Anomalous Monism Again

Posted on April 20, 2009
Filed Under Davidson, Philosophy | 4 Comments

What initially got me so interested in Davidson was Anomalous Monism. That is briefly the idea that the mental can’t be reduced to the physical yet the physical is all there is. It became rather popular the past 40 years as a way of recognizing the significance of the mental while remaining a materialist. Of course the main positions in philosophy of mind are still reductive even if it seems to many of us that such reductions of futile. But we also don’t want to embrace the kinds of dualisms usually offered as an alternative.

Later in his thought Davidson modified Anomalous Monism. (Primarily due to criticisms by people like Honderich and Kim) Where the early version of AM appears open to the rejection of ontological reductionism in the later version he explicitly accepts ontological reductionism. The distinction appears to be whether the mental can be reduced ontologically to the physical ontologically versus in terms of description.

A common attack on Davidson is that despite his claims the mental becomes impotent. That is it is a form of supervenience in which what is really going on causally is best described physically. That is the mental is causally effectuous but only as physical. The mental does no causes by virtual of its mental properties.

Davidson argues that if one event causes an other then it is a cause regardless of how described. I think this right but the big question I have (still) is how under a change of description one can maintain the same extension. That is take the extension of a mental description. Can one really find a physical description with exactly the same extension? I’m not sure one can.

That’s not the main point of attack for Davidson. Both in his original and revised versions of AM it seems that the issue is laws. There are causal laws that are strictly deterministic versus laws that are more general. By general I mean laws that are typically true but have exceptions. That is mental talk fundamentally is different from talk of causes in science.

To me this ends up tied to the prior point. One problem with mental talk might be that there is no stable extension. It is general in that it could cover some large set S of states. Given that we might call S1 the states compatible with mental description M1. There would then be a S2 which is the states compatible with mental description M2. Now for some state s1 in S1 there exists a lawlike relation to s2 in S2. That is the physical cause. But given the ambiguity in mental descriptions we can’t make the reduction to this physical state.

What I think Davidson wants to say is that for most s1 in S1 there is a relation to s2 in S2 that is nearly lawlike but that there will be some which don’t hold.

If this is so (and I’m not sure I’m reading Davidson right here) then I think the problem is that causality itself depends upon the kinds of descriptions offered. That is mental causes and physical causes are different sorts of talk. The problem is that I think Davidson wants causes to remain constant through translations of descriptions. I think this is because descriptions will pick out events but not solely those events.

Think of Davidson’s example of two identical events. “The assassination of the Archduke Ferdinand” is the same as “the event that started WWI.” The problem is that “the event that started WWI” isn’t only the assassination of Ferdinand. It picks out more individual occurrences. Any translation of description almost always entails that a valid interpretation of that description may pick out more individual occurrences.

Put it an other way. Consider there may be some ultimate description that picks out the occurrences forming an event perfectly (although I am skeptical of this). The kinds of descriptions we use simply can’t. Physical talk is this ideal description. Mental talk simply is “fuzzy” in a way that won’t let us pick out those occurrences.

Is this right? I’m not sure. I’m still thinking through this. The question someone posed to me earlier today is whether this is reducing Davidson to merely making an epistemological point. Perhaps but I don’t think it is, anymore than say Heidegger is only making an epistemological point. I think Davidson is ultimately making an ontological point about the mental. But in his revised view I’m not convinced that is the case.

If I read him right we have to distinguish between the event a speak intends picked out versus what the description as interpreted picks out. Davidson’s argument may rest on a certain equivocation between these two senses. (i.e. the event entailed by a description versus the event as intended)

Related posts:

  1. Yet More Anomalous Monism
  2. What is the Subconscious?
  3. Davidson: The Myth of the Subjective 2
  4. Vallicella on the Ontology of the Mental
  5. Davidson: What Is Present to the Mind
  6. Time and Chance

Comments

4 Responses to “Anomalous Monism Again”

“That is take the extension of a mental description. Can one really find a physical description with exactly the same extension? I’m not sure one can.”

I don’t see why Davidson would need this, stated this broadly. He just needs identities for particular events. So the question is just “Can an event picked out in a mental vocabulary also be picked out in a physical vocabulary”, where events are spatiotemporal particulars. (Davidson changes his views on the identity-criteria for events after criticisms from Quine; his mature position is that same spatiotemporal extent = same event. See the appendix to “Essays on Actions & Events”.)

It seems plausible to me that a region of space-time described in a mental vocabulary can always also be described in a physical vocabulary. Just describe all of the parts of it and build up an ungainly piecewise description of the relevant particular. Any problems in nailing down just which spatiotemporal region to identify with a given mental event will show up in the very attempt to pick out a mental event in the first place, so there’s no problem for anomalous monism, beyond a problem for materialism about mental events in general.

I’m not sure where you think Davidson revises his views on “ontological reductionism”, if that’s the question of whether the mental is something above & beyond the physical. He always held that all mental events are token-token identical with physical events. That’s already in “Mental Events” in 1970. And this makes clear what Davidson says about the matter: what’s there are events, which we can describe in myriad ways. All events must have a physical description which is true of them, if the “Nomological Nature of Causality” is true. But there’s no sense in which events are “really” or “basically” physical. All true descriptions of them are on a par, ontologically. They really are the things they’re truly described as being, and no description has a pride of place.

“That is mental talk fundamentally is different from talk of causes in science.”

It’s important to mark that Davidson does not oppose the mental to “talk of causes”; one of the premises on which Anomalous Monism is argued for is that of psychophysical causation (in both directions). What is opposed to the mental is the nomic: laws, not causes. And here again the distinction is only conceptual, not ontological: the causal relations themselves can be picked out in any number of vocabularies. Causality is an extensional relation between events. Physical vocabularies pick out causal relations in ways that are projectible (lawlike) as instances of general laws; mental vocabularies pick them out as things that make sense as part of a coherent overall picture of an agent’s life. So the fact that Davidson’s “physical talk” aims at an ideal physics doesn’t matter, because even false sentences in a physical vocabulary are aimed at something different than sentences in a mental vocabulary are aimed at. (Ambiguity has nothing to do with it, nor does the indeterminacy of translation — that is part of “Mental Events” that Davidson takes back, in “Three Varieties of Knowledge” and in the appendix to “Truth, Language, History”. The latter is a response to McDowell’s “The Constitutive Ideal of Rationality: Davidson and Sellars”, which is excellent on this topic.)

The causal relations themselves do not care about how they are described, and no true description of them is more or less true than any other.

I think one thing that might be tripping you out is talk of “properties” and “states”. Davidson does not have either in his ontology (talk of them is handled in predicates). What there is is objects and events. He sometimes talks of states such as beliefs as standing in causal relations, but what really stand in causal relations for Davidson are things like “changes of belief” or “gaining a desire while holding a certain cluster of beliefs”, which are events. (He’s more careful about this in “Indeterminacy and Antirealism”.) And properties are something Davidson sees no need for whatsoever. There are objects which are referred to by singular terms & picked out by quantifiers, and there are sentences which are true of them. Nothing in the world needs to correspond to the predicate in a subject-predicate sentence. (See the latter half of “Truth and Predication” for Davidson’s fullest account of this matter, though it’s a trivial consequence of a Tarskian semantics that only Fregean objects need positing, not Fregean concepts or Sinne or Frege’s “The True” or “The False”.)

“Think of Davidson’s example of two identical events. “The assassination of the Archduke Ferdinand” is the same as “the event that started WWI.” The problem is that “the event that started WWI” isn’t only the assassination of Ferdinand.”

I think you’re just objecting to Davidson’s example here: for the example to work, the event that caused WWI has to just be Ferdinand’s killing and nothing else. If there is more than one event that started WWI (say economic troubles in Prussia or whatever), then there is nothing which is “the event that started WWI”. You can replace Davidson’s example with “the firing of the gun” and “the starting of the race” (for some particular race & firing). Which are unamiguous: just the one firing, and just the one starting, are what are meant.

“If I read him right we have to distinguish between the event a speak intends picked out versus what the description as interpreted picks out. Davidson’s argument may rest on a certain equivocation between these two senses. (i.e. the event entailed by a description versus the event as intended)”

I have no clue what in Davidson’s writings on anomalous monism would make you think that anything like this is relevant. Also, events are not “intended” for Davidson; intentions are propositional attitudes. Propositions do not describe events (or anything else); they are not singular terms.

Davidson’s argument for anomalous monism is simple: Take the three premises of “Psychophysical causation”, “The nomological nature of causality”, and “The anomalism of the mental” and it follows straightforwardly. (You can skip part 2 of “Mental Events” entirely and the argument for anomalous monism remains untouched; part 2 is just an abortive attempt to argue for the third premise.)

1. There are mental events which stand in causal relations to physical events.
2. All causal relations are instances of strict laws (of the sort physics aims at).
3. There are no strict psychological laws (no strict laws relating mental events of a certain type to mental events of another type; our mental talk does not even aim at such a thing).

Hence each mental event is identical with a physical event, but there can be no reduction of mental vocabulary to physical vocabulary: monism, anomalous.

Wow. Long reply. Thanks.

Regarding picking out. The issue is ultimately how our descriptions pick out. Clearly they can (taken “literally”) pick out more than we need. Further they may pick out less, but due to our ability to interpretively make guesses we can pick out what is relevant. In that sense we can pick out the appropriate event. So in that sense we agree. I guess what I’m saying is other things can be drug along or missed. Perhaps, given his views on radical interpretation this isn’t a big deal.

The way I once saw this as problem is that it’s not clear what area of space-time gets picked out by mental descriptions. (The problem of externalism) I’ve come to realize that Davidson is less radical an externalist than I originally thought. Yet I don’t think he narrows mental descriptions just to the brain. (I need to go back and check on that point – I honestly can’t recall if he adopts a more Searle-like position here)

You appear to correct an error I made here and in the subsequent post. I was thinking that perhaps despite Davidson’s protestations that the same causes aren’t really picked out. But this is basically the same point in the first paragraph of this comment I made. Is the “a little more and a little less” significant? Clearly you don’t think so. And given an ideal interpretation I guess I have to agree.

This is the issue you object to most strenuously. Note by “intend” I meant the event intended in a description. That is I intend my description to pick out a particular event even if the words could pick out more. My confusion is over this ability of descriptions to narrowly pick out only the appropriate occurrences making up an event so as to constitute the same even under a translation of descriptions. Davidson just appears to take for granted this isn’t problematic. I’m very skeptical here. I need to reread Three Varieties of Knowledge here though. I remember it addressing it (although being unconvinced).

I think you’re right that what is confusing me is the issue of properties and predicates.

I suspect that what has happened (and why I’m not addressing AM in terms of the three points) is that I’ve become skeptical of (2). That is clearly we have no example of a strict law. Once, in the 19th century, we thought we had one. Now no one thinks we have one. If there are no strict laws then there are no causal relations. So in effect my doubts towards AM are doubts about causality in general. I think one can talk about descriptions of causality but they simply aren’t strict.

I expect this means that I’m coming to reject physicalism although it is itself such a muddled concept that I’ve long had trouble with it. I think one can accept most of physicalism without accepting that particular lawlike aspect.

To add, clearly descriptions by themselves can’t pick out. Rather it is descriptions in a context.

OK, I’ve been reading up on properties and I think that is indeed where I was getting confused. Part of the problem is that in his papers he does talk about them at times. But I forgot that for him it is more about predication by an interpreter. My bad.

Leave a Reply