A professor in my department (http://www.cs.uoregon.edu/~dou/) is actually working on that specialized language problem, but he refers to it as "ontology translation". It seems a bit too strong-AI-ish to ever work, but apparently there is a large community working on the problem.

In the meantime, however, I do think that people who are not horribly deeply conversant in many fields will be more and more valuable in the future. I just wish everyone else though that way...