---
name: brycewang-stanford/amr-writing-style
source: https://app.decimal.ai/s/brycewang-stanford-amr-writing-style@1/SKILL.md
source_sha256: 2ae8a8d40e43
---

# Writing Style: AOM House Style, Argument-Driven Prose (amr-writing-style)

## When to trigger

- The theory and contribution are settled; now the prose needs polish
- The draft reads like a literature review or an essay, not a built argument
- Sections are present but the manuscript does not "flow" toward the contribution
- You are unsure of AMR's structural conventions

## AMR's structure (durable form)

AMR conceptual articles follow a recognizable arc. Verify exact requirements on the current
AOM author page, but the durable shape is:

1. **Introduction — the theoretical puzzle.** Open with the tension, not a definition or a
   dictionary opening. State the puzzle, the conversation, and your move within the first
   two pages. Tell the reader what theory cannot currently do and that this paper fixes it.
2. **Theoretical development.** Constructs defined → assumptions → relationships +
   mechanisms → propositions (each earned by argument). This is the body and the bulk.
3. **The conceptual model / figure** integrated where it clarifies the argument.
4. **Discussion.** Contribution (what's new, differentiated), boundary conditions, and
   implications for theory, future research, and practice.

There is **no methods, results, or data section** — AMR is theory-only. The reasoning *is*
the contribution, so make the *Why* vivid (Whetten, AMR 1989, DOI 10.5465/amr.1989.4308371);
define constructs precisely and consistently (Suddaby, AMR 2010, DOI 10.5465/amr.2010.0419).
Exemplar to study for narrative: Oliver's "Strategic Responses to Institutional Processes"
(AMR 1991, DOI 10.5465/amr.1991.4279002) — puzzle → typology → propositions → implications,
no methods, no results.

## Voice and prose discipline

- **Argument, not survey.** Every paragraph should advance the theory. If a paragraph only
  reports what others found, it is review filler — cut or repurpose it to set up your move.
- **Define before you use.** A construct's first appearance carries its definition; use the
  exact same label every time afterward. Terminological drift reads as conceptual fuzziness.
- **Signpost the logic.** "Because... it follows that..."; "This implies..."; "We therefore
  propose..." Make the inferential moves visible — at AMR the prose *is* the proof.
- **Propositions in the text** are set off and numbered, each immediately following the
  argument that earns it.
- **Precision over flourish.** AMR prizes conceptual precision. Avoid vague intensifiers
  ("very important," "highly significant") and metaphors that do no theoretical work.
- **No hypothesis vocabulary.** Never write "we hypothesize," "H1," "results show," "we
  find," or "supported / not supported." AMR uses **propositions** because nothing is tested;
  that vocabulary is AMJ's and reads immediately as wrong-journal to AMR editors.
- **AOM style mechanics.** Use the Academy of Management Style Guide: author–date in-text
  citations, full reference list, Times New Roman 12-point body text, double-spaced manuscript,
  1-inch margins, no page numbers or running heads in the submitted manuscript, and an abstract
  of no more than 200 words. Treat "APA-style" as shorthand for AOM's author–date mechanics,
  not a claim about a named APA edition.
  Verify current manuscript length limits before submitting.
  Submission-stage facts such as AMR's current editor, no-fee policy, and ScholarOne link
  belong in `amr-submission` and `resources/official-source-map.md`, not in prose edits.

## Self-edit passes

- **Puzzle pass**: is the tension on page 1–2, or buried?
- **Survey pass**: highlight every sentence that only reports prior work; if it does not set up your move, cut it.
- **Definition pass**: is every construct defined once and labeled consistently?
- **Inference pass**: can a skeptical reader follow each step from premise to proposition?
- **So-what pass**: does the discussion differentiate the contribution, not restate it?

## Checklist

- [ ] Introduction states the puzzle and the move within ~2 pages
- [ ] No methods/results/data section exists
- [ ] Every paragraph advances the argument; no review-essay filler
- [ ] Constructs defined on first use; labels consistent throughout
- [ ] Inferential connectives make the logic explicit
- [ ] Propositions are numbered and follow their arguments
- [ ] Discussion differentiates the contribution and states implications
- [ ] AOM author-date citations and reference style applied (verify current guide)

## Anti-patterns

- A literature review masquerading as theory: paragraphs of "X found... Y found..."
- A dictionary or epigraph opening that delays the puzzle
- Construct labels that drift (same idea, three names)
- Buried propositions the reader must hunt for
- Empty intensifiers and decorative metaphor substituting for precise claims
- A discussion that summarizes the paper instead of stating what is newly thinkable

## Output format

```
【Puzzle on page 1–2?】yes / fix
【Review-essay filler】[paragraphs to cut/repurpose]
【Construct-label consistency】ok / drift at: ...
【Logic signposting】explicit / add at: ...
【AOM style】citations + references conform (verify)
【Next step】amr-submission
```