Early this year, Heidi Harley and I posted a few times about the job market in linguistics. I got the ball rolling with a post about the Linguist List's 2008 job ads. Heidi followed up with a comparison between the job-ad numbers and the indices at ProQuest, and then I put those numbers together.
The combined post told a dispiriting story about the theoretical areas: it looked like there was a serious mismatch between the number of PhDs and the number of jobs for them. I think now, though, that this was based on an unfair comparison. John McCarthy pointed out to me that the ProQuest counts we reported are not relativized to dissertations per se, but rather come from a more general search of ProQuest's databases.
If the ProQuest search term restricts always to (i) dissertations, (ii) the Linguistics subject area, and (iii) the last five years, then one gets much smaller numbers throughout, and the hits themselves seem generally reasonable. Here's a graph comparable to my earlier one, with the same job-ad numbers but the more restrictive ProQuest numbers.
This tells a different (and more heartening) story that the previous version. While it continues to support the notion that it is very smart to get computational or experimental training, it suggests that the theoretical areas have actually achieved a good balance between the number of PhDs they produce and the number of jobs available. (Of course, it is still a very tough market.)
There are lots of caveats (see the comments on the previous post for some important ones). In particular, the applied, cognitive, and computational areas are hard to assess, because of the numerous ways that scholars can identify their own work. I uniformly restricted the ProQuest searches to the subject area Linguistics, on the assumption that people browsing Linguist List for jobs will also be inclined to assign their work to the Linguistics subject area, and vice-versa. Here is a table of the ProQuest seaches I did:
|Language and cognition||