Hacker News Reader

Comment by techas | Hacker News Reader

I work in academia and have over 70 papers published. I Agree with most ideas in the article. Another dimension not covered is what I called “author engineering”. Many times it is very difficult to “get into” a new field if you don’t have an author known by the editors. I work in applied math (very transversal) and happen to me often to be rejected because “I don’t belong to the area”. PhD students usually don’t suffer from this as the supervisor is already a member of the community. But if not, try to bring a collaborator that is known in the area. This is usually done in conferences by chatting with people.

View on HN

That's not the way science is supposed to work...

And there's more where that came from. This particular sausage is made with tragically messed up incentives, and people will naturally always optimize the framework you put them in.

Thankfully the scientific process is incredibly resilient to nonsense, because a bad result will eventually screw up someone's future work when they come to rely on it. But it's not pretty.

> because a bad result will eventually screw up someone's future

Not if they isolate themselves enough from the outcome but I get what you're saying.

The world progresses despite these deeply flawed institutions (corporations or academia have these perverse incentive problems and all in all, they do create some value on average).

But in practice it absolutely is how the academic world works. Is politics all the way down.

Which is why it's so funny when you see non skeptical appeals to "the god of science" which apparently exists in a vacuum of correctness and ethical purity.

It's a pet peeve of mine, but "publications" and "peer-review" are not really part of the scientific process? Just like "academia" they have kind of grown onto the term, almost claiming it for their own. I find that a sad evolution.

The most fun in science can be had when done at home and shared with friends.

This!

Academia != science. It is a social construct and dominated by people with power within a given field. That being said, double blind review process improved the author engineering problem a lot.

Platonic ideal, meet human reality.

As long as It has some capacity to self correct, it’s a stable function.

It never really worked otherwise. Even before formal peer review and journals, social standing and political squabbling, finding patrons etc. definitely affected science. The grade school textbook ideal version is a literal lie-to-children. The problem is that threading the needle correctly, without falling over to the other side, of quackery and "the academic mafia is suppressing this perpetual motion machine" and "Big Science" doesn't want to admit that ESP exists etc. can be incredibly hard.

Every human endeavour reflects common human behaviors. Large groups of humans do not interact without political considerations arising.

Science has a few aspects that are distinct from non-Science enterprises, but more aspects in common.

The author elsewhere cautions “I write this because PhDs seem to attract a lot of smart, idealistic kids who are interested in doing Science 1 and don’t realize that they’ve signed up to do Science 2.”

https://maxwellforbes.com/posts/dont-try-to-reform-science/

> Many times it is very difficult to “get into” a new field if you don’t have an author known by the editors.

Although there's plenty of critique to go around about the review system, machine learning here typically uses double-blind peer review for the major conferences. That blinding is often imperfect (e.g. if a paper very obviously uses a dataset or cluster proprietary to a major company), but it's not precise enough to reject a paper based on the author being an unknown.

I thought that blind peer-reviews solved this?

Yes, with some caveats. Sometimes people can guess where a paper comes from based on the used datasets, even graphic design style, a skew in cited papers. Also, often people upload the preprint to Arxiv and so the reviewer may have seen the non-anonymous version already. Also, the having someone within the community among the authors will help with formulating the paper in a way that this community expects it. With the right turns of phrase, citing and praising those related works that are seen highly in the community, using the tone that is usual in the community etc. People should ideally try to counteract such biases in themselves, but humans are tribal and social. Especially if you scale this to tens of thousands of people, the average won't be a saint. People have careers, graduations, promotions, visas, green cards, job prospects, friendships and generally social standing in their professional community on the line. Academics aren't any more holy than people in finance, or politics or entertainment or startups or other ruthless and social-game heavy environments.

My wife works in a fairly niche field and can often guess at the very least which university or research group a blind paper is from, and quite often the author (or in the case of a PhD student, the authors supervisor).

But if they can't, surely they won't reject the paper because of that?

True. But that opens the game of writing a paper “with the style of” someone known and get accepted. Its a gamble game… for authors and reviewers.

Still many many journals don’t apply double blind reviews. There is no advantage of don’t doing it. There is no extra work in doing it.

So I assume that it is not done to keep outsiders out of your garden…

Honestly, I don’t find any other reason to don’t apply it.

There’s tons of extra work in double blind (for the author mostly, but also for peer reviewers and editors/chairs). Speaking from direct experience as author, reviewer, and chair at conferences. And generally the benefit is totally lost as you can easily circumvent it especially if you have already published in the field by just citing your previous work.

I don't see the extra work for author - how exactly?

Reply on news.ycombinator.com