Isolated Social Networkers

On the sometimes unfortunate effect of leaping between fields (I could not resist stealing the title) - thanks to Peter for sending me this - can't help but suspect there's an awful lot of this in some of the Agent, Semantic Web and Web Services research which is now going on.

Most of the respondees to the crooked timber post interestingly focus on wasted effort or the lost craft of looking up references. However, as one person points out, sometimes the route to innovation is to try to tackle a problem without checking out what was done before - without getting into the rut of previous approaches. Of course the problem arises when:

• Once solution in hand one doesn't check back to see if the "new idea" really is new.
  
• Academic plaudits (or even minor brownie points) are accumulated in one community on
  the basis of something well established in another.
  

Of course the later, if unchecked, might have a long run detrimental effect on the scientific community involved. I wonder if we're always as good at policing ourselves as we should be?

UPDATE: reading the excellent comments it seems that agent researchers are already popular with socilogists to "However, there is still a plague of mediocre physicists reinventing the wheel (and now, god help us, they have been joined by mediocre computer scientists producing “agent based simulations”)."