An Asymmetry between Criticism and Knowledge

(note: this is mb better in #elliot-temple – I am okay for the topic to be moved if so)

I had this idea replying to @deroj last night. It builds on ideas about asymmetry from @Elliot’s CF course (session 4 covers asymmetry between criticism and praise).

For context, here’s a short quote from the slides:

➤ If I criticize something, and I’m right about the criticism, then it won’t work
➤ If I praise something, and I’m right about the praise, then it could work or not work

Here is the idea from last night:

Knowledge has to be compatible with reality; it has to work. Additionally, knowledge needs to work reliably (at least for the {goal,context}). That means there has to be some convergence (overlap with other ideas that work) and some reach (b/c if it had ~0 reach then it would be too fragile to work).

But, for a criticism to work, it does not need to be compatible with reality (though it might be). Criticisms don’t need reach, and they only need to work once. They don’t need to be generalizable, or convergent.

Criticisms can work in a system of ideas that is completely disconnected from reality, but knowledge cannot.


One example I can see demonstrating this idea is something like two Christian scholars cycling through conjectures and refutations (about something god related) – some of those criticisms could work but could basically be completely unrelated to reality (besides like mb using some basic logic).

IDK of an example like that, tho. One thing I did remember tho was Socrates’ dialog with Euthyphro.

Euthyphro Dialog (page 7-8 in link)

Euthyphro’s idea:

Relevant bit of dialog:

Here, Socrates crits Euthyphro’s idea based on reasoning about gods. But gods don’t exist. The crit that the gods are inconsistent and/or contradictory is not knowledge in the sense that those gods don’t exist so they can’t be inconsistent. But the crit doesn’t need to be convergent w/ reality, it only needs to work for the reality that is in Euthyphro’s head.

I think criticisms with reach/generality – being able to criticize categories and reuse criticisms – is really important.

More broadly, I think you (like most people here) should either focus on learning activities and practice, or at least write shorter, simpler things and try to make your points really clear and right. Having a bunch of experience with small, correct, clear points would give you better building blocks to make more complex stuff with.

Did you think my post was unclear?

I thought it was good enough (not my best, but enough to communicate the idea to ppl familiar with CF stuff). I deliberately didn’t put too much effort in, that’s why the euthyphro dialog is screenshotted – copy-paste didn’t work well, it’s googleable, and I linked the PDF I used.

It was some of my warm-up writing for that day, too (I do this before work depending on the day and how I’m feeling)

I’m interested in feedback on clarity. I care a lot about it atm. (feedback from anyone welcome)

OK, some clarity things from the first main content paragraph (these don’t necessarily all needs to be explained – glossing over some issues to focus on others can be OK – but it’s too much):

“compatible with reality” typically means not contradicted by evidence (or laws of physics, and possibly laws of logic too), which is different than working (at something).

You don’t explain why knowledge needs to work reliably or what counts as reliable (reliability is typically seen as a matter of degree, and you haven’t specified a cutoff for how much reliability is enough for what you’re talking about).

You don’t explain why reliability requires convergence/overlap.

You don’t explain why no reach means fragile or why fragile means doesn’t work. And earlier you related working to compatibility with reality, so you seem to be saying fragile = incompatible with reality, which seems would require some explanation. In general, fragile means something does work but has a significant risk of being broken or failing in some way.

Also, my previous critical reply (“I think criticisms with reach/generality – being able to criticize categories and reuse criticisms – is really important.”) may be clarity related but I’m not clear on what’s going on there.

1 Like

I didn’t read that sentence as being critical (and I agree with it). I thought you were saying it mb so that other ppl weren’t mislead by the post (e.g. if someone was not so familiar with CF). I just looked up the defn of ‘critical’ and it matches what I thought it was. If that sentence reads as critical to most other ppl, IDK why I didn’t read it like that. It seemed mb dismissive more than critical.

I didn’t reply saying that I agreed because it felt like a low-qual (and obvious) reply; it didn’t seem substantial enough.

I was trying to disagree here. If you agree with me, then you could explain why there isn’t a conflict between what you said and what I said. Or if you can’t figure out any possible conflict to address, you could ask (I thought it’d be clear enough but maybe not).

Yeah okay. With hindsight, this is roughly what I would have said:

Yup, criticisms with reach are definitely more important, generally. A super specific, fragile[1] crit is only rly useful if you don’t have an existing (applicable) crit in your library of criticism. The reason I focused on that, tho, was because I thought it was easier to explain the asymmetry using that edge case (a specific, fragile, not-general crit). Or maybe because the asymmetry is only rly apparent in an edge case like that (all the crits w/ reach are going to have knowledge baked in). Either way, I thought that edge case would be useful for demonstrating the asymmetry.

  1. I still would have used “fragile” if I replied at the time, but wouldn’t now b/c of the “You don’t explain why no reach means fragile or why fragile means doesn’t work.” paragraph ↩︎

Were you intending strong emphasis on “need” like saying it isn’t a logical necessity?


OK that wasn’t clear to me. Also a criticism with zero reach is basically useless because it has no bearing on any of the infinitely many slight variants of whatever it’s criticizing.

1 Like

This wasn’t obvious to me. It makes sense (esp given some ideas from paths forward - I think from PF at least).

I don’t see that as a decisive criticism of the asymmetry existing. I do see it as a decisive crit of the goal of looking for those sort of criticisms particularly, tho. (edit: before that, I might have thought there were some times were mb it would make sense, but don’t anymore)

here are some ideas on things that (maybe) went wrong with my original post:

  • audience
    • I glossed over a lot, mb I wrote with too many assumptions about my audience (like that they’d agree with me based on familiarity with ideas – do error correction on my behalf).
    • Mb I should write for a broader audience in #unbounded more generally (it seems less important in #friendly or #other)
  • I didn’t see ET’s post as critical; and the thread is in #unbounded, so I mb should consider that posts which aren’t explicitly agreeing should be responded to (at least seeking clarity).
  • I didn’t treat the clarity crit with enough significance (I wasn’t even planning to ask about it); I understood what Elliot was saying, but didn’t know why he was saying it and I didn’t ask. If I had some idea of why he was making that sort of crit, at least I can make a somewhat-informed decision about following-up, but if I know I don’t know why the crit was said, then I never will unless I ask.
  • I mb didn’t put much effort in to the first paragraphs b/c I was too keen to get to the idea and thought it would make enough sense. Since I was using the post as a writing warm-up, but didn’t mention that, no one would know that I’d put less effort in than I might otherwise.
    • Also, probably not a good idea to use that sort of thing as a warm-up (mostly I do simpler things).
  • That’s the sort of topic I should be more serious about, or at least make it obvious when I’m not being too serious. The tone of OP doesn’t communicate that; I use proper capitalization and grammar (I do use ‘mb’ and ‘b/c’, but those are in parens so the effect on tone is diminished). This is made worse b/c I chose a well-known philosophical text as my example.

Some self-crits

I think this makes a point, but it also feels like an excuse. I shouldn’t be trying to excuse it.

This wasn’t clear – I should treat replies in unbounded that aren’t explicitly agreeing with more significance than I do, and I should follow up seeking clarity if need be. Agreements don’t need to be followed up, but an agreement requires both people to know they’re agreeing (in this case, I agreed with Elliot but Elliot didn’t know that). If I don’t follow up in those cases (asymmetric agreement) then – to everyone else – it appears that I just stopped discussing.

IDK if I did, actually. Before I replied: I re-read what he posted and “shorter, simpler things and try to make your points really clear and right” seemed more important than originally. I might have understood the words the first time I read them, but if I wrote a summary from memory 5 min later IDK if the clarity crit would have been included in that summar.

This quote is from an email with Elliot (posted w/ permission)

[Elliot said, regarding OP:]

i don’t think it’s efficient practice/learning.

Yeah, that makes sense. I think one reason that I don’t like that
crit is that I’m not claiming it’s efficient, and I don’t want all
my practice to be efficient b/c I still want room to explore. I
guess I could just split that up, tho – basically not call it
practice but exploratory writing instead (which is rly what it
is). The post probably would have been better if I’d deliberately
done it as exploratory writing, too, b/c I would have structured
it v different and put more effort into like clearer, dot-points
rather than writing some sorta-polished paragraphs.

I wanted to share the quote for the last sentence. My original post might have been okay if I’d done it as exploratory writing (it almost certainly wouldn’t have been as bad). One way to look at it: I would have been more honest w/ myself about how good the result was going to be. Instead, I put effort in to the wrong thing (one reason it’s inefficient as practice). doing it as exploratory writing means the project becomes smaller and there are fewer goals (so the goals that remain are easier to meet).

One thing that occurs to me is that this sounds related to my long-running conflict about learning styles/methods (another relevant microblog post). I might write more on that as I think of it.

Parallel comment to what I quoted:

edit: Note that the hard-wrapping is different between the two quotes b/c of different quoting indentations and email clients.