Writing a Thesis or Dissertation
copyright © 2004-2007 by Mary Bucholtz, University of California, Santa Barbara
Unlike other graduate requirements, the thesis or dissertation doesn't have clearly defined limits, and its scope and content are largely up to you. This makes the task more challenging than other parts of your graduate training, as well as more valuable to you. Many graduate students find the task of writing a dissertation most difficult not intellectually but psychologically--it's hard to know where to begin and whether you're doing it "right." The suggestions below therefore focus on overcoming obstacles of this kind. In what follows, I assume you've already collected your data but haven't yet worked through a detailed analysis, although in many cases you will have completed some preliminary analysis for your dissertation prospectus.
Establish a schedule with your advisor.
Depending on your department's culture and your advisor's philosophy, you may work largely independently or very closely with your advisor and/or other committee members throughout the thesis/dissertation process. You should talk to your advisor early on about what kind of schedule of meeting, writing, and consulting s/he recommends. Set up clear deadlines early on--your advisor won't necessarily expect you to meet these deadlines, but having a schedule will give you a goal to aim for, and it will break the otherwise daunting task of writing a thesis or dissertation into manageable chunks.
Find a writing partner.
Many students find it helpful to meet regularly with one or more other graduate students working on related projects (in their own or other departments). The idea is to exchange chapters or pieces of chapters and give each other feedback. This is especially useful if your advisor takes a more hands-off approach. It will give you deadlines to meet and useful suggestions from readers who may have a different perspective from your committee. For this arrangement to work, it's important to find people you really trust and can work well with.
Develop a broad support network.
It's important to be able to work independently on the thesis or dissertation, but it's also important to be able to ask for help when you need it. You may find that some people are especially useful for particular kinds of questions, so you might treat faculty (both on your committee and not) and fellow graduate students as a panel of experts that you can consult on particular issues. Most people are happy to answer a specific question, as long as you don't abuse their time.
Some schools require or allow one committee member to be from another university. This is highly recommended if it's practically feasible in your situation, not only in order to make sure that your dissertation will address a broad audience of scholars beyond your own department but also so that you can create social networks with scholars in other places, which will be valuable both intellectually and professionally throughout your career.
Find other sources of psychological support.
There's a whole industry of academic self-help materials: books, workshops, even dissertation coaches. Check the graduate office of your institution-- many schools sponsor workshops on how to write a dissertation, and it's useful to have on hand a book or two on the process that you can dip into when you're baffled or discouraged. A few recommended titles:
Create a structure.
Start with some kind of organizational scheme, which ideally you will have already developed in your dissertation prospectus. You don't have to stick to it, but it gives you a structure you can start putting things into. This may feel premature, especially if you haven't done much analysis yet, but you can always change it, and it makes a big difference in helping you to think about what you're looking at. Rough out a general table of contents for the dissertation, and as you start working through your data you can move the data into particular chapters. That way, when you sit down to write each chapter, you'll be able to assemble it relatively quickly and have a good sense of how long it will be. If you feel unable to put together chapters at this point, then sit down with your data and make a list of 20 to 50 themes or issues that you might want to address. Then start grouping these until a shape for the dissertation starts to emerge. Some things won't fit, but those you can always put into a separate article, so go ahead and flag those themes as well. (You'll be glad later that you did.)
Balance different kinds of tasks.
No matter how exciting the project, data analysis can be tedious and slow, especially in a big project like a dissertation. It's a good idea to balance that kind of work with other tasks, like doing background reading and writing up your literature review and methodology chapters, or other sections that don't require much data.
Work through the data in stages.
It's often helpful to go over the data in several stages: figuring out what you have, figuring out what to do with what you have, and then carrying out a detailed analysis. So for example, with audio- or video-recorded data it may seem that the ideal thing to do is to transcribe all your data before doing any analysis, but sometimes this isn't realistic. In that case, it's advisable to start out by creating an index: list, in great detail but without transcribing what speakers are saying/doing, and briefly note any particularly interesting linguistic phenomena. (This also helps to navigate through recordings when you're trying to find a particular excerpt.) Then go through the indexes (or transcripts if you have them) and code with keywords the things that you want to group together. Then go back over the data and do a detailed transcription which you can then analyze. (Chances are that some of your analysis will change as you look more closely at your data.) At many points, but especially early on, it's also useful to arrange a data session with your advisor and/or other committee member(s), or with one or more fellow graduate students.
Use writing to develop your analysis.
Analysis and writing go hand-in-hand, so don't put off writing anything until you're done analyzing the data. Writing up what you've done so far will help you to see what you still need, and maybe even help you figure out what's going on by forcing you to be precise and explicit. Even if you don't end up using everything you write (and you won't), it keeps you generating ideas.
Working regularly is the only way to keep the whole project in your head. If possible, try to work on the thesis or dissertation for at least 30 minutes a day every day (and ideally longer some days). Each week, make sure to give yourself a real day off--not just to do your laundry and dishes, but to relax and clear your mind. Writing is hard physical work--take regular breaks to stretch, rest your eyes, and get some exercise. Try to write something every day.
Have realistic expectations.
The thesis/dissertation is the capstone of an intellectual apprenticeship; treat it as such. Bear in mind that it is only the first step in what (we hope) will be a lifelong research process; it doesn't have to be the definitive work on your topic, just a solid original contribution to the field. Don't feel that you have to do an exhaustive analysis of all your data. Chances are that you have much more data than you could possibly use for a single project; it's okay to set some of it aside and deal with it after the thesis or dissertation. In fact, it's a really good idea to do so so that you can easily keep publishing without having to do a whole new project right away. Make your goals realistic and don't beat yourself up if you see problems and limitations with what you've done. Fix what you can, acknowledge what you can't (footnotes are a good place to put ideas you can't pursue in the dissertation/thesis itself), and plan to make improvements in the published version. Giving yourself permission to be imperfect will help you to get the thesis/dissertation written within a reasonable time frame. Remember that the best thesis/dissertation is a completed thesis/dissertation.
University of California, Santa Barbara | College of Letters and Science | Department of Linguistics
Last modified on: November 27, 2006.