I just stopped in for a quick visit and have not read this entire thread—so my brief comment may not be fully informed of what has already been covered.
We sometimes think in words but it depends upon the context and the person. I know that I tend to think in terms of ideas rather than words—but some people find the idea of thinking without using words in an “inner monologue/dialogue” unthinkable. (Pun only partially intended.)
I think it was psychologist Russell Hurlburt’s life’s work which helped establish that many people don’t think in terms of an inner monologue of words. And I think there are MRI studies which show that people don’t use the language areas of the brain at all when solving logic/mathematics problems which don’t rely upon words from a natural language (e.g., integrating a function.)
As to the title of this thread, “An argument for the immateriality of the intellect”, I’ve found that as I age the arguments concerning my intellect have become increasingly immaterial.
That may be a additional consideration, but it doesn’t affect my point. Feser’s arguments simply don’t apply to concepts like “green”.
No. Exact communication is only possible because the difficulties raised by Premise 2 can be overcome, and Premise 1 is irrelevant to that. And if the difficulties posed by Premise 2 can be overcome in the case of communication then there is no reason to think that they are necessarily a problem for the mind - even if it is physically based.
Only if you accept physicalism, which I think is the point of the argument. If both premises are true then physicalism is false. Although I don’t find this argument particularly convincing.
That is a good question, and the answer is neither. Or both, sort of. Feser is a Thomist, so he falls into a camp that is nowadays called “hylomorphic dualism”. See here for an interesting spin on how his position relates to substance or property dualism.
The relationship is highly contingent. In fact, most people in the world would not understand the word “adding” to mean adding or quadding, because they are non-English speakers - that alone demonstrates that there is nothing about the word “adding” itself that picks out any meaning at all.
The point of the semantic indeterminacy arguments referenced in support of Premise 2 is that even the whole context in which the word is used is insufficient to determinately fix its meaning; there will always be some ambiguity.
And no, this is not in conflict with the fact that we can learn the meaning of words with far less than the whole context in which they are used; again, since our thoughts can have determinate semantic content, we can associate meaning with words without the words themselves having determinate content.
There are pure concepts that we only attach to words after the fact: you can clearly understand, even experience, when two different words mean the same thing, or when the same word has more than one meaning. (And, as @AllenWitmerMiller pointed out, we don’t always think in words.)
Concepts and words (or words, mental images, and whatever else we might “think in”) are distinct. The question is if one depends on the other, and if so, which way the dependence goes. An upshot of the arguments referenced under premise 2 is that if our conceptual understanding depended on the meanings of words (etc), our thoughts would not have determinate semantic content. But they do - we can (at least sometimes, I claim) know what we are thinking without ambiguity. So the meanings of words depends on our concepts, not the other way around.
Sure, evidence is relevant to philosophy. Philosophy is all about evidence in the most general sense: reasons for believing that something is true. But I suspect you had a narrow sense of evidence in mind; I suspect you want empirical evidence, or something like it. Now, the support for the premises of this argument are empirical in a way. They are grounded in experience: experience of introspection of own thoughts, and experience of the ways that physically instantiated patterns (such as words) are related to the meanings of our thoughts.
But again, I suspect that doesn’t at all satisfy you; you want measurements, something we can do statistics on. But this isn’t that kind of question. What exactly are you expecting? If the conclusion of the argument I’m presenting here is true, what kind of measurements do you think I should be making to give your kind of evidence for it? I’m suggesting that there is some immaterial aspect to our minds that makes it possible for us to form concepts, to think abstractly, to reason. The observable consequences this would have if it were true would be… we could communicate with language, we could do math and science, we could understand the world around us in a way that lets us build incredibly complicated technology and get into arguments about metaphysics on the internet. It isn’t the kind of question that can be decided by any particular scientific measurement, rather, it is the kind of question the answer to which factors into whether science is possible in the first place.
Evidence (in the general sense) is relevant to philosophy; but here I think I should also point out that philosophy is relevant to evidence. You think we have evidence that brains exist? Not if all you consider is empirical evidence; the best that can do is show that reality appears to us as if brains exist. The question of whether external world realism or idealism or solipsism is true is a philosophical one. You think we have evidence that neurons firing produces mental activity? No, the best that empirical evidence can do is show that mental activity is correlated with neurons firing; if you want to say that either produces the other, you start getting into philosophy. You think there’s no need to take this argument seriously because it can’t be demonstrated empirically? Where is your empirical evidence backing up that belief?
This “evidence vs. arguments” kind of thinking invariably saws off the branch it’s sitting on (and then tries to pull itself up by its bootstraps).
The fact that it is possible to conceptualize the alternative interpretation, and know that it’s not what you actually meant, is sufficient for the argument to work.
No, they are not.
Premise 1 has the logical form “some A is B”. (A = thoughts, B = things with determinate semantic content)
Premise 2 has the logical form “no C is B”. (C = merely physical things)
There is no inherent contradiction, as a matter of logic. If you want to claim there’s an implicit contradiction, you need to show there’s some implicit premise that I’m assuming that makes these two statements inconsistent.
No, it really isn’t. It may be enough for Feser’s argument but if so it just shows that Feser’s argument is unsound.
So can you answer this. If a physical state corresponds to one and only one mental state is that sufficient to satisfy Premise 1? If not then why should we believe that Premise 1 is true?
If I’m following correctly, the argument you are referring to here is:
If that’s the case, then no, it doesn’t work. It works if you assume that when we think in words, the content of our thoughts is determined by the words. But that is false, as any word with more than one definition demonstrates.
I’m afraid I’m not sure what your point is. If I may step back a bit:
I did not respond to this before because I really don’t see actually see anything that gives reason to reject the premises, or that undercuts the support I presented for them. Could you expand on what you said here?
I can see no reason to believe this, and you haven’t provided any.
As far as I can tell, the difficulties raised by Premise 2 can be overcome in the case of communication in the exact same way that they can be overcome in general - namely, Premise 1 is true, and what makes it possible for our thoughts to have determinate meaning also makes it possible for us to attach determinate meanings to words that don’t have them (and to infer what meanings others are attaching to their words). And again, none of this implies that Premise 2 is false - unless you beg the question by assuming the conclusion of the argument is false.
How is it begging the question? If both premises are correct, then physicalism is false. As @structureoftruth pointed out, this argument takes the form “some A are B; all C are not B; therefore, some A are not C.” That is not contradictory or begging the question. I’m not sure you understand the argument.
However, I still don’t find this argument convincing. I’m not sure the arguments for premise 2 are correct, because I see no reason to think that (for example) adding and “quadding” cannot be represented by different physical states in the brain.
Looked above, didn’t see any justification for your assertion that the two premises contradict each other.
I will concede that I spoke a bit too quickly there; what I should have said is that the argument does not require anyone to actually engage in these kinds of alternative interpretations, just that it is logically possible to do so.
I’m not sure what you mean. Given everything I’ve said here, obviously I don’t think any physical state can correspond to one and only one meaning, since merely physical things have no intrinsic features that give them semantic content (feel free to chime in if you have a counterexample!), and any meaning derived extrinsically is going to be ambiguous (following the semantic indeterminacy arguments). The supposition that some merely physical state corresponds to one and only meaning is just the supposition that Premise 2 is false, and isn’t directly relevant to Premise 1.
We should believe Premise 1 is true for the reasons that I gave originally, and if you meant to ask about Premise 2, I just summarized the reasoning.
Don’t think it does. We of course know which meaning we intend when we think of a word. But does that mean that the word isn’t being the carrier of meaning? And how does this show that thought must be non-physical?
I think it is quite clear. I am suggesting that we do not require full determinacy in the sense that Feser means in Premise 1. If that is correct Premise 1 is false. If the only real requirement is that we know what we mean then it is obviously necessary to show that we cannot know what we mean if our thoughts fail to meet that standard.
That someone else can look at the (assumed) physical representation of our thought and impose their own interpretation on it does not in itself pose any threat to our ability to know what we mean. Yet such is taken as violating the determinacy that Feser requires.
Obviously the possibility of exact communication shows that the issues raised in support of Premise 2 can be overcome. And none of Feser’s examples assume that the participants do not meet the determinacy of thought requirement.
Feser’s way to overcome the problems is to insist that the communication is determinate. Feser’s support for Premise 2 denies this for verbal communication. And the point is not to argue that Premise 2 is false at all. The point is to show that the support for Premise 2 fails to show that the problems must occur if the mind is physical.
Yes, of course. The word is an arbitrary symbol to which meaning is attached by your brain (and within a speech community). I don’t think that’s relevant. When we think in words, as I do a lot, and you presumably do also, those words have determinative meaning. How that meaning is attached to the words is a question for neural scientists, I suppose. But what reason is there to think that it’s not physical?
Yes. I would suppose that a syllogism isn’t evidence, and premises and conclusions aren’t evidence. The evidence there would be whatever suggests that a premise is true. Could you give an example of non-empirical evidence?
That seems a bit more condescending than is warranted, as the rest of that paragraph and the two following.
That would indicate that Feser’s argument really is unsound. A logical possibility is simply insufficient - if determinacy in Feser’s sense is necessary, then it must be the case that anything falling short of that must have problems. If it almost certainly won’t, then Feser’s argument can’t be regarded as a success.
No, I am not arguing for the falsity of Premise 2, I am arguing for the inadequacy of the examples supporting Premise 2 in that they do not demonstrate a genuine problem for the mind (which argues for the falsity of Premise 1). The mere possibility of alternative interpretations does not show that such reinterpretations are actually possible within the mind. I have no doubt that someone else could look at the state and invent another interpretation. I just don’t see how that poses any problem at all.
It’s perfectly possible to invent alternate interpretations of DNA sequences or computer binaries. Reproduction still works and programs still reliably run (bugs excepted) because those alternative interpretations aren’t relevant to those processes.
So, the semantic indeterminacy arguments referenced under Premise 2 are not arguing that our thoughts about adding and “quadding” are in fact associated with the same physical state. Rather, they are arguing that there is nothing about the word “adding”, even considering the context of its use, that can determine that it definitely means adding and not something like “quadding” (even if you point out that our use of the word adding is associated with a particular algorithm rather than merely a finite set of examples, you can re-run the same kind of arguments for whatever terms you use to describe the algorithm, and so on) More generally, you can re-run the same kind of arguments for every word we ever use; no matter how much context you add, you can always keep on coming up with some alternative interpretation. (Yes, the reinterpretation will probably be weird, like Quine’s “undetatched rabbit part” example; but the point is that the facts about the word and it’s use in context are not sufficient to decide between the weird meaning and the one that seems more likely to us.)
The reasoning goes on to say that just as there is no intrinsic connection between patterns of sound waves and any particular meaning, there is no intrinsic connection between patterns of neurons firing (or any other physical system or pattern) and any particular meaning. But then the same kind of arguments work for physical patterns as for words; this gets us to Premise 2.
The semantic indeterminacy arguments don’t work if you start with something that does have intrinsic meaning, something the very nature of which is to have a particular semantic content - something like that just means what it means; it isn’t subject to reinterpretation. But given the contingent connection between neural structures and meanings (I doubt anyone here is suggesting that any time two people have the same meaning in mind, we could find exactly identical neural structures in their brains, especially given phenomena like neuroplasticity), such a thing can’t be physical.
If premise 1 is true, then premise 2 is true if and only if thoughts are not physical. But ‘thoughts are not physical’ is the thing to be demonstrated, so if premise 1 is true, then premise 2 is begging the question.
If premise 2 is true AND not begging the question, then it contradicts premise 1, or else forces premise 1 to beg the question instead. In other words, the only way to accept the argument is if you already think the argument is correct, so the argument must necessarily be incorrect. It’s just bad logic.
Not surprising, coming as it does from someone who evidently doesn’t know what the word ‘or’ means.
What I wrote was not directed at you personally, but at the reasoning you gave for why you found it impossible to take the argument seriously (and the implicit scientism therein). It is my explanation for why I find it impossible to take your dismissal of the argument seriously in turn.
Your supposition presumes physicalism, but anyways, the reasons are the reasons I’ve given for the premises of the argument. “What attaches meaning to words is not physical” is not a presupposition of the argument, but its conclusion.
(My most recent reply to @misterme987 contains an attempt to better articulate the support for Premise 2 in particular.)
None of the following can be justified purely empirically:
the laws of logic which let us transfer evidential force from premises to conclusions
rational intuitions underpinning our use of inductive and abductive reasoning
moral intuitions which factor into our moral and ethical beliefs
metaphysical intuitions which go into the kind of arguments for the reality of the external world or causation, which ultimately underlie conclusions like “we have evidence that brains exist”
And that is a discussion that I am sure has happened on this forum before, probably more than once. That is about as far as I feel like retreading on that ground.
Logically speaking, all that Premise 1 requires is that we have a single thought with unambiguous meaning. I think the example of adding is as good as any; I know exactly what I mean when I think about the operation of adding two natural numbers.
If you have a thought that you know has meaning A, then that thought unambiguously has meaning A. If your thought had indeterminate semantic content, it wouldn’t mean A - the meaning would be indeterminate between A or B or C and so on. Your belief that it meant A would be false, and therefore would not constitute knowledge.
The threat here is not epistemological, but metaphysical. The problem is that if your thought really is constituted by some physical structure, but the facts about that structure do not fix one particular meaning out of A or B or C, then there is just nothing to make it the case that your thought has meaning A rather than B or C.
And this is why all the argument requires is that any physical representation of meaning is going to have a logically possible reinterpretation - if such is possible, it shows that the facts about the representation do not logically entail the intended meaning, and so the facts do not fix one particular meaning.
Completely besides the point: DNA sequences don’t have semantic content (except where researchers make artificial DNA with some intended translation from groups of nucleotide bases to ASCII or something), and even if they did, their semantic content would be irrelevant to the physical process. Same with running a computer program: if the bits have any semantic content (say, they are intended by the programmer as a representation of a triangle), that meaning is irrelevant to how the program actually runs. So of course the alternative interpretation is also irrelevant.
But if, say, the outputs of the program are supposed to have semantic content, alternative interpretations for the internal states could imply alternative interpretations for the outputs. Or maybe more to the point: if you were telling me that the program constituted a genuine artificial intelligence, and that this AI had a thought with some determinate meaning constituted by that string of 1s and 0s, the reinterpretability of that bitstring would raise the question of how your claim could possibly be correct.
No, it presumes that we have no reason to reject physicalism, not quite the same thing.
I maintain that in order for the argument to work you must assume the conclusion.
Are you claiming that these are non-empirical evidence? Or is that not an attempt to answer my question? And I would say that some of these assumptions can indeed by justified by empirical evidence, i.e. that they appear to lead to conclusions that can be independently arrived at in other ways. This of course further assumes that consistency and consilience are useful criteria. Not sure what any of that has to do with non-empirical evidence.