Writing a PhD in the modern world — The scientists’ lament

Writing has changed. If you write for a living, this has been obvious for a long time. But for those of us who write for research — for whom writing is not an end in itself, but a medium to share results and ideas — the change is slower, less obvious, and maybe messier too.

Technology has served the professional writer well. Beautiful apps provide both cleanly presented source material and elegant tools for producing content. Publishers can turn out highly polished articles at the breakneck speed needed to satisfy voracious consumers.

The scientist, however, fares less well. Ours is a world constrained by complex bespoke requirements, obtuse practices, traditionalist environments, and complex tools.

My pet hate in this space is LaTeX— a typesetting system ingenious in its conception, powerful in its algorithms, but migraine-inducing in its implementation. The idea of a markup language which can dependably produce high-quality documents and meet the complex needs of scientists is laudable. Writing papers and theses involves tables, equations, figures, captions, and more. While word processors have slowly gained features to address these points, they introduce complexity and remain unreliable, especially for long documents.

But LaTeX is a system cobbled together over decades with little design vision or oversight, and programmed by scientists to boot (my gripes with coding in the scientific community will likely get a post to itself). A quick search for “LaTeX boilerplate” will throw up dozens of projects on Github, all containing a dozen or more files needed just to get a LaTeX document off the ground. Setting up a publishing workflow often requires finessing command line tools getting the options just right.

All this however is more of a symptom than anything else. Writing and publishing is a massively overlooked area of science. I can only guess as to the root causes. I suspect that research labs and institutions simply lack any drive to improve things. Postgrads are relatively cheap — no one cares if they spend days figuring out how to produce their reports. There’s no money in providing quality tools. The senior staff went through the same thing and see no reason why anyone else should have it easy — they probably think we’re all soft for being able to type on computers.

There is also an attitude in certain disciplines that considers learning complex and unpleasant tools to be a kind of rite of passage, granting access to a special club.

Really though, I’m guessing. The bare truth is that the software revolution has yet to reach scientific writing. Even the best, most cutting edge tools come with a raft of caveats and compromises.