“We’ve analyzed lots of orders and restaurants. What we find is that if you sit near a window, you’re about 80 percent more likely to order salad; you sit in that dark corner booth, you’re about 80 percent more likely to order dessert,” Wansink said.
The other day, I was explaining Writing Process Reengineering to a PhD student. I told him to articulate a key sentence at the end of the day and pick a specific time on the morning of the next to write. I told him not to let himself write at any other time. If the moment passes, he is not to write that paragraph at any other time the same day. He must make a new decision for the next day. And then do a better job of keeping his appointment.
He was puzzled about why. I told him it was jarring to his unconscious (the unconscious component of his prose, if you will), not just to be told it would be writing and then not get to write, but also to be forced to write at some unanticipated moment later that day. A double shock. I told him he would find his unconscious working more reliably if he kept his promises to it. He politely wondered whether I had any science to back that up. I did not claim that I did. I just asked him to take my word for it. To try it.
Andrew Gelman’s relentless critique of Brian Wansink is a good example of why I do this. It’s something I slowly realised after Freddie deBoer once pointed out that Arum and Roksa’s study isn’t, perhaps, as credible as one might hope. I realized I didn’t really know how credible it is. And I realized I didn’t really care. I had just found it useful to be able to cite a study that shows that writing makes you smarter and group work makes you stupider. It was a simple and effective way to present views I already held. Putting it the way I did, however, might have slightly overstated Arum and Roksa’s result. I didn’t really seem to care about that either.
Likewise, Wansink’s research, as Andrew keeps emphasizing, might very well be suggesting perfectly good health advice. The problem is just that he puts numbers on it and therefore might give the wrong impression about the effect size. Eat your rice on a black plate and you’ll probably eat less. But you’re probably not going to lose exactly 18 pounds. (And certainly not some even more precise number to three decimal places.)
I think I’m going to renounce science altogether in my area of expertise. I know what works for my own reasons. I’m not going to pretend to have “scientific” reasons to back me up. In fact, I think we should stop demanding scientific reasons for everything we do. Let science be what it can be. Mostly, we have to get by without it.
Tell your unconscious what you’re going to do the next day and then do it just as you said you would. I’m pretty sure that’s not going to do you any harm. And I’m also quite confident you’ll find it improves your writing. I could be wrong. But I’m not wrong about the science because I’m not claiming there is any science. If someone does come along with a study that shows I’m wrong, I will look at it very closely. But I promise I will not cite a study that confirms my views and say “See! I told you.” I’ll only do that when you come back to me and tell me my advice worked for you. As Van Morrison sings…
And if you get it right this time
You don’t have to come back again
And if you get it right this time
There’s no need to explain