Your reader already understands the basic concepts of your theory. So you don’t have to teach your theory to your reader; you simply have to explain how you are applying it to your particular object. In an important sense, your theory section constructs the objectivity of your object, it establishes a space of possible arrangements of things in the world, it “frames” your observations within this space. It is true that it identifies your “point of view” on your empirical material but it is no less true that this point of view is easily available to your readers. Their own training (in your discipline) allows them to easily step into it and see the world through your eyes. You share a perspective on the world and we call this perspective your “theory”.

In this talk I tried to bring together Ezra Zuckerman’s suggestion that you should strive to make your theory “compelling” with Steve Fuller’s understanding of theory as “presumption”. I lose track of my argument a bit near the end, when I accidently say (at 44:52) that you don’t have to help your reader to believe that your theoretical claims are true because the reader already believes them. It would be more precise to say that the reader presumes that they are true “for the sake of argument”, that is, they are willing to proceed (to your analysis) as though the theory is true. That’s why you have to make sure that your theoretical model is compelling — we should be in a certain sense “forced” to see the world as the theory suggests. By reminding ourselves that we no more believe in our theories than we believe that the accused in a trial is innocent (we may, indeed, believe the opposite!) we naturally check our biases — our “compulsion” to believe certain things without thinking carefully about them. Theoretical presumptions are not hypotheses to be tested (that comes later); they are frames that remain in place while the testing going on and are then removed when the testing is complete. Presumptions are “procedural” checks on our compulsions.

This ties in nicely with my other ideas about theory, such that a theory is a “system of expectations” or what Bourdieu calls a “program of perception”. Our theories lead us to expect certain things of our objects and therefore “program” us (like computers) to see the world in particular ways, noticing some things and ignoring others. When writing our theory section we’re being upfront about our programming. As I say in the talk, theories are ways of marshalling our passing “thoughts” about “things” in the more logical space of “concepts” and “objects”. Without theories, things are just lying around wherever and however they like. Our theories provide a “frame” in which to “model” their behavior, helping us to distinguish “normal” from “deviant” data points (yes, I’m trying to pun on statistical language there). This also become the frame in which you set up the hypotheses you will test. As Ezra reminds us, it is your “null” that is (presumably!) compelling. Your hypotheses (which, if true, will “reject the null”) are even more interesting.

There are softer ways about this. You can use your theory simply to shape your reader’s curiosity, or you can even leave the explication of your model until the end, after you have “grounded” it inductively. In such cases, I would strongly recommend writing a literature review that anticipates the conceptual resources you end up using, and putting it before your methods description. These are issues that we’ll touch on again next week, when we talk about method, and then again when we talk about how to structure the whole project.

Here are some more resources on the subject:

“Theory as Expectation”

“Theories, Concepts, and Models”

“Theory, Method, Style”

“Reading, Talking, Seeing”

“Consensus, Controversy, Contribution”

“Never Write Literature Reviews”

Here’s what I say about it in the “How to Structure” lecture:

And here is the video from 2022.