Contribution
It keeps the standard definition of a routine, “a repetitive, recognizable pattern of interdependent actions, involving multiple actors”, and argues that the definition itself forces a two-part ontology. The habit, program and gene metaphors that dominated the literature treat a routine as a fixed object and so cannot explain the variation and change that field studies keep finding. Splitting the routine into ostensive and performative aspects, recursively related, gives change a source inside the routine rather than only in external shocks.
Key insight
A routine is never only its script. Borrowing Latour’s distinction for power, “in principle” and “in practice”, the ostensive aspect is “the routine in principle” and the performative “the routine in practice”. No rule fixes its own application (Wittgenstein, via Blau), so every performance improvises; performances then feed back into what the routine is taken to be. Whoever can “anoint performances as legitimate” can “turn exceptions into rules” (p. 110). Stability and change are two outcomes of the same loop.
Assumptions
- The routine involves several participants whose information, goals and interpretations differ, so the ostensive aspect is “not monolithic” and “distributed unevenly” (pp. 101, 104).
- Actions are interdependent, so each performance is collective and each participant’s options depend on others’ prior actions (pp. 104–105).
- Structuration (Giddens) and practice theory (Bourdieu) supply the frame: structure is reproduced by action and constrains it.
- The unit of analysis is a routine within a given organization; the argument is not made for routines copied between organizations.
Key results
The paper’s results are conceptual.
- Ostensive to performative (pp. 105–107). Three uses: guiding (a template that never fixes the details), accounting (legitimating or delegitimating actions after the fact), referring (a gloss that lets people invoke “hiring” without knowing its details).
- Performative to ostensive (pp. 107–109). Performances create the routine (a pattern must repeat and be recognized as one), maintain it (an unperformed routine decays, like an old DOS manual or a medieval recipe) and modify it (variations may or may not be retained).
- Endogenous change (pp. 112–114). Campbell’s variation and selective retention applied inside one routine: performances are the variations, interpretation into the ostensive aspect is the selection, and the story of “how we do hiring” is the retention. The authors disclaim adaptiveness: the claim is only that “a given routine, within a given organization, has the inherent, endogenous capacity to generate and retain novel patterns of action”.
- Against the genetic reading (p. 114). “If the gene is changing itself, then the information it carries is not stable and unchanging, as required in ecological theory”; whether this matters for organizational ecology “is an empirical question”.
- Method (pp. 111–112). Survey measures of task variety and observed action sequences gave opposite rankings in Pentland’s bank study; the paper reads the first as measuring the ostensive and the second the performative.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | The core definition of a routine already entails a split into ostensive and performative aspects | moderate | argument from the definition’s four features, pp. 103–105 |
| C2 | Performances create, maintain and modify the ostensive aspect; the ostensive guides, accounts for and refers to performances | moderate | conceptual, illustrated by hiring |
| C3 | Routines can change endogenously, by variation and selective retention within them | moderate | conceptual; cites field studies (Feldman 2000, Pentland 2003) for the observation |
| C4 | Routines are poor carriers of unchanging genealogical information | weak | argued, explicitly left as an empirical question |
| C5 | Power shapes which variations are retained | weak | argued from Braverman and Crozier, no evidence of its own |
Concepts
- Organizational routine: “a repetitive, recognizable pattern of interdependent actions, involving multiple actors”.
- Ostensive aspect: “the ideal or schematic form of a routine … the routine in principle”, including participants’ divergent understandings and any artifacts (procedures, forms, past ads).
- Performative aspect: “specific actions, by specific people, in specific places and times … the routine in practice”.
- Endogenous change: change in a routine that results “from engagement in the routine itself”.
Connections
The paper sets itself against Nelson and Winter’s routines-as-genes and against Hannan and Freeman’s structural inertia, and inside Giddens’s duality of structure. Marquis and Tilcsik (LIT-1144, NOTE-1373) review imprinting, the opposite emphasis: features fixed in a brief sensitive period that persist. They list routines among the “building blocks” that can be imprinted, which this paper would expect to drift through performance unless performance keeps re-enacting the imprint.
Bearing on the record
- CLAIM-329. A6’s question for this paper was whether it separates “the continuity of a particular organization from reproduction of an organizational kind”. It does not. Its whole analysis is of one routine in one organization, and its persistence story is maintenance: an ostensive aspect survives only while it is performed, and fades when it is not. That is a form of the claim’s historical persistence: what carries the routine across a change of performers is the chain of performances and the shared understanding they keep alive, not similarity to a type. The paper’s one move toward the kind/individual question is negative: it doubts routines can serve ecology as unchanging genealogical carriers between organizations (p. 114), and leaves the matter as empirical. So the literature A6 hoped would already hold the claim’s distinction does not hold it here; the claim’s distinction remains the record’s own.
- QUESTION-118. The paper offers one candidate for “what must stay causally continuous”: the ostensive aspect as maintained by ongoing performance and by those with power to say which performances count. It does not discuss total turnover of performers, so it does not test whether a routine survives when every participant changes.
- No instruction for machine-learning practice; nothing for the anthology.
Limitations
- No new data; the hiring example is illustrative and hypothetical (the snowstorm videoconference).
- “Ostensive” is defined to include participants’ subjective, divergent understandings and also artifacts, so it is not one observable; the paper gives no measure of it beyond the survey versus observation contrast.
- The claim that endogenous change happens is supported by citation; the paper does not say how often or how fast, or when a routine resists it.
- Slips. The text cites Emirbayer and Mische (1998) as “Emirbayers and Mische” throughout, including in the reference list, and spells Narduzzo “Naduzzo” twice and correctly once (pp. 95, 100, 102), as the OCR reads the page images.
Open questions
- Does a routine survive a complete change of its performers, and if so what carries it: artifacts, overlapping tenure, or the people who certify performances?
- When the same routine appears in an unrelated organization, is that the routine persisting or a new one of the same kind? The paper’s doubt about genealogical transmission suggests it would say new, but it does not say.