Writing Up Your PhD

For PhD students, we reproduce an article by Dr Andrew Colman, published in 1993 in the BPS Postgraduate Section’s Newsletter. 


There is an apocryphal story about Stanley Milgram, the American social psychologist who died recently and who is remembered chiefly for his "shocking" experiments on obedience to authority, which are really examples of anti-social psychology. Milgram was a very clever man, but he was not at all practically minded. He used to play a little tennis now and again, and it is said that he managed to convince himself that he could beat John McEnroe (the strongest player in the world at the time) by winning just one point against him. He worked out that it didn't matter if he lost the first few games, or for that matter the first few sets, because players often win by coming from behind. All he had to do was to make sure he won the last game of the last set; and in fact it was quite unnecessary to win every point in the last game of the last set as long as he won the last point of the last game of the last set. This logic led him to the conclusion that, irrespective of what went before, he could beat John McEnroe by winning one single point - the last.

Doing research is a bit like that: if the final write-up is a winner, it doesn't matter what went before, although the catch is that you cannot make a silk purse out of a sow's ear. I am trying to make a serious point here. Research that is never written up might just as well never have been done in the first place. If it is not written up, there is no record of its findings, and without any record of the findings the research cannot make any enduring contribution to knowledge. Ph.D. or M.Phil. research that is never written up does not even achieve the much more limited goal of giving people letters to put behind their names. And yet it is a fact of life that lots of perfectly respectable research is never written up, and that includes lots of perfectly respectable Ph.D. and M.Phil. research. Psychologists have always paid much less attention to writing than to reading, in spite of the fact that writing is at least as important because if no one wrote anything there would be nothing to read (Hartley, 1980). Problems of writing are, in a sense, the most basic problems facing postgraduate students and other researchers; if they are not overcome, everything that might have been achieved in the research is wasted. That is one reason why I am interested in the psychology of writing. The other reason is that I know the problems of writing personally from long and bitter acquaintance. I was a journalist for a short while before taking not one but two Bachelor's degrees, both of which required me to submit quite big research projects, followed by a Master's degree by thesis and a Ph.D. plus various bits and pieces. My mother thinks I have almost as many degrees as a small thermometer. I have written quite a number of books and articles since I got my Ph.D., and I am not planning to hang up my biro just yet, so I think I am qualified by personal experience to write about writing.

What I propose to do in this article is to demolish certain "mythconceptions" (as I like to call them) about the process of writing. The ideas I am going to share with you have changed my attitude towards writing and helped to take some of the pain out of it, and I hope that they may help readers of this journal to write more easily and with greater enjoyment or at least less agony.

Numb de Plume

I suppose everyone will agree that many people, research students included, find writing difficult, distasteful, and even painful. Most people know the feeling; a survey by David Green and Peter Wason of University College London found that a dislike of writing was four times more common among experimental psychologists than either reading or talking (Wason, 1985). As a result, writing often gets put off again and again until it is finally confronted like a visit to the dentist -- and getting sensible words out can be remarkably like drawing wisdom teeth. In severe cases there is complete paralysis, usually accompanied by feelings of guilt, helplessness, and despair -- a condition called writer's block, although a better name for it is numb de plume. When you suffer from this you can sit in front of a blank, expressionless sheet of paper or blank, maliciously winking word-processor screen for days on end without any text emerging. In practice this seldom happens; usually you suddenly remember several pressing household jobs that simply must be done immediately or books that have to be read. Although most people will recognise this pattern of symptoms, there are exceptions, that is to say there are people who always write easily and fluently and with pleasure, and researchers have learnt a great deal from them.

Numb de plume is obviously a kind of inhibition, and if you think about it you will probably agree that there are two possible things that might inhibit you when you try to write: (a) you might be doubtful or confused about what to say, or (b) you might be unsure about how to say it. This probably looks obvious on reflection, but the experts on the psychology of writing have pointed out that the way these two things relate to each other is far from obvious.

Classical Model

The point is that according to the classical model of the writing process, what you have to say and how you say it are two separate things. This model also assumes that you know what you have to say before you put pen to paper (or finger to keyboard), so in its practical application it is entirely concerned with how best to express your pre-existing thoughts in writing. It ignores the doubts and confusions you might have about what to say. Nearly all textbooks on writing take this classical model for granted usually without even realising that they are doing so. One of the most famous texts, The King's English by Fowler and Fowler (1930), for example, starts off like this: "Anyone who wishes to become a good writer should endeavour to be direct, simple, brief, vigorous, and lucid. This general principle may be translated into practical rules in the domain of the vocabulary as follows: Prefer the familiar word to the far-fetched. Prefer the concrete word to the abstract. Prefer the simple word to the circumlocution. Prefer the short word to the long. Prefer the Saxon word to the Romance." The Fowler philosophy has been summed up in the mnemonic KISS, which stands for Keep It Simple, Stupid. My favourite writing manuals are The Oxford Guide to English Usage (Weiner, 1983) and the little (and inexpensive) Oxford Dictionary for Writers and Editors (Oxford English Dictionary Department, 1981). The preface of the first of these states: "This book is designed to answer the questions about English usage that ordinary people are constantly asking; it covers the known areas of difficulty and controversy, leaving aside the parts of the English language that cause most native speakers no trouble. Typical examples of the subjects covered are: whether to write forego or forgo, Jones' or Joneses', enrolment or enrollment ... the correct use of data, media, and strata; the difference between deprecate and depreciate, imply and infer, militate and mitigate; whether to say different from or different than; and when shall should be used rather than will."

Although this model is tacitly accepted by most people, I believe that it is false, and what is worse is that it is pernicious because it stops writers approaching the problems of writing in the best way. The fact is that most of us usually do not know exactly what we want to say when we begin to write. If we follow the classical model we are positively inviting an attack of numb de plume. The first thing that is wrong with the classical model, in my opinion, is that it is based on an oversimplified dichotomy between what is to be said and how to say it, between the ideas themselves and their expression in writing. The second and even more serious problem with it is that it lacks any psychological dimension. The human mind, after all, is a remarkable and complex organ: it starts working on full power as soon as you get up in the morning, and it does not stop until ... well, come to think of it, until you sit down to write.

Generative Model

A more convincing model of the writing process, formulated independently by Peter Elbow (1973) and Peter Wason (1970, 1985), is called the generative model. According to this model, you discover what you have to say in the actual process of writing. "Meaning is not what you start out with but what you end up with" (Elbow, 1973). If the generative model is right, then instead of first getting your ideas perfectly clear in your head and only then starting to write, you should start immediately and write as quickly as possible, without making any corrections and without being self-critical about some of the garbage that you will see spilling out as surely as weeds follow rain. You should continue writing like that until your meaning gradually emerges, or is generated, by the writing process itself. When this method of writing is working well, you discover what you wanted to say as if you had known it all along. Once you reach that stage, you can go back and edit what you have written. The editing process is likely to be much more radical than what you are used to: you may amend and amend the text like an old pair of knickers until there is hardly a thread of the original stuff left. You may think that this is likely to be a slow method of writing, but in fact the evidence suggests, on the contrary, that it is actually quicker than the classical method (Wason, 1985). And there is no doubt that it is much more fun.

David Lowenthal and Peter Wason (1977) carried out a survey of the writing habits of academics. And what did they find? Academics who plan their writing in detail before they begin generally dislike writing; those who develop their ideas in the process of writing generally enjoy it. There is a psychologist in London who writes so much that he is in danger of becoming a target of green activists worried about deforestation. He admitted to me that he is a writing junkie; he seems to have written more papers than he has read, and for him writing is a virtual addiction. His fingers literally itch when he has not written anything for a day or so. Needless to say, the way he writes fits the generative model perfectly. There is research evidence to show that some people find it hard to adopt the generative model in their writing. It requires self-discipline to suspend critical judgement of what you are writing and to keep going until you have produced a fair amount of text, ideally a rough draft of the whole chapter or paper. Some people just cannot bring themselves to leave a bad sentence alone and go on to the next one. I often have this problem myself, and I find it quite difficult to practice what I am preaching. Also, personality factors seem to make the generative model of writing much more beneficial to some people than to others. The ones who benefit most are what social psychologists call high self-monitors, that is, people who readily adapt their manner of speech and writing to suit different audiences and readerships. But not even the low self-monitors seem to perform better with classical methods. Above all, the generative method is less likely to leave the writer in a state of brooding distress, unable to produce anything at all.


The best way to write a thesis is undoubtedly to do it as you go along, at least in draft form. If, for example, your thesis is going to report a series of experiments, then you should write up each experiment as soon after you have done it as possible, using the generative method if you can. It is advisable to do it quickly for two reasons. First, it is always easiest to describe what you have done, and to understand the data you have produced and presumably discussed with your supervisor, when it is fresh in your mind. An interesting cognitive illusion operates here: it is very hard at the time to believe that in just a couple of months' time you could possibly forget any significant details, because when you have just done the work it all seems so clear and vivid and part of you that you think you will remember it forever. In fact, if you leave it too long before writing it up, you are likely to lose your grasp of it much more quickly and much more thoroughly than you can imagine.

The second reason why you should do the writing soon -- preferably as soon as you have analysed your data -- has to do with the psychology of motivation. The point is that if you procrastinate, even for a short while, you are likely to start feeling guilty. Once you get to that point, looking at your notes or your data or even thinking about them is likely to raise your guilt level, and then the easiest way to minimise the guilt feelings will be to put the work right

out of your mind. But it may be necessary to raise your guilt level temporarily, by going back to the work, in order to get the writing done and thereby eliminate the guilt entirely. In other words, you may have to make yourself feel a little worse now in order to make yourself feel much better later -- something we are all familiar with in the field of medicine. The longer you leave it, of course, the worse the guilt is likely to become and the higher the barrier over which you must push yourself to get rid of it. There may be a threshold beyond which you cannot bring yourself to do it at all. The writing, which may not have seemed at all distasteful before the guilt built up, and may have been put aside for entirely innocent reasons, gradually becomes almost too horrible to contemplate, and if you force yourself to do it in that state of mind the results may resemble not just a dog's breakfast but a dog's breakfast after it has passed through a dog. Your supervisor will probably think you are lazy, and that will be quite true, but the point I am making is that laziness is not necessarily a passive motivationless state: It can be an energetic stress-avoidance strategy.


It goes without saying that you should try, if possible, to prepare your draft chapters on a word-processor and keep them electronically stored (but do keep back-up discs and also hard copies as insurance against loss). Your department ought to provide you with the necessary hardware for this, and I think you have legitimate grounds for complaint if it does not. Storing your text in this way will enable you to amend it as often as you and your supervisor wish and update the material for the final draft when the time comes to submit the thesis. Another benefit is that you will be able to re-edit the text in a form suitable for journal submission without having to key everything in from scratch. And it is a very good idea to try (at least) to publish your findings as you go along, rather than to wait as most research students do until you have handed the thesis in, attended your viva, and (barring mishaps) had your photograph taken with your Mum and Dad on graduation day, by which time your work will probably be out of date, I regret to say, and your mind will probably be on quite different things.

Finally, remember that, even in this post-Thatcher era, presentation is very important. I cannot stress too strongly the importance of rigorous presentation, which includes accurate citation of references and very careful drawing of inferences from the literature you cite in introductory sections and from your own data in discussion sections. Having acted as either Internal or External Examiner for quite a number of theses, my impression is that they are more often open to criticism for unjustified inferences than for almost anything else. But do try to write in clear, simple, easy-to-read English. Lucid writing is highly valued by examiners but rarely found in psychology theses. Most psychology theses are about as exciting to read as the Tunbridge Wells telephone directory. Be rigorous, of course; but try also to be readable, bearing in mind that the external examiner, probably the only outsider who will ever read your thesis, will come to the material cold and will probably have enormous difficulty understanding it, still less enjoying it, unless you make a special effort. What you should aim for is rigour without the mortis.


Elbow, O. (1973) Writing without Teachers. Oxford: Oxford University Press.

Fowler, H.W. and Fowler, F.G. (1930) The King's English. Oxford: Clarendon Press.

Hartley, J. (Ed.) (1980) The Psychology of Written Communication. London: Kogan Page.

Lowenthal, D. and Wason, P.C. (1977) "Academics and their writing." The Times Literary Supplement, 24 June, p. 781.

Oxford English Dictionary Department. (1981) The Oxford English Dictionary for Writers and Editors. Oxford: Clarendon Press.

Wason, P.C. (1970). On writing scientific papers. Physics Bulletin, 21, 407-408. Reprinted in J. Hartley (Ed.), (1980) The Psychology of Written Communication (pp. 248-251), London: Kogan Page.

Wason, P.C. (1985) How to write an essay. The New Psychologist, May 1985,15-19.

Weiner, E.C.S. (1983) The Oxford Guide to English Usage. Oxford: Clarendon Press.

[Back to Top]