Hacker News Reader

Comment by Eliezer | Hacker News Reader

It might be that my standards are too high - I'm accustomed to reading how-to-write books written by the likes of Orson Scott Card - but I'm a bit worried that the article itself isn't as well-written as it should be. Maybe you need unnecessarily pollysyllabic linguistics for scientists to pay attention to you at all, but it's still a bit worrisome when it comes to my deciding how much to trust the advice.

View on HN

Maybe it could be better written, but their advice gels with how I try to approach technical writing. The most common problem I see are authors packing too many concepts or facts into a single sentence. The writing becomes difficult to read for a reason that I was aware of, but could not articulate until I read this piece: readers are searching for syntactic closure. The more stuff - no matter how important - you put inbetween the subject and verb, the harder it is to read. The way I think about it is that until the reader gets that syntactic closure, all new words go into a temporary buffer. If that buffer spills before reaching closure, the reader is lost.

I encounter this in CS papers I read and co-author frequently. My most common editing remark is "split into two sentences."

I regret to announce that I share your sentiments with regard to the circumloquacious fashion of the writing in that article.

I think we've made progress since it was written. Eye-tracking shows how we read and which words or letter combinations trip us up.

By the way, wouldn't that table naturally have time in the left hand column, considering it's probably an independent variable, while temperature is dependent?

As to your final point, the authors were trying to suss out why we want the independent variable first. Their answer: it provides the context to understand the significance of the dependent variable.

Reply on news.ycombinator.com