by Thomas Pyne
Suppose that the phenomena historically associated with demonic possession can be explained as psychotic symptoms. That would not tempt us in the slightest to adopt reductionist theoretical identities like:
Suppose that the phenomena historically associated with demonic possession can be explained as psychotic symptoms. That would not tempt us in the slightest to adopt reductionist theoretical identities like:
- Belial = Psychotic Condition X
- Asmodeus = Psychotic Condition Y
Instead we would just say, “There
are no demons.” And eliminate them from
our ontology.
What makes a sort of entity A a candidate for an
eliminativist program rather than reduction via theoretical identification? Two conditions: (i) A does no work in a literal and
complete account of Everything That Is So; (ii) A has dubious
credentials. That is, we have reason to
suppose that our acceptance of A was based on some cognitive error,
or confusion. An eliminativist program
then must account for the error by which we came, mistakenly, to think that
there were A’s.
Contemporary Eliminative Physicalism has been conscientious
in its attempt to meet both conditions.
First condition: it claims that attributing mental states is not needed
in a literal and complete account of the world. Rather, the place of those
attributions will be taken, as Paul Churchland puts it, by the employment of the
conceptual scheme of a matured neuroscience. It’s not that, say, ‘believing’ will be revealed as a brain process;
it’s that there is no such thing as
believing. With the science-based
conceptual scheme we will be able to talk about what is really going on in the
brain instead. Adoption of the new
scheme in place of the old will constitute a “quantum leap in self-apprehension.”
(It will accomplish that, of course, only if our ordinary mental attributions really
don’t do any work.)
Second condition: our
traditional attribution of mental states is consequent upon a conceptual scheme
that embodies a mistaken and inadequate theory.
This conceptual scheme, “Folk Psychology,” is the same kind of error or
confusion as invoking demons to explain the voices schizophrenics hear.
The two conditions on an eliminativist program are not
independent. If it turns out that
accepting a sort of entity A is not, after all, a cognitive error or
confusion (which is required for meeting the second condition), this weakens
our grounds for thinking that A will not figure in a literal and
complete account of Everything That Is So.
This
trope of characterizing our mental concepts as ‘Folk Psychology’should be
subjected to sterner questioning than it usually is. In particular we should question the
assumption that our ordinary mental concepts form a theory. Just
on the face of it, this is an implausible piece of historical revisionism, and
I have thought so from the very first time I encountered the phrase.
In
any language with which I was familiar the common verbs for cognitive activity
are of the same antiquity, and are as much semantic ‘roots,’ as the verbs for
other common activities. Liddell &
Scott’s Greek-English Lexicon thoughtfully prints semantic roots in caps.
GEN (‘become’: the root of ordinary
Greek epistemic terms, eg. gnosis) BOL
(‘desire’ or ‘intend’), PEITHO (‘overcome,’ ‘persuade,’ or in the middle voice
‘believe’) are basic to the language as EDO (‘eat’), PNEO (‘breathe’), BDEO
(‘fart’), and BINEO (‘mate’). In Old
English ‘think,’ ‘ween,’ ‘deem’ are “four-letter” words: as ancient as ‘walk,’ ‘sleep,’ and
‘shit.’
The
best abductive explanation of this fact is that mental terms, like the other
terms, designate common ordinary human functions and actions. There is no particular distinction made
between ‘physical’ actions and functions and ‘mental’ ones. Eating, sleeping, farting, thinking, believing,
and desiring are all just Stuff People Do.
To
describe someone as ‘believing that it will rain,” or “wanting to lie down” is
not to offer some sophisticated – though mistaken – explanation of what they’re doing; it’s simply to describe it. What is
a piece of philosophical sophistication is distinguishing between the ‘mental’
and ‘physical’ in a way that makes such attributions seem conceptually
troublesome. But this philosophical
sophistication doesn’t license our reading that distinction back into our ordinary
conceptual scheme.
To
use an analogy, ‘Zeus’s Spear’ is an explanatory
concept (in ‘Folk Meteorology’) of a more basic phenomenon, lightning. ‘Lightning,’ however, is not a term of Folk
Meteorology: it does not convey an explanation of anything. It names the phenomenon to be explained. Likewise, there is no more basic phenomenon
that ‘believe’ serves to explain: it is the phenomenon. ‘Believe’ is like ‘lightning,’ not like
‘Zeus’s Spear.’
Eliminative
physicalism regarding the mental became a popular strategy in the 80’s and 90’s
again when it grew increasingly clear that reductive physicalism was never
going to work. But candidates for
eliminativist strategies are entities with dubious credentials, our belief in
which is based on confusions. Thinking,
desiring, and believing hardly come with dubious credentials. They are common human functions, among the
most obvious and humdrum features of our being in the world.
They
are, when you stop to think about it, the least likely candidates imaginable
for an eliminativist program. After all,
there are no philosophers trying to eliminate ‘shit.’
That's because it makes an indispensable contribution to a literal and complete account of Everything That Is So.
Thomas F. Pyne
Professor
Department of Philosophy
Sacramento State
That's because it makes an indispensable contribution to a literal and complete account of Everything That Is So.
Thomas F. Pyne
Professor
Department of Philosophy
Sacramento State
To my mind, (which may need elimination too), a term like "belief" is a candidate for elimination when it stops being useful at identifying something that we are interested in or that appears to be a genuine, robust phenomena that warrants a description. Beliefs, once, we start probing a bit, seem to evaporate. It turns out that whether or not someone believes something is highly nebulous. The alleged owner of the belief isn't a reliable judge. He can be easily led to report that he does or doesn't believe lots of things depending on environmental prompts, priming, and other causal factors that he is completely unaware of. In one context, he'll claim to believe it, then 10 minutes later, he'll say something else, without any sense that he's contradicted himself. He'll report that he believes "I am a confident person" and then recant, and then waffle. He'll act in ways that are manifestly the opposite of what we'd expect if he did in fact believe what he claims to. He'll confabulate explanations about why he believes this thing that have nothing to do with his real actions. He'll construct rationalizations or justifications for the belief that have nothing to do with the actual factors that led to the acclamation from him in the first place. And so on. Whatever they are, if anything, they are vastly more mercurial than philosophers pretend when we look at them in the wild. Saying that Smith believes X turns out to be a handy, shorthand way for us all to allude to a complicated set of behaviors, and it's probably indispensable, at least in the short term for us to get by when we're talking casually. Primates need to make their grunts and snorts at each other. But as far as a robust entity or event that can be identified with some empirical conditions, it's proven to be suspiciously evasive. That makes them all sound more like evil demons everyday, in my book.
ReplyDeleteI am appreciating the fact that in the list of key words/labels for Tom's post, "fart" and "shit" are on the list. Let's hope some 13 year old in Des Moines is looking up "fart" and finds this post.
MM
My Reply to Matt (TP)
Delete(I'll have to do this in sections.)
Gee, this is great: A quick, serious, smart response. I feel honored. Matt doesn’t take up my main point, the status of Folk Psychology and the consequences for Eliminative Physicalism of (I argue) a more correct view of its status. My main argument concerned what I called the ‘second condition’: whether Eliminative Physicalist’s explanation of our mistake, namely our commitment to Folk Psychology, really does show that it is a mistake. Matt does not respond to that.
Matt instead aims straight at the ‘first condition’: whether the phenomenon of belief must be part of a literal and complete account of Everything That Is So. We needn’t take it to be so, he claims, because belief attributions ‘evaporate’ upon probing. His argument for this conclusion: The owner of the belief is not a reliable judge of what he believes. He will affirm and deny the same belief under different circumstances and in response to different promptings. Therefore belief attributions are ‘nebulous.’ Therefore they will ultimately be eliminated with no loss.
I accept Matt’s evidence, as would nearly everyone else: We are indeed not reliable judges of our own beliefs. But this does precisely nothing to support the claim that belief attributions are nebulous. Rather than casting doubt on the reality of belief, this point has been a commonplace in the literature regarding belief at least since Bertrand Russell’s treatment of the topic in 1905 : “I thought your yacht was larger than it is.” “No, my yacht is not larger than it is.”
Belief attributions, including self-attributions, are not nebulous. Rather, they are exquisitely sensitive to the particular proposition by which a given belief is connected to a piece of the world, as well as to how that proposition is expressed. Beliefs are, if anything, over-precise, too fine-grained, and so subject to misidentification even by those holding them. This has been an important issue since Frege proposed his ‘reference shift’ to account for the ancient Babylonian who believed that the Morning Star was visible just before dawn, but would deny that the Evening Star was.
Pierre, a Frenchman, has a belief he expresses as “Londres est jolie.” Later in life he moves to London and learns English by interaction with his neighbors. His poor fortunes would lead him to dissent vigorously from “London is pretty.” Given the sensitivity of belief attributions to how the proposition is expressed, does Pierre believe that London is pretty or not?
A shopper is pushing a shopping cart through the supermarket when he sees a trail of sugar on the floor. He comes to believe, “Someone is making a mess.” He follows the trail around the aisle in vain, the trail growing ever thicker. Finally he stops: something has happened to him. He reaches down, and straightens up the bag of sugar in his own cart. The rigor and intricacy of John Perry’s account of the state he must be in, and what he must believe, is the despair of my students in Philosophy of Language.
Continuing my reply to Matt...
DeleteIf any subject in philosophy during the period from 1892 to the present has required – and rewarded – the highest degree of rigor and precision it is the intensional logic and semantics of propositional attitudes. It must provide the necessary basis for any scientific investigation along functionalist lines of how those mental states are physically embodied.
Just because our word ‘shit’ is a semantic root, it does not follow that the processes of digestion and peristalsis are simple. Nevertheless, they are indispensable.
Consider what happened to the shopper with the bag of sugar. He underwent a conscious mental change; and there is ‘something that it is like’ to undergo such a change.
If we’re honest with ourselves we will acknowledge similar experiences.
How else will the bag of sugar get straightened up? What happened to the shopper is indispensable too.
That’s data.
To say that belief states are just complicated sets of behavior because they function in explanations of behavior is assertion, not argument. It’s like “Those who drive fat oxen must themselves be fat.” “Oh really? Why?”
Matt’s comments comparing human language and mental operation to primate grunts and snorts drastically understates the complexity of primate interactions.
If any 13 year-old boys are trolling the web, here’s another designator:
Booger.
I agree with the blog post that folk psychology isn’t a theory. Folk psychology is merely a collection of our intuitions and our common ways of speaking about psychology; it’s not a theory of psychology. Unfortunately, the blog says, “we should question the assumption that our ordinary mental concepts form a theory,” but who is making this incorrect assumption? The blog implies between the lines that important eliminative materialists make the assumption. Do they make it? Is there any evidence that they do?
ReplyDeleteWe do need a theory of psychology in “a literal and complete account of Everything That Is So.” This complete account had better contain well justified theories, not just intuitions and common ways of speaking and other non-theoretical remarks. Eliminative materialism is an empirical program about what will happen in our future theories of psychology. Being empirical, it doesn’t involve merely the analysis of concepts, or appeals to intuition, or any of the other typically philosophical products. It is about what basic concepts we will be using to explain the mind at some future day when we have the best theory of psychology. The emphasis here is on the future--not today. Churchland agrees that we need the concepts of thinking, desiring, and believing in order to produce our best explanations today, so when the blog argues that we really do need the concepts today, it is fighting a straw man.
Eliminative materialists do attack folk psychology, though. They believe we already know that when we use folk psychological talk to explain mental phenomena there will surely be some better way of explaining those phenomena in terms of a yet-to-be-found theory of the brain. This is a theory of neuroscience. So, they say that in finding a complete theory of everything we should be aiming for a good theory of neuroscience. To defeat Churchland, you need either to argue that neuroscience isn’t very important in psychology, or you need to argue that you have a good idea of what our future best theory of neuroscience will look like.
I don’t think we can see far enough into the future to know what a best theory of neuroscience will look like, so I recommend we be agnostic about eliminative materialism.
BD
Here's my response to Brad (TP)
DeleteThanks for your challenging response, Brad.
You asked for evidence that eliminative materialists make the assumption that Folk Psychology is a theory. The particular eliminative materialist I had in mind when writing the blogpost, Paul Churchland, certainly views Folk Psychology this way:
Eliminative materialism is the thesis that our commonsense conception of psychological phenomena constitutes a radically false theory, a theory so fundamentally defective that in both the principles and the ontology of that theory will eventually be displaced, rather than smoothly reduced, by completed neuroscience.
Granted he calls this theory “common-sense psychology” in this essay. But he clarifies in the paper I cited in the blogpost, where he makes this proposal:
Consider now that possibility of learning to describe, conceive, and introspectively apprehend the teeming intricacies of our inner lives within the conceptual framework of a matured neuroscience, a neuroscience that successfully reduces, either smoothly or roughly, our common-sense folk psychology. (italics mine)
I wholeheartedly agree that a literal and complete account of Everything That Is So (or as much of it as we are capable of formulating) should be a theory – or at least that we should strive for that.
But, what evidence does the thesis that there are no mental states rest on, given the fact – which is a fact – that the most obvious features of our experience are (can we say it any more simply) experiences? Along, of course, with desires, goals, beliefs, and hopes.
We would need a powerful reason to dismiss these as fictions, not to treat them as data. But in order for them to be data, we have to countenance them as real.
So I don’t understand the ‘Straw Man’ charge: To point out the consequences of a position is not to falsify it.
There are alternatives to eliminative physicalism that do not traffic merely ‘analysis of concepts’ and ‘appeals to intuition.’
(Continuing my response to Brad...)
DeleteYou are aware, if anyone is, that the literature on propositional attitudes and demonstratives was the most empirically-driven branch of philosophy throughout the 20th century, using as it did the methods of scientific linguistics. Those methods involved: gathering data in the form of utterances judged grammatical, meaningful - or deviant; proposing theoretical structures that sorted the utterances correctly and predicted how future utterances would be sorted; and conducting experiments. Experiments take the form of utterances which met the structural requirements but might be judged deviant. Then you set them before native or competent speakers to determine how they react.
There is no substitute, when studying a conceptual ability whose principles are not prima facie evident, to operating in this fashion. The responses you get from competent informants about whether the utterance is grammatical, whether it is meaningful, whether they believe this based on what their earlier belief reports were, differences in the truth conditions of closely-resembling utterances – all such responses – are not ‘intuitions.’ Rather they are the settled judgments of competent judges in “reflective equilibrium.” (Remember, Rawls borrowed this term from linguistics.)
How could we possibly “already know that…there will surely be some better way of explaining those phenomena in terms of a yet-to-be-found theory of the brain”? Because some physicalist theory must be the truth?
A religious believer could, with equal grounds, pronounce that there must be some reason why God permits evil too.
And why neuroscience? Why should the physical realization of functional mental states be at the neural level? Just on the face of it this is implausible. After all many of the important mental states, like belief, are relational states. They’re intentional connections to pieces of the world.
A relational state can have no neural realization. You’d just be looking in the wrong place. (Indeed, I have long suspected that among the last Cartesian internalist logical privacy epigones are the eliminative physicalists.)
Philosophers of religion make a distinction between ‘irenic’ agnosticism and a more ‘polemical’ version. On that principle, I guess I’m a polemical agnostic about eliminative physicalism. I don’t believe in it because the project has just not been shown to be coherent.
Tom, in responding to me, you make some interesting comments about Churchland’s views on eliminative materialism. It looks like Churchland really did say Folk Psychology is a theory. Churchland is surely wrong about this if he means the term “theory” in the sense of there being systematic and detailed laws of folk psychology. Probably Churchland means “theory” in the looser sense of a serious guess, as when we say, “Oh, that’s just a theory.”
ReplyDeleteLet’s stick to using “theory” in the sense of a collection of systematic and detailed laws. You make another interesting remark when you attack Churchland’s position that our best explanatory theory in the future will be based on neuroscience. You do this when you say:
How could we possibly “already know that…there will surely be some better way of explaining those phenomena in terms of a yet-to-be-found theory of the brain”? Because some physicalist theory must be the truth? …And why neuroscience? Why should the physical realization of functional mental states be at the neural level? Just on the face of it this is implausible.
Yes, some physicalist theory must be the truth. Non-physicalist theories of the mind—and here Descartes is the first important philosopher—have been shown to be unfruitful. You won’t get the science grant today if your grant application says, “I want to research how thought and belief have no physical basis.”
Let’s take a closer look at the question, “Why should the physical realization of functional mental states be at the neural level?” Churchland and I are not suggesting [as I think the question is implying] that we should replace talk about thoughts and beliefs with talk about brain cells (neurons). Churchland and I are suggesting that higher level mental states depend completely on lower level brain states. This suggestion might on the face of it be implausible, but that’s because we need to look beyond the face of it and look into the progress of science. Science is full of surprises.
An analogy might help. Let’s look at folk physics instead of folk psychology. In folk physics we might say, “The ice cubes lost their coldness with they were dropped into the glass of tap water.” The history of physics shows that the concept of “coldness” should be eliminated from our good explanations. We should talk instead about the cubes getting warmer as the ice cubes and tap water move toward a common temperature. This elimination of the term “coldness” is produced by advances in physics research at the level of atoms and molecules plus our theory of kinetic energy of the motion of the constituent molecules. These days when we want to talk informally but want to improve on folk physics, we should talk about “getting warmer” and “melting” and “reaching a common temperature,” but no longer talk about “coldness.” That latter concept has been eliminated by our advances in physics.
So, it is clear how today we would respond to the question, “Why should the physical realization of coldness states be at the atomic level?” By analogy, we know how to respond to the question, “Why should the physical realization of functional mental states be at the neural level?” Churchland would say, we know that advances in neuroscience will be required to give better-informed explanations of what is going on when people have what we today call thoughts and beliefs. But whatever those advances are, and regardless of whether they do or don’t require elimination of the terms “thought” and “belief,” they won’t require us to talk about neurons. But what most probably will be required is that the resolution of the dispute be found “at the neural level.”
For a fine, short YouTube video on why physicalism must be the truth and neuroscience is so important to psychology and why eliminative materialism has many virtues despite overreaching a bit, I recommend this link to an interview with Churchland’s wife: http://www.youtube.com/watch?v=vzT0jHJdq7Q
BD
Thanks, Brad. It’s really gratifying that my post, which after all was a pretty narrow-gauge argument about verbs, has produced such thoughtful and forceful responses from colleagues. I hope this discussion is helping others think through this tangle as much as it’s helping me.
DeleteThis reply will probably be more clarification of my points than disagreement with yours – in fact, I’m not sure we do in the end disagree.
You repeat your assertion that “some physicalist theory [of the mental] must be the truth.” Again, I deny that.
‘Physicalist’ means ‘employing only the concepts and principles of physics’ or it means something too vague to anchor a philosophical discussion to. Let’s expand ‘physics’ to include chemistry, including biochemistry and cell biology.
You say that “non-physicalist theories of the mind…have been shown to be unfruitful.” Evolutionary biology is a non-physicalist theory, since it employs concepts like ‘competition’ and ‘fitness’ whose application is not fixed by physical realizers, but depends on contingent external circumstances and relations, including social relations. It’s also not time-symmetric.
So explanations of the mental deriving from evolutionary psychology are not physicalist theories either. After all, the principles of evolution apply to cellular automata and other such creatures who exist as information states, and can account in some fashion for developments in those entities. Apparently you don’t have to be physical at all to be thus describable and explainable.
I continue to press my question, “Why should the physical realization of functional mental states be at the neural level?”
Your response to this question is that you and Churchland are not suggesting that that we should replace talk about thoughts and beliefs with talk about brain cells (neurons). I just don’t see how to interpret Churchland’s proposal that we should learn “to describe, conceive, and introspectively apprehend…within the conceptual framework of a matured neuroscience” as suggesting anything less than that.
So I think it’s your own position, not his, that higher level mental states depend completely on lower level brain states.
That is, you, but not Churchland, think that there are higher level mental states.
Well, we don’t disagree about that!
What do you mean by ‘depend’? Do higher mental states theoretically reduce to brain states? Supervene on brain states? Emerge from them?
DeleteWhat do you mean by ‘completely’? As a matter of physical necessity? If so, what kind? Furthermore, is the necessity logical, so that the dependence is analyzable via definitions, possibly a posteriori scientific definitions of the states?
And how can intentional mental states depend completely on lower level brain states? Mustn’t any dependence also include a relation to the objects of those states?
Here’s another non-physicalist theory of the mental. We see a coin lying on the table as round even though the shape it occupies in our visual field is elliptical. We continue to see the coin as round even when its visual shape changes as we watch. We see it as the same size even though the angle its edges subtend in our visual field changes as we approach it. That is, we see the coin correctly, as the same constant size and shape it is.
The concepts this theory wields are the concepts ‘representational constancy,’ and ‘representational accuracy.’ That is, the concepts at work are pre-eminently psychological.
How could the psychological concepts of representational accuracy and constancy be cashed out at the neural level?
Suppose we put a subject into an fMRI and observe the brain’s activity while the subject is telling us that the shape of the coin is the same, and while their retinal stimulation changes as we move the coin around. What would we learn?
More precisely, what would we learn if we were not already employing the psychological concepts of representation, representational accuracy, etc.?
Very little. For one thing we’d have to sneak in the fact that the subject is being confronted with something in the environment. That is, we would have to import the fact that the neural state has to do with representing. This fact can have no neural constitution, after all, since it must also involve something outside the brain. Thus it seems that working on the neural level is pointless without already having brought to bear the psychological concepts that the neural level is supposed to eliminate in favor of a physicalist account.
The problem is not with the science but with the philosophy: looking at the neural level is just looking in the wrong place. Tyler Burge has been saying for years that the really important developments in the study of the mind are happening in computational theories of vision. That program, not neuroscience, should be the model.
But those studies are deliberate and promiscuous employers of non-physicalist concepts and principles of explanation.
Tom, our discussion is getting interesting. You deny my remark that “some physicalist theory [of the mental] must be the truth.” Perhaps we are using the term ‘physicalist theory’ in different senses, but maybe not. When I say some physicalist theory of the mental ‘must’ be true, I don’t mean a logical must but only a physical must, and by ‘physicalist’ theory I just mean a naturalist theory as opposed to a supernaturalist theory. Descartes’ theory of the mind counts for me as supernaturalist, and so non-physicalist. I definitely don’t mean that some day philosophers of mind should define the thought that peanut butter is better than almond butter in terms of neurons or atoms. We have to pick the level of description appropriate for the topic. I have no complaint with Tyler Burge’s remarks that the really important developments in the study of the mind are happening in computational theories of vision more so than in neuroscience. Nevertheless, those theories of vision aren’t supernaturalist.
DeleteOur best theory of mind will physicalist in the sense that there can be no change in anyone’s mind without a change in their neurons and, ultimately, in their atoms. But atoms and neurons can change in a person’s brain without their thoughts changing. I’d summarize those last two sentences by saying that the brain causes the mind. We now know enough about neuroscience to say this ‘must’ be so, and that Descartes’ theory is dead. I think Churchland would agree. If someday, neuroscience progresses to the point (and I don’t know if it ever will) where we have a practical recipe for how to string together neurons and get a functioning mind, then Descartes will stay in his coffin forever.
You say, “You (Dowden), but not Churchland, think that there are higher level mental states.” Churchland believes that are levels of description and that at the highest level we will eventually discover better concepts to use than our present concepts of ‘mind,’ ‘belief,’ and ‘intention.’ However, he doesn’t have those replacement concepts in hand, so in order to talk to the rest of us he will surely have to say that “there are higher level mental states” of some sort.
Brad,
ReplyDeleteThanks for the clarification of your arguments. I was struck by one in particular and would like to comment on it.
It’s the clarification you issue here: “When I say some physicalist theory of the mental ‘must’ be true, I don’t mean a logical must but only a physical must, and by ‘physicalist’ theory I just mean a naturalist theory as opposed to a supernaturalist theory.”
This is why it’s so important not to lump physicalism and naturalism together. Evolution by natural selection is a naturalist theory, but (as I argued in the last post) not physicalist.
Descartes’ dualism was not a supernaturalist thesis. It was a thesis in natural philosophy: the cutting-edge science of his day. His arguments were not theological but scientific: given the prevailing conception of the physical (involving only the properties of extension, impenetrability, and mechanical necessity) physicalism could not account for illusion, false belief, and error. To explain those there had to be some entities that didn’t operate mechanically – but then they couldn’t be physical.
So Descartes’ dualist theory of the mind was likewise naturalist, but not physicalist.
But anyway supernaturalism is really irrelevant to the issue. We would still have to produce exactly the same account of how a non-physical entity can be in an ‘intentional’ or ‘representational’ state as we would for a physical one. There is no physicalist (or even naturalist) account of intentionality, and the two main figures here, Dretske and Fodor, have been ruefully candid on the recalcitrance of the problem.
I do disagree, though, with your observation that “Our best theory of mind will physicalist in the sense that there can be no change in anyone’s mind without a change in their neurons and, ultimately, in their atoms.”
Mental content notoriously fails to supervene on its physical realization. Two cases:
Case (1): Swampman is a molecular twin of Donald Davidson formed spontaneously by a lightning bolt at precisely the same instant that Davidson himself is killed by another bolt. Swampman totters out of the swamp producing utterances like “I hope that what I am doing may be described in part as defending the philosophical importance of Tarski’s semantical concept of truth.” By hypothesis Swampman’s neurons are in a given state if and only if Davidson’s were. In that state Davidson would have been thinking about Tarski’s semantical concept of truth.
Is Swampman thinking of Tarski’s semantical theory of truth? Obviously not. He came into existence only moments ago and has had no opportunity to take PHIL 154 or PHIL 176. He has no acquaintance, direct or indirect, with Tarski or semantics.
Case (2): Hilary Putnam is sitting by a lake observing the water and thinking about how clear it looks. A lightning bolt strikes him, propelling him timelessly through a portal to the exactly corresponding spot on Twin Earth (where lakes are full of XYZ, not H2O). There he continues his observations and ruminations. Has his mental state changed? Has he ‘changed his mind’? Yes and yes. He’s now observing and thinking about how clear the twin-water looks. Our best theory of mind will have to accommodate these.
You observe that “Churchland believes that there are levels of description and that at the highest level we will eventually discover better concepts…However, he doesn’t have those replacement concepts in hand, so in order to talk to the rest of us, he will surely have to say that ‘there are higher level mental states’ of some sort.”
I think the fact that Churchland utters sentences which sound the same as ours is too forgiving a standard for attributing to him an acceptance of higher mental states. We could imagine psychiatrists developing a shorthand language for psychotic conditions. “There,” they might say, “That’s Asmodeus!” An exorcist would have to be way too trusting to take this as an offer to light the candles and sprinkle the holy water!
Descartes is a naturalist? Not by today’s standards. He believed in physically transcendent minds, so that puts him outside the naturalist camp and into the supernaturalist camp. I think you are assuming that supernatural = religious. The fact that Descartes was called a natural philosopher in the old days doesn’t make him a naturalist.
ReplyDeleteMatt,
ReplyDeleteI found your comments about belief interesting. You write, “beliefs, once, we start probing a bit, seem to evaporate. It turns out that whether or not someone believes something is highly nebulous.” And at the same time, as per conversation from a couple of weeks ago about confirmation bias, people tend to cling to what they believe are beliefs; especially the emotionally charged or deeply entrenched ones. Do you think a desire to cling to a “belief” (whatever that might turn out to be) somehow compels one to try to be consistent in the purported beliefs? If our own beliefs are nebulous perhaps they create nebulous webs that catch us. Emerson comes to mind “a foolish consistency is the hobgoblin of little minds.” Based on what you wrote consistency (of either belief or action stemming from the belief) would seem to be a rather fruitless preoccupation.
Thanks Tom for this excellent piece and your thoughtful replies. It has taken me most of the morning to digest this discussion, but it has been well worth it.
ReplyDeleteTom, I would like to suggest that your views stand on a couple of problematic claims which have perhaps only been touched on obliquely so far.
First, you seem to me to make a categorical distinction between explaining and describing. I accept that there is a useful distinction between explaining and describing, but I think it is an entirely contextual one. We do not have two disjoint categories of entities labeled Things We Describe and Things We Posit for Explanatory Purposes. When we engage in explanation, we (defeasibly) accept that which we are explaining as a fact, and our responsibility as explainers is to describe the accepted facts clearly. But any phenomenon we accept and describe for the purpose of explaining it in one context may itself be posited within a theory to explain other facts we accept and describe in a different context. I may explain my persistent diarrhea by reference to a spell cast by a vindictive little wizard. But I may also describe spell casting as something that wizards do, and explain why in terms of a theory in which they are invested with certain magical powers. This is why, in my view, it makes perfect sense to speak of people as operating with a 'folk' theory of the world as consisting of 3-D objects in causal relationships, persons in moral relationships, and minds in believing relationships. These help us to better explain the way we experience the world, to gain more predictive power over it than we have without them, even though we may typically feel that 3-D physical objects, persons and beliefs are the sorts of things that are just presented to us for adequate description. In reality there is nothing that is just presented and we need posit no fundamental level of description just as we need posit no ultimate level of explanation. (These, I think, are two of the most serious philosophical errors in modern epistemology: The Given and Foundationalism.)
Secondly, even if you reject what I say above, your view seems to me to suggest that the descriptive level will not satisfy your condition 2, that A has dubious credentials. When you take on the eliminativist program, you deny its claim that beliefs and other mental states have dubious explanatory credentials. I want to reject this, but even if I accepted it,explanatory credentials are just one sort of credential (which I'm sure is what you intended to acknowledge in characterizing dubiosity broadly.) We often misdescribe the world in specific cases, and we can do so generally as well. There was, I'm sure you agree, a time when the existence of wizards and spells was almost as obvious to us as the existence of hands.
My own theory, for what its worth, is that the word 'belief' is not currently a serious candidate for elimination from scientific and philosophical discourse, but for different reasons than yours. Normal people are very receptive to the idea that we are to a large extent defined by our beliefs, meaning commitments that are simply not for sale. (This is why Scientific American can title a recent article by Michael Shermer "Why we should choose science over beliefs," and know it will make sense to the majority of its readers.) As Randy Larsen seems to be pointing out, the notion of confirmation bias seems to posit the existence of beliefs in this sense and it is an interesting, and well-enough-formulated-for-now (WEFFN) scientific question why people have them. But perhaps a satisfactorily clarified notion of 'meme' will ultimately better serve our purposes here.
Randy,
DeleteAs usual, I find that I don’t disagree with you.
I can live with your distinction between description and explanation as contextual. The ones placing serious weight on the reality-determining nature of such a distinction are eliminative physicalists, not me. Their program depends on it. I don’t.
(In her book Explaining Attitudes, Lynne Rudder Baker argues that their dependence is in fact much greater than I made out.)
My evidence from verbs in the original blog post was just intended as an abductive argument that the ordinary function of verbs of cognitive activity was to describe, not explain. I also suggested, perhaps not clearly enough, that ‘cognitive’ vs ‘physical’ activity is a philosophical sophistication that should not be read back into the language or conceptual scheme of the speakers. It is striking, I think, that languages spoken by hunter/gatherers have such a rich vocabulary for reporting (what we, not them, would call) cognitive states. Homeric Greek had an ‘optative’ mood, used for reporting counterfactuals which the speaker hoped would be the case.
Again, ‘conceiving of what might be the case’ and ‘pooping’ are both just Stuff People Do.
Do we operate with a ‘folk theory of the world’? Yeah, I agree we do. But you above all would not characterize this theory as ‘radically false,’ in the sense the eliminative physicalists intend. That would presuppose a realism which they hold but which you don’t. (I hold it too. I just disagree with them about what’s real.)
Not sure I understand your second paragraph. Are you rejecting my denial that beliefs and other mental states have dubious explanatory credentials? I don’t deny that they have dubious explanatory credentials; I deny that they have explanatory credentials at all. (At least in the way eliminative physicalism requires them to). What I’m claiming is that the motives eliminative physicalists provide for their radical revisionist program by appeal to the notion of ‘folk psychology’ don’t work – and don’t work in a way that diminishes the motivation for endorsing their project.
Maybe ‘believing’ will go the way of ‘cursing.’ Actually, I’m open to that, although the disuse into which believing may fall will have a different explanation. Cursing fell into disuse because the action it expresses is spurious: the world just doesn’t work that way. Believing may fall into disuse because what we were describing as belief is some other cognitive state instead.
Thanks Tom,
ReplyDeleteIn my second paragraph I am just saying that, while beliefs, etc. obviously have dubious explanatory value, they also are at least susceptible to having dubious descriptive value. I claim this simply on the basis that it is always possible for us to wrongly describe the world.
I suppose I don't completely grasp the significance of the statement that conceiving and pooping are just Stuff We Do(o). I think 'conceiving' is a word signifying something we think we do, but we may simply be wrong about that depending on how we conceive of conceiving. Same holds for pooping. We don't poop if what we think of as pooping is the elimination of miasmatic substances, since miasmas aren't how we explain disease. I know you mean more by this than that we have a commitment to preserving an existing referent for this term, but I am not sure I know what it is.
In your last paragraph it seems to me that you simply don't want to allow that believing will go the way of cursing because believing, like cursing, doesn't exist. (I would actually say that cursing exists, but curses don't.) But clearly it is possible that beliefs don't exist. Anything we think exists may not exist. I am open to arguments that this is highly unlikely, but those are the only kind.
I may be just misinterpreting you, but I would like to think we have a disagreement here. If what you are arguing is that EM's rejection of beliefs on the basis of their limited explanatory value is flawed because they are not explanatory, I think that is wrong because they are explanatory in ordinary contexts in which folk psychological explanations are useful. (I am not specifically agreeing with EM that beliefs will be eliminated, only that they do have an explanatory function.) On the other hand, If you are claiming that because they are not explanatory but descriptive they enjoy a privileged status that somehow guarantees their existence (though not that we have a proper grasp of them at any given time) then I think that is wrong because anything we think exists might not.
I think you are trying to pin me to some foundational statement (Cogito ergo sum, or sum such). Some commitment I make which as immune to doubt or question.
DeleteBut I’m with you on “It is always possible for us to wrongly describe the world.”
This is not a thesis in dispute.
I’m not going to say, “We are certain, when we poop, that we are pooping. Poop is immune to error through misidentification. The spade turns on poop.” Likewise for believing, hoping, or intending. There can be faux-poop; just as we can think we are believing something, but not really doing so.
But again the possibility of error presupposes something we can (sometime) be right about.
Believing and pooping are common to every language I know. And they require no glossing in those languages. Every competent user understands their application, and that mastery can be demonstrated even via misapplication (self-deception or rubber dog poop). Cursing usually does require further explanation: that is, different languages and different speakers will have different understandings of how a particular malign thought or form of words can have the desired bad effect. Witchcraft, while common to many cultures, also requires explanation. Indeed, believers in casting spells will provide different accounts of what happens and how it’s done. Witchfinders in early modern Europe thought witchcraft was associated with demons. Witches, apparently, had a different self-understanding. Probably it was a very early sort of technology: you had the specialists who made flint tools, the specialists who wove houses (the original meaning of TECHNE), and the specialists who Brought Down Bad Stuff on your enemies. In short, ‘curse’ and ‘cast (weave) a spell’ are different from ‘believe’ and ‘poop.’
My main point: the presence of verbs of cognition in all the languages I know, plus their root nature in the language, presents a powerful, though defeasible, argument that the functions they designate are indispensable features of human life.
Therefore they are among the least likely candidates imaginable for elimination.
Am I claiming some metaphysical necessity for this?
It’s an argument about verbs, for heavens sake.
I guess I'm not trying to pin that view on you as much as get you to explicitly distance yourself from it. I think that's at least useful for others to hear, who might have thought they were agreeing with you when they aren't. And it really helps someone like me to read your view formulated in explicitly empirical terms.
ReplyDeleteA couple of quick points.
"The least likely candidates imaginable for elimination" may still be quite good candidates for elimination, in the long run.
I guess I don't have a good sense of what sort of probability the fact that verbs of cognition and excretion are present in all languages (and require no gloss) confers on their correctly designating some indispensable feature of life. It sounds like you are proposing an abductive argument: the best explanation of their universality is that they are getting at something real. And, if so, that at least helpfully opens the floor to competing explanations.
I’m generally skeptical of monistic theories in both epistemology and ethics. If we take experience seriously (as the pragmatists urge us), we must recognize that the world of experience is a world of diversity, change and pluralism. If we consider a word/concept like “belief” in that context Randy’s phrase “well enough formulated for now” is quite helpful. It seems to me that how well formulated the thing ought to be depends on the specifics of the situation. Reminds me a bit of Aristotle writing (in words that surely would make his mentor turn over in his grave) that we are not interested in deliberating about what virtue is but rather in becoming virtuous. Perhaps beliefs are a bit like race. I’m not sure race exists in any pure form (as Matt was pointing to). But I’m pretty sure racism exists. So there is something riding on the word.
ReplyDeleteTo clarify: Matt was speaking of believes not race.
ReplyDeleteRandy,
ReplyDeleteTaking your last point first: that you “don’t have a good sense of what sort of probability” is involved in my argument.
What if my very small sample of languages is representative? (You know German, so you are aware that ‘denk-,’ ‘ken-,’ and ‘glaub-‘ are semantic roots.)
Any human language is likely to have short, simple root verbs (or verb equivalents) for the basic human functions and actions: things they have to talk about all the time and need brief messages for. No human language has ever managed to do without ‘believe’ and ‘poop,’ but have until quite recently (as we’re measuring time here) done without ‘calculate,’ ‘legislate,’ ‘ostracize,’ ‘experiment,’ and ‘twerk.’
Here I’d like to correct something from one of my responses. I should not have agreed that these verbs ‘describe’ instead of ‘explain.’ Obviously they don’t describe their functions at all; they simply ‘designate’ them. (And I used the terms ‘designate’ and ‘name’ in the post, not ‘describe.’) This is important because descriptions have satisfaction conditions, just as explanations have truth conditions. (BTW, Randy, these are different. Right? So how can descriptions and explanations be a matter of 'context'?) Designators lack satisfaction conditions. So the ‘description/explanation’ issue is irrelevant to me, but not to the eliminative phyicalists. They think the verbs explain rather than describe. I think they do neither.
Responding to Randy's first point:
ReplyDeleteAs you correctly point out my abductive argument ‘opens the floor’ to competing explanations.
You say that “least likely candidates imaginable for elimination may still be quite good candidates for elimination, in the long run.”
I’d like to know more about that “may.”
So let me rephrase:
Eliminative physicalism is epistemically possible, given my argument.
After all, it was just a little argument about verbs.
I did not to agree to:
Eliminative physicalism is epistemically possible FULL STOP.
Regarding that proposition I have no assurance whatever of its truth – and I don’t think anyone else does either.
But nagging doubts on the point may explain the striking fact that so few eliminative physicalists are thoroughgoing. Paul Churchland rejects belief and other propositional attitudes, but countenances conscious states, for which he has a reductive account. Dennett quines qualia (his examples are sensations, as I recall, thus avoiding the issue of intentional content); but he accepts propositional attitudes, which he thinks are necessary to explain behavior.
The only prominent eliminative physicalist of which I am aware who is prepared to go full-on is Alex Rosenberg, who calls intentionality a ‘powerful illusion.’
This has has caused him to produce bound sheaves of paper covered with ink marks, lacking any semantical properties, which we are under the illusion that we understand, and he is under the (presumably self-conscious) illusion that he believes.
It’s a commonplace of critical thinking that Strong Claims Require Strong Evidence.
What is the motive for adopting a view that approaches subjective idealism – even Parmenides – in the sheer amount of the world it renders illusory?
Is it because it’s more ‘scientific’? But lots of other views of the mental are every bit as consistent with the scientific evidence. So is reductive physicalism; and so are the many forms of supervenience-based non-reductive physicalism. Many theories of the mental have developed out of the very plausible assumption that mental concepts are functional concepts, and thus silent on the question of what they’re ‘made’ out of. Even Galen Strawson’s panpsychism is consistent, since brain science has nothing to say about the presence or absence of primitive properties it doesn’t test for. The list goes on.
Thus we must consider the possibility that the truth of eliminative physicalism may not – because it cannot – be settled by any future developments in science, since any future developments in science will be as consistent with its denial.
It is and will remain a thesis in philosophy.
But as philosophy, eliminative physicalism seems to me, at any rate, an obvious instance of what postmodern philosophers call a ‘premature totalization.’
Tom, that's a lot for me to think about, thanks. You are definitely right to ask about the significance of the 'may.'
ReplyDeleteThis does not address any of your questions, specifically, but I think one thing that is missing from this discussion is the origin of our intuitions of plausibility. You allow that it is possible, but highly unlikely that intentional verbs like 'belief' fail to refer to anything real based largely on their universality and pellucidness. You seem to predict that they will be with us forever (unless our brains just lose this capacity for some reason) and you explain this by reference to the fact that they designate (not describe, sorry) something real.
I regard this as an interesting, challenging and principled perspective. But there is a different perspective from which this basic view just strikes one as highly implausible. This is the perspective according to which all observation is theory laden, and that the observations of pre-scientific humans have largely turned out to be laden by very bad theories. Science has started replacing these theories by better ones and slowly changed how human beings both conceive of and experience the world as a result.
From this perspective, it seems like folly to draw a line in the sand somewhere, to say, yes, yes, I now all that, but there is still this class of terms or observations that remain largely immune from (not merely highly resistant to) that process. Should we really confidently assert something like that with any confidence at this point, especially as science is just barely gotten underway? What is the world going to look like to us in 100 or a 1000 years when we can easily experience the thoughts of others and instantly access and correctly interpret visual models of our own brain processes?
What seems more likely from this perspective is that we will eventually come to understand intentional descriptions as extremely useful heuristics. Introspection, which of course we still don't understand, is an amazing phenomenon, but it ultimately seems to amount to a highly simplified model of brain processes that allows for executive control. Intentional descriptions may persist, just as descriptions of the sun as rising and setting persist, but they will be understood for what they are, and when we require greater precision we'll need to access our more accurate models of the world. (Not unlike the way the Newtonian notion of gravity as an attractive force is still good enough to build bridges and send rovers to Mars, even though we know there is no force of gravity at all.)
My only other point would be that I don't tend to take charges of self-referential inconsistency all that seriously in this context. This is a standard and predictable feature of discussions of conceptual change and the substitution of one explanatory framework for another. There are myriad examples of what is now standard science that were accused of being twaddle until we started eating their fruit.
Tom, that's a lot for me to think about, thanks. You are definitely right to ask about the significance of the 'may.'
ReplyDeleteThis does not address any of your questions, specifically, but I think one thing that is missing from this discussion is the origin of our intuitions of plausibility. You allow that it is possible, but highly unlikely that intentional verbs like 'belief' fail to refer to anything real based largely on their universality and pellucidness. You seem to predict that they will be with us forever (unless our brains just lose this capacity for some reason) and you explain this by reference to the fact that they designate (not describe, sorry) something real.
I regard this as an interesting, challenging and principled perspective. But there is a different perspective from which this basic view just strikes one as highly implausible. This is the perspective according to which all observation is theory laden, and that the observations of pre-scientific humans have largely turned out to be laden by very bad theories. Science has started replacing these theories by better ones and slowly changed how human beings both conceive of and experience the world as a result.
From this perspective, it seems like folly to draw a line in the sand somewhere, to say, yes, yes, I now all that, but there is still this class of terms or observations that remain largely immune from (not merely highly resistant to) that process. Should we really confidently assert something like that with any confidence at this point, especially as science is just barely gotten underway? What is the world going to look like to us in 100 or a 1000 years when we can easily experience the thoughts of others and instantly access and correctly interpret visual models of our own brain processes?
What seems more likely from this perspective is that we will eventually come to understand intentional descriptions as extremely useful heuristics. Introspection, which of course we still don't understand, is an amazing phenomenon, but it ultimately seems to amount to a highly simplified model of brain processes that allows for executive control. Intentional descriptions may persist, just as descriptions of the sun as rising and setting persist, but they will be understood for what they are, and when we require greater precision we'll need to access our more accurate models of the world. (Not unlike the way the Newtonian notion of gravity as an attractive force is still good enough to build bridges and send rovers to Mars, even though we know there is no force of gravity at all.)
My only other point would be that I don't tend to take charges of self-referential inconsistency all that seriously in this context. This is a standard and predictable feature of discussions of conceptual change and the substitution of one explanatory framework for another. There are myriad examples of what is now standard science that were accused of being twaddle until we started eating their fruit.
Just a quick response, Randy. (Busy day!)
ReplyDeleteI don't think introspection is the issue.
The real issue is the reality, or non-reality, of the phenomenon of intentionality.
Current externalist accounts of intentionality are all about separating the introspectively apparent from the real. You think you're having thoughts about water, but you're not. It's twin-water instead.
Rosenberg has thought through the consequences of the eliminativist view more clear-headedly and honestly than most, and concluded that on his assumptions it does not form part of a literal and complete account of Everything That Is So.
The fact that there are organisms in contact with items in the world, where that contact must be characterized in terms of representational accuracy - and thus the possibility of misrepresentation - is an observable feature of the world requiring no appeal to introspection. We can see it in rats.
That's intentionality.
Why should we give that up when there are other accounts of the mental equally consistent with the evidence?
Tom, I think I agree with that about the rats. But rats can't represent their own brain states to themselves like we can. That's what introspection allows us to do, and to the extent that it is partly causally responsible for our accepting beliefs as veridical, I think the reliability of introspection for this purpose is relevant. The same point would hold for our visual system and our views about color, I think.
DeleteSorry, I should have said beliefs as real, not veridical. Kohlkopf.
DeleteOne other thought...
ReplyDeleteAt this early point in the history of our encounter with the world (I agree with you on this), it does not seem to me good practice to conclude that such salient and important features as intentionality are fictional or illusory.
Isn't it better to take the rich literature on intensional logic, propositional attitudes and try to build our theories of the mental taking it seriously?
If in the end we can come to no good functionalist account of intentionality then eliminativist physicalism will always be there.
I think you should place the blame for drawing lines in the sand on the other side of this dispute.
Tom, I like this point. My view is that all of this should be done and more. We need robust intelligent theorizing on all ends of the spectrum. I wouldn't want you to stop and I wouldn't want Churchland to stop either. I'm sure you agree. Your respect for the smartest of these people is palpable.You occasionally get frustrated, but you keep coming back with the attitude that there MUST be something you are missing, which is how philosophy is supposed to be done. And they (we?) should do the same.
Delete