or	enhanced	through	deliberation	among	rational,	autonomous	agents.	Habermas	(1985),	summarized	by	Warren	(2008:	184),	provides	a	clear	and	precise	definition	of	validity	that	summarizes	well	the	kind	of	decisions	that	proponents	of	deliberative	democracy	contend	that	deliberation	tends	to	produce.	Validity,	Warren	summarizes,	“…is	an	attribute	of	statements	assessed	according	to	whether	they	are	factually	true,	normatively	right	and	expressively	sincere	(or	truthful).”	He	goes	on	to	note	that	statements	influence	others	when	they	are	recognized	as	valid	in	the	three	“worlds”	outlined	by	Habermas:	the	empirical	(the	factual	world),	the	normative	(the	world	of	social	norms	and			 	103	rules),	and	the	internal	realm	of	feelings,	desires	and	thoughts	(the	world	of	internal	experience)	(Warren	2008:	184).		 The	underlying	force	of	validity	rests	on	the	quality	of	valid	reasons	and	arguments—the	idea	that	producing	good	decisions	requires	high-quality	evidence	and	reasons	(Chambers	1996),	corrects	or	avoids	mistakes	while	also	developing	understanding	(Gutmann	and	Thompson	2004),	and	publicly	links	objects	to	their	referents	in	order	to	generate	shared	understanding	(Warren	2008).	Validity	is	thus	a	function	of	the	process	of	generating	agreement	about	the	world	through	the	exchange	of	information	between	two	or	more	individuals.	Facts	about	the	world	are	embedded	in	statements	that	are	either	taken	up	by	others	(or	not),	and	thus	validity	is	established	or	not	in	the	process	of	coming	to	judgments	and	decisions.	When	it	comes	to	democratic	deliberation,	validity	is	not	about	faithfully	translating	facts	about	the	external	world	in	the	pursuit	of	some	ontological	uncovering	of	“reality.”29	Proponents	of	deliberation	claim	that	deliberative	democracy	often	allows	for	the	maximization	of	the	likelihood	that	valid	statements	will	be	made	and	that	a	coherent	shared	world	will	be	generated	by	those	assembled	to	deliberate—more	often	than	other	possible	decision	making	procedures	including	mere	voting,	technocratic	directives,	or	coercion—since	deliberation	is	directly	concerned	with	reasons	as	a	currency	instead	of	other	sorts	influence	such	as	impulse,	expertise,	or	force.				                                                29	For	more	on	this,	see	my	discussion	of	correspondence	theories	of	truth	and	consensus	theories	of	truth	in	chapter	two.			 	104	More legitimate outcomes  	As	a	process	of	decision	making	that	stresses	inclusion,	(non-cognitive)	autonomy,	self-determination,	reason	giving,	and	which	claims	to	produce	more	valid	and	reasoned	outcomes	than	the	alternative	decision	making	processes	noted	above,	proponents	of	deliberative	democracy	also	claim	that	the	outcomes	of	deliberations—or	the	outcomes	of	decision-making	procedures	in	which	deliberative	democracy	plays	a	role—are	more	legitimate	than	other	approaches	(Estlund	2008,	Fishkin	1996,	Warren	2008).		However,	the	core	justification	of	deliberation	as	a	legitimate	decision-making	process	focuses	on	participation	and	freedom.	As	Cohen	(2002)	puts	it,	“…free	deliberation	among	equals	is	the	basis	of	legitimacy”	(91).	So	deliberation—whatever	else	it	accomplishes—acts	as	a	source	of	legitimate	outcomes	through	its	commitment	to	providing	space	for	each	to	engage	and	to	be	a	part	of	the	decision-making	process,	so	that	each	is	bound	to	a	priori	to	outcomes	that	respect	the	inclusive,	free,	and	fair	deliberative	procedures	agreed	to	before	deliberations	begin.	Other	goods,	noted	above,	including	more	valid	outcomes	and	rational	outcomes,	further	support	deliberation.	Thus,	one	of	the	key	goods	generated	by	deliberation	is	a	sense	that	one	is	tied	to	and	bound	up	within	the	decisions	generated	by	deliberation;	the	further	assumption,	largely	borne	out	historically,	is	that	the	more	substantive	and	participatory	a	mode	of	decision	making	is,	the	more	it	will	be	respected	and	adhered	to	by	those	who	participate	in	it	and	are	affected	by	it—which	may	be	closely	intertwined	with	a	broader	desire	for	self-determination.					 	105	From evolutionary and social psychology to deliberation: dual process cognition in the system 1 and system 2 modes 	Consciousness	reigns,	but	doesn’t	govern.	–	Paul	Valéry		The	human	brain—the	site	of	consciousness	and	knowing—is	not	a	computer.	Indeed,	it	is	quite	far	from	the	computing	machine	that	it	is	often	explicitly	stated	or	implicitly	assumed	to	be	when	we	establish	expectations	about	how	individuals	should	engage	in	deliberative	political	discourse.	In	this	section,	I	will	discuss	some	ways	in	which	the	rational	brain	can	help	us	meet	the	standards	set	by	deliberative	theorists	and	some	ways	in	which	it	can,	upon	us	finding	ourselves	in	certain	circumstances	or	interacting	with	certain	institutions,	work	against	meeting	those	standards.However,	first,	I	will	explore	an	extreme	example	of	how	a	purely	rational	brain	might	operate,	the	brain-as-computer,	to	establish	some	parameters	for	when	I	discuss	how	the	brain	tends	to	operate.		Why	is	this	necessary?	Because	the	way	we	conceive	of	the	brain	is	linked	to	how	we	conceive	of	thinking.	Metaphors	and	analogies	are	powerful	and	it	is	important	that	we	understand	the	fundamental	architecture	of	thought	from	the	brain	to	behaviour.	Because	our	normative	expectations	are	conditioned	by	our	conceptions	about	how	we	think	and	how	we	ought	to	think,	we	should	pay	close	attention	to	those	conceptions;	and	since	they	are	rooted	in	metaphor	and	analogy,	it	is	useful	to	lay	them	bare	when	given	a	chance,	to	interrogate	them,	and,	when	necessary,	to	adjust	and	correct	them.	The	conceptual	model	of	brain-as-computer,	accepting	input	from	the	body	and	the	environment,	logically	and	serially	computing	a	response,	and	generating	a	smooth	output	is	incorrect	and	misleading,	and	I	am	not	suggesting	that	serious	theorists	are	suggesting	this	is	how	the	brain	functions,	but	it	worth	knowing	why	it	does	not	operate	this	way.	Edelman			 	106	(2006),	discussed	above	and	whom	I	will	quote	at	length	here,	summarizes	the	reasons	why	the	brain-as-computer	is	a	poor	analogy		First,	the	computer	works	by	using	logic	and	arithmetic	in	very	short	intervals	regulated	by	a	clock…the	brain	does	not	operate	by	logical	rules.	To	function,	a	computer	must	receive	unambiguous	input	signals.	But	signals	to	various	sensory	receptors	of	the	brain	are	not	so	organized;	the	world	(which	is	not	carved	beforehand	into	prescribed	categories)	is	not	a	piece	of	coded	tape.	Second,	the	brain	order…is	enormously	variable	at	its	finest	levels….no	two	brains	are	identical.	Last,	it	should	be	stressed	that	we	are	not	born	with	enough	genes	to	specify	the	synaptic	complexity	of	higher	brains	like	ours.	[Our]	gene	networks…are	enormously	variable	since	their	various	expression	patterns	depend	on	environmental	context	and	individual	experience	(21-22).		On	top	of	the	reasons	offered	by	Edelman	why	the	brain	is	not	a	computer,	there	is	one	more	that	we	ought	to	pay	attention	to—one	that	is	linked	to	the	phenomenon	of	neuroplasticity.	This	phenomenon	refers	to	the	changes	in	the	synaptic	(and	non-synaptic)	connections	in	the	brain	based	on	the	individual	experiences	of	each.	Through	one’s	changing	experiences,	new	synaptic	connections	are	generated	and	become,	along	with	other	connections,	more	or	less	likely	to	become	activated	so	that	different	patterns	of	neuronal	firing	will	become	more	or	less	likely	to	occur	based	on	changes	in	the	brain	brought	about	by	varied	individual	experience	and			 	107	environmental	changes.	In	essence,	the	brain	is	fundamentally	unlike	a	computer	because	it	is	constantly	changing	on	an	individual	basis	depending	on	the	very	particular	lived	experiences	of	each	human	being—even	though	shared	tendencies	can	and	do	emerge,	as	we	will	see	in	subsequent	chapters.	This	means	that	while	the	brain	is	not	a	purely	rational	computational	machine,	it	is	not	wholly	subject	to	manipulation,	either,	and	is	malleable	to	the	point	where	personal	practices	and	institutions	can	direct	behaviour	in	such	a	way	as	to	assist	in	improving	outcomes:	for	instance,	better	judgments	and	decisions.	Conversely,	as	we	will	see	below,	the	epistemic	force	of	deliberative	democracy	as	a	source	of	good	judgment	and	decision	making	is	undermined	if	specific	procedures	or	expectations	are	generated	concerning	the	brain-as-computer	model	(many	are	not,	but	some	are).	Part	of	the	reality	of	human	cognition	as	it	occurs	in	a	complex	world	is	that	both	internal	processes	and	the	external	environment	are	complex	and	often	enough	unpredictable,	and	highly	variable.		So,	moving	away	from	the	(deliberately	extreme)	analogy	of	the	brain-as-computer,	what	is	the	brain?	Moreover,	how	does	it	process	information	for	making	judgments	and	supporting	attitudinal	development	and	change?	Being	able	to	answer	this	question	is	important,	since	it	will	guide	not	only	our	understanding	of	how	and	why	individuals	behave	in	a	certain	way	when	engaging	in	democratic	deliberation,	but	also	because	it	will	provide	some	foundational	insight	into	how	we	might	design	personal	practices,	procedures,	and	institutions	in	order	to	improve	democratic	citizenship	in	general	and	deliberation	in	particular.	The	human	brain	is	a	bustling	clearinghouse	for	information;	it	is	a	site	for	processing	significant	amounts	of	internal	and	external	sensory	data,	for	interpreting	that	data,	and	for			 	108	constructing	a	world	in	our	minds	from	it	that	includes,	among	other	things,	values,	and,	when	required,	preferences,	and	judgments.	The	process	of	translating	raw	data	into	a	coherent	world	and	subsequent	worldview,	however,	is	not	a	clear	and	automatically	high-fidelity	endeavour.	Indeed,	the	process	of	data	management	for	the	brain	is	part	translation,	part	creation.	As	psychologist	Daniel	Kahneman	(2011)	notes	to	this	effect	“The	world	in	our	heads	is	not	a	precise	replica	of	reality”	(138).	And	when	it	comes	to	politics,	as	I	will	demonstrate	in	this	chapter	and	the	next,	cognition,	reasoning,	judgment,	and	decision	making	can	be	particularly	tricky	endeavours.	But	first,	let	us	look	at	how	specifically	the	brain	enables	the	individual	to	make	sense	of	the	world.	The	leading	theory	about	how	the	brain	processes	information	for	making	judgments	and	developing	or	changing	attitudes	is	grouped	into	a	series	of	models	of	cognition	known	as	“dual-process	models.”	While	there	are	variations	in	these	models	(for	a	summary	see	Eagly	and	Chaiken	1993),	one	long-standing	leader	is	particularly	incisive:	what	I	am	calling	the	systems	approach.	This	approach	to	dual-process	models	helps	us	understand	and	explain	how	we	make	certain	judgments;	it	is	most	famously	and	cogently	advanced	by	psychologists	Daniel	Kahneman	and	Amos	Tversky.	The	approach	is	defined	by	its	broad	division	of	thought	into	two	“systems”:	system	1	and	system	2	modes	of	thinking,	which	I	explain	below	(to	organize	four	models	of	cognition	and	attitude	change	that	I	will	use	to	evaluate	the	question	of	whether	citizens	can	deliberate	autonomously).	I	will	sort	elements	of	the	models	into	system	1	(which,	as	we	will	see,	is	automatic,	intuitive,	and	non-conscious	cognition)	and	system	2	(which	refers	to	slow,	conscious,	and	deliberate	cognition).	While	there	are	some	tensions	between	the	models	and			 	109	between	them	and	their	fit	in	the	system	1/system	2	dichotomy,	they	nonetheless	remain	consistent	with	the	approach	I	am	taking	here.	I	will	note	these	tensions	as	they	arise.	For	the	most	part,	I	operate	under	the	assumption	that	when	it	comes	to	democratic	deliberation,	participants	ought	to	be	operating	in	the	system	2	mode	as	much	as	possible	so	that	they	are	aware	of	the	sources	of	their	reasons	(i.e.	they	have	reasons	for	their	reasons)	and	can	thus	operate	autonomously.	The	models	I	explore	here—elaboration	likelihood,	system	justification,	automaticity,	and	social	intuitionism—under	the	overarching	division	of	system	1/system	2—cognition	are	essential	components	of	the	argument	that	follows	in	this	chapter,	and	indeed	in	the	remainder	of	this	dissertation.	From	here	onwards	I	will	refer	to	these	models	individually	by	their	particular	names	while	referring	to	the	systems	model	as	a	cognitive	system	or	a	systems	approach.	In	chapter	four	I	will	focus	on	motivated	reasoning	in	depth;	I	have	kept	motivated	reasoning	separate	from	the	other	models	since,	while	it	has	some	overlap	in	system	1/system	2	modes	of	cognition,	it	has	some	notable	differences	and	specific	implications	for	autonomous	deliberation	(since	motivated	reasoning	is	less	cognitive	and	more	motivational	in	most	cases)	that	warrant	separate	treatment.30				                                                30	There	are	some	elements	of	motivated	reasoning	that	fit	well	with	the	models	I	will	discuss	in	this	chapter—most	notably,	online	processing.	I	have	included	some	short	discussions	related	to	motivated	reasoning	in	this	chapter,	but	my	primary	exploration	of	motivated	reasoning	will	take	place	in	chapter	four.			 	110	The systems approach: how we generate judgments and beliefs 	What	I	am	calling	the	systems	approach	is	a	two-level	model	of	judgment	(including	attitude	and	belief	formation).	This	approach	was	developed	by	psychologist	Daniel	Kahneman,	alongside,	among	others,	his	academic	partner	Amos	Tversky.	It	is	based	on	decades	of	research	in	human	cognition	and	behaviour.	The	system	is	based	on	two	cognitive	sub-systems	bound	up	in	the	brain	that	can	be	grouped	into	system	1	and	system	2.	The	first	system—far	more	influential	on	our	behaviour	than	often	understood	or	admitted,	and	the	originator	of	many	of	our	judgments	and	beliefs—is	automatic,	rapid,	outside	of	conscious	control	(i.e.	non-conscious),	and	generally	effortless.	It	is	also	highly	subject	to	error	and	manipulation.	The	second	system	is	effortful,	conscious,	cognitively	taxing,	slow,	and	associated	with	cognitive	efforts	associated	with	agency,	choice,	and	concentration	(Kahneman	2011).	It	is	system	2	that	we	tend	to	think	of	when	we	talk	of	the	“self”	(Kahneman	2011).	However,	these	systems	are	meant	to	be	handy	metaphors—or	organizational	categories—rather	than	single	variables;	each	system	is	made	up	of	a	number	of	brain	regions	and	sub-processes.	So,	each	system	is	made	up	of	a	stylized	group	of	variables	that	serve	as	a	helpful	explanatory	cognitive	system	(e.g.	you	cannot	look	at	the	brain	and	“see”	the	system	1	or	system	2	regions).	Each	roughly	accounts	for	a	series	of	related	psychological	and	cognitive	processes	and	predispositions	(Kahneman	2011:	28-30).	Notably,	system	1	plays	an	important	role	in	generating	the	judgments	and	beliefs	of	system	2,	often	largely	outside	of	conscious	awareness.	System	1	generates	“impressions,	intuitions,	intentions,	and	feelings”	which,	if	endorsed	by	system	2,	can	become	adopted	and	integrated	into	more	complex	judgments	and	beliefs	(Kahneman	2011:	24).	And,	as	noted,	while	system	1	is	generally	efficient,	effective,	and			 	111	reliable,	structural	biases	and	shortcomings	render	it	susceptible	to	errors	in	specific	contexts,	threatening	the	quality	and	integrity	of	both	immediate	judgments	emerging	from	it	and	more	involved	system	2	judgments	in	which	it	plays	a	role	generating	and	sustaining	(Kahneman	2011).	Again,	proponents	of	deliberative	democracy	are	normatively	committed,	at	least	generally,	to	centering	deliberations	on	rational,	autonomous	actors	who	can	give	publicly	accessible	reasons	for	their	preferences.	The	proper	functioning	of	a	system	1/system	2	structure	does	not	a	priori	preclude	rational	and	autonomous	deliberation.	Indeed,	it	may	enhance	such	deliberations,	since	the	structure	of	system	1	is	such	that	it	can	pick	up	on	relevant	information—social	situational,	intuitional	information,	including	subtle	cues—and	use	that	data	in	the	conscious	reasoning	around	specific	points.	(It	is	important	to	keep	in	mind	that	system	1	still	represents	a	mode	of	cognition.)	However,	to	the	extent	that	system	1	generates	structural	biases	that	interrupt	or	distort	information	processing	carried	out	in	system	2,	it	can	undermine	the	integrity	of	the	judgments	that	follow,	threatening,	when	it	comes	to	judgment,	the	required	aforementioned	capacities	of	rationality	and	autonomy.		For	if	deliberation	is	based	on	system	2	reasoning	founded	on	non-conscious,	structurally	biased	system	1	cues—whatever	their	origins—the	autonomy	of	the	deliberator	might	itself	be	structurally	undermined.	So,	the	integrity	of	deliberation	can	be	said	to	rest	partially	on	the	extent	to	which	structural,	non-conscious	biases	find	their	way,	via	system	1	cues,	into	the	thinking,	reasoning,	and,	ultimately,	the	judgment	of	participants	in	a	deliberation.	It	is	thus	important	that	we	determine	the	conditions	under	which	such	biases	might	emerge	and	examine	how	specifically	those	biases	might	affect	autonomous			 	112	deliberation.31	In	this	dissertation	I	am	interested	in	sorting	cognition,	broadly,	into	system	1	and	system	2	modes	for	the	purposes	of	understanding	when	individuals	who	deliberate	are	best	able	to	engage	in	autonomous	judgment	and	decision	making;	I	am	particularly	interested	in	how	system	1	thinking	increases	the	chances	of	cognitive	distortion	and	bias,	and	how	it	undermines	an	agent’s	capacity	for	autonomy	and	rationality	as	they	would	be	expressed	through	particular	sorts	of	judgment	made	through	system	2	(which	is	itself	subject	to	bias	and	manipulation	insofar	as	an	individual’s	reasoning	can	be	undermined	at	any	point	in	the	thought	process).		What is cognitive distortion and bias? 	As	mentioned	above,	in	this	dissertation	I	will	explore	four	models	related	to	cognition	that	explain	how	judgment	is	susceptible	to	what	I	am	calling	cognitive	distortion	and	bias.	By	“bias”	I	am	referring	to	a	systematic—and	unreasonable—cognitive	privileging	of	some	perspective,	preference,	desire,	etc.,	or	of	series	of	these	that	undermines	rationality	and	autonomy.32	The	term	“cognitive	distortion”	requires	and	deserves	a	bit	more	commentary.	By	cognitive	distortion,	I	am	referring	to	a	phenomenon	in	which	there	occurs	a	non-conscious	transformation	of	the	content	or	meaning	of	data	during	the	process	of	cognition	that	would,	if	the	individual	were	aware	of	it,	be	unwelcome	and/or	would	contradict	their	understanding	of	                                                31	It	is	important	to	recall	that	these	systems	are	models	for	interpreting	behaviour	rather	than	specific	types	of	brains	or	brain	structures	or	personalities.		32	The	term	"bias"	is	typically	meant	pejoratively	insofar	as	it	is	associated	with	unfairness;	however,	bias	does	not	need	to	be	associated	with	unfairness.	Indeed,	we	can	imagine	bias	being	positive	since,	strictly	speaking,	it	refers	to	prejudice	in	favour	of	something	or	someone.	One	can	have	a	bias,	for	instance,	in	favour	of	healthy	food	over	junk	food.	My	concern	in	this	dissertation	is	with	unreasonable	and	hidden		bias	that	undermines	rationality	and	autonomy.			 	113	how	they	make	judgments	and	decisions.	To	distort	something	means	to	change	its	original	shape	so	that	one	is	left	with	some	product	or	perspective	different	than	before	the	distortion.	In	the	context	of	cognition,	distortion	occurs	when	some	initial	data	or	thought	is	transformed	during	the	process	of	transmission	in	a	way	that	would	be	unwelcome	to	the	individual	who	is	thinking	or	would	contradict	their	understanding	of	how	they	think,	judge,	and	decide	(this	is	what	separates	distortion	from	mere	change).			 In	the	context	of	the	two	systems	approach,	the	distorted	or	biased	cognition	that	I	am	concerned	with	here	occurs	under	system	one	processes,	outside	of	the	awareness	of	the	individual	(compared	to,	say,	deliberately	distorted	cognition,	as	when	someone	consciously	rationalizes	or	deliberately	chooses	information	selectively—as	is	the	case	in	some	instances	of	motivated	reasoning).	It	would	be	unwelcome	to	the	extent	that	if	an	individual	were	exercising	autonomy	and	processing	information	centrally	(e.g.	through	system	two)	she	would	either	deny	or	amend	the	choice	or	judgment	she	reached.	Again,	as	I	have	noted,	the	problem	at	hand	is	not	that	certain	information,	research	strategies,	or	cognitive	practices	are	bound	up	in	affective	cognitive	processes,	but	rather	that	these	are	hidden,	may	be	biased,	and	are	unavailable	for	individual	or	collective	evaluation	during	the	process	of	establishing	validity	through	deliberation.33	Admittedly,	there	is	little	cognition—perhaps	none—that	occurs	without	some	distortion	or	bias.	As	related	above,	the	world	in	our	heads	does	not	seamlessly	match	the	world	outside	of	it—which,	in	part,	is	the	why	correspondence	theory	of	truth	is	problematic.	                                                33	I	am,	of	course,	assuming	individuals	prefer	to	know	their	motivations	and	to	be	correct	about	why	they	reason	the	way	they	do.	For	this	dissertation,	I	am	bracketing	the	phenomena	of	deliberate	self-delusion	and	lying	in	the	contexts	of	democratic	deliberation.					 	114	Our	brains	and	minds	are	constantly	actively	constructing	the	world	by	processing	external	raw	data	and	transforming	it	into	something	coherent	and	useful	to	us.	However,	the	cognitive	distortion	to	which	I	am	referring	in	this	dissertation	is	structural	and	counter-productive—at	least	in	the	context	of	democratic	deliberation—since	it	undermines	the	quality	of	one’s	reasoning	(and	undermines	self-determination)	by	altering	(outside	of	one’s	awareness)	the	data,	thought(s),	or	reasoning	with	which	one	is	engaging	to	produce	a	judgment.	Moreover,	this	cognitive	distortion	is	unwelcome;	similar	to	the	definition	of	autonomy	offered	by	Christman	(1991),	and	discussed	above,	a	key	requirement	of	cognition	that	is	not	distorted	is	that	the	individual	who	is	thinking	would	affirm	any	transformation	of	data	during	the	process	of	cognition.		Whereas	much	of	our	cognitive	activity	is	creative—again,	we	are	constantly	building	a	world	in	our	minds	out	of	the	raw	stimuli	found	in	the	world	outside	us—cognitively	distorted	processing,	which,	unlike	the	brain	and	mind’s	act	of	assembling	a	world	from	raw	data,	can	in	principle	be	avoided	or	at	least	mitigated.	It	might	be	easiest	to	put	it	this	way:	all	cognition	is	creative,	but	cognitive	distortion	warps	data	and	the	process	of	reasoning	from	that	data	in	a	way	that	undermines	autonomy	and	threatens	our	ability	to	generate	valid	judgments	and	decisions.	This	effect	is	why	cognitive	distortion	earns	a	pejorative	status	within	a	particular	cultural	context	(e.g.	democratic	deliberation)	relative	to	certain	normative	expectations	(e.g.	that	those	who	deliberate	are	rational	and	autonomous).	We	could,	if	we	wanted	to,	agree	that	any	reason	based	on	any	consideration,	conscious	or	not,	would	be	fine	political	fodder	for	democratic	deliberation.	In	such	a	case,	distorted	cognition	would	be	perfectly	acceptable,	since	the	status	of	our	preferences	and	judgments			 	115	would	not	be	subject	to	evaluation	based	on	our	autonomy.	However,	since	proponents	of	deliberative	democracy	have	directly	or	indirectly	specified	the	need	for	rational	and	autonomous	actors,	cognitive	distortion	is	obviously	a	threat,	since	it	acts	as	a	challenge	to	both	the	desired	goods	of	rationality	and	autonomy.	Appreciating	the	ways	in	which	cognitive	distortion	occurs	or	is	manifest	is	key	to	understanding	the	failures	of	rationality	and	autonomy	to	explain	fractures	in	the	democratic	process.	Once	again,	our	cultural	assumptions,	desires,	or	requirements	for	democratic	deliberation	run	up	against	certain	natural	phenomena	and	the	limits	of	our	cognitive	realities	in	the	given	context	of	contemporary,	fast-paced,	complex	liberal	democracies.	Moreover,	while	that	challenge	does	not	imply	we	can	do	nothing	to	address	the	problem,	it	does	suggest	that	current	modes	of	democratic	deliberation	may	fall	short	of	meeting	certain	needs	and	that	the	status	quo	undermines	the	plausibility	and	reduces	the	quality	of	democratic	goods	offered	by	the	deliberative	democratic	approach.	For	now,	however,	before	addressing	potential	fixes	to	the	problem	of	cognitive	distortion,	let	us	turn	to	the	specific	threats	to	autonomous,	rational	judgments	that	are	explained	by	the	four	models	mentioned	above	and	see	specifically	how	such	distortion	might	occur	and	how	it	might	affect	democratic	deliberation	in	each	instance.	These	four	models	have	been	chosen	for	four	reasons	discussed	in	chapter	one,	but	two	reasons	are	particularly	important	and	warrant	repeating.	First,	they	are	representative	of	the	broader	literature	on	cognitive	distortion	insofar	as	they	are	generally	accepted	theories	in	social	psychology	and	include	the	majority	of	known	cognitive	biases	(i.e.	they	are	representative	and	comprehensive).	Moreover,	second,	they	are	of	specific	concern	to	democratic	deliberation			 	116	insofar	as	they	pertain	specifically	to	capacities	of	rationality	and	autonomy,	as	defined	in	this	dissertation,	required	for	good	judgment	and	decision	making	(i.e.	they	are	case-appropriate).		Models of cognition		The elaboration likelihood model (ELM): persuasion and opinion/attitude change The	ELM,	which	is	concerned	with	how	attitudinal	change	is	brought	about	in	individuals,	was	created	by	psychologists	John	Cacioppo	and	Richard	Petty.	It	has	remained,	more	or	less,	similar	to	the	form	that	they	developed	in	the	1970s	and	1980s.	The	ELM	helps	us	understand	how	political	judgments	are	formed	and	how	they	change	(if	they	change	at	all).	During	the	time	in	which	the	ELM	was	developed,	social	psychologists	turned	away	from	approaches	that	focused	strictly	on	rational,	deliberate	opinion/attitude	change	and	began	to	embrace	new	understandings	of	cognition	that	considered	less	effortful	and	conscious	mental	activity.	The	ELM	breaks	down	attitude	determination	into	two	routes	set	as	poles	along	a	continuum	measuring	one’s	motivation	for	elaborated	thought	(i.e.	thinking	capable	of	rationally	processing	a	complex	persuasive	message):	a	central	route,	which	is	characterized	by	cognitively	effortful	and	taxing	thought,	and	which	accounts	for	high	elaboration	likelihood;	and	a	peripheral	route	marked	by	the	use	of	affect,	cues,	and	heuristics,	accounting	for	low	elaboration	likelihood	(Petty	1999;	Petty	and	Cacioppo	1981,	1986).	We	can	place	central	processing	in	the	system	2	category	and	peripheral	processing	in	the	system	1	category;	and	we	can	understand	our	challenge	in	deliberative	democracy	as	one	concerned	with	encouraging	more	use	of	system	2	functions	to	encourage	central	processing	and	high	elaboration	when	changing	attitudes	and	generating	judgments.			 	117	For	example,	consider	an	individual	who	is	opposed	to	tax	increases	but	presented	with	a	situation	in	which	his	local	transit	authority	is	in	need	of	funding	for	infrastructure	spending.	Here,	a	specific	need	(i.e.	transit	funding	for	infrastructure)	meets	a	pre-existing	attitude	(i.e.	higher	taxes	are	undesirable).	Say	also	that	the	man	regularly	uses	public	transportation,	but	his	experience	on	transit	is	negative	because	of	the	inconveniences	he	faces	daily	because	of	inadequate	transit	infrastructure.	The	man	in	question	may	have	good	reason	to	engage	in	central	route	processing	around	this	issue	since	it	has	high	salience	for	him.	It	may	be	worth	his	time	to	engage	in	effortful,	cognitive	taxing	reflection	on	the	issue.	He	may	well	decide	that	a	tax	increase	is,	in	fact,	necessary	and,	on	balance,	desired	in	this	case.	He	would	presumably	have	good,	immediate	reasons,	too	(e.g.	commute	times	would	improve,	the	bus	would	be	less	crowded,	etc.).	Of	course,	he	directly	benefits	from	this	outcome,	but	it	could	just	as	well	be	the	reasons	that	have	changed	his	mind	(i.e.	the	force	of	argument/reflection)	rather	than	the	fact	that	he	benefits	from	those	reasons	bringing	about	a	decision.	Conversely,	in	the	same	situation,	a	driver	who	is	also	opposed	to	higher	taxes,	who	never	uses	public	transit,	and	who	is	not	affected	by	long	commute	times,	may	never	have	good	reason	to	examine	the	case	for/against	raising	taxes	to	generate	funds	for	transit	infrastructure	and	may	thus	engage	in	peripheral	route	processing.	To	simplify,	such	processing	might	look	something	like	this:	“They	want	to	raise	taxes	to	spend	on	transit.	I	don’t	like	taxation.	So,	I’m	opposed	to	increased	taxation	for	transit	funding.”	Note	that	in	the	former	case	reasons	are	central	(cued	by	the	salience	of	the	issue	to	the	first	man)	and	the	in	latter	case	reasons	are	absent	(instead,	affect	is	central	since	the	man	does	not	“like”	taxes).			 	118	Within	the	context	of	the	ELM,	researchers	find	that	situational	and	dispositional	factors	are	most	significant	in	determining	which	route	an	individual	will	take	when	processing	a	message	that	is	intended	to	persuade	them.	Those	who	“enjoy	thinking,”	for	instance,	and	thus	have	such	a	disposition	to	engage	in	prolonged	reflection,	are	more	likely	to	engage	in	central	route	processing	than	those	who	do	not	enjoy	thinking	(Cacioppo	and	Petty	1982).34	Situationally,	real-world	factors,	such	as	time-constraints,	can	also	effect	where	one	falls	along	the	continuum	in	any	given	circumstance	(and,	potentially,	they	can	shape	one’s	dispositions	in	the	first	place),	with	greater	constraints	decreasing	the	likelihood	that	one	will	engage	in	central	route	processing	(Moore	et	al.	1986).	However,	more	importantly	for	my	purposes	here,	given	that	central	processing	is	cognitively	costly,	issues	that	are	relevant	to	an	individual	and	which	they	are	motivated	to	engage	with	are	more	likely	to	be	centrally	processed	than	those	that	are	not,	with	the	latter	more	likely	to	remain	processed	through	the	peripheral	route.	Thus,	both	personal	psychological	and	broader	environmental	factors	are	at	work	in	determining	route	choice.	So,	when	it	comes	to	democracy	in	general,	and	deliberative	democracy	in	particular,	those	with	more	at	stake	in	a	given	decision	should	be	more	likely	to	engage	in	central	processing	(Heine	personal	communication	2015),	while	others	might	(at	least	initially)	be	more	likely	to	engage	in	peripheral	processing.	However,	as	I	will	argue	in	chapter	five,	to	improve	the	odds	that	an	individual	generates	good	judgments	and	that	a	group	                                                34	It	does	not	appear	to	be	the	case	that	high-elaboration	and	low-elaboration	tendencies	are	fixed,	say,	in	the	way	that	raw	athletic	talent	is	thought	of	as	fixed.	Accordingly,	generating	tactics	and	strategies	for	making	high-elaboration	more	likely	is	an	appropriate—and	potentially	quite	productive—research	objective,	and	one	I	will	take	up	in	chapter	six.			 	119	produces	good	decisions,	participants	in	deliberation	should	be	specifically	encouraged	and	motivated	to	engage	in	high-level	processing.		Notably,	research	on	the	ELM	also	suggests	that	there	are	differential	effects	on	attitudinal	change	and	persistence	based	on	where	one	finds	their	cognitive	effort	falling	on	the	ELM	continuum.	Unsurprisingly,	persuasive	messages	considered	along	points	closer	to	high	elaboration	likelihood	tend	to	be	more	durable	and	impactful	when	adopted	(Petty	et	al.	1995)	than	those	closer	to	low	elaboration	likelihood,	though	this	does	not	necessarily	account	for	bias	or	distorted	thinking	that	may	creep	into	higher	elaboration	states	(a	problem	that	will	be	discussed	later	in	this	chapter	and	again	in	the	second	half	of	this	dissertation).	The	key	takeaway	is	that	the	route	taken	when	processing	information	can	matter	a	great	deal:	the	same	combination	of	the	individual,	the	context,	and	the	information	presented	can	have	a	different	outcome	depending	on	where	one	falls	on	the	ELM	continuum	in	any	given	instance.	As	we	will	see	in	more	detail,	depending	on	potential	structural	tendencies	in	ELM	route	choice,	there	are	significant	potential	implications	for	political	opinion	and	attitude	formation,	including	different	real-world	political	outcomes,	for	how	democratic	deliberations	are	set	up	and	run,	and	for	the	types	of	judgments	that	emerge	from	them.	Obviously,	proponents	of	deliberative	democracy,	especially	those	who	advance	deliberation	as	a	means	of	making	epistemically	valid	decisions,	are	committed	to	and	expect	(at	least	relatively)	durable,	though	revisable,	attitudes—again,	these	are	among	the	cultural	assumptions	of	deliberation.	It	is	equally	obvious	that	these	proponents	are	likely	to	be	just	as	normatively	committed	to	encouraging	more	central	route	processing.	Even	though	scarce	cognitive	and	day-to-day	resources	make	choosing	to	expend	cognitive	effort	on	persuasive			 	120	messages	a	careful	art—when	indeed	we	have	some	control	over	the	choice—proponents	of	deliberation	have	good	reason	to	argue	that	deliberative	settings	are	highly	appropriate	locations	for	this	expending.	This	is	especially	true	given	that	central	route	processing	is	less	susceptible	to	cognitive	error	or	distortion,	and	deliberative	forums	are	expressly	meant	to	elicit	honest	and	reasoned	exchanges	among	participants	in	search	of	legitimate,	durable	decisions—that	is	to	say	that	critical	reflection	is	more	likely	to	elicit	reasons	that	must	be	defended	than	other,	less	engaged,	modes	of	decision	making.	The	key	concern,	then,	becomes	how	we	can	motivate	deliberators	to	engage	in	central	route	processing	through	reason	giving—again,	a	question	that	will	be	systematically	taken	up	and	explored	in	the	second	section	of	this	dissertation,	in	chapters	five	to	seven.		System justification 	In	social	psychology,	system	justification	is	defined	as	the	“process	by	which	existing	social	arrangements	are	legitimized,	even	at	the	expense	of	personal	and	group	interest”	(Jost	and	Banaji	1994:	2).		Evidence	from	decades	of	research	into	the	motivation	of	decision	making,	beliefs,	and	judgments	through	the	lens	of	system	justification	theory	has	shown	that	many	individuals	are	motivated,	psychologically	and	non-consciously,	to	hold	favourable	attitudes	towards	the	existing	social	structure—the	status	quo—even	occasionally	overriding	ego	justification	(i.e.	personal	interests)	and	group	justification	motives	while	doing	so.	As	Jost	et	al.	(2004)	note,	while	there	is	a	“general	psychological	tendency	to	justify	and	rationalize	the	status	quo,	we	do	not	assume	that	everyone	is	equally	motivated	to	engage	in	system	justification”	(912).	Nonetheless,	the	tendency	to	justify	the	status	quo	despite	strong	reasons			 	121	to	challenge	it	is	common	enough	to	warrant	study	and,	by	implication,	concern,	given	that	for	some	issues	(e.g.	climate	change,	poverty,	war),	the	status	quo	can	be	threatening,	costly,	or	both.	Indeed,	the	failure	to	challenge	the	status	quo	is	potentially	catastrophic	in	certain	instances—not	to	mention	critical	to	particular	populations	in	others	(e.g.	around	questions	of	redistribution	or	the	effects	of	climate	change).	Moreover,	advantaged	groups	tend	to	show	more	implicit	support	for	their	ingroup,	while	less	advantaged	groups	often	show	more	support	(which	can	be	either	conscious	or	non-conscious)	for	outgroups.	This	tendency	also	has	an	effect	on	judgments	made	by	such	individuals—for	instance,	leading	individuals	in	outgroups	to	work	against	their	group	and	its	interests	or	to	denigrate	their	group,	potentially	at	a	personal	cost	to	their	self-esteem	and	task	performance	(Batalha	et	al.	2007,	Dasgupta	2004).	Underlying	system	justification	tendencies,	according	to	Jost	and	his	colleagues,	are	a	series	of	processes	of	motivations	that	generate	beliefs	and	judgments	that	support	the	status	quo.35	These	include	ideological	motives	to	justify	the	status	quo	that	operate	implicitly	and	non-consciously	(Jost	et	al.	2004:	912).	One	of	the	key	motivations	for	these	tendencies	is	the	individual’s	drive	to	hold	favourable	attitudes	not	just	towards	themselves	(ego	justification)	and	their	ingroup	(group	justification),	but	also	towards	the	social	and	political	systems	that	surround	them	(Jost	et	al.	2004:	887).	In	a	case	study	on	political	conservatism,	Jost	et	al.	(2003)	found	common	traits	among	conservatives	at	personal,	epistemic,	existential,	and	ideological	levels;	they	distilled	the	core	of	conservatism	to	resistance	to	change	and	the	justification	of	inequality.	Findings	also	indicated	that	while	motivations	to	hold	conservative	ideological	                                                35	System	justification	may	ultimately	be	a	particular	example	of	loss	aversion/the	need	for	ontological	security/attempts	and	maintaining	frameworks	of	meaning.	One	of	the	common	threads	in	this	dissertation	is	the	pervasive	human	desire	for	stability	and	familiarity.			 	122	positions	varied	depending	on	particular	individuals	and	situations,	they	were	linked	to	the	needs	of	managing	uncertainty	and	threat	(Jost	et	al.	2003).	These	motivations—underwritten	by	a	process	known	as	socially	motivated	cognition—reflected	both	social	(e.g.	a	desire	to	maintain	the	status	quo)	and	cognitive	needs	(e.g.	to	protect	against	the	fear	of	loss	and	death),	though	these	considerations	do	not	necessarily	consciously	or	explicitly	factor	into	the	reasoning	or	reason	giving	of	individuals	when	they	are	explaining	their	beliefs	or	rationalizing	their	judgments.	Liberals,	on	the	other	hand,	tend	not	to	engage	this	palliative	function	of	system	justification	as	much	as	conservatives—and	thus	tend	to	be