Accept Defeat: The Neuroscience of Screwing Up

[Nasib]

I would be interested to hear of any professional or personal views that members here might have regarding this "unprecedented research project", the subject of this article published in 'Wired' Magazine 21.12.09

Accept Defeat: The Neuroscience of Screwing Up (by Jonah Lehrer)
　
The article can be found here:
http://www.wired.com/magazine/2009/1...t_defeat/all/1

[Harryprice]

Interesting article. The thing about science is that it relies on creativity which is a messy and poorly understood process. While outsiders and newcomers can produce breakthroughs they can also produce massive misunderstandings because they don't know the basic subject well enough. As the article says, it usually takes the expert insider to recognise that an outsider's question or idea could well be right. And the most creative scientists usually use metaphor or analogy to describe or examine their ideas. Famously, Einstein discovered relativity by imagining what it would be like to ride on a particle of light.

[Pebble]

The problem with advances in science, is that small incremental improvements in knowledge are rarely scrutinised, they will be accepted and as long as consistent findings are found when other labs repeat the experiment knolwedge will advance. By contrast substantial shifts in knowledge, will obviously be challanged, here it is not good enough to show that you have found something anomalous, you must prove that no methodological error explains the anomaly. This is difficult and tedious work, and without an explanation or at least a credible theory, why would anyone take any notice.
I think this article glosses over just how difficult it is to prove a negative. Showing that you can reproducibly repeat an experiment that is consistent with a new and better theory is one thing. Being able to produce an anomaly which in itself adds zero to current knowledge, and demonstrating that the result is truly at odds with current theories is extremely difficult.
In each of the examples provided, it was the provision of an acceptable explanation for the results that led to the advance, not the observation of an anomaly.