[rajin444]

> It's fundamentally wrong for one human (the PhD supervisor) to have so much unchecked power over another

Why is this fundamentally wrong? Short of a truly equitable society (which I think most agree would be a dystopian nightmare), I don't see this as fixable. The problem is that the power is abused.

The best way to fix this is high trust societies, not creating one more institution with a disproportionate amount of power.

[logifail]

> Why is this fundamentally wrong?

It's the unchecked nature of the power.

Politicians have upcoming elections to worry about. Tenured professors ... what do they worry about?

[ModernMech]

> Tenured professors ... what do they worry about?

Class assignments and number of classes taught per semester, teaching times and locations, class sizes, advising assignments, committee assignments, TA allocations, grader allocations, internal grants, withholding or revoking consulting permission, withholding internal department funding, lack of promotion (there is still a ladder after tenure), and most of all reputation, which in academia has higher currency than currency.

Tenured professors still have a job to do and they still have a boss (lots of bosses, as I said), and if they don't do their job they can still get fired (I've seen it happen), or their lives can be made miserable.

If there's an issue of academic dishonesty then there there are processes at journals, conferences, and funding agencies to address concerns. I've also witnessed these processes work as intended.

[logifail]

> Tenured professors still have a job to do and they still have a boss (lots of bosses, as I said)

We are apparently talking about very different kinds of professors at very different kinds of institutions.

Unless he were to actually commit a crime, I don't think there is anything my ex-boss could have possibly done which could have ended up with him being removed from his post. He was simply too important to the institution in terms of his reputation and his ability to attract external funding.

How he treated his students was really neither here nor there, as long as the money and the citations kept rolling in.

[ModernMech]

It sounds like you got a raw deal by a powerful asshole, and are now painting with a broad brush. Not every professor has so much power as your former boss, so saying that the problem is that professors have too much unchecked power, and the fix would be to bring more oversight, is missing the mark. Because what happens when the person at the top of the new oversight pyramid is an asshole? You've got the same problem, and the solution wouldn't be to again make sure he has more oversight.

> He was simply too important to the institution in terms of his reputation and his ability to attract external funding.

This reinforces my idea that you should always go with an advisor who doesn't have this kind of clout. Not that you should have known this ahead of time or you did anything wrong, but I have heard sad stories from large labs enough to be thankful for avoiding them in the past, and it's the advice I will pass on in the future.

I think if you wanted to get at the actual problem and a solution to it, I would say the problem is sad narcissistic jerks aiming to become kings of their research kingdoms, and then they act as tyrants when they get a hold of large sums of grant money. That may be the experience some people have had.

A solution to this that you may like if you want more oversight may be to treat a research lab more like a corporation with a board of directors (I can hear the researchers recoiling) or some other governing body over the lab and the students itself. What do you think of that idea?

[barry-cotter]

> A solution to this that you may like if you want more oversight may be to treat a research lab more like a corporation with a board of directors (I can hear the researchers recoiling) or some other governing body over the lab and the students itself. What do you think of that idea?

Rule by committee. So many examples of that working out great.

How about keeping the apprenticeship aspect just having more masters? One year of just comprehend exams, one year where you’re supposed to devote half your time to research with an advisor and half to comps, and three years where you work on six different projects with different advisors in six month chunks? Some of those will be written up as failures, but basically you end up with 3+ Master’s theses. That’s a great deal more love research academia than legal academia and they don’t seem to lack for scholarly work or impact.

[Fomite]

This would require revisiting a lot of things - it basically assumes that all projects can be broken down into one-year, coherent and publishable chunks. I love Masters students, but they are often not the most productive, as they rarely get to wade really deeply into a project, and almost never get to develop a project of their own.

[ModernMech]

I like your idea a lot, but I don't think it could replace the Ph.D. proper. You really do need all that time on one project to make a dent in the state of the art sometimes. One year I spent an entire summer generating data for a single figure in my dissertation. That's part of the reason why it took me 8 years to graduate.

[Fomite]

Tenured professor here:

- Promotion. Tenure is awesome, yes, but there are two full academic ranks at my institution above associate professor with tenure, and if I want them, I need to continue to do the same things that got me tenure. Graduate students, get grants, publish papers, etc.

There's also further ambitions like chair, dean, etc. that are often literally elections.

- Internal allocation of resources. Will one of my students be my department's choice for a fellowship? If there's a space crunch (there's always a space crunch) can I get the space I want? If there's internal funds to help support pilot projects or the like, am I on the short list for those?

- Reputation. This is important to me. I want it to be known that "my lab" turns out good people. And that we do good work. This is not only actually a criteria for promotion, but also something I want (I didn't go into academia for the money). It's what gets you invited to things I want to be invited to.

This also includes things like whisper networks. Students warn other students.

- Grants. A number of grants look at things like mentorship and students.

[watwut]

If there is nothing to stop the abuse, abuse escalates and gets normalized. This happens any time there is large system with unchecked large power.

That is why it is fundamentally wrong. Because it ensures widespread abuse.

[patrick451]

You can't have power exist and expect it to not get abused.

But this problem is totally fixable. We need to reduce the role that the advisor plays. Grad students should be able to switch labs and universities like most people switch jobs. Once they publish some set number of papers, they get a PhD. No thesis, no committee. The advisor basically is demoted to a fungible manager.

[ModernMech]

> Grad students should be able to switch labs and universities like most people switch jobs.

Maybe this fixes the problem of abusive professors wielding their power in wrong ways, but it creates new problems as well. For instance, it shifts the very nature of a Ph.D. from a deep exploration of a single topic to shallow explorations of many topics. If you tell me I need 4 papers to get a Ph.D., I'm going to get 4 papers from the easiest venues on the most shallow of topics, because why go any deeper?

And are these solo author papers? What if you were 1 author out of 30 on one paper, and 1 author out of 2 on another? Are each of those 1 paper credit?

Do paper credits transfer between universities? Do some schools require more papers than others? Who gets to decide? Do some disciplines require more papers than others? Again who gets to decide? What about journal papers versus conference papers versus workshop papers? Do presentations count? Do poster sessions count?

If you are in my lab and I've spent time and money training you, what's my recourse if you just quit on me to join another lab?

[patrick451]

> If you are in my lab and I've spent time and money training you, what's my recourse if you just quit on me to join another lab?

None. Every employer the entire world over deals with this. The fact that this makes you feel entitled to continued labor by "your" students is the entire problem.

[ModernMech]

Well that is going to fundamentally change the nature of the whole arrangement. A Ph.D. is not a work-for-hire kind of arrangement, it's more like an apprenticeship. If you want it to be work-for-hire then some other changes will need to be made.

When I hire a grad student I do so with the intention that I will keep them on for the duration of their education, which is a commitment of at least 4 years, sometimes up to 10 (do any employers have 10 year plans with you specifically in mind? No, as a 22 year old fresh out of college, you are fungible to them.) The bonds that are created between advisor and student can be one of the longest lasting and most important professional bonds a researcher makes in their whole life. It's so important that academic genealogy is actually tracked and recorded. There's a concept of a "grand advisor" or your advisor's Ph.D. advisor. So for all the negative stories you hear about a narcissistic asshole advisor wielding power and abusing underlings, there are far more stories of amazing intellectual relationships that can take a decade to fully bloom.

Your proposal would serve to virtually eliminate this relationship. The employee/employer relationship looks nothing like this. It's transactional and makes employees seem more like mercenaries than team members, no matter how much corporate culture wants to convince you that you're part of the team and a member of the family. You can actually find that environment during a Ph.D.; I know I did. You can make mistakes (sometimes huge mistakes) and not be fired. You can admit you don't know something or that you lack some skill and that's okay -- it's expected that you are still learning. Professors are happy to spend the time to teach such skills because they know they will be used on their own project, so it's worth while.

Here is what your proposal would do to my hiring practices and the way I run a lab: If you want to be an employee with the freedom to quit at anytime and move from lab to lab, I will conversely be quicker to fire you and to hire other better students from other labs to take your place. If you are going slow on a paper and you miss a deadline, I'm not going to shrug it off and say "Oh well, on to the next". I will probably fire you and hire someone else who won't miss deadlines.

If you make a huge mistake, like frying a $30k piece of equipment (true story), you'll be fired instead of met with understanding that you're just a student. Because that's what happens when you shift more onto the employee spectrum.

It's definitely also going to tighten the market for Ph.D. students significantly. This could be a good thing depending on your perspective, it would give you less competition. But it would also give new students fewer opportunities. I would think long and hard before hiring a grad student as opposed to research staff. If you make Ph.D. students too expensive, you'll just price them out of the market. Part of the deal of being a Ph.D. student is there's a general recognition that they will be working on projects they are not really qualified to work on yet.

As a 22 year old undergrad, you don't know enough about the domain to be useful. I need to invest years into you before you're even going to be publishing papers; if you can just leave me after that training to work on another project, I see no reason to invest that time in you in the first place. Instead, I'll just hire staff researchers who know what they're doing. Yeah they're more expensive monetarily than a grad student, but they also don't take up all of my time and make stupid mistakes that burn out expensive and one-of-a-kind electronics.

All in all, I believe your proposal is really throwing out the Ph.D. babies with the asshole professor bathwater. Honestly I joined academia to escape corporate hell, so as far as I'm concerned I'm going to keep treating my students as students rather than employees.