Orgstud Theory Development

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

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

【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

Source: brycewang-stanford/Awesome-Journal-Skills → Organization-Studies-Skills/skills/orgstud-theory-development/SKILL.md