the reality about specialization

[gusl]

Interests can be divided into "gee-whiz" interests (i.e. hobbies, etc) and "serious research interests" (areas where I'm trying to make an impact in the near future). As aspiring world-experts-in-something, PhD students have no time for hobbies, or so I'm told. The idea is that the former type of interest is important too, but only insofar as it's needed to keep us sufficiently entertained to be able to focus on the one-thing-that-people-will-associate-with-my-name-for-years-to-come.

To keep this vivid in my mind, I made these pictures:


Local expertise


Global expertise

Suppose red and light-blue are competing for a position in a department populated with the other colors.

As we can see, red is well-rounded, and has lots of knowledge that his/her colleagues don't have, which suggests that he/she could be an excellent collaborator. But light-blue has clearly made his/her name in one field.

Who gets the job? Who gets the grants?

Light blue does!

Apparently this is true even if his/her specialty is a very theoretical area. Red could potentially get lucky if the hiring committee isn't good at assessing world expertise, because it is outside of their areas and they didn't do a good job of eliciting world expertise, or don't care (e.g. because it's for a teaching position).

---

UPDATE: and then we read bios like this:

<< The research contributions of Professor Meng cover almost all areas of statistics, including statistical issues in astronomy and astrophysics, modeling and imputation in health and medical studies, and elegant mathematical statistics. >>

talk about a mixed message!

Comments

[infryq]

Shouldn't it depend on the position that's up for grabs? I would like to think that a department chair or other management heavy position would require moderate expertise in as many of the available local fields as possible, so as to make resource allotment decisions with aplomb. Whereas light blue should clearly be getting grants as a PI. Red would be a good second name on a grant, particularly for a more collaboration-heavy field like robotics or bio-CS, but wouldn't do well for top billing.

(of course what "ought" to work and what the actual politics are need not line up...)

[gustavolacerda.livejournal.com]

<< I would like to think that a department chair or other management heavy position would require moderate expertise in as many of the available local fields as possible, so as to make resource allotment decisions with aplomb. >>

Which resources do you have in mind, that would need knowledge of many subfields? Research money is in the form of grants, which go directly to the individual researcher (after the university takes a big cut).

Dept chairs need to be tenured which, in research universities, means that they need to have a rather pointy area, at least during their years as assistant prof.

[infryq]

Administrators, space, IT, teaching assignments, speaking engagements and talk/seminar schedules... Also, one needn't stay pointy forever. You have to be pointy to get a PhD, too.

[gustavolacerda.livejournal.com]

I don't see how the management of these resources would benefit from a chair's breath of knowledge beyond that of anyone with a PhD in the field.