This set up an unhealthy dynamic in which the student just wants to graduate, and they feel like they are fighting against the school and/or their advisor to be allowed to do so. It was like an endless cycle of begging "is this good enough to graduate?" and being rebuffed.
Conversation
Even at the very best schools, a PhD thesis typically makes only a tiny contribution. You are excited when it contributes to science meaningfully at all. Look. How many times you find yourself reading a PhD thesis out of interest, expecting to find new insights in it?
1
1
3
Aren't most "good" PhD theses in physics just collections of published papers?
1
1
How the thesis is built changes nothing to my argument. Most PhDs are granted on the basis of uninspired and useless contributions. If we really granted PhDs to people who made a genuine contribution to their field... there would be far fewer awarded.
2
1
If we did far fewer would gain the confidence to ever make major contributions. A PhD is a credential, not a prize. You’re right that most PhDs make no big contributions, just nominal ones that are enough to pass peer review. In fact most entire *careers* are like that.
1
1
I think this is fine. You need the low quality feedstock volume to get the high quality big bangs somewhere along the way. Your standard of competence is a way to kill the process statistics. The prize-like recognition you suggest is tenure or Fellow status at a company/society.
2
2
I was responding to the claim that we set the bar too high for a PhD thesis and people are stuck trying to reach that bar. The bar is not high. It is easy to get a PhD if you just do the work. Compared with creating a product or service people want to buy... it is a joke.
1
You might be conflating difficulty of work with difficulty of getting published results which vary ridiculously across fields. In CS, iirc, papers at good conferences are hard to get out and count towards PhD.
1
In my field of control theory conference papers are easy so committees/advisors demand journal papers which take 1.5-2 years to get out and capricious reviewers. Can delay graduation for years since committees at good schools won’t sign off until you have them accepted.
1
Fields also vary by the luck factor. You may work for 1-2 years to run experiments which turn out to be inconclusive through no weakness in your work. But then committees or journals don’t accept non-results as publishable and again you’ve spent 2 years for nothing.
1
1
I think your field of CS is outlier in STEM in a few ways: conference papers are accepted by committees which is a 3-4x speed up in iteration cycle (4-6 months instead of 1-3y), lower luck factor due to weaker experiment component, easier to do solo work. Net faster pipeline.
Physics and Chemistry, Life Science, Engineering have high PhD completion rates (about 60% to 70%). CS has a low completion rate (40%). These numbers don’t suggest a death march with unrealistic expectations.
1
I suspect CS deals with extreme industry poaching since that’s very lucrative. Most fields have much weaker plan Bs. Failure to complete is often actual failure, not being poached away.
1
Show replies
I would like to add my (very superficial and probably entirely wrong) observation on CS vs natural sciences like physics or bio. Since CS is on average closer to engineering, there is way more problems where one "just needs to do the work" and it's going to be fine. 1/3
1
2




