Terence Horgan

Terence Horgan (University of Arizona), “Materialism: Matters of Definition, Defense, and Deconstruction,” with commentary by Thomas Polger (University of Cincinnati).  Both the paper, commentary, and reply can be found here.

Comments

[Adam Pautz]

Thanks for the interesting exchange. I had a question for Terence Horgan about Moorean Emergentism and mental causation.

Suppose that experiential properties are not identical with physical properties. Suppose they are not even identical with functional properties. In fact, suppose that they are simple, primitive properties. This is a kind of super-non-reductionism. Some who call themselves ‘non-reductionists’ claim that experiential properties are not identical with physical properties but go on to identify them with functional properties. The view I have in mind goes further: it claims that experiential properties are not even identical with functional properties. There are no informative true identity statements concerning them at all. Call this view *Primitivism*.

Primitivists face a modal issue. They have two options:

(T) The primitive experiential properties supervene with metaphysical necessity on the physical/functional properties.

(T*) They supervene with only nomological necessity.

Ontologically, these views are identical. They differ only modally. Horgan calls (T) ‘Moorean Emergentism’.

It is worth noting that some actually seem to defend (T). As I read him, Nagel defends (T). I think Stephen Schiffer defends something like (T) as regards intentional properties. And John Campbell and Colin McGinn and *maybe* Yablo defend something like (T) as regards colors, holding that they are primitive (and so real ontological additions) but supervenient as a matter of metaphysical necessity. And, as I read him (maybe I read him incorrectly), Joseph Almog (in *What am I?*) attributes to Descartes a stronger view: the view that souls and bodies are (wholly?) distinct but somehow bound up as a matter of metaphysical necessity – a view which Horgan mentions as a possibility on p. 5.

One issue is whether (T) is a form of Physicalism or Dualism. One could count it a form of Physicalism because it upholds the kind of modal thesis that many now take to be definitive of Physicalism. One could call it “Primitivist Physicalism”. Alternatively, one could take it to be a form of Dualism because it upholds the kind of ontological thesis definitive of (property) Dualism. Then one could call it (to use Frank Jackson’s term) ‘Necessitarian Dualism’. Horgan argues that it should not be counted a form of Physicalism.

But I don’t want to speak to this issue. I want to speak to the different issue of whether (T) has any advantages of (T*). If we had to choose between these options, which should we choose? The question is not whether (T) deserves the label “Physicalism”. It is whether it deserves (or might deserve) belief.

Two preliminary points. First, Necessitarians about laws will think (T) and (T*) are the same. But set that view aside. Second, some may say that one could never be justified in accepting (T). For how might one be justified in accepting Primitivism? One might think that the only argument is some kind of conceivability argument. This is an argument for ontological distinctness via modal separability. So if one accepts this argument, then one must accept modal separability as well as ontological distinctness. In other words, one must reject (T). However, I think that there is another good reason to accept Primitivism: a kind of argument from failure. The project of reducing consciousness just does not work. If one accepts Primitivism for this reason, one can consistently accept the strong modal claim embodied in (T). The question remains whether one should.

Now some might think that (T) can be immediately ruled out apriori. It requires unHumean necessary connections between wholly distinct properties and those, it might be said, are unintelligible. But don’t we all have to admit some of those? Consider, for instance: necessarily, if something is red, it is extended. Or: necessarily, nothing can have a mass of 2 grams and 4 grams. And if we admit some unHumean connections, why couldn’t there be a lot of inter-level ones?

To see whether there might be any reason to accept (T) over (T*), I think it may be helpful to consider the question of whether the traditional arguments for Physicalism carry over to (T). In my view, the main arguments for Physicalism are the Inductive Argument (Smart suggests this argument when he says “That everything should be explicable in terms of physics . . .except the occurrence of sensations seems to be frankly unbelievable”), the Simplicity Argument (Smart suggests this distinct argument when he says that nomological danglers smell funny), and the Causal Argument (this argument came later, as I see it).

In my view, the Inductive Argument does not carry over. To get an inductive argument going, one would have to show that (T) is true for lots of other properties (being good, being a mountain), and I think that’s hard to do. The Simplicity Argument also does not carry over: where (T*) requires nomological danglers, (T) requires *modal danglers*: strange metaphysically necessary connections of a quite different from those typically recognized by philosophers, which in this sense dangle from the rest of the body of modal truths.

The question I’m interested in is: Does the Causal Argument carry over? Some will say that it does. They will say that, even if experiential properties are primitive, provided that they supervene with metaphysically necessity on physical/functional properties (and the right counterfactual/proportionality facts obtain, etc.), then our common sense beliefs about mental causation come out true. They will say that (T) provides a happy-face account of mental causation, while (T*) requires Epiphenomenalism or Overdetermination. And they will say that this provides a reason to accept (T) over (T*). My impression is that many will say this. So, they will say that (T) captures at least *one* of the motivations behind Physicalism, and so captures at least part of the spirit of Physicalism, even though in other ways it does not. But this seems wrong to me. Intuitively, (T) and (T*) are on a par with regard to mental causation as well. Both views face the dilemma between Epiphenomenalism and Overdetermination, in spite of their modal difference, because to get that dillemma going all you need is Primitivism. But I have not been able to figure how out to show this except by intuition-mongering.

I am interested in what Terence Horgan thinks about this in light of his own approach to the problem of causal exclusion, which views causal explanation as involving fitting a given phenomenon into some pattern of counterfactual dependence and which also incorporates the claim that the notions of causation and causal explanation are context-sensitive. On this view of causation/causal explanation, if (T) is true, do our ordinary beliefs about mental causation come out true or false? Thanks!

[Tom Polger]

From Terry:

Thanks for the question. I'm on vacation, so will now respond only very briefly.

I'm pulled two ways about whether T does better than T* in terms of vindicating mental causation--and both ways I'm pulled are reflected in things I've written on the matter. I do feel the appeal of arguments to the effect that if the mental is "modally separable" from the physical, then mental properties lack causal efficacy; I used this line of thought in defense of a thesis along the lines of T, in my Phil Review paper 'Supervenient Qualia' from about 1989. On the other hand, as Adam mentions, lately I've been attracted by a fairly liberal approach to mental causation, an approach that stresses (a) that causal efficacy of a given property is a matter of the property's fitting into a suitable, non-accidental, pattern of counterfactual dependencies, (b) that such patterns can exist at multiple levels of description, and (c) that contextual factors can operate to fix on a given pattern as the contextually pertinent one. Prima facie, this liberal approach could vindicate the causal efficacy of mental properties just on the basis of T*, even if T should turn out false.

As I say, I'm pulled both ways on the matter. I'm unsettled about this, as I am about the truth or falsity of physicalism/materialism itself.

TH