---
name: orgstud-theory-development
description: Use when building the theoretical engine for an Organization Studies (OS) manuscript — the mechanism, process model, or conceptual move that constitutes the contribution. Builds the argument; it does not select methods or run analysis.
---

# Theory Development (orgstud-theory-development)

## When to trigger

- You have a rich phenomenon (from `orgstud-topic-selection`) but no theoretical machinery that *organizes* it
- Your "theory" is a set of labeled themes or a box-and-arrow diagram with no generative mechanism
- A reviewer says the theory is "thin," "descriptive," "a relabeling of the data," or "doesn't engage the literature it claims to advance"
- You must decide whether the contribution is a new construct, a process model, a reconceptualization, or a boundary condition

## Theory is the deliverable at OS

This is the load-bearing skill in the pack. At OS the **theory itself is the contribution** — more so than the dataset or the estimator. The journal's European, sociologically-rooted tradition means process and interpretive theorizing are fully legitimate as the *main* claim (a different default from variance-template journals). Your job is to produce a **generative theoretical move** readers elsewhere can pick up and use.

State the **mechanism / process logic**, not just constructs:

- Name the **actors** and what they are *doing* (sensemaking, justifying, translating, resisting, categorizing, accounting, coordinating).
- Name the **social structure** they act within (institutional field, logic, hierarchy, network, category system, discourse, status order).
- Specify the **generative logic**: why does this actor, in this structure, produce this sequence or this outcome? For process work, specify the *sequence and transitions* (phases, feedback, turning points), not a static relationship.
- Specify **boundary conditions and scope**: where the mechanism holds, weakens, or reverses.

A test: can a reader *predict a non-obvious implication* — or *re-see a familiar phenomenon* — from your theory? If not, you have a label, not a contribution.

## Branch by contribution type

| Contribution type | What "good" looks like at OS |
|-------------------|------------------------------|
| **New process model** | Phases/turning points with explicit transition logic; a process figure; what each phase does that the prior one could not |
| **Reconceptualization** | An existing construct (e.g., legitimacy, identity, routine) redefined so that a puzzle dissolves or a new range of cases comes into view |
| **Extending a tradition** | A precise statement of what institutional/practice/sensemaking theory missed, and the specific amendment your evidence forces |
| **Critical/reflexive** | A theoretical account of how power, ideology, or taken-for-grantedness operates — critique that *theorizes*, not critique that only denounces |
| **Theoretical synthesis / essay** | Two conversations bridged into a usable new lens, with the move stated crisply enough to cite |

## For inductive / grounded work

If theory is *emergent*, this skill still applies — in reverse. The mechanism and process model are the **output**, presented after the data structure (see `orgstud-data-analysis`). Still required: a named mechanism, a process model (usually a figure with phases and feedback), and an explicit statement of novelty relative to existing organization theory. A grounded theory can be the *entire* OS contribution — it need not be a stepping-stone to a later test. Avoid "theory by adjective" (calling a pattern "dynamic," "recursive," "emergent" without specifying the process).

## For deductive / quantitative work

Hypotheses must be *derived*, each with its own causal-logic paragraph tied to the same core mechanism — not an asserted "bag of hypotheses." Moderators must be theoretically motivated, not data-mined. Even here, lead with the organizational mechanism the design illuminates; the estimator is in service of the theory.

## Engaging social theory

OS expects live engagement with social theory, used precisely. If you invoke institutional logics, Bourdieu's capital/field, Foucauldian discipline, Weickian sensemaking, or practice theory, deploy the concept's actual machinery — do not name-drop it as ornament. Misusing a borrowed framework is a fast reviewer pushback.

## Checklist

- [ ] A named generative mechanism (actors + structure + logic), not just constructs/themes
- [ ] For process work: sequence, phases, and transition logic specified (not a static X→Y)
- [ ] Boundary conditions / scope stated (where it holds, weakens, reverses)
- [ ] The move yields a non-obvious implication *or* lets readers re-see a familiar phenomenon
- [ ] Borrowed social theory used with its real machinery, precisely, not as ornament
- [ ] Inductive: theory is output, with a planned process figure and explicit novelty
- [ ] Deductive: each hypothesis derived from the shared mechanism; moderators motivated

## Anti-patterns

- Box-and-arrow or theme-list models with no mechanism ("constructs in search of a story")
- "Theory by adjective" — naming a pattern instead of explaining how it works
- Variance framing forced onto a process phenomenon (or vice versa)
- Critique that denounces without producing a theoretical account
- Name-dropping Bourdieu/Foucault/Weick without using the framework's actual logic
- A contribution stated as a finding ("we find X") rather than a usable theoretical move

## Output format

```text
【Contribution type】process model / reconceptualization / extension / critical / synthesis
【Core mechanism】actors + structure + generative/processual logic
【Process structure (if applicable)】phases + transition logic
【Boundary conditions】where it holds / weakens / reverses
【Social theory engaged】framework + how its machinery is used
【Novel move】the non-obvious implication or re-seeing
【Next skill】orgstud-literature-positioning
```