> ...establish beforehand whether it's a worthy thesis... I think you're describing the process of putting together a dissertation proposal, and I think it mig

> ...establish beforehand whether it's a worthy thesis... I think you're describing the process of putting together a dissertation proposal, and I think it might be the hardest part of getting a PhD. To do that, one first needs to acquire a broad knowledge of the field (which is what the qualifying exams are testing after one finishes all the required coursework) so that one can read papers and understand them. Then one needs to do a deep literature review of a narrow area to see where there are gaps or opportunities. And then one needs to come up with ideas to address those gaps and opportunities. The proposal is the presentation of all this, and the committee decides whether it is worthy (novel, impactful, etc.) or not. That can take a few years. Once the proposal is approved, one "just" needs to execute on the ideas; that's still a lot of work (like years of work), and could easily go off-the-rails or show that early assumptions were invalid or whatever, but I think it's more straightforward than trying to brainstorm worthy research ideas (that are appropriately scoped for a PhD) in the first place. One way to get in trouble at the defense seems to be deviating too much from the proposal without keeping the committee members informed. Or people have turnover on their committee, so the people who approved the proposal aren't the same people at the defense. In theory, these problems should still surface before a public defense is scheduled; like if committee members are surprised or have questions, that should come up in private and be resolved before there's any public embarrassment on either side. In the two ugly cases I saw, people basically weren't communicating with each other due to personal grievances, so this didn't happen. > I guess the defense is more about establishing whether the work is really your own and you didn't have other people do the work that you don't understand. I don't think so. Those things are too fundamental and shouldn't even be in question. I'd say that the defense is mostly about tradition. After that, I think it's mostly about one's ability to clearly communicate the work. In theory, a PhD should be able to teach, so it kind of tests one's ability to go out into the world and do that while representing the committee / department / school in a positive manner.