I have finally finished reading Robert Brandom‘s massive tome (650 pp.) Making It Explicit, and it’s given me a lot of new tools with which to think about the nature of concepts and the relation between language, perception, action, and the world.  This is my first attempt to do something with what I’ve learned from Brandom.

It is crucial to Brandom’s account that conceptual content — what our thoughts are about — is constrained in two different ways: normatively and causally.  Normative constraint is, for Brandom, essentially and fundamentally social and linguistic.  For a community of speakers, each speaker holds herself and the others accountable for what they say by keeping track of the compatibility and incompatibility of their commitments and entitlements. (If I assert p, and p implies q, then I am committed to q.  If I assert p, and p implies q, but I am already committed to ~q, then I am not entitled to assert p.  And so on.)  The various ways in which we keep track of our own commitments and entitlements, and our own, is a process that Brandom calls “deontic scorekeeping”: deontic from <I>deonta</I> (Greek, “duty”), what we ought to be committed to.   We keep score of what we ought to say.    Deontic scorekeeping is the only normative constraint on discursive statuses — what it is that we believe or desire.  The statuses — the beliefs and desires — are instituted by the attitudes of commitment, entitlement, acknowledgement, avowal, disavowal, and so on — and are only fully intelligible in those terms.

There is, of course, causal constraint as well, which is the only way that the world gets a vote on what we say about it.  And this only takes place through perception and action.  The crux of both perception and action is what Brandom calls “reliable differential responsive dispositions”: reliable tendencies to respond in similar ways to similar stimuli and different ways to different stimuli.   In perception, RDRDs are “triggered” by the impact of the environment on our senses and the RDRDs, thus triggered, in turn cause the right kinds of commitments and entitlements — such as, if one’s sensory receptors are exposed to the kind of object that normally causes the sensation of blue, then one is entitled to assert, “there’s something blue over there”.  In action, the story is parallel, except that the commitments and entitlements produce the RDRDs that in turn causally act on the world.

Now, what does all this have to do with the difference between science and metaphysics?   The relation between Brandom’s account of perception-language-action and science is, I would hope, pretty clear — science is just ordinary experience made rigorous and systematic.   Science is reliable because it takes advantage of the causal constraints on perception and action.  But what about metaphysics?

Ever since Kant (well, actually, ever since Leibniz, but Kant gets the credit), we say that science deals with what is actual or probable, whereas metaphysics deals with necessity and possibility.   (In technical terms, metaphysics deals with a priori truths, whereas science deals with a posteriori truths.)   But whence necessity and possibility?   On Brandom’s account, the only possible source of necessity and possibility is the compatibility and incompatibility between discursive statuses that are instituted by the attitudes of commitment and entitlement.  So the only constraint on metaphysical speculation is discursive, socio-linguistic constraint.

But that is a much weaker kind of constraint, because how the world really is doesn’t enter directly into the claims made about how it must be or could be.  The must-bes and could-bes — compatibility and incompatibility relations — are only assessed in light of the deontic scorekeeping (although ordinary background knowledge will be used as ‘auxillary hypotheses’ by each member of the prospective audience in interpreting the utterances, and so attributing content, what is said).   The line between metaphysics and fiction is, accordingly, not so much blurred as erased — since fictions similarly obey compatibility and incompatibility relations (if Frodo Baggins went to Rivendell before going to Mordor, then it cannot also be the case that he went to Mordor before going to Rivendell).  The difference between metaphysical speculation and fictions now appears to consist in how rigorously the rules are enforced.  If theology is “intellectual tennis without a net,” metaphysics has a very short net.

The only way that metaphysics could do any better than this would be if the speculations are constrained by the best available scientific theories, because scientific theories are constrained by the causal powers of the objects of inquiry.  Accordingly we could distinguish between prospective metaphysics (before taking science into account) and retrospective metaphysics (after taking science into account), or pre-scientific metaphysics and post-scientific metaphysics — however one decides on the terminology, it is only by taking science seriously that metaphysics can get any constraint other than the normative constraint built into deontic scorekeeping.