Monday, January 9, 2012

Big Wide Worl

Striving for, and defending, originality

Andrew Pontzen, astrophysics postdoc
rexfeatures_1403409a.jpg(Image: OJO Images/Rex Features)

The process of generating new scientific ideas can be difficult. Persuading other people that the ideas are novel and valuable is even harder. Where do we start?
Rather like art, new science must contain some mark of originality, but it also needs close connections with previous work if it is to find an audience. Of course being too original is a common problem only for geniuses: Van Gogh sold just one painting in his lifetime. Tickets to Schoenberg's premieres sold better - but mainly to protestors who liked to whistle through his performances. Physicist Hugh Everett struggled to sell his "many worlds" interpretation of quantum mechanics to anyone.
For the rest of us, the challenge is reversed: how can we be sufficiently novel? One approach is to combine many old thoughts in a new melting pot. With patience, it's then possible to be reasonably systematic about allowing new ideas to form.
The process is not always that methodical, though. Take one of my recent ideas - I was sitting in a conference in February and it just popped into my head more-or-less fully-formed. (That is, of course, a description of the subjective experience rather than a neurological analysis. The idea itself, as will become clear, was subconsciously triggered by previously accumulated knowledge. Rather like how J K Rowling says that the Harry Potter story came to her in a flash of inspiration, but she surely wouldn't deny its debt to earlier literature).
The idea, roughly speaking, was this. Let a marble roll around freely on a concave rubber sheet. Assuming no energy is lost to friction, the ball will keep moving around forever, the exact motion dependent on the contours of the sheet and the initial position of the ball. Now suddenly stretch the sheet, wait a while, then release back to its original shape. Is the marble's final motion equivalent to its initial motion? Or is it now on a different path?
The answer is that it's on a different path, a path that I could calculate mathematically. It turns out this kind of calculation can show how supernova explosions change the orbits of dark matter particles within galaxies. I tested the idea against detailed computer modeling of galaxy formation and, over a couple of months, wrote a paper showing that such a simple model can accurately explain some recent developments in state-of-the-art supernova simulations.
The reaction from the astrophysics community has been sharply divided. There are people - happily the majority - who are enthusiastic about the new calculations. The paper's initial citation rate has been higher than any paper I've written previously. But there is a small group who think that it's nothing new.
When I announced the result at a conference, one famous astrophysicist was heard to whisper loudly to his friend: "have we just travelled back to the 90s?" It's a favoured method to shoot down a young researcher: assert their calculation merely restates some earlier result, preferably a result that the gunner himself discovered.
Scientists are used to coming up with an idea only to find it's already known and understood. The internet makes it relatively easy to track down relevant prior publications and so abandon ideas that are adequately explained elsewhere. I had done this search and decided that my results were new, so was taken aback by the strength of the minority negative reaction. Things got worse when the referee, a scientist tasked with checking the paper for the journal, sided with the "not-new" camp. I had a few sleepless nights wondering whether I'd committed one of the original scientific sins: rebranding an old idea as my own.
But after re-reading the old articles I have concluded that I am innocent. Luckily a referee for a different journal agreed that the mathematical technique adds to our understanding. Yes, similar effects have been seen before, but my paper already explicitly refers to a vast number of previous studies. I now plan to update it with a clearer discussion in light of the more constructive complaints. In the end, though, the new approach really does allow for unprecedented precision in describing how explosive events are capable of changing the properties of dark matter.
Even if it weren't such a tangible step forward, what would have been wrong with publishing a new way to calculate an old result? Would that have been a creative failure? I've given up troubling over it. As Stephen Sondheim's musical Sunday in the Park with George suggests: "Stop worrying if your vision is new; let others make that decision, they usually do! You keep moving on."

http://www.newscientist.com/

No comments:

Post a Comment