Is moral deference reasonably acceptable?

Abstract

Advocates of epistemic conceptions of democracy sometimes argue that democratic decision-making is a more reliable guide to getting the issues at stake right than the decision-making of individuals. Such arguments give rise to the question of whether those finding themselves in the minority should defer to democratic outcomes. In this article, I discuss the bearing of the normative criterion of reasonable acceptability on this question. I thus ask, can the demand to defer to democratic outcomes be rendered reasonably acceptable to citizens who disagree with them? Using the example of David Estlund’s epistemic proceduralism, I show that the problem of moral deference is inherent to epistemic conceptions of democracy. I then discuss the kind of reasons that would make a demand of deference reasonably acceptable to the minority, arguing that they should be independent from the reasons supporting the majority position. Subsequently, I introduce the conciliatory conception of democracy as an example of what such reasons could be. The conception’s central idea is that disagreement between citizens conceived as equal epistemic authorities carries epistemic significance. This significance makes it rational for citizens so conceived to conciliate their conflicting judgments. Because democratic decision-making has an inbuilt tendency to track such conciliatory positions, the conception offers an independent reason to accept democratic outcomes as the more reliable guide to getting the issues at stake right.

Keywords:

• Moral deference
• reasonable acceptability
• epistemology of disagreement
• conciliatory democracy
• Estlund

Notes

1. The respective texts are, among others, Habermas (1990, 1996), Rawls (1996), Gutman and Thompson (1996), Benhabib (1996), and Cohen (1989). See also Bohman and Richardson (2009).

2. For a critique of this evasive maneuver, see Estlund (1998).

3. See e.g. Rawls (1996). With Habermas, things are a bit trickier as he tries to eschew such a standard. I agree with Rehg and Bohman (1996), however, that his counterfactual ideal of a consensus under idealized conditions is just that and with Estlund (1997) that we do not get around allowing procedure-independent standards of correctness to enter a ‘cognitive’ conception of democracy.

4. Bohman (1998) traces the admission and incorporation of majoritarian decision-making in deliberative conceptions of democracy.

5. Estlund (2008, 80) defines full anonymity as complete blindness to personal features, including judgments of the relevant people.

6. This procedure consists in the random and thus fair selection of any citizen with the power to make the decision according to their view of what the best decision would be.

7. Estlund employs “qualified acceptability” instead of “reasonable acceptability” to avoid any potential confusion about the meaning of reasonableness. I revert to “reasonable acceptability” in this text, assuming that it is by now standard usage in contemporary analytical political philosophy and no source of confusion for those interested in this text in particular.

8. “Reasonable” entails inter alia the acceptance of conceptions of justice which are expressions of an egalitarian morality and which have broadly liberal-democratic implications. Estlund conceives the criterion as a necessary, not a sufficient condition of legitimacy.

9. A lot depends on how we understand “partly” here. To what part and how precisely does the authority depend on its epistemic value? Estlund leaves this question open and seems to rely on our intuitive understanding of the point.

10. A better-than-random procedure is one that is more likely than not to pick out the right answer to an either-or question, i.e. the probability of selecting the best option is greater than 0.5. This value decreases proportionally to an increase in options.

11. Notice that Estlund makes much of the better-than-random criterion in the chapters establishing the foundations of his conception of epistemic proceduralism in his (2008). However, he later claims that performance in non-core cases need not be better-than-random for the epistemic argument for democracy to get off the ground. All that is required is better-than-randomness in avoiding primary bads (Estlund 2008, 235). This is puzzling because the Science/Democracy analogy aims to build the case for democracy on an extrapolation to non-core cases.

12. Enoch (2014) discusses, and ultimately rejects, a number of them.

13. From a passage in his doctoral dissertation cited in Estlund (2008, 104).

14. More to the point, the identification of moral experts with a right to rule is reasonably contested because “No one is so obviously better at these things that there isn’t some qualified point of view that denies it” (Estlund 2008, 36).

15. Notice that the majority, while judging the matter in accordance with the procedure’s outcome anyway, could still face a demand of deference in the sense that they would have to accept that the fact that their judgment has been vindicated by the procedure is a better justification for it than their reasoning as individuals.

16. We could also call them impartial or neutral reasons. However, this invites confusion with the impartiality or neutrality of justifying reasons as e.g. political liberalism demands it. See e.g. Rawls (1996).

17. This is where Wollheim (1962) locates his famous paradox of democracy.

18. The legitimate enactment of the law gives you a peremptory reason for action if the reason excludes other reasons from deliberation about what ought to be done. H.L.A. Hart (1982) introduced the term in the essay “Commands and Authoritative Legal Reasons.” In an effort to shift the emphasis from the level of deliberation to the level of action, Joseph Raz uses the term “pre-emptive reasons” (1986, 38 ff.).

19. See Pollock and Cruz (1986) for the distinction between rebutting and undercutting defeaters. See also Feldman (2005, 104) and Christensen (2010, 193ff).

20. For some samples of the discussion see “Special Issue: The Epistemology of Disagreement” Episteme: A Journal of Social Epistemology 6.3 (2009, 231–353) and the essays collected in Feldman and Warfield (2010) and Christensen and Lackey (2013).

21. For a definition along these lines see Elga (2007). The two points narrowing down the notion are mentioned in Kelly (2005, 10). It is crucial that the definition given here spells out epistemic peerhood in terms of an ascription of epistemic status by others, not in terms of an objective possession of that status.

22. On this point, see in particular Feldman (2006) and his (2009).

23. The debate is sometimes framed in terms of an all-or-nothing model of belief – see e.g. Feldman (2006, 235) – but mostly operates with various degrees of credence invested in one’s beliefs – see Christensen (2007, 189) and Elga (2007). I will proceed on the assumption that the degree-of-belief model is applicable to peer disagreement. This opens the possibility of being rationally required to revise one’s credence in a belief while nevertheless maintaining it. The standard Bayesian convention has it that the credence one invests in a given belief is assigned a numerical value between 0 and 1 inclusive. 1 represents maximal confidence that the proposition is true, 0 represents maximal confidence that the proposition is false. An agnostic stance is represented by 0.5.

24. It is also sometimes called The Steadfast View – see e.g. Christensen (2009).

25. The Equal Weight View is sometimes labeled Conciliationism which emphasizes the fact that it advocates “splitting the difference” in credence in cases of peer disagreements. In these cases, according to the view under consideration, peers ought to conciliate. This label, however, is misleading given the fact that the view also prescribes an increase in credence in cases of peer agreement. If two peers have a credence of say 0.95 and 0.9 in a shared belief they should, after becoming aware of this fact, arrive a credence higher than their original credence. Popular defenses of the Equal Weight View were undertaken by Elga (2007), Christensen (2007) and (2009), and Feldman (2006) and (2009).

26. The arguments revolve around an epistemological principle called Independence which I believe can be defended against the challenges which epistemologists – e.g. Sosa (2010) and Kelly (2010) – have raised against it.

27. The rationality at stake here is not grounded in practical but in theoretical reasons.

28. See especially Black (1948) and McDonald and Budge (2005).

29. I would like to thank an anonymous reviewer for this journal for offering this example and the objection for consideration.

30. See fn 25 above.

31. Note also Rousseau’s remark that unanimity comes about “when the citizens, fallen into servitude, no longer have freedom of will” (SC, IV.2.3).

32. If the voter distribution is multi-modal, the model predicts that the conciliation of these conceptions occurs at the subsequent stage of coalition-formation whose centripetal forces drive political parties to accept positions which correspond to conciliatory positions along the ideological spectrum (Budge 2006, 424; Gallagher, Laver, and Mair 2005, 382).

33. Another of their functions is to lighten the epistemic burdens of citizens to arrive at informed judgments about what policies most cohere with their abstract value judgments. The fact that in modern democratic systems the burdens on citizens is thus reduced lends further credence to the claim that they can regard each other as equal epistemic authorities (Ware 1996, 320).

34. See also Ebeling and Wolkenstein forthcoming.

35. The effect is more pronounced in electoral systems characterized by proportional representation but is discernible also in single member district systems when taking long-term effects into account.

36. To be clear, the demand of moral deference as conciliatory democracy conceives it is not superior or more acceptable because it involves only partial deference, namely meeting halfway. In fact, the degrees of deference will differ according to which point on an ideological spectrum citizens occupy. This means that some will have to “move” more than others. The “weakness” of the deference claim only refers to the temporal limitedness of moral deference to a decision. In the stand-off between a strong and a weak demand of moral deference, plausibility is what matters, not acceptability. If the claim that certain procedures are nearly infallible were plausible and supported by the right kind of reasons, nothing would stand in the way of a demand of strong deference.