[antognini]

There is a well documented trend in the sciences away from research performed by scientists on their own or in small collaborations towards research performed by very large collaborations [1]. My understanding is that this has been going on for a long time in biology and physics so that nowadays it's very difficult to make much of an impact without participating in a large collaboration. (This is more true with regards to experiment than theory.) This trend has also been noted in my field of astronomy, though there hasn't been nearly the same crowding out of individual astronomers as in physics.

I could imagine that this trend has made it more difficult for more eccentric scientists to succeed because they might be less able to "play nicely with others" in a big collaboration.

[1] http://www.sciencemag.org/content/316/5827/1036.short

[ihnorton]

In physics at least, isn't that because most of the low-hanging (low-energy..) fruit is gone? CMS or Atlas at the Large Hadron Collider are feats of engineering, not to mention the physics, and doing all of that takes a large organization. In theoretical physics, people seem to still be doing reasonably well with the lone-genius thing (although I suppose Peter Woit disagrees).

Biology is interesting because small, high-impact groups become arguably more sustainable every year due to cheaper, smaller instruments and the rise of contract services (e.g. DNA sequencing). On the other hand, many of the human problems biology is attempting to address are hard to inferentially study on a small-sample level, again necessitating large organizations.