Plantinga gives three main arguments against naturalism in the opening chapter of Knowledge of God. These are:
- naturalism cannot accommodate the idea of proper function (and since warrant essentially involves proper function, if naturalism were true then no one would have knowledge),
- naturalism leads directly to Humean skepticism, the condition in which you have a defeater for whatever you believe and cannot sensibly trust your cognitive faculties, and
- naturalism cannot accommodate belief; if naturalism were true, no one would believe anything (19).
Regarding (ii):
According to Plantinga, if you believe that naturalism is true,
you have a good and sufficient reason for doubting that your beliefs are mostly true. More exactly, you have a good reason for doubting that your cognitive faculties--your perception, memory, rational intuition, and the like--are reliable, provide you with mostly true beliefs.... You also don't believe that they are not reliable; you simply don't know what to think--about anything (30f).This argument is related to Plantinga's earlier presentations of the EAAN, but there are some new twists here (or at least twists that I think are new--I confess to not being up on all the extant literature on the EAAN).
We again see the connection Plantinga draws between materialism and naturalism. This previously came up in last week's discussion, and also comes up in Plantinga's discussion of (iii) below. Plantinga does have some things to say about dualistic naturalism (which he attributes to Russell, Broad, possibly Moore), but I'm going to pass over them.
Here, roughly, is the argument:
- If materialism were true, then beliefs would either supervene on NPs (neurophysiological properties) or beliefs would be reducible to NPs.
- If beliefs supervene on NPs, then it is unlikely that the content of those beliefs would be true, since the evolutionary processes would likely not produce reliable cognitive faculties.
- If beliefs reduce to NPs, then it is unlikely that the contents of those beliefs would be true, since the evolutionary processes would likely not produce reliable cognitive faculties.
- Therefore, if materialism were true, then it would be likely that our cognitive faculties were not reliable.
Once to this point, the more familiar EAAN comes in--so if naturalism plus materialism were true, then we would have a defeater for all of our beliefs, including our belief in the conjunction of naturalism and materialism....
The theistic materialism would seem to have at least two ways of rejecting this sort of argument (though I don't recall Plantinga mentioning either of these). One would be a guided evolution in which God providentially ensures that the beliefs that either supervene on or reduce to the NPs are reliable with respect to truth. The second would be do say that beliefs neither supervene on nor reduce to NPs, but instead are miraculously brought about by God in a way that corresponds with the NPs in the brains of those who have the beliefs in question, something along the lines of a miraculous occasionalism.
Regarding (iii):
Plantinga's final main argument is for a stronger conclusion than either (i) or (ii). If naturalism were true, then not only can you not have proper function (and thus warrant) or knowledge, but you cannot even have beliefs:
"[I]f you are a naturalist, then (so I say) you should reject the idea that anyone believes anything" (51).(I'm sure the Churchland's are pleased with this argument.) Plantinga cites Plato and Leibniz as giving versions of this kind of argument. Plantinga's version in this section goes roughly as follows:
- If naturalism were true, then either materialism or dualism (of some sort or other) about human persons would be true.
- If materialism about human persons were true, then human persons could not have beliefs, since the truth of materialism would mean that there is no content.
- If dualism about human persons were true, then naturalism could not account for content.
- Therefore, if naturalism were true, there would be no content.
- Therefore, if naturalism were true, then human persons would not have beliefs.
And why can't materialists account for content?
The answer, I think, is that one can just see upon reflection that these things are impossible.... One can see that a physical object just can't do that sort of thing [i.e., think]. This isn't as clear, perhaps, as that a proposition can't be red; some impossibilities are more clearly impossible than others. But one can see it at least to some degree. And the same doesn't go for an immaterial thing's thinking; we certainly can't see that no immaterial thing can think (57f).I would love to see Plantinga explain this to van Inwagen: "You see, Peter, you're just not thinking clearly when you hold both that you are a material being and that you think. I can just see that such a thing isn't possible." While trying to argue for the possibility or impossibility of something is difficult, surely more is needed here.
There is another problem with the argument for (2). Here is a key passage:
Presumably neither [electrons nor quarks] can think--neither can believe, doubt, want, fear, or feel pain. But then a proton composed of quarks won't be able to think either, at least by way of physical relations between its component quarks, and the same will go for an atom composed of protons and electrons, a molecule composed of atoms, a cell composed of molecules, and an organ (e.g., a brain) composed of cells. If electrons and quarks can't think, we won't find anything composed of them that can think by way of the physical interaction of its parts (53).
What is needed are reasons to think that this argument holds for thinking when a parallel argument for other kinds of properties is invalid:
Electrons and quarks lack property p.Run the argument with being visible to the naked eye or being able to shatter a window in the place of p and the argument form clearly fails. Plantinga then needs to give more reasons to think that the argument works when talking about being able to have beliefs rather than being visible to the naked eye or being able to shatter a window.
Therefore, atoms composed of electrons and quarks will lack property p.
Therefore, molecules composed of atoms will lack property p.
Therefore, macrophysical objects composed of molecules will lack property p.
Dualism does not face this content-problem, Plantinga claims since immaterial selves are simple and thus thinking is basic to them (58). But do all forms of dualism about human persons think that persons are simple? I don't think so. I would think that Aquinas' view of human nature would be a form of dualism, at least as that term is used by Plantinga, and human persons aren't simply on such an account.
[Disclaimer--I completed this post on 3 July. By the time it appears on the blog, I will be away on vacation, my first in 7+ years which does not involve me taking along a laptop. While this is a small moral victory for me, it does mean that I will not be able to participate in the ensuing discussion as much as I would otherwise.]
I can't think of a good reason to accept this premise:
If materialism were true, then beliefs would either supervene on NPs (neurophysiological properties) or beliefs would be reducible to NPs.
I know Burge himself is not an advocate of materialism, but his thought experiments show rather conclusively (to my mind, at least) that beliefs neither supervene on NPs nor could be reduced to such properties. The contents of our thoughts and the facts about what we believe depend (in part) upon relations between NPs and further properties found in the environment. Maybe there's a way of rewriting Plantinga's argument to accommodate anti-individualism in the philosophy of mind, but I think there had better be. Didn't individualism go out with Atari?
I agree with you, Kevin, in thinking that Plantinga's argument against materialism about believing creatures is too quick. In fact, I think that not only is there the worry that Plantinga's argument commits the fallacy of composition, there is the additional worry which we might state as follows. If anti-individualism is true, to the extent that it is proper to speak of mental properties as properties of narrowly individuated states of the brain, these will be extrinsic properties. Thus, even if an inspection of something could determine whether it had certain intrinsic properties (I think this is a dubious assumption), insofar as mental properties are extrinsic it is hardly surprising that an inspection of neurology would not tell us whether the subject was thinking of Vienna.
(Now, Plantinga might say that this just pushes the problem back. Inspecting the subject in her environment, he might say, might never put us in a position to see that the subject believes certain things. Moreover, inspecting the subject in relation to her environment, he might say, we will come to the conclusion that her beliefs just cannot supervene on her neurology and the relations between NPs and her environment.)
Maybe the materialist could push back (I'll save it for another post), but let me note something odd. On interactionist dualist views such as Descartes', a sensation will have a non-mental cause found in the subject's neurology. When I think about the properties of being a belief and causing in an immaterial substance a modification that is itself the forming of a belief, to the extent that I can "just see" that the former is neither an NP nor a property that supervenes on NPs, I'd say the same for the latter. I'd also say the same for the properties of being an experience of something red and prompting an experience with a certain qualitative character. So, to the extent that the argument has shown anything, are we to say that the argument has shown that interactionist dualism cannot be correct insofar as it takes properties like prompting an experience with a certain qualitative character to supervene or reduce to certain NPs (or certain NPs with laws specified)? If the causal theory of perception is true and you undergo an experience that is not causally related in the appropriate way (i.e., through the sense organs), you hallucinate. If you never could undergo such an experience, you're blind. Does Plantinga's argument give us the means by which we could determine we're blind from the armchair? That would be cool.
Kevin,
I would love to see Plantinga explain this to van Inwagen: "You see, Peter, you're just not thinking clearly when you hold both that you are a material being and that you think. I can just see that such a thing isn't possible."
I believe (from PvI's really nice book Metaphysics) that PvI agrees with the intuitions that it seems impossible for electrons or things composed out of electrons and atoms to be the sorts of things that are beliefs or are thinking. He instead tries to show that dualism has the same problem, and one of Plantinga's main contributions in the section is to respond to PvI by arguing that dualism doesn't have the same problem.
Electrons and quarks lack property p.
Therefore, atoms composed of electrons and quarks will lack property p.
Therefore, molecules composed of atoms will lack property p.
Therefore, macrophysical objects composed of molecules will lack property p.
I don't think Plantinga is presenting this argument. That first "therefore", I think, isn't be a part of his argument, since I think that Plantinga thinks that we have separate intuitions supporting both of those first two premises. We see that an electron couldn't be a belief, and we see that no matter how we combine a bunch of electrons, they aren't the sorts of things that couldn't be beliefs. I think that's why your argument with "being visible to the eye" isn't analogous.
Clayton,
I may have different intuitions than you, but while it seems impossible that an electron or arrangement of electrons (like the bicycle) cannot be think or have original content, it doesn't seem impossible to me that a nonmaterial thing be in causal relations with a material thing.
Think of Plantinga's quote: "To repeat: it isn't just that we can't see how it's done, in the way in which we can't see how the sleight-of-hand artist gets the pea to wind up under the middle shell. It is rather that we can see, to at least some degree, that it can't be done, just as we can see that an elephant can't be the number 7, and that the number 7 can't weigh seven pounds" (p. 56). This seems to me to be like your analogy in the following way. I can't see how a physical events could cause nonphysical events, but don't see that it couldn't be so. And it does seem to me that physical arrangements don't have content.
Andrew,
I don't think we can just "see" that it can't be that there's nothing more needed for belief than matter. Here's a quick argument.
(1) In general, inspecting the material parts does not put one in a position to see what could arise once the material parts are in place.
(2) As such, only people with special epistemic powers could see what Plantinga claims to see.
Because of (1), I'm generally unimpressed with the claim that a person can just "see" that something lacks a supervenience basis that is purely material. Let's change examples. Think about a computer. Do you think someone can just see by inspecting the computer's molecules that these molecules could be arranged in just such a way that a material object with the powers of the computer would arise? I think not. But, if (1) is true, I don't see how someone could have the epistemic powers to say that thinking about material parts and their properties we can just pronounce that material objects composed of these material parts cannot have further properties, powers, etc... I don't see Plantinga giving any argument whatsoever against materialism about the mental. Maybe he's willing to accept (2).
Clayton writes:
"I know Burge himself is not an advocate of materialism, but his thought experiments show rather conclusively (to my mind, at least) that beliefs neither supervene on NPs nor could be reduced to such properties. The contents of our thoughts and the facts about what we believe depend (in part) upon relations between NPs and further properties found in the environment. Maybe there's a way of rewriting Plantinga's argument to accommodate anti-individualism in the philosophy of mind, but I think there had better be. Didn't individualism go out with Atari?"
The dominant view is, as you say, that Putnam/Burge type arguments show that the *contents* of beliefs do not supervene on NPs, but you seem to equate this with the claim that beliefs do not so supervene. The inference from the former to the latter is, to put it mildly, not straightforward and does not fall right out of the standard thought experiments. Further controversial premises are needed and these won't be established in the classic Burge or Putnam papers though the issues are discussed in some of the relevant philosophy of mind literature.
Anon,
You are right that Burge's thought experiments do not show that the fact that a subject believes something or other is not a fact that holds in virtue of the intrinsic properties of that believer, but it seems to matter to Plantinga's argument that beliefs with a certain content either supervene on or are reducible to something narrow. I'll let Plantinga speak for himself (now that I am home and have the book before me):
In any event, what I suggested (p. 35) is that, given materialism, the plausible thing to think is that such mental properties as being a belief with such and such content are either identical with neurophysical (NP) properties or supervene upon them, either logically or causally. (p. 229).
So, you are right that there is a subtlety here that was glossed over in the comment above but I don't think that taking account of it will help with Plantinga's argument.
Anon,
Two more things quickly before I head out for the night. I think that the remark you quoted from me made it pretty clear that the sense of 'belief' I had in mind was one where it makes little sense to speak of there being a belief without it being a belief with a certain content. Anyway, I'm curious, are you of the view that Plantinga's argument can be made to work regardless of complications having to do with content externalist considerations? I'm not sure how you think this is going to further discussion.
Clayton, there are three issues here. What Burge and others show; what Plantinga claims; what Plantinga *needs* to claim to push his argument.
It appears we don't disagree about the first.But about this:
"I think that the remark you quoted from me made it pretty clear that the sense of 'belief' I had in mind was one where it makes little sense to speak of there being a belief without it being a belief with a certain content."
Everyone accepts that every belief is a belief with a certain content (right?), but it does not follow from that that belief states are individuated by their contents. This is where some of the subtleties about from earlier arise.
But ok, more importantly, about what Plantinga says and needs to say. Let me ask a question:
How obvious is it that Plantinga takes neurophysical properties to be "local" or "non-wide" properties?
After all, he is aware of externalism about content and, unless I'm mistaken, has nowhere rejected it. And so when he says that beliefs-with-specified-content supervene on NPs charity suggests he thinks the relevant physical supervenience bases are wide. He would not be alone in thinking this, though many of us shift, as I did just now, to calling such wide states "physical" rather than "NP" states.
This relates to the issue of what Plantinga (thinks he) needs for his argument. You ask how could any of this help with Plantinga's argument. Well, given that the argument is dominantly supported by his claim to be able to see that the material supervenience base *can't* suffice for mentality it seems to me that he'd make that very same claim about a wide-but-still-physical supervenience base for a belief-that-p as he (might) make for a narrow supervenience base for a belief state with but not individuated by its content.
The criticisms of Plantinga's claim to be able to see such a thing would still apply. As would the meta-philosophical points about the evidential or argumentative value of claiming to simply see things that many others confidently report they can't see. But the charge that Plantinga doesn't know some basic philosophy of mind would go away. So there's a certain consolidation of the dispute gained by avoiding the red herring of externalism that you started this discussion with. Seems like enough of an advance for a blog discussion to me.
A few comments in response to what Andrew says above.
Andrew writes: I believe (from PvI's really nice book Metaphysics) that PvI agrees with the intuitions that it seems impossible for electrons or things composed out of electrons and atoms to be the sorts of things that are beliefs or are thinking.
I'm away from my office and so would have to go back to look at what PvI says, but even if he agrees with the intuition here, he can't think that the intuition establishes what Plantinga thinks it does.
I don't think Plantinga is presenting this argument. That first "therefore", I think, isn't be a part of his argument, since I think that Plantinga thinks that we have separate intuitions supporting both of those first two premises.
Ok. But the argument is still fallacious even if the first two lines are premises, rather than a premise and a sub-conclusion.
Kevin,
Right, but that's because he thinks that the same intuitions plague dualism, so then (at least for PvI), it should depend on whether Plantinga's considerations in favor of dualism are insufficient. (You don't even have to check his book; Plantinga cites him as agreeing with the intuition that material things can't think.)
On the second point, I only said that for the second premise, but it's clear that Plantinga intends it for the other premises as well. It seems impossible for an electron to think, and it seems impossible that, even if we combine electrons and atoms and arrange them in various ways or whatever, it's not the sort of thing that thinks. Both intuitions are similar to the intuition that it's impossible that the number 2 is red. The number 2 isn't the sort of thing that can be red, and atoms and arrangements of atoms aren't the sort of things can think. Someone might disagree with the reliability of these intuitions (as Clayton does), but I don't think Plantinga's making the sort of inference you're saying he's making.
This relates to the issue of what Plantinga (thinks he) needs for his argument. You ask how could any of this help with Plantinga's argument. Well, given that the argument is dominantly supported by his claim to be able to see that the material supervenience base *can't* suffice for mentality it seems to me that he'd make that very same claim about a wide-but-still-physical supervenience base for a belief-that-p as he (might) make for a narrow supervenience base for a belief state with but not individuated by its content.
There are two arguments that need to be distinguished. The first is Plantinga's argument that a material entity cannot have beliefs. The second is the EAAN. The remarks about anti-individualism had to do with the second, not the first. In running the EAAN, Plantinga seems to be bracketing qualms about the possibility of belief given naturalist assumptions and asking about the reliability of the processes that produce those beliefs given the naturalistic assumptions. My remarks about wide content had to do with the second argument. When I first made my comments, they were based on Kevin's reconstruction but I've since gone back to look at the text and Plantinga says that he is going to bracket any discussion of wide content in his discussion of EAAN (This suggests that he does not think that NPs are individuated broadly). I think this is unfortunate. There has been recent work, which I happen to think is quite good, on the connection between externalism in philosophy of mind and epistemology that could be of use to addressing the EAAN. (See, for example, Sarah Sawyer's, "The Epistemological Argument for Content Externalism" for an argument to the effect that the conditions that individuate our thought contents underwrite the kind of connection to our external environment that would establish a reliable connection between our judgments and the features of the world that make them true.)
One final comment. You wrote:
the charge that Plantinga doesn't know some basic philosophy of mind would go away.
I never made such a charge and I think it's a bit ridiculous for you to accuse me of having done that. I don't think Plantinga doesn't know basic philosophy of mind. What worries me about Plantinga's discussion is that he does not seem willing to discuss the strongest versions of the naturalist/materialist view. This is a charge that I think Tooley essentially makes in his response.
I think I'll follow the lead of others in not arguing with anonymous commentators.
About this:
***
"One final comment. You wrote:
the charge that Plantinga doesn't know some basic philosophy of mind would go away.
I never made such a charge and I think it's a bit ridiculous for you to accuse me of having done that.
.....
.....
I think I'll follow the lead of others in not arguing with anonymous commentators.
***
I interpreted you as you saying Plantinga doesn't know some basic philosophy of mind
because you said that you see "no reason" for accepting this Plantinga premise:
"If materialism were true, then beliefs would either supervene on NPs (neurophysiological properties) or beliefs would be reducible to NPs."
And the reason you gave was that Burge's
"thought experiments show rather conclusively (to my mind, at least) that beliefs neither supervene on NPs nor could be reduced to such properties"
suggesting that an awareness of Burge leads to there being no reason to say what Plantinga did say. And you ended with the flourish:
"Didn't individualism go out with Atari?"
which at least seemed to suggest that you think Plantinga's argument is stuck back in pre-Burge / Atari days.
I hope it's not an obviously uncharitable or "ridiculous" interpretation to read you as having claimed that Plantinga would not assert the premise in question if he, like you, understood the conclusion of the Burge thought experiment (that is, the basic philosophy of mind).
But apparently that's not what you meant in saying that you see "no reason" to accept the premise given what we, but in your view apparently not Plantinga, have learned from Burge.
As to whether it's worth arguing with anonymous contributors, I think that sometimes what we say is worthwhile and sometimes it isn't and making case-by-case decisions about whether to argue with us seems like the only thing to do. Sorry if my comment contributions aren't up to local standards or are in some other way problematic.
One other small point. About this:
"I've since gone back to look at the text and Plantinga says that he is going to bracket any discussion of wide content in his discussion of EAAN (This suggests that he does not think that NPs are individuated broadly)."
I don't think that bracketing a discussion of wide content suggests any view about how NPs are individuated. It's only if he took up the additional complexities about wide content that he'd have to face the question about NPs so I think that bracketing the wide content issues is a way of staying silent about whether NP states are narrow or broad.
Why bracket the wide content issues? I assume he does this because, rightly or wrongly, he thinks that it adds a layer of irrelevant complexity to the discussion. That's a claim he's made in response to questions at public lectures on this material.
In the published versions it'd be nice to find at least a footnote explanation from him of why he thinks the additional complexities are irrelevant because it's apparently not obvious to some of us that this is right. I take it we've agreed about that much.
Clayton,
I'm getting to this a little late and I haven't studied the earlier notes (though I did read them). But it seems pretty clear that Plantinga is distinguishing beliefs from their contents in this argument. You see no reason to accept (a).
a. If materialism were true, then beliefs would either supervene on NPs (neurophysiological properties) or beliefs would be reducible to NPs.
But (a) is seems reasonable in light of (b). (b) makes it pretty clear (to me, anyway) that he is distinguishing beliefs from their contents. Here's (b)
b. If beliefs supervene on NPs, then it is unlikely that the content of those beliefs would be true, since the evolutionary processes would likely not produce reliable cognitive faculties.
Plantinga also makes this distinction in his discussion of "semantic epiphenomenalism" (in his EAAN argument). This is one view about how beliefs might be related to behavior on which beliefs are causally efficacious in behavior, though belief-contents are not. He says about this view of belief (obviously, not one that he endorses himself),
"On a naturalist or at least a materialist way of thinking, a belief could perhaps be something like a long-term pattern of neural activity, a long-term neuronal event"(Naturalism Defeated?, Beilby (ed), p.6 ff.)
This event has all sorts of syntactical properties (neurophysiological properties, electrochemical properties, etc.). It will also have semantical properties, especially the property of being the belief that p (for some proposition p). But only the syntactical properties that have any causal efficacy and this sort of materialist identifies belief with the syntactic side.
Hey Mike,
Plantinga does distinguish beliefs from their contents (although I thought my initial remark made it clear that I was concerned with beliefs with their contents) but makes it explicit on pp. 35 the view he's attacking:
But what about the content of a belief? If a belief is a neuronal event where does it get its content from? How does it get to be associated in that way with a given proposition? Materialists offer two (or possibly three) main theories here. According to the first, content supervenes upon NP properties; according to the second, content is reducible to NP properties.
He then goes on to add:
Note also for present purposes I ignore so-called 'wide-content.' [i.e., what some of us just call 'content'] ... In the interest of simplicity I ignore wide content; nothing in my argument below hinges on this omission
Well, I'm not sure that nothing hinges on this omission. Look at his remarks about indication and belief content on pp. 38. [I've now seen that Tooley picks up on this, too. See pp. 207.] On Tooley's reconstruction, Plantinga grants:
(6) If Darwinian evolution is true, then even if it is true both that neural states of type N are, in an organism H, reliable indicators of the presence of an instance of property P, and also that states of type N have content C, there is no reason why content need to be related to property P.
(7) If Darwinian evolution is true, then even if it is true both that neural states of type N are, in organism H, reliable indicators of the presence of an instance of property P, and also that states of type N have content C, there is no reason why content C need be related to property P.
Plantinga wrote on pp. 38:
Indication is one thing; belief content is something else altogether, and we know of no reason why one should be related to the other. Content simply arises upon teh appearance of neural structures of sufficient complexity; there is no reason why that content need to be related to what the structures indicate, if anything. The proposition constituting that content need not be so much as about the predator.
I take it that there will be some content externalists who balk at this and so reject (7). Tooley calls him on it and I guess we can later discuss Plantinga's response on pp. 232.
Thanks Clayton, I think I see your worry. I don't clearly see the difference between (6) and (7), since they seem to differ only wrt the variable 'C' in the last clause. In any case, it looks like the clause that bothers you is just that last one: viz., . . . there is no reason why content C need be related to property P. So, I observe some XYZ on TwinE and, as I observe it, I form some belief. It seems like Plantinga's claim is no more than that the content of my belief need not be about XYZ at all, even if the mental state I'm in reliably indicates the presence of XYZ. Your objection (tell me if I'm tracking you) seems (very roughly, of course) to be that IF the belief IS ABOUT that substance, then XYZ will be part of it's content. Plantinga could not agree with that?
Oops,
That's what I get for coming into the office to try to work. People find me, I get distracted, and (6) becomes (7). Backtrack. It seems Plantinga is willing to concede (6):
(6) If Darwinian evolution is true, then natural selection has given rise in animals to neural states that are reliable indicators of important states of affairs (e.g., the presence of predators).
But, he thinks that in light of (7) there is yet a serious problem for the naturalist. I think one worry for this argument is this: if the relevant neural state N is a reliable indicator of the presence of P, assuming that N is the narrow neural condition of certain thoughts there is going to be a reason to think that the thought concerns P rather than, say, some other property that does not obtain when the P-involving state of affairs indicated by N does obtain. I take it to show that it's not likely that our thoughts would be reliably correct in spite of (6), he'd have to argue that the contents of our thoughts do not depend on the wide causal relations between the inner neural properties and the external properties they reliably indicate. If, however, the neural property N that reliably indicates the presence of property P gave rise to a thought that attributed P, I can't see what is left of Plantinga's worry.
You can see Plantinga's response to this on pp. 230 and again on pp. 232.