How Do You Know How to Write a Thesis?

I wrote my PhD thesis so long ago it was typed for me by someone else based on my handwritten chapters; the diagrams were drawn laboriously and messily by me with a pen, ink and a fair number of smudges; and the micrographs were produced for me by the group photographer. I have absolutely no recollection of being told what the correct form for a thesis was; indeed I don’t even remember my supervisor reading it, but suspect that there my memory is at fault. Surely he didn’t just let me get on with it! These days most universities will run courses on it, Vitae has useful web pages on the topic and there are a fair number of books to help. Nevertheless, even with a supportive supervisor it is still a daunting task for students, made worse by far fewer opportunities to write essays, reports or even extended exam- question answers than in the past. Students don’t necessarily get opportunities to write prose. Picking the brains of previous students, reading their theses and learning by example all ought to be used to supplement self-help and formal supervisory advice as thesis writing gets under way, although it should also be borne in mind there is no truly ‘right’ answer, no single way to do it (not even, in general, a single structure that would suffice) although some ways are definitely better than others. These thoughts are in part prompted by reading John Ziman’s illuminating decade-old text Real Science in which he refers to

newcomers to research…. entering a self-perpetuating ‘tribe’, where their behaviour is governed by many unspoken rules.

This is true of much of life, of course, and one frequently gets it wrong in ways that matter to a greater or lesser degree. At least these days, faux pas are unlikely to lead to the sort of social ostracism the rigid social mores in the past about dress or speech used to provoke. Nevertheless, as with all these things, the scope for making a fool of oneself is ever-present, with the potential for consequent damage to one’s self esteem, even if a lot of red ink from your supervisor is not in itself life- or even career-threatening. However, before a chapter – let alone a whole thesis – reaches the supervisor, there are things that can and should be done to reduce the risks of that flood of ink being stimulated, such as those I mention above.

As both a supervisor, and an examiner for my own and other universities, I have read a fair few theses in my time. Although upon occasion the results are fascinating, it has to be said reading theses is rarely entertaining, often for reasons that are not the fault of the student. The form of the thesis, starting off with a lengthy literature review and meticulous description of experimental techniques, is frequently and necessarily very boring. These should be written to be comprehensible to the next generation of students who will be following up on the work described. That means the methodology is frequently described in exhaustive (for which read exhausting) if appropriate detail and the literature review can describe papers which are only too familiar, not least because the examiner may personally have written a fair proportion of them. In my experience the examiner may have a keen desire to skip a lot of it.

However, that would be unwise because our job is to see that the student understands what is going on; sometimes they don’t, or at least they don’t appear to from the text. This is the sort of place where the unwritten rules of what should be included often come to the fore – and it should be recognized, reinforcing Ziman’s use of the word ‘tribe’, that different disciplines have subtly different rules about some of the following, which can also confuse the examiner as well as the student. Students may cut and paste diagrams from other texts (such as a previous student’s thesis from the same group) without due reference, or with an inadequate caption; perhaps the figures lack annotations, which would make things so much clearer, or are reduced so small it is hard to see without a magnifying glass; sometimes the figure isn’t even referred to in the text at all. I have seen examples of all such as an examiner, and in the results sections as well as the introductory chapters. I have seen cases where the whole style of presentation throughout the thesis – perhaps in terms of detail, or referencing – lacks something. An indadequate bibliography is often a cause for examiners’ complaints (missing key references, not up to date, citing reviews solely rather than original papers would be examples of familiar complaints).  I would have expected that osmosis from reading other student’s theses would have put across many of these fundamental ideas, but clearly not always. (I do believe it is the supervisor’s responsibility to pick such issues up before submission, but that also doesn’t always seem to happen for a variety of reasons, both good and bad). Nevertheless, it is worth bearing in mind that often it is during discussion of these introductory chapters that the student goes adrift in the viva. Writing a thesis about some pet technique without having a good grasp of fundamentals of how it works is prone to lead to difficulties; be warned.

Before students start getting very, very nervous upon reading this, I should state that it is really rather rare for examiners to fail students – even to fail a student a first time, let alone irrevocably at the reconvened viva. There was one unfortunate occasion when I turned up to be told I had to fail a student so that their supervisor could be taken to task. It was a great relief subsequently to be told by this same student it was the best thing that had ever happened to him as he was then able to proceed with his research with a different and more suitable supervisor. (This student is, by the by, a professor now). I have only completely failed a student once. The external and I felt terrible about it, agonised over whether we were doing the student justice, but the corrections that had been made between the first and second submissions were insubstantial and certainly did not come anywhere near to what we had indicated were needed. Furthermore, in the second viva the student seemed no more on top of the problem than first time around, so we felt we had no choice. But when we told the student the outcome he shrugged his shoulders and said he would have been surprised if we had done otherwise. Bizarrely it seemed to be less of a deal for him than for us. But by and large examiners want to pass students – honestly.

Writing a thesis can feel very intimidating. It should represent the culmination of years of hard work and sometimes it can be incredibly difficult to see the wood for the trees. Often it feels natural to write things up chronologically, in the order the work was carried out. Since research rarely actually proceeds in a linear fashion, this is often not the clearest way to put across the outcomes to an outsider, but stepping back sufficiently to see how best to present the results logically can be a challenge. But as the style of thesis-writing is, as Ziman implied more generally, determined by custom it is worthwhile trying to establish what those appropriate customs are before starting. Some students clearly pick these up unconsciously, others probably have to work much harder, and seek advice much more often in order to produce something that fits the bill. It is to a large extent a case of learning by example. Perhaps one of the most satisfying compliments a student of mine ever paid me in a thesis acknowledgement section (these are always interesting  to read, whether as examiner or supervisor) was to say I had taught them the ‘importance of caution and communication’. When I quizzed him about this he said that I had made him think much more clearly both about how to sift the evidence and then how to put it down clearly in writing. Still, supervisors, however willing, can only do so much. It is not our place to write that thesis nor would we be much good at it – it is the student who has done the work after all and has mastery over all the twists and turns of the research. The thesis should be a once-in-a-lifetime chance to show off that mastery.