I’ve alluded to this in other comments, but I think worth spelling out more comprehensively here.
I think this post makes a few main points:
Categories are not arbitrary. You might need different categories for different purposes, but categories are for helping you think about the things you care about, and a category that doesn’t correspond to the territory will be less helpful for thinking and communciating.
Some categories might sort of look like they correspond to something in reality, but they are gerrymandered in a way optimized for deception.
You might sometimes wish something were a member of a category that it isn’t, and it is better to admit that so that you can actually communicate about the current state of reality.
I realize the three points cleave together pretty closely in the author’s model, and make sense to think about in conjunction. But I think trying to introduce them all at once makes for more confusing reading.
I think the followup post Unnatural Categories Are Optimized For Deception does a pretty good job of spelling out the details of points #2 and #3. I think the current post does a good job at #1, a decent job at #2, but a fairly confused job at #3.
In particular, these 2.5 paragraphs feel like a meandering vagueblog. I know what point they’re trying to make, but by trying to avoid the object level political disagreement, the post leaves me very confused about why I might be making these particular mistakes, or what to do about it.
If what you want isn’t currently true in reality, maybe there’s some action you could take to make it become true. To search for that action, you’re going to need accurate beliefs about what reality is currently like. To enlist the help of others in your planning, you’re going to need precise terminology to communicate accurate beliefs about what reality is currently like. Even when—especially when—the current reality is inconvenient.
(Oh, and if you’re actually trying to optimize other people’s models of the world, rather than the world itself—you could just lie, rather than playing clever category-gerrymandering mind games. It would be a lot simpler!)
(By contrast, the Unnatural Categories gives concrete real world examples that explain why you’d make this particular class of mistake. Somewhat oddly, it does use some politically charged examples, but I think it does a good job of laying out in as gearsy, not-too-politicized fashion. I think it actually could probably have gotten away with directly involving the original motivating example for the series)
If the point of Unnatural Categories is to replace this post, and this post is more like a first draft, then… seems potentially fine for Unnatural Categories to become the new canonical version of it and not worry overmuch about this one.
But, I think there’s a lot of distinct concepts here. Breaking them into multiple posts that deal with theseems worthwhile to me.
If this post is appearing in the 2019 Review as a standalone piece, I think it’d be clearer if it just cut the paragraphs I listed (and then reorganized itself slightly), rather than doing a rough, vague job of explaining them. When I got to like “even if it hurts” my reaction was “what? why would it hurt? what are you talking about? I think I know what the underlying political argument is and I’m still kinda confused about what’s going on here.”
I’ve also mentioned elsethread that I think the Bob the Vice President of Sorting example would be helpful to have earlier in the piece, to give a clearer example of when this whole problem might come up. But I realize people may vary in what pedagogy works best for them.
...
A sidepoint I notice while thinking about this: when I go back to the older sequence posts that this essay is referencing...
...well, they totally are vague and don’t spell out what real world examples you might run into that would motivate the philosophical confusion. But they are also much shorter, usually focus on one idea at a time instead of three, and are intermixed with other posts that do lay out more of the motivating-examples.
I’ve alluded to this in other comments, but I think worth spelling out more comprehensively here.
I think this post makes a few main points:
Categories are not arbitrary. You might need different categories for different purposes, but categories are for helping you think about the things you care about, and a category that doesn’t correspond to the territory will be less helpful for thinking and communciating.
Some categories might sort of look like they correspond to something in reality, but they are gerrymandered in a way optimized for deception.
You might sometimes wish something were a member of a category that it isn’t, and it is better to admit that so that you can actually communicate about the current state of reality.
I realize the three points cleave together pretty closely in the author’s model, and make sense to think about in conjunction. But I think trying to introduce them all at once makes for more confusing reading.
I think the followup post Unnatural Categories Are Optimized For Deception does a pretty good job of spelling out the details of points #2 and #3. I think the current post does a good job at #1, a decent job at #2, but a fairly confused job at #3.
In particular, these 2.5 paragraphs feel like a meandering vagueblog. I know what point they’re trying to make, but by trying to avoid the object level political disagreement, the post leaves me very confused about why I might be making these particular mistakes, or what to do about it.
(By contrast, the Unnatural Categories gives concrete real world examples that explain why you’d make this particular class of mistake. Somewhat oddly, it does use some politically charged examples, but I think it does a good job of laying out in as gearsy, not-too-politicized fashion. I think it actually could probably have gotten away with directly involving the original motivating example for the series)
If the point of Unnatural Categories is to replace this post, and this post is more like a first draft, then… seems potentially fine for Unnatural Categories to become the new canonical version of it and not worry overmuch about this one.
But, I think there’s a lot of distinct concepts here. Breaking them into multiple posts that deal with theseems worthwhile to me.
If this post is appearing in the 2019 Review as a standalone piece, I think it’d be clearer if it just cut the paragraphs I listed (and then reorganized itself slightly), rather than doing a rough, vague job of explaining them. When I got to like “even if it hurts” my reaction was “what? why would it hurt? what are you talking about? I think I know what the underlying political argument is and I’m still kinda confused about what’s going on here.”
I’ve also mentioned elsethread that I think the Bob the Vice President of Sorting example would be helpful to have earlier in the piece, to give a clearer example of when this whole problem might come up. But I realize people may vary in what pedagogy works best for them.
...
A sidepoint I notice while thinking about this: when I go back to the older sequence posts that this essay is referencing...
...well, they totally are vague and don’t spell out what real world examples you might run into that would motivate the philosophical confusion. But they are also much shorter, usually focus on one idea at a time instead of three, and are intermixed with other posts that do lay out more of the motivating-examples.