The following three sub-fields are highly specialized: Ancient philosophy, seventeenth/eighteenth century philosophy, and philosophy of physics. The following sub-fields have a low level of specialization: metaphilosophy, philosophy of religion, philosophy of probability, philosophy of the social sciences, decision theory, and philosophy of race and gender. Highly specialized sub-fields tend to require extensive knowledge in some area beyond the typical training of a philosopher, and outside of philosophy proper.--Brad Wray.
Brad Wray, a Kuhnian-naturalistic philosopher of science, has mined the PhilPapers data with an eye toward "the degree of specialization in each area of specialization" in the discipline. (Wray is a bit too confident that this is a "representative sample of the profession;" I worry about selection and, especially, geographical effects; even so the numbers are pretty large (3,226 people in total and 1,803 'philosophy faculty or PhD') so that the results can be illuminating if used with caution.)
Wray: "The degree of specialization of an area is a relative measure of how specialized a particular area is" and is calculated as follows: "The number of people who claim the area as their primary area of specialization/The number of people who claim the area as an area of specialization." I have posted a chunk of the abstract, which contains the core results, in the epigraph above. One of Wray's finding would not have surprised Adam Smith: "an analysis of the data suggests that the size of a specialization is correlated with the degree of specialization."
Wray's crucial result (which seems to have been explored at the prompting of a referee) is this one: "a high degree of specialization is the exception, not the norm in philosophical specialties. Many specialties seem to depend, to a significant degree, on the involvement of many who work in the area but who do not identify the area as their primary area of specialization." Of course, this says nothing about the way in which specialists set the agenda with a specialization.
Either way, this data suggests that there are still quite a few generalists in philosophy (it is amusing to me that I work in a 'specialist' area because us 'early modernists' cover two hundred years of systematic philosophy with ongoing discussions pertaining to M&E, value, science, and increasingly philosophy of religion). The question as to what degree Wray's pattern is born out by publication and citation-data is worth exploring in the future.