Against the belief that democracy may amount to “the rule of the idiots,” Democratic Reason: Politics, Collective Intelligence, and the Rule of the Many offers a welcome thesis: democracies are “smarter” than nondemocracies; they produce better outcomes. In this way, Landemore challenges understandings of democracy that assess democratic decisions only in terms of their procedural fairness and their justice-producing results. Following in the footsteps of theorists of epistemic democracy, she argues that democracies should be recognized for their “knowledge producing potential” (44). Specifically, the book focuses on two different forms of decision making: deliberation and aggregation. Arguably, democracy does epistemologically better than nondemocracies on those two fronts.

Landemore's work draws on a long-standing tradition of support for the “wisdom of crowds.” She does an excellent job of surveying the historical roots of the epistemic argument for democracy from Aristotle's remarks in favor of the intelligence of the many to Condorcet's celebrated “Jury Theorem” and Machiavelli's “Vox Populi, Vox Dei” dictum, among other examples.

Despite the great opportunity to make a case for democracy's legitimacy based on its epistemic benefits, Landemore sets a more circumscribed, yet valuable, goal: to explain which are the cognitive mechanisms responsible for producing smarter outcomes in a democracy. She says: “whether epistemic properties add to the legitimacy of democratic decisions in general or simply provide prudential reasons to abide by them is a question I will leave unaddressed” (47). In any case, this is a tall order, and Landemore's originality resides in providing an explanation of democracy's collective intelligence that is seemingly inconsistent with the empirical literature documenting apathy and lack of civic competency in voters. If many voters are indeed unqualified to vote, how come democracies are no less competent, and usually smarter, than nondemocracies (such as aristocracies and one-ruler systems)?

Landemore argues that the intelligence of the group does not equal the intelligence of its parts. Importantly, Landemore's book offers a masterful summary of the literature challenging claims that the regular voter is incompetent and misinformed. She also reviews critiques to the thesis that democracy is inconsistent because of its propensity to produce cyclical majorities. For these efforts only, the book is already highly valuable.

What are the specific cognitive mechanisms that make democratic intelligence possible? Landemore explains that there are two main mechanisms: cognitive diversity—“the difference in the way people will approach a problem or a question” (102)—and “distributed intelligence,” which refers to “an emergent phenomenon that can be traced not to individual minds but rather to the interaction between individual minds and between those minds and their environment” (19). Under the assumption that most democratic decisions can be likened to a problem-solving situation such as finding the way out of a maze, cognitive diversity turns out to be more crucial than individual ability because the larger the group deliberating, the better the chances for successful solutions. A small group of highly smart experts is less adept at finding fruitful solutions than a larger group of average citizens, assuming increasing the number of deliberators also increases the cognitive diversity in the group. The latter is an assumption that Landemore makes but does not substantiate fully. If democracies can be largely homogenous culturally and in terms of habit—as John Stuart Mill and Alexis de Tocqueville painfully regretted in their description of conformity-ridden mass societies—it is not clear that a small group of diverse enough experts will be worse than a large group of “average citizens,” who will not tend to offer different ways of seeing the world owing to their lack of originality, in Millian parlance.

In the same vein, Landemore appears to suggest that the epistemic argument for democracy applies both to direct and representative democracy. This thesis is partly supported by Josiah Ober's 2010 book documenting how ancient Athens was more efficient at pooling knowledge than oligarchic city-states. But democratic Athens contained elitist arrangements such as the Council of the Five Hundred, which Landemore mentions as an example of a knowledge-producing institution. The councilmen were experts by the standards of the period. This observation underpins a more general question: Is representative democracy equally competent as strictly pure democracy? Landemore is silent on this query. Could it be the case that representative democracy is smart because it gives “experts” some say, and it limits input from the masses in some respects?

In what follows, I would like to offer a few friendly observations that question the notion of cognitive diversity as a tool for democratic intelligence. My aim is to encourage the refinement of this long-awaited defense of epistemic democracy. Before I begin, it is imperative to understand what Landemore refers to as “the Argumentative Theory of Reasoning.” Landemore resorts to this new and ingenious psychological theory to “make sense of the successes and failures of empirically observed deliberations in a way that rescues the theoretical claim that deliberation has epistemic properties” (119). She employs this theory in order to respond to the criticism that deliberation polarizes participants and that it doesn't change people's minds because of notorious confirmation biases to which human beings are prone. On the classical theory of reasoning, deliberation should supposedly direct us to the truth, by way of “improving the correctness of our judgments and decisions” (124). On this classical approach, the empirical findings on the effectiveness of deliberation are profoundly dispiriting “because it seems that human reasoning is flawed and in need of correction” (125).

Landemore offers consolation. If empirical findings on deliberation show that people are reluctant to change their minds because they constantly and stubbornly seek to confirm their own beliefs as true, why not think of the function of argument in a different light? This is where “the argumentative theory of reasoning” enters the picture. Thus, it is claimed, reasoning's main function is to convince, not to seek truth. “As far as the production of arguments is concerned reasoning has and should have little concern for the pursuit of objective truth, since its main function is to derive support for beliefs already accepted as true” (126). In this way, “the argumentative approach is able to turn what seemed like vices into virtues. If the goal of reasoning is to convince others then the confirmation bias is actually useful since it leads to the identification of information and arguments for the side the individual already favors. Likewise, the fact that people are mostly good at falsifying statements that oppose their views is particularly useful if the goal of reasoning is to convince others” (126). The reason why people are so keen on confirming their own biases is evolutionary—so says the argumentative theory of reasoning. “Reasoning evolved to allow communication to proceed even where trust is limited: the production of arguments may convince people who would not accept others' claims on trust but who are able to evaluate the validity of an argument” (127).

The foregoing move is not helpful. Whatever the function of argumentation, the empirical evidence that deliberation leads to deadlock owing to people's confirmation biases remains intact. Maybe humans argue to convince each other—not to arrive at the truth—but deliberation does not become unproblematic because we have found another theory of why people argue. In a similar vein, there is empirical support for the idea that racial stereotyping is an evolutionary activity, even in well-intentioned egalitarian citizens. Racism could be traced to the fear of the different, which could have been necessary for protection against life-threatening dangers in the remote past. Does this evolutionary explanation remove the problem of racism? One would be hard-pressed to think so. Making sense of why people do what they do does not erase empirical difficulties resulting from collective action.

Landemore, however, adds that “the argumentative theory of reasoning does not imply that reasoning, because of its primarily argumentative function, has nothing to do with truth seeking. … Convincing others is more likely to work if the proposed arguments are sound and therefore have a connection to truth” (128). But this idea is inconsistent with the core of “the argumentative theory of reasoning”: If truth seeking is indeed an effective way of convincing others, then we are back to square one: the documented psychological biases that inhibit consensus in deliberation will ipso facto impair convincing. These empirical difficulties may explain, after all, why some democratic bodies end up making decisions not characterized by their smartness but by their toxic partisanship. So while the book does offer a pristinely organized and clear account of the epistemic benefits of aggregation, if aggregation is usually preceded by deliberation in most democracies, then those benefits may tend to be overshadowed by the biases that inhibit debate and rational discussion.

Using social-psychology tools that few scholars of democracy incorporate into their work, Democratic Reason represents a valuable interdisciplinary approach to understanding democracy. It invites us to expand our mental horizons in ways that are rarely seen in the world of normative political theory. Written with superb clarity and a masterful command of both the democratic-theory literature and the empirical literature on voting behavior, Landemore dares to think outside the box, honoring her very own concept of cognitive diversity.