I'm certain this isn't a dupe and that it follows the guidelines.
June 28, 2013 6:11 AM   Subscribe

Why it's good to be wrong. Unless, I'm wrong about that.
posted by Obscure Reference (12 comments total) 12 users marked this as a favorite
Would it be wrong to post a cleaver tautological comment here?
posted by sammyo at 6:44 AM on June 28, 2013 [1 favorite]

I only skimmed the article, so maybe it's more careful than I'm about to credit it with being, but it seems to me that there are two very problematic conflations in the piece:

(1) It conflates the claim that all our assertions are fallible (that is, possibly incorrect) with the claim that we are never warranted in making claims that are universal (that is, claims about what always happens or what one should always do).

(2) It conflates fallibilism (the view that any of our beliefs might be false and hence might need to be given up with future experience) with falsificationism (the view that scientific theories -- really, synthetic judgments generally -- are never inductively confirmed or verified, but that science proceeds by a process of conjecture and refutation wherein some theories are shown to be wrong but no theory is ever shown to be right).

Am I being uncharitable? I confess that I had very little patience for the wandering hobos and popes speaking ex cathedra about gravity.
posted by Jonathan Livengood at 6:44 AM on June 28, 2013

I think it was worth it if you made it past the weak hobo/pope thought experiment to the end; it's basically the Physicist's Guide to Post-Modernism, amirite
posted by Mooseli at 6:54 AM on June 28, 2013 [2 favorites]

And that concludes the opening lecture of Epistemology 100. Please sign up for a tutorial session on your way out, and I'll see you next Tuesday for our discussion of the fuzzy boundaries of a priori. Unless it turn out that the universe, including all sense of time, is merely a momentary epiphenomenon. Ha ha.
posted by 256 at 7:02 AM on June 28, 2013 [4 favorites]

I read it kind of disliking the style and being impatient with it but somehow it kept me reading till the end and I ended up thinking it was quite good.

If by "Post-Modernism" you mean the thing that angry people yell about destroying our society and intellectual culture, then yeah, this is kind of a Physicist's Guide to Post-Modernism.

(That thing doesn't bear very much relation to the several distinct things called "postmodern," "postmodernism," "postmodernity," etc, in actual fields of scholarship where those terms are relevant, but who cares? Not the yelling people!)
posted by edheil at 7:05 AM on June 28, 2013

Jonathan Livengood: I didn't personally read (1) in the text but it's possible. (2) doesn't seem to me to be a conflation at all; those views are pretty compatible. In fact I believe Karl Popper ruefully admitted something to that effect when made aware of certain earlier fallibilists (I'm afraid I don't have a source for that though).
posted by pixelrevolt at 7:21 AM on June 28, 2013

Compatibility is one thing. Conflation is treating them like they are the same position or doctrine. The latter is what I'm worried about.

Just for the record, I am a fallibilist, but I think falsificationism is crazily wrong.
posted by Jonathan Livengood at 7:26 AM on June 28, 2013

The concept of infallibility that most of the article is spent debunking seems to be a kind of strawman concept that literally no one actually buys into in real life. For example, no child anywhere has treated their parent as being absolutely infallible to the point of believing literally every thing they ever say and no one seriously suggests that they should. Even the ex cathedra concept, which is the closest any widely held view in real life gets to the idea of actual infallibility, is more of a practical concept of an organization having a type of statement that is given the highest possible priority level, in the same way that the US Supreme Court decisions are treated as being the final word on a particular issue.

When Deutsch actually makes points about things real people do, like his criticism of medical mistakes being considered malpractice, he seems to be focused more on abstract concepts around wrongdoing than about the practical reasons why systems need disincentives for negative outcomes to shape behavior. In practice everyone already has the same concepts fallibility that the article is talking about, they just don't spend too much time thinking about the crazy theoretical edge cases where something they general think of as true would actually be false.
posted by burnmp3s at 7:37 AM on June 28, 2013

Could you elaborate on falsificationism? Is it your position that scientific theories are not a class of beliefs?
posted by pixelrevolt at 7:39 AM on June 28, 2013

Would it be wrong to post a cleaver tautological comment here?

Occam's Cleaver suggests we'd make mince-meat out of you if you did.
posted by Potomac Avenue at 7:58 AM on June 28, 2013 [3 favorites]

Could you elaborate on falsificationism?

Sure. Falsificationism has two main components. The first component is a thesis about the demarcation of science and non-science. Falsificationists think that sentences (or propositions or hypotheses or judgments or whatever you think the truth-bearers are) count as scientific just in case they are falsifiable.

The falsificationist demarcation criterion looks pretty good at first. But when you start to push it, things stop looking so nice. One big problem here is that there are sentences that are not falsifiable but whose negations are falsifiable (e.g. at some future time, a particle will zoom into the observable universe from infinity). For starters, this seems to violate a principle that if a sentence is scientific, then so is its negation. But worse, in many of these cases (maybe in all of them), the unfalsifiable sentence seems genuinely scientific. I think that is the case in my parenthetical example.

The second component of falsificationism is a thesis about Hume's problem of induction. Recall that Hume argues that there is no non-circular justification for induction. Falsificationists like Popper think that Hume was absolutely right. So, they need some way for science to work without induction. Hard. The answer falsificationists give is that science works by bold conjecture and severe testing. But they explicitly deny that surviving a severe test ever makes a theory more likely to be true. For falsificationists, science never shows that any sentence is true or even likely to be true. All that science ever does is rule things out. But since there are infinitely many sentences to rule out, ruling out a finite number of them does not raise the chances that the not-ruled-out ones are true.

The falsificationist gets around Hume's problem, but in a Pyrrhic way. What one is left with is still a kind of skepticism. But it's now a skepticism about the ability of science to show that its claims are true. Personally, I think this is way, way too much skepticism. I agree with the fallibilism line that we should always be ready to jettison our theories if/when they conflict with experience. But I also think that experience (sometimes) gives us good inductive reasons for believing scientific theories to be true or approximately true.
posted by Jonathan Livengood at 8:02 AM on June 28, 2013 [5 favorites]

Unless it turn out that the universe, including all sense of time, is merely a momentary epiphenomenon. Ha ha.

I want to append this to the end of basically everything I say from now on. Unless, of course...
posted by gauche at 9:15 AM on June 28, 2013 [1 favorite]

« Older Focusing on the process   |   Look ma, no magnets Newer »

This thread has been archived and is closed to new comments