The Focused Introduction
Writing the Theoretical Framing
Preface
This guide grew out of frustration with a pattern I kept seeing in my own lab. Talented early career researchers would produce manuscript drafts that were theoretically exhaustive but argumentatively weak. Every relevant theory got a paragraph. Every related concept got a mention. The papers were comprehensive and hard to read.
The problem was not knowledge. My students knew the literature well, sometimes too well. The problem was a set of unwritten rules: how to decide which theories are load-bearing for your argument and which are not, how to write for a reader who needs to be convinced rather than impressed, and how to survive a review process that sometimes rewards the wrong things.
I work in cognitive psychology, at the intersection of perception, memory, and narrative cognition. It is a field rich in theory and, at times, poor in argument. We have more frameworks than we have crucial tests between them. I have watched gifted researchers spend months writing introductions that cover the landscape rather than staking a position in it, and I have read the reviews that follow: specific, detailed, and asking for three more theories. Those theories then get added. The paper that results is what the next cohort of PhD students reads as a model of how a strong introduction should look. What they see is breadth; what they cannot see is the revision history behind it.
This problem has become more visible with the arrival of AI writing tools. Ask one to improve your introduction and it will suggest theories you have not mentioned. The writing comes back polished and bloated, and it sounds authoritative enough that the missing argument becomes harder to notice.
This guide is my attempt to make some of those unwritten rules explicit. It is not a comprehensive textbook on scientific writing; there are better books for that. It is a practical guide for researchers who already know their science and are struggling to turn it into a paper that works. The final chapter addresses AI writing tools directly, and how to use them without letting them make the problem worse.
The problem this guide addresses is most acute in theory-rich fields like psychology, where phenomena attract multiple competing accounts and reviewing culture reinforces citation of all of them. In fields with tighter theoretical consensus, or where data more reliably rule out competing accounts, the pressure is lower. But psychology is not one of those fields, and neither are most of the disciplines adjacent to it.
The diagnosis is not new. Oberauer and Lewandowsky (2019) have argued that psychology’s theory crisis stems less from bad data than from a weak connection between theories and the predictions researchers actually test. The concept of strong inference, introduced by the physicist John Platt (1964), points toward a remedy. Platt observed that some fields were advancing far faster than others, and traced the difference to a specific habit of mind: devising competing hypotheses and designing experiments that could exclude at least one of them. He was writing for bench scientists. But his core insight applies directly to the problem this guide addresses. A theoretical framing that cannot say which result would count against which theory is not doing scientific work. The filters in this guide are, among other things, a way of asking Platt’s question at the level of the manuscript: does this theory belong here because it can lose, or because it feels relevant?
The examples throughout are drawn from event cognition research, which is where most of my own work lives. But the principles apply wherever you write a paper that makes a theoretical claim. If you work in an adjacent field, the norms around theoretical framing may differ from what is described here; treat the specific advice as a starting point rather than a prescription.
The theories are already in your head. This guide is about deciding which ones belong on the page.