Printer Friendly

Winning arguments.


The classical model of reason depicts a person who, like Rodin's Thinker, applies her higher mental faculties to the available evidence and arrives at the correct--or at least the best-answer to a given problem. Likewise, the dominant metaphor of the First Amendment envisions the ideal of a free speech marketplace in which rational consumers choose the best ideas from among those on display, ultimately resulting in the discovery of truth. Thus, the marketplace model as a justification for free speech protection closely tracks the classical view of reason and appears to depend on such a view as its foundation. Yet, as has become apparent in recent decades, people are very often irrational: we harbor a multitude of cognitive biases, engage in motivated reasoning, and frequently choose false and pernicious products from the marketplace of ideas. It is thus not surprising that most First Amendment scholars have abandoned epistemic rationales for freedom of speech in favor of justifications grounded in process-based models of democratic self-governance. Recently, however, a new theory has emerged positing that the evolved function of human reason is not to discover truth but, instead, to make and win arguments. This "Argumentative Theory of Reasoning" explains many of the cognitive biases and irrationalities long observed across many fields of academic study; as an alternative explanation of human reasoning, it has intriguing, if somewhat counterintuitive, implications for First Amendment law. Although reasoning may have evolved to serve the purpose of winning arguments rather than gaining accurate knowledge, it turns out that understanding the psychology of reasoning can illuminate the ways in which a free marketplace of ideas can actually lead to optimal epistemic outcomes. The Argumentative Theory of Reasoning ironically suggests that we should give truth another look as a solid foundation for freedom of speech protection.

       A. The Reign of Truth
       B. Truth is Dead: The Law Reviews
       C. Alternative Accounts in Legal and Democratic
       A. What is Reason For?
       B. Logical and Argumentative Skills
       C. Confirmation Bias and Motivated Reasoning
       D. Epistemic Vigilance and the Dialectical Nature of
       E. Group Decision-Making
       A. The Epistemic Case for Inclusive Deliberative
       B. The Deliberative Epistemic Case for Freedom of

"But all was false and hollow; though his tongue/ Dropt manna, and could make the worse appear/ The better reason, to perplex and dash/ Maturest counsels False" (1)

"Consideration of these issues inevitably recalls an aphorism of journalism that 'too much checking on the facts has ruined many a good news story."' (2)


Humans are rational beings. (3) Indeed, the capacity for reason is often said to set us apart from every other creature. (4) In the domain of cognitive science, the classical view "posits that the main function of reasoning is to correct misguided intuitions, helping the reasoner reach better beliefs and make better decisions." (5) This classical view tends to emphasize the individual thinker, envisioning a solitary intellect parsing information to arrive at the correct answer. Despite mountains of experimental findings that demonstrate serious and systematic defects in human reason so understood, (6) we cling to this ideal of our basic and inherent rationality. (7)

Over the past several years, there has been an explosion of research on how humans reason and make decisions. Scholars across many disciplines, from psychology to neuroscience to law, increasingly acknowledge that much of the decision-making process is "implicit," meaning outside of our conscious awareness. (8) The critical role of emotion in decision-making has also become apparent. (9) And a host of systematic cognitive distortions such as confirmation (or "myside") bias, (10) in-group bias, (11) framing effects, (12) hindsight bias, (13) and motivated reasoning (14) have been, by this point, exhaustively catalogued. (15) This "cognitive revolution" has had a profound effect within many academic fields, perhaps none more than economics, which is no surprise for a discipline explicitly modeled on the rational actor. (16) Within legal scholarship, where the law and economics movement has had such a significant impact, (17) many prominent scholars have incorporated analysis of these demonstrated irrationalities through the field of behavioral law and economics. (18) More generally, many works have explored the implications of various findings in the mind sciences to a wide range of legal doctrine and theory. (19)

And yet, the cognitive/behavioral revolution has so far largely bypassed the First Amendment. (20) This is surprising because the central metaphor of the First Amendment--the marketplace of ideas--is explicitly economic and thus the challenges to the rational actor model would seem to position the marketplace of ideas model as particularly low-hanging fruit. (21) Furthermore, First Amendment law is intricately bound up with assumptions about human decision-making and belief-formation processes; an explosion of knowledge about how people make decisions and form beliefs could not help but have implications for First Amendment theory and doctrine. (22) Thus far, however, the implications remain largely unexplored.

This Article ventures into this gap in the literature by considering recent work in cognitive psychology that offers a unifying theory to explain the many and various cognitive biases described in many different fields. This emerging theory, known as the Argumentative Theory of Reasoning (ATR), posits that the function of human reason is not-as in the classical view-to find the "right" or "true" or "best" answer; rather, the function of reason is to make and win arguments. (23) According to this view, "the primary function for which ... [reasoning] evolved is the production and evaluation of arguments in communication." (24) The ATR has caused a stir among cognitive scientists (25) in no small part because it offers a convincing way to make sense of much of the experimental results demonstrating cognitive biases and systematic irrationalities in human reason. (26) Under this alternative view of reasoning, in fact, the "irrationalities" make perfect sense: if the evolved function of reasoning is to make and evaluate arguments rather than to increase individual knowledge per se, then many of the biases that have confounded scholars across various academic fields are not defects in reasoning after all. (27) Rather, they are features we would expect to find.

This alternative view of reasoning has implications for how we think about protection of speech. In First Amendment jurisprudence, the "marketplace of ideas" (MOI) metaphor (28)--the most common incarnation of the "search for truth" (29) justification for freedom of speech--mirrors the classical view of reasoning. In the MOI, each prospective consumer of ideas surveys the merchandise being offered, and applying her reasoning capacity, selects the best ideas from among the many on display. When the similar actions of many individuals are aggregated, the invisible hand of the market leads to the flourishing of tine ideas and the withering of false ones. Given our reluctance to relinquish our self-image as rational actors, it is perhaps not surprising that, despite a myriad of criticisms, (30) the marketplace metaphor has kept so tenacious a hold on First Amendment jurisprudence. In case after case, the Supreme Court repeatedly has reaffirmed that a primary purpose of the First Amendment is "to preserve an uninhibited marketplace of ideas in which truth will ultimately prevail" (31) and that "[t]he theory of our Constitution is 'that the best test of truth is the power of the thought to get itself accepted in the competition of the market."' (32)

Despite the prevalence of these marketplace and other truth rationales in case law, it is no longer fashionable among scholars to defend freedom of speech on the ground that it is likely to lead to increased knowledge. Though the pages of the Supreme Court Reports continue to overflow with paeans to the marketplace of ideas and the search for truth, in the law reviews truth has fallen decidedly out of favor. Rather, proceduralist justifications (33) grounded in democratic discourse and autonomy have displaced epistemic rationales as the predominant explanations for free speech protection. The result is that there is currently something of a disconnect between Supreme Court jurisprudence and the academic literature concerning the vitality of the marketplace of ideas and truth-finding justifications for protecting free speech. (34)

At the same time--and this is surely no coincidence--process has also displaced epistemic rationales for democratic legitimacy in the political theory literature. (35) The predominant view among democratic theorists is that democratic legitimacy is grounded upon the equal dignity and respect to which every citizen's voice is entitled; the notion that participatory, deliberative democracy is desirable because it is more likely to lead to better decisions is, if not openly ridiculed, essentially ignored by most scholars in the field. (36)

And yet the idea that people--not just some people but "the people" as a collective--might be wiser (37) than any individual person has been gaining traction in recent years. A small cadre of epistemic democrats (38) has lately revived the argument that democracy is the best system not only because it equally respects and values the voices of all citizens but also because it is most likely to lead to the best decisions. (39) As relevant to free speech law and theory, one strand of this epistemic case for democracy rests on the benefits of deliberative decision-making among cognitively diverse (40) groups of individuals. (41) According to this argument, when large, inclusive (42) groups deliberate under conditions that the Argumentative Theory of Reason suggests are "felicitous"--that is, under a regime that understands that reasoning is for winning arguments--then these deliberating groups are most likely to reach optimal outcomes. This Article proposes that the same insights also apply in the context of the First Amendment. It is high time, in other words, to give the epistemic case for freedom of speech a fresh look. (43)

The Article proceeds as follows. Part II examines the strong trends in both the legal and political science scholarship to reject epistemic justifications for a system of democratic discourse and decision-making. Though the marketplace rationale remains alive and well in the Supreme Court Reports, (44) truth as a justification for speech protection has been dying a slow death in the law reviews. In its place, First Amendment scholars have turned to what I label "proceduralist" justifications for protecting freedom of speech. At the same time, there is a parallel dynamic among democratic theorists, most of whom reject the epistemic case for democracy just as most First Amendment theorists reject the epistemic case for freedom of speech. I suggest that these trends are likely related to the findings in economics and psychology that have undermined confidence in humans as rational actors.

Part III lays out the Argumentative Theory of Reasoning as developed in a series of papers by cognitive scientists Dan Sperber and Hugo Mercier. (45) According to the theory, our reasoning faculty evolved in the service of making and evaluating arguments in a deliberative context. Thus, people are more likely to gain epistemic benefits and to come to accurate or optimal decisions when they reason in such a context. Although increased knowledge is not the function of reason in the sense understood by classical models of reasoning (or of evolutionary theory), reason nonetheless can (and, according to the ATR, often will) lead to increased knowledge and better decision-making--particularly when the communicative environment is "felicitous" (i.e. it replicates the features under which reasoning evolved). In particular, and especially pertinent in terms of the First Amendment, the ATR illuminates and emphasizes the communicative, socially-embedded nature of human reason. According to this theory, groups can make good decisions despite the fact that they often behave "irrationally" according to the classical model.

After exploring these issues within the framework of cognitive psychology, Part IV sketches an epistemic case for freedom of speech based on the notion that we reason in order to win arguments. (46) Indeed, an appreciation of the epistemic benefits of deliberation is already present, though underdeveloped, in the Supreme Court's existing free speech jurisprudence. Drawing upon the model of reasoning advanced by the ATR, this Part proposes that under the right conditions, "more speech" can, in fact, generate increased knowledge and better decisions. In sketching out what these conditions might look like, this Part incorporates recent interdisciplinary work that elucidates the conditions under which groups are most likely to reach optimal decisions. (47)

A more nuanced understanding of how and why people reason could enable courts and scholars to approach difficult free speech questions in a way that would facilitate good epistemic outcomes rather than hinder them. Ironically, taking seriously the proposition that reason was made for winning arguments rather than for discovering truth may actually strengthen the case for a truth-seeking, epistemic justification for the First Amendment's protection of freedom of speech.


As noted above, the rational actor model of decision making that has been undermined by the cognitive revolution is closely aligned with the central metaphor of the modern First Amendment: The Marketplace of Ideas (MOI). "The marketplace metaphor neatly encapsulates the assumptions of the rational audience ideal, and it remains the dominant metaphor of First Amendment jurisprudence." (48) There has been much academic debate of late about the meaning and continued relevance of the marketplace metaphor, (49) and indeed criticism of one stripe or another has been a consistent feature of free speech scholarship for decades. (50) Nonetheless, the marketplace and other "search for truth" justifications for free speech protection remain firmly embedded in free speech doctrine (51) and also in the public imagination. (52)

A. The Reign of Truth

The notion of free speech as generating truth has played an enduring and crucial part in First Amendment doctrine and theory. Following the terminology in the political theory literature, (53) I refer to the range of truth-based arguments collectively as the "epistemic justification" for freedom of speech. Marketplace rationales are the most common embodiment of the epistemic justification, but all "arguments from truth" (to use Frederick Schauer's phrase) share the same fundamental assumptions and values. While it is certainly not hard to find support for alternative (or complementary) free speech justifications throughout the Supreme Court's First Amendment opinions, (54) the truth-seeking justification--often in the form of the marketplace of ideas metaphor--is not simply a historical artifact. It remains a vital component of the Court's free speech jurisprudence. C. Edwin Baker noted more than three decades ago that "[t]he Supreme Court steadfastly relies upon a marketplace of ideas theory in determining what speech is protected. Marketplace imagery ... pervades Court opinions and provides justification for their first amendment 'tests.'" (55) Some thirty-five years later, this remains an accurate description of Supreme Court First Amendment jurisprudence. As noted by one of the most respected First Amendment theorists, in a recently published book, "That the point of First Amendment doctrine is to 'advance knowledge and the search for truth by fostering a free marketplace of ideas and an "uninhibited, robust, wide-open debate on public issues'" has become more or less a constitutional commonplace." (56)

It is perhaps not surprising that free market imagery should have proven so salient and powerful in contemporary United States legal and popular culture. (57) Ironically, although he famously insisted that "[t]he 14th Amendment does not enact Mr. Herbert Spencer's Social Statics," (58) Justice Oliver Wendell Holmes, Jr. is widely credited with importing those very laissez-faire principles into First Amendment jurisprudence by invoking "the competition of the market" and "free trade in ideas" as the "best test of truth." (59) This market metaphor and related arguments from truth have most assuredly done significant work in First Amendment theory and doctrine to justify protecting a great deal of controversial expression.

The marketplace model assumes that, in a well-functioning free speech market, truth will ultimately prevail. (60) Even when market rhetoric is absent, arguments from truth are based on the premise that restrictions on speech are contrary to the goal of discovering the true or best answer. In Frederick Schauer's summary of the "argument from truth:"

[C]ertain core principles are found in all expressions of the doctrine. They all share a belief that freedom of speech is not an end but a means, a means of identifying and accepting truth. Further, they have a common faith in the power of truth to prevail in the adversary process, to emerge victorious from the competition among ideas. Finally, they share a deep scepticism with respect to accepted beliefs and widely acknowledged truth, logically coupled with a keen recognition of the possibility that the opinion we reject as false may in fact be true. A heavy dose of fallibilism is implicit in the view that freedom of speech is a necessary condition to the rational search for truth. (61)

Whether described in market terms or more generally, epistemic justifications share the fundamental assumption that a free and full airing of any and all speech will be most likely to lead, in the long term if not always in the short term, (62) to increased knowledge and a closer approximation of factual, political, artistic, historical, or moral truth. (63)

The epistemic justification--the argument from truth--is "the predominant and most persevering" of all of the reasons proffered over the years in favor of strong speech protection. (64) As a measure of the enormous power and reach of the marketplace of ideas metaphor, it has been said that "[njever before or since has a Justice conceived a metaphor that has done so much to change the way that courts, lawyers, and the public understand an entire area of constitutional law." (65) Though, as detailed in the following section, the marketplace rationale has in the past few decades fallen largely out of favor among leading First Amendment scholars, it has historically formed the bedrock of freedom of speech protection both theoretically and doctrinally. (66) From John Milton, to John Stuart Mill, to Thomas Jefferson, to Oliver Wendell Holmes, Jr., strong protections for free-flowing speech have been frequently and powerfully justified as the "best test of truth." (67)

B. Truth is Dead: The Law Reviews

In contrast to its continued vitality in the Supreme Court Reports, the epistemic rationale appears to be gasping its dying breaths in the law reviews. (68) While it is impossible to encapsulate the wide-ranging and complex body of work that is contemporary free speech scholarship, at the very least it is apparent that arguments from truth no longer predominate. (69) Frederick Schauer suggests that the decline of marketplace theory in academic scholarship--and the attendant rise of alternative theories such as autonomy and democratic self-governance--is the result of the bankruptcy of the conceptual and empirical claims of the marketplace metaphor. (70)

Many scholars are dismissive of truth-seeking justifications. James Weinstein, for example, has written that "the marketplace-of-ideas rationale is best described as a peripheral free speech norm." (71) The past several decades have instead been dominated by various strands of democracy-based free speech theories, most of which place minimal or no emphasis on epistemic values. Theories grounded in autonomy and selfrealization (72) likewise are concerned with values other than epistemic advancement. (73)

Perhaps the most telling evidence of truth's demise in the scholarly literature is that it is sometimes ignored altogether. An essay by Jack Balkin about the First Amendment implications of the digital technological revolution does not mention, even in passing, discovery of truth or the epistemic benefits of free speech. (74) For Professor Balkin, "The purpose of freedom of speech ... is to promote a democratic culture," which "is a culture in which individuals have a fair opportunity to participate in the forms of meaning making that constitute them as individuals, ... it is about each individual's ability to participate in the production and distribution of culture." (75) While it may go without saying that a culture infused with accurate ideas and knowledge is preferable to one filled with false and pernicious ideas, in Professor Balkin's essay it does (go without saying). The essay is entirely silent about any epistemic function of freedom of speech.

Other leading scholars do not just ignore but positively disparage the epistemic rationale. Discussing factual truth (but making a point that would also apply to political or moral truth), (76) Schauer writes:
   the persistence of the belief that a good remedy for false speech
   is more speech, or that truth will prevail in the long run, may
   itself be an example of the resistance of false factual
   propositions to argument and counterexample. No matter how many
   examples and how much serious scientific research refute the view
   that falsity typically yields to truth, or that the truth of a
   proposition is the dominant factor in determining which
   propositions will be accepted and which rejected, free speech
   claimants, sometimes in legal proceedings but far more often in
   public advocacy, trot out the tired old cliches that are little
   more than modern variants on Milton's now-legendary but almost
   certainly inaccurate paean to the pervasiveness and power of human
   rationality. (77)

To the extent that scholars do defend arguments from truth, they often seem to do so rather reluctantly. Thus, for example, Lyrissa Barnett Lidsky argues that despite strong evidence that the rational actor assumption at the heart of the marketplace of ideas model is very often "demonstrably false," it is worth retaining in order to protect citizen autonomy and democratic self-governance. (78) The "flawed ideal of audience rationality," (79) which she argues is a fundamental and necessary assumption of the marketplace of ideas model, (80) "is better than the alternative" because assuming that people are irrational would undermine the case for self-governance. (81) Note that Professor Lidsky does not advocate retaining the marketplace model for epistemic reasons, but rather in service of democratic and dignitary values. (82)

Even an article entitled "In Defense of the Marketplace of Ideas/Search for Truth as a Theory of Free Speech Protection" (83) does not offer a very enthusiastic defense of the epistemic properties of free speech. Eugene Volokh's symposium contribution responding to critiques by scholars advocating a (non-epistemic) participatory democracy theory (84) sounds largely in distrust of government and anti-paternalism, though it does also advance a limited epistemic defense of the marketplace of ideas. Indeed, Professor Volokh acknowledges that "[b]oth the academy and public debate have their flaws, which may allow falsehood to persist despite convincing proof that it is indeed false." (85)

In conjunction with the relative dearth of scholars stepping up to defend epistemic justifications for free speech, an increasing number of scholarly critiques of the truth rationale have appeared over the last several years. The marketplace metaphor has long been a particular target of criticism, which has proceeded on several fronts. One mode of attack has suggested that the metaphor itself is inapt, and that it is therefore unlikely that any expressive version of the invisible hand will emerge to render the truth triumphant. (86) As a recent article put it, "[e]conomic markets simply offer no support for the notion that deregulating speech would produce 'truth.'" (87)

Other critiques accept the basic analogy between economic markets and expressive markets, but assert that the Court's extreme laissez-faire approach with respect to the latter is misguided. According to this argument, if economic markets are to be highly regulated by government, so too should speech markets. (88) Some who advance this argument rely on the marketplace analogy to assert that the same market failures that require a regulatory response in economic markets also require (or at least allow) regulation in the context of speech markets. (89) For example, several scholars have relied on antitrust principles to argue for a right of access to the speech markets for those whose voices would otherwise be silenced or diminished by the monopolistic tendencies of powerful speech market actors. (90) And invoking the market distortion caused by externalities, others have proposed that tort remedies for victims of harmful speech should be upheld under the First Amendment. (91)

A more recent account of the marketplace model marshals the general behavioral critique of the basic assumptions of the classical economic model of homo economicus, suggesting that truth is unlikely to prevail where human decision-making is irrational, emotional, and fraught with systematic biases. Rather than suggesting regulation to correct for these irrationalities, this approach advocates abandoning the model altogether. (92) Because "we are anything but careful, calculating, rational consumers and users of information," the marketplace of ideas is a "failed paradigm" and "the proper response ... is to discard it as a tool." (93)

Just as the classical economic marketplace model relies on a rational actor assumption, it is common to assume that the speech marketplace model would require that actors be rational in order for truth to prevail. (94) This thinking has led some First Amendment scholars to suggest that there may be some specific speech sub-markets--those in which speakers and listeners are more likely to approximate rational actors--from which truth is indeed likely to emerge. The paradigmatic (more) rational market would seem to be, by these accounts, academia, where participants are "trained to think rationally and chosen for that ability." (95) According to Schauer, "[i]n systems of scientific and academic discourse, the argument from truth has substantial validity. Those who occupy positions in these fields may not always think rationally, but we are at least willing to say they should, and are inclined to try to replace those who do not think rationally with those who will." (96)

In a similar vein, Robert Post has argued that free speech doctrine based on the marketplace of ideas model is frequently misdirected because "'truth-seeking' ... requires much more [than simply unfettered communication]. It requires an important set of shared social practices: the capacity to listen and to engage in self-evaluation, as well as a commitment to the conventions of reason, which in turn entail aspirations toward objectivity, disinterest, civility, and mutual respect." (97) Dean Post offers the cultures of academia and scholarship, including the scientific method, as examples of cultural institutions that embody these social practice norms that are "necessary for a marketplace of ideas to serve a 'truth-seeing function.'" (98) Absent such shared social norms and practices, Post and others express serious doubts as to whether a free marketplace of ideas can result in epistemic benefits. (99)

Much of this doubt seems to stem from experience in the real world. One of the most serious criticisms advanced against truth-based justifications in current free speech jurisprudence is what has been called "the First Amendment's epistemological problem. (100) The central, defining trope of the marketplace metaphor is its faith that, in the end, truth will out; that true counsels are the remedy for false ones; that our democracy demands that we have faith that people will ultimately sift truth from falsehood, good ideas from bad. A fundamental premise of the behavioral critique of the marketplace metaphor rests on the assertion that these basic assumptions are simply wrong as an empirical matter.

It is discouragingly easy to compile a long list of patently false propositions--some silly and some quite consequential--that are embraced by large numbers of people. Professor Schauer makes the point quite powerfully:

President Obama was not born in Kenya. President Bush did not have advance notice of the September 11 attacks. The predictions of astrology have neither scientific basis nor the capacity to forecast the future. AIDS was not created by white physicians and multinational pharmaceutical companies in order to reduce the size of the African and African American populations. The Holocaust is not a myth fabricated by Zionists and their supporters. (101)

And yet too many people believe the demonstrably false propositions listed and refuted in this passage. (102) Despite the prevalence of truth rhetoric in First Amendment jurisprudence, "an extremely important social issue about the proliferation of demonstrable factual falsity in public debate is one as to which the venerable and inspiring history of freedom of expression has virtually nothing to say." (103)

Thus, almost from the moment that the marketplace metaphor so powerfully seized the judicial and public imagination, it has been the subject of multifaceted scholarly critiques. Along with a more generalized skepticism about the epistemic value of free speech, the marketplace of ideas has been the site of repeated attention and review. These critiques of truth-based rationales have intensified over time, such that it is now the rare scholar who defends freedom of speech protection primarily based on its contribution to knowledge or truth. Rather, scholarly trends have shifted toward justifications grounded in (proceduralist) democratic self-governance theories, along with other non-consequential autonomy and self-realization theories. (104) Truth, if given any weight at all, is confined to a minor role. Similarly, democratic theorists have, since the "deliberative turn," largely rejected epistemic justifications for democracy in favor of proceduralist theories.

C. Alternative Accounts in Legal and Democratic Theory

In place of truth-seeking as a normative justification (105) for freedom of speech protection, participatory democracy (106) has gained significant influence in First Amendment scholarship in recent years. (107) As put forward by two of its most notable proponents, Professor James Weinstein and Dean Robert Post, this theory can be characterized as proceduralist rather than epistemic--that is, it values democratic participation not because of any claim to good outcomes but because the process itself serves certain intrinsic values. (108) As Weinstein describes the basic, shared value of democratic self-governance, "this core precept recognizes the right of every individual to participate freely and equally in the speech by which we govern ourselves." (109) This free and equal participation right is necessary to the legitimacy of the system. (110) Thus, given our society's shared commitment to democratic self-governance, protection of speech that serves participatory democracy is warranted and, indeed, necessary. This account of freedom of speech as primarily driven by the value of democratic participation (rather than truth) (111) has much in common with the major trends in democratic political theory.

Within the field of political science, democratic theory is said to have taken a decidedly "deliberative turn" beginning about 1990. (112) As in the free speech literature, this deliberative turn has largely focused on procedural justifications for democracy rather than epistemic justifications. "With the exception of a small group of 'epistemic democrats,' the question of the cognitive competence of average citizens and the related question of the performance of democratic institutions either raises profound skepticism or is avoided altogether in contemporary democratic theory, both positive and normative." (113)

Rather than arguing that democratic governance is more likely than other systems to lead to better decisions, the strong trend among democratic theorists--including deliberative democrats--is to justify democracy based on intrinsic, proceduralist values such as fairness, autonomy, dignity, respect, and equality. (114) As summarized recently by Helene Landemore, "Democratic theory has evolved in a more participatory and deliberative direction, involving, ideally at least, the public as an active political agent. Yet this normative commitment is generally disconnected from a belief in people's competence." (115) The prevailing belief among democratic theorists and political scientists "is that [the] people are irrational, uninformed, and apathetic," (116) and therefore "normative theories of democracy--not only Schumpeterian, (117) but also participatory and deliberative theories--either solve the problem in a nondemocratic way or ignore it altogether." (118)

In Part IV, I propose an epistemic deliberative democratic theory that attempts to reconcile the truth-seeking and democracy-based accounts of freedom of speech. First, however, the next Part describes in some detail the Argumentative Theory of Reasoning, which has upended the idea of reason as a rational, solitary activity designed to discover truth and replaced it with a socially embedded cognitive capacity designed for making, evaluating, and winning arguments to and with other people.


The Argumentative Theory provides an alternative perspective on the relationship between the marketplace model and the epistemological problem. To the extent that First Amendment scholars have questioned epistemic theories based on the wealth of research demonstrating that humans are irrational and that groups often make biased decisions, the ATR provides an interesting counterpoint. In the first place, these irrationalities are to be expected--they are predicted--if the function of reason is to win arguments rather than to find the truth. Furthermore, when people communicate within a context that allows their reasoning ability to function as it was evolved to do, the very features that we now call "cognitive biases" turn out to generate good epistemic outcomes. Thus, as described in more detail below, the ATR predicts that people will be more likely to reason to correct answers under certain predictable conditions, which include engaging in deliberation with others of diverse opinion, and having sufficient access to arguments in support of the correct answer. (119) This is a version of the marketplace of ideas, but it is a very particular version in which the market perhaps resembles more a bazaar than a supermarket, in which buyers and sellers actively engage in a dynamic discourse before ultimately coming to terms. (120)

It is also an account that tends to reconcile the marketplace and democratic discourse rationales for free speech protection because it explains--and predicts--the conditions under which discourse is most likely to lead to optimal epistemic outcomes. Thus, this account lays the groundwork for an epistemic deliberative theory of freedom of speech, as opposed to existing deliberative or participatory democratic theories that are proceduralist in nature.

A. What is Reason For?

The classical view of reason in both philosophy and psychology sees human reason as a path to knowledge and views the function of reason--the why--as enhancing individual cognition. Thus, "[a] long tradition has regarded human thinking as a solitary, if not solipsistic, exercise aimed at facilitating behavior. This has privileged the assumption that reasoning is mainly for enabling individuals to seek the truth." (121) As described by cognitive scientist Dan Sperber, reason under this conception:
   is typically viewed as first and foremost a property of the
   individual Cartesian thinker. Its function ... is seen as that of
   allowing the individual to go beyond perception-based beliefs and
   to discover facts with which it happens not to have had perceptual
   acquaintance, and, more importantly, theoretical facts with which
   there is no way to be perceptually acquainted. On this view,
   reasoning is a higher-level form of individual cognition, a
   superior tool for the pursuit of knowledge. (122)

The human faculty of reason thus has generally been viewed as a tool for the individual to use in order to gain epistemic benefits.

The problem, according to the ATR, is that reason is often not very good at doing this job. The classical view of reasoning as a tool for individual cognition "to help people overcome the limits of intuition, acquire better grounded beliefs, particularly in areas beyond the reach of perception and spontaneous inference, and make good decisions" is at odds with "the massive evidence that human reasoning is not so good at fulfilling this alleged function." (123) In contrast, human intuition is relatively effective at performing many of these tasks. (124) Reasoning, as an extremely energy intensive, high-cost process is thus unlikely to have evolved "as a tool for individual cognition." (125)

The ATR, as formulated by Mercier and Sperber, sees the human ability to reason not primarily as a means to enhance individual cognition but rather as a tool for "the production and evaluation of arguments in communication." (126) According to this conception, reasoning evolved and persisted in human populations because it was adaptive (127) for both speakers and listeners: Speakers reason in order to persuade others so as to influence their behavior, and listeners reason in order to critically evaluate speakers' arguments. (128) The theory has gained much attention (129) in large part because it so elegantly makes sense of masses of experimental findings about cognitive biases that are much harder to explain under classical theories of reasoning.

To understand the ATR, some preliminary points are in order. First, given that this is a theory of reason and reasoning, it would do well to clarify what Mercier and Sperber mean when they use these terms. They offer the following definition of "reasoning" as used in their model: "[r]easoning can be defined as the ability to produce and evaluate reasons." (130) Lest this appear to be a tautology, they elsewhere offer this more detailed definition:
   [r]easoning, as commonly understood, refers to a very
   special form of inference at the conceptual level, where not
   only is a new mental representation (or conclusion)
   consciously produced, but the previously held
   representations (or premises) that warrant it are also
   consciously entertained. The premises are seen as
   providing reasons to accept the conclusion. Most work in
   the psychology of reasoning is about reasoning so
   understood. Such reasoning is typically human. (131)

Understood in this way, reasoning is roughly analogous to the so-called "System 2" in dual process models of cognition that currently dominate the cognitive and social sciences. (132) Dual process theories envision two distinct cognitive systems within the brain, typically referred to as System 1 and System 2. (133) At the most general level, this two-system view "distinguished] intuition from reasoning." (134) System 1 is characterized by fast, automatic, implicit thinking that is cognitively "cheap" (that is, requires relatively less cognitive energy). (135) It is the system that we rely on for most of our daily thinking. (136) System 2, in contrast, is characterized by comparatively slow, effortful, deliberate, and analytical thinking, which requires a great deal of cognitive energy. (137) System 2 is commonly thought of as a single, serially operating ability that allows us to do such things as process abstract concepts, plan ahead, and consider and evaluate options. (138)

Most dual process theories (139) posit that System 2 monitors and corrects System 1 as needed. (140) System 1 outputs, based on heuristics, are essentially "good enough" under most circumstances; under others, though, they may lead to suboptimal decisions. (141) It is here that System 2 is thought to step up and override the initial judgment. (142) According to Kahneman, "System 2 continuously monitors the tentative judgments and intentions that System 1 produces." (143) Thus understood, System 2--reasoning proper--functions to correct mistakes that may arise through the heuristics of System 1 so that the individual can make "better" decisions.

The problem, as noted by countless critics, is that System 2 often fails to correct the errors of System 1 ; indeed, one might say that System 2 is host to its own suite of systematic biases and irrationalities (144) that can often exacerbate these errors. As summarized by Mercier and Sperber, "[s]ince the 1960s, much work in the psychology of reason has suggested that, in fact, humans reason rather poorly, failing at simple logical tasks, committing egregious mistakes in probabilistic reasoning, and being subject to sundiy irrational biases in decision making." (145) It is difficult to accept the notion that reason evolved to enhance epistemic outcomes when it in fact performs rather poorly at this task.

Experiments have now revealed many systematic and widespread logical errors, reasoning flaws, and biases in human cognition. (146) Some of the latter are implicit--non-conscious--biases that are more fruitfully understood as belonging to System 1. (147) Others, however, are biases and logical defects that infect conscious, explicit reasoning. In general, the experimental literature shows not only that "reasoning falls short of delivering rational beliefs and rational decisions reliably, but also that, in a variety of cases, it may even be detrimental to rationality." (148) Mercier and Sperber suggest that their theory predicts many of the most well documented cognitive biases and also that it "explains wide swaths of the psychological literature within a single overarching framework." (149)

In the main paper in which they lay out their theory, Mercier and Sperber address claims and findings in three main areas of experimental research: (1) logical and argumentative skills; (2) confirmation bias; and (3) motivated reasoning. (150) Through a cross-disciplinary survey of the relevant experimental research in these areas, they demonstrate that an understanding of reasoning as having primarily an argumentative function very easily explains the findings in each of these areas. As Mercier puts it, "[t]he argumentative theory has been able to account for the reasoning performance reported by experiments in the psychology of reasoning and decision-making, developmental psychology, moral psychology, political psychology, and cross-cultural psychology." (151)

B. Logical and Argumentative Skills

Many studies have shown that individuals tend to make logical errors when solving relatively simple and straightforward problems. One early review of the literature found that 40% of participants incorrectly answered a problem based on the structure "if p then q; not q, therefore not p." (152) In the well-known Wason selection task, (153) which is another kind of basic logic problem, numerous studies over the past several decades have shown that many individuals perform quite poorly, even those who are highly educated. (154)

In recent years, however, there has been a movement within the field of psychology of reasoning to reevaluate the normative judgment that something must be wrong with human reasoning abilities if most of us cannot more accurately solve these abstract logical problems. (155) In particular, research has repeatedly shown that contextualizing logical problems such that they become relevant to subjects' experience often has a significant effect on accuracy. (156) Mercier and Sperber take this observation a step further to point out that an argumentative context dramatically increases performance on logic tasks. In addition, they note that when subjects are asked to produce arguments in such contexts, they frequently are able formulate valid "if p then g" type arguments. (157)

Mercier and Sperber also address the related "idea that people are not very skilled arguers," (158) acknowledging that, if accurate, this common assumption would be fatal to the ATR. (159) In terms of argument production, Mercier and Sperber review the existing studies with a view to how people would be expected to behave if they were trying to win an argument. (160) From that perspective, they conclude that the studies support the proposition that people in fact are quite good at producing arguments. (161) Similarly, with respect to the listener side of the equation, the authors examine the experimental findings and conclude "that, when they are motivated, participants are able to use reasoning to evaluate arguments accurately." (162) When evaluating the strength of arguments and spotting logical fallacies in others' statements, again the experimental literature seems to show that context is important and that an argumentative context boosts performance significantly. (163) In sum, Mercier and Sperber assert that the studies in diverse fields show that "people can be skilled arguers, producing and evaluating arguments felicitously. This good performance stands in sharp contrast with the abysmal results found in other, nonargumentative, settings." (164)

C. Confirmation Bias and Motivated Reasoning

The term "confirmation bias" generally refers to the tendency to look for, notice, and give undue weight to evidence that supports one's prior positions. This particular cognitive bias is "extensive and strong and . . . appears in many guises." (165) Though it can be implicit or explicit--and is well documented in both contexts--the term more often "connotes a less explicit, less consciously one-sided case-building process. It refers usually to unwitting selectivity in the acquisition and use of evidence." (166) Under the classical view of reason, in which a solitary thinker uses her higher faculties to reach the epistemically superior conclusion, confirmation bias is clearly both backwards and counterproductive.

There is a great deal of empirical evidence of confirmation bias; furthermore, it is evident across diverse groups and--unlike illogical reasoning--does not appear to be correlated with general intelligence. (167) Viewed from the vantage of the ATR, this tendency to search out, pay attention to, and give greater weight to facts that support one's positions (and, on the flip side, to seek evidence that disconfirms opposing views) is precisely what one would expect if the primary function of reasoning is to win arguments. Furthermore, the poor epistemic and policy outcomes that are often said to arise out of confirmation bias dynamics occur in situations that are not "felicitous" argumentative settings. (168) In contrast, "when reasoning is used in a more felicitous context--that is, in arguments among people who disagree but have a common interest in the truth--the confirmation bias contributes to an efficient form of division of cognitive labor" and has much less tendency to lead to poor outcomes. (169)

On the contrary, when people with diverse views deliberate--that is, reason in an argumentative context--each individual's confirmation bias ensures that he will gather the most compelling evidence and arguments that support his existing beliefs, as well as search for compelling evidence and arguments that undermine the points of those advancing opposing arguments. Rather than resulting in poor epistemic outcomes, in this context the collection of confirmation biases playing off against one another should tend to lead the group to optimal epistemic outcomes. Experimental evidence bears out this prediction. (170)

Motivated reasoning takes confirmation bias a step further: not only do we selectively notice and give disproportionate weight to evidence that tends to support our views, we also interpret evidence in a way that distorts it so that it tends to support our views. There are many sub-issues in the literature on motivated reasoning, (171) but at its core the idea is that "[w]e often ignore new contradictory information, actively argue against it or discount its source, all in an effort to maintain existing evaluations." (172) Indeed, recent neuroimaging experiments have shown that different regions of the brain are activated during motivated reasoning as compared to pure logical reasoning. (173)

Much of the research on motivated reasoning involves questions of emotion and affect and their relationship to cognitive processes. For example, a number of studies have shown that one very strong motivation that drives reasoning is the desire for positive self-evaluation. "In general, individuals tend to (1) give more credence to, and be more optimistic about, the validity of information that supports or confirms their standing as kind, competent, and healthy people; and (2) be more skeptical and cautious about information that threatens this standing." (174) And, particularly notable to those interested in freedom of speech and democratic theory, affective motivations regarding politics and political worldviews drive a great deal of political reasoning. (175)

As with the confirmation bias, this rather pervasive feature of human reasoning, which would be terribly bad design were reasoning meant to lead an individual to good answers, is to be expected if the function of reasoning is to win arguments. Given the mounting evidence that emotion is crucial to decision-making, it is not surprising that our emotions would push us to want to support our beliefs and to undermine opposing beliefs. And yet it is also the case that there are limits to the extent that people will allow their motivations to override epistemic outcomes. When faced with so-called "reality constraints," people will indeed reason toward a correct answer even if it contradicts their preferred views. (176) As with the confirmation bias, such intrusions of "reality" are much more likely to happen in the course of deliberation with others of diverse and contradictory opinions.

D. Epistemic Vigilance and the Dialectical Nature of Reasoning

In the model presented by the Argumentative Theory of Reason, the relative role of the person engaged in reasoning and the social dynamics of the reasoning process are significant. A person who occupies the role of attempting to persuade others to her view should be expected to search for and prefer arguments that support her position--in other words, to exhibit a strong confirmation bias. (177) As noted above, under this view such a feature of reasoning is not a design flaw but rather "a wonderfully designed argumentative device," (178) but only--or at least especially--when it is used within an appropriate argumentative context.

The person in the role of listener, on the other hand, should be expected to exercise "epistemic vigilance." (179) Because speakers and listeners (180) have divergent interests, (181) their cognitive mechanisms and behaviors should operate differently depending on their respective roles in any particular exchange. Epistemic vigilance refers to "a suite of cognitive mechanisms ... targeted at the risk of being misinformed by others." (182) These include two broad categories of vigilance mechanisms: those that involve vigilance toward the source of information (the speaker) and vigilance toward the content of information (its substance). With respect to the source of information, much of this evaluation appears to be implicit, with experimental data suggesting that people tend to make automatic and immediate--though frequently inaccurate--judgments about others' general trustworthiness as well as whether they are being truthful in a specific instance based on appearance and certain behavioral cues. (183) In addition, people rely strongly on reputation and other indications of "character" or "personality" to attribute traits such as trustworthiness to people so as to assess their behavior on a particular occasion. (184) Social context is also relevant to judgments of trustworthiness. The listener's assessment of a speaker's "benevolence" is affected by their relationship (biological or otherwise), as well as an assessment of what might seem to be more relevant factors: the speaker's general competence and her competence specifically on the subject on which she is speaking. (185)

In terms of reasoning as understood by the ATR, however, the epistemic vigilance mechanisms that focus on content, as opposed to source, are more central to explaining the process of reasoning on the listener side. (186) Primary among these is coherence checking: the comparison of communicated information against one's existing beliefs. (187) As summarized by Sperber et al., inherent in the interpretation of a communication is a process of checking the new information against available relevant perceptions and beliefs. (188) Sometimes this process will result in inconsistencies between the communicated information and the listener's own sensory perceptions, his prior substantive beliefs, or his beliefs about the reliability of the source. (189) "When such inconsistencies or incoherences occur, they trigger a procedure wholly dedicated to such assessment" of one's own beliefs. (190)

This is where the dialectical nature (191) of reasoning becomes crucial. In contrast to the classical image of the solitary reasoner parsing information and arguments to come to an epistemically superior conclusion, the Argumentative Theory imagines a dynamic social process that considers the mind and behavior of both (or all) parties in the relevant deliberative group. (192) From the speaker's perspective, there is a desire to persuade the audience but also an understanding that the audience will exercise some degree of epistemic vigilance. (193) If, in the speaker's assessment, the audience is likely to be resistant to her message either for reasons of source or content vigilance, she is going to have to come up with arguments-reasons--that might persuade the resistant listener to accept her position. (194) In so doing, as noted above, she should be expected to marshal the evidence in favor of her position rather than to be evenhanded--in other words, to exhibit confirmation bias. (195) The listener, for her part, may then assume the role of speaker (196) to muster arguments against the first speaker's position and in favor of her own or, having exercised epistemic vigilance, may be persuaded to adopt the first speaker's position. If there is, in addition to these two individuals, a group or other audience, this dynamic will take on additional features, as considered in the following section. (197)

E. Group Decision-Making

The findings of the experimental literature on the benefits of a group context for decision-making are decidedly mixed. (198) The epistemic benefits of so-called "crowd wisdom" have gained a great deal of attention in recent years; (199) much of the literature supports a general "truth wins" conclusion. (200) Sometimes, however, groups small and large make much worse decisions than could have been made by their best individual members- lay or expert. (201) They often succumb to dynamics of group polarization, (202) cascade effects, (203) and "group-think" (204) that represent the opposite of the rational deliberation ideal. Indeed, these effects may be even worse when the group members are experts: arguably, "groups of experts are responsible for some of the worst decisions and the weirdest beliefs in history." (205)

One aspect of the experimental literature that is important to this discussion is the distinction between simple aggregation of individuals' answers or predictions and answers reached following deliberation among group members. Under certain conditions, simple aggregation and averaging can indeed lead to remarkable epistemic outcomes. (206) The addition of some kind of interaction among the members of the group, however, consistently allows the group to perform better on both prediction and problem-solving tasks than groups in which members act independently. (207) With respect to logical, mathematical, or factual problems that have a demonstrably correct answer, the research overwhelmingly demonstrates that groups working together perform much better than individuals in coming to a correct result, and furthermore that "debates are essential to any improvement of performance in group settings." (208)

When suboptimal outcomes emerge from group decision-making, they are often predicted and explained by the ATR. Group polarization, for example, is most extreme when discussion occurs within homogenous groups in which all members start out with the same beliefs. Under those circumstances, confirmation bias and motivated reasoning occur in an echo chamber, causing the group to polarize such that they become even more strongly convinced of the original belief. (209) With nobody inclined to exercise epistemic vigilance nor motivated to marshal arguments for contrary positions, the position of the group tends to become more extreme. Since it is likely that various members of the group will contribute more facts and arguments than any one of them would possess on his own, after such a discussion the group will be left with an even stronger case in favor of its original position. If it was incorrect or unwise to begin with, the process of group discussion will exacerbate those flaws.

However, when groups work together under certain conditions, they are much more likely to realize epistemic benefits. Not only that, they frequently are smarter than even their smartest members. This phenomenon, by which the sum of the group can be greater than its individual parts, has been examined across several academic fields and is discussed further in Part IV. (210) As relevant to the ATR, the crucial point is that discussion, within an argumentative context, in which members of the group disagree and deliberate, can mean the difference between good epistemic outcomes and abysmal ones.


In this Part, I lay out the theoretical case for a new epistemic justification for freedom of speech. In doing so, I draw on insights from democratic theory, psychology, and the literature on group decision-making. Existing Supreme Court First Amendment jurisprudence values truth and knowledge, but it has approached questions of speech and truth with the bluntest of tools: classical economics and classical models of human reason. Yet, as this Part suggests, the seeds of an argumentative approach are already scattered throughout the landscape of First Amendment case law. A more nuanced model that accounts for how people actually reason in a communicative context would allow the Court to analyze many of these issues from a more solid theoretical foundation, bringing its doctrine into line with the empirical record.

A. The Epistemic Case for Inclusive Deliberative Democracy

As against the strong trend, described above, toward proceduralist deliberative theories and against epistemic justifications, a small group of scholars has reinvigorated the epistemic case for democracy. Focusing here on deliberative democratic theory, (211) the epistemic case for deliberative democracy has emerged in recent years in the work of theorists such as David Estlund (212) and empiricists such as Josiah Ober. (213) With respect to the theoretical case, Jose Luis Marti lays out a convincing argument that deliberative democrats must necessarily accept that the procedures they advocate have some epistemic grounding, though they rarely do so in explicit terms. (214) Estlund, a deliberative democrat who does make the case explicitly, offers a defense of a weak epistemic case for deliberative democracy. He proposes what he calls "epistemic proceduralism," which is a theoretical argument that deliberative democracy is superior to a random decision procedure--such as a coin toss--in its tendency to result in epistemically "correct" outcomes, and that it is the only such procedure that is justifiable based on the values of public reason. (215)

Helene Landemore, on whose work this section heavily draws, offers instead a sustained defense of a much stronger version of the epistemic case for democracy. (216) She argues that not only is deliberation better than a random decision procedure, but it is actually more likely than other procedures (217)--including those that don't necessarily satisfy the demands of deliberative democratic theory--to result in better outcomes. And as to what kinds of outcomes count as "better," Landemore goes beyond simply arguing that inclusive deliberation is more likely to avoid bad decisions; she argues that it is more likely affirmatively to approach optimal decisions. Her argument is grounded upon two emerging and interdisciplinary bodies of work: Sperber and Mercier's Argumentative Theory of Reasoning and findings regarding the existence of "collective intelligence," including most significantly work on cognitive diversity by Lu Hong and Scott Page. (218)

The concept of collective intelligence posits that the decisions of a group can be smarter than the decisions of even the smartest individual group members. (219) In animal behavior research, this is known as "swarm intelligence" and is a well-recognized phenomenon of group behavior in the social insects. (220) In humans, the idea is somewhat more controversial but has gotten much recent attention in the context of aggregative group decision-making such as prediction markets. (221) However, deliberation groups are often viewed as pernicious to good decisions. (222) Ultimately, social interaction among group members may sometimes lead to phenomena such as groupthink, polarization, and cascade effects, which are said to create the danger that group decisions will be much worse than the decisions of even average group members deciding alone (or at least independently of the group).

However, a growing body of work demonstrates that, under certain conditions, (223) group decisions made with the benefit of deliberation can be superior to those made without discourse (224) and the decisions of the smartest individual(s) in the group. (225) In addition to a deliberative context, a crucial condition that enables good decision-making is the existence of cognitive diversity within the group. (226) According to the work of Lu Hong and Scott Page, and under conditions that are not terribly unrealistic, (227) "a random collection of agents drawn from a large set of limited-ability agents typically outperforms a collection of the very best agents from that same set. This result is because, with a large population of agents, the [more expert] group, although its members have more ability, is less diverse." (228) They simplify, stating: "diversity trumps ability." (229)

"Cognitive diversity" here means something very specific; it is not coextensive with cultural, ethnic, or other types of social identity diversity, though the two forms of diversity may frequently be correlated. It refers to "the fact that people make predictions based on different models of the way the world works or should be interpreted." (230) Hong and Page distinguish "functional diversity," meaning "differences in how people represent problems and how they go about solving them" from "differences in ... demographic characteristics, cultural identities and ethnicity, and training and expertise." (231) The former is what they label cognitive diversity, while the latter is identity diversity. More specifically, cognitive diversity encompasses differences in the way that individuals internally represent problems (different "perspectives") and differences in the "algorithms that they use to locate solutions" (different "heuristics.") (232) Thus, "even if we were to accept the claim that IQ tests, Scholastic Aptitude Test scores, and college grades predict individual problem-solving ability, they may not be as important in determining a person's potential contribution as a problem solver as would be measures of how differently that person thinks." (233)

Landemore relies on these counterintuitive findings to develop a theory of inclusive democratic deliberation, according to which we can expect deliberation to lead to increasingly better epistemic results the more inclusive that deliberation is. This is because, as she argues, the more people who are included in the deliberative group, the more cognitive diversity there is likely to be. And the more cognitive diversity among the deliberative group, the more likely it is to reach better decisions. When Mercier and Sperber's Argumentative Theory of Reasoning is added to the mix, it helps to explain why cognitive diversity and deliberation can lead to epistemic benefits. A group composed of people who begin with different perspectives and different approaches to solving a problem, each motivated to marshal arguments for their views and against opposing views raised by other (cognitively diverse) individuals, will be most likely to put their built-in cognitive biases to good epistemic use. They will be more likely to reason in an argumentative context to the right answer. And when the group is inclusive and thus (on average) more cognitively diverse, it will be more likely to include individuals whose perspectives, heuristics, interpretations, and predictive models offer optimal global solutions to various aspects of the group's decision-making or predictive problem solving ability.

Furthermore, because every society is continually faced with novel and ever-evolving issues and problems, it can't easily be predicted in advance which particular individuals will happen to possess the best ideas and solutions at different times and to particular problems. In Landemore's view, this is why tendencies toward "epistocracy" are often misguided. While there is certainly a place for problem-solving and prediction by experts, (234) that place may be narrower than many people currently suppose. Conversely, the universe of problems amenable to optimal solutions based on inclusive deliberation appears to be much more expansive than critics of epistemic democracy and truth-based justifications for freedom of speech typically allow.

B. The Deliberative Epistemic Case for Freedom of Speech

Applying these insights in the context of the First Amendment, there is a powerful case to be made that there remains a place in theory as well as in doctrine for a truth-based justification for freedom of speech. Furthermore, that place is not confined to narrow realms of arguably "reasoned" (in the classical sense) discourse, such as academia or peer reviewed scholarship. (235) Rather, there are good reasons to protect a much wider range of speech on epistemic grounds. In contrast to proceduralist values of equality, autonomy, and legitimacy, (236) an epistemic approach is justified by the prospect of enhanced substantive outcomes. Consequently, while the "coverage" of the First Amendment under a deliberative epistemic theory is likely to map rather closely onto what a discourse-based theory of speech would consider within the First Amendment's purview, (237) the underlying value that drives such protection is distinct and its precise borders may therefore vary. The point is not that the intrinsic values embodied in democratic participation are unimportant, but simply that protecting democratic discourse--in the form of distributed deliberation--in this way can also serve epistemic values.

Reflections of an argumentative, deliberative account of reason can be found throughout much of the free speech case law. The cases involving the First Amendment's constraints on liability for defamation provide one striking example. This iconic passage from New York Times Co. v. Sullivan very closely tracks the features of the ATR: "[w]e consider this case against the background of a profound national commitment to the principle that debate on public issues should be uninhibited, robust, and wide-open, and that it may well include vehement, caustic, and sometimes unpleasantly sharp attacks on government and public officials." (238) Justice Brennan, speaking for the Court, most certainly is concerned with democratic legitimacy and participation. But there are also references that invoke the notions of argument, cognitive bias, and epistemic values. Consider this passage:
   To persuade others to his own point of view, the pleader,
   as we know, at times resorts to exaggeration, to vilification
   of men who have been, or are, prominent in church or
   state, and even to false statement. But the people of this
   nation have ordained in the light of history, that, in spite of
   the probability of excesses and abuses, these liberties are,
   in the long view, essential to enlightened opinion and right
   conduct on the part of the citizens of a democracy. (239)

A "debate" that is "uninhibited, robust, and wide-open" (240) sounds very much like the kind of deliberation that is most likely to lead, according to the ATR, to desirable epistemic outcomes. The reference to the "probability of excesses and abuses" (241) recognizes that confirmation bias and motivated reasoning are inevitable features of human reasoning. And the "right conduct" and "enlightened opinion" that are ultimately to be gained from such public deliberation are precisely the good epistemic outcomes that are predicted by the ATR and the epistemic case for inclusive democratic deliberation.

This view of reasoning also helps make sense of Sullivan's actual malice standard, under which false statements about public figures may only be punished if made with knowledge that they were false or reckless disregard for whether they were false. (242) Some scholars view this strong protection of falsehood as difficult to reconcile with epistemic notions of freedom of speech. But placed in an argumentative framework, with its attendant confirmation bias and motivated reasoning, it is to be expected that false or misleading statements will be made, and that very often these statements will not be purposefully false. Drawing the line at actual malice--that is, protecting falsehoods unless the speaker was actually conscious of the probable falsity of the speech--protects argumentative speech with its concomitant (and usually unconscious) confirmation bias. Thus understood, the robust debate protected by Sullivan and its progeny--rather than being a necessary evil aimed merely at giving "breathing room" to valuable speech (243) or keeping the government out of the business of deciding what is true--(244) is part and parcel of the search for truth.

The lesser protection for speech about private figures or on matters of private concern also makes sense in a way that is distinct from the very fact that they are not a part of "public discourse" as proposed by scholars who advance a proceduralist participatory democracy explanation. (245) More specifically, from an epistemic deliberative democratic perspective, such speech can be often thought of as a component of what Landemore calls "distributed deliberation," (246) which she defines "as a decentralized and fragmented deliberative process distributed over different people or clusters of people, ... as well as over time." (247) Along with distributed cognition, (248) distributed deliberation can generate epistemic benefits that are greater than the sum of the group's parts. False statements about private figures that do not involve issues of public concern are much less likely to comprise a component of even this broadly defined mechanism of distributed democratic deliberation. It makes sense, then, that they should enjoy lesser protection when considered from an epistemic deliberative perspective.

A similar point was emphasized in Snyder v. Phelps in the course of the Court's discussion of the relative First Amendment value of the Westboro Baptist Church's hateful speech directed toward the grieving family of a deceased veteran. (249) The reason that "speech on purely private matters does not implicate the same constitutional concerns as limiting speech on matters of public interest" is that the former does not threaten "the free and robust debate of public issues; there is no potential interference with a meaningful dialogue of ideas ...," (250) Other cases recognize a First Amendment "right to persuade ...," (251) Protection of this right to persuade was on display in McCullen v. Coakley, (252) which invalidated an abortion clinic buffer law that prevented anti-abortion speakers from engaging in face-to-face "counseling" and "conversation" with women entering the clinics. Though it did not reference truth as such, the Court did emphasize evidence in the record to the effect that other forms of speech available to the anti-abortion speakers had not been as effective in persuading women to forego abortions as were the face-to-face sidewalk conversations restricted by the Massachusetts law. (253) In so doing, the Court implicitly recognized the dynamics of persuasion and deliberation inherent in the epistemic deliberative account.

There are additional examples of the confluence of existing doctrines with the contours of the epistemic case for freedom of speech. For one, the protection and promotion of cognitive diversity--in the form of dissent--is a bedrock principle of the First Amendment. (254) A theory grounded in proceduralist democratic values would protect dissenters not because their speech is more likely to lead to the truth, but because of the importance of their equal right to participate in public discourse and self-government. Such an approach, however, would not affirmatively welcome dissent and debate: if those individuals agreed with the majority, their voices would equally be included and valued. Only an epistemic approach values dissent because it is contrary to the majority view. (255) As Justice Stevens put it (perhaps fittingly, in dissent) in Morse v. Frederick, "a rule that permits only one point of view to be expressed is less likely to produce correct answers than the open discussion of countervailing views." (256) Scott Page notes that dissent is a manifestation of cognitive diversity because it is generated by those with different perspectives and heuristics than others in the group. (257)

The "diversity trumps ability" theorem, as noted above, requires as one condition that the members of the group be "smart." Some might worry that this condition would doom its applicability in the real world. But again, First Amendment doctrine lines up with the epistemic case. The "smart" condition does not refer to IQ, but to knowledge: the members of the group must know something about the problem about which they are deliberating. They do not have to be very smart, but they cannot be completely ignorant either. To the extent that the Court protects-or even requires-the dissemination of information relevant to solving social, political, or scientific problems, it serves the purpose of making the group smart and therefore advances the epistemic value of free speech. (258)

Insofar as this epistemic deliberative argument for freedom of speech tends to suggest similar outcomes to the proceduralist participatory democratic argument, one might wonder whether it matters that First Amendment scholars have largely rejected the idea that freedom of speech can lead to truth. I propose that it does matter, because a necessary foundation for productive deliberative problem-solving, good predictions, and innovation is that the group have some common sense of purpose and desire to reach optimal results. Furthermore, cognitively diverse individuals must choose to engage with the group, and thus must believe that their contributions matter to the outcome. Dismissing truth as a fantasy in democratic deliberation and free speech scholarship might prove to be something of a self-fulfilling prophecy. In addition, research from the Good Judgment Project has shown that some of the individual cognitive abilities conducive to epistemic success can be cultivated with practice. (259) Admitting the possibility that people deliberating in groups, free to gather information and advance their confirmation bias-infected and motivated arguments, employing diverse perspectives and heuristics, might in many cases generate better epistemic outcomes than small groups of very smart experts would encourage people to cultivate these skills. (260)


Ironically, at the very moment that many First Amendment scholars, as well as many political scientists, have abandoned epistemic rationales for speech protection and deliberative democracy, the Argumentative Theory of Reasoning offers good cause to give truth another look as a basis for protecting freedom of speech. Examined through the lens of the Argumentative Theory, the conflicting behavioral data on individual and group decision-making supports the view that the ideal context for optimal epistemic outcomes is one of deliberation between and among cognitively diverse individuals, with dissent protected and indeed promoted in the service of better decisions. The marketplace of ideas may indeed be a place where truth is better realized, so long as these conditions are protected and promoted.

(1.) John Milton, Paradise Lost, Book II, lines 112-15 (1667; 1674).

(2.) Dun & Bradstreet, Inc. v. Greenmoss Builders, Inc., 472 U.S. 749, 764 (1985) (Burger, C.J., concurring).

(3.) As with much of what is to follow, even this simple statement hides much complexity. "Rational" here is used in the "weak," categorical sense of being based on reason. The opposite of "rational" used in this weak sense is not "irrational," but "arational." ... As Keith Stanovich points out, cognitive scientists use the term in its strong sense, meaning that a behavior is irrational when it "departs from the optimum prescribed by a particular normative model 49.. The scientist is not implying that no thought or reason was behind the behavior." See Keith e. Stanovich, Rationality and the Reflective Mind 4 (2011).

(4.) See, e.g., Michael S. Pardo, Introduction to the Meador Lectures on Rationality, 64 Ala. L. Rev. 141, 142-43 (2012) ("Our power to be rational ... helps to define what makes human animals special and, well, human"). There is, of course, much debate about whether other animals possess the capacity to reason, mostly turning on the definition of "reason." See, e.g., Rene Descartes, Animals are Machines, in environmental ethics: divergence and Convergence 281 (S.J. Armstrong and R.G. Botzler eds., McGraw-Hill. NY 1993); Hans-Johann Glock, Can Animals Act For Reasons?, 52 Inquiry 232 (2009) (entire issue devoted to the topic of human and non-human animal agency).

(5.) Hugo Mercier, Some Clarifications About the Argumentative Theory of Reasoning: A Reply to Santibanez Yanez, 32 Informal Logic 259, 259 (2012) [hereinafter Mercier, Clarifications]; see, e.g., Daniel Kahneman, A Perspective on Judgment and Choice, 58 Am. Psychol. 697, 699 (2003) ("[0]ne of the functions of System 2 [elsewhere described as "reason"] is to monitor the quality of both mental operations and overt behavior.").

(6.) See STANOVICH, supra note 3, at 7 ("A substantial research literature--one comprising literally hundreds of empirical studies conducted over several decades--has firmly established that people's responses sometimes deviate from the performance considered normative on many reasoning tasks.").

(7.) Or, when we recognize deviations from the ideal, we believe that those only happen to other people. See Jamie T. Kelly & Kristoffer Ahlstrom-Vij, Epistemic Perfectionism and Liberal Democracy, 29 Soc. PHIL. TODAY 49, 52 (2013) ("We may acknowledge that "among the general population various forms of irrationality and epistemic corruption prevail,' while still maintaining that.... it never seems like that from the inside."); Dan M. Kahan, Foreword: Neutral Principles, Motivated Cognition, and some Problems for Constitutional Law, 125 Harv. L. Rev. 1, 21-22 (2011) (discussing naive realism, whereby people--correctly--perceive that opponents are engaging in motivated reasoning but--naively--believe that their own judgments are based purely on objective fact).

(8.) See, e.g., John A. Bargh, Our Unconscious Mind, 310 Sci. Am. 30 (2014); John A. Bargh et al., Automaticity in Social-Cognitive Processes, 16 trends in cognitive Sci. 593, 593 (2012); see also Julie Seaman. Hate Speech and Identity Politics: A Situationalist Proposal, 36 Fla. St. U. L. Rev. 99, 119 (2008) [hereinafter Seaman, Hate Speech] (summarizing research on social identity and priming effects as well as other implicit situational effects on behavior and decision making).

(9.) See, e.g., Antoine Bechara, The Role of Emotion in Decision-Making: Evidence from Neurological Patients with Orbitofrontal Damage, 55 Brain and cognition 30, 37-39 (2004); Jonathan Haidt, The Emotional Dog and Its Rational Tail: A Social Intuitionist Approach to Moral Judgment, 108 Psychol. Rev. 814, 815 (2001).

(10.) See Raymond s. Nickerson, Confirmation Bias: A Ubiquitous Phenomenon in Many Guises, 2 Rev. Gen. Psychol. 175, 175 (1998).

(11.) See Miles Hewstone. Mark Rubin. & Hazel Willis, Intergroup Bias, 53 ann. Rev. Psychol. 575 (2002).

(12.) See Dennis Chong & James N. Druckman, Framing Theory, 10 Ann. Rev. Pol. Sci. 103 (2007).

(13.) See Neal J. Roese & Kathleen D. Vohs, Hindsight Bias, 1 Persp. On Psychol. Sci. 411 (2012).

(14.) See Ziva Kunda, The Case for Motivated Reasoning, 108 Psychol. bull. 480 (1990).

(15.) A Wikipedia entry currently lists more than one hundred such cognitive biases. See List of Decision-Making, Belief, and Behavioral biases, (last visited February 10, 2016).

(16.) See Behavioral Law and Economics 1-3 (Cass R. Sunstein ed., 2000); Christine Jolis, Cass R. Sunstein & Richard Thaler, A Behavioral Approach to Law and Economics, 50 Stan. L. Rev. 1471, 1471 (1998).

(17.) See. e.g., Anthony T. Kronman, The Lost Lawyer 166 (1993) ("[T]he intellectual movement that has had the greatest influence on American academic law in the past quarter-century ... [is] law and economics.").

(18.) See Joshua D. Wright & Douglas H. Ginsburg, Behavioral Law and Economics: Its Origins, Fatal Flaws, and Implications for Liberty, 106 Nw. U. L. Rev. 1033, 1054 (2012) (noting that, "law professors have produced hundreds of [publications incorporating behavioral economics] in a relatively short time").

(19.) See Donald C. Langevoort, Behavioral Theories of Judgment and Decision Making in Legal Scholarship: A Literature Review, 51 Vand. L. Rev. 1499, 1499 (1998) (describing the trajectory of this research through review of the literature through the late 1990s). Langevort groups the literature into eleven broad subject categories, including contract law, tort law, criminal law and litigation. Neither constitutional law generally nor First Amendment law in particular appears on the list. See also Francis X. Shen, The Law and Neuroscience Bibliography: Navigating the Emerging Field of Neurolaw, 38 Int'l J. Legal Info. 352, 352 (2010) (examining the implication of neuroscience breakthroughs to various legal issues).

(20.) See Paul Horwitz, Free Speech as Risk Analysis: Heuristics, Biases, and Institutions in the First Amendment, 76 Temp. L. Rev. 1, 6 (2003) ("[F]ew if any writers have asked what applications behavioral analysis may have for issues in constitutional law, including the interpretation of the First Amendment.") [hereinafter Horwitz, Risk Analysis]; Lyrissa Barnett Lidsky, Nobody's Fools: The Rational Audience as First Amendment Ideal, 2010 u. III. L. Rev. 799, 804 n.21 ("Although there is a large and growing body of literature on behavioral economics and cognitive psychology, only two scholars seem to have applied this literature in the First Amendment context.") (citing Derek E. Bambauer, Shopping Badly: Cognitive Biases, Communications, and the Fallacy of the Marketplace of Ideas, 77 U. Colo. L. Rev. 649, 651 (2006); Paul Horwitz, Free Speech as Risk Analysis: Heuristics, Biases, and Institutions in the First Amendment, 76 Temp. L. Rev. 1 (2003)). See also Kahan, supra note 7, at 24 (employing the First Amendment to illustrate his thesis about cognitive illiberalism and the neutrality crisis in constitutional law more generally); Jeffrey Stake, Are We Buyers or Hosts? A Memetic Approach to the First Amendment, 52 Ala. L. Rev. 1213, 1214 (2001).

(21.) It seems uncontroversial to observe that the economic interpretation of the market metaphor has been the opinion's main legacy, but see Vincent Blasi, Holmes and the Marketplace of Ideas, 2004 Sup. Ct. Rev. 1, 1-4 (2004) for a differing opinion about the centrality of economic thinking in Justice Holmes's Abrams dissent. Furthermore, as recently characterized by a leading First Amendment scholar, the purpose of freedom of speech under the "marketplace of ideas theory, is cognitive; [under this theory] the purpose of First Amendment protections for speech is said to be 'advancing knowledge and discovering truth.'" Robert C. Post, Democracy, Expertise, Academic Freedom: A First Amendment Jurisprudence for the Modern State 6 (2012) [hereinafter Post, Democracy, Expertise] (emphasis added).

(22.) See, e.g., Horwitz, Risk Analysis, supra note 20, at 6 ("Much of our current free speech jurisprudence is based on the assumption that the government should not regulate speech because, in an unregulated marketplace, people will be perfectly capable of responding rationally to speech.").

(23.) See Hugo Mercier & Dan Sperber, Why do Humans Reason? Arguments for an Argumentative Theory, 34 Behav. & BRAIN Sci. 57, 66-71 (2011) [hereinafter Mercier & Sperber, Argumentative Theory], More than forty years ago, sociologist Murray Davis proposed that theories gain traction not because they are correct but because they are interesting, and he went on to explore the features that make a theory interesting. See Murray S. Davis, That's Interesting! Toward a Phenomenology of Sociology and a Sociology of Phenomenology, 1 Phil. Soc. Sci. 309, 309 (1971). It may well turn out that part of what attracts people to accept an argument is that it is "interesting" based on these features.

(24.) Mercier & Sperber, Argumentative Theory, supra note 23, at 58-59. In evolutionary theory, the function of a trait is the purpose for which it evolved: "a function of a trait is an effect of that trait that causally explains its having evolved and persisted in a population: Thanks to this effect, the trait has been contributing to the fitness of organisms endowed with it." Id. at 59. This does not mean that a trait cannot also be used to do other things, but it will be most efficient at doing the thing it evolved to do. For one common example, some people may be able to use their hands for walking, but hands evolved not for locomotion but for manipulation and grasping and thus will be much more efficient at the latter tasks than the former.

(25.) The theory has received accolades from such noted cognitive scientists as Steven Pinker of Harvard ("The Argumentative Theory is original and provocative, has a large degree of support, and is strikingly relevant to contemporary affairs, including political discourse, higher education, and the nature of reason and rationality. It is likely to have a big impact on our understanding of ourselves and current affairs.") and Jonathan Haidt of NYU ("[T]he article is one of my favorite papers of the last ten years. I believe that they have solved one of the most important and longstanding puzzles in psychology: why are we so good at reasoning in some cases, but so hopelessly biased in others?"). The Argumentative Theory: A Conversation with Hugo Mercier, Edge (April 27, 2011), [hereinafter Mercier, Conversation], As another reflection of the theory's impact, it was featured in an issue of Behavior and Brain Sciences, a leading peer reviewed Cambridge University Press journal in the fields of the behavioral sciences and neurosciences. In addition to the "target article" by Mercier and Sperber, the issue contained some thirty-seven responses, as well as a reply by the authors.

(26.) See Mercier & Sperber, Argumentative Theory, supra note 23, at 58.

(27.) See, e.g.. Philip E. Tetlock & Barbara A. Meilers, The Great Rationality Debate, 13 Psychol. Sci. 94, 94-97 (2002) (reviewing Choice, Values, and Frames (Daniel Kahneman & Amos Tversky eds., 2000)). The ATR is thus responsive to one of the main critiques of the behavioral project, which is that it has largely failed to move beyond the descriptive phase of cataloguing irrationalities and biases to offer an integrative theory that can make sense of the masses of experimental findings. See, e.g., Wright & Ginsburg, supra note 18, at 1040 ("The theory-building exercise thus far has focused largely upon the effort to catalog circumstances in which economic decisionmakers appear systematically to depart from rational choice behavior. The second step required to make the theory of errors relevant to policy is to map the conditions under which specific errors are more or less likely to affect decisions and then to generate estimates of the social costs imposed by those errors.").

(28.) Though he didn't use the exact phrase, the source of the marketplace of ideas metaphor is Justice Oliver Wendell Holmes' famous dissent in Abrams v. United States, and in particular this iconic passage: "But when men have realized that time has upset many fighting faiths, they may come to believe even more than they believe the very foundations of their own conduct that the ultimate good desired is better reached by free trade in ideas- that the best test of truth is the power of the thought to get itself accepted in the competition of the market, and that truth is the only ground upon which their wishes safely can be carried out." 250 U.S. 616, 630 (1919) (Holmes, J., dissenting).

(29.) As elaborated in infra Part 11(C), the marketplace of ideas is a particular manifestation of the more general category of truth-based, epistemic justifications for freedom of speech protections.

(30.) See e.g., Lidsky, supra note 20, at 802; Frederick Schauer. Facts and the First Amendment, 57 UCLA L. Rev. 897, 91 I (2010) [hereinafter Schauer, Facts].

(31.) McCullen v. Coakley, 134 S. Ct. 2518, 2529 (2014) (quoting FCC v. League of Women Voters of Cal.. 468 U.S. 364, 377 (1984)).

(32.) United States v. Alvarez, 132 S. Ct. 2537, 2550 (2012) (quoting Abrams, 250 U.S. at 630. See also Bambauer, supra note 20, at 654-73 (providing a detailed summary of the influence of the marketplace metaphor in First Amendment opinions); Thomas W. Joo, The Worst Test of Truth: The "Marketplace of Ideas" As Faulty Metaphor, 89 Tul. L. Rev. 383, 396 (2014) [hereinafter Joo, The Worst Test of Truth].

(33.) This is the terminology used in the political science literature. A proceduralist account values democracy for its intrinsic, rather than instrumental (epistemic) qualities. Democracy is preferred because it is a process that reflects and respects the equality and autonomy of all citizens. See, e.g., Russell Muirhead, The Politics of Getting it Right, 26 Critical Rev. 115, 125 (2014) ("Elections, or counting citizens equally, are simply the most convincing way of recognizing the equality of citizens.").

(34.) That is not to say that proceduralist democracy-based rationales are absent from the case law--they certainly are not--but only that the truth-based rationales that are also prominent in the case law are virtually absent from contemporary scholarship.

(35.) See John S. Dryzek, Deliberative Democracy and Beyond: Liberals, Critics, Contestations 1 (2000); Helene Landemore, Democratic Reason 27-29 (2013) [hereinafter Landemore, Democratic Reason].

(36.) I frequently encounter skepticism among people outside the field of political science about this descriptive claim. As evidence, consider the claim of epistemic democracy advanced by David Estlund. Estlund states that if fair and equal process were really all that democratic theorists cared about, then a coin flip would be as acceptable as, say, majority vote because both are equally fair from an equality of process standpoint. That they would reject a coin flip as a process demonstrates, according to Estlund, that even proceduralists implicitly accept the premise that democratic decision-making is more likely than a random process to result in a good outcome. Notice how weak Estlund's epistemic defense of democracy is: it is preferable on epistemic grounds to a coin flip. That this claim is controversial demonstrates the extent to which political theory as a field has rejected a consequentialist case for democracy. See David Estlund, Democratic Authority: A Philosophical Framework 6 (2008).

(37.) In their volume on collective wisdom, Helene Landemore and Jon Elster prefer this term to "collective intelligence" for its broad connotations of diverse types of intelligence and because it "evokes a larger temporal horizon" of a collective intelligence "extending not just through space (including many people) but through time as well (including may generations), thus making room for both experience and memory...." Helene Landemore, Collective Wisdom: Old and New, in Collective Wisdom: Principles and Mechanisms 7 (Helene Landemore & John Elster eds., 2012).

(38.) Id at 5, 10.

(39.) "Best" in this context could mean wisest, most correct in terms of the facts, or even morally right. An epistemic conception presumes only that there is some process-independent good, better, or best answer to a given question, which is to say that the correct answer is not simply the one that the majority of people prefer. This assumption is a necessary feature of any epistemic theory of democracy or free speech. As Landemore makes clear, this idea does not preclude a "culturalist" standard of correctness that is not absolute but simply recognizes that there is some process-independent measure of fact, value, and truth that is "right" for a given society at a given time. As she explains, "[w]herever the basic set of public values comes from (history, moral 'facts,' intuitions, emotions, reason, or intersubjective agreement), what matters is that one may speak of a correct answer at the level of the group." See landemore, democratic Reason, supra note 35, at 218.

(40.) This notion of cognitive diversity is based on the work of Lu Hong and Scott Page. See Lu Hong & Scott E. Page, Groups of Diverse Problem Solvers Can Outperform Groups of High-Ability Problem Solvers, 101 PNAS 16385, 16386 (2004) [hereinafter Hong & Page, Groups].

(41.) The other strand of the argument relies on the benefits of aggregative "crowd wisdom" even absent deliberation, as in the famous natural experiment by Sir Francis Galton in which the average of the predictions of several hundred visitors to a county fair, guessing at the weight of a butchered and dressed ox, came within one percent of the correct answer. See Francis Galton, Vox Populi (The Wisdom of Crowds), 75 Nature 450, 450 (1907); see also James Surowiecki, The Wisdom of Crowds: Why the Many Are Smarter Than the Few xi-xv (2004). This non-deliberative component is tangential to the Argumentative Theory of Reasoning and its relationship to freedom of speech; thus, its potential implications for the law of free speech are beyond the scope of this Article.

(42.) Surowiecki, supra note 41, at 83.

(43.) I don't mean to claim that truth is the only, or even necessarily the most important, value supporting the constitutional protection of freedom of speech. Rather, I advance the more modest claim that truth has a space in free speech theory and doctrine that is much more extensive than most current scholarship supposes.

(44.) See Bambauer, supra note 20, at 654-73 (providing a detailed summary of the influence of the marketplace metaphor in First Amendment opinions); Joo, The Worst Test of Truth, supra note 32, at 383.

(45.) Dan Sperber first proposed a version of the Argumentative Theory in 2001. It has been developed over the past several years primarily by Sperber and Hugo Mercier, with contributions by other coauthors as well. See generally Hugo Mercier & Dan Sperber, The Enigma of Reason (2017) [hereinafter Mercier & Sperber. Enigma of Reason]; see also Dan Sperber, An Evolutionary Perspective on Testimony and Argumentation, 29 Phil. Topics 401 (2001); Mercier & Sperber, Argumentative Theory, supra note 23; Dan Sperber et al., Epistemic Vigilance, 25 Mind & language 359 (2010); Mercier, Clarifications, supra note 5; Hugo Mercier, Using Evolutionary Thinking to Cut Across Disciplines: The Example of the Argumentative Theory of Reasoning, in comparative decision Making (t. Zentall & p. Crowley, eds., 2013); Hugo Mercier & Dan Sperber, Intuitive and Reflective Inferences, in In Two Minds: Dual Processes and Beyond (Keith Frankish & Jonathan St. B T. Evans eds., 2009) [hereinafter Mercier & Sperber, Intuitive and Reflective Inferences]; Hugo Mercier, On the Universality of Argumentative Reasoning, 11 J. of Cognition & Culture 85 (2011); Hugo Mercier & Helene Landemore, Reasoning is for Arguing: Understanding the Successes and Failures of Deliberation, 33 J. Pol. Psychol. 243, 249-51 (2012) [hereinafter Mercier & Landemore, Reasoning is for Arguing].

(46.) One legal scholar has already considered the more general implications of the ATR in the context of legal argumentation and decision-making, offering lessons for judges, moot court participants, and lawyers. Timothy P. O'Neill, Law and "The Argumentative Theory, " 90 Or. L. Rev 837, 837-43 (2012) (containing an excellent summary of the Argumentative Theory and of some of the cognitive biases it seeks to explain).

(47.) The work of democratic theorist Helene Landemore, including her collaboration with Hugo Mercier, has been instrumental in my thinking about the epistemic properties of deliberation as it relates to democratic decision making. See, e.g., Mercier & Landemore, Reasoning is for Arguing, supra note 45, at 249-51.

(48.) Lidsky, supra note 20, at 802 n.8.

(49.) See, e.g., Blasi, supra note 21; Eugene Volokh, In Defense of the Marketplace of Ideas/Search for Truth as a Theory of Free Speech Protection, 97 Va. L. Rev. 595 (2011); James Weinstein, Participatory Democracy as the Central Value of American Free Speech Doctrine, 97 VA. L. Rev. 491, 502 (2011) [hereinafter Weinstein, Participatory Democracy]; Robert Post, Participatory Democracy and Free Speech, 97 VA. L. Rev. 477, 487 (2011) [hereinafter Post, Participatory Democracy]; Gregory Brazeal, How Much Does a Belief Cost?: Revisiting the Marketplace of Ideas, 21 S. CAL. Interdisc. L. J. 1 (2011); Schauer, Facts, supra note 30, at 899.

(50.) The classic treatment is Stanley Ingber, The Marketplace of Ideas: A Legitimizing Myth, 1984 Duke L. J. 1, 6 (1984) (arguing that "the present marketplace [of ideas] simply fine-tunes differences among elites while defusing pressure for change and fostering a myth of personal autonomy essential to the continued popular acceptance of a governing system biased toward the status quo"). Professor Ingber anticipated many of the recent behavioral arguments about irrationality, but he rejected as "unworkable, dangerous, and inconsistent with the articulated purpose of the first amendment" the idea that the solution, as in economic markets, was government regulation aimed at correcting these market flaws.

(51.) A Westlaw search for "marketplace /3 ideas" in the Supreme Court database generates seventy-two separate cases, some with multiple mentions within opinions or across opinions, concurrences, and dissents. And this certainly undercounts instances of marketplace rhetoric, since the source of the theory, Justice Holmes' Abrams dissent, itself would not have been captured by this search.

(52.) A Google search for "marketplace of ideas" returns approximately 6,780,000 results. On American free speech exceptionalism, see Robert C. Post, Community and the First Amendment, 29 Ariz. St. L. J. 473, 483 (1997) [hereinafter Post, Community] ("No other country allows such a breadth of defamatory, indecent, abusive, and outrageous utterances.").

(53.) See, e.g., Schauer, Facts, supra note 30, at 909-10.

(54.) See Weinstein, Participatory Democracy, supra note 49, at 491; Post, Community, supra note 52, at 482.

(55.) C. Edwin Baker, Scope of the First Amendment Freedom of Speech, 25 UCLA L. Rev. 964, 968 (1978); see also Ingber, supra note 50, at 2* n.2 (noting that "[t]he marketplace of ideas permeates the Supreme Court's first amendment jurisprudence" and citing numerous examples). But see James Weinstein, Democracy, Sex and the First Amendment, 31 N.Y.U. Rev. L. & Soc. change 865, 875 n.46 (2007) [hereinafter Weinstein, Democracy, Sex and the First Amendment] (arguing that the marketplace of ideas "is not a core or even important free speech norm," and offering examples of Supreme Court doctrine that support this argument). Weinstein claims that the shape of the doctrine does not readily conform to marketplace of ideas values, which is not a claim that truth and marketplace rhetoric are insignificant in the Supreme Court's opinions.

(56.) Post, Democracy, Expertise, supra note 21, at x-xi; see also Joseph Blocher, Institutions in the Marketplace of Ideas, 57 Duke L. J. 821, 854 (2008) [hereinafter Blocher, Institutions]; Bambauer, supra note 20, at 658; Joo, The Worst Test of Truth, supra note 32, at 389. All of these authors argue that the metaphor exerts a strong force on current doctrine.

(57.) Joo notes the parallel timing of the rise of the marketplace of ideas metaphor and the rise of free market economics. He points out that "[t]he market metaphor has been particularly robust when free market ideology has been at its strongest: the early twentieth century, the Cold War Era, and today." Joo, The Worst Test of Truth, supra note 32, at 396.

(58.) Lochnerv. New York, 198 U.S. 45, 75 (1905) (Holmes, J., dissenting).

(59.) Abrams v. United States, 250 U.S. 616, 630 (1919) (Holmes, J., dissenting).

(60.) On an alternative reading of the famous passage in Justice Holmes's Abrams dissent, he may have had in mind the more pragmatic idea that truth is simply what most people believe, i.e. whatever prevails in a free marketplace of ideas. See Robert Post, Reconciling Theory and Doctrine in First Amendment Jurisprudence, 88 Cal. L. Rev. 2353, 2360 (2000) [hereinafter Post, Reconciling Theory and Doctrine]; Thomas Healy, The Great Dissent: How Oliver Wendell Holmes Changed His Mind--and Changed the History of Free Speech in America 218-19 (2013). However, such a tautological, proceduralist understanding of what counts as "truth" would not be an epistemic rationale for freedom of speech. As detailed in Part IV, an epistemic account requires some process-independent account of truth.

(61.) Frederick Schauer, Free Speech: A Philosophical Enquiry 16 (1982) [hereinafter Schauer, Free Speech],

(62.) John Stuart Mill acknowledged that false ideas might take hold for a period of time but argued that, over time, a liberal approach to free speech would be most likely to correct such errors. See John Stuart Mill, On Liberty 29 (Batoche Books 2001) (1859) ("The real advantage which truth has, consists in this, that when an opinion is true, it may be extinguished once, twice, or many times, but in the course of ages there will generally be found persons to rediscover it. Until some one of its reappearances falls on a time when from favourable circumstances it escapes persecution until it has made such head as to withstand all subsequent attempts to suppress it.") Of course, as John Maynard Keynes famously quipped, "[i]n the long run, we are all dead." See, e.g., Victor Sperandeo, The Long Run Is Here: Let's Bury John Maynard Keynes, Forbes (Mar. 30, 2012), http://www.forbes.eom/sites/victorsperandeo/2012/03/30/the-long-run-is-here -lets-bury-john-maynardkeynes/.

(63.) See Schauer, Free Speech, supra note 61, at 18-19 ("All 1 have said about the function of 'truth' in the argument from truth applies, mutatis mutandis, to any cognitivist theory of ethical or moral knowledge. Our use of descriptive language in ethical discourse presupposes that we can be wrong as well as right. And if being right is better than being wrong, then what I have said about factual knowledge applies to ethical knowledge, and applies as well to questions of social or political policy."); see also Landemore, democratic Reason, supra note 35, at 208-31 (laying out a defense of "political cognitivism," such that "it makes sense to presuppose and aim for a procedure-independent standard of correctness, whether we decide to call it truth, lightness or otherwise, when we engage in political deliberations"). Note that the question whether truth should necessarily take precedence over other values, such as privacy or reputation, is distinct.

(64.) See Schauer, Free Speech, supra note 61, at 15.

(65.) Blocher, Institutions, supra note 56, at, 824-25.

(66.) See, e.g., Joseph Blocher, Nonsense and the Freedom of Speech: What Meaning Means for the First Amendment, 63 Duke L. J. 1423, 1441-42 (2014) [hereinafter Nonsense] ("[T]he marketplace of ideas--the first and perhaps still most prominent First Amendment theory--rests on the notion that, if left unregulated, good ideas will eventually win out over bad ones.").

(67.) Abrams v. United States, 250 U.S. 616, 630 (1919) (Holmes, J., dissenting).

(68.) For an exception from 2008, see Blocher, Institutions, supra note 56, at 821 (applying ideas from the New Institutional Economics to First Amendment analysis of speech markets to "rehabilitate the marketplace of ideas"). For a less recent example of a strong scholarly defense of the marketplace rationale, see Christopher T. Wonnell, Truth and the Marketplace of Ideas, 19 U.C. Davis L. Rev. 669 (1986). Others have also pointed out that the truth-finding, marketplace rationale for protecting freedom of speech has become much less prominent in the scholarly literature. See Paul Horwitz, The First Amendment's Epistemological Problem, 87 Wash. L. Rev. 445, 453 (2012) [hereinafter Horwitz, Epistemologica! Problems]; see also Schauer, Facts, supra note 30 , at 909-10.

(69.) Paul Horwitz makes the interesting point that "[t]here is a kind of hydraulic dynamic at work here. As the truth-seeking justification for the protection of false statements has receded, other justifications for protecting them, such as those focused on democratic self-governance or distrust of government, have flowed in to take its place. Our theories have changed, but the consensus on which speech should be protected remains roughly the same." Horwitz, Epistemological Problems, supra note 68, at 469-70. Though Horwitz was speaking specifically about the First Amendment protection of false statements of fact, his insight also applies to other areas of free speech doctrine. This is a curious feature of free speech scholarship: there is often an explicit acknowledgment that intuition and/or normative opinion about results does or even should drive theory and doctrine, rather than the other way around. See, e.g., Seana Valentine Shiffrin, A Thinker-Based Approach to Freedom of Speech, 27 Const. Comment. 283, 285 (2011) ("[A] decent regime of freedom of speech must provide a principled and strong form of protection for political speech and, in particular, for incendiary speech and other forms of dissent, for religious speech, for fiction, art--whether abstract or representational and music, for diaries and other forms of discourse meant primarily for self-consumption, and for that private speech and discourse, e.g. personal conversations and letters, crucial to developing, pursuing, and maintaining personal relationships.").

(70.) It is no small irony that while scholars treat the marketplace of ideas model "as a relic that has not survived exposure to modern science," in the public mass market people are extremely bullish on the idea. See Schauer. Facts, supra note 30, at 909-10.

(71.) Weinstein, Democracy, Sex and the First Amendment, supra note 55, at 875 n.46.

(72.) See, e.g., David A. Strauss, Persuasion, Autonomy, and Freedom of Expression, 91 Colum. L. Rev. 334, 334 (1991). Also note the feminist critique of the traditional truth theory: Susan H. Williams, Truth, Autonomy, and Speech: Feminist Theory and the First Amendment (2004). See also Susan H. Williams, Feminist Theory and Freedom of Speech, 84 Ind. L. J. 999. 1001-04 (2009).

(73.) Schauer notes that "the free speech literature appears increasingly to have detached itself from the empirical and instrumental epistemic arguments, either by relying increasingly on arguments from democratic decision making or deliberation or participation, or on arguments from individual autonomy and self-expression, or on some number of other less-prominent free speech rationales." Schauer, Facts, supra note 30, at 909-10; see also Horwitz, EpistemologicaI Problems, supra note 68, at 453 ("It is worth noting that the truth-seeking justification, and its accompanying marketplace of ideas metaphor, have become far less prevalent in contemporary free speech scholarship.").

(74.) See Jack M. Balkin, Digital Speech and Democratic Culture: A Theory of Freedom of Expression for the Information Society, 79 N.Y.U. L. Rev. 1 (2004).

(75.) Id. at 3.

(76.) See schauer, free Speech, supra note 61, at 18-19.

(77.) Schauer, Facts, supra note 30, at 911.

(78.) Lidsky, supra note 20, at 802 (2010).

(79.) Id. at 804.

(80.) As explained in infra Part IV, however, the new models of collective wisdom do not rest on a rational actor model and therefore challenge this assumption.

(81.) Id. at 804. It is not clear to me why a flawed ideal of citizen rationality would undermine the case for self-governance under the prevailing process-based justifications for democracy.

(82.) Id. at 835.

(83.) See, e.g., Volokh, supra note 49, at 595.

(84.) See Post, Participatory Democracy, supra note 49, at 489; Weinstein, Participatory Democracy, supra note 49, at 497.

(85.) Volokh, supra note 49, at 597.

(86.) See, e.g., Strauss, supra note 72, at 348-49 (arguing that, unlike with economic market theory, "the real problem with the market metaphor" is that there is no real theory about the mechanisms by which speech markets might lead to desirable outcomes); Baker, supra note 55, at 96566; C. Edwin Baker, Ideology of the Economic Analysis of Law, 5 PHIL. & PUB. Aff. 3, 4 (1975); Joo, The Worst Test of Truth, supra note 32, at 409-14; Robert Sparrow & Robert E. Goodin, The Competition of Ideas: Market or Garden?, 4 Crit. Rev. Int'l Soc. & Pol. Phil. 45, 45 (2001).

(87.) Joo, The Worst Test of Truth, supra note 32, at 388.

(88.) Interestingly, this argument was first advanced by Chicago-school free market economists, most notably Ronald Coase. See R. H. Coase, The Market for Goods and the Market for Ideas, 64 AM. Econ. Rev. 384, 386-89 (1974). Coase argued that it was hypocritical and irrational to favor strong regulation of economic markets while at the same time rejecting regulation of speech. Coase asserted that the two markets demanded a consistent approach; he argued that "[t]here is no fundamental difference between these two markets and, in deciding on public policy with regard to them, we need to take into account the same considerations." And in view of the relevant considerations, he proposed that "the case for government intervention in the market for ideas is much stronger than it is, in general, in the market for goods." In a recent article, Thomas Joo points out a similar tension in later Supreme Court decisions, arguing that free speech liberals employed free market rhetoric in First Amendment decisions that they would not have accepted in an economic context. Joo, The Worst Test of Truth, supra note 32, at 399. See also R. H. Coase, Advertising and Free Speech, 6 J. Legal Stud. 1 (1977); Aaron Director, The Parity of the Economic Market Place, 7 J. L. & ECON. 1, 3 (1964).

(89.) See, e.g., Owen M. Fiss, Liberalism Divided: Freedom of Speech and the Many Uses of State Power, 9-20 (1996); Cass R. Sunstein, Democracy and the Problem of Freedom of Speech 17-51 (1993).

(90.) See, e.g., Fiss, supra note 89, at 9-20; Sunstein, supra note 89, at 17-51; Jerome A. Barron, Access to the Press--A New First Amendment Right, 80 Harv. L. Rev. 1641, 1641-2 (1967); Jerome A. Barron, Access--The Only Choice for the Media?, 48 Tex. L. Rev. 766, 782 ( 1970).

(91.) See Marianne Wesson, Girls Should Bring Lawsuits Everywhere ... Nothing Will Be Corrupted: Pornography as Speech and Product, 60 U. Chi. L. Rev. 845, 858 (1993) ("if the transaction that encompasses the creation, distribution, and consumption of pornography is one that creates a serious external harm, the logic of the marketplace dictates that the pornography industry should internalize the harm"); see also Frederick Schauer, Uncoupling Free Speech, 92 Colum. L. Rev. 1321, 1326 (1992) (framing protection for libelous speech as a public good and exploring various options for shifting the cost from victims to media or the public more generally).

(92.) See Bambauer, supra note 20, at 650. Whereas Professor Bambauer is explicit in his rejection of the marketplace of ideas model, as discussed further in this Part, many other scholars implicitly reject it by rejecting more generally the idea that unregulated speech will lead to epistemic benefits.

(93.) Bambauer, supra note 20, at 708-09.

(94.) See, e.g., Lidsky supra note 20, at 813-14.

(95.) See SCHAUER, FREE SPEECH, supra note 61, at 26. Perhaps this is an ironic example of the "everyone else is irrational except me" syndrome. In his early critique of the hypocrisy of liberals in supporting regulation of economic markets but not speech markets, Aaron Director made a similar point, suggesting that those intellectuals who embraced the free market as an ideal in the speech context while at the same time supporting government regulation in the market for goods were driven less by principle than by self-interest and elitism. They were self-interested because they rejected government regulation of the market in which they themselves tended to operate: the marketplace of ideas; they were elitist because they failed to recognize that "the bulk of mankind ... freedom of choice as owners of resources in choosing within available and continually changing opportunities, areas of employment, investment, and consumption is fully as important as freedom of discussion and participation in government." Director, supra note 88, at 6.

(96.) Schauer, Free Speech, supra note 61, at 26.

(97.) Post, Reconciling Theory and Doctrine, supra note 60, at 2365.

(98.) Id.; see also Volokh, supra note 49; Post, Democracy, Expertise, supra note 21, at 20; Post, Participator Democracy, supra note 49, at 478-79.

(99.) As I argue in Part VI, infra, this notion of the importance of crucial shared norms and practices is indeed a large part of the story of speech and epistemic benefits; where I disagree with Professor Post and others is about what these norms and practices should be. The interdisciplinary work discussed in this Article suggests that norms of deliberation, rather than rationality in its classical definition, are most important.

(100.) Horwitz, Epistemological Problems, supra note 68, at 445.

(101.) Schauer, Facts, supra note 30, at 897. Scorn for the popular tendency to believe false and ridiculous things is not new. See, e.g., Charles Mackay, Extraordinary Popular Delusions and the Madness of Crowds (1841) (describing "financial bubbles, religious crusades, witch hunts, the fashion of poisoning, and admiration for Robin Hood" as among the "popular follies" of the times); Landemore, Democratic Reason, supra note 35, at 28-29 n.5. Landemore adds to this list "Socrates' killing, [democracy's] tolerance of slavery both in ancient and modern times, and. arguably, its responsibility for bringing Hitler to power."

(102.) It may be worth noting that compiling such a list is probably an exercise in the availability heuristic (and possibly an illustration of confirmation bias and motivated reasoning). Certainly, there is a much longer list of true facts that many or most people in this country believe.

(103.) Schauer, Facts, supra note 30, at 908.

(104.) Note other theories such as Blasi's checking theory and Lee Bollinger's tolerance theory. Vincent Blasi, The Checking Value in First Amendment Theory, 3 Am. B. Found. Res. J. 521, 527 (1977); Lee Bollinger, The Tolerant Society 248 (1986).

(105.) Its proponents also argue that democratic participation represents the most accurate descriptive justification for the contours of current free speech doctrine. See Weinstein, Participatory Democracy, supra note 49, at 491 ("[descriptively, no other theory provides nearly as good an explanation of the actual pattern of the Supreme Court's free speech decisions"); Post, Participatory Democracy, supra note 49, at 482 ("[i]n my view, the best possible explanation of the shape of First Amendment doctrine is the value of democratic self-governance").

(106.) Along with participatory democratic and truth theories, the other major category of free speech justification is grounded in autonomy. See, e.g.. Post, Participatory Democracy, supra note 49. at 478 ("There are presently three major candidates for such [constitutional] values: (1) the creation of new knowledge; (2) individual autonomy; and (3) democratic self-government."). Of course, any such summary necessarily ignores a great deal of nuance and overlap within and among these three general categories. Free speech scholarship is broad and deep, with many theories that defy easy labels, and this article makes no claim to even mention them all. 1 focus here on participatory and deliberative democratic theory both because it has attracted a great deal of attention among scholars in recent years and because it provides the scaffolding for my construction of an epistemic deliberative theory of freedom of speech.

(107.) Of course this free speech theory has deep roots going back decades. See, e.g., Alexander Meiklejohn, Political Freedom (1960); William J. Brennan, The Supreme Court and the Meiklejohn Interpretation of the First Amendment, 79 Harv. L. Rev. 1, 11-12 (1965). 1 mean here only to suggest that it has, in recent years, risen above other theories and seems currently to enjoy "lexical priority" among scholars. See Robert Post. Recuperating First Amendment Doctrine, 47 Stan. L. Rev. 1249, 1272-73 (1995) (arguing that when the two primary rationales of truth and democracy lead to conflicting results in a particular case or context, democracy does--and should--enjoy lexical priority and thus trump).

(108.) Weinstein, Participatory Democracy, supra note 49, at 497.

(109.) Id.

(110.) Id. at 498.

(111.) Note, however, that Robert Post's conception of the constitutional value of "democratic competence" does consider the judicial protection of knowledge creation--mainly outside the public sphere--in the service of the "democratic legitimacy" advanced by content neutrality in the public sphere. See generally. Post, Democracy, Expertise, supra note 21, at 27-60. There is some confluence between Professor Post's idea of democratic competence and epistemic justifications of democracy, but there are also significant differences. Most importantly. Professor Post sees a strong tension between the value of democratic competence and that of democratic legitimation, arguing that courts should (and do) preference the former in certain professional and academic spheres and the latter in the public sphere.

(112.) See Dryzek, supra note 35, at 1 ("The final decade of the second millennium saw the theory of democracy take a strong deliberative turn. Increasingly, democratic legitimacy came to be seen in terms of the ability or opportunity to participate in effective deliberation on the part of those subject to collective decisions.").

(113.) Landemore, Democratic Reason, supra note 35, at 29. It is unlikely that this parallel is a coincidence; there is much cross-pollination between political theory and First Amendment theory. In fact, though, it may be more accurate to view this state of affairs among political theorists more as a continuation of a broad historical skepticism than as a recent trend. See also id. at 27-29 (noting the "longstanding prejudice against democracy, not only among political theorists and philosophers but, more surprisingly, among the people themselves, based on a general suspicion of people's capacity for self-rule").

(114.) Some have argued that even proceduralist deliberative democrats must necessarily accept at least a weak epistemic case in order for their position to be coherent. In a widely discussed passage, David Estlund reasoned that, if pure procedural fairness was the only criterion, then a coin flip should be just as acceptable as majoritarian democracy. That it is not, for him, revealed that "[t]here is something about democracy other than its fairness that contributes to our sense that it can justify authority and legal coercion." See Estlund, supra note 36, at 6; see also Jose Luis Marti, The Epistemic Conception of Deliberative Democracy Defended: Reasons, Rightness and Equal Political Autonomy, in Deliberative Democracy and its Discontents 27, 28 (Samantha Beeson & Jose Luis Marti eds., 2006). In any event, though, for most deliberative democrats the epistemic benefits of deliberation are either ignored or at least are secondary, if they are not explicitly rejected.

(115.) Landemore, Democratic Reason, supra note 35, at 31.

(116.) Id.

(117.) Joseph Schumpeter advanced a theory of a minimally participatory democracy in which leaders compete for votes as in an economic market, but then implement their own conception of the good rather than that of the larger citizenry. See generally Joseph Schumpeter, capitalism, Socialism, and Democracy (1942).

(118.) Landemore, Democratic Reason, supra note 35, at 32. The "non-democratic way" of solving the problem would be to resort to a decision-maker believed to be wiser, more rational, and/or more informed. This could be an "epistocracy," a benign autocrat, or some other replacement for the incompetent masses.

(119.) Where reasons can't be marshaled in support of a position, individuals often take the position that is publicly supportable, even if it is not correct. This is known as "reason-based choice." See, e.g., Eldar Shafir et al., Reason-Based Choice, 49 Cognition 11 (1993).

(120.) I use the image of bargaining here with some hesitation, because in democratic theory the classic definition of deliberation is often contrasted with bargaining behavior. "In that view, deliberation is not supposed to involve, in particular, threats, promises, sophistry, or any form of 'strategic' rather than 'communicative' action." Landemore, Democratic Reason, supra note 35, at 93. The bargaining I refer to here ideally is based on an exchange of arguments and reasons.

(121.) Roy F. Baumeister et al., Arguing. Reasoning, and the Interpersonal (Cultural) Functions of Human Consciousness, 42 Behav. & Brain SCI. 74, 74 (2011).

(122.) Sperber, supra note 45, at 408.

(123.) Sperber et al., supra note 45, at 378.

(124.) Id:, see also Malcolm Gladwell, Blink: the power of thinking without thinking 16 (2005).

(125.) Sperber et al., supra note 45, at 378; see also Julie A. Seaman, Form and (Disjunction in Sexual Harassment Law: Biology, Culture, and the Spandrels of Title VII, 37 Ariz. St. L.J. 321. 331. That is not to say that it cannot perform this function. Many adaptations are co-opted in the service of functions distinct from those that seem to have driven their evolution what in evolutionary biology is known as an "exaptation." In perhaps the most well-known example, noses are very useful for holding up glasses.

(126.) Mercier & Sperber. Argumentative Theory, supra note 23, at 58.

(127.) Evolutionary biologists look at species-typical traits and consider how and why those traits would have been adaptive for the animals' ancestors such that they came to be disproportionally selected and passed on through the process of Darwinian natural selection. The adaptive function of a particular trait--whether physical or behavioral--refers to the purpose that the trait evolved to serve in the organism. Of course, traits may also come to serve additional functions beyond those for which they originally were selected. See generally Seaman, supra note 125, at 340-47; see also Mercier & Sperber. Argumentative Theory, supra note 23, at 59 ("[A] function of a trait is an effect of that trait that causally explains its having evolved and persisted in a population: Thanks to this effect, the trait has been contributing to the fitness of organisms endowed with it.").

(128.) As Mercier and Sperber point out, if the trait benefited only speakers at the expense of audiences, or vice versa, complex human communication would not have evolved. Mercier & Sperber, Argumentative Theory, supra note 23, at 60. And, of course, speakers and listeners are the same individuals at different times and in different roles.

(129.) ATR has been discussed in major news sources and on numerous science blogs. See, e.g., Patricia Cohen, People Argue Just to Win, Scholars Assert, June 15, 2011, at C1; Sharon Begley, Why Evolution May Favor Irrationality, Newsweek (August 5, 2010), whyevolution-may-favor-irrationality-71535; John Horgan, Winning Argument: As a "New" Critique of Reason, Argumentative Theory Is Trite hut Useful. SCI. Am. (May 30, 2011), -critique-of-reason-argumentative-theory-is-trite-but-useful/; Patricia Cohen. Reason Seen More as Weapon Than Path to Truth. N.Y. Times (June 14, 2011),; Jonah Lehrer, The Reason We Reason, Wired (May 4, 2011). The science website, best known for its "annual question" series, summarized the excitement about the ATR as follows: "[Mercier and Sperber's] Argumentative Theory has already generated much excitement in the academic community.... The paper has created a storm of interest and controversy and has ... attracted attention well beyond academic circles.... In addition, many leading thinkers have taken note." See Mercier, Conversation, supra note 25.

(130.) Hugo Mercier & Dan Sperber, Reasoning as a Social Competence in Helene Landemore & Jon Elster, Collective Wisdom: Principles and Mechanisms, 381, 381 (Helene Landemore & Jon Elster eds., 2012).

(131.) Mercier (fe Sperber, Argumentative Theory, supra note 23, at 57.

(132.) Mercier and Sperber distinguish between intuitive and reflective inferences (and also intuitive and reflective conclusions). Mercier & Sperber, Intuitive and Reflective Inferences, supra note 45, at 156-57. They note that their idea of reflective inferences and conclusions is roughly analogous to System 2 in the sense that both "are characterised [sic] by control, effortfulness, explicitness and, (at least virtual) domain-generality." Though there are differences between their model and other dual process models, for purposes of this analysis it is probably not too misleading to think about "reasoning" in the ATR as roughly akin to a System 2 process. For a discussion of the disanalogies, see id. For one thing, though, Mercier and Sperber posit that some argumentation is intuitive: "We contend in particular that the arguments used in reasoning are the output of a mechanism of intuitive inference." Mercier & Sperber, Argumentative Theory, supra note 23, at 58.

(133.) The terms were originally used by Stanovich and West. See Keith E. Stanovich & Richard F. West, Individual Differences in Reasoning: Implications for the Rationality Debate?, 23 Behav. & Brain Sci. 645, 658 (2000). The terms have subsequently become widespread and standard in the field. See e.g., Daniel Kahneman, thinking Fast and Slow passim (2011).

(134.) Daniel Kahneman, A Perspective on Judgment and Choice, 58 Am. Psychol. 697, 698 (2003) (article based on its author's Nobel Prize lecture).

(135.) Mercier & Sperber, Intuitive and Reflective Inferences, supra note 45, at 148.

(136.) See Jonathan St. B.T. Evans, In Two Minds: Dual-Process Accounts of Reasoning, 7 Trends in Cogn. Sci. 454, 454 (2003). Note that System 1 is not typically viewed as a single system but as a complex set of cognitive subsystems that allow us to quickly process information about our environment using emotion, instincts, and previous learning. See Kahneman, supra note 133, at 22-26.

(137.) See Kahneman, supra note 134, at 698. Note, too. that practice and expertise may convert what for most people is a System 2 process into a System 1 process.

(138.) See id. (noting that practice and expertise may convert what for most people is a System 2 process into a System 1 process).

(139.) Dual process models have come to dominate many related but distinct fields. "There is a wide variety of evidence that has converged on the conclusion that some type of dual-process notion is needed in a diverse set of specialty areas not limited to: cognitive psychology, social psychology, neuropsychology, naturalistic philosophy, decision theory, and clinical psychology." Keith E. Stanovich, Rationality & the Reflective Mind 16-17 (2011). There are variations among these different dual process models, but they all make a basic distinction between what may be thought of as non-rational (System 1) and rational (System 2) processes. For a table comparing the various formulations, see id. at 18 (Table I.I).

(140.) Kahneman, supra note 134, at 699. Kahneman uses the following problem, drawn from a well-known experiment to illustrate the differences between System 1 and System 2 processes: If a bat and ball together cost $1.10, and the bat costs $1.00 more than the ball, how much does the ball cost? System 1 leads most people to intuit an answer of 10 cents, which is incorrect. System 2 is needed to recognize and "override" this error of intuition and to perform the analysis required to arrive at the correct solution of 5 cents ($1.05 + .05 = $1.10).

(141.) Id.

(142.) Id.

(143.) Id. at 710.

(144.) Stanovich, supra note 3, at 7-8. There is a great deal of debate within cognitive science about how to characterize these departures from the rational normative ideal. This disagreement is known as "The Great Rationality Debate." Id. As described by Keith Stanovich, the "so-called Meliorists ... assume that human reasoning is not as good as it could be, and that thinking could be improved." Id. Panglossians, in contrast, argue "that the research in the heuristics and biases tradition has not demonstrated human irrationality at all" because they reject the normative ideal of rationality advanced by. for example, classical economics. Id. at 345-46.

(145.) Mercier & Sperber. Argumentative Theory, supra note 23. at 58 (internal citations omitted).

(146.) Recently, some very interesting work has begun to focus on the minority of individuals who do not exhibit certain of these cognitive biases, or not to the same extent. There is some evidence that resistance to certain biases is correlated with general intelligence as well as some more specific cognitive and psychological qualities. See Jeffrey J. Rachlinski. Cognitive Errors, Individual Differences, and Paternalism, 73 U. Chi. L. Rev. 207, 216 (2006); stanovich, supra note 3, at 34758. See also Thomas Talhelm et al.. Liberals Think More Analytically (More "WEIRD") Than Conservatives, 41 Personality & Soc. Psych. Bull. 250, 250 (2015).

(147.) The question of the relevance of implicit biases to law has given rise to a large literature. For just a small sampling of this rich body of work, see, e.g., Jerry Kang, Trojan Horses of Race, 118 Harv. L. Rev. 1489, 1497 (2005); Christine Jolis & Cass R. Sunstein, The Law of Implicit Bias, 94 Cal. L. Rev. 969 (2006); Linda H. Krieger & Susan T. Fiske, Behavioral Realism in Employment Discrimination Law: Implicit Bias and Disparate Treatment, 94 Cal. L. Rev. 997 (2006). For a critique of this project, see Gregory Mitchell & Philip E. Tetlock, Antidiscrimination Law and the Perils of Mindreading, 67 Ohio St. L.J. 1023, 1030 (2006).

(148.) Mercier & Sperber. Argumentative Theory, supra note 23, at 72.

(149.) Id.

(150.) Id. at 57.

(151.) Hugo Mercier. When Experts Argue: Explaining the Best and the Worst of Reasoning, 25 Argumentation 313, 315 (2011) [hereinafter Mercier, When Experts Argue] (internal citations omitted) (referencing several articles by Mercier, Sperber, and other cross-disciplinary scholars in these various fields).

(152.) This argument form is referred to in formal logic as modus tollens. Mercier & Sperber, Argumentative Theory, supra note 23, at 61 (citing Jonathan St. B.T. Evans et al.. Human Reasoning: The Psychology of Deduction) (1993).

(153.) See Jonathan St. B.T. Evans, Logic and Human Reasoning: An Assessment of the Deduction Paradigm, 128 PSYCH. BULL. 978, 980 (2002). In the original version of this paradigm, "people were told that four cards each had a letter on one side and a number on the other and that the following rule applied to the cards: If there is a vowel on one side of the card then there is an even number on the other side of the card. The four cards displayed might have the values A (vowel), T (consonant), 4 (even number), and 7 (odd number) on their visible sides. The instruction was to choose only those cards that needed to be turned over in order to decide whether the rule was true or false." The correct answer "is typically given by 10% or less of university students participating in such experiments." Id.

(154.) Sperber and Girotto have summarized the almost magnetic appeal to researchers of the Wason selection task thus: "Why has Wason's selection task been, for almost forty years, so extensively used in psychology of reasoning? Because, it has a simple, logically compelling solution, and yet, in most versions, most participants fail to solve it." Dan Sperber & Vittorio Girotto, Use or Misuse of the Selection Task? Rejoinder to Fiddick, Cosmides, and Tooby, 85 cognition 277, 277 (2002) (internal citation omitted). I should note, however, that Girotto and Sperber do not discuss the selection task in the logical fallacy section of their paper, likely because Sperber in this and other papers has argued that the task is methodologically flawed and subject to manipulation, which in his view accounts in large part for its popularity in the literature. They do, however, discuss the selection task in their discussion of confirmation bias because some results evidence this distinct cognitive bias. I mention the Wason selection task here, though, because it is a stalwart in discussions of the defects in human logical reasoning ability. E.g., Evans, supra note 153, at 980 ("The selection task has become probably the most investigated single problem in the whole literature on the psychology of reasoning.").

(155.) See Evans, supra note 153, at 980; Stanovich, supra note 3, at 5. The challenge is to the assumption on the part of researchers that rationality is defined by logic, and therefore that logical errors are equivalent to irrationality.

(156.) See. e.g., Leda Cosmides & John Tooby, Can a General Deontic Logic Capture the Facts of Human Moral Reasoning? How the Mind Interprets Social Exchange Rules and Detects Cheaters, 1 Moral psychology 53, 64 (Walter Sinnott-Armstrong, ed. 2008). Evans, supra note 153, reviews this literature.

(157.) Mercier & Sperber, Argumentative Theory, supra note 23, at 61 (citing Nancy Pennington & Reid Hastie, Reasoning in Explanation-Based Decision-Making, 49 cognition 123 (1993)).

(158.) Id.

(159.) Id. at 61-62.

(160.) Id.

(161.) Id. For example, some researchers had found that their subjects in simulated debate scenarios failed to anticipate and respond to counterarguments. Mercer and Sperber point out that in these studies there was no back-and-forth between the two sides and therefore the participants had no occasion to address counterarguments but on the contrary were behaving appropriately if their goal was to convince others and win the argument. They point out that in other studies, in which participants were required to challenge others' positions, participants were in fact very adept at finding counterarguments against a claim.

(162.) Id. at 61 (citing literature review and multiple sources).

(163.) Id. See also Sperber et al., supra note 45, at 35-36.

(164.) Mercier & Sperber, Argumentative Theory, supra note 23, at 61.

(165.) Nickerson, supra note 10, at 177 (1998).

(166.) Id. at 175.

(167.) See id. at 202; see also Daniel B. Klein, I Was Wrong, and So Are You, The Atlantic (Dec. 2011), -and-so-areyou/308713/. This article is libertarian economist Klein's mea culpa for an earlier and widely-read Wall Street Journal op-ed in which he reported his experimental findings that progressives and liberals were not as smart as conservatives and libertarians when it came to understanding economic issues. As it turned out, when the study questions were re-written to reflect progressive values (though the logical problems were similar), conservatives performed poorly and progressives did much better. His conclusion: "The proper inference from our work is not that one group is more enlightened, or less. It's that 'myside bias'--the tendency to judge a statement according to how conveniently it fits with one's settled position--is pervasive among all of America's political groups. The bias is seen in the data, and in my actions." On the subject of politics and logic, see also Talhelm et al., supra note 146, at 252.

(168.) In terms of adaptive benefits and the evolutionary story, the explanation of the epistemic benefits of reason under the ATR view seems to flirt with group selection theory, which is controversial. See, e.g., David Sloan Wilson, A Theory of Group Selection, 72 PNAS 143, 143 (1975); David Sloan Wilson, The Group Selection Controversy: History and Current Status, 14 Ann. Rev. ECOL. SYST. 159, 182-85 (1983); Steven Pinker, The False Allure of Group Selection, Edge, Hugo Mercier disagrees that the ATR depends on any model of group selection. Skype Conversation with Hugo Mercier, (Nov. 2012); Comments from Hugo Mercier, (Nov. 2012) (on file with author).

(169.) Mercier & Sperber, Argumentative Theory, supra note 23, at 65.

(170.) See Mercier & Sperber, Enigma of Reason, supra note 45, at 211-21. This description of the ideal argumentative context likely to lead to discovery of truth happens to closely track the American adversarial model of litigation.

(171.) For a comprehensive recent summary, see Daniel C. Molden & E. Tory Higgins, Motivated Thinking, in The Cambridge Handbook of Thinking and Reasoning 295, 295-96 (Keith J. Holyoak & Robert G. Morrison eds., 2005).

(172.) David p. Redlawsk. A Matter of Motivated "Reasoning, " N.Y times, (Apr. 22. 2011, 3:55 PM), psychology-of-the-birther-myth/a-matter-of- motivated-reasoning; see Dan M. Kahan, Ideology, Motivated Reasoning, and Cognitive Reflection, 8 judgment and decision Making 407, 407 (2013); Z. Kunda, The Case for Motivated Political Reasoning, 108 psychol. bull. 480, 480 (1990); David P. Redlawsk et al.. The Affective Tipping Point: Do Motivated Reasoners Ever "Get It?, " 31 pol. psychol. 563, 564 (2010). The idea of motivated reasoning is closely akin to cognitive dissonance theory, as Redlawsk notes. See generally Julie A. Seaman, Cognitive Dissonance in the Classroom: Rationale and Rationalization in the Law of Evidence, 50 St. Louis L. J. 1097, 1109-12 (2006) [hereinafter Seaman, Cognitive Dissonance] (summarizing research on cognitive dissonance).

(173.) See Drew Westen et al., Neural Bases of Motivated Reasoning: An FMRI Study of Emotional Constraints on Partisan Political Judgment in the 2004 U.S. Presidental Election, 18 J. Cogn. Neurosci. 1947, 1955 (2006).

(174.) Molden & Higgins, supra note 171, at 298.

(175.) See Drew Westen, RATS, We Should Have Used Clinton: Subliminal Priming in Political Campaigns, 33 Pol. Psychol. 631, 632 (2008); see also The Cultural Cognition Project at Yale Law school, (last visited Oct. 14, 2016).

(176.) Molden & Higgins, supra note 171, at 303-04. People are less likely to be moved by "objective reality" (in Molden & Higgins' phrasing) in situations where that reality is less certain and more vague. Thus, the overall results of the experimental literature "suggest that thinking and reasoning inspired by directional outcomes does not so much lead people to ignore the sometimes disappointing reality they face as it inspires them to exploit the uncertainties that exist in this reality to their favor." Id. at 304. This suggests that a robust deliberative context that can reduce these uncertainties will tend to minimize the epistemic failures caused by motivated reasoning.

(177.) "If reasoning is designed for arguing, it can be expected to easily produce such arguments-reasoning should display a strong confirmation bias. And it does: the confirmation bias may be the most prevalent and robust bias ever evidenced by psychologists." Mercier, When Experts Argue, supra note 151, at 315.

(178.) Mercier, Clarifications, supra note 5, at 262.

(179.) See Sperber et al., supra note 45, at 360.

(180.) In a recent paper on the topic of expert reasoning, Mercier seems to distinguish the "audience" as listener from actors who are in active debate or conversation with the speaker. See Mercier, When Experts Argue, supra note 151, at 322. This makes sense because those actively engaged in the argumentative deliberation are most likely to invoke their own cognitive biases to counter the arguments of the speaker. More passive listeners are less likely to do so.

(181.) As Sperber et al. explain, "[w]hile the interests of others often overlap with our own, they rarely coincide with ours exactly. In a variety of situations, their interests are best served by misleading or deceiving us. It is because of the risk of deception that epistemic vigilance may be not merely advantageous but indispensable if communication itself is to remain advantageous." Sperber et al., supra note 45, at 360. And. as Sperber et al. point out, "interaction among epistemically vigilant agents is likely to generate not only psychological but also social vigilance mechanisms." Id. at 361. The evidentiary doctrines surrounding witness impeachment and credibility are an institutional instantiation of epistemic vigilance, as are peer review and other academic and scientific processes. Cf. Post, Participatory Democracy, supra note 49, at 478-79.

(182.) Sperber et al., supra note 45. at 359.

(183.) Id. at 369-70 (citing sources). A particularly troubling recent study found that defendants with less trustworthy faces were much more likely to be sentenced to death. This occurred even in cases where the defendants were later exonerated. In other words, innocent defendants who happened to have facial features associated with untrustworthiness are much more likely to receive a death sentence. See John Paul Wilson & Nicholas O. Rule, Facial Trustworthiness Predicts Extreme Criminal-Sentencing Outcomes, 10 Psych. Sci. 1, 5 (2015). There is much experimental (and real world) data showing that people are especially poor at detecting deception and, furthermore, that experts are no better than laypeople in doing so, though their confidence level is higher. See Julie A. Seaman, Black Boxes, 58 Emory L.J. 427, 437 n.36 (2008) (summarizing the research). While Sperber et al. acknowledge the questionable utility of such "split-second judgments of trustworthiness," they note that at the very least the literature shows "that looking for signs of trustworthiness is one of the first things we do when we see a new face." Sperber et al., supra note 45, at 370. Thus presumably providing strong support for the existence of vigilance as an important cognitive mechanism. In addition, there is "wide agreement over which faces look trustworthy and which do not," providing further support for this proposition. Wilson & Rule, supra, at 1.

(184.) This emphasis on personality over situational factors in predicting and interpreting behavior is generally referred to as the "fundamental attribution error." See, e.g., Philip E. Tetlock, Accountability: A Social Check on the Fundamental Attribution Error, 48 Soc. Psychol. Q. 227, 227 (1985). There is much evidence that situational factors have a greater influence on behavior than people generally suppose. See Seaman, Hate Speech, supra note 8, at 102-03.

(185.) Sperber et al., supra note 45, at 369-73. See also Nicholas M. Hobson et al.. When Novel Rituals Impact Intergroup Bias: Evidence from Economic Games and Neurophysiology, (Mar. 27, 2017), https://papers.ssrn. com/sol3/papers.cfm?abstract_id=2835548 (demonstrating that shared rituals--even those that are novel and meaningless--have a drastic effect on levels of trust among otherwise unrelated individuals).

(186.) It appears that Mercier and Sperber consider some of the mechanisms of epistemic vigilance to be distinct from "reasoning" as they define it, because for them reasoning is a very particular exercise in producing and evaluating arguments, i.e. reasons to accept a particular position. Thus, in their model (and according to their definition), "[r]easoning is used when other mechanisms of epistemic vigilance would have led to the rejection of some communicated information." Mercier, When Experts Argue, supra note 151, at 320; see also Hugo Mercier, Our Pigheaded Core: How We Became Smarter to Be Influenced by Other People, in cooperation and Its Evolution 373, 373 (Kim Sterelny et al. eds., 2013). In other passages, however, they refer to reasoning as "a tool for epistemic vigilance, and for communication with vigilant addressees." Sperber et al., supra note 45, at 378.

(187.) Sperber et al., supra note 45, at 374-76. In addition, the authors posit that listeners may monitor the coherence of the speaker's statements with other statements the same speaker has made, though this checking is probably more usefully seen as a species of source vigilance rather than content vigilance. See Mercier, When Experts Argue, supra note 151, at 322.

(188.) Sperber et al., supra note 45, at 376.

(189.) Id.

(190.) Id. According to cognitive dissonance theory, any incoherence that arises during this process will cause uneasiness, with the result that the person will attempt to decrease the dissonance by bringing the dissonant cognitions into consonance. See generally Seaman, Cognitive Dissonance, supra note 172, at 1 109-12.

(191.) This is my own term; the authors use the terms "dialogic" and "deliberative." See Mercier & Sperber, Argumentative Theory supra note 23, at 61; Wim De Neys, The Freak in AU of Us: Logical Truthseeking Without Argumentation. 34 Behav. & Brain Sci. 75, 76 (2011). However, "dialectic" seems quite descriptive of the reasoning process as they describe it because the actors arc employing theory of mind to imagine the desires, behaviors, and likely reactions of those with whom they are in dialogue, thus influencing their behavior in a recursive and dynamic way.

(192.) Mercier, Conversation, supra note 25.

(193.) Id. (indicating the type and degree of vigilance exercised will depend on factors such as the salience and relevance to the listener of the information, the relationship of the listener to the speaker, etc.).

(194.) As the authors point out, "from the communicator's point of view, a vigilant addressee is better than one who rejects her testimony outright," because at least there is the opportunity for persuasion if she is able to muster convincing arguments: "the addressee's reliance on coherence as a criterion for accepting or rejecting her claim may offer the communicator an opportunity to get past his defences and convince him after all." Sperber et al., supra note 45, at 376; see also Mercier, When Experts Argue, supra note 151, at 321-23.

(195.) Sperber et al., supra note 45, at 376. In addition, the speaker can take advantage of the listener's coherence checking by explicitly reminding listeners of beliefs they hold that are consistent with her position (or that are inconsistent with a contrary position).

(196.) The person might also, or instead, engage in an inner dialogue with these same features. There is some evidence that doing so improves epistemic performance, but it is probably not as effective as true deliberation in a social context. For a discussion of this question, see Helene Landemore & Hugo Mercier, Talking It Out With Others vs. Deliberation Within and the Law of Group Polarization: Some Implications of the Argumentative Theory of Reasoning for Deliberative Democracy, 205 Analise social 910, 920-25 (2012) [hereinafter Landemore & Mercier. Deliberations]; see also Robert E. Goodin, Reflective Democracy 169 (2003).

(197.) In the group decision-making literature, two individuals count as a group. Moreover, in Experts, Mercier distinguishes between the parties to a debate and the audience, i.e. between the "listener/addressee" as described in the text above and more diffuse listeners who do not actively participate in the discussion. See Mercier, When Experts Argue, supra note 151, at 322.

(198.) See Mercier & Landemore, Reasoning Is for Arguing, supra note 45, at 243- 44 (2012). They summarize the state of the literature thus: "[g]roup deliberation sometimes homogenizes attitudes and sometimes polarizes them. Taking part in discussions can increase or decrease engagement in political activity. Decisions made in groups will sometimes be better and sometimes worse than decisions made by individuals ... 'the general conclusion of surveys of the empirical research so far is that taken together the findings are mixed or inconclusive."' Id. (internal citations omitted) (quoting Dennis F. Thompson, Deliberative Democratic Theory and Empirical Political Science, 11 Ann. Rev. Pol. Sci. 497, 499 (2008)).

(199.) See generally SUROWIEC'KI, supra note 41.

(200.) See id. However, certain conditions arguably must be satisfied for these benefits to accrue. See Helene Landemore, Majority Rule and the Wisdom of Crowds: The Task- Specificity of Majority Rule as a Predictive Tool (2010), [hereinafter Landemore, Majority Rule] (outlining the theoretical conditions necessary under the Condorcet Jury Theorem, Mays Theorem, and other explanations for crowd wisdom). But see R. Scott Tindale, et al., Good and Bad Group Performance: Same Process--Different Outcomes, 15 Grp. Proc. & intergrp. Rels. 603, 604 (2012) (suggesting that the primary difference between good and poor group performance is the degree and kind of "social sharedness" among group members and its fit with the specifics of the decision under discussion).

(201.) Cass Sunstein has been a leading voice on this subject. See, e.g., Cass R. Sunstein, Deliberative Trouble? Why Groups Go to Extremes, 110 Yale L.J. 71, 107 (2000); Cass R. Sunstein, 2.0, 3 (2007).

(202.) Cass R. Sunstein, The Law of Group Polarization, 10 J. Polit. Phil. 175, 176-77 (2002).

(203.) Timur Kuran & Cass R. Sunstein, Availability Cascades and Risk Regulation, 51 Stan. L. Rev. 683, 685-91 (1999).

(204.) See Irving Janis, Groupthink (1982); Robert S. Baron, So Right It's Wrong: Groupthink and the Ubiquitous Nature of Polarized Group Decision-Making, 37 Advances in Exper. Soc. Psych. 219, 219-27(2005).

(205.) Mercier, When Experts Argue, supra note 151, at 320.

(206.) See Surowiecki, supra note 41, at xiv.

(207.) See Lyle Ungar et al., The Good Judgment Project: A Large Scale Test of Different Methods of Combining Expert Predictions. AAAI Technical Report FS-12-06 at 4 (finding that "[w]orking in groups greatly improves prediction accuracy"). Barbara A. Mellers et al., Improving Geopolitical Forecasting with Teamwork. Training, and Algorithms (paper on file with author).

(208.) Mercier & Sperber, Argumentative Theory, supra note 23, at 63.

(209.) As Landemore notes, this is less of a problem if the prior belief was correct, but in any case the group discussion will not have led to any epistemic improvement in the outcome. See Helene Landemore, Deliberation, Cognitive Diversity, and Democratic Inclusiveness: An Epistemic Argument for the Random Selection of Representatives. 190 Synthese 1209, 1201, 1227(2013).

(210.) In animal behavior, it is known as swarm intelligence; in various fields of psychology, cognition, and computer science it goes under the labels collective intelligence, distributed cognition, extended mind, and group mind. See, e.g., Dervis Karaboga & Bahriye Akay, A Survey: Algorithms Simulating Bee Swarm Intelligence, 31 Artificial Intel. Rev. 61. 61 (2009); Sandra Fernandes-Machado et al., Social Cognitions About Food Choice in Children Aged Five to Eight Years, 105 Appetite 144, 144 (2016); Epaminondas Kapetanios, Quo Vadis Computer Science: From Turing to Personal Computers, Personal Content and Collective Intelligence, 67 J. Data & Knowledge Engineering 286, 286 (2008).

(211.) Contemporary democratic theorists are sometimes categorized based on their aggregative versus deliberative focus--as Landemore labels them, the "counters" versus the "talkers." Landemore, Democratic Reason, supra note 35, at 88. The remainder of this section focuses on the epistemic position of deliberative democrats--the "talkers"--because of the centrality of democratic discourse theories as justifications for freedom of speech in contemporary First Amendment theory (and also, obviously, the centrality of speech to deliberation). Questions about the epistemic benefits of aggregation apart from deliberation are more tangential to free speech issues. Among the "counters," there is an active conversation about the epistemic value of majoritarian decision-making. See Robert Goodin & Christian List. Epistemic Democracy: Generalizing the Condorcet Jury Theorem, 9 J. Pol. Phil. 277, 277 (2001).

(212.) See, e.g., Estlund, supra note 36, at 6.

(213.) See Josiah Ober, democracy and knowledge: innovation and learning in classical Athens 2 (2010); Josiah Ober. Epistemic Democracy in Classical Athens: Sophistication. Diversity, and Innovation, in collective wisdom: principles and mechanisms 54, 54 (Helene Landemore & Jon Elster eds., 2012).

(214.) See Jose Luis Marti, The Epistemic Conception of Deliberative Democracy Defended: Reasons, Rightness, and Equal Political Autonomy, in Deliberative Democracy and Its Discontents 27, 28 (Samantha Besson et al. eds., 2006).

(215.) As summarized in a review of Estlund's most recent book, epistemic proceduralism rests on the following argument: "[governments exercise legitimate authority in virtue of making decisions according to procedures that, among those that can be justified in terms acceptable to all reasonable points of view, have some modest tendency to get the right answers. They are better than chance and better than qualified alternatives." Elizabeth Anderson, An Epistemic Defense of Democracy: David Estlund's Democratic Authority, 5 Episteme 129, 131 (2008) (reviewing David Estlund, Democratic authority (2008)). Epistemic Proceduralism is thus a hybrid epistemic and procedural theoretical justification for deliberative democracy.

(216.) See Helene Landemore, Why the Many Are Smarter than the Few and Why It Matters, 8 J. of Pub. Deliberation i, 2 (2012); Landemore & Mercier, Deliberations, supra note 196, at 913, 1209; Mercier & Landemore, Reasoning Is for Arguing: Understanding the Successes and Failures of Deliberation, 33 Pol. Psychol. 1467, 1467 (2012). I focus here on Landemore's defense of epistemic democracy in the deliberative context because of its particular relevance to freedom of speech. She also mounts a robust defense of the epistemic properties of aggregative or majoritarian democracy, see Landemore, Democratic Reason, supra note 35, at 145-206 (2013).

(217.) Other procedures could include decisions by experts, deliberation by a smaller group of people chosen for their superior intelligence, decision by one smart individual, etc.

(218.) See Lu Hong & Scott E. Page, Micro-Foundations of Collective Wisdom, in collective Wisdom: Principles and Mechanisms (Helene Landemore & Jon Elster eds., 2012); Hong & Page, Groups, supra note 40 at 16386; Scott E. Page. The Difference: How the Power of Diversity Creates Better Groups, Firms, Schools, and Societies (2007) [hereinafter. Page, The Difference]; Scott E. Page, Making the Difference: Applying a Logic of Diversity, academy of Mgmt. Perspectives, Nov. 2007, at 6.

(219.) E.g., Jane McGonigal. Why / Love Bees: A Case Study in Collective Intelligence Gaming, in The Ecology of Games: Connecting Youth, Games, and Learning 199 (Katie Salen, ed., MIT Press 2008); Jens Krause et al., Swarm Intelligence in Animals and Humans, 25 Trends in Ecology and Evolution 28. 29 (2010) (literature review). Krause et al. propose the following definition: "[w]e propose an overarching definition of SI [Swarm Intelligence] that covers phenomena observed in both animals and humans: two or more individuals independently, or at least partially independently, acquire information and these different packages of information are combined and processed through social interaction, which provides a solution to a cognitive problem in a way that cannot be implemented by isolated individuals."

(220.) See Krause et al., supra note 220, at 28.

(221.) See Surowiecki, supra note 41, at 83; Stefan Krause et al., Swarm Intelligence in Humans: Diversity Can Trump Ability, 81 Animal Behav. 941, 941 (2011).

(222.) See supra Part III (D).

(223.) One such condition is that the task the group is attempting to solve must be reasonably difficult. Simple problems do not benefit from group problem-solving. See page, The difference, supra note 219 (specifying as a condition of the "diversity trumps ability" theorem that "the problem must be difficult"). See also Takao Sasaki et al., Ant Colonies Outperform Individuals When a Sensory Discrimination Task is Difficult but not When it is Easy, 110 PNAS 13769, 13769 (2013); David Sloan Wilson et al., Cognitive Cooperation: When the Going Gets Tough, Think as a Group, 15 Hum. Nature 225, 231 (2004).

(224.) Landemore has argued that aggregative majority rule is well-suited to predictive tasks, whereas inclusive deliberation is most conducive to problem-solving tasks. See Landemore, Majority Rule, supra note 200.

(225.) Page, The Difference, supra note 219, at 48.

(226.) Interestingly, another condition that seems to improve group decision making is the presence of females in the group; furthermore, this correlation is linear, such that the more females that are in the group, the better the group's decisions. See Anita Williams Woolley et al.. Evidence for a Collective Intelligence Factor in the Performance of Human Groups, 330 Science 686, 688 (2010). Woolley and Malone attribute this result to the correlation between gender and social sensitivity. See Anita Woolley and Thomas Malone, Defend Your Research: What Makes a Team Smarter? More Women, Harv. Bus. Rev. 1, 1 (June 2011).

(227.) These conditions are "(1) the problem must be difficult; (2) the perspectives and heuristics that the problem solvers possess must be diverse; (3) the set of problem solvers from which we choose our collections ... must be large; and (4) the collections of problem solvers must not be too small." Page, The Difference, supra note 219, at 10.

(228.) Hong & Page, Groups, supra note 40, at 16386. See also Page, The Difference, supra note 219.

(229.) Hong & Page, Groups, supra note 40, at 16386.

(230.) See Landemore, Majority Rule, supra note 200, at 13.

(231.) Hong & Page, Groups, supra note 40, at 16385.

(232.) Id. Hong and Page note that the "challenge has been how to encourage people with diverse identities and backgrounds to work together productively." Id. (citing J.T. Polzer et al., 47 Admin. Sci. Q. 296 (2002)).

(233.) Id. at 16389.

(234.) As Scott Page puts it, diversity must be relevant to the particular problem: "we cannot expect that adding a poet to a medical research team would enable them to find a cure for the common cold." page, The Difference, supra note 219, at xxix. In addition, fascinating work by Philip Tetlock. Barb Meilers, and others at the Good Judgment Project demonstrates that while certain individuals are consistently better at predictions about particular issues than others, these are not necessarily those who we might have considered ex ante to be the "experts." The Good Judgment Project, (last visited Oct. 22, 2016).

(235.) See Post, Democracy, Expertise, supra note 21, at 6.

(236.) To be clear, I do not propose that these proceduralist values are absent from, or unimportant to, free speech and democracy theories and practices. I propose only that epistemic values should also be recognized and promoted.

(237.) I invoke here Frederick Schauer's very useful distinction between First Amendment "coverage" and First Amendment "protection." See Frederick Schauer, The Politics and Incentives of First Amendment Coverage, 56 Wm. & Mary L. Rev. 1613, 1617-24 (2013); Frederick Schauer, The Boundaries of the First Amendment: A Preliminary Exploration of Constitutional Salience, 117 Harv. L. Rev. 1765, 1769-74 (2004); Frederick Schauer, Categories and the First Amendment: A Play in Three Acts, 34 Vand. L. Rev. 265, 267-96 (1981); see also Post, Democracy, Expertise, supra note 21, at 1 ("Following Frederick Schauer, I distinguish between First Amendment 'coverage' and First Amendment 'protection.'").

(238.) New York Times Co. v. Sullivan. 376 U.S. 254, 270 (1964).

(239.) Id. at 271 (emphasis added).

(240.) Id. at 270.

(241.) Id. at 271.

(242.) See id. at 279-80.

(243.) See, e.g., Sullivan, 376 U.S. at 272.

(244.) Id.

(245.) Post, Democracy, Expertise, supra note 21, at 12; Weinstein, Participatory Democracy, supra note 49, at 491.

(246.) Landemore, Democratic Reason, supra note 35, at 91 (contrasting distributed deliberation with simple deliberation among individuals, and focusing on the latter as a cleaner type from which to develop her model of inclusive deliberative democracy).

(247.) Id. (providing a definition of "distributed deliberation" that is somewhat broader than that advanced by Robert Goodin). E.g., Robert E. Goodin, Sequencing Deliberative Moments, 40 Acta Politica 182 (2005).

(248.) Distributed cognition is a component of collective intelligence that makes groups smarter through division of cognitive labor. See Landemore, Democratic Reason, supra note 35, at 91.

(249.) 562 U.S. 443, 452 (2011).

(250.) Id. (quoting Dun & Bradstreet, Inc. v. Greenmoss Builders, Inc., 472 U.S. 749, 760 (1985)) (emphasis added).

(251.) Hill v. Colorado, 530 U.S. 703, 717 (2000) (citing Thornhill v. Alabama, 310 U.S. 88 (1940)) (emphasis added).

(252.) 134 S.Ct. 2518, 2535-36 (2014).

(253.) See id.

(254.) See Texas v. Johnson, 491 U.S. 397, 414 (1989).

(255.) See Cass Sunstein, Why Societies Need Dissent 22-23 (2003).

(256.) 551 U.S. 393, 448 (2007) (Stevens, J., dissenting).

(257.) Page's examples include the practice of enlisting outside committees to review academic departments and corporations hiring management consultants. As he notes, these outsiders are not smarter, but they bring different ways of thinking--they are cognitively diverse from the existing groups. See PAGE, THF. DIFFERENCE, supra note 219, at 342-44.

(258.) See Barry P. McDonald, The First Amendment and the Free Flow of Information: Towards a Realistic Right to Gather Information in the Information Age, 65 Ohio St. L.J. 249, 249 (2004).

(259.) See Philip E. Tetlock & Dan Gardner, Superforecasting: The Art and Science of Prediction 105-09(2015).

(260.) Cf. Lee Bollinger. The Tolerant Society 242-43 (1986).

Julie A. Seaman, Associate Professor, Emory Law School. JD Harvard Law School. This article benefited from excellent research assistance by Ian Longacre and Kevin Engleberg. I also received helpful feedback during colloquia and workshops at Emory Law School. The University of Texas Law School; the annual SEAL scholarship conference held at the University of Pennsylvania Law School; the NYU Business, Ethics & Evolution roundtable, and the SUNY Binghamton Evolutionary Studies seminar series. This Article has benefited from spirited discussion with many cognitively diverse people; in particular, Hugo Mercier was unfailingly generous with his time and comments. Among the many others who offered feedback, I'd especially like to thank Tom Arthur. Kay Levine, Michael Perry, Fred Schauer, Liza Vertinsky, and David Sloan Wilson. Any false ideas that remain are likely due to motivated reasoning in a context not conducive to good decision-making.
COPYRIGHT 2017 The Law & Psychology Review
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2017 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Seaman, Julie A.
Publication:Law and Psychology Review
Date:Jan 1, 2017
Previous Article:From Bob Jones to Obergefell: what the history of the gay rights movement means for the future of religious tax-exemption.
Next Article:Cognitive emotion and the law.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |