In an attempt to get as much work done as possible before the inevitable decline in cognitive abilities due to infant-induced insomnia, I am now working maniacally to finish multiple white papers, start drafting grants for the May through Sept deadlines, and prepare all nearly-ready journal manuscripts for submission.
The latter activity has of late been the bane of my existence.
1) When I spend many, many hours editing your excruciatingly comprehensive paper in equally excruciating detail, and you give me the next version in which large chunks of text have been untouched so I have to fuckin' do it all over again, I will bite your head off. Do not bat your eyelashes in dismay at my wrath: ENTER THE GODDAMN CHANGES WHEN I REQUEST THEM.
2) If I tell you that the introduction sucks so badly that nobody can tell what the hell we did that was new or why we bothered with the work to begin with or where our work falls in the grand scheme of things, that means you HAVE TO COMPLETELY REDO IT.
When all I write in big, screaming letters is REVISE THOROUGHLY, and then talk to you to give you pointers along which it needs to be revised, that does NOT mean that the writeup is perfect. In fact, it means that it stinks so bad that the only way to save humanity from such a pathetic attempt at scientific writing is to burn it, and no amount of ink would have been enough to specifically mark all that is wrong with it. REVISE COMPLETELY.
3) Have mercy towards the reader. Do not bestow on him/her artificial, painfully cumbersome compound adjectives that no one in their right mind would use in speech. When techniques Awesome, Breathtaking, and Crucial are combined into an über-technique for a measurement, please, please don't say "Awesome-Breathtaking-Crucial-measured quantity". Saying "quantity measured by the combined Awesome-Breathtaking-Crucial (A-B-C) technique" is fine.
4) Do not keep your paper at arm's length, like it's a poisonous reptile that will hurt you if you get too close. Don't write in a detached, (cumbersome adjective alert!) passive-voice-heavy style. If you hate your paper, it will hate you back, and the reviewers will hate both of you even more. Clear, fluid, and engaging writing is absolutely critical. I cannot emphasize enough how strongly the good quality of writing correlates with short review times and overall better review outcomes.
5) Just because you found a specific analytical derivation or a computational intricacy or an experimental protocol particularly daunting and were proud of yourself for surmounting this obstacle, that does not necessarily qualify said obstacle for a central position in your paper. The paper must present a story with the technical details necessary to describe the work and support your message without obscuring it. For the 20 pages of details that are only likely to be read by a poor grad student soul entrusted with reproducing your data, there are appendices and online supplementary documents.
6) Don't be lazy with references. One of the most common complaints I have regarding the writing of my students, even those who write compelling technical prose, is underciting. I am looking at a manuscript written by a recent PhD grad of mine, who is sticking around for a short-term postdoc while interviewing for jobs: there is a large amount of text in the introduction, ironed out through multiple conference abstracts and his own PhD dissertation writing process; in the text, there are statements we know to be true, but they are not trivially obvious and should therefore be accompanied by appropriate references. He did not cite anyone.
7) Random pet peeves: I hate "impact" used as a verb. Not sure why, I just do. I hate reading about stuff being "negatively impacted" by the leprechaun turd production. I have a collaborator who savors it (in contrast to "to influence" or "to affect," for instance) and it's driving me crazy. Then again, I also used to feel passionately against "thus," but have since developed a tolerance to it. And I hate it when people take liberties with abbreviating journal titles as they see fit (and not as they are standardly abbreviated). I am religious about the serial comma.
I invite you to share your own technical writing pet peeves.