I actually disagree with this. I haven’t thought too hard about it and might just not be seeing it, but on first thought I am not really seeing how such evidence would make the post “much stronger”.
To elaborate, I like to use Paul Graham’s Disagreement Hierarchy as a lens to look through for the question of how strong a post is. In particular, I like to focus pretty hard on the central point (DH6) rather than supporting and tangential points. I think the central point plays a very large role in determining how strong a post is.
Here, my interpretation of the central point(s) is something like this:
Poverty is largely determined by the weakest link in the chain.
Anoxan is a helpful example to illustrate this.
It’s not too clear what drives poverty today, and so it’s not too clear that UBI would meaningfully reduce poverty.
I thought the post did a nice job of making those central points. Sure, something like a survey of the research in positive psychology could provide more support for point #1, for example, but I dunno, I found the sort of intuitive argument for point #1 to be pretty strong, I’m pretty persuaded by it, and so I don’t think I’d update too hard in response to the survey of positive psychology research.
Another thing I think about when asking myself how strong I think a post is is how “far along” it is. Is it an off the cuff conversation starter? An informal write up of something that’s been moderately refined? A formal write up of something that has been significantly refined?
I think this post was somewhere towards the beginning of the spectrum (note: it was originally a tweet, not a LessWrong post). So then, for things like citations supporting empirical claims, I don’t think it’s reasonable to expect very much from the author, and so I lean away from viewing the lack of citations as something that (meaningfully) weakens the post.
I don’t think it’s reasonable to expect very much from the author, and so I lean away from viewing the lack of citations as something that (meaningfully) weakens the post.
I feel like our expectations of the author and the circumstances of the authorship can inform our opinions of how “blameworthy” the author is for not improving the post in some way, but shouldn’t really have any relevance to what changes would be improvements if they occurred. The latter seems to me to purely be a claim about the text of the post, not a claim about the process that wrote it.
Hm. I hear ya. Good point. I’m not sure whether I agree or disagree.
I’m trying to think of an analogy and came up with the following. Imagine you go to McDonalds with some friends and someone comments that their burger would be better if they used prime ribeye for their ground beef.
I guess it’s technically true, but something also feels off about it to me that I’m having trouble putting my finger on. Maybe it’s that it feels like a moot point to discuss things that would make something better that are also impractical to implement.
In particular, I like to focus pretty hard on the central point (DH6) rather than supporting and tangential points.
This is a recipe for Gish gallops.
It also leads to Schrodinger’s importance, where a point is important right up until someone looks at it and shows that it’s poorly supported, whereupon it’s suddenly unimportant. If it’s important enough to use, it’s important enough to be refuted.
The Gish gallop (/ˈɡɪʃ ˈɡæləp/) is a rhetorical technique in which a person in a debate attempts to overwhelm an opponent by abandoning formal debating principles, providing an excessive number of arguments with no regard for the accuracy or strength of those arguments and that are impossible to address adequately in the time allotted to the opponent. Gish galloping prioritizes the quantity of the galloper’s arguments at the expense of their quality.
I disagree that focusing on the central point is a recipe for Gish gallops and that it leads to Schrodinger’s importance.
Well, I think that it in combination with a bunch of other poor epistemic norms it might be a recipe for those things, but a) not by itself and b) I think the norms would have to be pretty poor. Like, I don’t expect that you need 10⁄10 level epistemic norms in the presence of focusing on the central point to shield from those failure modes, I think you just need something more like 3⁄10 level epistemic norms. Here on LessWrong I think our epistemic norms are strong enough where focusing on the central point doesn’t put us at risk of things like Gish gallops and Schrodinger’s importance.
Focusing on the “central point” in the midst of a lot of other “unimportant” points is a recipe for Gish gallops because you can claim that any point which has been refuted is an unimportant one. This forces your questioner to keep refuting point after point until you run out of them. That amounts to a Gish gallop.
If the point was important enough to strengthen your argument—and presumably it was or you wouldn’t have used it—it’s important enough that refuting it weakens the argument.
I actually disagree with this. I haven’t thought too hard about it and might just not be seeing it, but on first thought I am not really seeing how such evidence would make the post “much stronger”.
To elaborate, I like to use Paul Graham’s Disagreement Hierarchy as a lens to look through for the question of how strong a post is. In particular, I like to focus pretty hard on the central point (DH6) rather than supporting and tangential points. I think the central point plays a very large role in determining how strong a post is.
Here, my interpretation of the central point(s) is something like this:
Poverty is largely determined by the weakest link in the chain.
Anoxan is a helpful example to illustrate this.
It’s not too clear what drives poverty today, and so it’s not too clear that UBI would meaningfully reduce poverty.
I thought the post did a nice job of making those central points. Sure, something like a survey of the research in positive psychology could provide more support for point #1, for example, but I dunno, I found the sort of intuitive argument for point #1 to be pretty strong, I’m pretty persuaded by it, and so I don’t think I’d update too hard in response to the survey of positive psychology research.
Another thing I think about when asking myself how strong I think a post is is how “far along” it is. Is it an off the cuff conversation starter? An informal write up of something that’s been moderately refined? A formal write up of something that has been significantly refined?
I think this post was somewhere towards the beginning of the spectrum (note: it was originally a tweet, not a LessWrong post). So then, for things like citations supporting empirical claims, I don’t think it’s reasonable to expect very much from the author, and so I lean away from viewing the lack of citations as something that (meaningfully) weakens the post.
I feel like our expectations of the author and the circumstances of the authorship can inform our opinions of how “blameworthy” the author is for not improving the post in some way, but shouldn’t really have any relevance to what changes would be improvements if they occurred. The latter seems to me to purely be a claim about the text of the post, not a claim about the process that wrote it.
Hm. I hear ya. Good point. I’m not sure whether I agree or disagree.
I’m trying to think of an analogy and came up with the following. Imagine you go to McDonalds with some friends and someone comments that their burger would be better if they used prime ribeye for their ground beef.
I guess it’s technically true, but something also feels off about it to me that I’m having trouble putting my finger on. Maybe it’s that it feels like a moot point to discuss things that would make something better that are also impractical to implement.
This is a recipe for Gish gallops.
It also leads to Schrodinger’s importance, where a point is important right up until someone looks at it and shows that it’s poorly supported, whereupon it’s suddenly unimportant. If it’s important enough to use, it’s important enough to be refuted.
I just looked up Gish gallops on Wikipedia. Here’s the first paragraph:
I disagree that focusing on the central point is a recipe for Gish gallops and that it leads to Schrodinger’s importance.
Well, I think that it in combination with a bunch of other poor epistemic norms it might be a recipe for those things, but a) not by itself and b) I think the norms would have to be pretty poor. Like, I don’t expect that you need 10⁄10 level epistemic norms in the presence of focusing on the central point to shield from those failure modes, I think you just need something more like 3⁄10 level epistemic norms. Here on LessWrong I think our epistemic norms are strong enough where focusing on the central point doesn’t put us at risk of things like Gish gallops and Schrodinger’s importance.
Focusing on the “central point” in the midst of a lot of other “unimportant” points is a recipe for Gish gallops because you can claim that any point which has been refuted is an unimportant one. This forces your questioner to keep refuting point after point until you run out of them. That amounts to a Gish gallop.
If the point was important enough to strengthen your argument—and presumably it was or you wouldn’t have used it—it’s important enough that refuting it weakens the argument.