Fun article. I’m with you on these silly word games, and w/ a little philosophy as a way out. But I think focusing on conceptual disaggregation w/r/t the answers isn’t always the solution. Like, you write on the “tree and sound” problem: “The reason this question sounds difficult is that the word “sound” actually covers two separate phenomena that coincide in every case we observe. The one is the auditory experience, and the other is the physical phenomena that cause the auditory experience (pressure waves and all that).” Do we really have a concept of “sound” as an auditory experience at all??? It seems in the grammar of “sound” that it’s a thing that you can hear. (Otherwise, we can say something like “inaudible sound.”) I can’t think of a “sound” of which it wouldn’t make sense to say, “I heard/didn’t hear the sound.” In other words, it doesn’t seem like “hearing” is ever built into the concept of sound. —The “tree and sound” problem is obv silly. But isn’t it because the question is silly? The answers are only silly because they take the question seriously. What the actual fuck does someone who asks “is the tree making a sound?” want to know?? The question is almost essentially “philosophical.” It has no ordinary context, at least for people who know what trees are and sounds are. I feel like philosophy, or at least folk philosophy, went astray by permitting questions that nobody actually knows what it would mean to actually ask. Imagining these questions as asked by kids personally helps me, because they might actually be able to mean them. “Do trees make sounds if nobody hears?” Of course, kiddo! And then you explain to them what sounds are, or whatever.
Thanks! I may not quite be following you, but it sounds to me like you're just defending the "physical phenomena" specification. That's of course fine, and I have no objection to going with one route over the other necessarily. Though I also don't think someone who takes the "auditory experience" route is conceptually confused. For example, if I have an auditory hallucination, it doesn't seem particularly odd to say that it was a sound I heard. Also, just when referring to a "sound" it seems to me like I can just as well mean the experience I'm having as the thing that caused the experience.
But that's really unimportant, and (as I say) a non-substantial disagreement. You may be making a further point too that I'm missing.
I wouldn’t defend the physical phenomena specification! You’re right about the hallucination thing. I’m just defending the idea that sounds are always things we are in a position to hear—even hallucinations. But it’s a quibble, as you say. The more important thing is where the word game goes wrong: in the answers? or in the posing of the problem?
I'm not sure that they go wrong anywhere per se. I mean, it's not clear that a question can "go wrong" as such. Rather I think what is wrong is thinking that there is a genuine mystery to be had, and in there where it goes wrong is in not being precise about the terms involved. Whether that's in the question or the answer, I'm not sure, but if you really want to choose one, I guess the question.
I do have one slight objection to this: I think conceptual analysis can be done even on vague properties to delimit the boundaries of meaning. For example, although it is vague what counts as "red", a laser with a wavelength of 500 nm is unambiguously *not* red. So any attempt to define what it means that be red that includes 500 nm light in the category of red things can be rejected as definitely not what we mean by the word "red". The Gettier problem is an example of this sort of argument in philosophy: Although the concept of knowledge is probably too imprecise for philosophers to ever succeed at unambiguously defining it, it seems that just about everyone shares the judgement that, in Gettier cases, a person does not have knowledge, and therefore, analyses of knowledge that imply that we would have knowledge in Gettier cases are no longer candidate meanings. They can be ousted from the "superposition" of possible precisifications of the term.
A more controversial example: I think that there are probably many possible meanings of the term "free will," but I come down on a definite side of the compatibilist-incompatibilist debate. I'm a compatibilist because I think that there are coherent conceptions of free will (i.e., possible specific properties that would fit with what we mean by the term), but there are no coherent conceptions of libertarian free will. Thus, any coherent conception of free will must be a compatibilist one.
I suppose I agree to some extent. I agree in the sense that there are certain specifications of a term that would be extremely strange as they would clash with what anyone would ever want to communicate with the term. Thus any specification which includes these strange examples (like 500 nm light, or Gettier cases) would be "wrong" in the sense that they would not be capturing what we want to communicate with the term.
But I'm not sure I would say that they are strictly speaking wrong. For them to be wrong, there would need to be some class of not-wrong specifications and not-not-included instances of the term, and I just think that isn't the case. Like with something like knowledge, for a Gettier cases to be definitively not knowledge, there would still need to be *some* non-vague class of definitely-not-knowledge. But I just don't think that's the case--it's vagueness all the way down.
Putting it another way, it seems that your view would require that there is at least some set of necessary conditions for the term in question that are objectively right. If this weren't the case, well then no case could be definitively excluded, as being definitively excluded just means not living up to the necessary conditions for counting as an instance (I don't know how persuasive this is, it sounds sorta circular as I'm writing it). But I want to deny that there are any objectively right necessary conditions for the term. Firstly because I just can't see why we should expect there to be such, or what would explain why they were these rather than others; and secondly because I just don't see uncontroversial necessary conditions being formulated (I guess with knowledge something like being a belief might count, but that's very minimal, and I would still want to say that it isn't *objectively* right).
I'm actually not sure it's a problem for the line between "unambiguously wrong" and "within the range of possible meanings of a vague term" to itself be imprecise.
Well, I think that unless you have some non-vagueness at some level of description, you will be unable to distinguish anything. Say that 700 nm is definitely red and 500 nm definitely isn't, but the cutoff is vague. That will mean that there will be some set of wavelengths for which it is vague whether they are red or not.
But what is this set of wavelengths? If it is well-defined then that seems just as arbitrary as picking a definite wavelength as the cutoff for red--why is this wavelength definitely-red and the wavelength 0.00000001 nm shorter not-definitely-red. Alternatively it's itself vague (as you suggest) which suggests that there is some set of values which are not-definitely-red and also not-definitely-vague. But then the regress continues.
If you never reach a point where there is some definite cutoff, then I don't see how you can justify your assumption that 500 nm definitely isn't red and 700 nm definitely is. If there is never any point where any boundary is clearly drawn, then there just seems to be no things that are clearly within the concept or outside it.
And if we do reach a sharp cutoff at some point, then I would want to know what determines/grounds its value. Its hard for me to see what could be the reason for there being some very clear sharp cutoff in the application of a word that we just made up as we went along. It seems much more simple to me to just say that there really isn't any definite fact of the matter as to what the words we use pick out. Rather there are just certain ways of precisifying them that would be more or less useful/clear for communication.
Perhaps you want to say that the vague vagueness-cutoffs somehow asymptotically approach some value, lower than which all things are definitely not red, but this just looks like an equally arbitrary and ungrounded choice--why isn't this asymptote slightly higher or lower?
"Now, I think that we should clearly map the word “woman” onto the social category, since not doing so would just make trans people unnecessarily miserable for no reason."
The meaning of language isn't determined by the necessity of misery of certain groups, but by the nature of thought.
I discuss this question of language (and answer "what is a woman") as well as what knowledge is (addressing Gettier problems) below.
Fun article. I’m with you on these silly word games, and w/ a little philosophy as a way out. But I think focusing on conceptual disaggregation w/r/t the answers isn’t always the solution. Like, you write on the “tree and sound” problem: “The reason this question sounds difficult is that the word “sound” actually covers two separate phenomena that coincide in every case we observe. The one is the auditory experience, and the other is the physical phenomena that cause the auditory experience (pressure waves and all that).” Do we really have a concept of “sound” as an auditory experience at all??? It seems in the grammar of “sound” that it’s a thing that you can hear. (Otherwise, we can say something like “inaudible sound.”) I can’t think of a “sound” of which it wouldn’t make sense to say, “I heard/didn’t hear the sound.” In other words, it doesn’t seem like “hearing” is ever built into the concept of sound. —The “tree and sound” problem is obv silly. But isn’t it because the question is silly? The answers are only silly because they take the question seriously. What the actual fuck does someone who asks “is the tree making a sound?” want to know?? The question is almost essentially “philosophical.” It has no ordinary context, at least for people who know what trees are and sounds are. I feel like philosophy, or at least folk philosophy, went astray by permitting questions that nobody actually knows what it would mean to actually ask. Imagining these questions as asked by kids personally helps me, because they might actually be able to mean them. “Do trees make sounds if nobody hears?” Of course, kiddo! And then you explain to them what sounds are, or whatever.
Thanks! I may not quite be following you, but it sounds to me like you're just defending the "physical phenomena" specification. That's of course fine, and I have no objection to going with one route over the other necessarily. Though I also don't think someone who takes the "auditory experience" route is conceptually confused. For example, if I have an auditory hallucination, it doesn't seem particularly odd to say that it was a sound I heard. Also, just when referring to a "sound" it seems to me like I can just as well mean the experience I'm having as the thing that caused the experience.
But that's really unimportant, and (as I say) a non-substantial disagreement. You may be making a further point too that I'm missing.
I wouldn’t defend the physical phenomena specification! You’re right about the hallucination thing. I’m just defending the idea that sounds are always things we are in a position to hear—even hallucinations. But it’s a quibble, as you say. The more important thing is where the word game goes wrong: in the answers? or in the posing of the problem?
I'm not sure that they go wrong anywhere per se. I mean, it's not clear that a question can "go wrong" as such. Rather I think what is wrong is thinking that there is a genuine mystery to be had, and in there where it goes wrong is in not being precise about the terms involved. Whether that's in the question or the answer, I'm not sure, but if you really want to choose one, I guess the question.
I do have one slight objection to this: I think conceptual analysis can be done even on vague properties to delimit the boundaries of meaning. For example, although it is vague what counts as "red", a laser with a wavelength of 500 nm is unambiguously *not* red. So any attempt to define what it means that be red that includes 500 nm light in the category of red things can be rejected as definitely not what we mean by the word "red". The Gettier problem is an example of this sort of argument in philosophy: Although the concept of knowledge is probably too imprecise for philosophers to ever succeed at unambiguously defining it, it seems that just about everyone shares the judgement that, in Gettier cases, a person does not have knowledge, and therefore, analyses of knowledge that imply that we would have knowledge in Gettier cases are no longer candidate meanings. They can be ousted from the "superposition" of possible precisifications of the term.
A more controversial example: I think that there are probably many possible meanings of the term "free will," but I come down on a definite side of the compatibilist-incompatibilist debate. I'm a compatibilist because I think that there are coherent conceptions of free will (i.e., possible specific properties that would fit with what we mean by the term), but there are no coherent conceptions of libertarian free will. Thus, any coherent conception of free will must be a compatibilist one.
I suppose I agree to some extent. I agree in the sense that there are certain specifications of a term that would be extremely strange as they would clash with what anyone would ever want to communicate with the term. Thus any specification which includes these strange examples (like 500 nm light, or Gettier cases) would be "wrong" in the sense that they would not be capturing what we want to communicate with the term.
But I'm not sure I would say that they are strictly speaking wrong. For them to be wrong, there would need to be some class of not-wrong specifications and not-not-included instances of the term, and I just think that isn't the case. Like with something like knowledge, for a Gettier cases to be definitively not knowledge, there would still need to be *some* non-vague class of definitely-not-knowledge. But I just don't think that's the case--it's vagueness all the way down.
Putting it another way, it seems that your view would require that there is at least some set of necessary conditions for the term in question that are objectively right. If this weren't the case, well then no case could be definitively excluded, as being definitively excluded just means not living up to the necessary conditions for counting as an instance (I don't know how persuasive this is, it sounds sorta circular as I'm writing it). But I want to deny that there are any objectively right necessary conditions for the term. Firstly because I just can't see why we should expect there to be such, or what would explain why they were these rather than others; and secondly because I just don't see uncontroversial necessary conditions being formulated (I guess with knowledge something like being a belief might count, but that's very minimal, and I would still want to say that it isn't *objectively* right).
I'm actually not sure it's a problem for the line between "unambiguously wrong" and "within the range of possible meanings of a vague term" to itself be imprecise.
Well, I think that unless you have some non-vagueness at some level of description, you will be unable to distinguish anything. Say that 700 nm is definitely red and 500 nm definitely isn't, but the cutoff is vague. That will mean that there will be some set of wavelengths for which it is vague whether they are red or not.
But what is this set of wavelengths? If it is well-defined then that seems just as arbitrary as picking a definite wavelength as the cutoff for red--why is this wavelength definitely-red and the wavelength 0.00000001 nm shorter not-definitely-red. Alternatively it's itself vague (as you suggest) which suggests that there is some set of values which are not-definitely-red and also not-definitely-vague. But then the regress continues.
If you never reach a point where there is some definite cutoff, then I don't see how you can justify your assumption that 500 nm definitely isn't red and 700 nm definitely is. If there is never any point where any boundary is clearly drawn, then there just seems to be no things that are clearly within the concept or outside it.
And if we do reach a sharp cutoff at some point, then I would want to know what determines/grounds its value. Its hard for me to see what could be the reason for there being some very clear sharp cutoff in the application of a word that we just made up as we went along. It seems much more simple to me to just say that there really isn't any definite fact of the matter as to what the words we use pick out. Rather there are just certain ways of precisifying them that would be more or less useful/clear for communication.
Perhaps you want to say that the vague vagueness-cutoffs somehow asymptotically approach some value, lower than which all things are definitely not red, but this just looks like an equally arbitrary and ungrounded choice--why isn't this asymptote slightly higher or lower?
"Now, I think that we should clearly map the word “woman” onto the social category, since not doing so would just make trans people unnecessarily miserable for no reason."
The meaning of language isn't determined by the necessity of misery of certain groups, but by the nature of thought.
I discuss this question of language (and answer "what is a woman") as well as what knowledge is (addressing Gettier problems) below.
https://neonomos.substack.com/p/sense-reference-and-what-is-a-woman
https://neonomos.substack.com/p/what-is-knowledge