Very good, thanks Brad! Let me see.

1. regarding reply (a) about individuation,
I don’t think I failed to distinguish sense and reference. In fact, to confess my own prejudice, I usually understand the notion of content in terms of ‘sense’: it strikes me that a proposition is distinct from the fact or state-of-affairs which makes it true (in fact, this is slightly inexact: there is also the reference – referent distinction to take into account).

Still, I don’t see how one can specify the sense or cognitive content of a mental state without making reference to external facts. A sense is a kind of representation, it seems to me (or a mode of presentation, or a way in which things are represented, to use the traditional metaphors). So how can you specify the nature of a representation without saying anything about what it represents or purports to represent.

And to repeat, to say that one must refer to some external facts in specifying a sense is not to say that that’s all one needs to refer to. One can also refer to lots of internal facts. I suspect that two versions of Externalism are sometimes conflated:

Externalism: content is to be individuated by reference to some external facts.

Strong Externalism: content is to be individuated by reference to external facts only.

Strong Externalism is a version of Externalism, but Externalists don’t have to accept Strong Externalism. Strong Externalism strikes me as false, by the way.

2. regarding reply (b) about individuation
Lots of good things here. I’m not sure my notion of individuation really captures the issue either. But I’m not sure what really does.

You suggest that “if what makes it the case that my mental state refers to wombat is metaphysically independent of my external relations to wombats, then this should count as a form of Internalism”. I’m not entirely sure what you mean by ‘metaphysical independence’ here, but I don’t think that’s right anyway. If social externalism is true, then even though my community is independent from my relations to wombats, the content of my thoughts about wombats is to be individuated by reference to some external factors.

But your point captures something important. Part of the issue is: what makes it the case that mental state M represents O? So it’s a question about how mental states get to have their content, I take it, not about the content itself. And the issue is whether an answer to that question only involves internal facts or not. And I still don’t see how an informative answer could involve internal facts only (more on this below).

Should we stick with supervenience? Your point about Davies’ argument strikes me as incorrect in one respect: assuming a modest form of Empiricism about the mind, the circumstances of the case are not unusual at all (perhaps, the co-variance between internal factors and their external causes will be a matter of nomological necessity only, but this is still problematic for Internalists like Segal who understand the relevant supervenience theses in terms of such necessity anyway!). Anyway, even an unusual case suffices to show that supervenience  and individuation come apart: that the first doesn’t entail the second.
This difference between individuation and supervenience seems important. If we agree that supervenience doesn’t entail Individuation, and hence is distinct from it, there’s a question as to what individuation is exactly. The difference between the two might come down to this, maybe: it seems as though there are two readings of the dispute, one thick and one thin:

Thin: what constitutes mental content (or the fact that a mental state has a given content)? Only internal facts or internal and external facts?

Thick: how to fully characterise the content of a mental state, and the conditions required for a mental state to have that content (or how that mental state has that content)? Only by reference to internal facts, or be reference to external facts too?

The thin reading is in terms of supervenience. The thick question demands a different kind of explanation: among other things, it seems to require a specification of the content itself (a semantic question), and of the conditions required for there to be such content (a meta-semantic question). By the way, recall that Burge, Davies, and even Segal, are concerned with issues in scientific psychology, and it’s not entirely clear the thin reading would be of much interest to Psychologists.

My first worry in the post was that Internalism is implausible as a response to the thick question. What about the thin question? As I said, if the dispute is about the nature of content itself, then Internalism seems like a non-starter. So, what about the thin question concerning how mental states have their content?

Consider an analogy. A picture is a kind of representation too. Though it isn’t a sense or content, rather it’s a vehicle for such a sense or content (the same goes about linguistic representation). Suppose we agree contents or senses are abstract objects and that the question is how a mental state has the content that it does. The analogous question would be: how does a picture get to have the content (or representational properties) that it does. It’s not a question about what makes up such a vehicle. (The answer is easy and uninteresting: paint on a canvas. Likewise, in the mental case, the question is not: what makes up the internal vehicles which convey or carry the content had by a given mental state? That’s not a very interesting philosophical question). The question is how such a picture gets to represent what it does.

But there’s nothing about the picture itself (the bits of material making up the picture and nothing else) which can help provide an answer to that question. Rather, an informative answer will mention the relation of resemblance between the picture and its object (which involves something about the picture too, of course), or the intention of the artists, or the concepts of the community of arts lovers, etc.

Similarly, in the mental case, I don’t see how an answer that doesn’t involve some relevant causal relations with the environment (broadly construed), but restricts itself to internal facts only, can even get started (particularly in the case of perceptual content, about which I was thinking initially). Perhaps, the postulation of a little homunculus might help, but this doesn’t seem very plausible. More importantly, it’s hard to see how causal relations only between different elements or mechanisms (or parts of mechanisms) in the brain can offer an adequate answer to our question. But then, what else can provide an explanation according to the Internalist? (of course, the analogy may be misleading: a picture is not as complex as a brain. But it provides one way in which to make a start on the relevant question. It seems that a natural way to answer the relevant question about other kinds of representations is externalist and that Internalism is a non-starter in these cases. I don’t really see how the case of mental representation is any different.).

3. about the metaphysics of content. I like your suggestion.

4. about the issue of explanation. I’m not sure I understand what you say, if we’ve agreed there is no such thing as a mental content that is internal. Content as an abstract object isn’t internal at all. The question, as you seemed to agree, is whether a mental state has a content purely in virtue of some internal facts. But then, the Internalist cannot suggest that such internal facts or elements are a mental content. Can’t have it both ways. So even on the internalist picture, a content and the internal facts which make it the case that a mental state is related to such content are distinct.

Second, if the explanation of behaviour is supposed to be causal, and if the proximal cause of behaviour is supposed to be something in the head, then the content of a mental state itself cannot play such a causal role, even for the Internalist. Again, this depends on how you construe abstract objects, but usually they don’t play any causal role. In any case, contents aren’t in the head. What is in the head, according to the Internalist, are certain facts according to which a given mental state has the content that it does. So, it must be these facts which play a causal role, according to the Internalist. Perhaps, the facts in question involve certain kind of internal concrete representations, or vehicles.

But then, my point goes through. Externalists can agree that what makes it the case that a mental state has a given content is determined by some internal facts: they only insist that some external facts play a role too. But now, the Externalist can agree that such internal facts can figure in the causal explanation of behaviour.

Left by Philippe Chuard on January 15th, 2007

Hello Pete, how are you?

Fair enough! My main problem is that I’m not sure I really understand the kind of project that is at issue between the Internalists and the Externalists. And if I do understand the project, then I find myself completely blind regarding the motivations to provide an internalist solution (but you’re right: this might be a prejudice, though I’m not sure it is).

Second, I think you’re very right to mention conceptual role semantics. The only plausible story to be told on an Internalist account is a holist one, I suspect. But I don’t find holism very plausible. Remember that the point of my post was to ask: what reasons would one have to be an Internalist? It wouldn’t do to mention content-holism as a motivation, and then, when I ask why would anyone want to be a holist, appeal to Internalism to motivate the latter commitment. And there is this lingering worry about the project itself: what do you mean by a specification of a content?

Let me try to say a little bit more about this, and you can tell if that helps (warning, though: this is going to be as hand-wavy and vague as my response to Brad’s comments).

When dealing with content (linguistic, mental, pictorial, etc.), we can usually distinguish 2 kinds of projects. There is a semantic project, which is theoretically crucial, but rather mundane philosophically: the point of the project is simply to specify what the content of a representation is, by describing (or paraphrasing) what that representation represents. In other words, in engaging in the semantic project, one just tries to characterise the particular content of a given mental state, by saying what that content is (not the metaphysics of content, as with the dispute between Frege, Russell, and Stalnaker, but simply what the state represents and refers to).

The second project, much more interesting philosophically, is meta-semantic. It consists in uncovering the conditions both necessary and sufficient for a given state (or agent, or speaker, etc.) to have a particular content.

Now, the two projects are not independent. In particular, the meta-semantic project is not independent from the semantic one. To have an appropriate answer to the meta-semantic question, a meta-semantic theory must explain why a given state has the particular content that it does. In other words, the meta-semantic story must deliver the right results in respect to the semantic story.

At first sight, the Internalism-Externalism debate has something to do with the meta-semantic project: whether the conditions determining how a mental state has a certain content are all internal, or include some external facts as well.

But it seems to me that part of what needs to be done in realizing that project is first to give us the semantics of the contents at issue. And this was the reason for my intuitive “but surely”. I find it very difficult to see how one could specify the content of a mental state (i.e., describe what that state represents) without referring to the worldly objects and properties such a state purports to represent.

You might reply: ok, but the Internalist can say the following. We can refer to external facts in doing semantics. We don’t deny this. Our point is simply that we can solve the meta-semantic question without mentioning such facts.

Is this neutral enough?

Here, though, I don’t see how feasible the internalist project is. Focusing on a holistic account, as you suggest, I just don’t see how relations with other contents can deliver the right results concerning the semantic description of the contents under consideration. Surely, not all other contents a subject is related with are relevant to determining a particular content p. If that’s right, what constraints to we impose on which contents play that determining role? So I don’t really see how the project can be carried out in detail. It seems a bit hopeless!

Also, it seems as though there is always the kind of problem facing analytical functionalists about the mental. There could be two distinct contents related in exactly the same way to two different sets of contents, themselves related in the same way to two bigger sets of contents, so that there is nothing (in particular, no asymmetry) in the holistic network to which the Internalist can appeal in explaining why the two contents under consideration are distinct.

But, more importantly, to go back to my initial problem, what I still don’t see is what motivation exactly one would have to follow through with such a project. What sort of considerations would make on think: hey, let’s account for how mental states get to represent certain worldly things without mentioning anything outside the head?

Left by Philippe Chuard on January 17th, 2007

Maybe, Pete. If my characterisation of the issue is not neutral enough for your extreme internalist, then what is the issue exactly?

In this respect, you’ll need to tell me more about the distinction you make between (i) meaning and (ii) representation, and explain how exactly it is more neutral. I don’t see what the distinction is exactly. In particular, if mental content is not representational (at least in some sense), then I’m not sure exactly what we mean by content anymore. Worse, one way to read your point is as a suggestion that the dispute between Externalists and Internalists (or extreme internalists at least) collapses: they try to explain different kinds of things.

Regarding the extreme internalist, you might be right about their motivations. But I’m not sure that’s very interesting (not to me, in any case). What is interesting are the (putative) motivations for the rational internalist, who like most of us, is not swayed by sketpticism, idealism, or bad philosophical arguments like the indeterminacy of translation. After all, wouldn’t it be surprising if the only way to be a Content-internalist with decent motivations for your internalism would be by being a skeptic or idealist? Not sure many Internalists will be happy with that suggestion!

Now, I’m not claiming, by the way, that there is no room in logical space for a stable internalism, or that there is no good argument for Internalism, or that there cannot be a fairly complex but ultimately powerful theoretical motivation to espouse content-internalism. What I’d like to know is: what are the pre-theoretical (intuitive) motivations for such a view? And are they any good? So I’m not asking for a conclusive argument either. Just plausible motivations which do give some support for Internalism.

Of course, I also would like to know if there is a good argument (even if not entirely conclusive) for Internalism. I’m not sure which argument that would be!

Now, about Sellars: I’m no great fan of Sellars. Here’s why. First, I don’t see why exactly the best model for a theory of mental content has to be an account of meaning for linguistic terms. Any account of linguist meaning? Presumably not!

Second, it’s true that there is a usage of “means” which is univocal between ‘rot’ means ‘red’ and ‘und’ means ‘and’. Intuitively, it’s the usage of ‘means’ where ‘means’ means ‘is synonymous with’. No argument here. But there is also a usage of ‘means’ where Pete means synonymy by ‘means’ or Jenny means rabbit by ‘gavagai’, where we specify the referent of a term with the complement of ‘means’. Why think that the first usage is more important than the second when it comes to having a theory of meaning? After all, synonymy captures sameness of meaning, but nothing more. It can’t help us to actually specify that meaning (by providing a semantic description), or say what such meaning consists in.

Finally, it’s quite plausible to think that ‘means’ in “‘und means and” does express a word-object relation: the object in question here could be an abstract object, or a concept, or a linguist item, …

So I don’t think these sellarsian moves help much.

And brevity is fine (I wish I had more of it!!).

Left by Philippe Chuard on January 17th, 2007

Adam Pautz posted yesterday and the administrators (me) failed to approve his comment on the blog: usually we get an email warning us a new comment has been posted. Sorry about that, Adam. I thought I’d paste Adam’s comment here again so that it appears at the current end of the thread and isn’t missed by anyone. Adam wrote:

Very interesting post. Let me say a few things about your earlier comments, then I will say something about one of your more recent comments.

(1) What is Internalism? Like Brad, I think it should be taken to be a thesis of supervenience, not individuation. (I would also add that you must keep the laws of nature fixed: a point made by Lewis in “Reduction of Mind”.) Anyhow, that is how I will take it in what follows.

(2) What is the case for it? Philippe mentions an argument about explanation. Just thought I’d mention a couple of other arguments.

(a) In “Reduction of Mind” Lewis gives an argument from ordinary examples: he thinks that some propositional attitude properties just aren’t Twin-Earthable. I think he mentions something about square objects (I am without my Lewis bible at the moment). It may also be that this is how McGinn argues for Internalism about some prop attitudes in *Mental Content*. This argument is meant to be pretheoretical. So, not all arguments depend on heavy duty assumptions about explanation, etc.

(b) Another argument is empirical. This argument only works in the case of sensory content. I think that, in many sense modalities, there is evidence that phenomenology is really poorly correlated with external stuff (for instance, the character of color experience is very poorly correlated with the character of the reflectance property tracked, even under optimal conditions; the intensity of pain experience is very poorly correlated with the intensity of the stimulus, even under optimal conditions, etc.) and really well correlated with internal stuff. This I think is good evidence for at least “Dependence”: roughly, the thesis that internal factors play a role in determining phenomenology (in a certain non-trivial sense I will not try to spell out here). This thesis is compatible with external factors playing a role too (a ‘two factor theory’) and falls short of Internalism. But, although I do not myself endorse Internalism (for me what is important is Dependence), one might think that the same evidence supports Internalism: that the external factors don’t play a role at all in determining phenomenology, only internal factors matter. Given Intentionalism, this would get you Internalism about sensory content. But, again, to my mind the important thesis is Dependence, not Internalism. Nevertheless I’ll pretend I accept Internalism here.

(3) Now I will discuss two arguments against Internalism. One argument against Internalism that seemed to come up was that the relevant propositions are outside the head, involve external properties, etc. So how could bearing some prop. attitude relation R to such a proposition p supervene on internal stuff only? Just thought I would make a couple of points.

First a sociological point. This issue is addressed by Lewis in *Reduction of Mind* and was addressed by Michael Tye back when he was an Internalist about sensory content (‘Qualia, Content and the Inverted Spectrum’, p. 169). I think that what they say in reply is right: there is no conflict between the claim that content is external and involves external properties and Internalism. (Or, to put it in McGinn’s terminology, you can accept Weak Externalism but reject Strong Externalism.) But, second, an analogy that I find helpful: Numbers are “outside” objects, but, necessarily, if two objects have the same mass-property, then they bear the relation *x has mass-in-grams y* to the same number.

(4) A second argument that came up was this:

“in the mental case, I don’t see how an answer that doesn’t involve some relevant causal relations with the environment (broadly construed), but restricts itself to internal facts only, can even get started (particularly in the case of perceptual content, about which I was thinking initially). Perhaps, the postulation of a little homunculus might help, but this doesn’t seem very plausible. More importantly, it’s hard to see how causal relations only between different elements or mechanisms (or parts of mechanisms) in the brain can offer an adequate answer to our question. But then, what else can provide an explanation according to the Internalist?”

This is the argument (or question) I really wanted to discuss (because I sympathize). (The other stuff I said was more or less covered already perhaps.) The Internalist might say one of two things in reply:

(a) Give an Internalist theory of sensory content and use that to explain the supervenience. Lewis does that. *Very* roughly he says that the right assignment of content is the one that best rationalizes behavior. Since behavior is internally-supervenient, so is content. (Then he gives a story about wide content as derivative.) This theory would “explain” the narrow supervenience.

(b) Live with the mystery. The above argument (I will treat it as an argument) against Internalism seems to implicitly rely on something like the claim that “where there is supervenience it is explainable”. I take it that the relevant notion of explanation is reductive explanation: this is what the Internalist must provide but cannot, according to the argument (but I may be wrong here). (By a reductive explanation I mean an explanation that relies on some identity statement like “intentional relation R = physical/functional relation R”. As I understand Reduction, Reduction is an ontological claim, not an epistemic claim; it is compatible with B-type materialism.) So, this premise seems quite strong to me, requiring as it does a kind of global Reductionism about the entire manifest world. The Primitivist (about pretty much anything) would reject it. He would reject the theory of Lewis and say that no explanation in this sense is possible. (Schiffer holds that the supervenience of belief on the physical-functional is not explainable in this sense in *The Things we Mean*. Others take similar views on color – e. g. Campbell and McGinn. Of course, many take a similar view about experience.) Is that bad? I guess. It provides an argument against Primitivism. But the Primitivist might say that the argument for Primitivism is even stronger.

Left by Philippe Chuard on January 17th, 2007

Thanks for that, Adam!

And thanks for the references. As I said, this is not an issue I work on and so I’m rather ignorant of the literature (though I read this piece by Lewis as an undergrad a long time ago). But maybe one day I will read this stuff again. Or maybe I should have read it before writing this post!!!!

(i) About individuation and supervenience, let me say something a bit more general. Earlier in response to Pete, I used the distinction between (a) semantic and (b) meta-semantic questions. Now: however we construe the dispute between Internalists and Externalists, it’s pretty clear that it’s about (b) at least: which conditions determine whether or not a given mental state has a given content. And supervenience seems like the appropriate notion here.

Further, you might think that the notion of ‘individuation’ is not only unclear, but that one thing it suggests is an answer to a semantic question: to individuate a content is simply to give a semantic description of that content. And clearly, we shouldn’t confuse (a) with (b).

Part of my confusion about the nature of the dispute is this I fear I’m going to repeat myself). I agree with the 2 paragraphs above. On the other hand, there is this line of thought. First, many theorists in this dispute often present their view in terms of individuation. What do they mean by that?

Second, and more importantly, there is the Davies’ point. Assuming that the relevant internal facts are nomologically dependent (to suit Adam’s suggestion) upon certain external factors, then even if Internalist supervenience is true, this is not enough to show that Internalism is true: the former doesn’t entail the latter. I think that’s a good point. And, unlike Brad, I think the assumption made is quite plausible. Earlier, I suggested that a moderate form of empiricism would provide one reason for such an assumption. But in fact, even if one is a nativist and accounts for the acquisition of innate neural structures (and other innate traits) in terms of evolution, that assumption is plausible too.

But if that’s right, the point suggests that supervenience cannot be all there is to the Internalism-Externalism dispute. Of course, if Internalism is true, it will entail some internalist supervenience thesis. The point is that the converse doesn’t hold. So now the question is: what else do we need to add to a supervenience thesis to get full-blown Internalism (or Externalism) and can a specification of the notion of ‘individuation’ help?

This is why I’m puzzled about the nature of the dispute. If you want to stick to supervenience, I’d like to know how to dismiss the Davies’ point and what it seems to suggest about the dispute.

(ii) about your empirical argument (is this from your Nous paper?). Let’s grant that (1) poor correlation with external factors and (2) great correlation with internal factors suggest that phenomenology has to be dependent on internal factors at least (I’m not sure I want to grant (1), because it’s going to depend on what the relevant external factors are and whether it’s right to include only reflectance properties rather than some complex disjunctions of sets of reflectance properties together with the environmental and lighting conditions, but let’s leave this complex issue aside). As you say, it’s compatible with the claim that phenomenology is dependent on external factors too. So it’s compatible with Externalism. So I don’t really see how, by itself, this provides a motivation for Internalism, unless one confuses Externalism with what I called earlier Strong Externalism (only external factors in the supervenience basis).

Of course, talk of “initial”, “plausible”, “intuitive”, “pre-theoretical” motivation for a view is a bit tricky. As I said, I’m not asking for a conclusive argument. But it should be the case that such a motivation is good enough to sway (or at least contribute to) someone like me, who has no stake in this dispute, and is in fact quite neutral (if you don’t believe me, pretend I’m an externalist-internalist virgin).

Imagine that I’m completely new to the dispute. First, you explain the terms of the dispute to me: Internalists say content is entirely determined by internal facts; Externalists deny this and say that some external facts also determine content. Then, give me the motivation. It seems that the not-too-daft novice who is not too confused about the dialectic should say: hang on, that’s not good enough! That motivation doesn’t provide any reason at all (conclusive or not) to think that Externalism is false about phenomenology. Maybe such a standard is too high, but intuitively, this just doesn’t seem like a good motivation for Internalism. Let’s do some empirical philosophy and try it on our undergrads!!!!

(iii) your point about the metaphysics of content is essentially Stalnaker’s point, right? It’s misleading to talk about internalist and externalist accounts OF content. It’s an issue about what it takes for a mental state to HAVE content, not an issue about the metaphysics of content, right? I take it that’s the point of the analogy?

(iv) about my “I can’t see how” rant, with which you say you sympathise. I don’t really understand the motivation offered by Lewis. First, granting that the right assignment of content best rationalizes behaviour, how does this favour Internalism? (This partly repeats my point about explanation of behaviour in response to Brad.) Second, behaviour is internally supervenient? I don’t understand this. If behaviour is partly constituted by bodily movements, such events don’t seem to me to supervene on internal facts only. If they did, it seems that it would be possible for my newly envatted brain and my old self to behave in exactly the same way. That’s going to be tough on my envatted brain!

But there are less mundane reasons why this seems false. Is it the case that Oscar on earth and twin-Oscar on twin-earth behave in the same way. Well that depends on how you individuate behaviour and how fine-grained the individuation is. But it seems as though, they only behave in similar ways. After all, while Oscar drinks water, twin-Oscar engages in a different kind of behaviour by drinking twater.

As for the point of explanation, I agree that supervenience should be explainable, but I don’t think it should necessarily be a reductive explanation in the sense that it relies on the kind of identity statements you mention. There are other (weaker) notions of reductive explanations, by the way. But more importantly, I’d be happy with any kind of explanation, even non-reductive (whatever that is exactly). I don’t think the worry I tried to give voice to relies on the assumption that the explanation has to be reductive in your sense.

Suppose the question is: how can we explain how a given mental state M has the content that it does, namely p. The internalist says the necessary and sufficient conditions for M to have p are all internal factors.

What I want to know is what kind of internalist explanation (and which conditions) make this internalist claim plausible. I’m not sure it even makes sense to suggest that the fact that M has p is identical with these internal factors necessary and sufficient for M to have p.

Similarly, what explains the fact that A causes B: that is, what conditions must be satisfied for A to cause B. Does it make sense to think that A causing B is simply identical, ontologically speaking, to the set of relevant background conditions? Doesn’t seem to be the case.

Furthermore, the internalist-externalist dispute is independent from the mind-body problem and should arise whether one is a dualist or physicalist. So I don’t think I was assuming anything like you suggest.

All I’m asking is: what plausible story can be told so that internal factors are solely responsible for the fact that M has p as its content? And in particular, which internal factors would figure in such a story. I phrased the point assuming that the story would be compatible with naturalism, because although a supernatural internalist story is always possible, it’s not very interesting. But that’s all I was assuming. I think.

Still no brevity here, I’m sorry!

Left by Philippe Chuard on January 17th, 2007

Thanks everyone for all the help! (Keep helping: we’re not done yet! I’ve had a very nice chat with Brad and Robert who raised some very interesting points which they’ll hopefully post soon!)

I’ll try to reply to all the recent comments in one post, but hopefully with some order. Don’t expect any brevity!!!

First, a caveat: I said I was looking for a plausible pre-theoretical motivation for Internalism, not for a (conclusive) argument. Of course, this is a very fine line to draw here, and it may be that it can’t be drawn. After all, it is usually the case that motivations for a view can be turned into arguments and that arguments are motivations. But at least, one argument which is plausible (if not conclusive) and seems to support Internalism rather than Externalism would be good to have.

So now, let’s go through some of the proposed motivations. I’ll deal with the issues about explanation and supervenience below.

ARGUMENTS FOR INTERNALISM

(1) non-existent objects and empty thoughts (Kati)
We can think about unicorns but they don’t exist. Why is this an argument for Internalism? Because it shows that the fact that our thoughts have contents about unicorns cannot supervene upon (or be determined by) external factors. Why not? After all, how do children get to have thoughts about unicorns? By being presented with images (pictorial or literary) of unicorns, or by being told stories about unicorns. And the telling of stories or the production of images are external factors.
Here, I think, we can offer a diagnosis of the mistake. Appeal to unicorns provides an argument against a direct-reference theory of meaning in favour of descriptivism. But again, such a dispute should not be confused with the Internalism – externalism dispute about mental content. Even if the relevant contents about unicorns end up being descriptive, they could still be external. Here, for instance, is an example of an empty thought, the content of which is descriptive, but plausibly to be individuated externalistically:

The water that has arthritis is in this glass.

In general, if the concepts figuring in a descriptive thought have their content in virtue of some external factors, then Descriptivism provides no support for Internalism.

(2) evil demons and BIVs (Kati)
One can have mental states with representational content even if the external world (or at least part of it, I guess) didn’t exist, as if one was a BIV or deceived by an evil demon. Again, though, how does this support Internalism? It seems noteworthy that in both scenarios we do need to imagine something external (the mad scientist stimulating my brain via electrodes, the evil demon) which is causally responsible for the thoughts that I have. So, this seems to suggest, we need to posit some external conditions to imagine that I have thoughts with contents.

Suppose, in contrast, that you just create a brain (or a head) by molecular manipulation of some cells and put it in a jar with jello which ensures that the brain functions properly (blood flow, electrical current, etc.), but with no connection to any external stimulation of any kind. Is it plausible to imagine that such a brain will end up having psychological states with content? Unless some magical or supernatural intervention takes place, I find this highly unlikely.

Again, it seems we can explain the mistake made in the argument. It is designed to disarm a crude version of a causal theory of reference, according to which it is necessary to have a mental content about an object o that a mental state with such content be caused by o itself. But Externalists don’t have to accept that.

(3) content & rationality (Ignacio)
BIVs and non-BIVs, if externalism is true, must have beliefs with different contents. Yet, it seems, they can have very similar contents, and this similarity is important to explain rationality and inference. (Not sure I get the point about the laws of nature being probabilistic, you’ll have to say a bit more!).

Is the suggestion that such a similarity of the contents of the BIV and non-BIV support Internalism? Why exactly? Externalists agree that the contents in question supervene (among other things) on the same internal factors, while they supervene on some distinct external factors. Why isn’t the fact that such contents supervene on the same internal factors sufficient to account for the relevant similarity?

I’ll have to hear a bit more about this motivation before it makes exact sense to me. But at first sight, it doesn’t seem to be too different from the issue about explanation of behaviour mentioned earlier.

Furthermore, we have to be careful about the similarity in question. It may be plausible to assume that the BIV and non-BIV are both rational to the same extent, but that doesn’t mean they make the same inferences, even if their inferences are valid to the same degree. There will be similarities between their inferences. But we don’t have to think they are exactly the same: if the contents are different, they won’t be.

Finally, there is another reply available here, which I don’t like that much (it’s ugly!). Some externalists will deny that BIVs have mental states with contents, or at least with the right contents. Indeed, this kind of Externalism has been taken by some to provide a response to Skepticism. So if the claim that one must be able to explain the fact that BIVs and Non-BIVs have the same content is meant as an argument against Externalism, it doesn’t seem completely illegitimate to reply that the argument is question-begging.

(4) explanation of behaviour again (Brad)
I agree with you Brad: the point that content, being abstract, cannot serve in a causal explanation of behaviour, has nothing to do with the issue as to whether behaviour itself supervenes on internal factors only. But notice that in your example, it’s rather hard to see how the behaviour of reaching for a cookie, grabbing it, and putting it in your mouth can be construed internalistically. Is the cookie entirely irrelevant? What about the movements of your hands? Further, it seems that, for certain actions (for instance, killing Alfred), their performance requires success, so that the relevant behaviour in question cannot be individuated only by reference to internal factors.

About the putative difference (and advantage) between the internalist explanation and the externalist one, it’s an interesting point. I’m not sure I can see the advantage. For one thing, why are causal directness and causal sufficiency advantages in these kinds of psychological explanations?
Further, the externalist can say that my wanting to eat a cookie causes me to grab one because (i) I have the sorts of desires typically caused by types of experiences themselves caused by the presence of cookies, and (ii) the neurophysiology underlying the desire causes me to move in such a way as to grab a cookie.

I’m really not sure about how and why one explanation is supposed to be better than the other. But doesn’t the externalist explanation have a virtue lacked by the internalist one. The externalist explanation can explain why I succeed most of the time in getting what I want (absent funny conditions, or irratinonality). On the internalist explanation, the content of my mental state can contribute to explain why I act in a certain way, but there is no guarantee that I succeed (say I’m a brain in a vat), or that my success is not due to a constant form of luck. The externalist seems to be in a better position to explain my success, because I’m going to have the relevant contents only in the presence of cookies.

I’m not too sure about these points, because I’m not too sure how we really measure such advantages and whether these really are advantages.

ABOUT THE DAVIES POINT

(1) a parallel argument against Externalism (Kati)
To repeat, Davies’ point is not meant as an argument against Internalism, but rather as a methodological point: internalist supervenience doesn’t entail Internalism. As I said, the observation goes both ways. We don’t need the meningitis case to see that, even in Martin’s example, we could assume that Internalism is true, but Externalism supervenience is true too (because of the assumed correlation between internal and external factors). Hence, Externalism supervenience doesn’t entail Externalism. I don’t think this counts as an argument for either view. It’s much more interesting than that, as it raises the question what we need to add to supervenience to get a complete Internalism or Externalism!!

So far, however, everyone seems to want to stick to supervenience without providing any ground to dismiss Martin’s point. (and of course, supervenience will be fine insofar as we are trying to state necessary conditions for Internalism or Externalism).

(2) more meat? (Adam)
Suppose that content (what an experience represents) supervenes on the differential firing rate of certain particular cells in the visual cortex. Suppose that it is nomologically true that such cells cannot have the particular firing rate that they do (distinct from their base rate) unless ganglion cells in the retina have a matching firing rate (distinct from their base rate), and suppose that it is nomologically true that the latter can have a firing rate different from their base rate only in the presence of certain distal environmental causes. By transitivity, the relevant internal structures (processes, events) in the cortex (upon which content supervenes) will be nomologically dependent on the presence of certain external factors. Internalist supervenience is still true: any change in content comes with a change in the underlying internal factors. But externalist supervenience is true too: any change in content comes with accompanying environmental changes. And even if this claim is true in a small subset of possible worlds, that’s all it takes for supervenience to fail to entail Internalism (if we stipulate Internalism true in this case) or Externalism (if we stipulate its truth).

(3) logical necessity & BIVs (Brad)
Could an internal duplicate of me be created such that, from its very inception, it would lack all the relevant external relations to the environment, and nevertheless have the same mental states with the same contents that I have. Not without a miracle! And a miracle involving no external causal influence (from God, say) makes it very hard to imagine how this could possibly be so. How does the miracle work to have that effect?

I’m not sure I can really imagine such cases. Maybe I can in the same way as I can imagine worlds where God exists and where there are souls. But these are far away worlds. And their possibility, being so remote, doesn’t seem to suffice to bring any plausibility to Internalism. At least not if the issue is how our mental states get to have the content that they actually do!
As I insist below, it seems to be that Naturalism (a very weak kind of Naturalism: no supernatural explanation) is a crucial asumption when we ask: what needs to be in place for a particular mental state M to have content p? Even if Dualism were true in our world (a Dalism that is Naturalistically acceptable), the explanation would still have to be non-supernatural and informative.

And to be honest, I’m a bit puzzled by these kinds of motivations (BIVs, Idealism, and the like). The structure of these arguments seems to be like this: (1) it’s possible there is this kind of creature with contents but no relations to the environment, (2) the contents of such creature must be explained, (3) externalism cannot provide any such explanation.

First, as I said, I’m not sure whether (1) is true. It certainly seems as though the Externalism could deny such creatures: in which case, the argument would be question-begging. Now, it could be that when we claim to imagine such creatures, we succeed only by imagining worlds where Internalism is true. These could be very remote worlds. So it’s unclear they show anything about how things are with us and our contents.

Second, there is something epistemically weird about these motivations if I’m asking for a motivation making Internalism a plausible account of how our mental states get to have their contents. Internalism has a certain initial plausibility. Not very high, according to me, but at least it seems much higher than the existence of BIVs, miracles, souls, or the truth of Idealism. It seems to me that most of us, even if we disagree about the plausibility of Internalism, would agree that Internalism’s more likely to be true than these mere possibilities. But then, what’s the point of a consideration supposed to raise the probability of Internalism, if its own probability (or credence) is way lower than that of Internalism? Furthermore, whether or not Externalism is true, what’s the point for a theory of how our mental states get to have their content that it should explain these very improbable things? (and, in any case, it’s not as if there is nothing the Externalist can say about such cases, in which case (3) doesn’t follow anyway).

So it seems to me that if these are the only motivations for Internalism, then they lower the probability of Internalism rather than raise it.

Finally, since we care in the first place about what’s true in our world and we assume Naturalism, it seems nomological possibility is relevant here. You say that logical possibility is more apposite because it’s a guide to metaphysical distinctness. I don’t understand this point: I don’t see how metaphysical distinctness applies here. The distinctness of what from what?

IS IT AN ISSUE ABOUT IRREDUCTIVE EXPLANATION?

(1) a magical theory of intrinsic intentionality (Kati, Adam)
The suggestion is that Internalism might be seen as the rejection of a reductive theory of representation. So the fact that a mental state has a content is irreducible. The advantage, Kati argues, is that at least internalism doesn’t face decisive counter-examples.

4 things to say here. First, to repeat, explaining how a given mental state M gets to have the content that it has (i.e., p) doesn’t need to involve a reductionist explanation (whatever we mean by ‘reductionist’ here). But it has to mention some necessary and sufficient conditions for M to have p, such that M having p supervenes on the satisfaction of these conditions.

Second, is this already too reductionist? Is the suggestion that no such explanation can be given of intrinsic intentionality? But hang on, Internalists appeal to supervenience too and mention internal conditions which presumably are distinct from the fact that M has p. So is this just all in bad faith?

Third, if no explanation can be given of why M has p which provides necessary and sufficient conditions for that fact, then it’s true that the Internalist account doesn’t face counterexamples. But there is a good reason for that: because it doesn’t really say anything. In what sense is this even an account of how M has p. Rather, it seems to be a claim to the effect that no explanation can be provided for how M gets to have p.

Fourth, show me the decisive counter-examples to externalist explanations (the sophisticated enough ones, of course, not the intro-philo-of-mind examples)!

(2) a motivation for Externalism? (Adam)
I can see a few motivations. For instance, causation seems to be a necessary conditions for perception and for perceptual experience. It seems quite plausible that such condition plays a role in how experiences get to have their content. This applies to sensory content, by the way.(I don’t mean just causation by the object of the experience, but by a whole set of environmental factors in different situations).

Another motivation might go like this. Either empiricism is true, or Nativism is true. If Nativism is true, we must appeal to the evolution of the species in certain environmental conditions to account for the development of mental capacities. Either way, what’s in the head seems to depend (at least causally and nomologically) on the presence of certain environmental conditions at certain points in our history. It’s hard to see, if these types of explanations work generally in psychology, why they don’t work for mental content.

But I don’t really want to speak for Externalists. I’m supposed to be neutral, remember. ☺ I just don’t really understand why people want to be Internalists in the first place!

However, like you, I don’t find the insistence on the need for a reductive or reductionist theory of how states get to have their content (again, depends what you mean by ‘reductionist’) very appealing. However, Naturalism seems to be true (or at least a crucial assumption to make, if only as a working hypothesis), and some natural explanation for the fact that M has p would be nice. I guess psychologists would agree. Such explanation doesn’t have to be reductive in your strong metaphysical sense that we reduce the supervenient properties to the subvenient ones by identity. But it must deliver some conditions (necessary and sufficient) which are informative (maybe not an exhaustive list, but the most important one).

To be honest, I’m worried that mention of a reductionist explanation is a bit of a straw-man here. Surely, it’s not as if there are only 2 alternatives: either we offer a reductive explanation of how M gets to have p as its content, or it’s just a primitive fact which cannot be explained!

(3) how about sensory content? (Adam)
You ask: what about externalist motivations about sensory content, the content that determines phenomenology, according to Intentionalism? And then you ask: in cases of inverted spectra, we do not have the intuition that the phenomenology and the sensory content have to be different? By the way, isn’t this supposed to be a case refuting Intentionalism? If so, not sure I get the question.

But suppose the phenomenology is the same. Should we want to say that the content is the same? Not if the respective sensory contents of the inverts are both veridical, I take it (since their respective worlds are supposed to be different, right). So again, I’m not entirely sure which aspect of this example you want to focus on.

You claim that, if it’s an a posteriori question, the empirical facts support Internalism. Again, I take it you refer to the case of colour and reflectance properties mentioned earlier. Suppose that the particular visual phenomenology humans experience is determined, in an important respect, by a bunch of mechanisms in their visual cortex. Still, it will also depend on the kind of worldly properties which (edges, motion, shapes, colour, etc,) the cones and rods in the retina are sensitive to. But more importantly, whether a given colour experience of mine represents a given colour (in normal conditions) will depend (admittedly among other things) upon the wavelength of the light hitting the retina. Is that an internal factor? Are you sure you’re not objecting to what I called Strong Externalism (content is determined only be external factors)? You seem to suggest that much when you say your target pure ‘input’ accounts of intentionality.
Anyway, I honestly don’t see how the empirical facts could possibly support Internalism. To me, it seems that Externalism obtains its highest plausibility in the case of sensory content! ☺

Furthermore, and more interestingly, it seems to me that this kind of argument commits the same mistake (or fallacy, I’m tempted to say) as the BIV argument above. If we focus on a very narrow set of external factors (the presence of the object thought about, reflectance properties, etc.), then it’s easy to falsify Externalism. But what’s more difficult to see is why plausible versions of Externalism would restrict the set of external factors (relevant for the supervenience or individuation of content) to such a narrow list, rather than include a broader list of causally relevant environmental factors.

Finally, behavioural dispositions. I have them too. I have the disposition to go and play ice-hockey, which I will do provided I desire it srongly enough. But you don’t need to mention any content of my actual (occurrent and dispositional) mental states to explain such dispositions. I’ve never had the desire and never played ice-hockey. (And certainly, the alleged explanation doesn’t work in the usual causal way that my actual mental state with a given content causes a certain behaviour. Behavioural dispositions are not caused that way, even though their manifestations are.) Apologies for this rant: I just don’t see how this is supposed to help exactly!! If the BIV doesn’t behave, then there is no need for the content of her actual mental states to explain anything relative to her behaviour, since no such behaviour has been caused. Possible mental states with content could possibly explain possible behaviours, but the latter is external too.
VOILA!
I hope I have addressed most of the points raised. If not, tell me. And again, apologies for this inexcusably long comment.

Left by Philippe Chuard on January 20th, 2007

Hi Ignacio, thanks for your reply. Some very interesting questions here, which deserve a non-brief answer (alas!!). 3 things:

1) EXTERNAL CONDITIONS & THE BIV-MOTIVATION FOR INTERNALISM
You agree that, typically, when we imagine BIVs or evil demon-worlds, we do not imagine worlds where the ‘external part’ of the world is empty. For one thing, these worlds include mad scientists, vats and electrodes, or evil demons. (For another, what usually matters in these scenarios is not that the long list of familiar objects we are acquainted with doesn’t exist, but that what we take to be our normal epistemic relation with such objects has been severed.) More importantly, we do not imagine creatures with mental contents, where their having states with such contents is not caused by external conditions. Rather, we imagine situations where their mental states with contents are caused by very unusual external conditions.

But you object that: (1) this isn’t enough to support Externalism because (2) external causes of thoughts in the BIV-world cannot serve to individuate the contents of their thoughts and (3) the BIVs’ mental states are about their simulated world (the world that looks like ours, I take it), not their real world (the world in which they spend their life in jars in a lab at MIT).

I think that’s right (to some extent: see below). But remember, the question, to begin with, was whether BIV-scenarios offer a plausible motivation for Internalism. Now, I think BIV-scenarios fail to provide support for Internalism, but that’s not to say that they support Externalism either. In general, you don’t have to think that if one motivation fails to support a view, it thereby supports (or fully supports) the other view.

There are 2 ways to think of the structure of such motivations, one (i) where a BIV-scenario offers a positive reason for Internalism, another (ii) where the reasoning is negative.

Positive Version:
(1) there can be apparently similar mental contents had by creatures in completely different environments.
(2) Hence, this suggests that the having of such similar content must supervene on internal factors only.

Negative Version:
(1) there can be apparently similar mental contents had by creatures in completely different environments.
(2) Externalism cannot explain why such contents are similar.
(3) Therefore, Externalism is false and Internalism must be true.

Now, what does the observation that BIV-scenarios posit some external causes of mental states with content mean for these two motivations?

Regarding the first positive (and weaker) motivation, the fact that the mental states of the BIV depend on the mad scientist or evil demon for having their content seems to suggest that Internalism doesn’t provide a complete answer to the question: what conditions must be in place for M to have content p? So this scenario doesn’t really support Internalism, since Internalism seems to be the claim that internal facts are all one needs to answer that question.

After all, Internalism seems to entail that it’s possible to have mental states with content without having (or having had) any interaction with any environment. And yet, when we try to imagine such situations, we keep positing some external factors, which do a fair amount of work. Even if this point doesn’t support Externalism, it undermines the alleged intuitive appeal of Internalism.

What about the negative (and stronger) motivation for Internalism? Well, as I noted, Externalists could always reject premise (1). They could insist that BIVs only have illusions of content (for instance, they have the same vehicles of content that we have, but no real content is conveyed or carried by such vehicles).

What about premise (2)? I think it’s false too. Externalists can explain why the contents of BIVs and non-BIVS appear similar. The explanation (or a template for it, at least) mentions 3 facts: (a) the contents are not in fact the same, they are only similar, Externalists will insist; (b) the contents supervene on the same internal factors, and this explains in a large part why they are similar; (c) the reason why the evil demon or the mad scientist succeeds in causing in the BIV similar contents to the ones had by the non-BIV is that she has found a way to simulate the normal external causes of such thoughts and contents. All this is compatible with Externalism (I’ll say something about individuation below). In which case, this negative construal of the BIV motivation for Internalism doesn’t work either.

2) MORE ON SUPERVENIENCE & INDIVIDUATION
Unlike your fellow Internalists on this thread, you do seem willing to appeal to the notion of ‘individuation’: you insist the external factors posited in conceiving a BIV world (the evil demon, or the mad scientist plus some machinery) are not “enough to individuate thought contents”.

So what is this notion of ‘individuation’ and what is its relation to supervenience? This was the question I was interested in at the very beginning of this thread.

First, I’ll assume you agree that supervenience is not sufficient to get individuation either. Insofar as the BIV is concerned, it can be true that any difference in content between her thoughts co-varies with a difference in the external conditions: the evil demon or mad scientist has to do something different to generate a thought with a different content in the BIV. Hence, it seems, externalist supervenience can be true in this case. But your point is that the external conditions figuring in the supervenience basis in this case cannot serve to individuate the content of the BIV’s thoughts, since her thoughts are not about things in her environment (what you call the ‘real world’), but about things in our environment (the ‘simulated world’).

By the individuation of content, I take it you mean something like this: a complete and accurate description of the semantic content of a mental state, which specifies what the mental state is (or purports to be) about? What about the relationship between individuation and supervenience?

Perhaps, individuation is meant to be a constraint on supervenience, such that the content of a given mental state M can be read off from the list of conditions figuring in the supervenience basis. In other words, the conditions that figure in that supervenience basis must serve to characterise the semantics of the content itself.

Hence, the external fact that the evil demon is responsible for causing in me a thought about Dijon mustard cannot serve to individuate the content of my thought, because such an external fact has nothing to do with Dijon mustard. I take it that’s the idea.

Suppose that’s right. This doesn’t help the Internalist much. As I said in my response to Pete, Internalism strikes me as particularly hopeless in meeting that constraint. Suppose Internalism is true and the having of content supervenes on some machinery in the brain. It’s hard to see how a description of the semantics of the relevant content can be extracted from a description of the elements of the relevant neural machinery. Suppose alternatively that Internalism is nothing more than Holism and the fact that M has p is determined by relations between p and other contents had by other mental states of the subject. Again, I seriously doubt that a complete description of the relevant relations will provide you with an accurate description of the semantics of p. This is the permutation problem: (a) different contents might have the same (or at least symmetrical) relations to the same sets of contents, or (b), across different individuals, the same contents might be related to different contents. But, in the first place, I find it difficult to imagine how such an account would look like.
So this point doesn’t help much in trying to find a good motivation for Internalism. This is so even for the kind of Internalism you suggest which says that the individuation of content must refer only to structural facts about the brain, although the structural facts themselves depend for what they are upon external factors. It strikes me that, insofar as individuation of content is concerned, these sorts of internal facts are particularly unlikely to deliver. And if, as a result, we go back to supervenience-based formulations of Internalism, then the fact that structural facts about the brain are themselves causally dependent on some external facts will open the possibility that differences in content co-vary with external factors (even though they co-vary with internal facts too).

Now, meeting the individuation constraint in the BIV-world is going to be pretty difficult for Externalists, admittedly. Though I don’t think the situation is as dire as it is for Internalists. Perhaps, the relevant external factors are going to be something like this: the object or property x which, in normal circumstances, would have caused, together with the relevant background conditions, a mental state to represent x, where, in some abnormal conditions, x’s effect can be simulated by some alternative cause y. (If I remember correctly, some other proposal relevant to this case is suggested by Burge in his paper in the McDowell/Pettit collection).

3) PHILOSOPHY OR SCIENCE?
That’s a good question. Suppose we agree that the issue between Internalists and Externalists is: what conditions must be in place for a mental state M to have a content p, and which of these conditions determine what the content p is.

What would a philosophical account of these conditions look like? Supposing that some of these conditions involve particular neural states, how could philosophy help in discovering which states?

One popular answer about philosophical methodology in the philosophy of mind amounts to the following: philosophy analyses and systematises (cleans up) folk intuitions, assumptions, and/or tacit beliefs about the mind. Ok, but what are the folk’s intuitions about the conditions necessary and sufficient for a particular mental state M to have a particular content p? I’m not sure one can go very far with such a project. And if that’s not the methodology, then I’d like to know what is.

Notice also that, in the philosophical literature, there is currently no detailed or complete internalist account of what it takes to have a particular mental state M with content p. No Internalist even seems to be in a position to say: in order to have M with p (as opposed to q), the following necessary and sufficient conditions must be satisfied. This suggest that the philosophical dispute is rather a dispute about the kind of account (Internalist, Externalist) that a complete science of the mind will ultimately deliver.

For this reason, I think Adam is right on the money with his empirical motivation for Internalism about sensory content. I think the motivation fails, but it’s the right place to look.

The same goes for Externalism (as some of my suggestions indicated). The various intuitions to the effect that the environment plays an essential role in determining the kinds of minds we have motivates Externalism, it seems to me. So what about twin-earth? As some Externalists point out, these sorts of cases merely provide intuition pumps in favour of a view. So all they can do is help test your intuitions whether you are an Internalist or Externalist: they exploit certain logical consequences of Internalism or Externalism to see whether you accept them.

But I don’t think these cases provide very good motivations or arguments for Externalism (or Internalism). Whether twin-earth cases are used to support Internalism or Externalism, it’s always possible for the other side (a) to reject the intuition, (b) to provide an alternative description of the example which is friendly to their view.

Finally, the Davies point (at least as I understand it): to repeat, it’s simply that internalist supervenience is not enough to entail Internalism, so even if internalist supervenience is true, something else is needed to show that Internalism is true. And the same goes for Externalism, it seems. Methodologically, it suggests that a supervenience thesis is not quite the same thing as an individuation thesis.

The point about explanation of behaviour is a different point.

Left by Philippe Chuard on January 21st, 2007

Hello Adam, hello Ignacio. Thanks a lot for these replies. Again, I’ve ordered things thematically. (The issues I have the least to say are at the bottom.)

ON EMPIRICAL MOTIVATION FOR INTERNALISM ABOUT SENSORY CONTENT (Adam)
Thanks for the details, Adam. I think I understood where the point was going the first time, but this helps. I don’t know anything about gustatory experience so I’ll take your word for it. You might be right that this provides some reason for Internalism over Externalism. It strikes me that you acknowledge it’s not going to be a very strong reason, though. In fact, given what you acknowledge in the end, it doesn’t really seem to be a reason (but I’m repeating myself!). In any case, I think I still have at least 2 or 3 predictable worries:

(1) as I said earlier, you seem to assume that the relevant externalist factors mentioned by the externalist are going to be either the reflectance properties or whatever properties are tracked by the visual system (and only these kinds of properties). But why assume that? The relevant external factors could be disjunctions of such properties in particular conditions (say, the overall lighting conditions). The external factors relevant to the externalist thesis can be complex and don’t have to include only properties that are actually represented/tracked/indicated by the visual system, but also external conditions for such tracking to take place. The externalist could then say that colour experience does correlate with such disjunctions of complex properties. I’m not sure you have addressed this point. And I haven’t looked at my Hardin again, but I’d like to know if there is any empirical reason why this move is not available to Externalists.

(2) Again, haven’t looked at my Hardin, but only one kind of bad external correlation is going to be problematic for externalism, right? If there are differences in chromatic perceptual content which are unaccompanied by differences in some external conditions, or if the very same external condition fail to ensure the same chromatic perceptual content: is this what you mean by bad correlations? And again, allow for the external conditions to be a bit more complicated than you suggest. But, if there are differences in external conditions which give rise to the same chromatic contents, then we agree that this is perfectly compatible with externalist supervenience, don’t we? So the bad correlation you have in mind had better not be that, right?

(3) finally, a more general point about the dialectic. Even if you’re right that chromatic perceptual content correlates better with internal factors than it does with external factors, is this sufficient to show that externalist supervenience is false and that some external factors do not figure in the supervenience basis for such contents? That’s going to depend on the kinds of correlations at issue, I assume. But note also that if we’re only interested in supervenience without individuation, then it seems as though any change in content might come with some change in internal facts as well as some change in external facts (no matter what these facts are). Forget about dual component theory, then: it seems your data could be compatible with Externalism tout court, provided it’s not Strong Externalism. But it seems we might agree on that, given that you take this to support your Dependence thesis, rather than Internalism. And your dependence thesis might be compatible with Externalism, it seems.

ON DAVIDSONIAN/WITTGENSTEINIAN MOTIVATIONS FOR INTERNALISM (Ignacio)
You’re right Ignacio, I’m not too keen on these sorts of considerations, but I don’t think they help the Internalist. Appeal to (i) the normativity of meaning, (ii) triangulation and interpretation, (iii) interpretationist theories of content, and other Davidsonian or wittgensteinian considerations, all seem to focus on the fact(s) that mental contents are had by creatures who live in a social context, with certain interpretative and normative practices (constituted in an important sense by the subject’s behaviour), and that such practices and context determine content. But these things are external conditions. If anything, such considerations seem to support the kind of social externalism defended by Burge rather than any version of Internalism.

About Naturalism, I think that such approaches are in fact compatible with Naturalism, albeit a different kind of Naturalism. But remember, my use of the assumption that a supervenience thesis should be explainable in naturalistic terms only relied on a weak kind of Naturalism: no supernatural explanation (no magic or miracles, and no primitivist talk). Moreover, explanation is supposed to be quite weak too: just some necessary and sufficient conditions and some story why they determine the having of content. But the sort of approach you mention strikes me as perfectly naturalist in this sense.

I don’t think Anti-naturalism really helps here either. Internalists do claim that a mental state M having content P supervenes on internal facts. Which internal facts exactly? That seems to be a fair question. Presumably, an internalist theory can provide an answer, if only a vague one (otherwise, what is the point of such a theory? Furthermore, externalists clearly can provide a rather precise answer even if it faces some objections). That answer will come with some story as to how the having of a given content does supervene on a particular set of internal facts. That’s naturalist enough for me (in my sense). And it doesn’t have to be reductionist in any way (again, whatever that means exactly, but certainly not in Adams’ sense). But then, notice, it’s not unlikely that the view may be subject to counter-examples. The permutation problem I mentioned earlier is one recipe for cooking up such counter-examples.

ON BIVS AGAIN (Ignacio)
You’ll have to explain to me exactly how BIV-worlds prevent Externalists from providing a set of conditions upon which the having of a mental content supervenes. I’ve mentioned quite a few options in the past posts, so why aren’t these any good exactly? What’s the argument here?

ABOUT CONTENTS AS ABSTRACT OBJECTS (Ignacio)
Sure, you can deny that contents are propositions or abstract objects. But tell us what they are and how this accounts for the various motivations advanced for thinking that contents are abstract objects. Even if you can, I fail to see what difference exactly this makes. Appealing to the relation to contents as abstract objects (a point from Stalnaker) only helps to make clear what the dispute is really about (not about a kind of content, but about the having of content by a mental state) and to specify the issue in what seems like a neutral way.

ON THE DAVIES POINT (Adam)
The scenario I described is only one reason (albeit a plausible one) for thinking that internal factors could be dependent upon external factors. There are plenty other ways to motivate that assumption. An occasionalist might say that God necessitates certain internal facts in our minds/brains when God himself perceives certain external facts. Not very plausible, but relevant. Even if there is one world where such necessitation holds (even a remote world), that’s enough to show that supervenience (internalist or exernalist) doesn’t entail Internalism or Externalism.

ON INVERTED EARTH (Adam)
Sorry I let my ignorance get the better of me. Obviously, we disagree about what’s intuitive about sensory content. But I think you might be right about the general lesson from such cases: psychosemantics plays a big role in our intuitions about intentionalism.

Left by Philippe Chuard on January 25th, 2007

Hi Ignacio, apologies for the belated reply: things are sort of busy, here. Thanks for the reply: I’ll try to conclude my posting on this thread in my response. I fear that I may be repeating myself: sorry about that!

I did ask for an intuitive motivation for Internalism and did indeed get plenty or replies. But I disagree with you on one point. I don’t think the proposed motivations are any good, and I suspect many rely on some confusion or other (as I’ve tried to argue). In particular, I don’t think any would sway my fictitious character: the clever and annoying undergrad student sitting at the back of the class who questions everything the (internalist) lecturer says. We can assume that that student is not an Externalist: he/she just doesn’t find the alleged motivations to be very good.

You mention an epistemic motivation for Internalism: same epistemic situation, same content. Again, I don’t think that helps. First, unless we side with the skeptic, it doesn’t seem as though I am in the same overall epistemic situation as my envatted duplicate. Second, if the first point is right, you must have a small subset of epistemic facts in mind. Maybe something like this: consciously available evidence determines thought-content. Does this really help? Well, for one thing, the consciously available evidence might itself depend on external factors (even in the BIV-world), so by transitivity, … More importantly, the question was: what sort of factors determine that a mental state has this content rather than that content? I’m not sure how appeal to consciously available evidence (something like perceptual content?) helps. Is this a kind of content-holism? Finally, if the point is supposed to depend on considerations of privileged access, I think that’s pretty hopeless too (see my reply to Katy on the other thread).

You agree with me that external factors seem to be required in an answer to our question (even in BIV-worlds). And I don’t see how the epistemic point above helps the internalist to defuse that challenge.

I am also confused by your response to my demand for a theory providing conditions determining what content a mental state has. One the one hand, you simply deny the request. But that doesn’t seem to be an internalist response: if the having of content is said to supervene on internal factors, it’s fair to ask which internal factors. Rather, your response seems to suggest a third position: that the dispute between internalists and externalists doesn’t make sense, or that it collapses, or that it’s a pseudo-question. This sounds like a form of quietism. I don’t see what’s internalist about it.

On the other hand, you mention a kind of Dennettian Interpretationism or Interpretivism. Fine. But that seems to be providing necessary and sufficient conditions, though these conditions mention facts about the interpreters and the practice of interpreting other subjects. Notice that that’s a lot of external facts! To repeat, my guess is that the account is going to look a lot like Burge’s social externalism. That’s also a naturalist account in the sense I offered (I suspect Dennett would agree). It may not be reductive (in some sense), but, to repeat, I never insisted on it being so.

And I’m not sure I understand the distinction between the metaphysical and methodological disputes. I thought the way the dispute arose about perceptual content, for instance, was: look at the methodology of psychologists like Marr, doesn’t it rely on externalist metaphysical assumptions?

Left by Philippe Chuard on February 1st, 2007