Philip Goff --- The Cartesian Argument Against Physicalism

[New Waves in Philosophy of Mind]

The Cartesian Argument Against Physicalism

Philip Goff

In my undergraduate lectures, Descartes’ arguments against materialism were presented as objects for target practice rather than serious evaluation. At the time it seemed to me that there was more to the arguments than they were being given credit for. I now think Descartes’ Meditations provides us with the resources for a sound argument against standard contemporary forms of physicalism. In what follows I shall present this argument.

A PDF of the paper is ready to view and download in the attachment below. 

A direct link to the PDF: http://goo.gl/K7cCb

[Ian Phillips]

Hi Philip

Thanks for a super stimulating paper. A couple of quick questions:

1. Why shouldn't the a priori functionalist simply deny that doubting that our mental life is realised in any 'first order' states is coherent? Why can't they say that whilst it's true that we can doubt that we have a body, or any physical nature, we can't coherently doubt that there is *any* way in which our mental life is realised? In other words they should agree that the argument shows that we are a priori ignorant of our specific nature, but insist that we can know a priori that we must have some nature.

I take it that you reply to something like this worry in the a posteriori section as follows: "However, it is not plausible to suppose that there is a conceptual connection between mental and causal role properties which it is too subtle to be noticed without some incredibly sophisticated reflection. We are not dealing with complicated mathematics here. Rather the analytic functionalist proposal is that causal role properties constitute the basic a priori content of mental concepts, that to suppose that someone is in pain just is to suppose that someone has an inner state that plays the pain role. If this were true, then at the end of the second meditation Descartes would be contradicting himself in the most perverse and straightforward way. This is simply not plausible."

I wondered whether this was a little too quick. If it was true, wouldn't it be hard to understand *any* disagreement about whether analytical functionalism was true? If true, those who deny it must be conceptually confused in a really basic way. Won't the analytical functionalist more likely say that we are all in a position to figure out the incoherence but fail to, no doubt because the a priori isn't the same as the obvious? They might also appeal to the idea that the relevant knowledge is tacit, revealed in our practices of attribution, not explicit. I guess I'm just pushing Arnauld's point a bit.

2: Am I right in thinking that in your view nothing in Decartes would tell against a view which held that (necessarily) fixing the physical facts fixes the mental facts (which I take it is consistent with there being worlds in which there are mental facts but not physical facts)?

3. In your final remark you suggest that Descartes argument targets cognitive states directly, thus avoiding the issue of whether there is something it is like to be in them. But suppose someone said that the fact that Descartes argument could be run on a state showed that there was something it was like to be in the relevant state. After all, doesn't the argument really only run on conscious judgements, imaginings etc. not on non-episodic states like beliefs and desires etc.?

Cheers -- and thanks again for a really thought provoking paper.

-- Ian

[Declan Smithies]

Hi Philip - thanks, I very much enjoyed the paper. Can I draw you out a bit on the argument against analytic functionalism? The key claim is that I can coherently conceive of myself as a pure and lonely thinker in a world in which nothing exists except me and my experiences. But how exactly does this pose a problem for analytic functionalism?

 

As you explain, Lewis thinks it’s a priori that something is in pain if and only if it has a state that plays the pain role for some appropriate population. I see the problem here – assuming that populations must be larger than one and must exist within a single possible world – but then why not just abandon the appeal to populations?

Perhaps the problem is that the pain role includes dispositions to engage in certain kinds of bodily behavior, but since a lonely and disembodied thinker does not have any bodily dispositions, it’s not a priori that something is in pain if and only if it has a state that plays the pain role. I can think of two responses here.

First, you might say that the lonely and disembodied thinker does have dispositions to engage in certain kinds of bodily behavior, although those dispositions are not manifested because the manifestation conditions for the disposition include having a body. But if the disembodied thinker were to be embodied, then it would manifest the relevant bodily dispositions. (I think David Armstrong says something like this about the envatted brain.)

Alternatively, if the first option is not compelling, you might simply deny that the role that is a priori associated with the concept of pain includes dispositions to engage in bodily behavior. Instead, perhaps it includes various mental dispositions, such as the disposition to believe that one is in pain, to intensely dislike it, to try to get rid of it given appropriate background beliefs and desires, and so on. In order to refute this kind of analytic functionalism, it seems that you need to appeal to the conceivability of a pure and lonely thinker who is also mad in Lewis’s sense, i.e. its experiences play highly eccentric causal roles. But it’s questionable whether such cases are genuinely conceivable and certainly analytic functionalists will deny this.

Perhaps the problem lies elsewhere, since what you actually say is this: “For the straightforward analytic functionalist, it is a priori that something has a given mental state if and only if it has the higher-order state of having some other state that plays the relevant causal role. However, a pure and lonely thinker has no states other than the mental states themselves: its mental states are not realised in anything more fundamental.”

I’m just wondering why the analytic functionalist can’t say this: it’s necessary and a priori that if one is in pain, then one has some state that plays the pain role – namely, pain. But it’s contingent and a posteriori whether this state is realized in anything more fundamental at all, such as matter or ectoplasm. This would be to grant the conceivability and the possibility of the pure and lonely thinker.

[Kevin Ryan]

Hi Philip,

 

Thank you for your paper.  I have a few thoughts and a question for you about a posteriori physicalism and your claim that "Once Descartes has guided me to a conception of myself as a pure and lonely thinker, it is evident that… I am having a rich and substantive conception of my nature, from which I can infer that the semantic externalist model of mental concepts favored by most contemporary physicalists is false".

From a broadly a posteriori perspective, the move from the conclusion of meditation two to the idea that we gain any rich concepts may be disputed.  Examples abound in the scientific and philosophical literature concerning the fallible nature of introspection.   Peter Carruthers, for instance, has suggested that our best scientific evidence points to an “interpretive sensory-access” (ISA) theory in place of any direct introspective access.  On the ISA account, self-knowledge is gained in a manner akin to how we gain knowledge about the mental states and emotions of others.  A defender of this theory would most likely suggest that there is a noticeable dearth of quality and consistency to introspection, so much so that self-access may not provide a suitable idea of my current emotional state let alone be the basis for understanding a "rich and substantive" account of my nature. 

As a result of the above, on your account, what do you think of current scientific evidence and philosophical accounts against the reliability of introspection and Descartes' introspective method?  Moreover, assuming introspection is fallible and something like Carruthers' ISA is true (at least for the sake of argument if you're not convinced by his account), can the results of the second meditation nevertheless be used to disprove an external semantic theory of language?

[tom_mcclelland]

Hi Philip, thanks for a really interesting paper.

My first reaction is in line with Declan’s final comment above: “I’m just wondering why the analytic functionalist can’t say this: it’s necessary and a priori that if one is in pain, then one has some state that plays the pain role – namely, pain. But it’s contingent and a posteriori whether this state is realized in anything more fundamental at all, such as matter or ectoplasm. This would be to grant the conceivability and the possibility of the pure and lonely thinker.” I take it that this would mean that functionalism is not damaged by Descartes’ argument, and further that the possibility of a disembodied minds need not entail that minds are distinct from bodies in our world.

This line of thought seems to raise two key questions: i) does it ever make sense to say that a role is instantiated without that role being occupied by something more fundamental and ii) if so, does it make sense to say specifically that a mental role is instantiated without that role being occupied by something more fundamental (whether it be physical or ectoplasmic)?

I think the answer to ‘i’ must be “Yes”. It’s plausible that science defines entities in terms of the causal role they play. Sometimes these roles turn out to occupied by entities at a more fundamental level, but it’s credible that there must eventually be a bottom level of entities that perform a role, but are such that this role is not realised by anything more fundamental. The interesting thing about this is that you can’t tell just by looking at your concept of a kind of physical entity whether it’s fundamental. For instance, it’s an a posteriori question whether electrons are basic entities, or whether the electron-role is performed by something more fundamental (strings?).

This encourages a “Yes” to ‘ii’ as well. Just as it is conceivable/possible for electrons to be fundamental and for them to be realised by something more basic than electrons themselves, so it is conceivable/possible for pains to be fundamental and for them to be realised by something other than pains themselves. That said, concepts like ‘head of state’ seem to call out for realisation by something more fundamental. Perhaps you could argue that, according to functionalism, mental properties are the kind of property that have to be realised by something more fundamental in order to be instantiated. If this were the case, then thought experiments in which pain exists without any such lower-level realisation would still count against the functionalist.  

[Philip Goff]

Thanks for great comments!!!

Ian:

1. I’m not sure you’ve connected up the bits of the paper in the way I intended. With regards to the first point, when we reach the end of the second meditation we have doubted the existence of *anything* other than our conscious states, ergo we have doubted the existence of some nature that realises our conscious states. That doesn't mean we are conceiving as ourselves as having no nature; it means that we are conceiving ourselves as having a purely mental nature.

 

With regards to the second point, I suppose I do think analytic functionalism is obviously false, and think philosophers of the future will look back and find it bizarre than serious minded people took it seriously. I think we are in a peculiar period of history in which we are blown away by the success of science, and consequently think that giving natural science due respect requires taking it to be what its creators (Galileo, Descartes, Newton) never took it to be: a complete description of reality. 

2. I don’t here give a conclusive argument that the mental doesn't supervene on the physical, but I try to argue against the specific contemporary ways of making sense of this thesis, i.e. by identifying conscious states with neurophysiological or functional states.

3. I would be happy to draw this implication, but just wanted to make the point that even if this implication were resisted we would still have an argument for the non-physicality of cognitive states.

Declan:

I think we have to clearly distinguish (i) classical analytic functionalist (pain is analytically the property of having another state that plays the pain role), (ii) Australian analytic functionalism (pain is analytically whatever plays the pain role in the relevant population), (iii) some other form of analytic functionalism. I take it you agree that the second meditation provides resources to dismiss the first two views, which is already a significant conclusion. In this paper I’m only trying to refute standard and popular forms of physicalism, rather than physicalism as such.

But as regards the versions of (iii) your propose (I've seen a talk by Kati Farkas proposing something similar in response to similar arguments of mine), I think it’s pretty plausible that at the end of Cartesian doubt I’m conceiving of a world in which there is no causation (after all, surely epiphenomenalist is coherent?), which I take it would be inconsistent with any form of analytic functionalism.

Kevin:

I wonder whether focusing on the reliability of our introspective reports is a red herring. We can ask quite generally: suppose I know that someone is ‘suffering terribly because of the inevitable destruction of her reputation, and regretting lacking the discretion not to trust someone without good reason’. Don’t you think it’s implausible that the concept expressed by the description in quotes is a ‘blind pointer(s)’, i.e. refers to a state without revealing anything about its nature? That’s all I mean to claim.

Tom:

Interesting stuff. My intuitions are that functional/dispositional concepts are like ‘head of state’ in calling out for realisation, and hence I don’t find the idea of a ‘pure power’ world intelligible (there must be more to what something *is* than how it’s disposed to behave!), but the trouble is that it’s difficult to demonstrate this (which doesn't mean it isn't the case!).

But my argument doesn't rely on this. As I point out to Declan above, it still has the resources to refute standard and popular forms of analytic functionalism. And I think there’s considerable plausibility to the idea that at the end of Cartesian doubt I am conceiving of my mental states lacking causal powers, which would be inconsistent with any form of analytic functionalism.

[Ted Parent]

Hi Philip,
Thanks again for the paper.

I wonder how you might react to the following: Analytic functionalism does not claim that functional properties are apriori knowable from mental properties and vice-versa. Rather, it claims only that mental properties are apriori knowable from functional properties. Think of the Laplacean demon: It infers facts about the higher level from the lower level...not vice versa. (I might note that you seem to characterize things in this latter way at the outset of your paper. But when it comes to the central argument, it seems you're concerned to show that the opposite direction fails: that the functional is not apriori knowable from the mental.)

A follow up: Suppose you say that you're just concerned with an analytic functionalism where the apriori knowability go both ways. Then, there is a further issue. It is important that the analytic functionalist starts with knowledge of *all* fundamental physical facts before claiming to infer apriori anything further. But in the second meditation, we don't have all that physical knowledge. So if Descartes cannot know apriori the functional facts, why should that be notable?

[Ted Parent]

Just to clarify the last point: I am adopting the assumption, at the beginning of the paper, that analytic functionalism is a species of apriori physicalism. I'm not entirely sure of that, but the point above requires that as background.

[Kevin Ryan]

Hi Philip,

While some, or perhaps even most, reads of the reliability issue are likely red herrings (and in rereading my previous post I think I'm guilty of pointing to one there), I still believe a more focused application is not.  Specifically, I see the reliability claim as putting a lot of stress on the "clear and distinct" clause of Descartes argument; even if introspection refers to (some) transparent concepts, is the "nature" it reveals strong enough to ground a modal argument?  You mention the importance of justifying possible modal ramifications - and your analysis was not only clear but definitely helped me to understand the stakes better than I did before - nevertheless, it seems to me that the issue is not about the transparency/opaqueness but the justification we can draw from using its results.  More importantly, I think the fallibility of introspective access plays is an integral part in responding to your burden of proof charge towards the end of section II. 

 

Also, in regards to your response to Ian, can our Cartesian doubter ever doubt the existence of the evil demon or being in a dream, i.e. that which is the cause of having doubts in the first place?  If not, it seems that the role played by the demon/dream would parallel the world on content externalist accounts, thus fixing referents to something outside our conscious states.  Then again, perhaps not.

[Philip Goff]

Ted

I agree that the *definition* of a priori physicalism is the one you give, and that a priori physicalism so defined is not obviously inconsistent with anything in the Meditations. However, the a priori physicalist is obliged to give some account of mental concepts from which it follows that the mental facts are a priori entailed by the physical facts. And this inevitably involves giving some kind of causal analysis of mental concepts. It is this analysis that is inconsistent with the state we end up conceiving of at the end of Meditation II.  

Kevin: Will get back to you tomorrow.

[Philip Goff]

Kevin:

Note that the argument works against analytic functionalism and standard forms of a posteriori physicalism even without moving from conceivability to possibility. Standard forms of a posteriori physicalism give some kind of semantic externalist account of the reference of phenomenal concepts, according to which their reference is determined by causal or semantic facts. These accounts seem to straightforwardly imply that our phenomenal concepts are opaque. Therefore, if I can show that phenomenal concepts are not opaque, I can refute these views.

The general argument against physicalism does require a move from conceivability to possibility. I think that metaphysical possibility just is conceivability under transparent concepts. The advantage of this view is that we have a non-mysterious account of what metaphysical possibility is, and a clear account of how we know about metaphysical possibility. On the assumption that this view is correct, we can move from conceivability under transparent concepts to possibility. Of course this is not a knock down argument, but it my move from conceivability to possibility is routed in an attractive account of possibility, which has no obvious exceptions. 

I think we can doubt that there is a cause of our experiences. It's conceivable that I am the only thing that exists. Unless we accept the principle of sufficient reason, some things have to exist unexplained. Why not me?

[Richard Brown]

Hi Philip, and everyone, nice discussion going on in here!

I want to follow up on some of the things that Ian and Declan brought up. Is it the case that you (Philip) think that 'causal power' must pick out something physical? I mean, when I am engaging in this Cartesian doubt I still think of myself as having (the experience as of) a body, thoughts, sensations, etc. I still experience pain distracting me from formulating the argument, and I still experience it as being produced by experiencing my foot being stepped on, etc. Aren't these causal connections? If we, so to speak, take a Kantian perspective on this debate then it looks like we are conceiving of a world where the true nature of cause is different from our actual world (and in which there are purely mental beings). 

On another note, I wonder what you would make of the following kind of argument. On the kind of view you have it seems conceivable to have cases of inversion for pain and pleasure. We should be able to conceive of a creature who experienced what I experience when I am burned alive when I would experience pleasure from a warm shower, or a nice sunny day, or whatever. But of course, these two creatures would have to be functionally identical, even physically identical. So there would be no difference at all in their behavior at all. Is this even remotely plausible? I am not sure I can make any sense of experiencing excruciating pain (as I would if I were burned alive) and yet smiling and saying how nice it felt. That these kinds of scenarios seem incoherent to me seems to be very strong evidence that there is an a priori connection between function and consciousness.   

[Philip Goff]

Richard:

As I said in reply to Declan, I think I can reject standard analytic functionalism/Australian functionalism without committing to the conceivability of a world without causation, and this is still a significant result. However, I am inclined to think that I can conceive of a world without causation, in which my mental states are causally impotent, but just happen to flow in an orderly manner (isn't epiphenomenalism conceivable?), which would refute less orthodox versions of analytic functionalism. 

However, I think the points you make against that latter bolder claim are exactly the points to press me on. Hedda Hassel argues that we find our idea of necessary connection in the link between pain and wanting to avoid, in part by alleging the inconceivability of pain/pleasure inversions. I suppose I do bite the bullet and accept that such pain/pleasure inversions are conceivable. Our resistance to them is that they are normatively inappropriate pairings of conscious states and desire, but this has nothing to do with what can be fitted together causally. 

Here's a bit of an argument in my favour. Don't you think Humeanism is conceivable? If so, and if I can conceive of a world where only I exist, and only for a short space of time, then I can conceive of a world where my conscious states are causally impotent (as there do not exist the relevant regularities to ground laws of nature). 

[Adam Pautz]

Hey Philip

I think these are interesting issues. Don't have any real substantive points. I just wanted to point out some helpful (to me) discussions of these issues and make some small points.

(1) Philip said "I am inclined to think that I can conceive of a world without causation, in which my mental states are causally impotent, but just happen to flow in an orderly manner", while nothing that that's not required for his argument. Interestingly, John Hawthorne (though no functionalist) argues that this is *inconceivable*, in his "Why Humeans are out of their minds". He uses that in an argument that Humeanism about causation can be ruled out more or less a priori.

(2) I find Richard's pain/pleasure inversion case (which btw Noa Latham discusses in a paper on Chalmers on psychophysical laws) to be totally coherent, no problem! (There are actual people seek pain and avoid pleasure, etc.)  At the same time, I am inclined to think there is *some kind* of a priori necessary connection between pain/pleasure and *dispositions* to have certain desires. (Harold Langsam talks about this in his recent book on consciousness and provides lots of references - e. g. to discussions by Sprigge.) These two claims aren't obviously incompatible. It's just that, if you accept both, you have to say that in Richard's case, even if the guy doesn't have the desire for pain to stop (e. g.) he has the disposition to have that desire. And maybe that's ok.

(3) I think Philip could accept the aforementioned claim that *some kind* of a priori necessary connection between pain/pleasure and *dispositions* to have certain desires (which, as I've said, I think is the most plausible claim in the vicinity of what Richard is saying). True, Philip claims (as I understand him) that there is no a priori necessary connection between consciousness and causation. But maybe the a priori nec. connection between pain/pleasure and *dispositions* to have certain desires is not "causal" but "constitutive". We shouldn't say that the pain causes you to have the disposition but is part of what makes it the case that you have the disposition, or something. Of course, not even clear what "causal" etc. means but maybe there is a tiny bit of daylight here.

(4) I think it is not plausible that there is an a priori nec connection between pain and desiring to engage in any particular behavior, like *avoiding* the flame (not sure but maybe that's what Hedda meant - you seemed to say that she thinks there's a nec connect between pain and avoidance behavior). I think that's not plausible because of the holism of the mental. Maybe you believe that by putting your foot more into the fire you will relieve the pain (you are very, very stupid). Maybe you live in a funny counternomic world where you believe that because it's true!!

Take care, Adam

[Hedda Hassel]

Hi, 

jumping in on the pain/pleasure inversion in the last minute!

Philip: Seems to me it's only plausible that there's a necessary connection between pain (given absence of other motives, i.e., ceteris paribus) and mentally trying to avoid the pain. The inconceivability of pain/pleasure inversions when it comes to mental tryings should be compatible with the denial of physicalist functionalism, because there is no further necessary connection between mental tryings and any physical/causal manifestation. One can try and never succeed, for all kinds of reasons. But if you accept this then you are committed to believing that pain in a sense is essentially relational/dispositional, being directed toward its own cessation, as it were. But that does not take away from the fact that it also has intrinsic qualities and can be picked out by phenomenology alone. Seems to me one can also experience that pain has this relational property in virtue of its intrinsic qualitative feel, so it's no use to the functionalist who wants mental states to be exhausted by functional relations. So I don't think you need to bite the bullet!

Richard: Your scenario of being in pain and finding yourself smiling and saying everything is okay, etc. allows for two different interpretations: in one interpretation you are also trying mentally to smile and pretend it's all fine, and all this is motivated by the feel of the pain alone. That seems incoherent/inconcievable, but that is compatible with denial of physicalist functionalism because the seemingly necessary connections are all within the mental. In the other interpretation you are mentally trying to, say, scream and run away from the pain, but you have no control over your body. What's going on here is failure of supervenience/epiphenomenalism rather than pain/pleasure inversion, given that there is no such inversion from the mental side of things. 

Adam: One can have mental holism along with a ceteris paribus necessary connection between eg. pain and avoidance attempts. The ceteris paribus clause allows that if pain is combined with other motives, such as a belief that the pain will lead to something good, then this might break the connection and even influence the character of the pain itself.  But this does not trivialise the claim (turn it into just saying "pain leads to trying to avoid except when it doesn't"). If there is a case where an agent is in pain and has nothing else on her mind, except a belief about how to try and avoid the pain, then it seems inconceivable that this should not lead to trying to avoid it, which seems substantial enough. 

Hedda 

[Adam Pautz]

Thanks, Hedda, that helps. Btw, there are puzzling cases of pain aymbolia that might be a problem both for the claim that pain necessitates (in those with the capacity for desire) the disposition to desire that it stop (Langsam, maybe me), and the related claim that ceteris paribus (no countervailing desires, no irrationality or weakness of the will, the presence of beliefs about how to make it stop) pain leads to attempts to behave in a such way that it stop (you, Sprigge?). (Colin Klein has a really helpful paper about such cases on his website.) But maybe the claims should only be made about "normal" or "non-asymbolic" pain (which is connected with desire in the right way). But now we do run a real risk of trivializing. I guess this is getting kind of far away from Philip's concerns. But, to repeat, the basic thought we're pressing is this: even if we're with Philip in rejecting functionalism as a reductive theory of mentality (with the result that non-mentally characterized functional stuff should necessitate mentality), there is room for thinking that there is an element of truth in some stuff the functionalists said - for instance, some mental states necessitate certain functional-dispositional stuff. And maybe Philip can agree.

[Richard Brown]

Hey Adam, I like the general direction you are pushing in but I am not so sure about the pain/pleasure stuff being totally coherent! So, take someone who does seek pain and avoid pleasure. These people will tell you what it is about the pain they like. They might, say, for instance, that they like the pain because it distracts them from their emotional pain, or maybe that they feel that they need to be punished, or they like the sense of control (there is an extensive literature on this)...that is they seem to cite the causal connections that pain has as part of their reason for preferring it or seeking it out. So yeah, we can make sense of people seeking pain and avoiding pleasure but that doesn't seem to support the idea of inversion, rather it seems to count against it!

[Richard Brown]

Ack! So much to say and so little time left! I wish this had been held in January instead of during finals, but c'est la vie :)

I wonder about the appeal to epiphenomenalism as a way to get at a world where there is no causation. On those views it is not the case that the the mental states 'just happen to flow in an orderly manner'. Rather, they are caused by physical states and flow the way the do because the causal states flow the way they do. It seems like you would need something more like occasionalism or parallelism but then I lose my grip on the intuition that there is no causation in said world...

[Richard Brown]

Hi Hedda!

Actually I had (what I think is a) third interpretation of the pain/pleasure inversion scenario. In my scenario there is no pretending. One is smiling, and one truly believes that one is experiencing pleasure (warm sunlight on one's arm, say) but is really experiencing the agony of being burned alive. And this is because the (qualitative) pain state is playing the functional role that the (qualitative) pleasure state plays in me. This strikes me as incoherent. But since that is what Philip's view is committed to, his view seems incoherent. 

[Richard Brown]

One final thought about Adam's appeal to Pain Asymbolia. It seems to me that the take home message from those kinds of cases is that we need to be careful to distinguish the *pain sensation* from the affective component that typically accompanies the pain sensation. We typically elide this distinction and just talk about 'pain experience' but once the distinction is made all of the points can be reformulated in terms of the latter rather than the former. 

[Jeremy Dunham]

Hi Philip,

Thank you for a very interesting (and compelling) paper. I have a few comments. I thought that Richard Brown’s reference to occasionalism was quite an interesting one and it seems from the discussion that has followed your paper that someone like Malebranche might be quite a good historical ally for your account. This seems confirmed by the discussion of the causal impotence of mental states that has arisen in the comments. Malebranche argues that there is no necessary connection between the sense-perception of pain and any neurophysiological state, and, theoretically, pain-experiencing minds could exist without bodies. It seems to me that the dualism that results from your argument is in this respect closer to Malebranche’s than Descartes’. My reason for thinking this stems from the fact that the main example of a mental state which you give throughout the paper is ‘pain’. However, it does not seem clear that for Descartes ‘pain’ would really be an example of a mental state that could be thought of clearly and distinctly without relation to the body, and actually he may well wish that we avoid reading the second meditation when we have a headache if it is to bring about the desired result! For sure, Descartes writes in the second meditation that I am a ‘thing that doubts, understands, affirms, denies, is willing, is unwilling, and also imagines and has sensory perceptions’. However, ‘imagines and has sensory perceptions’ appear after the ‘and also’ for a reason. After all, Descartes has distinguished sense-perception from thought earlier in the same meditation and told us that the former ‘surely do not appear without a body’ while it is ‘thought’ alone which is truly inseparable from me. He has to be extremely careful with sense-perception and the imagination in the second meditation because too much emphasis on either will introduce what he conceives to be the tight connection between the mind and the body, since both sensations and the imagination are used in the sixth meditation in aid of his proof of the existence of the external world. Descartes’ dualism is not as extreme as is often made out. The ‘intellect’ and ‘reason’ could exist without any connection to a body but for other types of experience Descartes requires a very close relationship between the mind and the body. Malebranche’s dualism, however, is much more extreme; he does deny that there is any necessary connection between a mental state such as 'pain' and the body and also that the former could be used as proof of the latter; thus he would be entirely on your side.

I enjoyed your paper very much!

[Philip Goff]

Sorry for delay replying...

On the general discussion...I'm still inclined to think that there is a normative rather than a necessary causal connection in these cases. In the inversion cases what worries us that the person is highly irrational. But people can be set up to be highly irrational. I don't think we discern in pain some necessary connection to trying to avoid the pain; rather we discern that it is rationally appropriate to avoid pain. Interested by Adam's suggestion that it might be constitutive rather than causal, but I'd want to hear more (will also check out the Hawthorne argument referred to). 

Jeremy: Thanks. I must admit I'm no Descartes scholar, and my aim is to suggest that the Meditations provide the resources for a certain anti-physicalist argument, not that Descartes would himself endorse that argument. But I am interested in this point about Descartes (people have said this to me before). Would you mind pointing me in the direction of where Descartes makes this explicit? (I get from reading the Meditations the view that pain could not exist without a body)

[Adam Pautz]

Hey Philip, you said "Interested by Adam's suggestion that it might be constitutive rather than causal, but I'd want to hear more (will also check out the Hawthorne argument referred to)."
Here is one idea. Maybe, as you say, necessarily, if you have severe pain it is rational to desire that it stop (contra Hume desires can be rational). That's just a basic fact. And maybe Davidson and Lewis and others are right that normative stuff about rationality helps fix prop attitudes. Give these two ideas, necessarily, if you have pain, then, ceteris paribus, you desire that it stop (cuz that's really, really rational). Indeed, in those cases having pain is part of the story of what makes it true that you have that desire. That might help to give sense to my claim that the relation might a constitive or truth-making one, not a causal one. Consciousness necessarily connected to rationality/normativity and rationality/normativity necessarily connected to belief/desire. A similar story might help explain something Hawthorne says: necessarily, if you have a bluish exp, a purplish exp and a yellowish exp, then ceteris paribus you'll be disposed to believe the first two are more alike than the first and third. When you have those experiences that belief is super rational, so it is very hard not to have that belief (at least in a standing-dispositional sense). This is a broadly a functionalist view about prop attitudes, but it's not a causal story and it's not meant to go with a functionalist theory of experience (which I think can't be functionally explained). That's rough, but in case you're interested in discuss a view like this in the last section of a paper 'Does Phenomenology Ground Content?' (forthc. in *Phenomenal Intentionality*; on my website).

[Philip Goff]

Oh yer, I've read your paper on this. Very elegant way of resolving the problem, but I'm afraid I'm too much of a fan of cognitive phenomenology to adopt it.

[Philip Goff]

P.s. Probably a bit late in the day, but does anyone think my arguments against standard/Lewisian analytic functionalism don't work? This has been a very interesting discussion, but no one seems to think those arguments work as they stand?