My work is on narratives of rice in literary texts in Kerala, 21st and 20th century. I treat it as goal in itself, but the long-term plan is to use the observations from my work to speculate on the formation of the entity called the Malayali, especially through material interactions described in narratives.
The dissertation is the thesis T which is in turn supported by arguments O, P, F.
O, P, and F are arguments of the constitutent chapters 1, 2, and 3 respectively.
Each of O, P, F are in turn supported by arguments A, B, C.
These A, B, C are the slightly abstract arguments of a single chapter, say, chapter 1 that makes up the argument O.
Each of A, B, C are in turn supported by a topics identified as X, Y, and Z. (X, Y, Z are observations made from the text studied; is embedded in the text and is concrete.) X is related to A, Y is related to B, and Z is related to C.
Each of X, Y, Z are in turn supported by R1 and R2, which are two instances where X (or Y or Z) is made more complex, is dealt with in a particular way, is played with and so on.
Thus, T=O+P+F=(OPF)·(ABC), where A, B, C are of the structure X+R1+R2
R1 and R2 are the most grounded, then comes X Y and Z which are still based on the text, then A B and C (goes out of the text to society and history), then O P and F (a figuration, a way of seeing), and then T.
Literary studies dissertations, although they have an inner schematic, is built through rhetorical moves using certain phrases. Mastering this is key to good writing. There are academic phrasebanks available online for reference. Or you can write a basic regex-searching script in python and run a dissertation through it to find phrases such as “I read…” “The novel examines…” and the like. You might be able to find a few on my github page (there is a long term plan to provide a front-end for this text parsing).