‘Intuition’ (by Allegra Goodman) takes place in a cancer research lab where one of the junior workers, Cliff, thinks he’s found a virus which can cure cancer in mice. The lab is in financial trouble, so its directors are grateful for the chance to publicise these results and use them to raise funds. However, it proves impossible to replicate the results, and another post-doc, who happens to be Cliff’s ex-girlfriend, suspects him of fraud.
Some reviews of this book have suggested that it’s never made entirely clear whether Cliff has knowingly committed fraud, or if he has unwittingly made a mistake. But there is a clue quite late on, when we get a glimpse of Cliff’s reasoning;
‘He had not chosen to discuss every piece of data, but had run ahead with the smaller set of startling results he’d found. Still, aspects of his data were so compelling that in his mind they outweighed everything else. He had sifted out what was significant and the rest had floated off like chaff.’
From a scientific point of view, the first sentence of this quote is damning. Cliff just doesn’t seem to be a very good scientist. When you run an experiment you cannot pick and choose where your results start and end. If you start with a hundred mice, you must discuss the results of the experiment on all of those mice, not just the few that happen to show a good result. This is because there’s always the possibility that your good result happened by chance. If you spot a good looking result in a subset of your data, it may just be a random fluctuation. (That is also true of the whole dataset, of course, and should be quantified as far as possible.)
It’s bad science to select a subgroup of interesting results after the event, but it’s depressingly common. When I was an astronomer, I worked on quasars. There’s a so-called ‘controversy’ about whether or not these objects are at the incredibly large distances as implied by their redshifts, if the Big Bang model is correct. This controversy was a genuine problem when quasars were first discovered in the sixties, but has now fizzled away. Only a few ‘maverick’ astronomers, such as Halton Arp, now believe quasar redshifts aren’t cosmological. The controversy is an artefact of only choosing the data that fit the hypothesis and ignoring all the other data, in this case of looking at quasars that appear to be near to much lower redshift galaxies. Statistically this doesn’t happen any more than you would expect – but superficially it looks ‘interesting’.
One nice aspect of ‘Intuition’ is that the point of view is omniscient and no one character is particularly favoured; rare in modern literature. The reader gets to inhabit the minds of all the major characters and understand their view of the drama unfolding around them. Even so, with all this information that we are provided, it proves impossible to understand the reasons behind the characters’ actions.
This failure of the omniscient narrator to get to the truth of the matter could be read as a warning – are we deluded in hoping/expecting science to be an impartial tool for understanding the external world? Or am I confusing science with omniscience?
No comments:
Post a Comment