believe in methodological and theoretical pluralism. What matters most is that
the work is interesting and well-executed. However, given my own experience, I
am a suitable adviser for only positivist dissertations. That is, I expect
students to develop falsifiable arguments that seek to generalize about some
kind of political phenomena. I am not an appropriate advisor for someone who
wants to do a post-modern critique or for someone who is interested in a single
case of something but not in thinking more broadly about it. Likewise, the
dissertation should not be normative (not about should or should not) but
analytical (why, why not). If you are interested in a normative dissertation,
see the political theorists in the department.
students do not have to be rationalists or rational choice theorists (these
terms and groups of people are not identical), but my work largely considers how
actors (individuals, groups, states) face difficult choices and try to pick the
best ones. I prefer deductively derived theories, but induction is appropriate
if done well.
interests cross fields—both IR and Comparative Politics. However, I am not an
expert in all questions in both fields. In IR, my expertise lies in comparative
foreign policy, intervention into civil conflicts, and security in general.
There are others in the department who are more expert in some security issues
(proliferation, arms control) and in all IPE issues. In terms of Comparative
Politics, I study ethnic conflict, particularly separatism and elite political
competition as mediated by institutions, but am not an expert on social
movements. I know some about Eastern Europe, specific African and South Asian
countries for certain periods of time, but I am not a country expert of any
largely puzzle driven—that is, there is some question, some anomaly that needs
explanation. It could start with theory—does this theory explain something
else? How do these theories compare to each other? Or, it could start with
some empirical anomaly. What is this a case of? Why does this pattern of
A dissertation (or a
thesis) is not a speculation about future events, but an examination of past
behavior (case studies, quantitative analyses, etc) to determine the patterns of
causation. The dissertation should not examine an event in progress, as this
will make completion difficult or impossible.
dissertation should be interesting—that it address a question that is relevant
for theory and for policy. My view of social scientists is of scholars who seek
to understand reality not just for their own benefit but to improve their
surrounding society. However, the goal of policy relevance should not drive the
findings, but that the student ought to consider the policy implications of the
should be interesting to you—it will consume much of your time and energy not
only until you finish it but beyond, as your first several years of publications
and teaching will still focus on your dissertation.
will probably take two or more years to complete. Rushing a dissertation is a
bad idea, as it is the foundation for your career. It establishes your
reputation, sets your research agenda, and may provide more than a few
publications for yourself. However, one should not dawdle either, as funding
has a nasty tendency to be time-based and to run out.
will not include everything that you researched. Some stuff, it turns, will
not be relevant. Just because you read something or ran a particular analysis
does not mean that it is relevant.
Do not expect instant feedback. Do not turn
in drafts expecting comments immediately or expecting a defense of the proposal
or the dissertation within days. The professors on your committee take their
own duties and your dissertation seriously, so expect that they will have other
things to do and that they will take some time reading your stuff thoughtfully.
Take seriously the feedback you receive. If
you have a fixed idea of what you are going to do and how you are going to do
it, find someone else to rubber stamp it. Social science is an incredibly
social effort—you will get feedback your entire career (reviews from journals to
say the least). If you do not respond to constructive criticism in a productive
manner (you can disagree, but you need to disagree constructively), then your
work will rarely develop and improve.
Do not appear in my office five minutes before
I have a class. Schedule your appointments so that we can focus on your work,
and make sure that you hand in the written work with enough time so that it can
be read before the appointment (see #1).
Please double space all drafts, bind, include
a header on each page that has your name, a chapter title/phrase, and
page numbers. If you are at McGill, please print out the chapters and give
them to me or put them in my box. Email attachments are ok if you are
working elsewhere, as long as I can open them (MsWord, PDF files are
easiest) and virus-free. If you send me one file with a virus, I will no
longer accept attachments from you—it will have to be paper from then on.
** New rule: No chapter should be more
than 45 pages. If any chapter is significantly longer than what an article
would be, then revise. Do not submit chapters to me before you have edited
them--which not only refers to improving the writing but seriously reading the
chapter and dropping all that which is not relevant. The less you waste my
time, the quicker I can get comments back to you.
I check my email very frequently, so
that is the easiest way to reach me. I check my office phone messages less
frequently, and my mailbox in the political science office much less frequently.
If you want me to write you a letter of
recommendation, please give me a CV/resume, a self-addressed envelope and
whatever forms that are necessary. If it is for further educational work,
internship, or job, please provide a paragraph describing your goals and the
program/position. And see my
instructions for letters on this website, although letters for grad and
undergrad students will differ.
Expect both emails and handwritten comments on
your drafts. The latter will be hard to read and may need some de-ciphering.
A literature review—not just a summary of
the stuff you read, but rather an assessment of how the question has been
asked before--the strengths and weaknesses of previous attempts to answer
the question (again, not everything that you thought would be relevant is).
The review also sets the context—how does your question fit into larger
questions of interest to scholars in the field.
An answer—your theory. It should not
float from nowhere, but should emerge logically from the literature
review—building on previous work, borrowing theory from other areas (other
issues in the subfield, other subfields, other disciplines). Your theory
can have more than one or two variables in it, and it does not have either
be Realist or anti-Realist, but it needs to be logically coherent. It
should be falsifiable—under what circumstances would your theory be wrong?
Testable hypotheses (If x, then y).
These should be derived from your theory.
A methodology. The proposal should
explain and justify how you will test your hypotheses. No single method is
always the best for all topics and questions. You will have to figure out
which method or methods will work best. If doing case studies, you should
consult the literature on qualitative methods (see
my old syllabus for some hints) to determine strategies of case
selection. If doing quant work, research the availability of datasets and
think carefully about how you will operationalize your concepts. Consider
which kinds of techniques you will have to learn.
Some thought about policy implications.
Iterations—your first dissertation
proposal will not be your last. It will evolve over time, after you receive
comments on previous version.
Since moving to McGill, I have received many emails
from students in Canada and around the world asking me about the graduate
program and seeking to work with me. While this is flattering, our system does
not work this way. Students apply to the department. If they are accepted
and decide to come, then they find advisors generally. I do not "Cherry-pick"
files and push them through the process. So, do not email me about yourself
since it will not help you. The department and university websites have a great
deal more accurate information about the graduate programs here than I
possess. If you get in, then feel free to email me then.