On the Utility of the Eliminativist/Reductivist Distinction

If we think carefully about examples such as the following, I think we can come to agree that it is useful to make a distinction between eliminativist and reductivist claims.  The distinction is useful because it allows us to disambiguate claims that otherwise would be ambiguous.  Roughly, the distinction is between claims of the form There are no Fs and of the form There are Fs but Fs are Gs.

1. God is an anthropomorphic projection. (Feuerbach)  This could be construed as implying that there is an x such that x = God and x is an anthropomorphic projection.  This would be a reductivist construal.  It has absolutely nothing going for it given that the concept of God is the concept of a being that exists a se, and thus independently of human beings and their thoughts and projections.  If you think that God could be an anthropomorphic projection, then you simply do not understand the concept of God -- whether or not anything instantiates the concept.The Feuerbachian claim is to be read eliminatively as  implying that there is no x such that x is God.

So here we have a very clear example of a sentence which, though it appears to be predicating something of God, and thus presupposing the existence of God, is really a negative existential in disguise. 

2. The self is a bundle of perceptions. (Hume)  This example, unlike the first, can be read either way with some plausibility.  I would argue that a mere bundle of perceptions (or of other mental data) cannot constitute a self because a self is that which supports and unifies and is aware of such data. But other philosophers will disagree.  So for present purposes I judge this example to be susceptible of both readings.

3. Causation is regular succession. Someone who claims that events e1, e2 are such that e1 causes e2 iff the e1-e2 event sequence instantiates a regularity is arguably leaving out something so fundamental -- the notion that the cause produces or brings into existence the effect --  that the claim is tantamount to a denial of causation.  Or so I would argue.  But regularity theorists will vigorously disagree.  They will take the dictum (suitably expanded and qualified) to express the nature of causation.  They will insist that there is causation but that what it is is regular succession.  So, in an irenic spirit, I will classify this example as open to both readings.

4. Properties are sets.  (David Lewis)  Accordingly, the property of being red is the set of all actual and possible red things.  A cruder form of the theory is that the property of being red, e.g., is the set of all red things.  The theory in either form is hopeless, but that is not the question.  The question is whether it is eliminativist or reductivist.  Is it tantamount to a denial of properties, or does it imply that there are properties but that what they are are sets?  I say the former, but David Lewis is one formidable opponent!  Here is a quick little argument:  Properties are instantiable entities by definition; no set is instantiable; ergo, no property is a set!  My considered opinion is that 'Properties are sets' boils down to a denial of properties.  If you understand the concept property, then you know no property could be a set.  It is just like #1 above: if you understand the concept God, then you know that God could not be an anthropological projection.

5. Mental events are brain  events.  I suddenly remember an evening spent on the banks of the River Charles with a pretty girl . . . . That sudden remembering is a mental event token.  There are those who want to say that it is identical to a brain event token. These philosophers speak of 'token-token identity theory.'  The philosophers who maintain this do not intend to deny the existence of mental events; their intention is to inform us as to the nature of mental events on the presupposition that  they exist.

But although their intention is reductive identification and not elimination, one can reasonably wonder whether the reduction does not collapse into an elimination.  Indeed, that is what I would maintain.  For if every mental state is identical to some brain state, and if the identification is supposed to be a reduction  of the mental to the physical, then what you have in the final analysis is just the brain state: the mental state has been eliminated.

Even if you disagree with me that in this case the reduction collapses into an elimination, to even understand what the debate is about you must understand the distinction between elimination and reduction.

6. The tree in the quad is a cluster of ideas in the mind of God. (Berkeley)  The good bishop is not denying that there are physical things; he is telling us what he thinks physical things are.  They reduce to clusters or bundles of divine ideas.  It shows a complete lack of understanding to think that stone-kicking is so much as relevant to the idealist thesis.  So this is a clear case of a reduction.

Could one argue that in this case too the reduction collapses into an elimination?  If the mind-brain identity thesis collapses into an elimination of the mind (as was claimed in #5), then why shouldn't the Berkeleyan identity thesis (Physical objects are a clusters of divine ideas) collapse into an elimination of physical objects?  Perhaps we can say the following.  That the existence of a tree is its existence for the divine mind is consistent with everything we know about trees. (We do not know about trees that they can exist independently of any mind.)  But that a mental state is identical to a brain state is not consistent with what we know about mental states.  Thus we know that they exhibit intentionality while physical states do not.  Mental states cannot be identical to brain states; therefore, a materialist about the mind must be an eliminativist.  But an idealist about physical objects needn't be an eliminativist.

Comments

[William]

A very good post. I like the way you characterise the distinction in the first para. "Roughly, the distinction is between claims of the form There are no Fs and of the form There are Fs but Fs are Gs."

Now I'm off to bed. I recommend the Churchland paper I mentioned earlier.

http://psych.dbourget.com/readings/churchland.pdf

[bob koepp]

Bill V - Thanks for this post looking at how a variety of identity claims intersect with a rough and ready distinction between eliminativist and reductionist views. However, in your final paragraph I think you go well beyond that primarily explicative project and allow your evaluation of the soundness of certain theses to skew your judgment that in some cases, reductions collapse into eliminations. For example, you might expect "blowback" when you say that we know that physical states do not exhibit intentionality. A great many philosophers and cognitive scientists, even those sympathetic to your position, would insist that we know no such thing.

[Bill Vallicella]

Thank you, gentlemen. The fine points can be debated, but I think I have said enough to show that there is a useful distinction between eliminativist and reductionist claims.

[Joseph A.]

Bob Koepp,

"For example, you might expect "blowback" when you say that we know that physical states do not exhibit intentionality. A great many philosophers and cognitive scientists, even those sympathetic to your position, would insist that we know no such thing."

The response I've more commonly read is that if there really is intentionality in the material world - if this physical state really is "about" this or that, etc - then we've left physicalism/materialism behind. Bill says he thinks in the case of the mental, reduction is elimination. But I wonder if he would agree that another possibility exists: That 'the physical' could have been redefined to something else in the process of the argument. Ed Feser has argued this, arguing that to assert intentionality is really present in the physical world is to embrace Aristotilean metaphysics.

I mention this because a good exchange as popped up over how elimination is or could be distinct from reduction. You say it's not clear that physical states do not exhibit intentionality. I could, given Feser's views (I find them compelling, really), agree with that. But how far does that uncertainty about the physical go? It seems to me Berkeley could join in and say that "We don't know the physical isn't what I take it to be", and argue he's a physicalist after all.

[William]

>>For if every mental state is identical to some brain state, and if the identification is supposed to be a reduction of the mental to the physical, then what you have in the final analysis is just the brain state: the mental state has been eliminated.

Actually I'm not sure I agree with this. If every mental state is identical to some brain state, you haven't eliminated any mental state, unless you have eliminated some brain state.

Either the definition of 'mental state' includes some feature that is inconsistent with being a physical state (let's say, intentionality). In that case, no mental state can be a physical state. Or: intentionality is a non-essential feature of mental states. In that case, nothing is eliminated, except for mental states under some description ('mental states that exhibit intentionality').

[William]

Another question (I discuss this here http://ocham.blogspot.com/2010/08/reducing-hesperus.html in case anyone prefers carry all discussion through a 'chairman'). If a reductive assertion is merely an identity, why isn't it symmetrical? The statement 'A=B' has the same truth-value as 'B=A'. But the statement 'mental states can be reduced to brain states' does not seem to be equivalent to 'brain states can be reduced to mental states'. That suggests that something stronger than simple identity is going on. The 'something stronger', I suggest, is the implication that something has been lost in the reduction.

[William]

PS: adding a further example:

(E) There are no walls. There are only bricks arranged wall-wise.
(R) There are walls, but walls are only bricks arranged wall-wise.

[bob koepp]

Joseph A - My personal view is that the only way the mental will ever be successfully reduced to the physical is for our ideas about physical reality to undergo some pretty radical revisions, so we come to attribute to "brute" matter properties that are not presently thought of as physical. It's happened before. Gravity, as conceived by Newton, was an "occult" property in the eyes of many of his contemporaries. After Newton, matter was far from being "inert." Perhaps we will someday have a new Newton, after whom matter will be "mind infused" -- without ceasing to be matter. I view this as a live possibility.

[Joseph A.]

Bob Koepp,

I'd be tremendously sympathetic to that view, and I'm not singling out you or your view on this matter. My problem is this: Just as it can be argued that a reduction is really an elimination, I wonder if a materialism that embraces "a pretty radical revision" of matter is still materialism.

Again: Is Berkeley a physicalist/materialist after all? He (as per Bill) doesn't deny physical objects exist - for him it's a question of what those things reduce to. And is what Berkeley reduces them to just physical reality with "some pretty radical revisions"? If not, why not? And if so, how can anyone be anything BUT a physicalist according to that definition?

I guess another question would be: If I say I'm a reductive or eliminative materialist, yet my view of matter is radically revised, have I really reduced or eliminated anything?

[bob koepp]

What sorts of changes our idea of matter can undergo and still be about matter probably doesn't admit of a "principled" answer. But as I said, we went from matter as inert to matter infused with a very strange power that acts at a distance and isn't attenuated in the process. That's sufficiently weird, even to my modern sensibilities, that I understand why it was once rejected as "occult" by very smart people.

As for Berkeley... if we take him on his own terms, in his own historical context, there's no doubt that he was an idealist and not a physicalist/materialist. If we attach radically different senses to such terms we arent' really discussing his views.

ps - I've already had my say about whether reduction is eliminative, so I'll leave it at that.

[T. Hanson]

Bill, you said:

"But although their intention is reductive identification and not elimination, one can reasonably wonder whether the reduction does not collapse into an elimination. Indeed, that is what I would maintain. For if every mental state is identical to some brain state, and if the identification is supposed to be a reduction of the mental to the physical, then what you have in the final analysis is just the brain state: the mental state has been eliminated."

It depends on what the terms in the initial folk psychology refer to. If the reference of "pain" in the folk psychology is the qualia of pain then physicalism would arguably eliminate pain, and the mental.

But what if "pain" as we commonly use it refers to qualia plus brain states plus? Then collapsing "pain" into brain states would only eliminate the qualia, and leave the other element of the meaning intact. This would be reductivist, since it would not eliminate pain in the original meaning completely.

Since whether a view is reductivist or eliminativist depends on the semantics of the original terms, in many cases it will never be settled. Though I have not read Stitch on this, it looks like (from the SEP article) this is the point of his criticism of the use of this distinction.

[Bill Vallicella]

T. Hanson,

>>If the reference of "pain" in the folk psychology is the qualia of pain then physicalism would arguably eliminate pain, and the mental.<< Since the context is the mind-body problem, the reference of 'pain' is of course the pain quale, the felt pain, the pain as felt, the phenomenal pain, call it what you will. The claim that felt pain is a brain state (or any physical state)is eliminativist: if implies that there is no felt pain. And so I say that identity materialism with respect to qualia is a lunatic position because it contradicts that which is directly evident.

I also rather doubt that there is such a thing as folk psychology. Words like 'pain' and 'desire' are not theoretical terms, they are 'datanic' terms. If I am in a state of desire, it is directly and indubitably evident to me that I am. There is nothing theoretical about thirst or lust.

>>But what if "pain" as we commonly use it refers to qualia plus brain states?<< I rather doubt that 'pain' as we ordinarily use it has this reference. Aristotle thought the brain was an organ for cooling the body. But I'd bet that he used 'pain' or rather the Greek equivalent in roughly the same way we do.

What is your typical runner referring to when he uses 'knee pain'? Even if he is not referring merely to the quale, he is not referring to something going on in his skull but to something going on in his knee.

[Bill Vallicella]

>>PS: adding a further example:

(E) There are no walls. There are only bricks arranged wall-wise.
(R) There are walls, but walls are only bricks arranged wall-wise.<<

Right. The question is whether, with respect to artifacts, there is a y such that the xs compose y. Van I. says 'No.' So he is an eliminativist about artifacts.

But he is not a lunatic open to a Moorean rebuttal because he does not, like the lunatic, deny the xs.

Does (R) collapse into (E)? I think it does.

[Bill Vallicella]

Perhaps we could put it this way. For van I., an inventory of the 'ultimate furniture of the world' would not mention any human artifacts.

[William]

>>Does (R) collapse into (E)? I think it does.

yes. There is no difference between the 'ontology'. Both could agree that there are only bricks, arranged in certain ways. But there is a profound disagreement about the meaning of 'wall'.

[William]

>>Perhaps we could put it this way. For van I., an inventory of the 'ultimate furniture of the world' would not mention any human artifacts.

Even that is problematic. Van I visits my house, and mentions the bricks beautifully arranged wall-wise behind the bricks arranged house-wise. Has he not mentioned (in so many words) a human artefact? He would refuse to qualify it as a human artefact of course, at least with the singular 'a'. But has he not mentioned it, for all that? Highly confusing.

[Mike Rand]

On the Feuerbach example, your thesis revolves around the referent of God being the concept of God and that concept of God being something like an Aristotelian or Thomistic concept of God. God was presented, via scripture, church and theology, to most nineteenth century Germans with pretty much the same sense, but not all nineteeth century Germans assumed the same referent. If we suppose that God has the same sense but a different referent for Feuerbach, then "God is an anthropomorphic projection" becomes a reductionist statement, although it remains eliminative of the Thomistic concept of God.

[T. Hanson]

Bill, you said:

"What is your typical runner referring to when he uses 'knee pain'? Even if he is not referring merely to the quale, he is not referring to something going on in his skull but to something going on in his knee."

Okay, but I think the physicalist could easily grant that pain is not just a brain state, but the neural state part of which is occuring in the knee.

You may be right about the typical runner, but what if the runner is a doctor or a physical therapist or anyone familiar with physiology who would know pain is intimitely linked up with physical processes and whose concept of pain includes these (as well as qualia)? Would the Churchlands, for example, then be eliminativists with respect to the typical, uninformed, or child runner, but not for the informed ones? And if informed people have a more complex concept of pain what does this say about pain being datanic?