Image credit: xkcd https://what-if.xkcd.com/76/
Warning: long post!
Writing is an absolutely essential skill for grad students. I’ve seen many flounder (and some fail to finish the degree) not because their research was poorly conceived or their experiments didn’t yield data, but because they just couldn’t convert their data into theses and publications. With writing skills so crucial, it might seem surprising that most grad students never receive any formal instruction in writing. There isn’t even any consensus as to how or whether grad students should receive such instruction (and there’s little formal research on the topic).
So how should grad students learn to write? Let’s look at some possibilities. Spoiler alert: I’m going to suggest that none of them works by itself.
(1) They could enter our programs having learned to write as undergrads. I think this one fails for three reasons. First (as recently emphasized by Terry McGlynn and Manu Saunders) few undergraduate STEM programs provide extensive training in writing. Second, the kind of writing undergrads are usually asked to do is very different from the kind of writing needed in grad school. But most importantly, writing just isn’t something one can master as an undergrad – or as a grad student, or an early-career researcher, or a late-career one. It’s a craft that we practice, and keep improving, throughout our careers. So even with the very best undergraduate experiences, our grad students would still have more to learn.
(2) They could take grad-level writing courses we offer in our own departments. This is an obvious solution, and one that grad students often agitate for. Such courses can definitely help when they’re offered (Torrance et al. 1993); but they’re seldom offered. This is partly because we encourage grad students to transition away from learning via coursework and towards self-directed learning. But even if we wanted to offer writing courses, we’d face two problems. First, writing often isn’t in the teaching wheelhouse of available faculty members the way statistics, or bioinformatics, or chemical synthesis might be. Second, teaching writing is an enormous amount of work. Students can’t learn writing by being told how to do it – they have to write, and have that writing evaluated and commented on, and revise it and have it re-evaluated, and so on. There are ways to make this more efficient, especially through peer commenting – but “more efficient” doesn’t mean easy. Having enough faculty complement to offer good writing courses may be a luxury very few departments enjoy.
(3) They could take writing courses taught in English departments, writing centres, etc. We might presume it’s best to have writing taught by professionals, but this idea has some hitches. To start, it only displaces the workload issue: now it’s the English department that must be generously staffed, and I rarely hear my English colleagues complain of doing too much hiring. But a more interesting issue is how often non-science units can teach scientific writing effectively. There’s a lot about writing that’s universal; but there’s a fair bit about scientific writing that isn’t universal, and it may ask a lot of a non-scientist to master our writing forms well enough to teach them. The issue goes beyond content and form, though. I wonder if non-scientists can teach scientific writing with authority – the kind of authority that comes with first-hand involvement with the craft and with the language and culture of science. I think scientific writers respond well, for instance, to being taught some of the science behind good writing practice, or some of the history-of-science reasons for the way we write. Such aspects of writing are rarely taught by non-scientists. Now, of course there are some valuable writing courses offered outside science departments, especially in units that hire staff with science credentials. They’re not common, though, and not likely to be the whole solution.
(4) They could take online writing courses (MOOCs). I’m skeptical of MOOCs in general, and deeply skeptical of the idea that writing could be taught via a MOOC. So I wouldn’t even bring this up, except that I’ve heard from a couple of grad students that Stanford’s Writing In The Sciences MOOC is worthwhile. Motivated and self-directed grad students might even be the ideal MOOC audience. But another flavour of the workload problem applies. Teaching writing effectively has to involve evaluating and commenting on lots of writing through repeated revisions, and if this is enormously hard with a small class, it seems completely infeasible in a MOOC model. A course structure that relies almost entirely on peer commenting might make a MOOC approach feasible, but if it’s possible to learn writing with so little direct contact with the instructor, couldn’t we just remove the instructor altogether? I could be proved wrong, but I doubt that MOOCs will ever be a big part of the way we teach scientific writing.
(5) They could learn via writing assignments in other grad courses. This is essentially the usual undergrad model brought to the grad level, and it brings all its problems with it. Single courses make it hard to involve ambitious projects or multiple rounds of revision, and decentralizing the teaching of writing doesn’t make it less total work – just work for more people. Of course, it’s better than not teaching writing at all!
(6) They could learn through rounds of thesis revision with their own advisors. This is the dominant model in practice; we all help our grad students learn writing this way, and we always will. But we can’t expect advisors to do it all. It’s inefficient, and a bit soul-crushing, to correct all the same mistakes for the 35th grad student of one’s career; and until you’ve done it, it’s hard to imagine the amount of time it takes. Each student produces (one hopes!) a lot of writing, and each draft of each piece by each student needs thorough feedback that identifies problems and possible solutions, and points the writer towards generalizing them. If we’re to get anything else done at all, we need to escape some part of this work. (Note to my own students reading this: it may sound like I resent reading your drafts. I don’t! But if students can learn writing with fewer comments on fewer drafts, that has to be good for both students and advisors).
(7) They could learn from writing books. Given that I’ve just written such a book (details here), you’d think I’d be all about this option. And I am, in a way. I’ve learned a lot from guidebooks myself – not just books on science writing, but more general guidebooks like Strunk and White’s Elements of Style and Stephen King’s On Writing, and also technical references like the Little, Brown Handbook. Every scientific writer should own and study a couple of these*. However, we shouldn’t expect grad students to master scientific writing just by reading books about it any more than we’d expect them to master cabinetmaking that way. Writing takes practice and it takes feedback – lots of both.
(8) They could learn in self-organized writing groups. On the face of it, this seems like a dumb idea: if a bunch of people don’t know something, how can they teach it to each other? But peer commenting can be hugely valuable for writing. As a writer you hope to reach readers with less expertise than you have, and that makes a peer in a different lab a particular good source of comments. On top of that, it’s almost certain that members of a writing group will have complementary writing skills – one person may be have tricks for writing discipline, another may excel at grammar or at jargon control. A “writing group” could mean large, formal meetings, or it could mean a pair of students agreeing to exchange first drafts; either can help. But it would be naïve to think that writing groups can do the whole job. There are a lot of techniques to writing (“writing” in both its textual and behavioural senses) and it’s unreasonable to expect each writing group to independently rediscover them.
(9) They could be encouraged to practice. Ultimately, we get better at writing as we do more of it. Grad students could be encouraged to develop their writing skills by writing review papers outside of the thesis, by writing outreach pieces for local media, by blogging, and so on. There is no question that doing so will improve their writing, but there is a balance to be struck here: a little blogging is absolutely a good thing, but daily blogging that competes with writing the thesis is surely not.
Now, I tipped you off to this right up front: every one of these options can help, but none suffices on its own. So how should grad students learn to write? Grad students should direct their own learning, via combinations of (1) – (9). Now, that’s irritatingly vague, so let me try to do better.
First, grad students should take primary responsibility for their own learning, in writing just as they should in everything else. This is one of the major transitions of grad school, away from undergrad coursework-style “push” learning and towards the “pull” learning that will typify the rest of their careers.
Second, no single blend of options (1) – (9) will suit every grad student. But in general, I think grad students should:
- own and read writing guides (note the plural)
- look for available writing courses but not expect them to do too much
- organize themselves to take advantage of peer learning
- see thesis comments from advisors more as ways of being alerted to general issues and being directed to resources, and not so much as “corrections” or “edits”.
- see writing as more than just producing a thesis, and so practice writing in a variety of media
And third, and most important of all: all of us should realize that scientific writing is actually something to be learned! Maybe that seems obvious (especially at the end of this post) but it took me a very long time to figure it out – as a grad student I thought I would sit down and writing would just happen, and it was very frustrating when I discovered writing was hard. (It still is.) Writing is a craft that all of us need to practice throughout our careers, and we need to pay deliberate attention to it as much as we do to statistics, or chemical techniques, or whatever else our own fields entail.
© Stephen Heard (firstname.lastname@example.org) April 20, 2015.
Thanks to Auriel Fournier, Julia Mlynarek, and Chandra Moffat for comments on this issue and this post.
*One of them should be mine, of course. But you can also use your library to try out a couple or others, because they approach writing from different perspectives and no single book is best for everyone.