> I'd prefer something like "The purpose of PhD programs is to advance the field".

If you read Wikipedia under 'Doctor of Philosophy', you will find that a Ph.D. was once more of a prestigious title you got after doing the scholarship:

"The first higher doctorate in the modern sense was Durham University's DSc, introduced in 1882. This was soon followed by other universities, including the University of Cambridge establishing its ScD in the same year and the University of London transforming its DSc into a research degree in 1885. These were, however, very advanced degrees, rather than research-training degrees at the PhD level—Harold Jeffreys said that getting a Cambridge ScD was "more or less equivalent to being proposed for the Royal Society."

It is still possible to get a doctorate in this manner. Please see wikipedia under 'Doctor of Philosophy by publication'.

"A Doctor of Philosophy by publication (also known as a Ph.D. by Published Work, PhD by portfolio or Ph.D. under Special Regulation) is a manner of awarding a Ph.D. degree offered by some universities in which a series of articles usually with a common theme are published in scholarly, peer-reviewed journals to meet the requirements for the degree, in lieu of presentation of a final dissertation. Many PhD by Publication programs require the submission of a formal thesis and a viva voce."

It is offered in several countries in Europe. The wikipedia entry is incomplete: it is not just offered in the UK.

Furthermore, it is relatively common to get advanced degrees from well known universities (e.g., Harvard) without having an undergraduate degree.

I see your point and I think that brings us a bit closer to alignment. But I think if someone is __incapable__ of writing such a paper there would likely be many larger flags and they probably should not have been able to pass their undergraduate curriculum.

I do want to make it clear: I'm not opposed to arbitrary filters when there is a high number of applicants and you simply need to reduce the number. I am opposed to pretending that such a filter is not arbitrary. I think we need to be clear about how strong of a signal any filter is, and be quite explicit that they are not all equal indicators. That is my main point: being explicit about the strength of a signal.

On regards to training, I do agree that schooling isn't __just__ training, but I'd fully disagree that this isn't one of the most important aspects of it, even in grad school. Your first two years (in US systems) are nearly identical to a masters and highly focused on classes. What are classes but training? Even being a TA or lecturer is, in part, training (full instructor of record would not be). Post conditional, I still think you are in training at least up until candidacy. That is much more arguable given the variability of advisors, with some being very hands on (training) and some being very hands off (on your own).

I'd prefer something as "The purpose of PhD programs is to train people to advance the field." Because by all accounts, it seems like you've done this (even with the self-deprecating humor. That is exceptionally common in PhDs too lol). I still maintain training because this isn't the end, but the beginning. Post PhD is where you can choose to go to be an academic researcher or industry researcher (or abandon research). Those are the actual jobs (which should have continued training) but your degree is more akin to a certification from your institution. You do come out with a body of work that is distinct from the institution, but the institution's goal is not to keep you around and continue performing work. They are explicitly formed to graduate you. To educate you. And what is education except a form of training?

Fwiw, I think we are decently aligned, but sometimes text is hard to communicate, especially post by post. I do think your critiques are valid, even where I disagree.