I think I have one counterargument to this: Most terms philosophers do conceptual analysis on are ones that they have decided identify some important concept, and the conceptual analysis they do is actually already pretty similar to conceptual engineering as you've described it. Philosophers could just stipulate that knowledge is defined as justified true belief, for example, but in light of Gettier cases, they've decided that that definition doesn't fully get at the concept they want to describe. It doesn't seem useful to describe cases where someone justifiably believes a true proposition by accident, only because they justifiably believe a false proposition that entails it, as knowledge, so philosophers look for a better definition of the concept. The same could be said for, say, free will. I think there are many possible definitions that all roughly fit the concept because the concept is ambiguous, but some might be more useful than others. One of the reasons I'm a compatibilist is because I think libertarian free will is metaphysically impossible by any definition that even comes close to fitting the intuitive concept of free will, and therefore any useful definition of free will must be compatibilist.
Well, I'm not sure that's actually disagreeing much with my position. I think that in many cases conceptual engineering and conceptual analysis will look quite similar, since words like knowledge and free will have often been introduced because they capture something we care about, and we should try to get clear on what exactly that is. The difference is that I would dispute that there is a right answer to be found--the origins of the words are such that we shouldn't expect them to pick out any precise concept we could find.
In many cases this won't matter, but it many cases it also will. For example, if we can't find a precise definition of knowledge that captures all our intuitions, we should just settle for something that does well enough. In fact I doubt that any definition would avoid any intuitive counterexamples.
Another example is "disease." There's apparently a pretty significant literature devoted to analyzing this concept. But I think this is misguided. We shouldn't be trying to capture our conceptual or linguistic intuitions when figuring out what to count as a disease--rather we should figure out what features we actually care about capturing with the concept, and then try to define it accordingly.
Yeah, I'm not so much disagreeing with the usefulness of conceptual engineering - I guess the disagreement is really just, "It might not be such a revolutionary idea if current philosophical practice is already doing it more than you think."
I definitely agree on disease. For some reason it feels even more like a concept that ought to be conceptually engineered than knowledge or free will to me, maybe because it's not a psychological concept and therefore I don't have the intuitive feeling of, "No, this is what it's pointing to and I can tell by introspection," and maybe because it's more obvious that what counts as a disease is defined by convention given how much the standards for disease have changed in the past.
I think I have one counterargument to this: Most terms philosophers do conceptual analysis on are ones that they have decided identify some important concept, and the conceptual analysis they do is actually already pretty similar to conceptual engineering as you've described it. Philosophers could just stipulate that knowledge is defined as justified true belief, for example, but in light of Gettier cases, they've decided that that definition doesn't fully get at the concept they want to describe. It doesn't seem useful to describe cases where someone justifiably believes a true proposition by accident, only because they justifiably believe a false proposition that entails it, as knowledge, so philosophers look for a better definition of the concept. The same could be said for, say, free will. I think there are many possible definitions that all roughly fit the concept because the concept is ambiguous, but some might be more useful than others. One of the reasons I'm a compatibilist is because I think libertarian free will is metaphysically impossible by any definition that even comes close to fitting the intuitive concept of free will, and therefore any useful definition of free will must be compatibilist.
Well, I'm not sure that's actually disagreeing much with my position. I think that in many cases conceptual engineering and conceptual analysis will look quite similar, since words like knowledge and free will have often been introduced because they capture something we care about, and we should try to get clear on what exactly that is. The difference is that I would dispute that there is a right answer to be found--the origins of the words are such that we shouldn't expect them to pick out any precise concept we could find.
In many cases this won't matter, but it many cases it also will. For example, if we can't find a precise definition of knowledge that captures all our intuitions, we should just settle for something that does well enough. In fact I doubt that any definition would avoid any intuitive counterexamples.
Another example is "disease." There's apparently a pretty significant literature devoted to analyzing this concept. But I think this is misguided. We shouldn't be trying to capture our conceptual or linguistic intuitions when figuring out what to count as a disease--rather we should figure out what features we actually care about capturing with the concept, and then try to define it accordingly.
Yeah, I'm not so much disagreeing with the usefulness of conceptual engineering - I guess the disagreement is really just, "It might not be such a revolutionary idea if current philosophical practice is already doing it more than you think."
I definitely agree on disease. For some reason it feels even more like a concept that ought to be conceptually engineered than knowledge or free will to me, maybe because it's not a psychological concept and therefore I don't have the intuitive feeling of, "No, this is what it's pointing to and I can tell by introspection," and maybe because it's more obvious that what counts as a disease is defined by convention given how much the standards for disease have changed in the past.