A common minor writing problem is the misuse of 'time' words like while, when, and often. Often is often used to indicate commonly or typically or in most/many cases, but often implies time and the other words don't.
In fact, I saw an example in a New York Times op-ed essay this week: ".. they are often elderly." Are these people elderly on Mondays, Wednesdays, and Fridays, or for certain hours of the day?
This is not the worst writing error in the world. If someone uses often to mean commonly, the reader knows what the writer intended to say.
(Mis)use of constrain is of a similar order of magnitude minor error. If you say that you (or your data) constrained something, everyone will know that you mean that you measured or determined something or that you have reduced the number of reasonable interpretations by figuring something out. Technically, however, constrain means to compel, confine, restrain, inhibit, or limit by force (physical force or more conceptual force). Even if you have wrestled (metaphorically) with your equipment, data, or graduate students to get a result, it is unlikely that you have constrained anything.
When I edit a manuscript written by someone else, my first priority is to make the text understandable. The extent of my editing depends of course on circumstance -- what is my role in this manuscript (editor, co-author, reviewer, advisor)?, is it better to provide advice rather than major edits (e.g., for someone who can learn from a general comment and make the necessary corrections)? does the author need help writing in English? have I edited this document before? how much caffeine have I had today?
In situations in which it is required or appropriate for me to be an active editor, I make corrections of minor things like often and constrain in manuscripts that need the most and the least editing, and I don't bother to make such minor corrections in manuscripts that need a medium amount of editing.
Why the worst and the best and not the middling manuscripts?
The worst manuscripts (in terms of writing, not content) need total rewriting, so I fix everything. The entire effort can take a lot of time, but it doesn't require additional time to fix the trivial problems.
The best manuscripts are a pleasure to read and edit. If, however, I see something not-quite-right, I can't help but fix it. I think writing/editing is the only thing for which I have some perfectionist tendencies, but I'm probably not the one to ask about that.
Manuscripts that require an intermediate level of editing are those for which I don't need to rewrite many many sentences and paragraphs and pages, but they may require a substantial amount of fixing of writing errors such as lack of subject-verb agreement, inconsistent verb tense use, misplaced modifiers, vague or ambiguous wording, lack of topic sentences at the beginning of paragraphs, or lack of parallelism of items in a list, like I just did in this list. If I have to fix things at that level, I let the minor problems slide as long as the meaning of the text is clear.
Some of my colleagues and students would probably disagree with this assessment -- they would likely say that I comment on everything, no matter how minor -- and in some cases that is true (but not as often as they may think).
5 weeks ago