I have some thoughts about the terrific discussion, so far (through comment 7), of my Chapter 2. I can’t take everything up, but I hope this offers some clarification where people asked for it.
In addition to his excellent summary, Jonathan raises some good questions about my defense against the charge that I beg the question in favor of democracy. Let me review my argument briefly, and then respond to the challenges that have come up. Before reviewing the argument, it might help avoid confusion to remind us all of the distinction between authority (the power to require action), and legitimacy (permissibility of enforcement). The reason is that it’s not entirely clear which concept is mainly at stake here. This points to something that will probably come up later, that it’s not very clear what role, if any, the general acceptability requirement plays in my account of authority, as distinct from legitimacy. In the passages at issue here, I use the term “authority” often, without clearly acknowledging the difference between that and legitimacy. All I can say here is that, despite appearances, I think the general acceptability requirement also plays a role in questions of authority, though I will not try to say exactly how that works here. For present purposes, don’t worry too much about the legitimacy/authority distinction, and then we can see later whether bringing that distinction in cause further difficulties for what I say.
The argument in question is my response to an anticipated challenge, and so I’ll start there. I argue that invidious comparisons with respect to political wisdom are open to qualified disagreement, and that universal suffrage is the default in that context. But, someone might object that there could also be qualified disagreement about universal suffrage, holding that some form of epistocracy is epistemically superior. I grant that there might well be. If so, what entitles universal suffrage to the position of a default? If both arrangements fail the test of general acceptability (and since they are logically exhaustive (partial suffrage or universal suffrage, with dictatorship counting as very partial suffrage) no political coercion would be justified. I grant this last conditional point as well: if no political authority relation passes the general acceptability test then there is no political authority, but a political state of nature. So this leaves the question: what entitles universal suffrage to the position of a default?
The answer I offer is this, though I reformulate it here: the general acceptability requirement is a constraint on proposed authority relations. Here’s one kind of authority relation: each person has a say in majority rule, but each person is also under the authority of the majority decision (within limits, etc.). I think this kind of authority must pass the general acceptability test. That is, there must be some justification for that kind of authority that does not rely on any doctrines or claims that are open to qualified disagreement. I believe there is, and the book as a whole develops the idea of epistemic proceduralism, normative consent, and so on, in order to try to show this.
Here’s another kind of authority relation: certain specified people (specified either by name, definite description, or by membership in a defined class) are given authority over others on the basis of alleged epistemic superiority, but those others are not given symmetrical authority over them. This, I think, cannot be justified to all qualified points of view, for the reasons I give in my discussion of epistocracy. If this kind of authority can’t be justified, I say universal suffrage, which can be justified, is the default.
But this doesn’t fully answer the challenge. Once the issue is put in terms of qualified objections to claims of epistemic superiority we have to admit that some qualified points of view might believe in the epistemic superiority of epistocracy over democracy. Why, then, should democracy be the default? The reason I mean to be appealing to is this: the two forms of authority are not on a par. While I don’t take for granted, without argument, that democracy is justified, I do contend that the kind of authority epistocracy involves is of a greater magnitude of subjection to others. I will return to the question of how to support this, but first I’m trying to clarify the structure of the argument. We can think of it this way: there are three possibilities I’m considering. One is a state of nature, with no political authority/legitimacy. A second is democratic authority, with universal suffrage in some form. A third is non-democratic authority in which only some specified subset of people have authority, concentrating on arguments for epistocracy. My argument depends on the three kinds of proposed authority being antecedently ordered in the magnitude of authority they are proposing. The state of nature proposes the least, epistocracy proposes the most. And it depends on there being a presumption against higher magnitudes of authority. That, I trust, is not a democratic claim, and so it doesn’t beg the crucial question. My claim is that the case for democracy over the state of nature can be made to all qualified points of view (as I argue in the book as a whole), thus overcoming the presumption against more (or any) authority. The case for epistocracy over democracy cannot. This might seem to leave a tie between democracy and epistocracy, since there is qualified disagreement in both directions about which is epistemically better. But since epistocracy is a greater magnitude of authority, there is a presumption against it, and it loses. This is not a presumption in favor of democracy. It is a presumption in favor of less authority, and democracy benefits from this presumption. (I think Andrew’s way of glossing my argument in comment 3 is more or less the way I see it.)
So, how might we decide which kinds of authority relations are of a higher magnitude than others, warranting the presumption against them? (Here I go beyond the argument in the book, and I appreciate being pushed to do so.) In a certain sense, under universal suffrage no one’s formal authority is greater than anyone else’s. This is still authority, and it needs justification, but under partial suffrage, there is the additional thing: some people’s formal authority is greater than others. Of course, there are sometimes de facto “permanent majorities” even under universal suffrage. But any argument trying to establish those particular permanent majorities as authoritative would be seeking to establish an extra magnitude of authority, a greater magnitude than exists when they are merely de facto permanent majorities. (None of this denies that there can be reasons to worry about permanent majorities.) The same goes for epistocracy. It could happen that there is a de facto rule of the wise even under universal suffrage. The wise people could be in the majority. But an argument purporting to establish the rule of the wise (the very same people) would be trying to add an extra magnitude of authority, and faces a presumption against it. (My argument here would seem to cover the case, mentioned by Ben in comment 7, of a moderate epistocracy: more votes for some than for others.)
Boiling the point down, consider three of us deciding on a movie. We decide to vote, and agree the majority will be authoritative. You two end up voting against me. This is authority and needs justification (here it seems to rest on my consent, unlike in politics, but that’s beside the point). Now compare that to a proposal to give the two of you votes, but not me. This plainly involves an extra magnitude of authority over me. That’s the basis for my claim that epistocracy involves a greater magnitude of authority than universal suffrage. If that’s right, then there is a presumption in favor of universal suffrage over epistocracy. Then the fact that there’s qualified disagreement as to whether universal suffrage or epistocracy is epistemically better favors universal suffrage.
Paul (comment 4) suggests that I might have in mind an analogy between the individual and the collective case: in each case the default is self-rule. My own view is that there’s no moral significance to “collective self-rule.” The morally relevant unit is the individual. So my talk of defaults above is all about a presumption against more authority(/legitimacy) over the individual. The original default is none at all, etc. I don’t defend this moral individualism, but just assume it, so others might try to make something out of the idea of a moral immunity, held by a collective, against being ruled by other agents or collectives. I don’t mean to be denying, by the way, that it can be meaningful to speak of a collective’s actions or choices, etc. I just don’t think there’s any moral relevance to whether a collective is self-governing. I suppose it could turn out that such a thing is morally significant because of some specified relation between collective self-rule and the rights and liberties of individuals. I just don’t know how that can be worked out, so I pitch it all at the level of the moral claims of individuals.
I turn next to Jonathan’s worry about the boundaries of the group of “qualified” points of view. As he points out, if there were no qualified disagreement about who the experts are then my argument would imply that epistocracy would be justified and the case for democracy would fail. As I say in the book, I opt for the term “qualified” rather than Rawls’s term “reasonable” partly in order to discourage the temptation to fill in the content with the way the term is used in ordinary language. I hoped that the bland “qualified” didn’t suggest any particular content. Jonathan appeals to the idea of a “qualified political philosopher,” but this is just the kind of thing I’d like to avoid. I don’t mean the word to suggest anything much to do with the ways we’d apply the term outside of my theory. So it’s a term of art, almost as if I’d made up a whole new word, and that’s how I encourage readers to think of it.
So Jonathan’s question whether qualified points of view must be able to identify who the political experts are might look unanswerable in the absence of any content for “qualified.” That’s not quite true, since I give the term a sense, even though I give it no reference (to use the Fregean distinction in a basic way). To be qualified (look out, long sentence coming) is to be such that your views and convictions, even though they might be false, are not ones that disqualify you from only being permissibly coerced politically when that arrangement can be justified in terms that do not contradict your views. (This might make the qualified acceptability condition look like a tautology. To avoid that appearance, that principle itself can be put this way: certain views are such that, even though they might be false, their holders are not thereby subject to permissible coercion without a justification that is compatible with their views. Call those views “qualified.”) So, when we ask whether a certain view is qualified or not, we are asking whether or not holders of that view are disqualified from the normal burdens we have to justify our coercion in terms they can accept. This is still a long way from telling us which views are qualified in this sense and which are not, but we do now understand what the question is.
How does any of this address Jonathan’s challenge? Once the issue is specified in this way, the issue would be whether a person is only entitled to acceptable justification if they have the ability to identify the politically wise people, the potential epistocrats. Here I simply assert that this is not so. I don’t know where the burden of proof lies, but I see no reason to think that an inability to identify the wise would disqualify a person in this way. I offer no theory that would generate an answer here, so I simply assert it, and welcome substantive moral argument about it. It is no argument against me here simply to ask, “what if we counted people as qualified only if they could identify the wise?” The answer to that is obvious: my argument against epistocracy would fail. But why think it actually does? Another way of formulating my stance is to put it more weakly as a conditional: if, as I think, a failure to identify the wise is not disqualifying, we have a promising argument against epistocracy. (As you’ll see in the next chapter, I do think it’s quite plausible to say that the qualified must have an incorrect view about what views count as qualified. I put that assumption to heavy use. It is a very different issue, however, although they could easily be confused.)
I’m now anticipating issues from the next chapter, so let me leave that matter here and we’ll surely be coming back to it. Jonathan’s question about whether the existence of qualified disagreement is contingent or not should be addressed. Quickly, I think the question is the possibility, not the existence, of qualified disagreement. That would not be contingent. But more next week if people would like. On another point, I’ll briefly note that I think Ben (comment 1, his 3rd point) is right that my view does not suggest that all political decisions should be beyond reasonable disagreement, at least as we are likely to hear that. I’m not sure, though, that his “constitutional essentials” is right either, and here I might diverge from Rawls. Addressing both of those points, I’d say, rather, that any instance of political coercion must have a justification that is beyond reasonable disagreement. I apply it to “any instance,” not just special matters, but I say only that it “must have a justification” that is acceptable, though the decision itself could be open to qualified disagreement (as I assume many or most political decisions are).