a.k.a. Musings about predictability in science
One of my favourite books is Death and the Penguin, by Ukrainian author Andrey Kurkov. The protagonists of the book are Viktor, a young writer who’s struggling to survive in post-Soviet Ukraine, and Misha, a king penguin, whom Viktor adopted from the local zoo after it went bankrupt and who now lives in Victor’s bath tub. To pay for bills (and frozen fish for Misha) Viktor accepts a job writing obituaries for the local newspaper. However, there’s a twist: the people he writes about aren’t dead – yet. Once he writes about their sad passing, they mysteriously die. For example, a senior politician he writes about, falls from a sixth-floor window: “was cleaning it for some reason, although apparently it wasn’t his. And at night.”, the editor of the newspaper comments.
Recently, I was reminded of this obscure plot in a somewhat surprising setting: as an avid listener to pretty much all Nature podcasts, I also follow Backchat, a monthly podcast produced by the Nature team, where they discuss some behind-the-scenes of various stories. One of the things I have found most fascinating is how (or rather when) some of the stories are written. Yes, you guessed it: they’re often written before they happen!
I first heard about it, when the team was discussing Brexit, and they mentioned that they had prepared two sets of articles (one for a “Yes” and one for a “No” outcome). At the time, I thought it was a lot of work, but kind of made sense – after all, it was an important event to cover, and there were only two possible outcomes. But more recently, two other issues with pre-written stories were mentioned (a Nobel for LIGO that never happened and an (un)successful European Mars probe landing), which really made me wonder…
Surely, writing a story before it has happened is a risk (you’re investing time/energy without knowing for certain you’ll be publishing it), so this practice must reflect the extreme pressure to stay ahead of the curve. But in addition, I wonder if it reflects how predictable science is? And if it is predictable, on what timescale (a week, a month, years…?), does it apply to all fields (obviously some projects, such as space missions, have an aspect of long-term predictability, but others?) and does it have any implications for how we do science?
Clearly, this is somewhat philosophical and far removed from an amazing news team covering stories in a timely matter, but with what certainty could a good analyst predict where science is headed? And are we, the people who do the science, just following a pre-set path, without even realising it?