Phew. You made it past the midcourse crisis. You reaffirmed your commitment to research, and now the data are just pouring from your computer or your test tube or your sequencer. About this time another question--"When am I going to publish all this?"--will probably start to dance in your head. And even if you have yet to get around to thinking about papers, I guarantee your advisor has. Refereed journal papers are the coin of the academic realm. The sooner you start filling your purse, the better.
But how exactly do you do that?
Good question. There are no simple rules for turning a jumble of numbers, hypotheses, and hunches into a clearly argued scientific paper. And even if such a recipe existed, you probably wouldn't know about it. As far as I know, not a single graduate science program in the world demands that their students take a course in science writing. So the responsibility for learning how to write about research inevitably falls to the student, hopefully led by their advisor.
This time-honored system is not the most consistent way to teach scientific writing, because it depends so heavily on the abilities (or otherwise) of the PI. If you don't believe me, read any science journal. Bleah! Most of the writing borders on incomprehensible. But if you get lucky, the set up can work. As I have previously described in The Spy , my advisor taught me to write. I was lucky. Not only is he a solid expository writer, his wife is an editor. I suspect that many of the lessons he taught me originated with her.
The most important thing he taught me was to read before you write. As you prepare to write your first paper, gather up some articles written by a variety of scientists and read through them. Stick to papers in your own field, and if you know what journal you want to submit to, stick to that journal. Each field, and sometimes each journal within a field, has its own conventions that dictate everything from what goes in an abstract to the format for references. Some papers will seem totally incomprehensible from the very first word. These are poorly written. Others will be models of clarity. These are the well-written papers. Imitate them. Look at how the authors construct their arguments and try to use this same structure to present your own data.
The Writing Coach
If your advisor is indifferent to writing, or indifferent to teaching you how to write, don't despair. Just hire a writing coach. Universities are filled with great writing teachers. Humanities professors and graduate students often tutor students who are having trouble getting their thoughts on paper. Some enlightened universities even hire a few writing consultants to help scientists. Their primary charge is to bring nonnative speakers up to speed, but writing coaches can be a great help to everyone.
A writing coach usually works one-on-one or with a small group. Each student brings in a piece of writing. The coach reads through it and points out places where the writing can be improved. They also suggest general strategies for strengthening the weak sections. But just as a tennis coach doesn't play the matches, your writing coach won't rewrite the paper for you. That is your job.
While that may sound simple, in practice it is often hard for novice writers to separate the good from the bad and then divine the techniques that good writers use. If you are having trouble with this, ask your advisor to recommend some well-written articles and what it is that they particularly like about them. But in the hit-and-miss world of learning scientific writing, you may find that your advisor is as mystified as you, leaving you with two options: muddle through or look for some outside help. I recommend the latter course.
Beginning writers can choose from a number of educational options. Every college offers several basic writing courses every semester, but these may be too general and time-consuming for a scientist's immediate needs. One glaring exception to this rule is Rice University's Cain Project , which aims to teach scientists at Rice how to improve their communication skills. Many universities also have writing labs where you can have your writing critiqued by an experienced writer or editor. The best way to find these is usually to surf over to the English department Web site, but if that doesn't work, you can also call the registrar's office.
You might also thumb through one of the many books devoted to writing. There are lots of these books ... so many, in fact, that a cynic might be tempted to paraphrase the old saying, "Those who can write, do. Those who can't, write a book about writing." Cynicism aside, many of these books offer valuable advice (a personal favorite is Writing with Style, by John R. Trimble; see the Survive and Thrive Resources  page), although I have never come across one that spoke directly to the problem of writing a paper for a refereed journal.
Writing instructors have also taken to the net, and the number of Web sites with helpful information for writers has exploded. Instead of trying to give you a comprehensive overview of these sites, I'll direct you to the professionals at the National Writing Centers Association . And if you're after science-specific information, then try the Links section on the Cain Project  Web site.
Now that you are producing papers like a professional, the obvious next step is to get the information out where your colleagues can see it and be duly impressed. I will address that topic soon. But first there is a more immediate concern: the intense pressure that all graduate students feel as they struggle toward graduation.