The epistemology of inclusiveness
- 1.2k Downloads
In a strikingly pessimistic passage for a man that came to represent one of the most significant epistemological defenses of inclusiveness, Nicolas de Condorcet made “a rather important observation”:
A very numerous assembly cannot be composed of very enlightened men. It is even probable that those comprising this assembly will on many matters combine great ignorance with many prejudices. Thus there will be a great number of questions on which the probability of the truth of each voter will be below 1/2. It follows that the more numerous the assembly, the more it will be exposed to the risk of making false decisions (Condorcet 1976/1785, p. 49).
Setting aside the difficulty that this contention poses for the applicability of the theorem for which Condorcet came to be known, the idea that there is an inverse relationship between inclusiveness and accuracy is not altogether uncommon. The idea, in turn, fuels the intuition that, were it not for important moral and political reservations, the proper approach to all matters epistemic would be that of a minority of entrenched experts.
But all of this begs an important question: What is the relationship between inclusiveness and accuracy of judgment? Surely, it cannot be a strictly inverse one. Matters have to be more complicated than that. But what is the relationship, then? That is the central question of the present special issue, the purpose of which is to bring together essays from a variety of philosophical perspectives, unified by an attempt to better understand the epistemic virtues and vices of inclusiveness.
One natural starting point is to consider how inclusive we should be when encountering someone who disagrees with us. This is the topic of Sanford Goldberg’s contribution. Goldberg considers the epistemic implications of following a policy of inclusiveness in contexts of systematic disagreement, i.e., roughly, persistent disagreement among reasonable people on a broad and interconnected set of issues, the contours of which are well-known to the disagreeing parties. Goldberg acknowledges that there might be epistemic costs to a policy of inclusiveness in such contexts. For example, the policy might result in people giving up on beliefs that in fact constitute knowledge, or lowering what is in fact a fully rational credence. However, Goldberg argues, even such costs would not rationalize putting restrictions on inclusiveness. In so far as we have reason to impose such restrictions, the relevant considerations fall outside the purview of epistemology proper, in pertaining to costs framed in terms of time, effort, and other non-epistemic resources.
Leaving contexts of systematic disagreements aside, it seems reasonable to expect that taking into consideration the views of others will often be epistemically beneficial. This expectation is borne out in Hélène Landemore’s contribution. Landemore considers the epistemic implications of inclusiveness in contexts of deliberative assemblies of citizens getting together to solve problems. Building on Lu Hong and Scott Page’s “diversity trumps ability” theorem, she argues that, under certain conditions—more specifically, that the problem is difficult enough, and the deliberators sufficiently smart, different, and large in number—inclusiveness is a virtue, on account of a greater number of deliberators being better than a smaller number of deliberators. Moreover, Landemore argues that, in cases where there is an upper limit to the number of deliberators that can participate, the argument for more deliberators being better suggests that we should randomly select the participants of deliberative assemblies.
In his contribution to this issue, Christopher Thompson, too, is interested in the power of numbers and the benefits of inclusiveness. However, Thompson is concerned not with people deliberating but with people voting. Moreover, he points out that, prior to voting, people need to collect good information. The quote from Condorcet above highlights the fact that, for Condorcet’s Jury Theorem to imply that the probability of a correct verdict increases with group size, the voters need to be competent on the relevant issues. Assuming competency, Thompson argues that the same goes for groups searching for good information. The bigger the group, Thompson argues, the better. In that sense, inclusiveness is not merely a virtue in contexts of majority rule—again, assuming competency on the part of the voters—but also in the search for the information that is to inform such rule.
Goldberg, Landemore and Thompson (and many others with them) assume that, as far as epistemology is concerned, belief-forming processes are only good to the extent that they make for accurate judgments. Under that assumption, the epistemic value of processes is exhausted by their instrumental epistemic value. In her contribution to this issue, Fabienne Peter calls into question this assumption. She argues that, in contexts of disagreement, being prepared to adjust one’s beliefs in the face of a person with whom one disagrees is epistemically valuable independently of whether it renders one’s judgment on the matter more accurate. Exporting Stephen Darwall’s idea of a second-person standpoint in ethics into epistemology, Peter argues that adjusting one’s beliefs thus is of procedural, as opposed to merely instrumental, epistemic value, on account of the mutual accountability that characterizes properly conducted deliberation.
In some cases, many of us will for one reason or the other be excluded from the relevant decisions, be they arrived at on the basis of voting or deliberation. J. D. Trout’s contribution concerns one such case, namely that of policy-making. Given the technical nature of many policies as well as practical constraints on citizen input, we cannot always decide the relevant questions ourselves. Instead, some form of deference is called for. To ensure that our will is nevertheless represented in the relevant deliberations, however, we do not want to defer to just anyone, but rather to elected proxies. Trout considers a minimal epistemic condition that it seems reasonable that we would want such proxies to satisfy, namely that they be sensitive to the evidence. That might sound trivial, but as Trout goes on to show, it’s a condition that all too often fails to be satisfied by elected representatives. For this reason, Trout also puts forward a number of ameliorative recommendations for how to enable those representing our will in policy-making contexts to at the very least satisfy aforementioned minimal epistemic condition.
In some cases, the relevant decisions are neither in our own hands nor in the hands of our elected representatives, but instead in the hands of (non-elected) experts. What is the role of inclusiveness when it comes to expert knowledge? Boaz Miller considers this question in the context of experts having reached a consensus. A group of expert might arrive at a consensus for many different reasons, including shared biases. So under what conditions can we infer that an expert consensus is based on knowledge, as opposed to on a mere convergence of judgment? According to Boaz Miller, when the experts involved are committed to using the same evidential standards, the consensus is based on a varied but convergent set of evidence, and the experts are socially diverse, i.e., diverse with respect to social background and perspective. As such, the inclusiveness involved in opening up the scientific community to a wide set of experiences and perspectives is an important factor in explaining how a scientific consensus may be based on knowledge as opposed to on a mere agreement.
Questions about inclusion and exclusion raise questions about power, which brings us to Miranda Fricker’s contribution. Fricker considers not so much the value of inclusiveness as the vice of its opposite. More specifically, she argues that the kind of epistemic exclusion involved in people suffering an epistemic injustice—i.e., the kind of injustice involved when people are wronged in their capacity as knowers—renders people unable to contest decisions, and as such also compromises their freedom, on the republican notion of freedom as non-domination. In fact, Fricker goes further than this, in arguing that non-domination, properly understood, is not relevant only to republicans; to the contrary, non-domination is a condition on freedom to which even negative libertarians are committed. Consequently, epistemic justice is a constitutive condition of non-domination, which in turn is to be understood as a central political ideal of freedom.
If any conclusion can be drawn from the contributions to this issue, it is that there seems to be a strong and (fittingly enough) diverse case for the epistemic benefits of inclusiveness. Contrary to what we might take to be the implication of Condorcet’s observation quoted above—i.e., that the relationship between inclusiveness and accuracy is an inverse one—greater inclusiveness is often epistemically beneficial. Not in all cases, of course—as noted at the outset, matters are likely to be complicated, which is exactly why we need a more fine-grained conception of the variety of considerations that determine when and in what sense inclusiveness is epistemically beneficial. As guest editors, it has been a pleasure to read the sophisticated and thoughtful suggestions put forward by the contributors to this issue, and it is our hope that compiling them here will further the discussion of the many important issues in the epistemology of inclusiveness.
- Condorcet, M. J. A. N. C. (1976). Essay on the application of mathematics to the theory of decision-making. In K. M. Baker (Ed.), Condorcet: Selected writings. Indiana: Bobbs-Merrill. (originally published in 1785).Google Scholar