brycewang-stanford Literature Positioning (psychrev-literature-positioning)
When to trigger
- The model is forming but you cannot name whose account you improve on
- Reviewers (you anticipate) will ask "how is this different from [established model]?"
- Your introduction cites broadly but engages no specific theory
- You are unsure whether your idea is new or a re-derivation of a known model
The core move: name the rivals, not just the topic
At Psychological Review the central positioning question is not "what is the area?" but "which existing theories make claims about this explanandum, and what does yours do that they cannot?" Editors and reviewers are typically drawn from exactly the modeling tradition you enter. A paper that does not engage the standing models reads as unaware of its own field.
For each phenomenon in your explanandum, build a rival map:
| Phenomenon (diagnostic) | Rival theory/model | What it predicts | Where it fails / is silent | Your account |
|---|
The most persuasive Review papers select diagnostic phenomena — cases where rivals make different predictions — and show their theory gets these right while saying clearly where rivals do not. "Systematic evaluation of alternative theories" is in the journal's own scope statement; positioning is where that evaluation begins.
Three positioning stances (pick the honest one)
- Subsumption — your theory recovers the rival as a special case and explains more. The strongest stance; show the limiting condition under which yours reduces to theirs.
- Adjudication — two rivals conflict; your theory (or a formal contrast) settles which is right, or shows both are special cases of a deeper mechanism.
- Extension into new territory — rivals are silent on phenomena yours explains. Weaker unless the new territory is important and the rivals genuinely cannot be stretched to cover it.
Avoid the false-novelty stance ("no one has modeled this") when a standing model could be applied — a reviewer will apply it for you.
Reading the conversation honestly
- Cite the current state of each rival, not its 1990 form, if it has since been revised.
- Represent rivals at their strongest (steelman). Beating a weak version is the classic reviewer complaint at this journal.
- Distinguish a rival's core commitment from its implementational details; do not claim victory by defeating a replaceable detail.
- Locate yourself relative to formal rivals where they exist — a verbal contrast with a formal model invites the question "but what does your model predict quantitatively?"
Checklist
- Every diagnostic phenomenon has a named rival theory/model
- The rival map states what each rival predicts and where it fails or is silent
- Rivals are represented at their current, strongest form (steelmanned)
- The positioning stance (subsumption / adjudication / extension) is explicit
- Where a rival is formal, your contrast addresses it on formal terms
- The "before → after" the field can do is implied and ready for contribution-framing
Anti-patterns
- A topic-level literature review with no specific rival models named
- Beating a strawman or an outdated version of a standing theory
- Claiming "no one has done this" when an existing model obviously applies
- Citing rivals only in the introduction, then never contrasting predictions
- Confusing positioning (whose account) with contribution (what is new) — keep them separate
- Ignoring formal rivals because your own theory is verbal
Output format
【Conversation】[the modeling tradition / theoretical debate entered]
【Rival map】[phenomenon → rival → its prediction → its failure → your account] (table)
【Stance】subsumption | adjudication | extension — with the limiting/diagnostic condition
【Steelman check】rivals stated at strongest, current form: yes / fix
【Next step】psychrev-theory-construction (if model not built) → psychrev-argument-development