Towards A Kinder, Gentler Verificationism
A Sketch of a Research Project
It often happens, for various reasons, that philosophers defend radical views which, first, are too radical to be plausible, and second, are such that a less radical and more plausible view would satisfy the underlying motivations. Here is a historical example. The logical positivists famously sought to eliminate traditional metaphysics by arguing that the statements metaphysicians make are meaningless because of being unverifiable. Much of the ensuing discussion concerned whether verifiability is really necessary for meaningfulness. But clearly, even if the logical positivists were wrong about this, they could still have a strong case for the elimination of metaphysics. For already if they could establish that the statements made by metaphysicians are unverifiable, they could argue for the pointlessness of the enterprise. If we cannot obtain good evidence for or against the statements of metaphysics, surely metaphysics is a pointless enterprise.
—Matti Eklund, “Rejectionism About Truth”, p. 1, https://courses.cit.cornell.edu/me72/reject_truth.pdf
I think it may be advisable to follow a wise point that lies buried in Logical Positivism: Though it may very well be that unverifiable statements are meaningful, and so true or false, this does nothing to prove they are legitimate objects of inquiry. For the statements the positivists sought to ban are, after all, unverifiable, and where there is no possible method of verifying a statement, there is seemingly no means of resolving disputes concerning it. Save in those cases where our psychology compels our universal assent, such as, perhaps, our belief in the existence of other minds, there is, short of force, coercion or sheer coincidence, no way to reach a consensus with respect to unverifiable statements, and given this it appears unwise to argue over them. For if we do not aim to achieve consensus, to what purpose do we argue? However, it must be admitted that verifiability comes in degrees, and there are perhaps no statements which are conclusively verifiable. Still, one can say that the harder it is to verify or refute a given statement, the less reasonable it is to argue over it. And of course to say that we shouldn’t bother arguing over something is not to say that we cannot have our own, private opinion on the matter.
None of this is to say that the positivists were correct in determining which statements are verifiable and which are not, only that we should not try to assess the truth value of statements which are genuinely unverifiable. I differ from the positivists in my estimation of the scope of the unverifiable. I think there may be a great many metaphysical, religious, and ethical claims which are verifiable. Furthermore, the methods of verification employed need not be empirical; they need only be capable of deciding the issue under examination, of resolving it one way or another. In light of this, I will henceforth talk of decidability instead of verifiability.
In spite of the above qualifications, one might think that any verificationist-inspired project is doomed to failure for much the same reasons the Logical Positivists’ was. As hard as they tried to hammer out a workable verification criterion, everything they came up with proved to be either too strict or too liberal to suit their purposes. If one makes the criterion too strict, it will turn out, for example, that the notion of a natural law is meaningless, since one cannot verify that it holds in an infinite number of instances. Statements about the remote past, the far future, and theoretical entities would also go by the board. But if one liberalizes the criterion so that we count as meaningful statements about things that experience only renders probable, then one will have to admit things into one’s ontology that would make a positivist shudder, such as the “élan vitale” of the vitalists or the Absolute of the Absolute Idealists, if their existence would have any implications, however slight, for our experience. The verification criterion would then seem to impose no real constraints on the meaningfulness of statements. The simple fact is that no verification criterion could do what the positivists wanted it to do. One objection to my proposal, then, is this: Is not my “decidability criterion” dangerously close to a liberalized verification criterion? What, exactly, does the “decidability criterion” rule out?
My reply is that we should not try to deduce a priori that certain classes of statements are undecidable; for there is, I think, no single form or subject matter that unites all undecidable statements and makes them undecidable. In my opinion, the positivists’ failure to recognize this point is a major reason why their repeated attempts to formulate a workable verification criterion failed. And even if there were a single form or subject matter which all undecidable statements had in common, it seems unlikely that a priori reflection could discover it. (One reason to think this is that the sustained a priori reflection of the positivists failed to discover it.) Instead, we should look to the past to see what kinds of statement have proved easy to decide and which have not. My criterion, then, would counsel us to reject statements which are of the same kind as those which history has shown to systematically resist attempts at resolution. In future work, my task will thus be to show that this criterion imposes non-trivial constraints on philosophical practice.
I intend to develop my own kinder, gentler verificationism by using the views of Rudolf Carnap as a foil. Though I sympathize with several facets of his position, and with the motivation behind them still more, I think there are a few important points about which he was mistaken. First, his conception of what philosophical practice should be is excessively formal; it focuses too much on language, and it seems to allow no room for the vagueness or open texture of words and concepts; and it also seems to rule out Wittgensteinian “family resemblances”.
Second, his distinction between “internal” and “external” questions is too sharply drawn. If we consider “frameworks” that are actually in use, in both the sciences and in the humanities, I think we will find that it is not always so clear what counts as part of a framework and what doesn’t. Also, I doubt very much that pragmatic considerations are the only things relevant to the selection of a framework. Furthermore, there may be an objective truth—one transcends any particular framework—even if we can’t know what it is.
Finally, Carnap’s views were bound up with the philosophy of language of his day, before the advent of externalism and causal theories of reference, and stand in need of substantial modification.
As for my own account, I will try to accommodate the lessons we learned about language and categorization from Wittgenstein and the theories in cognitive science that his work helped to inspire, such as prototype and exemplar theories of concepts. Also, on my view, we might be able to attain knowledge of objective truth without having to “step outside” of our various frameworks—though we might also be able to know the objective truth by doing so, if doing so is possible—by comparing our frameworks themselves against each other and seeing what, if anything, they have in common. If all the viable, mutually comparable frameworks concerning a given subject matter agree about something, I think that is good evidence that it is probably true, provided that the different frameworks come close to exhausting the possibilities. (This is, in essence, a generalization of my approach for getting eithical guidance from ethical theories.) And I see no reason why our frameworks need to be (solely) linguistic. I will also try to give an account which is consistent with causal and externalist views on reference and mental content. And on my view, what counts as evidence for what is in part determined by what “context of inquiry” one belongs to, among other things, and so my view of evidence is in some sense externalist.
In spite of the differences I have just spelled out, Carnap and I have something important in common: we both put great stress on engaging in disputes only if there is some way of resolving them.