The basic phenomenon of figure-ground reversal is one in which there are two distinct ways in which to see an image. In famous examples, the same image can be seen in two different ways. For example we might see a vase, or we might see a pair of facing profiles.
Whatever we see, we see it one way at a time. We may be able to see the image in both ways, but we switch between them; we don't see both at the same time. I won't attempt a neurological explanation, but I don't doubt that there is one.
George Lakoff speaks of "biconceptualism"--the ability to use two different views of the world--which is a similar concept, but not visual. We can explain situations in different ways--for example, how does one respond to an enemy or one who has wronged you? The Bible says both to take retribution (an eye for an eye, etc.) and to treat that person with love (love your enemies). We can take these as two different models on how to deal with the same problem: we don't want someone to hurt us and ours repeatedly, so what do we do? The hawks say that the answer is to go to war; the doves say the answer is to seek other means to reduce the aggression.
So working with a writing project (or any project, really) one of the questions we're faced with is how to look at the situation: what view are we going to use. The examples above, and other simple examples, use binary comparisons, e.g., is the glass half full or half empty? But as a matter of practice, there are any number of different views that one can take of a subject. Horst Rittel, late professor of architecture at the University of California, Berkeley, argued that many problems that we face can be described in many different ways--"there is no definitive formulation of a wicked problem," he wrote--a wicked problem being what he called problems that did not suit the rationalistic problem-solving techniques of his time (I'm digressing a little here).
Think of what it means if there is no definitive formulation to a given problem. What if the problem is to get a dissertation done? Well that's one very general formulation. More specific formulations would be "to convince your professors that you're done" or "to write a great dissertation on my topic of study." These two formulations are distinct, though not mutually exclusive.
The other day Sarah asked for common flips that might occur to grad students or researchers. I mentioned one example: the flip between seeing the project as an expression of your own interests and your own wisdom, and seeing the project as an attempt to say what is necessary to please the professors.
Another flip might be between having too much research material and having too little. This, I think, afflicts a lot of people in the literature review sections, or in their work if they don't have a clearly defined research agenda, or if their project is open-ended, as a project in literature or history might be. At one moment a person can be saying "I have so much material that I have to read and organize that I'm overwhelmed." The next moment the same person might say "I need to do more research." Admittedly, these are perspectives that might reasonably co-exist--sometimes we do have too much material to handle and still have to acquire more, if, for example, a professor requested more--but from a pragmatic point of view, we can only act effectively on one of these perspectives at once: either we can try to get a better handle on the material we have, or we can try to gather more. Both courses might be productive, but vacillating between them is not.
I don't know that I can think of other specific perspective shifts that might be common to the graduate student writer, but I think that the writer who learns to deal with these shifts, and learns to manage the different perspectives is well-positioned to respond to the myriad different demands of the long-term research project.