Let’s discuss Amie Thomasson’s paper A Nonreductivist Solution to Mental Causation. I’ll save my thoughts for the comment thread.
114 thoughts on “Amie Thomasson on nonreductive physicalism”
Leave a Reply
You must be logged in to post a comment.
The Skeptical Zone
"I beseech you, in the bowels of Christ, think it possible that you may be mistaken."
First, an interesting but tangential observation:
When I see the name “Amie”, particularly in a smaller font, my brain tends to separate the “m” into “rn”, giving “Arnie” instead of “Amie”. I can even make it flip back and forth like a Necker cube.
I think it’s because most Amys/Amies spell their name with a ‘y’, while most Arnies spell it with an ‘ie’. The ‘ie’ reaches over and interferes with my perception of the “m”.
Taking a look at this paper again, I’d call what she advocates for “Unrepentant Property Dualism”: I take this description of her view to entail that:
The attractions of such a view are obvious: The nonreductivism should ensure that mental properties are preserved in all of their distinctive character…
Now, admittedly, she calls her position a type of “physicalism” because it
ensure[s] that we nonetheless offer a view of mind compatible with a scientific world-view.
But, as I’ve said, the (hardy) physicalists I’ve dealt with would shun her with as much fervor as a unicorn could expect to receive if he denied the existence of Candy Mountain. Anyhow, it should be obvious that these titles are unimportant. I consider Thomasson’s solution and the resulting position very attractive, whether it’s really physicalism or not.
There are also these from her paper.
As I mentioned in my reply to Keith in the other thread, I think the important distinction is between “reduce”, which an NPR denies, and “depend”, which an NPR accepts for some sense of depend (ie supervene).
But I will stop now on that; I just wanted to make a bit more of my case about why I read her the way I do.
I think the more interesting issues are raised in her concluding sections:
1. How do we handle mental causation for physical events? What notion of causation is needed to consider us agents whose mental events cause physical events (eg is what I am thinking reflected in what I type in my posts?).
Kim has a later paper on that which I’ll summarize in a day or so.
2. Is the level vision of science needed or helpful? I won’t try to summarize L&R’s sophisticated (ie complicated) attack on it, but I will try to present Bechtel’s view, which I think reconceptualizes the issue, giving a physicalists (my sense!) a better way of looking at emergentism and downward causation, and also presenting a better picture of what scientists actually do when using concepts from different sciences. I’ll try to gather my thoughts on that as well and post over the next day or two.
I do not believe there is a useful definition of causation, a definition that would make discussions like this worthwhile.
Causation is one of those intuitive concepts that just doesn’t hold up to scrutiny.
Non-reductive Physicalism – like most Philosophical issues- is imaginary issue with no real world purpose.
P.S: I love Lawrence Krauss ( Only when is talks about philosophy 🙂 )
What is a “real world purpose”? Do anthropological studies that respond to our curiosity about some culture have a real world purpose? How about an archeological dig that informs us what kind of tools some society used 3500 years ago? Does classics studies have a real world purpose? Does art history?
Archeological digs are not useful ? – that coming from a Darwinist ? 🙂 ‘Classical Study’ is a collection of subjects – not a single subject. As for Art History, at least they can be used to put a value to the art object and paintings- it makes some business sense.
In the concluding sections of her paper, AT admits that there is more work to be done to understand causation between mental events.
Furthermore, she admits this will be particularly challenging for thoughts causing physical action:
As example of mental causing physical, consider: I wanted a beer so went to the fridge. How can my mental desire to have a beer cause the physical action of walking to the fridge? How can one be an agent whose thoughts direct voluntary physical action?
In a 2007 paper, Causation and Mental Causation, Kim looks how we might understand causation for agency.
First, he asks whether simple regularity can be the explanation of cause for agency. Regularity is often considered the basis for natural law. But is regularity acceptable for agency? No, because it could result from spurious correlations, not cause. For example, consider the shadows cast by a moving car. The shadow for the car in a later position will be correlated with the shadow from a previous position, but it is not correct to say it was caused by the previous shadow. Indeed the shadows are determined by the cars, to use the language of the AT paper, but no one would say that the first shadow caused the second. So regularity does not work for causation.
Counterfactuals provide another approach to causation. If I had not wanted a beer, I would not have gone to the fridge. In general, we can say (roughly) x causes y means if x had not occurred, then y would not have occurred. And it seems to work for the beer/fridge case. But consider these other situations: If I had not wanted to check the light, I would not have gone to the fridge. If I had not wanted to sweep the kitchen floor, I would not have gone to the fridge. Those seem true but not relevant to why I went to the fridge.
Kim points out that the usual understanding of counterfactuals uses the ideas of possible worlds. To focus on a simpler example, he asks us to consider:
The possible worlds analysis of counterfactuality asks us to find the closest possible world to ours where the match was not struck and to confirm that it does not light there. But how to determine the closest such world? Kim argues that the way we do that is to determine a possible world which has the same natural laws as ours, the only difference being that the match was not struck. But if we have to use the idea of natural laws, that is regularity, then we are faced with the issue of mistaking spurious correlation for cause, as in the moving car and its shadows. So he says the counterfactual approach to cause does not work for agency.
Finally, Kim considers productive causation. For this case, we say event c causes event e if c helps to generate or bring about event e. For example, we can think of energy flow between c and e and conservation of energy applying to the sequence c then e. Kim argues that this is the way to think about agent causation. My desire for a beer productively causes the action of me going to the fridge.
Kim admits he does not have a knockdown argument to prefer productive causation over counterfactuality, but believes this approach to causation is the best to explain how we think about causation for both for the manifest image of agency in common usage and also the scientific usage of causation for psychology and neuroscience.*
But if we insist on some physical quantity flowing or being conserved for causation, how can mental events cause physical events. Only if mental events are physical events, that is the mental reduces to the physical, at least by realization if not by identity.
Still, there still seems something to be said for having different “levels” of explanation. Philosophers talk about “folk psychology”; psychologists call this “theory of mind”. We explain peoples’ behavior by relating their desires and beliefs to their actions: I wanted a drink, I believed there was beer in the fridge, therefore I walked to the fridge. Even if these types of simple explanations disappear from a mature psychology, does that mean that all mental events terms in psychology must be eliminated and replaced by neuroscience? Do we have to give up on levels?
One problem, I think, is in the concept of distinct and separate levels of science and explanation, which AT says in her conclusion is something important to keep. But maybe the relation between sciences is more complicated than simple separation to distinct levels. Bechtel and others think that levels of science is a vague and unhelpful concept, and that mechanistic explanations relating different sciences are often used in biology, psychology, and other special sciences. I’ll share his ideas in another post.
——————————–
(*) Kim does make it clear that he understands that physics may not need the concept of causation, but he believes that “higher level” sciences rely on causation in their explanations.
I repeat my question. What is a “real world purpose”? How can I know, e.g., whether an archeological dig has one if you don’t tell me what you mean by the expression?
“Real World Purpose” is what you can use in real world – Knowledge that can be put to use. If I find bones in the dig, I can date it. A “nonreductive physicalism” is of no practical use. Why does it matter if the mind is reductive or non-reductive or if the something supervenes something else ?
You may want to take a look at this video of him versus two philosophers, skipping right to the intro by him using the video links on that page.
“This may surprise you”, he says (paraphrasing slightly), “but I think philosophy has uses for scientists, by formulating questions in fields where we have not begin to understand the key issues, like consciousness and neuroscience.”
The rest of the video is mainly he and the philosophers talking at cross-purposes, so I don’t recommend the whole thing. But if you are interested in more detail, listen to the Krauss pitch section which you can jump to.
(ETA: edited paraphrase after another listen to the video).
If you find bones in the dig, you can date it–yes. But If you find bad arguments in the paper you can publish something on it. If you find good ones, you can learn something. I take it you don’t think that latter stuff is the “real world”?
I don’t know where you hail from, but I note also that a good hunk of the U.S. Constitution and Bill of Rights–some documents that at least SEEM to have effects on the real world, came almost directly from a philosophical treatise by Locke. The Libertarian Party platform (and a follower of that doctrine may be the Repub. nominee for president in a couple of years) comes largely from a bunch of philosophers, including two recent ones, Hospers and Nozick. Marx had some “real world” effects too, I’d imagine.
Leaving political philosophy and turning to other stuff, there a bunch of other thinkers whose philosophical writings (some worked in other fields as well) would seem to have had some real world effects. A few that pop immediately to mind are: Buddha, Plato, Aristotle, Augustine, Machiavelli, Leibniz, Laplace, Hegel, Nietzsche, Chomsky, Poincare, Turing, Russell, Whitehead. The point is, it’s hard to know in advance what will have “real world” use. Human beings study stuff that’s interesting and let the chips fall where they may.
Hold on, there.
Sure Turing wrote papers that could be called philosophy (eg the Turing test one).
But he was a mathematician, not a philosopher.
I’m not sure about Chomsky, but I don’t really care much about where he ends up.
But keep these philosophical mitts off Turing!
That’s what I was thinking of. It’s been highly influential (for good or ill). But yes, I agree that he was not mainly a philosopher (unlike say, Frank Ramsey).
BTW, Bruce, not sure if I’ve mentioned it, but I was in a Chisholm seminar with Kim once, when he was visiting at Brown in the 70s.
FWIW, he also wrote a pioneering paper in developmental biology.
Played some kind of role in WW2 as well, if memory serves. Was it hospital orderly?, No, wait, I am thinking of another guy there I think.
He was a code cracker. (In the title to the movie about him, this is a double-entendre.)
Some more recent examples:
Bert Dreyfus used Heidegger and Merleau-Ponty to criticize “expert system” approaches to A.I. in the 1970s, and that had a big impact on A.I. research. Alva Noe and Anthony Chemero use Merleau-Ponty and Dewey (respectively) in their on-going research in cognitive science of perception. Cheney and Seyfarth use Dennett on the intentional stance in their work on vervet monkeys (How Monkeys See the World). Ilya Prigogine co-authored Order Out of Chaos with Isabelle Stengers, herself a well-known philosopher in France and a former student of Deleuze. It takes a while for these ideas to trickle out of philosophy and into science, art, and politics, but it does happen.
I agree that there’s a law of diminishing returns when it comes to figuring out the differences between non-reductive physicalism and property dualism in the metaphysics of mind. I worry that that kind of metaphysics teeters on the brink of what Ladyman and Ross attack as “neo-Scholasticism”. But that’s no objection to metaphysics as such. Metaphysics can be extremely helpful for guiding future empirical research when conducted in close conversation with science.
Doesn’t seem to have deterred IBM from building Jeopardy playing systems and attempting to sell medical diagnostic systems. Nor has it deterred Google from attempting to build self-driving cars, no Tesla and others from building crash avoidance systems.
At the moment, no one knows how to build anything except expert systems. Some of them can learn, but their learning is confined to refining their expertness.
petrushka,
Neither Dreyfus nor I would deny that expert systems are useful! His point, in the 1970s, was that expert systems would never be able to do what AI researchers at the time hoped that they would eventually do: be sentient, sapient beings in the exact same sense that we are.
I believe I’ve said the same thing about a dozen time here, with no comment received. I have a lot of unpopular opinions in this area, but my opinions are neither original nor rare.
I believe what I have said is that hard AI will have to evolve, and if we achieve it, we may not understand it any better than we understand brains. (we would know that it is possible, which would end a lot of philosophical debate.)
But brains seem to have started out as expert systems. Very simple reflexes and tropisms. What seems to evolve is layers of nuance.
I suppose Dreyfus would insist on the difference between a connectionist system that models a brain — e.g. by learning — and a symbol-processing A.I. that has all the rules it needs explicitly coded in, which is what AI researchers were trying in the 1970s. In any event, I only mentioned this as an example of how even the obscure and arcane texts of “Continental” philosophy have made a real-world difference.
I rather stubbornly refuse to take the bait on disputes over how the brain works.
We will know how it works when we can make one. Not necessarily a human class brain, but at least a mammal class brain.
My intuition, guided by my education in psychology, is that evolution has produce a non-obvious kind of learning machine. One with many pathways loosely wired. It’s not a blank slate and it’s not hard wired. We will know what this means when IBM is manufacturing them.
petrushka,
You have received comments on this, but apparently you’ve forgotten.
Look here, for instance.
It has been over a year, exceeding my attention span.
I agree that humans will be able to shortcut the evolution of AI. It’s still possible that we will not fully understand consciousness, if it should emerge.
The Tesla car guy says we are summoning the demon. Unintended consequences. I think there have been several movies on the subject.
What do you mean by ‘Humans will shortcut the evolution of AI’ ? Are there others helping Evolution of AI ? As far as I know there are no aliens or random unguided process involved.
No kidding….!
The hospital orderly was, in fact, Wittgenstein.
The stuff Dreyfus, Searle, etc criticized in the 70s and 80s is lovingly called Good Old Fashioned AI these days. I agree that expert systems, which were built using those ideas, are not helpful in understanding how human cognition evolved.
As you say, the GOFAIs start by assuming the hard part — abstract symbols — is already in place. To avoid this, connectionism or perceptual systems approaches to representation seem to offer a better chance of understanding how mind evolved.
These two approaches are also better connected to embodiment, which I agree must be part of the picture. But I don’t think embodied approaches can replace the need for representation, although they do help define the nature of representational implementations that will work by specifying how representation is to be used for the success of the organism in its environment.
One more comment on the AT paper:
AT suggests a layered view is needed to support the non-reductivity of causation she argues for. She quotes Kim’s description of what she means:
She goes on to add:
But as I pointed out in previous post, it will be hard for her to complete her project of explaining separate notions of causation at separate levels, especially for the mental to physical causation needed for human agency.
The mechanism approach to levels and reduction is a superior alternative to her levels approach. The following discussion is based on Bechtel’s Mental Mechanisms..
Start with her project. How can we define levels of science, to separate physics from biology from economics? Size of the entity is not enough; after all, physics with atoms and galaxies, biology with germs and ecosystems.
Simple composition also is inadequate; physics deals with the composition of atoms and biology with the composition of genotypes.
But suppose we divide the sciences into four groups: physical (physics, chemistry), biological, behavioral (eg psychology), social (eg anthropology). Now instead of simple compositions, consider modelling of the same entities from different scientific viewpoints. For example, the entities of biology are a special case of the entities modelled by physics.
This type of consideration leads to the idea of mechanism. A mechanism is a structure which performs a function defined as a whole in virtue of its parts, their properties, their organization, and their operation.
Function can only be defined in terms of the overall structure performing a function in the context where that function matters. Function cannot be reduced to the parts. Further, although the causal properties of the components matter, these only matter when the parts are organized and operating as the mechanism.
Mechanisms provide a detailed and scientically useful alternative to AT’s vague ideas of ” determination, dependence and material constitution”.
Rather than worrying about wholesale reduction between levels, mechanisms allow for interdisciplinary modelling, local reduction, for functional analysis and synthesis in specific, scientifically testing circumstances. Rather than different levels of causation, the mechanism idea links causation of the parts operating in the mechanisms with the functionality and hence causal powers of the whole operating in the context of the mechanism.
Emergence arises simply and naturally from the causal powers of the whole mechanism. No new physics is needed.
Neuro-economics provides one example of the application of the mechanism idea in scientific study of mind/brain. Here the scientists model psychological choice by mechanisms with components that are neural activity in specific areas of the brain.
To get back to the original worries of the AT paper on mental/physical, Bechtel does some initial work in his book to extend his ideas to understanding the brain/mind by looking at representation in the mind in terms of feedback mechanisms in the brain. But a much more detailed application of mechanism to brain/mind, specifically consciousness, can be found in Prinz’s The Conscious Brain.
As my last word on the subject of how the term NPR is used in philosophy, I’ll quote T. Horgan, an NPR adherent, on what the term means (as quoted by Kim from Horgan’s 1996 article on Reduction, Reductionism in The Encyclopedia of Philosophy):
I’m not saying I think that is a correct or even metaphysically defensible idea.
Only that it is what NPR means in common philosophical usage.
I’m not sure what you mean by “expert systems”. You seem to be using the term as a synonym for any non-biological approach to learning and intelligence, ie for any implementation of machine intelligence.
In fact, the technologies underlying Google drive or learning from “Big Data” are much different from those used to build most of the expert systems built before 2000.
I mean it will take less than billions of years, because we can reverse engineer a work-alike.
My other point is that we may be able to reverse engineer consciousness without understanding it.
I would define an expert system as one engineered to produce commercially useful results. I’m not asserting that there is a bright line distinction between expert systems and AI. I think commercial systems will evolve into AI.
As for Google, I can see it getting smarter. And scarier. The ads i see on the internet are increasingly targeted to me. Unfortunately for the advertisers, they tend to be for things I just bought.
BruceS,
I don’t see that that requires that size of entity be the sole determinant as you suggest in your commentary. In fact I’d take Bechtel’s position to be a spelling out of Thomasson’s, on which, incidentally, there is more info in her SEP article on categories. Furthermore, you even quote this from Bechtel:
This type of consideration leads to the idea of mechanism. A mechanism is a structure which performs a function defined as a whole in virtue of its parts, their properties, their organization, and their operation.
That seems at least as big/little focused as anything Thomasson says.
Finally, you’d like not to “worry about wholesale reduction” between levels. My take is that without such reduction (in the sense of determination/constitution) the whole question that Thomasson is attempting to address, both here and in her book, will be left unanswered. I.e., either there is a contribution of my throwing the red socks into the wash or not. If there is–what is it? If not–is it something extra that has no causal efficacy? Both of those put us back where we started. Thomasson’s solution is both complete and elegant precisely because it requires that which you and Bechtel don’t want to worry about.
I think that LionKitty has actually given a better defense of the Non-reductive-physicalism-is-not-a-species-of-property-dualism thesis than Horgan does here. But I remain unmoved by either of them.
LK says, “What matters is whether one could have just these physical properties in some other possible world but not have the mental properties that go along with them here (and perhaps not have any mental properties at all). If one thinks there IS such a possible world, one is a PD, if one does not, one is a N-RP.” I get it, and, certainly, that is one way to parse these positions, but I think it’s obvious that “physicalism” once meant–and still means to a lot of people–that every entity and property in the universe is physical or is reducible to stuff that is. The wee-wee physicalist says “Oh, no, ‘physicalism’ just means that nothing that isn’t physical is IMPORTANT to anything: these mental characterists are NECESSARILY connected in some mysterious way to a certain batch of physical stuff.” But epiphenomenalists always thought that, and nobody ever cared about THEM!
Again, this classificatory scheme you (and, apparently others) prefer just seems to me a degradation of the physicalist programme. But, of course, it makes it easier to be a physicalist and, as I’ve said, everybody who’s anybody wants to be one of those! In other words, it’s a cheat.
So, as I said, I’ll try to adapt, but I’m not going to stop snickering.
walto,
Just wanted to clarify for anybody who didn’t understand this that the stuff after my “LK says” in the above post was a paraphrase, in spite of the quotation marks. Sorry if there was any confusion about that (and I hope I got her right)!
It’s all about reduce versus depend.
I understand many physicalists philosophers are still convinced that multiple realizability breaks reduction without breaking dependence (supervenience).
I’d be happy if Betchel was just finishing the work stated by the paper, but Bechtel does not think levels is important and she does, it seems.
I skimmed the SEP categories article but did not see anything directly relevant. What did I miss?
What Bechtel adds is the key point that organization and interoperation of the components is critical, not simply constitution. For example, software analysts use the mechanism approach to analyse software, an abstract object where size does not enter the picture.
“My throwing the red sock” is not part her example: there is no mention of agency (at least in the first mention of the example, I did not re-read the whole paper).
Her example is about explaining the state of the sock now based on its physical composition now versus its history as a sock. I take that as an analogy to explaining a mental state now as a physical brain state versus its role in a history of mental states. She makes it clear that she is not covering agency in the sense of mental causing physical in her conclusion.
I am not sure what sections of the book you are referring to. In the first section, she states that her argument on baseballs versus atoms does not work for mental events; she references the paper for one that she thinks does. Is there another chapter that you think addresses the issue?
I wonder if your use of the phrase ” My take is that without such reduction (in the sense of determination/constitution)” is one source of our difference.
I think reduction is not just constitution in the sense of dependence (=supervenience), and its reverse, determination, if determination means the opposite of constitution. As I posted in the other thread, I don’t know if reduction is in general possible between sciences; it seems unlikely. But in specific cases, with a lot of hard word and use of ideas like mechanisms, scientists are able to create interfield explanations.
I also think that mental states versus brains states is a special case because it is about subjective versus objective, whereas for every other case in reduction between sciences, no subjectivity is involved. So possibly there will be a lot of progress in the reduction of specifying psychology function in terms of brain mechanism.
Switching gears: I think the most interesting question noted in her paper is the tension between mental causation and mental content. It is commonly accepted that content supervenes on more than brain state, eg as per twin earth or even externalist explanations of intentionality like Millikan’s. How does one reconcile that with the intuition that physical causes of movement/agency should be intrinsic to body/brain?
Obviously, she was not attempting to answer that question but she does say the determination approach to separating levels and causes might help. I suppose she wants to include both the brain state and the physical state of the world now and in the past as part of the determination of the mental state that causes me to open the fridge. But that seems to lump together a lot of different types of causal explanations.
Indeed, the fMRI research program claims to be doing useful science by ignoring all but the local brain state to understand psychological causes and actions.
Yet I agree that externalism about content makes sense.
So it’s a fascinating issue to me.
I don’t really understand much of what you’ve written above, Bruce. You ask, e.g., where in Ordinary Objects or in Thomasson’s SEP article does she discuss these issues of causal agency and determination that we are discussing? And I just want to say “everywhere”, that that is the whole point of this paper, and much of the point of her book and her article. In a word, she is arguing that the categories of common sense and ordinary language–what Sellars called the manifest image (where it is the putting of the sock in the wash that causes the fading) only apparently conflict with micro-physical explanations. Because causation has been confused with determination, it may seem like there’s over-determination/uselessness in one of the explanations, but she explains why that view is incorrect.
The question of whether one science is reducible to another will depend, as we’ve often discussed, on what “reducibility” means. I’m not certain, but I’d think that Thomasson’s answer would be “yes and no” (Yes, if we’re talking about complete determination, No if we need some kind of analytical reduction.)
I agree with your closing remarks–regarding the connections between externalism, constitution, and content–that those issues are confusing, complicated, important and controversial. IIRC, Pautz gets into some of that stuff. I remember struggling with a couple of his papers a year or so back. I don’t know whether Thomasson has opined on that stuff or not. But, as you say, it’s connected, and can’t really be ignored.
Her words in the 98 paper:
I read the paper as solely about Kims discussion of mental-mental versus physical-physical. That is, she discusses how there can be a causal connection between two mental events without having either overdetermination of causes and without denying the causal completeness of physics. I think her argument only works (if at all) for mental to mental causation and that is what she means by the above quote. Because she avoids downward causation, she avoids mental to physical causation.
I don’t want to retype a quote from Ordinary Objects, but the last two paragraphs of section 1.4 read to me as explicitly denying that the arguments in this book are sufficient for the mental/brain case, although she says the do provide some suggestion on how to start. She cites the 98 paper in those paragraphs as providing more on how one could proceed. But then she admits in the 98 article that mental to physical is not covered.
Of course, it may be possible to extend her arguments to mental to physical. But I don’t think it is explicit in either work.
Can you point me to something more specific in the SEP categories article, even just a subsection? There seems to be a lot of overview of how various philosophers have understood the term or developed varies means to create categories, but I’m having trouble making the connection.
I take the following to be the key point of the paper we are discussing here:
I take that argument to be basically the same argument that runs through the first chapter of _Ordinary Objects_ as well as her paper, “Metaphysical Arguments against Ordinary Objects.” I mean, there are differences, because the issue of analytic connections isn’t involved (or at least so obviously involved) in the paper on non-reductionism, presumably because the mental and physical worlds are so….different. But the moral seems to me to be basically the same–there are different levels or categories at play, and it is a kind of category mistake to claim that one causal connection blows away the other.
Agreed, but that does not help directly with thoughts causing physical action, which to me requires an argument from M to P with P then being the start of a causal sequence leading to physical movements.
Again, it is her own words that I am using at the end of chapter one of OO for saying the argument there is not enough even to do what the 98 paper does.
I could try to explain why that is in my own words. Maybe tomorrow. It actually bears on the content versus causation stuff, I think, maybe, if I understand her reasoning, …
Again, I see what she means: so long as layers remain within the physical world, where the realization is from one physical “layer” to another, we have a somewhat different creature than an argument meant to move from mental activity to physical results. My point is only that the basic nature of these arguments is the same. In both areas, she focuses on clearing up a confusion between causes at the same level and determination between levels. It is that confusion that produces claims regarding overdetermination, violations of causal closure, and the like.
The arguments are, as they say, of a piece.
I realize that I didn’t respond to this:
What Bechtel adds is the key point that organization and interoperation of the components is critical, not simply constitution. For example, software analysts use the mechanism approach to analyse software, an abstract object where size does not enter the picture.
I guess I don’t feel competent to get into that. I’m not entirely sure where Thomasson is on that subject, and I haven’t thought about it myself (supposing such activity would help). I’d like to hear her respond. Maybe I should take your approach on the other thread and pass along this link to her?
Oh, I also wanted to respond to your claim above that Thomasson doesn’t even attempt to discuss mental-physical causation in this paper, limiting herself to Kim on Mental/Mental and Physical/Physical. Actually, she writes this:
She talks about physical to mental in the first paragraph you quote but punts on mental to physical, saying in the next two that it is not covered by the paper, but non-reductionism and the multi-level view is too good to give up, despite that.
In the second and third paragraphs, she argues for upward causation of people in social sciences, which is a bit of a contradiction for the separation of levels, it seems to me (why not allow upward causation physical to mental, then?). But let that go.
As far as I can see, she does not address how one can provide a good explanation of mental causation of body movement without crossing levels in some way she has not allowed for.
If you wanted to use a parallel argument to the first paragraph to cover mental to physical, you’d have to say that that there is a mental to physical determination going on (the opposite direction to the usage in her whole paper). That is not supervenience because you have two way one to one dependence. It is closer to identity theory, which she wants to avoid.
In the end, all she is saying, I think, is that folk psychology provides people with useful explanations. Which is fine, but that is not a good enough notion of cause or explanation for science, which is what Kim is more concerned with, I believe.
By the way, Putnam’s latest blog postcovers his views on this and related issues.
As I read him, he is an NPR functionalist, NPR because the substrate can vary, hence he is still (!) a believer in multiple realizability as the basis for NPR.
Further, he calls himself a long arm functionalist because the functional states he believes in reach out to the environment to cover affordances, intentionality, and (I suspect) his current approach to qualia (he has become a qualia realist in Block’s sense of qualia, it appears)..
On another note, there are some great videos here of some the big names discussing qualia, consciousness, free will, philosophy of mind. Chalmers, Dennett, the Churchlands, Humphreys, Perebroom, Clark, Prinz, and others were there.
Scroll down at the linked page to look for the Greenland discussion videos.
I found them interesting because it is philosophers talking to philosophers, not a popular audience, and so they don’t try to keep the language non-technical. Also it is an informal, wide-ranging discussion, not simply the reading of some technical paper at a philosopher conference followed by Q&A focused on its narrow technical points.
Most of them do quite well at extemporaneous discussion. As usual, Chalmers comes off very well. He really is an intelligent, articulate person.
Thanks for the mention of the new Putnam entry. I hadn’t seen the series on Tarski either. Good stuff, as usual. (Neil might enjoy the nod to Gibson’s “affordances.”)
I wonder, though, why Putnam doesn’t just update the Wikipedia entry on Hilary Putnam himself instead of grousing that it’s incomplete and explaining how one ought to fix it in his blog.
Yes, that occurred to me too. Especially for a guy who obviously knows his way around the internet (I’m assuming he types the blog entries himself — I once had a boss in the early days of email who had his secretary print off all of his incoming emails, use scissors to cut off the printer headers at the top of messages, and then put the paper in his in box!).
Is it considered a faux pas to update your own Wikipedia entry? Especially if you do it anonymously? Although I could imagine anonymous updates leading to a situation like the one where Woody Allen pulls Marshall McLuhan out from behind a billboard in Annie Hall. That is, some Wiki editor comments on an anonymous update written by HP saying, “that is not what HP thinks!”. Then comes the big reveal.
The long arm functionalism stuff is also supposed to help with content externalism, I believe. That is, intentional states include both brain states as well as world states (eg for natural kinds like twin water) and language community states (for Burge’s stuff on externality depending on one’s linguistic community).
But as I mentioned in another post, real world science seems to be able to make progress by considering just brain states. In effect, the natural world and language community stuff is factored out of the experiment, I guess by assuming it is constant or perhaps by suitable randomization and control.
So I wonder if there is a way to mathematically or philosophically split the long arm function states into to local (brain-based) and external components which can then be studied independently to reflect that scientific approach.
I think there are some philosophical approaches hybrid approaches to mental content that may be similar to that idea — at some point, I need to try to understand them.
Maybe KN’s book will speak to related points when he talks about naturalizing intentionality.