Graeme Harper says postgraduates need good supervisors, but they also need more recognition from universities and the Government

What makes a good supervisor? Do you want a no-holds-barred answer? OK then, just between you and me, here's my response: the good supervisor is passionate, unflinching, wary of purely instrumental value, mostly individual but partly institutional. The good supervisor is unable to be moved from the paramount human position he or she holds dear: knowledge matters, postgraduate students are committed to new knowledge, every manner of political, economic and social change will affect universities, but postgraduate study must continue, will continue, must flourish, and must be supported, above all else.

Speaking in confidence, that is what I would say. But on the record, I'm cautiously enigmatic, offhandedly quoting the tagline from the 2006 film The Good Shepherd: "Who is the good shepherd? The one who looks after his family or the one who looks after his country?" Postgraduate students are part of our particular subject "families", and good supervisors have a key role to play in shepherding family members forward through the sometimes rugged country of the contemporary academy. Subjects cannot expect to travel anywhere without postgraduates and those who support them: good supervisors keep subjects going forward so that new discoveries can be made.

Isn't the really atrocious supervisor just the stuff of university legend, an academic bogeyman designed to generate graduate devotion to the cause of independence, commitment to self-management and adherence to the concept of researcher self-reliance? The alarmingly detached supervisor with more kudos than a Nobel prizewinners' office party but about as much engagement with postgraduates as the office hatstand - isn't this awful spook merely academic hoo-hah?

É«ºÐÖ±²¥

Unfortunately not. Supervisors sometimes get it horribly wrong. And yet, read around and you'll soon discover articles on just about every supervisory practice: choosing a good supervisor, monitoring progress, responsibilities, the supervisor as guide, the supervisor as leader, the supervisor as manager. The impression you get is that there's so much guidance out there that even the least-engaged academic couldn't help but be half good at it. Perhaps, however, it is simply that we far too easily give the general impression of good supervision. We write about it in a professional manner, create policies to support it, even monitor it to the point of obsession, and in doing so we generate a great deal of academic paraphernalia associated with well-considered supervisory processes, even if we don't truly offer good supervision. Might it even be that we don't actually have the right word for what we're trying to accomplish?

The word "supervise" means something akin to watching over, overseeing. But speaking personally, that's certainly not what I'm looking to achieve. I imagine if I began to watch over my doctoral students they'd almost certainly run for the hills. More accurately, I don't think I'm supervising at all; rather, I'm encouraging, challenging, supporting, challenging further. My final objective is to help form fully fledged researchers who don't merely map their research on to what I have undertaken or may suggest - that approach is not a process of developing an individual or a subject, but mere indoctrination. I'm aiming for something far better than that - a higher kind of learning.

É«ºÐÖ±²¥

Of course, good supervision isn't a one-size-fits-all affair, and the variety of interactions, if I dare use that somewhat overused word, maps relatively well on to the various levels of study. The range of freedom extends as students progress from undergraduate to advanced postgraduate level, with freedom to fail as well as succeed. In supervising doctoral students, we don't relinquish our duty to advise and assist, but do need to give students a good sense of what makes a developed researcher. Some of that involves encouraging a candidate to take risks, to challenge established ideas or seemingly immovable subject paradigms, to be willing to be entrepreneurial and innovative in practices and critical assessments. If we return to the "subject family" notion, that's fairly difficult stuff for both student and supervisor - after all, as a supervisor you may be sending a family member out into some wild educational weather.

But that's the point. Doctoral study is really the last time that somebody is going to give you a big creative or intellectual nudge. Future colleagues may challenge you. They may even support you in taking a further, bolder step. But it is at the doctoral level that the good supervisor can judge how far candidates can take things, and encourage them to go as far as they're capable of going. The relationship works best this way; it's mutual, you walk the next step together.

Such an ideal does not fit well with the generic definition of "supervision"; the word "mentoring" seems more accurate. But even that suggests a one-way exchange, and the relationship I'm talking about is far from unidirectional. The word I'm looking for is perhaps more akin to "advocate" or "good counsel".

And here's the other thing: good supervisors are adventurers. I don't necessarily mean that they've climbed Mount Everest or flown a hang-glider around the world. But they do see new knowledge as something to explore and universities as places in which it should be explored.

Good supervisors are not people who are going to tell you all you need to know. They're going to suggest what you may like to find out, and they will want to come with you (certainly metaphorically if not literally). They're going to try to ensure that you both survive the journey.

Partly this is because they want someone else working in the field who is equally passionate, but also because they want the pleasure of seeing someone achieve. Even if they tell you their own period of doctoral candidature was hell, they'd do it all again anytime. Hell is no challenge for the good supervisor - it's merely a little hotter and less fluffy than ordinary academic life.

A key problem with defining the good supervisor is not simply that supervision is decidedly individual and complex, but also that defining it involves making some decisions about the nature of a university and therefore the nature of higher education. With this in mind, for my definition we'd need to agree the following points.

First, a university is a social space, and supervision (if that is the word we continue to use) is a social activity, involving at least two individuals and sometimes more. Social spaces are human habitats and can't be maintained or developed without social interaction. This means that universities are at least partly responsible for creating and supporting good supervision.

É«ºÐÖ±²¥

If universities open their doors only at certain hours and close them religiously at certain hours, thereby preventing a natural flow of action, contemplation and investigation, or if they severely limit the resources needed for critical assessment and creative exploration, then they fail to support good supervision and good supervisors. "Invest in good supervision" is the motto, and if you take it to heart, you will produce excellent postgraduates.

A university is a place devoted to a certain kind of learning - broadly speaking, "higher learning" - and that involves the desire for and the aim of extending knowledge, not just confirming or applying it. If universities engage slavishly in the pursuit of immediate results, short-term achievements or research tailored to modest expectations, the possibilities for good supervision will be limited.

A university is not a place you enter at a certain age and leave after a certain period, never to reference again. My definition of a good supervisor depends on a lifelong exchange - even if not directly or continuously. That exchange relates to the expansion of a love of knowledge. A university is not merely a collection of buildings and courses fixed in time and place. If universities fail to send out this important message, they also fail to support good supervision.

Finally, a university is a place of creativity, invention and pioneering. The risk-averse university is, therefore, a difficult place in which to ensure good supervision. The university that fails to provide intellectual and creative leadership is even more of a problem for good supervision and good supervisors. Likewise, the university that doesn't engage in international as well as national, regional and local knowledge exchange has a negative effect on good supervision.

If my definition of the good supervisor is going to hold water, we would need to agree these things. But, as may already be obvious, we also would have to agree that not every academic or indeed every university should be involved in supervision. I would not back away from defending that conclusion. But perhaps such a conclusion does not address the more specific question: can a good supervisor be found in a university that has not, as yet, provided the best habitat for supervision?

Good supervisors and good supervisory environments don't always go hand in hand, just as fabulous sporting facilities and the development of world-class sportspeople don't, but it helps. I'd thus like to see the Government commit more to world-class supervisory habitats in British universities.

I don't blame individual institutions for the lack of those habitats; I blame the Government for not sending out the right messages or providing the appropriate level of funding.

For far too long, we've treated postgraduate study as an addendum; an adjunct. We've directly doled out millions to increase the number of school leavers going to university, but then often leave graduates to fight for themselves.

We've put pressure on research councils to fund only the best graduates, and then reduced funding when financial times are tough. We've said an awful lot about encouraging overseas graduates to come to Britain for postgraduate study - and to pay British universities the fees that go along with this - but relatively little about the importance of postgraduate study in general to both Britain and the rest of the world. This isn't good enough.

I mentioned the film The Good Shepherd, starring Robert De Niro and Matt Damon. It's actually about the founding of the Central Intelligence Agency. What I'm talking about here is very far from clandestine, but it is, however, most certainly about intelligence, a kind of intelligence to which the good supervisor is devoted. This is the intelligence associated with capacity and comprehension, with clear perception and good judgment. It is not a great thing to supervise someone into matching expectations - either their own or their university's.

Good supervisors are capacity builders, stargazers, speculators. They suspect that there is more out there than immediately meets the eye and believe their supervisees can help find it.

Every now and then, someone will ask how to build a training programme to develop excellent postgraduates and brilliant supervisors. There's a personal temptation simply to say: "Tell them to believe in the wondrous and to have the courage to seek it out." But 21st-century universities deal in results as much as in ideals.

Therefore, without asking us to abandon that belief in the wondrous, the answers may be these: introduce opportunities where there is unequivocal support for freedom of thought and action - sometimes this will mean physical spaces and places in which to explore openly; if your university talks mostly to itself, open portals to other places, schedule connections with the world, connections that can be physical or virtual. This will strengthen the work of both supervisors and supervisees and build capacity in supervision - the more people there are supervising, the more who will become good at it.

É«ºÐÖ±²¥

We should also encourage greater discussion of projects and progress; it's not just economies of learning scale, it's also the establishment of communities of thought and practice. We should reward good supervisors by ensuring that local workload models acknowledge the importance of their efforts and resist simplistic equations (eg, "completions are good, in and of themselves"; "postgraduates can be marshalled into a responsive group in much the same way as undergraduates").

If we support the family, the success of the country will follow; place the importance of graduate subject provision first, and a successful university graduate culture will form. We must also lobby the Government for more support for postgraduates and postgraduate habitats in Britain - this should be a national priority.

I would be the last one to suggest that I am a good supervisor, although I would like to believe I am. I care a great deal about my students, their projects, and their current and future successes. And I fear that a lot of what has been said lately about postgraduate study in Britain doesn't engage with the core values needed for good supervision: the desire to increase understanding and truly extend knowledge.

Mostly, however, I fear that we think far too little about the human interaction that good supervision involves and don't give enough support to universities to create the habitats and policies that support good supervision and good postgraduate experiences. Postgraduates are the future of higher learning in all our subjects. Perhaps it is time we attended more concertedly to that fact.

Graeme Harper is professor of creative writing at Bangor University, and director of research for Bangor's College of Arts and Humanities.

É«ºÐÖ±²¥ VIEW FROM INSIDE É«ºÐÖ±²¥ FLOCK

What makes a good PhD supervisor? There have been many attempts, both serious and comic, to classify the role.

I have apparently been allocated "Professor Never There", an academic whose carbon footprint is comparable to that of some small nations. Yet despite her intimidating timetable, she affords me a significant amount of her time, and were it not for all her tales from the road and the near-constant presence of a moderately sized wheelie suitcase by her side, I might not know she was ever away.

Although the classifications of supervisor are typically constructed in jest - indeed, "Professor Never There" appears on one such list, alongside the likes of "Dr Never Satisfied", "Professor Different Planet", "Dr Overbearing Interferer" and "Dr Slave Labour No Research" - there appears to be at least a grain of truth in them.

I believe it is supervisors' awareness of how their own particular situation or modus operandi can affect students that allows them to transcend these stereotypes and tailor their supervision accordingly.

As the trend moves towards prospective graduate students applying to a department or research group rather than to a particular academic, students are often subjected to a kind of supervisory Russian roulette, where a supervisor is assigned with little or no input from the student. As this allocation is the first event on the critical path to obtaining a doctorate, it is important to understand what makes a good supervisor.

There is arguably no standard yardstick against which all PhD supervisors can be measured; what makes a good supervisor is subjective and somewhat dependent on the supervisee. I do believe, though, that there are a number of traits that can be universally accepted as desirable.

It is also important to recognise that the supervisor is only one half of the student-supervisor relationship and that the ultimate success or failure of a PhD can depend just as much on the student's own strengths and weaknesses.

From the beginning, a good supervisor will be open and honest, and will treat the initial meeting with the student as one of the most important events of the whole PhD process, second perhaps only to the viva. It is the supervisor's opportunity to understand the student's expectations, to lay out what is expected of them and to dispel (or confirm) any myths about the challenge about to be undertaken. Students are looking for confirmation, both at the start and throughout the process, that their expectations are not only ambitious but also realistic.

Possibly the most valuable gift a supervisor can bestow on a student is regular, honest and detailed feedback. The only thing worse than being told you are heading down the wrong path is being unaware of this fact until it is potentially too late.

However, feedback should be carefully weighed and should guide, but not lead, students, offering them the freedom to make independent decisions. Moreover, it should allow students to follow aspects of their research that interest them, encouraging them to become autonomous academics rather than mere extensions of their supervisors.

A good supervisor will also enjoy the pedagogic aspect of supervision and exude an infectious enthusiasm. At times when a student hits a wall or begins to flag, an animated and inspiring conversation with a passionate supervisor can be the perfect remedy to revive morale, if only until the next crisis.

Ideally, there will be some conversations between the student and the supervisor that are far removed from the topic of the student's thesis, and these are just as important.

While discussions about the weather won't produce outcomes that feed directly into a write-up (unless the chosen field happens to be meteorology), a supervisor who is human, approachable and willing to build a friendship can make the process feel a lot less daunting for both parties.

Great supervisors will keep doctoral students in the back of their minds at all times.

It's rewarding and motivating to know that, while "out in the wild" conducting "real" research, a supervisor has kept their student in mind, and at the appropriate moment during a discussion with peers interjected with a short promotion of them and their work.

These mentions, however brief, can have a significant impact, initiating collaborations and opening up opportunities that may otherwise be unavailable to young academics.

While a good supervisor can guide a student through a successful PhD, it is a great supervisor who prepares them for a successful academic career.

Matthew Gamble is a doctoral candidate in the School of Computer Science, University of Manchester.

É«ºÐÖ±²¥