2011-04-04

You shall read Knuth et al.'s minicourse for technical writing

When i am reading papers from one field, i cannot finish a page in an hour. I can read a single sentence for 10 times, and i still don't understand what that sentence means. The worst part is, that i couldn't remember what their papers say after a week.

Tonight, I suddenly felt that it was not my problem. It's because they like to introduce new "awkward" notations, and they don't like to use the daily meanings of words.

By awkward, I mean not following conventions. For example, in math, a superscript, most often, means the power. In the authors' language, it could mean something has nothing to do with power.

By daily, i mean the meaning most people will come up in mind when hearing/seeing that word. What's the purpose of using the 5th or 6th meaning of a word (unless it is its meaning under the technical context)? For example, why do they use the word "reduct", an obsolete word in modern English (according to Merriam-Webster,  Wiktionary and my Firefox spell checking)? Can they use "reduce" or "reduction"?

To verify whether i was wrong, I googled out Knuth et al.'s Minicourse for Technical Writing. Right, it's not my problem.

For instance, point 6: "Think of a dialog between author and reader" Well, I completely cannot feel any dialog from sentences without "we" or connections.

Another rule related to this point is, point 12: "Motivate the reader for what follows." Well, i cannot think about what's next if i can't even follow.

Another example, point 16: "Display important formulas on a line by themselves." I really hate people squeezing formulas into lines with English, especially when formulas are more than half line width long. 

Last example, point 13: "your sentences should flow smoothly when all but the simplest formulas are replaced by 'blah' or some other grunting noise." This is very important. Please give every new notation you introduce a name, intuitive ones.

So, i recommend anyone who is writing papers with a lot of math to read that doc, which is published in 1987. A copy is here:  http://math.stanford.edu/~rubin/110/mathwriting.pdf
You can google "Knuth Minicourse on technical writing" if that link above doesn't work.

A note: The word "reduct" is called "reducing agent" or reducer in modern chemistry.

No comments: