Is advice that can be reversed even advice?
an extended semantic nitpick on the law of equal and opposite advice
There’s a classic blog post from SSC about how people should consider reversing any advice they hear (“except feedback directed at them personally”). I also hear people talk about the Law of Equal and Opposite Advice, and I’m not sure where that came from, but here’s an example of someone referring to it.
In the post, Scott makes a point about how people are often drawn to groups based on ways they already are, the group continues to upvote the virtue of being that way, and then people end up more that way than is good for them. I think this is a good and substantive point! That said, I have seem to have a big objection to the framing.
In my ontology, nonindexical imperatives and “should” statements aren’t even advice.
Epistemic status: thing I want to tell my past self
“You need to stop being so hard on yourself, remember you are your own worst critic,” the first example Scott cites, is, IMO the sort of thing that past me would have probably considered a type of advice. I might have considered whether I ought to follow it.
But now I read it and think ~“this statement is not even wrong”. It’s not a claim I can evaluate one way or the other, and I want to relate to it as ~noise.
The “you” part of the statement is a common English construction (that afaict is called the “generic you”), and I want to call it a lie. By which I mean for me, it belief reports not to “true”, or even to “false”, but to “lie”. (I am well aware that the statement is not any sort of lie in the standard sense.) I see it as someone trying to have some effect on my mind, perhaps well meaning, but by saying words that aren’t true.
The above “should” statement also has no scope. Who should? everyone? Why should they? because it would achieve more good in expectation (for everyone) than it would cost to do it? How long should they try it for if they can’t tell whether it’s working? indefinitely and assume they are doing it wrong if it doesn’t work?1
My other big issue is that the “should” statement has no moving parts. A version with moving parts might look like:
“There’s a common thing where people perpetually compare themselves to their ideal self. This can be great for orienting, but it’s typically not helpful for judging, since it’s a moving target that often specifically moves away as you approach it, like the horizon. When judging yourself, make sure that you’ve picked a more functional standard than comparing yourself to an ideal.”2
Someone could take issue with that as advice. Maybe it’s not really common to do what I described! Maybe the thing with the horizon is a bad analogy for reasons, but it isn’t nothing! And I’m not sure what it would mean to reverse it. I suppose someone could always consider that the framing I laid out above was actively harmful and make a point of avoiding it? But I think the same thing that makes it substantive makes it a stretch to reverse it.
Obviously, this is all a big semantic nitpick. I think most people would call the “you should” statement I’ve been complaining about advice, and there’s a lot to be said for using language in the normal ways. I’ve also heard people say that they are happy to hear a bunch of “advice” of this type—if it hits, it hits, and if it isn’t for them, they will happily move on.
Those people are of course entitled to their opinion, but I disagree. I think see these sorts of statements as polluting the what-does-anything-even-refer-to commons, and IMO society would be healthier if most people had more contempt for it3.
So that’s my advice for people who are interested in trying out making sense of the world in ways that I do now, wish I had started doing earlier, and found worth it to write a short thing about. Consider thinking of generic, reversible advice as “not even advice”.
I think it’s a little silly for me to have written up this entire complaint, but I have been trying to write some blog posts I thought I believed in more, and kept stalling. This one seemed pretty easy to say, so I wrote it instead.
People are frequently like, “well why don’t you ‘just’ make the most obvious translation then? Why treat it as noise?” I could, and I sort of do sometimes… and I mostly think it leads to confusion. Because IME people who say the one thing mean a really different and less contentful thing from people who say the other thing.
I am reminded of this passage from Foundation:
I took that record, had it copied out and sent that to Holk for analysis, also.”
Lundin Crast said, “And where is the analysis?”
“That,” replied Hardin, “is the interesting thing. The analysis was the most difficult of the three by all odds. When Holk, after two days of steady work, succeeded in eliminating meaningless statements, vague gibberish, useless qualifications—in short, all the goo and dribble—he found he had nothing left. Everything canceled out.
“Lord Dorwin, gentlemen, in five days of discussion didn’t say one damned thing, and said it so you never noticed.
I got this from Bill Harris the Holosync guy, who was probably not the originator of the framing. Also he recommended that people judge progress relative to their past self.
Not for any of the people giving it! I am very anti people having contempt for entire people.
I think it's well-formed enough to be something like a command - there's something it would be like to submit to it - but as you point out it's not a conditional prediction, and even real commands have to implicitly predict differential consequences depending on whether you obey. So it's a third-degree simulacrum of a command, masking the absence of an underlying reality.