Archive for June 17th, 2007

How science is done: the nuts-and-bolts edition

June 17, 2007

Clifford has a nice post at Asymptotia:

Yes, part of my job is to sit and think about how the universe works. People hear this, and they wonder exactly what that entails. Well, it entails a lot of things – sometimes there’s the grand thoughts and the thought experiments and the like that you hear of from documentaries and books about Einstein and other famous scientists – but more often that not it is grungy nuts and bolts.

Some of the things are only too familiar to me:

I could not work out what the source of the two was. Was it physics? Did I mis-transcribe an expression at some point and lose it somewhere? Did I lose a two in computation somewhere in the notebook? These can be hard to isolate. I flipped to earlier parts of the big notebook from early in the week and found that I’d somehow multiplied at some point where I should have divided.

There is also more–like for example, nailing a problem (or, rather thinking that you have nailed it), only to realise later that you are wrong; and, sometimes, some freak accident which gives right answers in spite of mistakes–this happens to me especially while coding–I have seen obviously wrong codes giving me right answers making me believe that my code is right.

Take a look!

Electron microscopy, Tyndall effect and monkey scrotums

June 17, 2007

I just could not resist linking to this; via John Hawks.