What are PhDs for anyway?

A few weeks ago Nature published a piece, which asked “The world is producing more PhDs than ever before. Is it time to stop?”. A couple of weeks later the Royal Institution in London arranged a panel discussion about careers in science, a panel which included the UK science minister. Many of the problems discussed buzz around the problems junior researchers have in finding permanent jobs, and in how science is funded by providing short-term grants, which don’t give any job security.

Looking at these discussion, I’ve concluded that the root cause of the problem is the way the PhD is viewed and used by Society, and how this has changed over the decades.

The Academic’s View

Historically the PhD has been a research degree; it has been seen as a form of apprenticeship for people wanting a career a researcher in science. Someone with with a PhD is seen as qualified to carry out independent research. They could continue in a scientific career in a university, research institute or in industry. Supporting them through grants is thus a means to train the next generation of scientists.
On this view, the demand for PhDs places would be driven by the demand for trained scientists. Not every new doctor will end up a tenured professor – some will go into industry, others decide to leave science. But ideally all those who want to stay in research (and who are good enough) will be able to.

The man on the Street’s View

But I think this view ignores one big part of the tenured professor’s job: education. Universities professors are expected to teach; to profess. And this connection between education and the PhD is viewed as wider than just an apprenticeship. Society at large sees education as important (which is why we put money into schools, universities and colleges). In most countries the PhD is the highest degree, which gives it a level of importance as a status symbol, as showing that the holder is educated. The only other professions that get a title are medical doctors and people in religious posts, both also professions that are (generally) well respected. So, for the general public, the “use” of the PhD is a secondary one, to indicate status. I’m uneasy with this elitism – as someone with a PhD I’m not intrinsically more deserving of respect than anyone else who is highly trained – a nurse, an electrician or an accountant, for example.
This status is, I think, largely promulgated in universities because the teachers themselves have PhDs. This reinforces the view that the most educated people are those with PhDs, even though the degree is not strictly about learning stuff. This is a historical relic: firstly because the PhD used to be about scholarship (which is learning), and modern scientific research is very different to traditional scholarship. And secondly, because the home of a lot of scientific research is the university, where it grew out of scholarship. There is really no reason, in practice, why this should be so. There are many research institutes throughout the world (I have worked in several of them), so there is no necessity for researchers to also lecture to all levels of undergrad.

The Politician’s View

But the PhD is being undermined, not because of any deliberate act, but as a by-product of the way research is increasingly funded. Governments want to fund research, and a lot of that is done using soft money: paying for short projects that then need to be renewed with new short projects. The research is done through funding junior scientists, who are cheaper than more experienced researchers. In practice this means a lot of PhD students. A side benefit is that this can be claimed as “education” as well as research (and hence be paid for by the education department, which is a nuts way of paying for research). But the supply of scientists now becomes dependent on how much research the government wants to pay for, and that is a depends on the economy, and the political priorities of the party currently ruling. The connection to the demand for students is broken: there is over-supply for the traditional careers that hire PhDs.

When viewpoints collide

But we, as academics, are still selling the PhD under the old assumptions about what it’s for. We are telling students that they should do a PhD to become researchers, even though there may not be the jobs afterwards. And we are training them to be researchers, to carry out specialised tasks that might have nothing to do with the career they end up in (ironically, this training wasn’t been in teaching, even though universities were a major employer of PhDs). So, in order to get research done we are training up researchers, many of whom then have to leave the career they are trained for. This is just silly – these students aren’t being prepared for the workplace they’ll end up in, and become frustrated because they can’t find a job in the career they’ve pursued.

What to do?

It seems to me that there are two solutions:

  1. Keep the traditional meaning of the PhD, as a research apprenticeship, and limit the number of positions available. This is the solution I like, although creating a mechanism to do (i.e. to decide how many PhDs to give out) might not be easy. It also creates a second problem – how are you going to get the science done? This relates to the discussion at the RI last month, and Jenny Rohn (formally of this parish) has blogged about one solution which I like : creating a career structure for people who want to do science, but not be in charge. In essence, super-technicians. This used to exist (and may still do) in the civil service, where there were several grades of “scientific officer”. the problem is that the more senior SOs cost more, and need some long-term job security, so they are more expensive. On the other hand, they should also be more productive, as they have already gone through the process of learning how to do research (either as a more junior SO, or through getting a PhD).
  2. Change what the PhD is for. Rather than make it an apprenticeship for research, expand the scope. This would mean teaching more “transferable skills”, e.g. in communication. This raises a few problems. One is that the science may suffer (as the Nature article acknowledges may be happening here in Germany), as students spend their time learning other things. So, if the government wants to pay students as a way of increasing scientific productivity, this will be diminished. Another problem is deciding what the PhD should really be about. If it is not a research apprenticeship, what is it? What should we be teaching the students? I don’t know.

At some point one we, as a society (or rather as several societies), will have to chose which path to go down. My guess is that inaction from the top will mean that there will be little change in the funding of science, so that the purpose of the PhD will change, and with it the form of education: it will pick up a more general component, teaching skills that will be useful in the wider world. I actually think it’s good that there is a component of this sort of teaching, but where’s the point where we dilute the research to the level so that a PhD isn’t able to carry out independent research? How close do we want to get to that point?

About rpg

Scientist, poet, gadfly
This entry was posted in Science Blogging. Bookmark the permalink.

11 Responses to What are PhDs for anyway?

  1. Henry Gee says:

    I have a suspicion that the Ph.D. degree is as much a victim of grade inflation as any other qualification. When I was a lad, and even a student, some of the people who taught me didn’t have Ph.Ds.
    Even now, there are quite a few senior people at Your Favourite Weekly Professional Science Magazine Beginning With N who don’t have a Ph.D. – the former editor, John Maddox, didn’t have one. When I joined YFWPSMBWN I didn’t have a PhD either (I was writing up). These days, because the market is so competitive, to get a job here you pretty much have to have a Ph.D. and postdoc experience too.

  2. John Wilkins says:

     Interestingly I make a similar point in the Times Higher Education Magazine in two days…

  3. Bob O'Hara says:

    That’s true, Henry. I guess it’s the excess of supply that does this. Do/did your colleagues at YFWPSMBWN without PhD/post doc experience do a worse job because of that? IOW, does it matter that they have research experience? (I think we can except John maddox as being exceptional).
    Ha, John – great minds and all that. I’ll look forward to seeing what you wrote. I’m sure it’ll be as erudite as always.

  4. Henry Gee says:

    Do/did your colleagues at YFWPSMBWN without PhD/post doc experience do a worse job because of that?
    No. Research experience is always desirable so one can put oneself in an author’s shoes, and have some knowledge of the research discussed. But in the last analysis the job is about making decisions, which can be done by any person with a sense of judgment and some experience.

  5. Tom Webb says:

    Rather tangential to your post, Bob, but what I found rather depressing about the debate in Nature and on here (but was too busy at the time to stick my oar in) was that (almost) everyone seemed to be starting from the premise that a PhD was first, foremost, and (I got the impression) perhaps exclusively a form of training / apprenticeship. Now, of course, that’s important; but it’s also an education, a chance to spend a intensive period of time finding out stuff that nobody else has ever known. If you end up not using that expertise in your post-PhD profession, I’m sort of inclined to say, so what? You haven’t wasted your time, you’ve contributed to the sum of human knowledge, and have become a more rounded human being as a result. And while I take all your points about the govt. looking to fund research on the cheap, if this gives more people the chance to experience primary research, I think that’s a good thing. As long as we are honest from the start about the low probability of (and sacrifices implicit in) going on to an academic career.
    Or is that me being impossibly idealistic?!

  6. Mike Fowler says:

    A very fine summary, Bob. Have you any access to the changes in the number of PhD positions that have been offered in science (or biology, or mildew ecology) over the last 30-50 years? IIRC, NERC (in the UK, for the uninitiated) changed their approach to PhD funding as I was finishing mine in 2002 (fewer grants of ever so slightly greater financial value), which was thought to be a ‘good thing’. Time series analysis is well known to be the solution to all things mysterious…
    But even back then (which almost certainly wasn’t as back when Henry’s was), they were attempting to foist ‘transferrable’ skills on us. As a student who was dead keen on continuing with academic research following my PhD, I absolutely felt this time taken away from my investigations of the real, in silico world, was not time well spent. As has been mentioned elsewhere, it’s probably worthwhile making some of these tranzferrable skillz optional, at the students’ discretion, innit? They are, after all, adults (by most other reckoning) by the time they start a PhD.

  7. Bob O'Hara says:

    Tom – yes, that’s an interesting point. I suppose one could argue that becoming a more rounded person (must be all the cake) means giving transferrable skills.
    Your comment implies that there is some intrinsic value to finding out something new, but that’s obviously a value judgement (is it really worth discovering useless knowledge for the sake of it?). I’m not sure how far to go with this – it would need another blog post…
    Mike – I don’t have the data to hand, but it must be possible to find it. I’m not sure what you mean about PhD students being adults, unless you’re going to claim that you’re the exception.

  8. Mike Fowler says:

    Bob – I was merely suggesting that by the time anyone starts a PhD – which in the UK could be as early as 21 years old -they are old enough to make certain decisions for themselves.
    For example, having to attend a compulsory, 3 day "Introduction to Statistics" course, covering such statistically challenging topics as ‘What is correlation’ and even ‘What is a p-value’ (in the numerical, rather than philosophical sense), was not the most profitable use of time for someone who had calculated 2-way ANOVAs by hand some years earlier and was more interested in learning about Principle Component Analyses. (Bitter? Moi?)
    Making attendance at the ‘transferrable skills’ courses optional seems like a reasonable approach. Students can discuss with their supervisors what would be a good way for them to gain useful knowledge, depending on their specific career goals.
    And yes, I are the exception, except in this case.

  9. Tom Webb says:

    Mike – sounds like you would have benefitted from Sheffield’s patented ‘Doctoral Development Programme’ – which we’ve recently introduced (amidst much confusion) to replace our old, more formal, training programme for PhD students. Now the dust is settling, it looks like being a good way of doing things – students and supervisors sit down together to discuss training needs, and how to go about achieving them. So hopefully we avoid sending prodigies on basic stats courses!
    Bob – isn’t there a saying that it’s worth re-inventing the wheel from time to time, if along the way you learn something useful about wheels? By which I mean, generating useless knowledge can still have benefits, if it switches you on to the whole process of how we generate knowledge. And who’s to say what’s useless? (On reflection, yes, this is quite a large issue!)

  10. Mike Fowler says:

    Tom – definitely sounds like an improvement. And I in no way meant to suggest I was a prodigy. Everyone who went through an undergradtuate Zoology, or Psychology degree at Glasgow Uni (at the tail end of last century, at least) already had access to a very good (frequentist) statistical background.
    And until I started my PhD in 1999, the ‘Zoology’ department (aka Division of Evolutionary & Environmental Biology by then) had offered a very good advanced stats course for its PhD students (covering PCAs et al.). Due to faculty restructuring though, they decided to scrap that, and send all Life Sciences PhD students to the compulsory "Stats for Dummies" course. It may have been useful for microbiologists, or medical bioscientists, but apparently not for anyone with an organismal biology background.
    However, I would support any innovation that allowed for flexibility in how students can choose to develop their careers – good on Sheffield – especially in the face of the apparently constant shifting faculty structure that some Uni administrators seem to delight in.

  11. Tom Webb says:

    Mike – I didn’t mean to suggest that you had in any way meant to suggest that you were a prodigy, my tongue was firmly in cheek! (Er, not that I want to suggest that you weren’t, either, mind… I’ll stop digging, now!)