I'm going to be giving a talk on academic writing next month, so I am starting to think about what I might say.
I plan to put my thoughts together in this thread, partly in case they're of use to anyone else, but mainly so that other people can add ideas for me to steal.
1/n
I plan to put my thoughts together in this thread, partly in case they're of use to anyone else, but mainly so that other people can add ideas for me to steal.
1/n
First thought: academic publishing is increasingly competitive. That makes it more important than ever to show that you've got something novel to say. You need to make your 'so what?' clear from the start. 2/n
One of my PhD supervisors, editor-in-chief of a journal, called this the 'worthy but dull' problem. So many submissions are methodologically rigorous, address an important issue, but completely fail to catch the eye or capture the imagination (at least at first glance). 3/n
Lots of things are important to this: covering letter; abstract; even title of the paper. In terms of the paper itself, I think what it requires above all is an interesting research question. And that means asking something that the existing literature hasn't yet answered. 4/n
Very occasionally, that gap in the literature is glaring. Much more commonly, it requires work from the author. That means not just reviewing the literature, but constructing it: curating, arranging, critiquing, finding oppositions and consensuses. (5/n)
This is a process that requires craft. You need to shape the literature, but carefully. Show command and authority, but be aware of sculpting a beautiful straw man. A false gap, tension or consensus tends to be obvious to the educated editor or reviewer. (6/n)
You can't typically cover everything, even within a generous word count, so acknowledge the big ideas and the key references you're expected to cite, and focus on the existing literature that's most relevant, showing that you know its strengths and weaknesses well. (7/n)
Use the literature to take you to what want to focus on: the debate, gap, or received wisdom that you think your study addresses. Then articulate this as clearly as you can, state your research question, and explain (briefly--without giving too much away) what you will add. (8/n)
Right, that's all I've got for now. I'll return to this later with further thoughts. (9/n)
Hello! With the presentation looming, I’m picking this up again. Next, I want to take a step back and think about some of the more general aspects of academic writing, starting with planning a paper. (10/n)
The first step, of course, is to think about your intended audience and your preferences with regard to journal, and consider carefully what both would want from your paper. Lots has been written on this, so I won’t say too much… (11/n)
…other than that you’ll get a lot more from reading recent, relevant papers in your journal of choice than just reading instructions for authors. It’s important to conform to journal requirements, but on their own they won’t tell you much about what is really published. (12/n)
Also, have a few journals in rank order of preference. Generally clusters of journals don’t differ too much in requirements, so it’s not as heartbreaking if (when) you’re rejected as it can be with a grant application. (13/n)
But do ensure that your submission is tailored to the journal you’re submitting to. Try not to do this. (14/n)
In terms of writing the paper itself, I imagine there are as many ways as there are authors. Here’s what I do, though, in case it’s any help. (15/n)
First of all, I have a rough idea for a paper—usually driven by data, but occasionally by theory. I can start by working it up in my head, but sooner or later I need to splurge* it down on paper. (16/n)
*Technical term. You won’t find it in non-specialist dictionaries.
*Technical term. You won’t find it in non-specialist dictionaries.
The key challenge at this point is that, although I may have some idea of the beginning and the end, I have very little sense of what the middle will look like. I usually have some of the right words, but definitely not in the right places. (17/n)
And excluding the odd postmodern display of pretence, papers need order. They need linearity. For me at least, the only way of reaching this linearity is by chucking all the ideas down and trying out different ways of arranging them. (18/n)
My preferred approach is spider diagram after spider diagram, on successive pieces of plain A3 paper. Here’s one I made earlier. (19/n) 
What I like about doing it this way is that it saves me from settling on a suboptimal structure prematurely. Once I’ve started writing, I get very attached to what I’ve written (more on this later). This way, I can play with ideas and orders without getting stuck to them. (20/n)
Eventually (and this can take a long time), I arrive on a structure that I think will work well. Next, I start to put it together in a more linear form—still on line-free paper, and still with lots of arrows and scribblings. (21/n)
Only after several more iterations along these lines do I start to write. I then tend to write quite quickly, because I’ve already done so much of the planning of the paper in my mind and on those A3 pages. (22/n)
As I say above, I think this is an individual process, and everyone needs to find their own way. What I do think is important to most people, though, is having an ordered, systematic means of getting from a bunch of nice-sounding ideas to a linear story that works. (23/n)
Other things that I’ve heard people use to this end are storyboards (with or without pictures), post-it notes, and simple bulleted lists. Some people are much better at discarding their drafts than I am, and so can get stuck straight into writing. (24/n)
