Saul Greenberg

back to The Thesis Oral

How to survive a thesis defence (extract)
by Joe Wolfe, School of Physics, The University of New South Wales, Sydney

This page was reformatted from the text originally available at

The thesis defence or viva is like an examination in some ways. It is different in many ways, however. The chief difference is that the candidate usually knows more about the syllabus than do the examiners.

  • Some questions will be sincere questions: the asker asks because s/he doesn't know and expects that the candidate will be able to rectify this. Students often expect questions to be difficult and attacking, and answer them accordingly. Often the questions will be much simpler than you expect.
  • In a curious relativistic effect, time expands in the mind of the student. A few seconds pause to reflect before answering seems eminently reasonable to the panel, but to the defender it seems like minutes of mute failure. Take your time.
  • For the same reason, let them take their time. Let them finish the question.

The phrase "That's a good question" is exceedingly useful. It flatters the asker and may get him/her onside, or less offside; it gives you time to think; it implies that you have understood the question and assessed it already and that you have probably thought about it before. If necessary, it can be followed by a bit more stalling "Now the answer to that is not obvious/straightforward..." which has the same advantages.

  • If the nightmare ever did come true, and some questioner found a question that put something in the work in doubt... mind you this is thankfully very rare.... then what? Well the first thing would be to concede that the question imposes a serious limitation on the applicability of the work "Well you have identified a serious limitation in this technique, and the results have to be interpreted in the light of that observation". The questioner is then more likely to back off and even help answer it, whereas a straight denial may encourage him/her to pursue more ardently. Then go through the argument in detail - showing listeners how serious it is while giving yourself time to find flaws in it or to limit the damage that will ensue. In the worst caese, one would then think of what can be saved. But all this is hypothetical because this won't happen.
  • What usually happens is that the examiners have read the work perhaps twice, and looked closely at some parts that interested them most. These are usually the good bits. They are not out to fail you. It is a lot more complicated to fail you than to pass you. In general, they feel good about the idea of a new, fresh researcher coming into their area. You are no immediate threat to them. They have to show that they have read it and they have to give you the opportunity to show that you understand it (you do, of course). And they usually have a genuine interest in the work. Some of them may feel it is necessary to maintain their image as senior scholars and founts of wisdom. Judicious use of the "Good question", "Yes, you're right of course", "Good idea.." and "Thanks for that" will allow that with a minimum of fuss and a maximum of time for champagne drinking.
  • If one of the examiners is a real ..., your thesis defence is probably not the best place and time in which to do anything about it, except perhaps for allowing him/her to demonstrate it clearly and thus to establish the support of the rest of the panel. If you want a major dispute, save it up for when you are on even ground, unless you are very very sure of yourself and think that you have nothing to lose.
  • Be ready for a 'free kick'. It is relatively common that a panel will ask one (or more) questions that, whatever the actual wording may be, are essentially an invitation to you to tell them (briefly) what is important, new and good in your thesis. You ought not stumble at this stage, so you should rehearse this. You should be able to produce on demand (say) a one minute speech and a five minute speech, and be prepared to extend them if invited by further questions. Do not try to recite your abstract: written and spoken styles should be rather different. Rather, rehearse answers to the questions: "What is your thesis about, and what have you done that merits a PhD?".
  • Read points the first two bullet points again. Keep calm - and good luck!

Opinions expressed in these notes are mine and do not necessarily reflect the policy of the University of New South Wales or of the School of Physics.

Joe Wolfe / 61-2-9385 4954 (UT + 10, +11 Oct-Mar)