Friday, June 09, 2006

What Physicalism Can't Allow

A couple of posts ago I said “…an account of genuine ontological emergence…must fail, given an assumption of physicalism.” Fortuitously, this issue was explored in depth in a paper posted by Terry Horgan in the last week of the recent Online Philosophy Conference entitled Materialism: Matters of Definition, Defense, and Deconstruction. There was also an interesting commentary on the paper by Thomas W. Polger. Of the several related topics explored in Horgan’s paper, one crucial issue was whether necessary “inter-level” metaphysical relations which go beyond physics are allowed or not for the materialist. (Please note that “materialism” and “physicalism” refer to the same thing – the older term has been ‘winning out’ in recent times it seems).

In the first part of his paper, Horgan sets out to take the pre-theoretic conception of materialism which he thinks most people share and create a rigorous definition.

I’m going to summarize the first parts of his formulation quickly, so please refer to the paper for the real thing. His effort is in the spirit of ideas from David Lewis and Frank Jackson, among others. Horgan utilizes “physically possible worlds”, which are possible worlds which have the same physical laws and entities as the actual world (“worlds” here mean maximal propositions: the actual world instantiates a possible world). Then to help screen out physically possible worlds that have superfluous non-physical stuff, he utilizes also uses the notion of a “minimal physical duplicate” of a world, which duplicates a world’s physical stuff but contains nothing beyond whatever is needed for it to be a physical duplicate. His definition of materialism then says that the actual world is such a minimal physical duplicate of itself, and for any 2 physically possible worlds, their minimal duplicates are identical. So, the actual world has no extra non-physical stuff; further, all physically possible worlds (which like the actual one lack superfluous non-physical stuff) must be the same in all respects.

[UPDATE/CORRECTION - JAN 26,2007: A helpful e-mail correspondent has pointed out a couple of problems in the the last two sentences of my summary above. First, where I say "for any 2 physically possible worlds, their minimal duplicates are identical" -- this will only be true only if the worlds selected are alike in all physical detail, not just laws and types of entities (Horgan's text doesn't say this, but it is implied) Then my last sentence has two mistakes: I should not have said all physically possible worlds must be the same in all respects; further, while the definition specifies that the actual world doesn't have superflous stuff (since it is a minimal duplicate of itself), there is no attempt here to rule out superfluous stuff at all physically possible worlds. I apologize for the errors.]

Horgan further adds a loophole-closer ruling out worlds having more than one spatio-temporal regions within them which are physically identical but otherwise differ.

So far this is somewhat familiar territory, but Horgan next explores one of the thorniest and most interesting aspects of defining physicalism.

Even the formulation given above isn’t enough to meet our pre-theoretic concept of materialism, says Horgan, and he adds a final part to his formulation. He worries that someone could propose a necessary metaphysical supervenience connection between certain high level non-physical properties and low level physical properties where such connection is brute and could not be derived from other facts. Such a “strong emergentist” proposal is not ruled out by the modal appeal to minimal physical-duplicate worlds in the definition (the necessary connections would exist in all worlds).

So he adds another clause to his formulation of materialism which says: “there are no brute inter-level relations of metaphysical necessitation linking physical particulars or properties to non-physical particulars or properties.”

Polger wonders if this clause amounts to “throwing out the baby with the bathwater.” Horgan is led to the clause by considering some examples of strong emergence which seem obviously unacceptable to the physicalist. These include G.E. Moore’s meta-ethical position that “intrinsic goodness” necessarily supervenes on the physical; and, of course the example of immaterial Cartesian minds brutely and necessarily emerging when a human brain is formed. Polger says: “is it generally true that being a relata of a metaphysically fundamental necessitation relation disqualifies a property from being physically acceptable? I don’t think so.” He then lists some examples which he thinks would be OK. These include: causal relata; the whole individual’s relations to its parts; identity relations between higher level and lower level concepts (water and H2O). Polger argues that just because some examples of brute metaphysical relations are not compatible with physicalism doesn’t mean they all are.

Polger concludes that the extra clause goes too far. He concedes it would remain an interesting and important question which metaphysical relations are physically “kosher” and which are not, and how we decide this. Further he says it remains a particularly difficult problem to see how phenomenal consciousness will be made to fit into the picture. “But it should be clear enough that we need not answer that question by consulting physics alone…Why think that physics itself could be the source of not only the basic entities and properties but also the source of all ontological relations?”

Well, I think that most materialists/physicalists do think physics itself is (or will be) adequate and therefore Horgan is right to include his additional clause. I find Polger’s objection to be very interesting because, apart from serious metaphysicians themselves, one never sees advocates of materialism conceding that it is an inadequate metaphysics without explicitly adding relations (or entities) which go beyond current or contemplated physics. While scholars continue to debate competing theories of causality and ontology, most folks appear to assume one can be a materialist without troubling too much about these difficult topics.

There is a good deal more to Horgan’s paper (also Polger’s response) than what I have discussed here; the heart of the paper concerns how one can go about defending materialism as defined, and what are the prospects for success. This focuses on finding “kosher” or materialist-friendly ways to conceptualize the phenomena in the world which we describe using non-physical vocabulary. He notes how difficult this is especially when it comes to phenomenal concepts, and concludes that philosophers have yet to successfully defend materialism by providing adequate accounts of these concepts.


marco said...

For any physical theory fitting the phenomena, we can contruct as many more as want that fit equally well. So considering and choosing the simplist, elegant theory has to be part of physical enquiry.

Horgans conceives 'physical dupliate' in terms of a completed physics. So that means all less simple, less elegant etc theories are ruled out. Why not believe that attention to similar theoretical virtues rules out exacly those positions that Horgans want to rule out by introducing his third clause? It seems to me at least that thus, dualism and Moorean non-naturalism are ruled out.

Steve said...

I think that's right, except that it is still an interesting question as to whether any metaphysical relations beyond those within the domain of physics will be needed to make physicalism work. You bring up a good point also with regard to the idea of a future completed physics. If physics itself ends up expanding to encompass entities or relations which do more work explaining high level phenomena like conscious experience, then the need to go beyond physics falls away.

marco said...

That's my hunch: if I see physics and philosophy sharing general tools of rational enquiry, and if when we talk of a completed physics we mean the completed exercise of this rational enquiry, the thesis of physicalism just becomes somethng like: that we should believe that something exists if we have reason to believe it. I'm aware this is an existing problem: do you know if it's discussed anyhere, and how one might respond?

Steve said...

Well, I'm not sure. I can't remember seeing anything relating the discussion of physicalism to a more epistemological statement like that.