Wilson's piece is dripping with metaphysical turns of phrase, but much of what she says could be re-expressed in a conceptual or linguistic key. I think this goes for a good deal of contemporary metaphysics. However, converting metaphysically-expressed ideas and claims into a conceptual or linguistic key may make them a bit fiddlier to think and express. And if you're doing hard philosophy and need to think and express a lot of things, this extra cost is going to pile up. Sometimes, having things in a conceptual or linguistic register may make things clearer, and sometimes it may be essential. But for many purposes the metaphysical register does fine, and often has the benefit of being less resource-hungry.
Yes, some metaphysics may not be capturable in conceptual or linguistic terms, and perhaps even in favourable cases the capturing will not be complete or perfect. And there are doubtless other important things going on behind the sociological phenomenon of the resurgence of metaphysics. But maybe this is part of the story.
UPDATE: Brandon Watson (at the end of a post on Fitch's paradox) links to this post, writing: 'I'm very interested, of course, in accounts of how philosophical scenes get transformed, how ideas transmogrify, and the like. This hypothesis for the rise of analytic metaphysics makes considerable amount of sense, and is probably true.' This is encouraging!
Hi Tristan, interesting thoughts. Expanding on that, I'd like to propose three aspects in which metaphysics both avoids and creates problems: Behaviorism, vagueness, and essentialism.
ReplyDelete- If we go full (late) wittgensteinian, and ask only "how do we use the terms of our language?" we avoid many metaphysical questions, in the way that they are (supposedly) created by misuse of language and therefore confused. But this easily lets us treat philosophy as mere language behaviorism, as a form of linguistic anti-philosophy. One could say that this seems to underestimate the expressibility of our philosophical language. Metaphysics avoids behaviorism in explicitly stating: I don't talk about language usage, I actually talk about reality.
- Metaphysics seems to naturally avoid vagueness: This makes things clear and understandable, it leads to clear cut distinctions, in the best case to categorical hierarchies which help to structure our reasoning. But this also can make us prone to assume implicitly that vagueness is merely secondary, that our simplifications aren't really simplifications, that behind the imprecision of our language there is some ultimate fact which makes (in the most extreme case) a heap either a heap or not. Williamson is such an extreme case. Many philosophers doing metaphysics would not go that far in clear cases like heaps, but they still may be tempted to do the same thing in other (more abstract) cases where it is much less obvious that borderline cases might actually be a product of semantic vagueness than missing metaphysical knowledge.
- Metaphysics sometimes seems to lead to "too much essentialism": We then assume that there is a way in which things really are even when we are using (to some extent) arbitrary conceptualizations. We then treat necessary/sufficient conditions of our concepts as if they where essential properties of the things we name. The opposite would be "too little essentialism", e.g. in a form of extreme constructivism we find sometimes in the social sciences: When we think that some things only depend on our arbitrary concepts even though there actually is a difference in the world which is reflected by the difference we make in our language.
Thanks for this intelligent comment.
DeleteBy the way, in your third point, the middle sentence ('We then treat...') doesn't seem, to me, to be about the same phenomenon. To invoke an example close to my heart and relevant to philosophy of language (especially Kripke's puzzle), consider questions about whether some X and Y mean the same as each other or not. I think that sometimes there is no single context-independent answer - that 'means the same as' exhibits a certain kind of flexibility. (I've discussed this under the heading 'semantic granularity' in my PhD thesis and on this blog.)
DeleteOn this kind of view there is a kind of philosophical mistake in asking, after having considered both sides of the issue, whether or not some pair of expressions or expression occurrences mean the same or not, in the sort of case where one answer is natural in some situation or from one point of view, but the opposite answer is natural from the other - for instance, Kripke's case of Pierre's different attitudes to 'Londres est jolie' and 'London is pretty'.
This sort of mistake seems to be closely related to what you're talking about in the first sentence of your third point and again at the end, and seems definitely to be a type of instance of your second point. But then the second sentence of the third point doesn't seem to apply.
All in all, I think it would be good to separate things out a bit in the "too much essentialism" category. There seem to be a few quite different things going on there.
Regarding talk about different kinds of concepts, and whether the difference is arbitrary or reflects something in the world, I think this is very confusing and runs different things together - talking generally now, not just about what you wrote.
Sorry for the late reply. Yes, the London/Londres case seems to be more a issue of vagueness or of granularity.
DeleteMy third point indeed wasn't very clear. What I wanted to highlight:
How we "conceptually divide" the world depends to a great extent on us as language users, but it isn't entirely arbitrary either. Language is a means of communication, it has to be useful. We therefore draw with many concepts distinctions which correspond to differences in the world. Social constructs are exceptions to this. There is no special physical difference which corresponds to how we distinguish some sheets of paper we call "money" from others which we call "counterfeit money". Similarly for professors and non-professors, a nation and foreign nations etc. In other cases (e.g. gender theory) it is a matter of controversy to which extent a distinction is a social construct and to which extent it corresponds to a difference in the world. E.g. some biologist might say that there is a difference in the world which corresponds to a distinction which some social scientist might take to be a mere social construct. But nonetheless it isn't controversial to state that our concepts (at least) often mark differences in the world. Because our concepts depend in part on differences in the world. An example is the difference between our concept of "dog" and our concept of "animal". Being an animal is a necessary (though not a sufficient) condition for us to apply the term "dog" to something.
But one could take things in a completely reversed manner. Instead of speaking of necessary conditions for a concept, one could speak of necessary properties of a class of things. Note that "necessary" means something very different here. When we talk about conditions, we distinguish between those which are necessary, sufficient, both, or neither. When someone talks about properties, this person probably distinguishes between necessary ones and contingent ones. The latter distinction is the same which is usually applied to propositions, where we say a proposition is either necessary, contingent, or impossible. I don't see a good reason to postulate the existence of "necessary properties": Necessary conditions of concepts seem to be sufficient.
Take these examples:
(1) All dogs are animals.
(2) All bachelors are humans.
It seems as if they talk about things in the world. But they are just about necessary conditions of the concept of "animal" and "bachelor" respectively. The apparent difference between the two stems from the fact that the concept of "bachelor" is a social construct. This becomes even clearer when we consider this example:
(3) All bachelors are unmarried.
Now the concept of "being married" (or generally "marriage") is a social construct, too.
A quite different example is this sentence:
Delete(4) All humans are mortal.
I think it was Kant who took this to be analytic. Russel said it is synthetic. But was there disagreement? For Kant was "being mortal" a necessary condition to apply the term "human". For Russel this was not the case. They where talking about the same sentence, but for Russel it expressed a different proposition than for Kant, because they didn't have the same concept of "human".
For me the case is different. I'm unsure whether I would call a being, which is immortal but otherwise looks like a human, still "human". My concept of "human" is vague. One could say (4) therefore expresses a vague proposition for me. One which is half analytic and half synthetic. And therefore half necessary and half contingent. I think this makes perfectly sense. But I don't think it makes a lot of sense to talk about "vague properties" which humans have. Even less to talk about properties which are partly necessary and partly contingent. At least it isn't necessary to postulate such "meta properties" (vagueness, necessity/contingency) which properties could have.
Thank youu for sharing this
ReplyDelete