In recent years, growing numbers of UK universities have been offering postgraduate students a less conventional way of gaining a PhD. While the PhD by thesis remains the most likely path to a doctorate, one granted on the basis of published works is becoming increasingly popular, although no statistics are available.

The critics allege that the PhD by publication is a back-door route for those who are incapable of earning a higher degree in the conventional way. But while it is clearly not without its problems - like any degree pathway - it is perfectly sound, and standards are just as rigorous as for a PhD by thesis.

The PhD by publication allows the applicant to submit a number of published papers instead of an original bound thesis or dissertation. Of course, not just any papers will do. Although regulations differ from institution to institution, candidates must demonstrate evidence of an original and distinct contribution to knowledge, within the discipline, in peer-reviewed journals. They are expected to present a portfolio of published research papers - between five and seven seems to be average - and these must adhere to a coherent interrelated theme. The portfolio must also be submitted with a supporting statement that, typically, critically appraises the submitted works, explains how they fit together, accounts for and supports their chosen methodologies, and highlights the significant and original contribution to scholarship. The publications can be papers, chapters, monographs, technical reports or scholarly editions of a text.

The role of the supervisor or adviser, who will usually be expected to have a substantial record of publication, is to provide support and guidance to the student in writing the supporting statement. The supervisor guides the candidate in the choice of publications to be submitted, indicates whether the existing works are good enough - both in quality and quantity - to form a viable submission, and confirms that they represent a coherent theme.

³ÉÈËVRÊÓƵ

Once the submission has been put together, the candidate is invited to present it to an academic committee. When the committee is satisfied that all the conditions have been met, the student is recommended to undertake an oral examination, just as conventional PhD students are subjected to a viva.

The majority of the objections to the PhD by publication concern technical problems. For example, the route is available only to specific candidates. Usually a university will accept only its own graduates; even more commonly, it will accept only its own staff. Even here, further restrictions may be made. The University of Surrey, for example, takes on only full-time members of staff; in contrast, Roehampton University is a rare case of an institution offering its doctorate by published works to most graduates (subject, of course, to other criteria).

³ÉÈËVRÊÓƵ

As a result, candidates hoping to complete their PhD by publication at their alma mater might have difficulty finding a tutor. They may be making their application several years after they completed their last degree. Some universities even require that applicants wait a number of years following graduation, before they enrol. However, by then, the professors they were familiar with may have moved on or retired, so it will not always be possible to draw on the same support as was offered at the time of graduation. This admittedly also applies to the PhD by thesis, but the great majority of students trying to follow the published PhD route cannot apply elsewhere.

Take Stephen Armstrong, a Canadian expert in change management who gained a BSc (hons) degree in mechanical engineering at a London university in 1980. He went on to found management consultancy AMGI, where he used his engineering expertise to improve the operations of blue-chip firms around the world. He has published widely in professional journals and authored or co-authored two books published by Cambridge University Press. He has recently taught two master's courses at the University of Toronto on managing engineering innovation, and wants to use his extensive writings to obtain his PhD.

But when he asked about a doctorate by publication at his London alma mater, he was told by an administrator that the institution "has changed significantly since you graduated. We no longer have a mechanical engineering department nor do we run courses related to what they were doing."

There was almost no chance of finding an appropriate tutor, the university added, although Armstrong is still looking.

Another problem is that regulations can differ considerably between universities, particularly for the supporting statement. While the norm seems to be about 12,000 words, it can range from 2,000 to 25,000 words. At the University of Edinburgh, the statement must be between 10,000 and 25,000 words, putting it at the top end of the scale; the University of Leeds stipulates about 5,500.

Often the regulations are ambiguous (see box below) even though in the past few years, universities seem to be tightening up. In earlier years, it was apparent that the supporting statement was, in some cases, no more than a glorified application form, whereas today universities are aware that what is needed is a document suitable to be bound and kept in university and copyright libraries, presenting a lasting record and containing a fundamental analysis of the research question under review, to be communicated to other academics throughout the world.

The main problem, though, is that producing six or seven papers worthy of publication is highly demanding - as tough as completing one thesis, if not more so. There can be time delays if a journal article must be extensively revised, or if a paper is rejected by the editorial board. The time period between starting the research and seeing the finished product in print can be lengthy, and it takes a good deal of stamina to see the process through to completion.

In a sense, though, this is also the main argument for PhDs by publication. It is emphatically not an easier path than the PhD by thesis. As anyone who has ever submitted a paper - let alone six or seven - for publication in a peer-reviewed journal will know, producing the papers takes a great deal of determination, time and effort. Crucially, the same quality assurance applies to PhDs produced in this way as to PhDs by thesis.

³ÉÈËVRÊÓƵ

As the papers used for PhD by publication are peer reviewed, examiners would be hard-pressed to reject an article on the basis that it does not meet the standard for a PhD. This is also true of academic books, which are often reviewed by other academic experts before being accepted by a publisher for publication. In fact, the process can actually help to sharpen the PhD; although considerable delays can occur when a paper is being reviewed, the referees' suggestions for improvement can strengthen the presentation and argument of the paper's thesis.

As a spokeman at the University of Exeter says, candidates have to ensure that their publications are of "doctoral standard".

Detractors might claim that the PhD by publication route is beset with difficulties that are hard to overcome; however, it can be argued that this shows that it is not an easy route for those unqualified to achieve a "proper" doctorate.

Indeed, the PhD by publication has one particular advantage over its more traditional rival, one tied up with the rationale for universities introducing it. It was intended to help, and its main beneficiaries appear to be, professionals who take up lectureships and research positions later in life, who may have produced scholarly publications at doctoral level in other occupations, but whose lack of a final qualification puts them at a disadvantage in academia. This is particularly relevant in vocational areas such as law, nursing and accountancy, where professional experience is especially valued.

We need to give credit and support to those who embark on academic careers following a productive life in another profession, and this is a good way of recognising their true educational worth after they have amassed many publications to their name. Not only is it good for them personally, it also helps universities, who can expand the ranks of those who are qualified to supervise doctoral students, and make sure that more senior staff members have PhDs.

Mark Doel, emeritus professor of social work at Sheffield Hallam University, spent 20 years as a practising social worker and published widely in scholarly journals, although he never produced a PhD thesis. The PhD by publication route enabled him to enter a scholarly life later in his career. Jennifer Law, principal lecturer in public sector management at the University of Glamorgan in Pontypridd, argues that a PhD by publication is especially useful if, like her, "you have not entered academia via a traditional route with a PhD already completed".

"You can focus on building up publications earlier than you might if you were doing a traditional PhD," she says. "You can produce publications - and in doing so contribute to the research excellence framework - but also gain your PhD."

But the process is not easy, she emphasises. "Although you use your existing publications as part of the PhD, you need to have a number of good-quality publications to submit, and after this there is still a considerable amount of work to be done to gain the PhD. I spent about a year developing the overview that I needed to submit with my publications."

At the moment, the US and Canada do not offer PhDs by publication. Yet there is growing support for the pathway in Australia, where the demands of the universities can be greater than in the UK - for example, in terms of the number of published articles submitted.

Before securing her doctorate by publication at the Fenner School of Environment and Society at the Australian National University, Lisa Robins submitted 11 peer-reviewed published articles. The PhD gave her an edge over competitors in seeking full-time postdoctoral employment, she says, supported by the fact that she had a substantial body of published works as well as the postgraduate qualification.

So while the traditional route to a PhD remains dominant and vital, the advantages of the PhD by publication should not be dismissed. The majority of those who follow this route may have been unable to attend a university to conduct postgraduate study, but they have managed to do the necessary research in other ways, for example, as part of their employment outside academia. And the standards are as rigorous as for a conventional doctorate. A PhD by publication is the equivalent of any other PhD.

³ÉÈËVRÊÓƵ

Doctorates by publication: consensus and guidance would benefit candidates, advisers and examiners

Apart from technicalities relating to process and claiming the relevant fee and expenses, most experienced PhD examiners probably pay little attention to the paperwork that arrives from the academic registry or graduate school.

After all, regulations provide a framework, but the way PhDs are assessed practically is secured by the academic conventions, vague though they might be, that put flesh on the skeleton. New PhD examiners might look at the regulations quite carefully, but you will soon find them taking soundings from more experienced colleagues about how to actually go about their task.

However, my experience at several different universities suggests that even experienced PhD examiners find themselves less than sure-footed when it comes to a PhD by publication.

Partly this is caused by a lack of individual and collective experience, but another factor is the diversity in universities' regulations and a lack of specificity.

The latter is understandable, given the nature of regulations, but perhaps greater congruity and detail would enable the sector to grow in confidence in dealing with PhDs by publication.

This would no doubt help candidates - and their advisers, too, who often struggle to pin down exactly what is required.

Perhaps the greatest area of uncertainty for candidates, supervisors and examiners alike is how much is required.

In particular, if the candidate is relying on refereed journal papers, how many should be submitted? I have been asked this many times, but it really is a "length of a piece of string" question.

An answer is unlikely to appear in a set of regulations. After all, it depends on the quality and scope of the individual papers and the degree of overlap between them, among other things.

There are a couple of areas where some regulations could be more helpful though.

First, what is the nature of the accompanying commentary, and what is it supposed to achieve?

For example, given that the papers will contain methodological justification and reviews of literature, how much detail is required in the commentary?

My preference is for an overview of the methodological approach and the key points of contact with the literature, but I have seen some cases where examiners seem to want much more, all within the constraint of just a few thousand words.

With some convergence on the appropriate role of the commentary, it would make sense for the sector to settle on an appropriate maximum or guideline length. Then all parties will more easily become accustomed to what such a commentary is expected to look like.

Second, how should jointly authored papers be treated?

At one level, this is an issue familiar from the PhD by thesis. It is not unusual for a candidate, especially in the sciences, to bind with the thesis (or otherwise refer to) papers co-authored with a supervisor or with other members of a research team.

It is then for the examiners to reassure themselves about the individual's contribution to the body of work.

However, in a PhD by publication, matters can be more complicated. There may be several more papers referred to and, more significantly, they might involve different authorial teams.

It would therefore be helpful for regulations to require a statement from the candidate of their contribution to such publications and, if possible, to have that statement signed off by the co-authors.

This is by no means universal practice, but it would be a great help to examiners.

Christopher Cowton is dean of the University of Huddersfield Business School.

³ÉÈËVRÊÓƵ