January 23, 2007

Papers are grants, papers are blood, papers are life!

My little title phrase today is from the LabLit article Do you have what it takes to get your scientific article out there? by Keren Boren. Boren is a post doc in France and the article recounts her efforts to get papers published in her field. The trauma of finicky experiments, the fear of journal rejection, the capriciousness of the reviewers, the joy of the feeling of validation that publication brings, the career worries, the blood sweat and tears just pour threw the pixels of the article. It's a great insight into the lives of all those grad students, post docs and junior faculty we see scurrying around campus.

We postdoctoral scientists don’t just slave over grant applications and research 24/7 – we have to write papers about our experiments and submit them for publication. And these aren’t just some sort of formality, tying up the loose ends and documenting a discovery – they frequently make or break your scientific career. It truly is publish or perish: papers are grants, papers are blood, papers are life!

*snip*

But one day, all the experiments are done, you have read all the related scientific literature and the paper is written.

You think you’re done? Not at all – in fact, it’s only just begun.

First, there is the eternal discussion with your co-authors about that all-important decision: which journal? At least one co-author will always recommend aiming for the very top: Nature. Yet others, the more pessimistic ones, go to the opposite extreme and advocate the lowest of the low, just to be sure to get it in with as little pain as possible. When you finally agree on a journal, you need to look at their ‘Instructions for Authors’ webpage to put the manuscript into the approved format. It’s some sort of law of the publishing universe that, no matter how short your paper is, you will find that it is two-thirds too long for the journal in question.

It's really a great article. But it also gets me thinking, especially after listening to Jean-Claude Bradley's Open Science presentation from the Blogging Conference. If blogs and wikis and other social software constructs begin to seriously supplement/replace traditional journal article publishing, will these lab rats lives be easier, less pressured to produce these articles. Or, will they be harder, with the completely interactive demands of the Net demanding that they produce more and more, faster and faster to keep up? Who knows.

On the other hand, if you believe Gregory Benford, there is a way to produce the perfect scientific paper. Perhaps this is the way to go?
In this paper a new scheme for paper-organising is proposed. It does not rely on weaning scientists away from the passive voice sentence, like that last one. Instead, we should recognise how scientists actually read.

Our calculations, statistics, and closely-reasoned analysis appears in the body of the main text. First we summarise our results with merciful brevity.

Conclusions
While reading a scientific paper, scientists are led by two needs: (a) ego and (b) desire for information. Our research shows that the former always predominates. Therefore, papers should be organised to satisfy this. The preferred scheme follows:

1. TITLE
Maximise trendy buzz words, even if irrelevant. (Indeed, some will misread this non-connection as going over their heads.) Try to include many verbs that end in -ise and -ishness.

2. AUTHOR'S NAME
Avoid initials. People remember actual names. Let your students be represented by their initials if they want; readers will assume they are nobodies.

3. REFERENCES
The most important part of the paper, yet the most neglected.
References cited must contain a broad spectrum of sources, to insure the greatest probability of naming the reader, and especially, of saluting the referee. Use multi-author papers to maximise the number of people mentioned. Corral any paper even slightly related to your field; Nobel winners' papers are of course preferred, no matter how thin the connection.

A scientist will always give greater attention to colleagues who cite him, if only to find where in the text you mention him. Thus the best strategy is to cite everybody you can but place the citations in an unlikely place in the paper. They would then have to read carefully to find it, and so might even discover what the paper is about. The highest-risk strategy is to cite someone in the list of references but not in the text. Then he will have to read the whole paper. The disadvantage, of course, is that he will be livid with rage and frustation by the time he finishes. But at least he will not forget you!

4. ACKNOWLEDGEMENTS

Another important ego-feeding ground. Thank the big names in your field, even if your sole contact with them was toting coffee at a conference three years ago. The list should be lavish, implying close connections with all the movers and shakers. Avoid mentioning dead people; they can do you no more good, and their rivals are still around. If space permits, include those who actually helped you.

5. GRANT REFERENCE
Your grant monitoring officer will always look for this, so put it early. Others will want to know what agency got suckered into paying out.

6. INTRODUCTION

Here you explain what you plan to do. Promise a lot. Few will reach the main text to see if you actually did it.

7. CONCLUSIONS
Always overstate your results. This is a firm rule - everyone will expect it, anyway.

Claim certainty where you have vague suspicions. Use statistics as an art form, not as a serious check on your work. Why be a sceptic about error bars, after all?

Graphs proudly showing agreement between theory and experiment should be prominent. Only in a footnote (tiny type!) should you explain that the theory has been scaled to the experiment in the first place, the coordinates multiplied by a fudge factor, or other artful dodges.

8. MAIN TEXT
With any luck, there will be no need to actually write this section. Everyone will have turned to the next paper.

I really enjoyed Benford's self-deprecating article, it has as much insight into the scientific (academic?) mind as any twenty sociology articles while also having the benefit of being laugh-out-loud funny.

These two articles together provide a very interesting insight into the world of scientists, one that often gets overlooked by more formal treatments. A little angst, a little joke, a little fun, makes the message a little easier to take. Benford's novelist's eye for human foibles is really evident. Read the full text of both.

Update: I forgot to mention that the Benford article was via Locusmag blinks. As well, Jane reminded me of her post from a couple of weeks ago, The little paper that could, which is certainly in the same spirit as the the Keren Boren article above.

3 comments:

Jean-Claude Bradley said...

To comment on whether life is easier or harder for lab workers using Open Notebook Science, I think the challenge has been modified. For people who already know that your experiments are posted in real time, it gives an unprecedented opportunity for peers (and future employers as well hopefully) to interact with and appreciate the strengths of individual researchers. But for others who are not plugged into the RSS world there is still the work to be done in letting them know that this is available. That means giving (and going to) talks, poster sessions, emailing authors, etc.

Jane said...

Thanks for the links! The first one certainly rang true to me with my recent experience with getting a paper (finally) accepted. And the second....well, let's just say I'm in the middle of writing a journal article, and this is just the levity I needed.

John Dupuis said...

Jean-Claude, Jane, thanks for the comments. It's interesting that you both got something entirely different from the post! As for stress levels, I guess it is this transitional time that's the worst as you really have to have a foot in both the old and the new worlds. Interesting times.