Hacker News Reader

Comment by untilHellbanned | Hacker News Reader

Desperately need this in academia. The academic literature is full of hyped-up phrasing. Probably safe to say the more hyped-up something is presented, the less novel it actually is.

View on HN

I used to think I wasn't smart enough to read academic articles. Then I realized that most of the authors are just terrible writers or deliberately make their papers as difficult to read as possible to obscure or exaggerate their findings.

This always irks me, not just with academic articles but also with other "high-brow"/"intellectual" writing (including everything from literature and philosophy to tech writing). There's a lot of intentional or accidental obfucasting or complicating of things that when you call them out will just get you accused of being too stupid to understand it.

Its better today than it used to be. Go read some renaissance alchemical or natural philosophy texts. These are often called occult. The word "occult" means "hidden" as in "concealed behind a wall of unbelievably long winded flowery bullshit."

Today's academic writing is a paragon of brevity and clarity.

I've known a couple academics who argued to me that this problem is systemic, rather than something that can be blamed on indiviual authors. A friend of a friend submitted a biology paper that she tried to make as jargon-free as possible, and it was rejected on the grounds that she hadn't shown a familiarity with the field. She edited it to include the dense jargon she had originally avoided, and the article was accepted (though I'm not sure if the acceptance was from the original journal or a different one).

It definitely seems systemic. Scientific papers have a pretty distinct style. "Familiarity with the field" is something rather hard to check when reviewing the paper, but one of the easiest proxy for it is whether your paper sounds like every other paper. The expected side effect of immersing yourself in a field is that you pick up its style.

It's often said that if the writing is dense, the writer probably doesn't understand what they are saying. Being in academia for 20 years now this seems spot-on.

The following is one of the most well written PhD dissertation I've read. Its all written in simple clear to understand words.

Bulldog: A Compiler for VLIW Architectures by John R. Ellis http://www.cs.yale.edu/publications/techreports/tr364.pdf

A huge amount of them can be summarized as "if your X fits these highly unlikely criteria, a highly specialized Y can give you 1% better results than a long-standing general algorithm that already works and you can probably use within 5 minutes".

No. Really? shutupandtakemymoney.gif

Also, in physics at least, it seems like at some point in the 80s, the writing in academic papers became much worse than before.

When I had to (had the pleasure to) read some older papers, it was like they were from a different genre. They were clear and pleasant to read. Maybe it was before academics were measured by "hard numbers" (the academic version of PageRank - resulting in link farms) that some lazy fools made up.

My former field, in the humanities, prizes clear, direct writing and doesn't suffer bullshit or needless density. It's one of my favorite aspects of the discipline and its culture.

I haven't noticed that with computer vision papers, maybe it depends on the field? There is sometimes difficult to understand jargon but it's generally not gratuitous.

Not wkth bullshit.js takes less than ever to understand and benefit from

Reply on news.ycombinator.com