I ask this in all sincerity, because after talking to other students in other schools and other fields, I don’t seem any closer to an answer. Maybe it’s just because I think my department is weird, because we don’t assemble dissertation committees until we propose, and we propose fairly late (it’s pretty common for people to propose only a year before they plan on defending).
The closest thing to a consensus answer I can find is that committees exist to make sure advisors aren’t just handing out degrees. But if that is the case, it seems like there isn’t really a guarantee the average committee that doesn’t do much more than read the proposal and the dissertation would be effective at that. A group of less than half a dozen people who typically have two weeks to read a ~200 page summary of what is usually years of research can’t really independently verify the results that are presented. And if a professor really was intent on just handing out degrees to their lab, they could help make that data look more convincing. (I’m not saying this happens a lot. I don’t know for sure, but I don’t think so. My point is just that it seems easy to work around the supposed purpose of committees.)
I thought the point of a dissertation committee was to be a real committee, which in my mind means that at least part of it’s power comes from the fact that it is a group. Advisors can be great and all, but sometimes you need the perspective of other people to plan an experiment or help think through an interpretation of results. I thought the committee could help mediate part of the intellectual relationship between the advisor and student. Say a student wants to redo or alter some experiment but the advisor doesn’t think that it is worth the time; the student can try to convince the committee as a group of intellectual peers, and if they agree, they can essentially override the advisor’s wishes on behalf of the student. I think this is key because it can help diffuse some negative feelings in conflicts like this away from the student. (I don’t think the committee should take on issues that rise to the point of breaking up the advising relationship. Though if this works, I also think fewer issues should lead to the break up of the relationship.) I’m not sure if the converse matters as much because advisors do generally have a lot of control over what their students do, but if an advisor felt the student wasn’t doing something well, he or she could have the committee make it clearer.
So I’ll close with two questions I would love to hear answers from people in other graduate programs. First, when does you first assemble your committee? Second, what does your committee do?