Thursday, April 17, 2014

Reference magnetism and anti-reductionism

According to reference magnetism, the meanings of our terms are constituted by requiring the optimization of desiderata that include the naturalness of referents (or, more generally, by making the joints in language correspond to joints in the world, as much as possible) and something like charity (making as many real-world uses as possible be correct).

Suppose we measure naturalness by the complexity of expression in fundamental terms—terms that correspond to perfectly natural things. (In particular, we can't talk of what cannot be expressed in fundamental terms, since reference magnetism would presumably not permit reference to what is infinitely unnatural.) Consider the reductionist thesis that the vocabulary of microphysics is the only fundamental vocabulary about the natural world. If this thesis is true, then our ordinary terms like "conscious" or "intention" or "wrong" are going to be cashed out in terms of extremely complex sentences, often of a functional sort. But I suspect that once these expressions are sufficiently complex, then there will be many non-equivalent variants of them that will fit our actual uses about as well and are about as complex. Consequently, we should expect that the meaning of terms terms like "conscious", "intention" and "wrong" to be highly underdetermined.

If we have reason to resist this underdetermination, we need to embrace an anti-reductionism on which the terms of microphysics are not the only fundamental ones, or else have another measure of naturalness.


Jonathan D. Jacobs said...

Why think this is true? "[O]nce these expressions are sufficiently complex, then there will be many non-equivalent variants of them that will fit our actual uses about as well and are about as complex."

Brian Cutter said...

I agree that considerations like these pressure us either to "embrace an anti-reductionism on which the terms of microphysics are not the only fundamental ones, or else have another measure of naturalness." But it seems that similar considerations apply to predicates which no one wants to say correspond to fundamental properties, e.g. "table," "mountain," "backpack," and indeed, just about any English predicate. This suggests that considerations of this kind provide more support for thinking that this Lewisian measure of naturalness is inadequate than for thinking that there are fundamental properties beyond those mentioned in physics.

Alexander R Pruss said...


Take some clause in the very complex expression, which clause is very rarely actually relevant, and replace it by its conjunction with a very simple fundamental truth. Complexity goes up by a tiny fraction, and fit to actual use is about the same.


Yeah, but there is no cost (or negative cost, i.e., benefit) to a view's leading to the conclusion that "table" has many candidate meanings that are about as good. But for "conscious" and "obligatory" that's a cost.

Brian Cutter said...

Maybe I'm misunderstanding the kind (or rather degree) of semantic indeterminacy you have in mind. On the one hand, a theory might entail that a predicate like "table" is "radically" indeterminate in meaning, such that, for example, among its candidate extensions is the set of hard-boiled eggs. (Of course, there are interpretations of our language which assign this extension to "table" that do just fine in terms of charity, at least as measured by truth maximization, so long as other expressions receive coordinated deviant interpretations.) Such a consequence would be a very serious cost to a theory, I think. On the other hand, a theory might have the consequence that "table" only has the kind of "mild" indeterminacy associated with vagueness. On this option, very roughly, the candidate meanings would all be very similar to one another in some intuitive sense, as the property of being at least 5'11" tall is intuitively very similar to the property of being at least 5'11.1" tall. If "table" is indeterminate in this way (e.g. because one eligible meaning excludes such penumbral items as the tree stump on which I lay out my picnic lunch, and another includes it), then that would not be a cost.

The corresponding form of indeterminacy in the case of "conscious" might have it that different candidate interpretations disagree on the hard cases, e.g. honey bees, worms, or humans on the border between full wakefulness and full sleep, but agree on the easy cases, e.g. dogs, chimps, humans (while awake).

Is your claim just that this more mild form of indeterminacy results from reductionism + the Lewisian measure of naturalness? Or that something more like the "radical" indeterminacy described above would result?

By the way, Geoff Lee has a really great paper on related issues called "Alien Subjectivity and the Importance of Consciousness." (His "Materialism and the epistemic significance of consciousness" is also relevant.) If I recall, his basic idea, for which he argues very persuasively, is that if reductive materialism is true, then in many cases the question of whether an organism is conscious is not "substantive" (in roughly Sider's sense, which is understood in terms of naturalness, though significantly there's no reliance on the "length of definition" measure of naturalness). Lee goes in for a modus ponens here, but arguably a modus tollens is more reasonable, given the intuitive plausibility that questions of whether a thing is conscious are always substantive.

Alexander R Pruss said...

It's the milder form I had in mind (though my choice of words "highly underdetermined" was unfortunate). That's why I think it's not a cost at all in the case of "table", but I am inclined to think it's very much a cost in the case of "wrong".

Heath White said...

“we need to embrace an anti-reductionism on which the terms of microphysics are not the only fundamental ones”

I would go for that. Mental and moral predicates are good candidates.

“or else have another measure of naturalness.”

I would go for that too. Epistemically easy cases, like mountains and backpacks, are good candidates.

One lesson we might take from this is that “carving at the joints” is the wrong metaphor. It makes us focus on defining the edges of meaning, the necessary and sufficient conditions, rather than defining the center, the exemplars and paradigm cases. What if we said: we want the bones of our language (the non-vague, paradigm cases) to match the bones of the world?

Also, I don’t think it’s critical that, say, “wrong” be a non-vague predicate. Paradigmatic venial sins are good cases to think about: I spent 15 minutes the other day playing online chess when I probably should have been grading papers. Was it wrong? What if I had only spent two minutes? How long a break from grading can I take (or how long can I procrastinate) before it’s wrong?

Alexander R Pruss said...


I wonder if vagueness is the right worry.

If you thought you were doing wrong, you were doing wrong. (At least against your conscience.) :-(

That said, one might think that "wrong" has some vagueness around the edge, but it doesn't suffer from indeterminacy of reference. I.e., there aren't multiple properties wrongness1, wrongness2, ..., such that it is indeterminate which provides the referent for "wrong" as used in its central moral sense. Indeterminacy of reference is supposed to be different from vagueness: the multiple properties might themselves be vague in different ways.

Brian Cutter said...

It seems that some instances of vagueness are just a matter of indeterminacy of reference (or content). It's probably important to distinguish between vague *predicates* and vague *properties.* I would want to say: a necessary condition on a predicate's being vague is that it is indeterminate which property it expresses, whereas a necessary condition on a property's being vague is that, possibly, there is something such that it is indeterminate whether it instantiates the property. It seems that the predicate "wrong" (as used in its core moral sense) is not vague, even if the corresponding property is. By contrast, "tall" seems to be a vague *predicate.* Some evidence for this: Suppose Joe is a borderline case of "tall." We feel as though a dispute about whether Joe is tall is merely a verbal dispute. By contrast, (it seems to me that) no dispute as to whether a particular act is morally wrong seems to be merely a verbal dispute, even when it seems to have no determinate answer.

Alexander R Pruss said...


That sounds like a reasonable way of drawing the distinction.

Maybe one way to argue for it is this. Wrongness plays a unique rational role -- I am thinking here of the overridingness of moral reasons -- that it could not play if there were more than one property playing that role. But it could perhaps still play that role while remaining vague.