Friday, September 28, 2007

Is non-reductive materialism an oxymoron

BDK: Epiphenomenalism is a special case of reduction. Just because X is epiphenomenal, doesn't mean it isn't reducible to lower-level bits (e.g. your shadow is epiphenomenal wrt your movement along the street, but it's still reducible).

VR: If so, doesn't it follow that the phrase "non-reductive materialism" is an oxymoron in a class with jumbo shrimp and compassionate conservatism? If that's what you think I'm inclined to agree.



At 9/28/2007 10:49:00 PM , Blogger Blue Devil Knight said...

I'm not sure how what you say would follow from what I said. For one, eliminative materalists are non-reductive materialists, and while they may be wrong, there is no oxymoron. I don't understand your inference there.

I'm not even sure what you are saying. Are you saying non-reductive materialism isn't really an oxymoron because 'jumbo shrimp' is perfectly legitimate, and only a superficial reading of the word juxtapositions suggests it is an oxymoron? (I consider 'reductive elimination' as a true oxymoron, as we went over a few posts ago).

I was technically incorrect in the bit you quoted, at any rate, for another reason. The question of epiphenomenalism is orthogonal to the question of reduction/elimination. That is epi-p is not always a special case of reduction, which is what I sloppily suggested: I should have said it can be epiphenomenal and reducible. You can have an epiphenomenon that is reducible (e.g., a shadow). You can also have a nonreducible epiphenomenon (e.g., externalists about propositional contents tend to think contents are not reducible to brain states, and these externally fixed contents are epiphenomenal (me and my twin on twin earth have different contents, but we have the exact same behavior and brain states: hence the content doesn't affect behavior or brain states)).

So, basically, I think you are wrong to say that epi-p, reduct, and eliminativism are exhaustive of materialism. There are materialists who are none of the three! Dretske, for example. Searle, under fairly natural interpretations.

Perhaps you could argue that they are holding an incoherent position. But to make that argument, we'd need a lot more in favor of the claim that reduct., epi., and eliminativis are exhaustive of the naturalist's options.

At 10/01/2007 05:12:00 AM , Blogger Jason Pratt said...

Well!--I go away for a few months, and when I come back I find someone saying that eliminative materialists are non-reductive materialists. While I don't know if I would say either of those in themselves is an oxymoron, the position as stated (where one identified as identical with some class of the other) sure looks like one.

Had you said eliminatives are reducts, I'd have had no problem. But--if eliminative means there are no levels or phases of behavior to trace back through, then I can see why it would count as a special class of _non_-reduct.

In any case, if Victor is "wrong to say that epi-p, reduct, and eliminativism are exhaustive of materialism", I'm not sure what option you presented here as an alternative. (The "So, basically" tends to imply you think you've given a distinct fourth option.) It can't be non-reduct, because you've classified that together with eliminativism. You corrected yourself about epi-p being always a special class of reduction, but the correction involves it being sometimes non-reduct, and sometimes reduct.

Maybe you meant that Victor was wrongly identifying epi-p as a type of reduct and so in a list of three Victor was forgetting non-reduct? i.e. the list of three actually _would_ be exhaustive in your reading, but because elim is a type of non-reduct and non-reduct needs to be in the list?


At 10/01/2007 06:51:00 AM , Blogger Blue Devil Knight said...

JP: when I said that EMers are nonreductive materialists, I just meant that they are not reductionists (because for them propositional attitudes will be eliminated, not reduced), and they are materialists. However, I should have been more careful, as 'nonreductive materialism' is usually reserved for people who believe X exists but is not reducible. That was why I mentioned Dretske in the original comment. He believes in propositional contents, thinks they are not reducible, not epiphenomenal, but is a materialist. If you ignore the first para of my comment, my comment reads much better :)

Note I said that eliminativists are nonreductive materalists, not that all nonreductive materalists are eliminativists. I probably shouldn't have used that language anyway as it does go against how people talk.

Also, the important point was I had no idea how Victor got from my claims about epiphenomenalism to what he said, even if we use the more typical understanding of nonreductive physicalism about mind.

At 10/01/2007 07:39:00 AM , Blogger Rino said...


I think you would find your last comment is correct: an epiphenomenalist believes the mental is irreducible due to various differences between the mental and the physical (ie, normativity, rationality, intentionality, multiple realizability).

Ironically, a reductionist usually ends up having to admit that the mental exists. Ie, since the tool of ontological reduction is usually identity, and identity is symmetrical, so if the mental is physical that means the physical is mental as well, and the mental must be added to one's ontology - a most unwelcome guest for those who don't believe there is order to the universe.

At 10/01/2007 10:59:00 AM , Blogger Blue Devil Knight said...


Shadows are reducible and ephiphenomenal, so I don't understand why you think an epiphenomenon isn't reducible. That seems plainly false. That was the main point I was trying to make (I just made it too strong, as the fact is they are orthogonal issues).

Reductionists wrt mind all think the mental exists, as they are not eliminativists. Victor has a post about this a couple of weeks ago where we hashed this out.

The claim about identity is more interesting in a clever-word-play kind of way. Reduction is not symmetric: thermodynamics reduces to statistical mechanics, but not vice-versa. Within the context of this reduction, we find correspondences, such as that between temperature (in thermodynamics) and mean kinetic energy (in stat mech). That means we think that temperature exists, even though it is (roughly) identical to mean kinetic energy. But that doesn't mean we think that temperature exists separately from kinetic energy, or is a fundamental part of the universe.

Similar for mind: if mind ends up being reducible to some physical property, then it still exists, but isn't a fundamental feature of the universe but a derivable feature.

In my defense, I wrote my above comment right before I wrote this post, perhaps the least coherent thing I've ever penned.

At 10/01/2007 05:14:00 PM , Blogger Victor Reppert said...

I was looking at your comments about shadows and was thinking about what is sufficient for a reduction. And I thought "Gosh, if what is fundamentally real is physical there has to be some connection that makes the mental physical in the final analysis, and why wouldn't whatever that is be a reduction?

There are arguments, coming largely from Jaegwon Kim, which say that non-reductive materialism cannot avoid epiphenomenalism. I'll provide a JSTOR link at least for that one.

At 10/01/2007 06:59:00 PM , Blogger Rino said...


Thanks for the comments. A reason people choose epiphenomenalism is because they want to keep the mental separate from the physical. That is why I think you would be hard-pressed to find someone who chooses epiphenomenalism, but at the same time chooses reduction, since reduction takes away the separate status of the mental. What reason would there be to adhere to the troubling doctrine of epiphenomenalism if they are going to give away the only benefit of epiphenomenalism in the next step?

As for the symmetrical nature of identity: I mentioned ontological reduction via identity, you responded by discussing inter-theoretic reduction. This brings about a confusion. Sure, the theoretical framework of thermodynamics can be translated into the language of statistical mechanics, but this has said nothing about the world yet. If you want to say that both really exist in the world, then you either have to say there are two universes, or that they are ontologically identical. This means that both would be added to one's ontology, and out in the world both actually exist. Anyway, this is a complicated topic, hope I've clarified myself a little.

At 10/02/2007 06:59:00 AM , Blogger Jason Pratt said...

BDK: {{when I said that EMers are nonreductive materialists, I just meant that they are not reductionists (because for them propositional attitudes will be eliminated, not reduced), and they are materialists.}}

Nice to talk with you again, btw.

Okay, now we're getting into the reason for why I've always considered reductionists and eliminativists to be vs. non-reductionists: because at the end of the day, unless the materialist is a naturalistic theist (i.e. a pantheist of some sort), those propositional attitudes are in fact going to be eliminated in a reduction. This is why I've always understood the distinction between reductionists and eliminativists per se to be one of stage-stepping, not of final result.

The reductionist (so I've tended to find) believes propositional attitudes are real but reduce down to non-propositional explanations at bottom. (Contra Rhino, I don't typically find reductionists ironically having to admit propositional attitudes really exist as such. That's part of their whole point.)

The eliminativist, seeing the problematic contradiction of claiming that intentive actions are-and-are-only non-intentive behaviors (i.e. only automatic reactions), solves the problem by eliminating the intentive actions as a proposal in the theory at all.

The non-reductivist, seeing the contradiction of the reductivist but also the contradiction of the eliminativist (at least insofar as the claims of the eliminativist in regard to her own theory are concerned!--this is where Rhino's note about ironic admission would have to come in), solves the problem by not reducing actions down to mere reactions and claiming instead that it's impossible to do so.

But then the non-reduc takes flak from the other two, because insofar as all three are supposed to be atheistic (including atheistic materialists, though technically the same problem would arrive if they were supernaturalistic atheists instead) those actions have either got to be reducible down to mere reactions after all or else denied to exist at all.

As you noted, some epi-ps go the non-reduct route (Rhino agrees), some go the reduct route. (If it comes to that, I get the impression that some eliminativists are basically epi-ps, too! {g}) So that theory group is irrelevant for categorization, though epi-ps can have debates among themselves as to which route is more legitimate for epi-p.

Thus, the distinctions arrive due to a trilemma over what to do with intentional actions or propositional attitudes in an _atheistic_ theory or reality. (The AfR proponent solves the problem by concluding that not-atheism, and maybe supernaturalistic not-atheism too, should be considered to be true. {s})

That being said, I can also see how eliminativists might be considered non-reducts, because in fact they do no reducing in their theory: they don't have to, because they deny there's anything that has to be reduced. But this doesn't solve the propositional trilemma, and it's the propositional trilemma Victor is talking about (perhaps not quite in the way that I've put it) when discussing the topic in relation to the AfR.

{{Note I said that eliminativists are nonreductive materalists, not that all nonreductive materalists are eliminativists.}}

True; and I wasn't critting you on that.

Also, you still don't seem to have answered the question about how Victor is supposed to be wrong in saying that epi-p, reduct and elim are exhaustive of materialism. Was it because Victor forgot to include non-reduction as a proposal option?


At 10/02/2007 09:34:00 AM , Blogger Blue Devil Knight said...

Jason: see the discussion a couple of weeks ago: reduction is not elimination. That discussion should let us avoid much more on that topic here. That's a confused way to put things.

I criticised Victor because there are naturalists who are none of the three, as I said (and as I discuss more in the next post of Victor's).

At 10/02/2007 12:26:00 PM , Blogger Rino said...


I agree with you that the reductionist says that the mental exists. My contention from several weeks ago was that the reductionist violates the law of non-contradiction when he asserts that the mind both does and does not exist. To use you as an example:

You have stated the mental exists: "Reductionists wrt mind all think the mental exists, as they are not eliminativists" -BDK

You have also stated that the reduced property does not exist: "But that doesn't mean we think that temperature exists separately from kinetic energy, or is a fundamental part of the universe."

This is the problem that reductionists have. The mental both exists and doesn't exist. And, to make matters worse, they pick and choose when the mental exists based on whatever is convenient at the moment. I think the reductionist owes an explanation here: Does the mental exist or not? This is a yes/no.


At 10/02/2007 01:02:00 PM , Blogger Jason Pratt said...

{{Jason: see the discussion a couple of weeks ago: reduction is not elimination.}}

Insofar as a description of theory process, I agree. (And said I could see how it wouldn't be.)

Unless the reduction is eventually arriving (at least in principle) at X=not-X, though, then we land on a fundamentally mental characteristic of reality that cannot in principle be explained by non-intentional operations. There isn't much way to avoid theism from that conclusion, and supernaturalistic theism to boot (insofar as Nature is agreed to be non-intentive in its operations); which is why reductivists and eliminativists disagree with non-reductionists of the sort you yourself agreed existed.

Eliminativists and non-reductivists each in turn (and each in their own admittedly non-reductive but differently non-reductive ways) agree with each other over against the reductive contention that at bottom X really exists but really is not-X instead of X.

Non-reductivists (those who aren't eliminativists) and reductivists meanwhile agree with each other over-against the claims of eliminativists, that the X of intentional agency has to be considered real and that the eliminativists (who would eliminate it as a factor, not reduce down to it by theory process) must themselves be putting that agreement of X's distinct existence as intentional agency into play.

But, I'll go back and look at the discussion from a couple of weeks ago, too. {s} And I'll check your reply to Victor's new post, to see if you gave a clearly distinct fourth option there, too.


At 10/13/2009 01:51:00 AM , Anonymous Mark said...

It seems that the cheif issue is what it means to reduce.

as I understand it, to 'Reduce' something is to explain its existance and all its properties in terms of other thing/s.

e.g. Water (and its properties) is nothing but lots of Hydrogen bonded H2O molecules (and their properties) in a liquid state.

I think its also possible to produce a 'reduction' for something in terms of things that do not exist. eg. Gravity is nothing but the influence of the flying spagetti monster.

however it seems that Naturalism and particularly Physicalism is devoted to a Realism, whereby the only things that are Real are the things that could be described by the sciences.

now if i define a 'total reduction' as a reduction that is totally in terms of things that are Real, then Reductive Physicalism seems to say that any 'total reduction' must be in terms of physics.
because any 'total reduction' belongs to physics, then any other true reduction must be in terms of things that are unreal.

for instance: a Physicalist can happily talk about "Panadol helps to take away sickness" however such a thing is an unreal statement, for 'sickness' is an object that is not itself described by physics and therefore is not itself Real.

in this way it seems that if a Naturalist wishes to reduce the mental to the physical, and believes that the only things that are real are physical things, then he/she must relegate the mental to the unreal.

Note that it is still quite valid for such a person to talk about the mental existing (just as it is valid for them to talk about 'sickness' existing) but not as if they were things worthy of direct reality itself.

it seems to me as if the Non-Reductive Physicalist believes that the only things that are Real are the things of Physics, but some (namely mental) properties cannot be reduced in terms of Physics.
the Reductive Physicalist seems to further believe that there does exist a reduction for the mental in terms of physics.
and it seems as if the Eliminative Materialist further posits that we should never talk of the remaining unreal mentalities in neuroscience.

in my analysis i am led to believe that 'non-reductive physicalist' is synonimous with 'property dualist' - both believe in irreducible properties that can be described as 'mental'.

in this way i can see ephiphenominalism (the thesis that the mental does not affect the physical, but the physical the mental) is compatable with both non-reductive and reductive physicalism.

hope this helps.


Post a Comment

Subscribe to Post Comments [Atom]

<< Home