Against Deference: Epistemic Privilege Considered

Following up on my claim in the last post that “honor produces error,” in my view, the problem is epistemic privilege, i.e. deference.

By deference, I mean the epistemic privileges that some people receive or earn through demonstrating their erudition or looking like they know what they’re talking about. When you think of deference, you should think of the feminist critique of uptake conventions among men and women. Consider: as a male I often notice that whether or not I know what I’m talking about, people tend to pay attention when I speak. With some of my equally smart (or smarter) colleagues, I see this reversed: they get talked over, get fewer turns to speak, and their expertise is challenged more often. Usually, the factors that go into these differences in uptake aren’t epistemically salient: things like gender, bearing, dress, physical fitness, or facial symmetry aren’t particularly accurate signals of knowledge. It’s not just unfair to my colleagues, but a bad way to get the right answer!

Epistemic Injustice

Miranda Fricker has argued in her work on epistemic injustice that it is possible to wrong another in their capacity as a knower. I agree! But Fricker goes on to argue that inappropriate lack of deference entails appropriate deference: if deference can be withheld unfairly, then that implies that there will be times when justice obligates us to defer to others. Yet here, I believe Fricker conflates testimonial deference (“You were there, I wasn’t, what did you see?”) with deference to expertise. (“You have studied this matter, and I haven’t, what do you know?”)

If deference can be appropriate, then epistemic privileges can be justified. This means that some forms of uptake and inattention are moral: sometimes, for Fricker, the right thing to do is to ignore one person and pay attention to another. Every act of deference requires an act of dismissal. When we deal with experts, this can even mean dismissing ourselves! The moment we grant epistemic privileges, we risk suspending our own critical faculties and ignoring other potential knowers in favor of the expert knower. When we grant doctors, lawyers, and academics deference in their respective fields, we do so without making independent inquiry into the matter. At best, we only hold ourselves responsible for evaluating the prestige or reputation of a particular putative expert. Whether in the case of disease, a lawsuit, or the interpretation of experimental data, when we bow to the professionals’ authority, we do so at the expense of epistemic equality. Is this justified?

Those who champion epistemic dependence generally do so in the name of epistemic interdependence. No one, we are told, can be a jack-of-all-trades and still function well within a complex society like our own. Breadth comes at the expense of depth. This produces a vision of egalitarian deference: so long as we defer to each other equally, as each of use takes up her areas of expertise in turn, our mutual deference need never produce a larger domination or inferiority.

The Promise (and Perils) of Expertise

Unfortunately, like capital or charisma, knowledge and expertise tend to pool and accumulate. Just as the accumulation of capital produces societal inequalities, so the accumulation of expertise and the attendant deference is also likely to produce error. Remember that testimony is often less accurate than we have a tendency to believe. In much the same way, expert opinions are granted more accuracy than they deserve. In fact, experts are prone to a kind of motivated skepticism that can render them less likely to make correct predictions in their areas of expertise than a simple coin flip. This is despite the fact that experts will have large amounts of correct information (and that they tend to underestimate their abilities while amateurs overestimate): when it comes to making predictions under conditions of uncertainty, experts can’t tell which of their views are correct and which biased.

To make sense of this, we must distinguish between a single expert and the consensus of experts in a field. As a group, experts are trustworthy. Alone, they are no better than a coin flip. The point here is that individual experts don’t deserve the epistemic privileges we grant them, even though expertise itself is useful and does deserve our attention. Deference is only due to a community of inquirers, not to persons: as an attitude towards individual persons, deference is both immoral (for producing epistemic injustice) and likely to lead to error. Even those who grant epistemic authority in principle will act on this suspicion of experts in practice. Instead of trusting a harried doctor to interpret our symptoms and the medical literature, we’ll frequently seek a second opinion or research her diagnosis online. When they feel unjustly convicted by the supposedly-arcane court system, some inmates spend their incarceration researching the legal precedents in an effort to file their own appeals, and in so doing, learn a good deal of law. When our lives or freedom are on the line, we’re able to see the problem with this epistemic dependence.

Using Authority to Deflate Privilege

Because our interactions are shot through with status games and signalling, perhaps epistemic and moral authority are inevitable. We might think that in the best we can hope for is to channel these forms of authority in egalitarian ways. For instance, a version of the deference dynamic sometimes also plays out in my classroom: although I try to set and maintain conventions of mutual respect, I must sometimes enforce them authoritatively. In doing so, I believe I put my epistemic privilege to an anti-privilege use, but thereby I also preserve it: the students respect each other because I remind them to do it, and they respect my reminders because they respect me. This is probably the way sentiments of mutual respect are generated, but I’m not sure it’s particularly moral.

Nor is my goal in demanding mutual respect in the classroom that each student’s opinion be treated as equally correct. Just the opposite: the hope is that after all opinions are weighed and most found wanting, some opinions will be preferred. That is to say, the goal of mutual respect in the classroom is that we will all defer to the best arguments rather than the loudest voices or most confident speakers. This is true even when the loud voice is my own. If the students simply took what I said as gospel, that would be as stultifying as if they believed their most charismatic comrades. So the norm of equality and non-deference is designed to produce a community of mutual inquiry, where the best argument wins.

Here, it’s the argument that gains authority, not the person. A teacher or leader with some kind special insight into the best arguments ought to lose her authority simply by failing to act or speak consistently with the principles and arguments she espouses. Yet we cannot revoke our deference unless we count ourselves as equal in the evaluation of the argument and the teacher’s mastery of it. Isn’t this the core of Enlightenment in which we “emerge from our self-imposed tutelage“?  The only legitimate epistemic authority is the world’s own authority to report its facts.

(Steven Maloney and I made some related arguments in our paper “Foresight, Epistemic Reliability, and the Systematic Underestimation of Risk.” There, we evaluated the problem of expertise under conditions of potential existential threats to the regime, and thus ducked the question of the *morality* of epistemic privileges.)

Beyond Utopophobia

The newest issue of The Good Society has been released, with a symposium my friend Steven Maloney and I put together on epistemic proceduralism. It features contributions by James Bohman, Corey Brettschneider, Noëlle McAfee, and Robert Talisse and Michael Harbour.  The ‘utopophobia’ in the title comes from David Estlund’s book Democratic Authority, which invokes epistemic grounds in defense of democratic legitimacy: because democratic procedures get things right more often then competing regimes, the decisions they make are legitimate. This puts Estlund in the company of contemporary epistemic liberals like Cheryl Misak and Robert Talisse.  Contrary to Rawlsian liberalism’s distaste for substantive epistemic and normative questions, the epistemic liberals suggest that we ought not to fear public deliberation and contestation about first principles, end-states, and matters of fundamental concern.

Steve and I also contributed a paper, “Foresight, Epistemic Reliability, and the Systematic Underestimation of Risk,” in which we struggle with democracy’s capacity to foresee ‘regime-busting’ cataclysms like war, famine, natural disasters, or economic shocks and either prevent them or prepare for them adequately. As the title suggests, we believe that democratic polities systematically underestimate such risks, at their own peril.

Check it out!