Contribution
Sewell keeps Giddens’s thesis that structures are both the medium and the outcome of practice, and repairs the concept so that it can carry agency and change. He replaces “rules” with “schemas”, argues that Giddens’s “virtual” structures cannot include resources, which are actual, and defines structure as schemas and resources that “mutually imply and sustain each other over time” (p. 13). Against Bourdieu’s habitus, which he finds reproduces itself with no room for change, he gives five axioms under which the ordinary working of structures generates their own transformation.
Key insight
Knowing a schema is being able to apply it to a case it was not learned on, and that ability is agency. “Knowledge of a rule or a schema by definition means the ability to transpose or extend it-that is, to apply it creatively. If this is so, then agency, which I would define as entailing the capacity to transpose and extend schemas to new contexts, is inherent in the knowledge of cultural schemas that characterizes all minimally competent members of society” (p. 18). Since the results of a transposition are not predictable, and schemas live only if their results validate them, reproduction is never automatic.
Assumptions
- Schemas are virtual. “they cannot be reduced to their existence in any particular practice or any particular location in space and time” (p. 8). Codified rules (statutes, contracts) are resources, not schemas.
- Resources are actual, human and nonhuman, and “media of power” unevenly distributed, though “some measure of both human and nonhuman resources are controlled by all members of society” (pp. 9–10).
- Mutual dependence. “Schemas not empowered or regenerated by resources would eventually be abandoned and forgotten, just as resources without cultural schemas to direct their use would eventually dissipate and decay” (p. 13).
- Society is plural. The change axioms require “a far more multiple, contingent, and fractured conception of society-and of structure” (p. 16).
Key results
No data. By section:
- Three problems with “structure” (pp. 2–3): it implies rigid determinism, makes change awkward, and means opposite things in sociology (hard, material) and anthropology (culture).
- Rules to schemas (pp. 6–8). Giddens’s rules, read as “generalizable procedures”, become “the informal and not always conscious schemas, metaphors, or assumptions presupposed by such formal statements” (p. 8), at depths from Lévi-Strauss’s oppositions to etiquette.
- Resources (pp. 9–12). Giddens’s definition rendered as “resources are anything that can serve as a source of power in social interactions” (p. 9). Resources take their value from schemas: Hudson Bay blankets are warmth, or, in a potlatch, the chief’s power (pp. 11–12).
- Duality kept (p. 13). “Structure, then, should be defined as composed simultaneously of schemas, which are virtual, and of resources, which are actual.” A factory’s gate and punch clock “teach and validate the rules of the capitalist labor contract”; “Resources, we might say, are read like texts”.
- Against habitus (pp. 14–16). Bourdieu’s habitus “retains precisely the agent-proof quality that the concept of the duality of structure is supposed to overcome” (p. 15); Kabylia “produced a momentous anticolonial revolution shortly after Bourdieu returned to France” (p. 15); what misleads is his “unrealistically unified and totalized concept of habitus” (p. 16).
- Five axioms of change (pp. 16–19). The multiplicity of structures (never all homologous); the transposability of schemas; the unpredictability of resource accumulation (schemas “will in fact be differentially validated when they are put into action”, p. 18); the polysemy of resources (“Any array of resources is capable of being interpreted in varying ways and, therefore, of empowering different actors and teaching different schemas”, p. 19); and the intersection of structures. “Structures are at risk, at least to some extent, in all of the social encounters they shape” (p. 19).
- Agency (pp. 20–21). “Agency is implied by the existence of structures” (p. 20). It differs “in both kind and extent” with position (p. 20). It is “collective as well as individual”, but Sewell rejects applying “agent” to corporate units in the same sense as to persons (p. 21).
- Depth and power (pp. 22–26). Structures vary in depth (schemas: how pervasive and taken for granted) and power (resources). Language: very deep, little power, very durable. States: usually high power and low depth, so “natural targets for open struggles” (p. 24); the United States “has had a single constitution since 1789” yet “at least five sharply distinct party systems” (p. 24). “Durability, then, would appear to be determined more by a structure’s depth than by its power” (p. 24). Capitalism combines both: the commodity form is “exceptionally transposable”, and its surface instability “actually reinforces its deeper structures” (pp. 25–26).
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Giddens’s structure is incoherent, since virtual structures cannot include actual resources | strong as an argument about Giddens’s definition | textual analysis (pp. 9–11) |
| C2 | Structures are mutually sustaining virtual schemas and actual resources | interpretive: a definition, argued from examples | factory, Mass, potlatch (pp. 11–13) |
| C3 | Bourdieu’s habitus, as stated in Outline, cannot generate change from within | moderate as a reading of Outline (1977) | quotation and the Algerian revolution (pp. 15–16) |
| C4 | Multiplicity, transposability, unpredictable accumulation, polysemy and intersection let ordinary enactment transform structures | interpretive; illustrated, not shown | examples throughout (pp. 16–19) |
| C5 | Agency is the capacity to transpose schemas and remobilize resources, inherent in all competent members, varying by position | interpretive | Goffman cited (p. 20) |
| C6 | A structure’s durability depends more on its depth than on its power | weak: a conjecture | language, states, capitalism as illustrations (pp. 22–26) |
Concepts
- Schema: a virtual, transposable procedure or assumption, “not formally stated prescriptions” (p. 8).
- Resource: anything that can serve as a source of power, human or nonhuman, always actual; its value is set by schemas.
- Structure: “sets of mutually sustaining schemas and resources that empower and constrain social action and that tend to be reproduced by that social action” (p. 19).
- Transposition: applying a schema to a new case, “but in such a way that the rule will have subtly different forms in each of its applications” (n. 9, p. 17).
- Depth and power: the schema and the resource dimensions along which structures vary (p. 22).
Connections
It works from Giddens (1976–1984) and Bourdieu’s Outline (1977), with Sahlins on how categories are “burdened with the world” (n. 11) and Goffman for everyday agency. Swidler (LIT-1367) is thanked and her “tool kit” is close to his schemas, but she is not cited. Wacquant (LIT-1377, 2016) answers the later version of this critique (Sewell 2005, ch. 4) by saying habitus needs no supplement, without engaging the argument. Sewell is right about Outline’s sentence (“all the actions … and no others”); Wacquant’s habitus, layered and able to misfire, is a later reading the 1992 essay does not meet.
Bearing on the record
- CLAIM-417 (selves as organizational perspectives). The LIT says the exchange cited this as an account of how structures are reproduced and transformed through activity. That is accurate. Three points bear on the claim. First, Sewell’s agency is the claim’s carrier with a capacity, not a perspective: agents are persons, and structures are multiple and intersecting in each, “a wide range of different and even incompatible schemas” (p. 17), so one carrier enacting several structures is his normal case. Second, he says what the claim also says, that collective agency is real without collectivities being agents in the same sense as persons (p. 21); that is the claim’s “Not that perspectives are agents”, reached from structure. Third, the defeater that established accounts “explain the same transitions as well with fewer assumptions” bites hardest here: transposition and remobilization are a theory of how one organization changes into another, which is CLAIM-417’s “transformation of organizational perspectives”. What Sewell lacks is the claim’s distinction between a lineage and a character; “transformations of the old” (p. 27) are recognized, he says, but he gives no criterion for when a transformed structure is the same one.
- CLAIM-416 (competition over tests). The polysemy of resources is the claim’s reflexive part stated in general. Sewell’s example: the gains of a successful cavalry charge “may be attributed to the superior discipline and elan of the cavalry officers and thereby enhance the power of an aristocratic officer corps, or it may be attributed to the commanding general and thereby result in the increasing subordination of officers to a charismatic leader” (p. 19). What a success shows is itself contested, and the reading that wins redistributes power. And unpredictable accumulation says success validates the schema that produced it (“A brilliantly successful cavalry attack on a new terrain may change the battle plans of subsequent campaigns or even theories of military tactics”, p. 18). The claim could cite this as the general mechanism of which competition over tests is a case.
- CLAIM-423 (timescale and authority in a political collective). Sewell offers a rival account of the same durabilities. Where the claim puts timescale, Sewell puts depth: state structures change because they are powerful and visible, and deep political structures (the American constitutional system, the French bureaucracy, English community law) last because they are taken for granted as power-neutral (p. 24). His “single constitution” under “five sharply distinct party systems” is the claim’s “a state can stay continuous as an actor while the organization producing its positions changes under it”, stated in 1992. The claim’s defeater asks whether existing accounts already state it; Sewell states the continuity, not the link between speed and authority.
- CASE-133 (party realignment). The same sentence: party systems change sharply under a continuous constitution (p. 24, citing Burnham 1967). It supports the case’s carrier-and-character distinction for the state, not for the parties.
- TERM-072 (perspective). A Sewell structure is not a perspective in any of the term’s senses: it is a schema-resource set, virtual on one side and material on the other, and he denies that schemas alone are structure. The nearest object in the term’s vocabulary note is a kind (the schema, transposable across carriers) coupled to a realization (the resources in one place). His transposition is how a kind acquires a new realization.
- THEORY warranted? None filed. C1–C5 are a conceptual framework and C6 is a conjecture with three illustrations; neither is a finding the record can hold as a claim about the world.
- No instruction for machine-learning practice; nothing for the anthology.
Limitations
- Conceptual throughout. The examples (factories, the Mass, potlatch, cavalry charges) show the concepts can describe cases; nothing shows they predict any.
- “Depth” is defined by pervasiveness and taken-for-grantedness, but no measure is offered, and C6’s comparisons (Britain and France, Costa Rica and its neighbours) are listed, not analysed.
- The critique of Bourdieu rests on Outline (1977); later Bourdieu and Wacquant’s restatement (LIT-1377) answer some of it.
- Transposition makes every competent person an agent; the essay does not say what would count as a schema that failed to transpose because of the schema rather than the situation.
Open questions
- When is a transformed structure the same structure transformed, and when a new one? Sewell says transformations are “recognizable as transformations of the old” (p. 27) and does not say by what.
- Can depth be measured independently of durability, so that C6 is not circular?
- In the polysemy of resources, what decides which interpretation of a success prevails? The cavalry example names two readings and no mechanism for choosing between them.