Hacker News new | ask | show | jobs
by dnautics 3436 days ago
Peer review most certainly does not determine scientific truth.
4 comments

This is a vacuous statement. Science is an inductive exercise and suffers from the problem of induction. It's difficult to make the case that real truth can come from it at all. In lieu of this, peer review is a very good heuristic for sifting out bunk science.
I don't think it's fair to call a correct statement vacuous. More nuanced than it appears, and easily misinterpreted maybe.

I think it's fair to point out that peer review doesn't determine scientific truth, because that would get the causal relation in reverse.

Consensus has little to do with objective truth, but it is a useful heuristic as you say.

Yeah, let's put peer review into perspective. In a topic about Trump, censoring scientists. I can see the "give science back to the people" anti-intellectual slogans coming up.
Those "anti-intellectual slogans" I think you mean can also be seen as arguments for a return to enlightenment values instead of an entrenched, elitist dogmatism.

Behind those slogans is the idea that intellectual life doesn't belong to a chosen few, but to everyone who's interested and willing to engage their capacity for independent thought and critical thinking, and those who are willing and eager to engage with other people with different opinions.

A lot of arguments can be seen as sensible until one realizes they were pushed by people with a hidden agenda. It's often just the motte part of motte-and-bailey arguments.

That's why there are people (eg. in Europe) who aggressively promote secularism, and many agree with that, until you find out they belong to some hateful anti-islam organization. And that's why some people put science into question, like, how can we be sure of anything, which is defensible from an epistemological standpoint, until you discover they are creationists.

So you want to be careful before giving random people the benefit of the doubt.

does 'being pushed by people with a hidden agenda' automatically invalidate something? I'd say that that the American (and presumably European) political establishment (both the right and the left) have a 'hidden agenda' of pushing science because it creates a higher authority that they can appeal to justify questionable policy choices. That doesn't mean I think all science is wrong.
>And that's why some people put science into question, like, how can we be sure of anything, which is defensible from an epistemological standpoint, until you discover they are creationists.

Honestly, if skeptical arguments in epistemology can be used to strengthen Creationism, I think that's a major point against taking them seriously in epistemology. Good epistemology ought to have a low enough false-positive rate (disbelieving things when they're true) to throw out "all the evidence is wrong, because I got this book someone wrote a while ago".

It's not a useful heuristic any more. I'd say about 40% of the 'reputable' scientific literature in chemistry and biology is flawed to the point of uselessness.
i think it's perfectly appropriate to call a true statement vacuous. for instance, a counterfactual conditional is always true; that fact makes the particular truth of a particular counterfactual conditional completely uninteresting, or vacuous. it's true, but in a hollow, empty, vacuous way. what use would there be calling a false statement vacuous?
Okay sure, I made a bad generalisation. Some true statements can be uninteresting. The point I was trying to make was more that the parent commenter was making a true statement about epistemology that is relevant and useful to the discussion.

Maybe it could have been better elaborated, but I don't think it deserves being down-voted and replies of what seems like moral disgust.

"you're not wrong you're just an asshole"
Arguable, but lack of peer review certainly should inspire a lack of confidence in results.
Do you not believe grigori Perelman's proof of the poincare conjecture? Do you believe in arsenic life? The first is still not peer reviewed, and the second very much was so.

These are some more well known cases. Having been there I'd say probably about 40% of peer reviewed material in chemistry and biology is seriously flawed.

We really have this problem where peer review is held up as a scientific standard. It shouldn't be. The scientific standards are independent replication and confirmation through prediction of a derivative result.

That is not to say all scientific peer review is flawed. If it appears in the journal organic syntheses I'll believe it, every time. (In order to get published an editor has to repeat the experiment in their lab, there are often liner notes)

>Do you not believe grigori Perelman's proof of the poincare conjecture?

Yes, because his peers spent serious time checking it and found no flaws. Peer reviewed doesn't need publication in a journal.

>Do you believe in arsenic life?

No, because peers found it flawed.

You're conflating peer review with simply getting past initial peer reviews. The more peers that review a claim, the more likely it is correct. The flawed ones are almost always not reviewed much at all.

And I'd bet there is a significant gap in correctness between things that are barely peer reviewed over things that are not.

It does a better job than anything else you can come up with.
What?