Last year my friend Craig Martin made an interesting post on the subject of normativity – what we might, for lack of a better word, call value judgements – in academic religious studies. I disagree with almost all of it, and I think it’s helpful to spell out the reasons for doing so.
Craig situates himself as a poststructuralist who does not accept “the dream of objectivity or objective truth”, but nevertheless deems it “both important and useful to appeal to intersubjective verification (of the sort we see in the work of the American pragmatists)…” The problem is that the “intersubjective verification” described in this post sounds, to my ears, almost exactly like the old-fashioned empiricism that poststructuralists are (rightly) supposed to be rejecting. As it is applied in this particular context, “intersubjective verification” seems to be little more than a fancy way of maintaining the empiricist’s fact-value distinction: “intersubjective verification” is something we can reach about empirically verifiable facts, but not about those silly insubstantial value judgements.
The basic problem with such an approach, for Craig as for the empiricists, is that such a standard of intersubjective verification is itself a value judgement of exactly the kind that it urges we avoid. The problem may be best captured by repeating one of the post’s last statements: “I think we should attempt to avoid using praiseworthy or pejorative evaluative terms, as well as ‘should’ statements about our objects of study.”
Notice something amiss here? “We should avoid ‘should’ statements” is a rather problematic claim! The specification Craig adds is “we should attempt to avoid… ‘should’ statements about our objects of study.” So one could theoretically defend this claim by saying it is fine to make “should” statements about ourselves as scholars and the ways we study, but not about our objects of study. But that is exactly the sort of distinction that poststructuralists have rightly worked hard to break down! We scholars should be among our own objects of study; indeed we need to be, if we are to do that study well. Our concepts that we use as scholars of religious studies have their own history, and we get into trouble if we are unaware of that history. Our methodology, our discipline, our scholarly identities must be among our objects of study, if we are to study anything else well. (From what I have seen of Craig’s work so far, I can’t imagine he could disagree with this much.) And that means that any “should” statement about “us” is a “should” statement about our objects of study. So the statement “we should attempt to avoid… ‘should’ statements about our objects of study” is self-contradicting, and therefore false.
Craig recognizes some of these issues of reflexivity in the closing statement: “Of course, this normative conclusion could be intersubjectively valid only for those scholars who, like me, value intersubjective verification.” But the problem goes deeper than that. The problem is the same problem that vexes logical positivism and related empiricism: this normative conclusion, that we should avoid should statements about our objects of study, is itself not intersubjectively verified! Or rather, it is not intersubjectively verified by the standards that Craig demands for intersubjective verification, with respect to normative recommendations about what scholars should do (the fourth variety of normativity that his post identifies):
the fourth case involves a type of normativity that would not be intersubjectively verifiable by those with competing sympathies. “We should promote equality between men and women” could only be agreed upon by those who share feminist sympathies. Individuals or groups who hold patriarchal norms cannot intersubjectively verify the truth of this “should” claim.
Here, Craig appears to be saying that scholars should only make normative claims that everybody can agree on. He is fine with scholars being motivated by more contestable normative claims (the first case he lists), but they should not make normative claims in the course of their study because those cannot reach universal agreement. Except, it would seem, for the kind of normative claim that does not fall under one of the four cases he lists – norms about the proper conduct of historical-critical study (such as “we should avoid making claims that are falsified by documentary evidence”). Those, he seems to want to accept. But to say that we should only accept such historical-critical or empirical norms is itself a norm – a “should” statement. And that norm is not as widely accepted as the historical-critical norms are themselves. It is relevant here that the norms of historical-critical inquiry are not universally accepted either – they won’t get very far with King-James-Only Christians or young-earth creationists. But the kind of methodological norm that Craig is advocating is even less universally accepted than those. I certainly don’t accept it. And neither does Thomas A. Lewis, whom Craig cites as an opponent at the beginning of his post. Lewis and I cannot and will not intersubjectively verify the truth of Craig’s “should” claims about what scholars do, any more than patriarchal groups can or will verify the truth of “We should promote equality between men and women.” If the normative standard of gender equality fails by the normative standard of intersubjective verification, then so does the normative standard of intersubjective verification itself.
So the norm of intersubjective verification, as described in Craig’s post, is self-refuting. The kind of normative approach taken by Lewis, in which normative claims are fine but require their own justification, provides a much surer way. What is striking to me in this context is the way that, despite his appropriate suspicions toward empiricism, Craig nevertheless manages to reproduce empiricism’s single least defensible feature, the one that makes the whole empiricist research program fall: namely, one cannot empirically verify empiricism. So the kind of strong empiricism advocated by the logical positivists, according to which non-empirical and non-tautological statements are meaningless, would itself be either meaningless or a tautology – if it were true, which of course it isn’t.
Now it is possible to intersubjectively verify some forms of intersubjective verification as a standard. Aristotle’s dialectical reasoning works on grounds one could reasonably call intersubjective verification: one looks at the various positions accepted by those one judges to be wise, and tries to reconcile them in a way that saves a significant amount of the reasoning underlying them. That standard is not self-contradicting. But the reason it is not self-contradicting is it does not aspire to universal intersubjective verification. The fact that certain groups of people could not accept the claim, from the standpoint they currently occupy, is not a reason to reject it. Yes, there are people who refuse to accept that we should promote equality between men and women. There are also people who refuse to accept the theory of evolution, and people who refuse to accept that the Qur’an was collected by human beings over time. They are all wrong. That doesn’t mean there is no truth in their ideas; it does not mean we should refuse to listen to them. It does mean that we have good reason to accept the contraries of their ideas, and are justified in declaring their ideas false and the contraries of their ideas true. I noted a while ago that a universal truth is one that everyone should accept, given appropriate reasoning from their original premises and assumptions; it is not necessarily one that everyone does accept.
But there is no significant difference on these points between statements of fact and statements of value. The latter are as intersubjectively verifiable as the former. There are as many good reasons to believe in women’s equality as there are to believe in the theory of evolution; the latter is no less controversial than the former. And so the rejection of “should” statements should be rejected.