On rhetoric in scientific writing

Submitted by drupaladmin on 8 November 2011.

Scientific papers are infamous for being dry and technical. There are good reasons for that. A scientific paper is supposed to convince the reader of its conclusions on the objective basis of the evidence it presents. What matters, or what is supposed to matter, is what you say, not how you say it.

So is rhetoric—writing or speaking so as to persuade—ever legitimate? An influential line of thought, tracing all the way back to Plato, says no. But there are alternative views, tracing all the way back to the Sophists and Aristotle. Which I'm not going to get into beyond linking to Wikipedia, because this is only a blog post.* I just want to toss out some ideas on when it's ok for a scientific paper to be written so as to persuade.

First, I think that rhetoric should only ever be a supplement rather than a substitute for substance. If you go too far in trying to use rhetoric as a substitute for good data and good arguments, you end up with something like "The Spandrels of San Marco". All hat and no cattle. (Of course, I’m sure that Gould and Lewontin thought they had good arguments, which their rhetoric was merely supplementing.)

Second, assuming you have good data and good arguments, I think you should only supplement them with rhetoric if you have some reason for thinking that your data and arguments won’t be sufficient. Ideally, what matters in science really should be what you say, not how you say it. Plus, scientists as a group often are suspicious of rhetoric, suspecting that the author is trying to use style to cover up lack of substance. So for a scientific audience, the best rhetorical technique is ordinarily to avoid rhetoric. But if there’s some reason for thinking that what you say isn’t going to be sufficient on its own—even though, objectively, it ought to be—then I think its legitimate to try to write so as to persuade. If you have good reason to think that your audience won’t be objective, then I’m fine with using non-objective means to try to persuade them to be objective.

For instance, this is what I’ve done in my zombie ideas posts. In attacking zombie ideas, I’m attacking very well-established views that have already withstood numerous, objectively-correct attacks. Further, I don’t actually have any new data or arguments to bring to bear—I just want my audience to reconsider previous arguments which, objectively, should already have carried the day. So if I’m to get my audience to even seriously consider what I have to say, I need to shock them a little bit. And once I’ve got their attention, if I’m going to get them to give up what are apparently very strongly-held views, I need to use strong language. Hence the zombie metaphor and various other persuasive techniques I use in my zombie ideas posts. I suspect that Elias (1958) was thinking along the same lines when he resorted to mockery in order to stop the bandwagon of bad applications of information theory. And the same line of thought was probably one reason why Gould and Lewontin chose to write as they did.

I bet many scientists would buy this line of thinking in other contexts. For instance, if forced to debate a committed creationist, I suspect many scientists would have no problem supplementing logical argument and empirical data with The Flying Spaghetti Monster. What I’m suggesting is that the analogous situation—writing for an audience that you have good reason to think will be at least somewhat resistant to what you have to say—can crop up within science as well, albeit in a milder form.

*You get what you pay for on this blog.

AdviceNew ideas