Hacker News Reader

Comment by vaylian | Hacker News Reader

My impression is that many new academic writers try to imitate the style of established academics. The results are very unnatural and hard to read texts. And journals are fine with terrible writing style, because most academics are not good writers. Furthermore, unclear and convoluted writing is often seen as fitting due to the complexity of what is being written about.

View on HN

Also consider that by the numbers, most of academia wasn't even born native English speaking, whether they now live in an English speaking country or not.

A personal note, I'm a native english speaker living abroad, with a non-native-english speaking superiors. While their writing is quite good, i find myself battling constantly for what we might consider 'better, more clear' writing, in favor of the current scientific writing 'trends' in our field.

A successful paper or two with a weird phrase or typo will end up reshaping that field to adopt this language, as authors of subsequent work attempt to align themselves to the original topic.

The "concise and clear" style of writing that's prevalent in the Anglophile world isn't nearly as popular in other cultures. In primary school in post-communist Bulgaria, we were given bonus points for constructing compound sentences with 3 or 4 clauses, for using different words for referring to the same entity (rich vocabulary, supposedly). Repeating the same noun in a consequent sentence was a big no-no. Writing in a "concise" way made you look "simple" to the teacher, and complex sentences made a good impression even when the content is factually or semantically unsound.

It's partly the same in much early writing instruction in the UK and there are pedagogical advantages to encouraging learning writers to embrace ambitious word choice and grammar and risk making mistakes. This should not be the final goal of writing instruction.

Former academic here. Have to rather strongly disagree with you.

At least in fields I was familiar with, a lot of Ph.D. training focuses on making people into good writers. Clear communication with both peers and the general population is rewarded; it is damn near impossible to get ahead without these skills.

I don't know what you mean by "hard to read." A good scientific paper is supposed to be very information rich. It condenses 100 other very dense papers into 4 pages, leaving just a few pages for the new contribution of the current work.

Even for a very clear paper, for a field I am well versed in, a good paper takes 10 readings to understand. Because of the information density.

I think the argument is that it shouldn't take 10 readings to understand a very clear and good paper in your field. There's likely an easier way to communicate dense information.

You should try writing papers. Most papers take many months to write, hundreds of hours, go through many revisions and rewrites. People try really hard to write good papers. Everyone knows well written papers get cited more.

But synthesizing newly discovered knowledge that only a handful you know (and often don't even understand all aspects of), into something others can understand is far harder than you think it is.

If something is clearly written, it shouldn't take 10 readings. Information density often is at odds with readability. In programming material, there is a difference between learning oriented explanatory documentation and dense reference documentation. Dense only works after you already understand the concepts of the whole. Before that you need clear, non-dense explanations to learn those concepts. Nearly all academic papers I've read are quite inaccessible to read. So either the phd training you're familiar with is a diamond in the rough, or the training isn't working to help write papers that are written clearly for people who aren't already phds.

It's all too human an instinct to attempt to signal in-group membership by imitating and adopting the culture of said in-group (see shibboleth [1]). It's also common to go overboard in these attempts to the point of hypercorrection which, if turnover rate is high enough, may then become accepted usage for the next generation…

[1] https://en.wikipedia.org/wiki/Shibboleth

Reply on news.ycombinator.com