56 4 Making a Strong Start
for a mature researcher—the sort of person the PhD student wants to become—but
can be an effective way of learning.
There is a balance between respecting a supervisor and thinking independently.
Supervisors are mature, and are in their role because their knowledge and instincts
and experience are reliable; but as your expertise grows, you will sometimes ques-
tion your supervisor, and some of those times you will be right. Every supervisor
has stories of bull-headed students who insist on ignoring their advice and getting it
wrong, or, less often, ignoring their advice and getting it right;^3 and, conversely, of
students who mindlessly follow their advice and never try to think for themselves.
As I said, you need to find a balance.
A colleague, Robert, told me about a student (let’s call him Tom) who came to his
office and reported the outcome of an obviously silly chemical experiment—Tom
should have noticed that the equipment simply wasn’t capable of detecting the ef-
fect he was looking for. After some discussion Tom suggested that he may have
made a mistake. Robert then asked Tom what would happen if the reagents were
changed (although they were irrelevant to the failure). A week later, Tom reported
back that the experiment still hadn’t worked. When Robert asked whether he had
expected anything different, Tom suddenly looked troubled; it was clear he hadn’t
ever stopped to ask whether what he was doing was sensible. Tom asked why Rob-
ert had suggested the experiment at all; Robert asked back, ‘Why did you do the
experiment without thinking?’ The lesson for Tom strikes me as a harsh one,^4 but
incredibly important: all students must learn the skill of critical thinking, not just
every now and again but in every aspect of their research.
Most of the anecdotes in this book concern cases that led to a happy outcome,
but not all PhDs go well, and there are lessons to be learnt from the failures. Two
cases that are of relevance here are PhDs that went off track due to problems in the
student–supervisor working relationship.
Hasrim did not settle into an effective working relationship with me. He initially
felt that commencing research would be as straightforward as (if longer-term than)
the task of, say, getting ready to teach a new subject: find resources, do some read-
ing, follow a schedule of experiments and investigation, and write up. He had previ-
ously completed a Masters thesis at another university, but it quickly developed that
this previous experience had not prepared him well, as it had largely consisted of
uncritically repeating an earlier investigation undertaken by his supervisor there; he
had not even had to search for background literature. (Without the context of having
read the supervisor’s earlier paper, Hasrim’s thesis looks like a sound, independent
piece of work, but when they are read together the lack of depth is obvious.) In
(^3) A student of mine persisted with work on an algorithm that I ‘knew’ was foolish, and ended up
with a strong result and a paper in a top journal. But this case was a rare exception.
(^4) This was in the 1980s. In my view Robert should not have deliberately wasted a week of his
student’s time, although it does seem that the lesson was an effective one. I sometimes use this
same anecdote as an example of the kind of treatment of students that was once common but now,
happily, seems to be dying out.