[anouk_anca]

> the apprentice must simply produce three publishable research papers within the committee's set time limit

One big issue with the current PhD system is reliance on published articles to determine a student's worth, which transforms academia into a paper-churning factory. In your scenario, having a set time limit combined with the low pay will still make candidates rush to publish papers, perpetuating the current issues.

Ideally, I would prefer a system like in the Netherlands, where PhD students are considered employees of the university, and get paid close to median wage, and get contributions to a retirement plan. It's a demanding job, and should be paid as such. I would also like to see the recommended period for finishing a PhD extended to 7-10 years, ideally spent in 2 or more research institutions. Funding might be a concern, and universities might have to reduce the number of accepted students, but it's worth it if it reduces the current pressure to constantly publish. The process of scientific publishing itself needs to be revamped, to encourage publishing of failed hypotheses, or replication of old experiments.

[dagw]

I would also like to see the recommended period for finishing a PhD extended to 7-10 years, ideally spent in 2 or more research institutions.

That's basically describes a PhD plus a postdoc today. Which is essentially the minimum requirement for kicking off an academic career.

I completely agree that PhD students should be full employees and get paid basically the median wage of a first job in the relevant field.

[pcrh]

>I would also like to see the recommended period for finishing a PhD extended to 7-10 years

That is way too long. A PhD should only be long enough as is required for the average competent person to demonstrate their competence in research. Adding extra bells and whistles, more requirements, etc, does not improve their abilities and only prolongs the apprenticeship to the advantage of their supervisors.

[anouk_anca]

A PhD is not so much an apprenticeship though, its goal should be to bring a novel contribution to the state of the art in the field. The current system does not need more requirements, but it does need more time and freedom for PhD students to develop truly novel ideas, and not fall prey to the "publish or perish" mentality that is so pervasive in academic culture. Dependence on supervisors can be sorted out with proper labour contracts.

[pcrh]

>its goal should be to bring a novel contribution to the state of the art in the field

I would qualify that statement by saying that the goal of a PhD should be to demonstrate the ability to bring novel contributions, etc.

It is impossible to predict the outcome of a truly innovative project, which ideally the PhD candidate is engaged in, so the failure of any such project shouldn't be a reason to deny graduation.

The candidate should be judged instead on their ability to do research correctly, i.e. with the goal of discovering facts, however inconvenient they may turn out to be, rather than be coerced into churning out yet more useless publications with overblown claims.

[eli_gottlieb]

>A PhD is not so much an apprenticeship though, its goal should be to bring a novel contribution to the state of the art in the field.

I think here we have a disagreement over goals. I've always thought of a PhD as being the period in which you prove you can compose and carry out a "Minimum Viable Sustained Research Project." But it's intentionally minimal: you're supposed to get it done and start your real career as a qualified scientist.