Marriage from a Dialectical Generative Relational Economics Perspective - A Preliminary Discussion of the Relation Between Allocative and Generative Accounts 【（Preliminary）Draft】

Posted on 2026-08-01 In Economics , Philosophy Views: Disqus: Word count in article: 12k Reading time ≈ 44 mins.

Abstract

Discussion of marriage, in ordinary evaluation and in a large part of family economics, proceeds from an allocative question: what each party holds, what each party supplies, and how holdings and duties are to be arranged between them. This paper states a second question that the first leaves open, namely how the coupling of two parties alters the set of actions, self-descriptions and relations available to each of them over time. The relation between the two questions is treated dialectically. The allocative account is reconstructed as the adequate abstraction of determinate material conditions, principally scarcity, the bearing of marital risk by kin, and the unobservability of future capacity; the generative account is treated as the question that becomes formulable once those conditions shift and once dynamical instruments exist to state it. Five propositions are advanced: coupling changes feasible action sets in ways that conserved-sum representations misdescribe; endowments stand as outputs of a relational trajectory as well as inputs to it; generative change carries no evaluative sign, since expansion, constriction and redirection are generative in the same sense and the mode is defined for each party separately; allocative indicators are verifiable in advance while generative capacity is verifiable in retrospect, which grounds both the rationality of external evaluation by senior kin and the standing risk that generative language serves to defer distributive claims; and distributive guarantees stand among the conditions under which the expansion mode is available. The paper closes with an admissibility condition in constraint form: a generative description of a relation holds for evaluative purposes where the party bearing the cost of the present allocation retains the standing to contest it and a mechanism through which contestation can alter it. The condition places distributive institutions among the enabling conditions of generation, and it blocks the aggregative reading under which one participant’s expansion offsets another’s constriction.

Keywords: marriage; family economics; generative relational economics; distribution; capability; care labour; coercive control; dialectical materialism.

A note on the standing of this paper. This paper states a position and does not establish a theory. It supplies no measurement procedure for the quantities it names, no estimated model, and no ranking of relations. Section 3 concedes at length that the observation that relationships produce what their parties lack at the outset is thoroughly anticipated in household production theory, in the capability approach, in feminist economics of care, in relational sociology, and in social psychology; the residue claimed here is correspondingly narrow and is stated as Claim 3.1. The propositions of Section 4 are formulated so that they can be shown false, and Section 8 states what would show each of them false. Objection, correction, and counter-evidence are welcome at huangwanhong@serendip.ngo.

1. Introduction

This section states the question the paper addresses, situates that question in the development of economic analysis, fixes the two accounts whose relation is at issue, states the contribution claimed together with the limits of the claim, and describes the method and the order of what follows. The material of the section is expository throughout, and every substantive statement made here is established later at the place indicated.

The occasion of this paper is an observation about evaluative vocabulary. Where a marriage is assessed in advance, whether by the parties, by their families, or by a body of social scientific work, the assessment proceeds through a characteristic inventory: earnings, education, employment stability, housing, health, domestic competence, temperament, and the division of household and care duties the pair is expected to adopt. The inventory has an implicit structure. Each item names something one party holds at the moment of assessment, and the assessment consists in arranging the holdings of both parties into a workable configuration. The question so posed is a question of arrangement, and it presupposes that what is to be arranged is already in existence.

The inventory is applied with particular consistency by senior kin, and the reason is stated in §2: a family that underwrites the risk of a marriage requires indicators it can verify before the commitment is made, and the items of the inventory are indicators of that kind. The consistency is therefore a feature of the evaluator’s position rather than a defect of the evaluator’s perception, and any account that treats it as a failure of vision will misdescribe both the evaluator and the disagreement.

A second question receives markedly less attention. Over the decades a marriage occupies, the parties acquire capacities, expectations, contacts, tolerances and descriptions of themselves that neither of them held at the outset, and some of what they acquire is attributable to the coupling. Consider a pairing in which one party holds few material resources at formation. Under the inventory the contribution of that party is low, and the assessment is correspondingly unfavourable. A different description of the same pairing is available: the conduct of the better endowed party alters the other’s expectations of the world, which alters what the other attempts, which alters what the other obtains, until both occupy a position that neither held at formation. The status of that description is the subject of this paper, and §6 states the three qualifications under which it may be advanced. What it establishes here is that the inventory leaves a question open, and that the open question concerns the production of the very quantities the inventory takes as given.

The structure of the inventory has a counterpart in the constitution of economics as a discipline, and the parallel is worth stating because it identifies the form at issue as a general one rather than a peculiarity of thinking about marriage.

Classical political economy took the factors of production as given and made their apportionment the central problem. Ricardo opened his Principles by naming the determination of the laws regulating the distribution of the produce of the earth among landlords, capitalists and labourers as the principal problem of the discipline (Ricardo, 1817). The marginalist apparatus that followed refined the treatment of apportionment and retained its form, and Robbins gave the form its canonical statement by defining economics as the study of conduct under scarce means that have alternative uses (Robbins, 1932). Under that definition allocation is constitutive of the subject: where the means are given and the uses compete, the analytical problem is the assignment of the former to the latter.

Accounts of the production of the means arrived later and, for a considerable period, from the margins of the discipline. Schumpeter placed the recombination of existing elements at the centre of economic change (Schumpeter, 1934), and the evolutionary tradition that developed his orientation replaced equilibrium optimisation with routines, search and selection (Nelson & Winter, 1982); Arrow made productivity a function of cumulative activity (Arrow, 1962); Hirschman argued that development proceeds by eliciting capacities that were latent and unrecognised (Hirschman, 1958; Hirschman, 1967); Ostrom established that commons governance sustains the conditions of continued joint production (Ostrom, 1990); Arthur assembled the treatment of increasing returns and path dependence (Arthur, 1994); and endogenous growth theory made the accumulation of nonrival ideas the engine of sustained growth (Romer, 1986; Romer, 1990). The movement is from the apportionment of a stock toward the production of the conditions under which a stock arises.

Two remarks fix the standing of this history for present purposes.

The first concerns instruments. The apparatus for treating systems whose components alter one another over time, which includes nonlinear dynamics, the analysis of feedback and of self-organisation, and the formal treatment of increasing returns, matured after the allocative form of economic analysis was fixed. A discipline states the problems its instruments allow it to state, so the sequence observed is in part a sequence of available formal means. It is also, and more fundamentally, a sequence of material conditions, and §2 gives that account for the case of marriage.

The second concerns priority, and it is a concession made here so that it is not extracted later. The question of the production of human powers is present in the classical tradition itself. Marx treats the development of human capacities as an end in the Grundrisse and elsewhere, and the passages are among the most cited in his corpus (Marx, 1857). What the present paper claims is a determinate formulation of the relation between the two questions in the analysis of intimate relations, and no claim is made that the second question is unprecedented.

Family economics inherited the allocative form and gave it a powerful development, which §2 reconstructs, and the bodies of work that have already stated the second question in their own terms are surveyed in §3.

Two accounts are distinguished throughout the paper, and the distinction is drawn at the level of the question each poses rather than at the level of the answers each returns.

Definition 1.1. (Allocative account). An allocative account of a relation takes the endowments, capacities and obligations of the parties as given at the moment of analysis, and treats the analytical problem as the arrangement of those given quantities between the parties and across the tasks the relation requires.

Definition 1.2. (Generative account). A generative account of a relation takes as its object the change, over an interval, in the set of actions, self-descriptions and relations available to each party, where that change is attributable to the coupling; endowments enter such an account both as initial conditions and as outputs of the trajectory.

Claim 1.3. (Central position). The two accounts answer distinct questions, stand in a relation of mutual dependence, and admit of a determinate ordering of dependence in one direction: distributive guarantees function among the conditions under which generative descriptions of a relation are true, and generative processes continually reconstitute the objects over which distribution operates.

Claim 1.3 states the position the paper defends, and its two halves are established separately. The reconstitution half is the content of Propositions 4.1 and 4.2. The dependence half is the content of Proposition 4.6, and it is the half that carries the dialectical commitment announced in the title, since it holds the generative account to a condition supplied by the account it is tempted to supersede.

Figure 1 sets the two forms of the question beside one another. They share their material and differ in the direction of one arrow: the endowments that the first takes as data of the problem appear in the second as an output of the process as well as an input to it.

Figure. Two forms of the economic question. Under (a) the endowments are data of the problem and the analysis terminates in an arrangement of them. Under (b) the same endowments enter a coupling whose dynamics alter the reachable sets of the parties, and the altered sets stand as the endowments of the following interval. The difference between the panels is the return arrow, and Proposition 4.2 is the statement of that arrow. Neither panel carries an evaluative sign, since the return may enlarge a reachable set, narrow it, or reorganise it, as Figure 2 states.

The paper is a position paper, and three things are claimed for it.

The first is a joint statement of the two accounts with an explicit ordering of dependence between them, in preference to a treatment of generation as an amendment to allocation or as its successor. The second is a valence-free definition of generative change, under which expansion, constriction and redirection are generative in the same sense and the mode is determined for each party separately; the definition severs the vocabulary of generation from its optimistic reading, and it is what allows the framework to describe a coercive relation without embarrassment. The third is an admissibility condition on generative descriptions, stated in constraint form, which gives the position operative content in the cases where the two accounts recommend opposite conclusions.

The limits of the claim are correspondingly definite. The paper supplies no measurement procedure for the quantities it names, no estimated model, and no ranking of relations; the position rules descriptions inadmissible and ranks nothing. It establishes no distributional claim about how often relations of any described kind occur, since its cases are illustrative. And it claims no priority for the observation that relationships produce what their parties lack at the outset, which §3 shows to be established in five separate literatures. Claim 3.1 states the residue by subtraction, and §9 states what remains unmeasured, unformalised and unresolved.

The method has three components. The reconstruction in §2 is historical and materialist: the allocative account is explained by the conditions that made it adequate, in preference to being assessed against a standard external to those conditions, and §7.4 applies the same treatment to the account this paper advances. The statement of position in §4 and §5 is conceptual and minimally formal, with an explicit account of how much formalisation the material warrants and an explicit refusal of notation that carries the appearance of derivation without its substance. The cases in §6 are illustrative and are labelled as such, since a position paper establishes the availability of a description and leaves its frequency to empirical work.

The order of what follows is: the historical constitution of the allocative account (§2); prior formulations of production within relationships, with the residual claim stated against them (§3); the five propositions (§4); the warranted formal register (§5); five worked cases (§6); the admissibility condition and the self-critique that motivates it (§7); consequences for evaluation and for empirical work, with the conditions of refutation (§8); the limits of the present statement (§9); and the standing of the dialectical formulation (§10).

One affiliation may be declared and then set aside. The paper belongs to a programme applying a generative relational ontology to the domain of value generation and flow, which this series designates GR-Eco, and it stands beside a study of luxury space in the same series (Huang, 2026). Nothing below depends on the wider framework, and a reader who declines it entirely may still hold that Definition 1.2 names an object and that Condition 7.3 is owed.

2. The Historical Constitution of the Allocative Account

This section reconstructs the conditions under which the allocative account of marriage became the adequate abstraction of its object. The objective is to establish that the account holds a determinate historical validity, so that the position of this paper can be stated as an extension of scope in preference to a correction of error. The method is a materialist reconstruction, working from the productive and risk-bearing functions of the household to the analytical form those functions imposed. Three components are treated in turn: the household as a unit of production and insurance, the theoretical form this received in family economics, and the position of the external evaluator.

2.1 The household as a unit of production and of insurance

Across the greater part of recorded agrarian and early industrial history, the household operated as a productive enterprise and as the primary institution of insurance. Subsistence margins were narrow, and the failure of a harvest, an illness, a death, or a lost season of labour carried consequences that had no institutional absorber beyond kin. Under such conditions the pairing of two adults performed identifiable functions: pooling of labour across tasks with distinct seasonal and physical profiles, pooling of risk across shocks that were partly independent, transmission of land and tools across generations, and the production and rearing of the labour on which the enterprise would depend within a decade. Engels engels1884 placed the historical transformations of the family in relation to property and its transmission, and the general point survives independently of the particular reconstruction he gave: the form of the marital tie tracked the productive and proprietary functions the household carried.

Where these functions dominate, the allocative question is the operative question. The quantities at issue, which were land, tools, stock, hands, hours, and claims on kin, existed prior to the marriage and were countable. Marriage arranged them. An analysis that took the endowments as given and asked after their configuration was addressing the feature of the situation on which survival turned.

2.2 The theoretical form given by family economics

The middle of the twentieth century supplied this configuration with a formal statement of considerable power. Becker modelled marriage as an assignment problem in which gains arise from specialisation and from the complementarity of the parties’ comparative advantages, with sorting on the marriage market determining who pairs with whom (Becker, 1973; Becker, 1981). The framework absorbed a substantial body of phenomena: assortative matching on education, the historical division of market and domestic work, the relation of divorce to the size of marital surplus (Becker et al., 1977), and the timing of family formation.

Subsequent work internalised the distribution question that the unitary model had set aside. Bargaining approaches modelled household allocation as the outcome of a bargain whose terms depend on each party’s fallback position (Manser & Brown, 1980; McElroy & Horney, 1981); the separate spheres model located the threat point inside a continuing marriage (Lundberg & Pollak, 1993); and Sen formulated the household as cooperative conflict, in which cooperation and divergence of interest coexist and in which perceptions of contribution shape the outcome (Sen, 1990). Empirical work established that the identity of the recipient of household income affects expenditure and child outcomes (Thomas, 1990; Duflo, 2003), which settles the question whether intra-household distribution is a real object of study.

Three features of this literature deserve emphasis, since the present paper depends on all three. The literature is analytically serious, and an account that treats it as a naive error will be worse than it is. Its distributional branch established that marriage contains conflict of interest as a permanent structural feature, which is a result the generative account has to accommodate. And the framework carries a determinate scope: it explains the arrangement of quantities, and it is at its weakest where the quantities themselves change as a consequence of the arrangement.

2.3 The position of the external evaluator

Ordinary evaluation of a prospective marriage by senior kin exhibits the allocative form in concentrated shape. The usual explanation, that elders observe resources while remaining unable to observe relational dynamics, is correct as far as it goes and stops short of the operative variable. The position of the elder is more precisely characterised by the risk it bears.

Historically, the kin group underwrote the marriage. It bore the cost of failure through the return of a daughter, the support of a widow, the care of children, the loss of a bride price or dowry, the maintenance of the pair through periods of dearth, and the standing of the lineage among its neighbours. A party who underwrites a risk demands indicators that can be verified before commitment. Earnings, land, employment, health, family reputation and domestic competence are precisely such indicators: observable at low cost, stable over the horizon of the decision, and correlated with the losses the underwriter would otherwise absorb. In the vocabulary of information economics these are screening devices operating under asymmetric information about an unobservable future variable (Spence, 1973), and they perform the function that collateral performs in credit markets under unobservable productivity (Stiglitz & Weiss, 1981).

Claim 2.1. (Rationality of external allocative criteria). Where a third party underwrites the risk of a relation whose future generative capacity is unobservable in advance, reliance on verifiable present endowments is a rational screening response to that unobservability, and it remains rational under a full acknowledgement that generative capacity exists and matters.

Two consequences follow, and the second supplies the material of §6.3. Screening on observables in the presence of substantial unobserved heterogeneity produces systematic misclassification with a determinate direction: pairs whose present endowments are low and whose generative capacity is high are the cases the screen rejects, and their rejection is invisible to the screener, since the counterfactual trajectory goes unobserved. And where the underwriting function migrates from kin to the state, to insurance markets and to independent individual income, the standing of the external evaluator changes while the criteria persist, so that a screening technology outlives the risk-bearing position that justified it. The persistence of a form beyond the conditions that produced it is a familiar figure in materialist analysis, and it applies here to the evaluative criteria of senior kin with no need for any imputation of failure to them.

3. Prior Formulations of Production Within Relationships

This section identifies the bodies of work that have already stated, in their own terms, that relationships produce what their parties lack at the outset. The objective is to fix by subtraction the residual claim available to this paper. The method is a survey organised by the discipline in which each formulation arose, followed by an explicit statement of what survives the subtraction. A position paper that presented relational production as a discovery would be making a false claim about a well populated field, and the credibility of the propositions in §4 depends on this section being generous.

3.1 Household production and human capital

Becker’s own framework contains a production theory. In the household production model the household combines purchased goods and time to produce the commodities that enter utility (Becker, 1965), so that the household is a producer by construction. Human capital theory makes capacities themselves the object of investment (Becker, 1964), and marriage-specific capital names capital whose value is contingent on the continuation of the particular relation (Becker et al., 1977). Any claim that mainstream family economics treats value as fixed in quantity is therefore false as stated. The feature to isolate is different: the production function has a fixed input set and a fixed output space, the technology is exogenous, and the accounting remains that of an allocation of given time and goods across given uses.

3.2 Care, its production, and its accounting

Feminist economics established that the labour performed within households produces value and that standard accounts systematically omitted it (Waring, 1988; Folbre, 1994; Himmelweit, 1995). Folbre’s analysis of care as producing capacities with public-good characteristics is a statement, in economic terms, of value generated within relations and unrecorded by the instruments that measure output (Folbre, 2001). This literature also supplies the strongest available reason for caution about generative vocabulary, since it documents at length the mechanism by which uncounted production becomes uncompensated production. Section 7 builds directly on it.

3.3 Capabilities and the shift from holdings to doings

The capability approach relocated the evaluative space from resources held to what persons are able to do and to be, and made the conversion between the two an explicit object of analysis, with conversion factors that are personal, social and environmental (Sen, 1985; Sen, 1999; Nussbaum, 2000). Much of the work that the phrase “the endowment is an output” is asked to perform in this paper has an established home there. Sen’s analysis of cooperative conflict within households (Sen, 1990) further anticipates the joint treatment of cooperation and divergent interest that §7 requires.

3.4 Relational sociology and the sociology of intimacy

Relational sociology proposed the transaction and the relation as the primary units of analysis in place of the substance and the attribute (Emirbayer, 1997; Donati, 2011; Dépelteau, 2008). Coleman treated family and community relations as capital productive of educational attainment (Coleman, 1988), and Bourdieu treated forms of capital and their conversion, with social capital defined by durable networks of relations (Bourdieu, 1986). On intimacy specifically, Giddens described the pure relationship as one sustained by what it delivers to the parties (Giddens, 1992), and Beck and Beck-Gernsheim analysed the transfer of biographical construction onto the couple (Beck & Beck-Gernsheim, 1995). Illouz supplies the critical counterpart, and her account of emotional capitalism bears directly on the self-critique of §7.4 (Illouz, 2007; Illouz, 2012).

3.5 Psychological work on coupling and capacity

Social and developmental psychology has documented the specific mechanism at issue. Attachment theory’s secure base construct holds that the availability of a responsive partner supports exploration away from the partner, with adult evidence linking secure base support to the pursuit of goals outside the relationship (Bowlby, 1988; Feeney, 2004). The self-expansion model treats the inclusion of the other in the self as a mechanism by which relationships enlarge the resources, perspectives and identities available to each party (Aron & Aron, 1986). The Michelangelo phenomenon documents partners sculpting one another toward their respective ideal selves, with movement toward the ideal self predicting personal and relational outcomes (Drigotas et al., 1999). Erikson established the term generativity in a distinct sense, concerning the adult’s care for what succeeds them (Erikson, 1950), and the usage of this paper requires explicit differentiation from his: the term is used here for change in what a coupling makes available to its parties, with no reference to succession and no favourable evaluative content.

3.6 Evolutionary economics, increasing returns, and generative justice

Outside the family, several literatures treat the production of the conditions of production. Endogenous growth theory made the accumulation of nonrival ideas the engine of sustained growth, with increasing returns arising from that nonrivalry (Romer, 1986; Romer, 1990); learning by doing made productivity a function of cumulative activity (Arrow, 1962); Schumpeter placed the recombination of existing elements at the centre of economic change (Schumpeter, 1934); Hirschman argued that development proceeds by eliciting latent and previously unrecognised capacities (Hirschman, 1958; Hirschman, 1967). Ostrom established that commons governance sustains the conditions of continued joint production (Ostrom, 1990). The evolutionary tradition carries the same orientation furthest: Veblen asked why economics had failed to become an evolutionary science (Veblen, 1898); Nelson and Winter built a theory of economic change on routines, search and selection in place of equilibrium optimisation, with the capacities of the units transformed by the process the theory describes (Nelson & Winter, 1982); and Hodgson traced what follows from taking the evolutionary metaphor seriously (Hodgson, 1993). Eglash formulates generative justice around the return of generated value to those who generate it (Eglash, 2016), and that formulation anticipates the direction of the criterion in §7.

The standing of this group in the present paper should be stated exactly, since the vocabulary is close enough for the relation to be misread. These traditions are drawn on for orientation, and none of them is treated here as a rival to be displaced or as a source of results the argument depends upon. What the evolutionary tradition supplies is the general form of an analysis in which the units and their capacities are transformed by the process under study, which is the form Definition 1.2 adopts for a two-party relation. What generative justice supplies is a criterion of return, which §7 develops in the narrower setting where the generating parties are two and the returning mechanism is a distributive institution. The debt in both cases is a direction of attention, and the claims of §4 stand or fall on the grounds given there.

3.7 The residue

Given the above, the following claims are unavailable to this paper: that relationships produce value; that capacities are endogenous; that resources omit much of what matters; that the household is a site of production; that intra-household distribution is contested. Each is established elsewhere and is cited here as such.

Claim 3.1. (Residual contribution). Three claims remain available. The first is the joint statement: the allocative and the generative account are formulated together, with an explicit ordering of dependence between them, in preference to a treatment of generation as an amendment to allocation. The second is the valence-free formulation: generative change is defined so that constriction and redirection are generative in exactly the sense expansion is, which severs the vocabulary from its optimistic reading. The third is the admissibility condition: a criterion is stated under which a generative description of a relation is licensed, which gives the position operative content in the cases where the two accounts diverge in their practical recommendation.

4. Statement of the Position

This section states the propositions that constitute the position. The objective is to give the position a form in which each element can be assessed and contested separately. The method is analytic: each proposition is stated, its ground is given, its scope is bounded, and the inference it blocks is identified where one is blocked. The five propositions concern, in order, the conservation assumption, the standing of endowments, the evaluative sign of generative change, the asymmetry of verification, and the ordering of dependence between the accounts. Table 1 summarises the comparison they establish.

4.1 The conservation assumption and its failure

Proposition 4.1. (Non-conservation). For a coupling of two parties over an interval, the set of actions, self-descriptions and relations available to each party at the end of the interval can contain elements available to neither party at the outset, and the availability of those elements can be attributable to the coupling; representations that hold a total constant across the parties therefore misdescribe the object over intervals of that length.

The ground of the proposition is the ground of increasing returns in growth theory. Where an output of a process enters as an input to the same process on a nonrival basis, an accounting that treats the process as the division of a fixed total gives the wrong answer, and it gives the wrong answer by a margin that grows with the horizon (Romer, 1990; Arrow, 1962). Within a couple, trust, procedural knowledge of the other, shared reference, jointly held networks and the confidence to attempt what would otherwise go unattempted all have this character: their use by one party leaves them available to the other, and their accumulation depends on the history of the pair.

The scope of the proposition is limited in two directions. Over short intervals and for physical goods the conservation description is accurate, and household allocation of a given month’s income across competing uses is a genuine division of a fixed total. And the proposition establishes the failure of an accounting assumption; it establishes nothing about the sign of the change, which Proposition 4.3 addresses.

4.2 The standing of endowments

Proposition 4.2. (Endogenous endowments). The endowment vector of each party is an output of the relational trajectory as well as an input to it; consequently the explanatory standing of endowments measured at formation declines with the age of the relation, and inference from formation endowments to later states requires the trajectory as an intermediate term.

The ground here is partly documented in §3.5 and partly a matter of ordinary observation over decades: earning capacity, health behaviour, social range, tolerance of risk, and the description under which a person understands their own situation are all revised in the course of a long relation. The proposition carries a methodological consequence for empirical work, since it identifies the standard cross-sectional design, which regresses outcomes on characteristics measured at formation, as one that attributes to selection whatever the trajectory produced.

4.3 The evaluative neutrality of generative change

Proposition 4.3. (Directional neutrality). Generative change carries no evaluative sign. Where a coupling alters the set of actions, self-descriptions and relations available to a party, three modes are available and all three are generative in the same sense: expansion, where the later set contains the earlier; constriction, where the earlier set contains the later; and redirection, where each set contains elements absent from the other. The mode is defined separately for each party, so that a single coupling can expand one party’s set while constricting the other’s.

This proposition carries the critical weight of the paper, and its ground is straightforward. A relation that reorganises what a person can do, say and become is doing so whether the reorganisation enlarges their range or closes it. The coercive relation is generative in the full sense of Definition 1.2: it produces a subject with novel capacities, including highly developed anticipation of the other’s states, and it produces them through the coupling mechanism that the benign case employs. Stark documents the pattern under the name coercive control (Stark, 2007), and the legal recognition of that pattern, in the offence of controlling or coercive behaviour under section 76 of the Serious Crime Act 2015 in England and Wales and in comparable course-of-conduct provisions elsewhere, records the constriction of the victim’s space for action as the harm, in preference to any single event within it.

Redirection deserves separate emphasis, since it is the mode most easily mistaken for its neighbours. Under redirection the volume of change can be large and the capacity to act can remain substantial, while the range within which the party can constitute themselves has been reorganised around a position the relation requires. A party may become highly capable as the one who manages the other’s moods, and correspondingly less able to occupy any description in which that management is absent. Measures of activity, of effort and even of skill acquisition register redirection as generation, which is one reason the vocabulary requires the discipline this proposition imposes. Figure 2 states the three modes.

Figure. The three modes of generative change, stated as relations between the reachable set of party $i$ at the beginning and at the end of an interval. The mode is defined for each party separately, so the state of a couple is a pair of modes and the diagram is to be read twice for any relation. Nothing in the figure carries an evaluative sign; the diagram fixes what is to be observed, and §7 states the separate conditions under which a claim about it is admissible.

Claim 4.4. (Blocked inference). From the premise that a relation is generative, no evaluative conclusion follows, and in particular no conclusion follows about what any party should bear. Generativity is a dynamical property of a coupling; evaluative standing arises from the conditions stated in §7, which concern contestability and the separate treatment of each party.

The inference blocked here is the one most readily available to the vocabulary this paper uses, and it is available in a superficially attractive form: the marriage generated a great deal, therefore the arrangement that produced it is vindicated. That inference aggregates across the parties, and Proposition 4.3 removes the ground for aggregation by defining the mode for each of them separately.

4.4 The asymmetry of verification

Proposition 4.5. (Verification asymmetry). Allocative indicators of a prospective relation are verifiable by third parties in advance of commitment; generative capacity is verifiable in retrospect and, for the parties themselves, is subject to interpretation controlled by the relation whose properties are in question.

The first half of the proposition grounds Claim 2.1 and explains the durability of allocative evaluation in ordinary life. The second half carries a warning about the position advanced here. A property verifiable only in retrospect supports narratives constructed after the fact, and the construction is available to whichever party benefits from it. Where the interpretation of a relation is asymmetrically controlled, the party who controls it can supply the description under which the relation appears generative, and the party bearing the cost can be left without the terms in which to state what has occurred. Fricker analyses the general form of this condition as testimonial and hermeneutical injustice (Fricker, 2007); its application to marriage is direct, and it is the reason §7 states its criterion in terms of standing and mechanism in preference to terms of correct description.

4.5 The ordering of dependence

Proposition 4.6. (Distributive precedence in the order of conditions). Distributive guarantees, understood as the fallback position, the exit option, the independent claim on income, and the recognised standing of the contribution each party makes, function among the conditions under which the expansion mode of Proposition 4.3 is available; where those guarantees are absent, coupling remains generative and its available modes narrow toward constriction and redirection.

The ground of this proposition is theoretical and empirical. The theoretical ground is the bargaining literature of §2.2: the terms on which a couple cooperates depend on the fallback position of each party, so a party with a closed exit negotiates on terms set elsewhere (Manser & Brown, 1980; Lundberg & Pollak, 1993; Sen, 1990). The empirical ground includes evidence that expansions of exit options changed outcomes inside continuing marriages: Stevenson and Wolfers report declines in domestic violence and in female suicide following the adoption of unilateral divorce provisions across the United States (Stevenson & Wolfers, 2006). The usual cautions about quasi-experimental estimates apply, including sensitivity to specification and to the timing of adoption, and the proposition accordingly rests on the direction of the finding in preference to its magnitude.

Proposition 4.6 supplies the dialectical content the title of this paper announces, and it does so in a determinate direction. The generative account holds that distribution operates on objects that generation has produced. Proposition 4.6 adds that generation of the expansive kind requires distributive guarantees already in place. The two accounts are therefore related as moments of one process, each supplying conditions for the other, and any reading under which the generative account supersedes the allocative one loses precisely the guarantee that makes generative descriptions true.

Table. The two accounts compared across seven dimensions. The rows are dimensions of analytical form and carry no ranking; the last row states the failure characteristic of each account when it is applied beyond its scope.

Dimension	Allocative account	Generative account
Primary unit	The party and the holding	The coupling and its trajectory
Standing of endowments	Given data of the problem	Input at formation and output over time
Object explained	Arrangement of quantities across parties and tasks	Change in the set of available actions and self-descriptions
Temporal orientation	The present configuration	The interval and its history
Mode of verification	Observable in advance by third parties	Verifiable in retrospect, interpretation contestable
Characteristic institution	Property regimes, maintenance, division on dissolution	Practices sustaining trust, repair and joint reference
Characteristic failure	Attribution to selection of what the trajectory produced	Deferral of distributive claims by appeal to what the relation generates

5. The Formal Register and Its Warranted Scope

This section fixes how much formalisation the material supports. The objective is to supply a minimal apparatus adequate to state Proposition 4.3 precisely, and to identify the notations that carry the appearance of derivation without its substance. The method is to decline three notations with reasons, to introduce one construction and state what it does, and to state where the construction connects to the formal work of this series and where it fails to.

5.1 Notations declined

Three notational forms appear in informal discussion of this topic and are declined here.

The first is the conserved sum, in which the value held by two parties is written as summing to a constant. The expression presupposes that the quantities in question admit of addition, which is the substantive claim at issue; adopting the notation settles that claim by stipulation. The point is general and has been made for the formal treatment of heterogeneous social states: an assumption of order presumes commensurability, and an assumption of a distinguished zero presumes that loss is a magnitude (Huang, 2026).

The second is the substitution of a relational term into a production function, in the form of an output written as a function of relation in place of capital and labour. The expression is an announcement, since the function is unspecified, its arguments are undefined, and nothing is derived from it.

The third is the identity statement, in which a slogan is written with an equals sign. Formal apparatus earns its place by supporting inference, and apparatus that supports none reduces the credibility of the formal work carried out elsewhere in the same series.

5.2 The reachable set and the three modes

For a party $i$ and a time $t$, let $A_i(t)$ denote the reachable set: the collection of actions, self-descriptions and relations available to $i$ at $t$. The set is introduced as a bookkeeping device carrying no metric, no ordering and no distinguished element, in keeping with the treatment of ontologically heterogeneous social states developed for this series (Huang, 2026).

A coupling $C$ is generative for party $i$ over the interval $[t_0,t_1]$ where the symmetric difference $A_i(t_0),\triangle,A_i(t_1)$ is nonempty and the difference is attributable to $C$. Three modes follow from the inclusion relations:

expansion, where $A_i(t_0) \subset A_i(t_1)$;
constriction, where $A_i(t_1) \subset A_i(t_0)$;
redirection, where $A_i(t_0) \setminus A_i(t_1)$ and $A_i(t_1) \setminus A_i(t_0)$ are both nonempty.

Three features of this construction do the work the paper requires. The mode is defined for each party separately, so the state of a couple is a pair of modes and the nine resulting combinations include the cases in which one party expands as the other constricts. The definition is stated in set relations alone, so it commits the paper to no measure of how much was generated and permits no summation across parties, which is the formal counterpart of Claim 4.4. And the allocative account appears within this apparatus as the projection of $A_i(t)$ onto its resource coordinates, which locates the relation between the accounts as one of projection in preference to one of contradiction.

5.3 Connection to the formal series, and the open problem

The construction connects to the formal work of this series at one point and fails to connect at another, and both should be stated. Constriction has a formal home: it corresponds to the collapse of distinctions under a mechanism, which the metamodel treats as an irreversible and cumulative feature of social dynamics (Huang, 2026). Expansion has no such home, since that treatment leaves the generation of novelty unformalised by explicit declaration. The set-theoretic language above is therefore provisional in the expansion case, and it stands as notation for a phenomenon whose formal treatment remains open. Stating this plainly is preferable to concealing it behind apparatus, and it identifies where further formal work on the object of this paper would have to begin.

6. Marriage under the Two Accounts

This section applies the propositions of §4 to four cases. The objective is to show what each account registers and what it omits in situations where their descriptions diverge. The method is illustrative: each case establishes that a description is available and that it is missed by the alternative account, and no case supports a claim about how frequently such situations arise. The cases are ordered so that the first two display the reach of the generative account and the last two constrain it.

6.1 Support and the expansion of the reachable set

Consider a party who, over several years within a marriage, changes occupation, undertakes training, or begins work that had previously appeared beyond reach, and who attributes the change to the presence of a partner whose responses made the attempt survivable.

The allocative account registers this as a transfer: the partner supplied support and possibly income during the transition, and the recipient supplied whatever the reciprocal arrangement specified. The description is accurate and incomplete, since it accounts for the flow and omits the change in the reachable set. What altered was the class of attempts that could be made at all, which is a change in $A_i$ and appears in the allocative account later and indirectly, through the earnings that eventually result.

Section 3 shows that this case is documented, and the honest statement of what the generative account adds is therefore modest: it makes the change in the reachable set the primary object of analysis in the economic register, where the psychological literature has kept it in the psychological register (Feeney, 2004; Aron & Aron, 1986) and the economic register has recorded its downstream resource trace.

6.2 Asymmetric initial endowment, with its limiting conditions stated

Consider a pairing in which one party holds substantially fewer material resources at formation. The allocative account has a clear reading: the contribution of that party is low, the terms of the arrangement should reflect it, and the pairing is questionable. A generative reading runs otherwise: the conduct of the better endowed party alters the other’s expectations of the world, which alters conduct, which alters outcomes, until both parties occupy a position that neither held at the outset.

Three qualifications are required, and their absence would leave this case doing more harm than good to the position.

The case establishes an existence claim. Such trajectories occur, and the allocative screen of §2.3 misclassifies them; that is the whole of what the case shows. A reading of it as a general recommendation to disregard resource asymmetry at formation converts an existence claim into a distributional claim, and the distributional claim is unsupported.

The case is exposed to survivorship. Narratives of the ascending spiral are collected from the relations that produced one, and the coupling structure that produces the ascending spiral produces the descending one, in which a withdrawal of confidence reduces attempts, which reduces outcomes, which reduces confidence further. A theory that predicts the ascending case and remains silent on the descending case has predicted nothing, and Proposition 4.3 exists so that both are available to one apparatus.

The case runs on a mechanism available to the interested party. A party who benefits from the present arrangement can supply the ascending narrative as a description of a relation that is in fact constricting for the other, and Proposition 4.5 identifies why such a description is difficult to contest from within. Section 7 addresses this directly.

6.3 The evaluative position of senior kin

Consider a family that assesses a prospective marriage by the criteria catalogued in §1: earnings, employment, housing, health, domestic competence, standing.

Section 2 establishes that these criteria are a screening technology appropriate to a risk-bearing position under unobservable future capacity, and this reading yields a more useful account of the resulting disagreements than one attributing the criteria to limited perception. Three consequences follow. The criteria are appropriate in the measure that the risk is still borne by the family, and the migration of that risk to the state, to the labour market and to independent income reduces their standing in the same measure. The screen misclassifies in a determinate direction, and its errors are unobservable to the screener, since the trajectory of a rejected pairing goes unobserved. And the screen has a genuine claim that the generative account should concede: under Proposition 4.5 the parties themselves have no verified access to their own generative capacity at the moment of decision, so the confidence of the couple is weak evidence, and the senior party is entitled to say so.

The disagreement is therefore between a verifiable criterion with a known misclassification profile and an unverifiable claim with a known rationalisation profile. Stating it in that form serves both parties better than an argument in which one side holds the concept of dynamics and the other holds the facts.

6.4 The deferred claim on jointly produced capacity

Consider the case in which one party supports the other through an extended period of training, with income, domestic labour and the deferral of their own occupational development, and the relation dissolves after the training has begun to yield returns.

The generative description is accurate: the qualification is an output of the coupling, produced by a trajectory in which both parties participated, and the individual who holds it would hold less of it alone. That description settles nothing about who holds the asset, and the case shows why the allocative question survives the generative account intact. The asset sits in one person, is inseparable from that person, and continues to yield.

Legal systems have addressed this in allocative terms and have reached varying answers, which is itself informative. The New York Court of Appeals in O’Brien v. O’Brien (1985) treated a professional licence obtained during the marriage as marital property subject to distribution; the Supreme Court of New Jersey in Mahoney v. Mahoney (1982) declined the property characterisation and awarded reimbursement to the supporting spouse; regimes of deferred community and of statutory division on dissolution, including the division-of-property provision of Japanese family law and the accrued-gains regime in German law, reach the question by other routes. The variation shows that the question is contested; the universality of the question shows that it has to be answered.

Claim 6.1. (Persistence of the allocative question). Where a jointly generated capacity is embodied in one party and continues to yield after the relation ends, the generative description of its production leaves the question of entitlement open, and the question is answerable in allocative terms.

Claim 6.1 is the strongest available check on the position of this paper, and it is stated here in preference to being conceded under pressure. A framework built on generation is under standing temptation to treat the question of entitlement as a symptom of the superseded paradigm. The temptation should be refused, since the party who bears the cost of leaving the question unanswered is identifiable in advance and is generally the party who supplied the support.

6.5 Coercion as a generative relation

Consider a relation in which one party’s range of action, contact and self-description narrows progressively over years, through a pattern of monitoring, isolation, and the regulation of ordinary conduct, with no single event of a kind that the criminal law of an earlier period would register.

This case is generative under Definition 1.2 in the full sense. The coupling alters the reachable set of both parties; it produces novel capacities in the constrained party, including highly developed anticipation of the other’s states; and it produces novel self-descriptions. Its mode, for the constrained party, is constriction accompanied by redirection: the party can become the one who manages, and progressively less able to occupy descriptions in which the managing is absent.

Two conclusions follow. Any account of intimate relations that treats generativity as a favourable property is refuted by this case, which is why Proposition 4.3 is stated in the form it takes. And the case identifies where the evaluative content of the position has to come from, since it comes from neither the presence nor the quantity of generation. Stark’s analysis and its legislative uptake locate the harm in the constriction of the space for action (Stark, 2007), and the criterion of §7 follows the same route by attending to the standing of the constrained party in preference to the properties of the relation as a whole.

7. Conditions on the Admissibility of Generative Claims

This section states the condition under which a generative description of a relation is admissible for evaluative purposes. The objective is to give the position operative content in cases where the two accounts recommend opposite conclusions, and to close the ideological function that the generative vocabulary otherwise performs. The method is to state the hazard, then the wall between explanation and warrant, then the condition, then the application of the paper’s own historical method to itself. The section closes with what the condition leaves undone.

7.1 The hazard specific to generative vocabulary

Generative description carries a characteristic practical implication: present arrangements are to be assessed by what the relation will have produced. The implication converts a present distributive claim into a matter to be settled later, and the conversion is available to whichever party benefits from deferral.

The mechanism is documented. Section 3 records that care labour was long described as an activity whose value lay in what it sustained, and that the description accompanied its exclusion from compensation and from measurement (Waring, 1988; Folbre, 1994; Himmelweit, 1995). The structure repeats wherever contribution is described in generative terms while entitlement is settled in allocative ones. Proposition 4.5 explains why the party bearing the cost has difficulty contesting the description: the generative claim concerns a trajectory verifiable in retrospect, and its interpretation is asymmetrically controlled inside the relation.

Claim 7.1. (Ideological availability). A generative description of a relation is available as a justification for deferring the distributive claims of the party bearing the greater present cost, and the availability increases with the asymmetry of interpretive control within the relation.

7.2 The wall between explanation and warrant

Claim 7.2. (Explanation and warrant). From an accurate account of what a relational configuration generates, no warrant follows for the configuration and none for any distribution of its costs.

The claim states a wall that a generative framework has particular reason to maintain, since it uses one vocabulary for its descriptive and its evaluative work, and the two therefore slide together with little friction. The prohibited sentence takes the form: this arrangement, whatever its cost to one party, sustains the generative capacity of the relation, and it is therefore warranted. The sentence is available in the framework’s own terms, and it can be true in its descriptive part while supplying nothing in its evaluative part. The corresponding prohibition on aggregation follows: the expansion of one party’s reachable set is unavailable as an offset against the constriction of the other’s, since Proposition 4.3 defines the mode for each of them separately and supplies no operation for combining the two.

7.3 The admissibility condition

Condition 7.3. (Admissibility of a generative description). A generative description of a relation is admissible for evaluative purposes where two conditions hold jointly: the party bearing the cost of the present allocation retains the standing to contest the description and the allocation, in terms of their own; and the contestation can reach a mechanism capable of altering the allocation.

Four features of the condition should be marked.

The condition is structural. It fixes what a party must be able to do and remains silent on what any party ought to believe about their relation. This keeps the position clear of the paternalism that any theory of unrecognised constriction otherwise invites, since such a theory supplies the correct description and someone must then be authorised to supply it.

Standing is allocated asymmetrically, to the party who bears the cost. Symmetrical formulations, under which the couple jointly reflects on its arrangement, fail in the cases the condition exists to cover, since asymmetric interpretive control is what those cases consist of.

The mechanism requirement carries most of the practical weight, and it identifies the institutions of the allocative account as the enabling conditions of admissible generative description. Independent income, an exit that leaves the departing party with a livelihood, recognised claims on jointly produced assets, and the legal recognition of care as contribution are the mechanisms through which contestation can alter an allocation. Proposition 4.6 and the evidence cited with it are the argument that these mechanisms change conduct inside continuing relations, which is what makes them conditions of generation in preference to remedies for its failure.

The condition is a constraint and supplies no objective function. It rules descriptions inadmissible and ranks nothing. Two relations that both satisfy it are unranked by anything in this paper, and no advice about the conduct of a marriage follows from it. Section 7.5 states the cost of this deliberately.

7.4 The historical situation of the generative account itself

The method declared in §1 requires that the historical reconstruction applied to the allocative account in §2 be applied to the account advanced here, and the result is unfavourable in a way worth stating.

A description of marriage as a site for the production of capacity, meaning and possibility becomes available under determinate conditions: the migration of subsistence risk away from kin, the availability of independent income to both parties, extended life expectancy, employment whose returns depend on capacities formed over decades, and a cultural repertoire in which the relation is expected to underwrite personal development (Giddens, 1992; Beck & Beck-Gernsheim, 1995). Illouz analyses that repertoire as emotional capitalism, in which the vocabulary of self-realisation organises intimate life along lines congenial to market forms (Illouz, 2007; Illouz, 2012).

Two consequences follow for this paper. The generative account is a historically situated abstraction on the same footing as the allocative account, with its own conditions of adequacy and its own expiry; presenting it as the discovery of what marriage always was would repeat, in a new vocabulary, the error this paper attributes to universalised allocative models. And the account is differentially available by material position, since a couple whose margin of subsistence is narrow has grounds for evaluating a marriage by the security it supplies, and a theory that describes their evaluation as a failure of vision is producing the ideology of a class that can afford the alternative.

Claim 7.4. (Situated adequacy). Each of the two accounts is adequate to determinate material conditions; where subsistence risk is borne within the relation and exit is unavailable, the allocative account describes what the parties confront, and the generative account applies in the measure that those conditions are relieved.

Claim 7.4 completes the dialectical formulation the title announces. The relation between the accounts is one of conditions and moments, with the historical movement running from the first toward the second and with the second retaining the first as its enabling condition.

7.5 What the condition leaves undone

Condition 7.3 rules out and ranks nothing. It is silent on which of two arrangements is better where both are contestable; it advises no couple on the conduct of their relation; it gives no guidance where no party bears an uncontestable cost. This is the standing cost of constraint form, and it is accepted here for the reason given in Claim 4.4: a formulation that ranked arrangements by what they generate would reintroduce the aggregation that Proposition 4.3 removes, and the aggregated formulation licenses exactly the sentence Claim 7.2 blocks.

8. Consequences for the Evaluation and Study of Marriage

This section draws the consequences of the position for three audiences, namely the parties to an evaluation, the participants in the intergenerational disagreement described in §6.3, and empirical researchers, and it states the conditions under which each proposition fails. The objective is to say what changes and, equally, what stays in place. The method is direct statement, with the evidential status of each consequence marked.

8.1 Consequences for evaluation by the parties

The position adds one question to the evaluative inventory and removes none. Alongside the question of what each party holds, the question of what the coupling has produced over the intervals already observed becomes available: what each party has become able to do since the relation began, which descriptions of themselves each has acquired and lost, and whether the changes run in the same direction for both. Under Proposition 4.5 the answers available in advance are weak, and the honest form of the consequence is correspondingly modest: the second question has a determinate object, it is answerable in retrospect over short intervals already lived, and confidence about a future decade is unwarranted for the parties as much as for their families.

The consequence with the most practical content follows from Proposition 4.3. Since expansion, constriction and redirection are all generative, the question of intensity is uninformative and the question of mode is the operative one. A relation in which a great deal is happening supports no inference; a relation in which the range of one party’s possible descriptions of themselves is narrowing supports one.

8.2 Consequences for the intergenerational disagreement

Section 6.3 supplies both parties to this disagreement with something to concede. The senior party’s criteria are a screening technology with a rational basis and a known misclassification profile, and their standing declines in the measure that the underwriting risk has moved elsewhere. The junior party’s generative claim concerns a property that neither party can verify in advance, and confidence in it is a claim about a future made by the parties with the strongest interest in its truth.

The productive form of the disagreement locates it where the two claims meet: on what would have to hold for the screening criteria to be misclassifying this case, and on what fallback the junior party retains where the generative claim proves false. The second of these is Proposition 4.6 in practical form, and it is a question the senior party is well placed to press.

8.3 Consequences for empirical work, and the conditions of refutation

Four consequences bear on research design, and each carries a corresponding condition of refutation.

The endogeneity of endowments (Proposition 4.2) implies that designs measuring characteristics at formation and outcomes decades later attribute to selection what the trajectory produced. Panel designs with repeated measurement of capacity, and designs exploiting variation in the timing of relational events, address this. Proposition 4.2 fails where change in capacity within couples is fully accounted for by characteristics measured at formation, in which case the object named here has no separate existence.

The neutrality proposition (Proposition 4.3) implies that measures of relational activity, of effort and of skill acquisition register constriction and redirection alongside expansion, so measurement requires the reachable set in preference to the intensity of change. Instruments recording the range of a party’s independent activity, contacts and self-descriptions over time are the appropriate form. Proposition 4.3 does no empirical work where intensity and mode covary strongly enough that the distinction is idle.

The verification asymmetry (Proposition 4.5) implies that self-reported generative claims from within a relation carry the interpretive control of that relation, so designs relying on them measure something other than what they intend where control is asymmetric. Corroboration through observable changes in independent activity is available, and Proposition 4.5 fails where such corroboration and self-report coincide under conditions of known asymmetry.

The precedence proposition (Proposition 4.6) is the most directly testable element of the position, since variation in exit options, in the recognition of care contributions, and in property regimes on dissolution occurs across jurisdictions and over time, with Stevenson & Wolfers (2006) as a template. Proposition 4.6 fails upon a finding that variation in the fallback position leaves the within-relation trajectory of capacity unchanged, which would remove the ground for treating distributive guarantees as conditions of generation.

8.4 Scope beyond intimate relations

The propositions are stated in terms of coupling and reachable sets, so their form extends to organisations, to teaching relations, to communities and to cooperation between states. This paper establishes nothing about those domains, and the extension is recorded here as the reason for the domain-neutral formulation, with the substantive work left to separate treatment.

9. Limits of the Present Statement

This section states the limits of the position with sufficient specificity that a reader can judge which of them are remediable. The objective is to prevent the propositions from being read as more than they are. The method is enumeration by type: what remains unmeasured, what remains unformalised, what remains unresolved in principle, and what risk the vocabulary carries.

The reachable set of §5.2 has no measurement procedure. The construct is defined in set relations, and nothing here individuates its elements or determines their availability at a time. Everything in §8.3 that turns on measuring the mode of change is therefore a research programme in place of a result.

Expansion has no formal home, as §5.3 states. The apparatus of this series formalises the collapse of distinctions and leaves the generation of novelty open, so the central positive phenomenon of this paper is at present notation.

Attribution to the coupling remains unresolved. Definition 1.2 requires that a change in the reachable set be attributable to the coupling, and separating that contribution from concurrent employment, health, ageing and social change is the standard problem of causal inference in relational settings, which this paper addresses by declaring it.

The valence-free definition purchases discipline at a cost, since it leaves the position with a descriptive vocabulary in which the most important distinctions among relations are invisible, and it relocates the entire evaluative burden onto Condition 7.3. Whether a single structural condition can carry that burden across cases is the point at which the position is most exposed.

The vocabulary carries a standing risk, identified in §7.1 and repeated here because the risk survives the condition designed to contain it. Terms of generation, production and emergence are attractive, they describe relations in a register their participants find flattering, and they are available to whoever benefits from deferral. A reader who takes from this paper the proposition that relations produce value, and leaves Condition 7.3 behind, will have taken the part that was already established and left the part that constrains it.

10. The Standing of the Dialectical Formulation

This paper has advanced a position on the relation between two accounts of marriage. The allocative account, which takes endowments as given and treats their arrangement as the analytical problem, has been reconstructed as the adequate abstraction of determinate conditions: scarcity, the bearing of marital risk by kin, and the unobservability of future capacity. The generative account, which takes as its object the change in what each party can do, say and become as a consequence of the coupling, has been stated in five propositions concerning non-conservation, the endogeneity of endowments, the evaluative neutrality of generative change, the asymmetry of verification, and the precedence of distributive guarantees among the conditions of generation.

The dialectical form of the position consists in the two directions of dependence it asserts. Generation continually reconstitutes the objects that distribution arranges, which is the sense in which the allocative account has a scope limit. Distributive guarantees stand among the conditions under which coupling expands the reachable sets of both parties, which is the sense in which the generative account depends on what it is tempted to supersede. The temptation is the point of maximum danger for the position, since the sentence that a relation generated a great deal is available as a reason for deferring the claims of whoever paid for it, and the sentence is available in the framework’s own vocabulary.

The operative content of the position is therefore a constraint. A generative description of a relation is admissible where the party bearing the cost of the present allocation retains the standing to contest it and a mechanism through which contestation can alter it. The constraint ranks nothing and advises no one; it identifies the institutions of the allocative account as the enabling conditions of truthful generative description; and it blocks the aggregation under which one party’s expansion is offered as an offset for another’s constriction. Marriage, on this account, is a system in which value is continually produced and continually apportioned, with the terms of the apportionment among the conditions of the production.

References

Aron, Arthur, and Elaine N. Aron (1986). Love and the Expansion of Self: Understanding Attraction and Satisfaction. New York: Hemisphere.

Arrow, Kenneth J. (1962). The Economic Implications of Learning by Doing. Review of Economic Studies, 29(3), 155–173.

Arthur, W. Brian (1994). Increasing Returns and Path Dependence in the Economy. Ann Arbor: University of Michigan Press.

Beck, Ulrich, and Elisabeth Beck-Gernsheim (1995). The Normal Chaos of Love. Cambridge: Polity Press.

Becker, Gary S. (1964). Human Capital: A Theoretical and Empirical Analysis, with Special Reference to Education. New York: Columbia University Press.

Becker, Gary S. (1965). A Theory of the Allocation of Time. Economic Journal, 75(299), 493–517.

Becker, Gary S. (1973). A Theory of Marriage: Part I. Journal of Political Economy, 81(4), 813–846.

Becker, Gary S. (1981). A Treatise on the Family. Cambridge, MA: Harvard University Press.

Becker, Gary S., Elisabeth M. Landes, and Robert T. Michael (1977). An Economic Analysis of Marital Instability. Journal of Political Economy, 85(6), 1141–1187.

Bourdieu, Pierre (1986). The Forms of Capital. In J. G. Richardson (ed.), Handbook of Theory and Research for the Sociology of Education. New York: Greenwood, 241–258.

Bowlby, John (1988). A Secure Base: Clinical Applications of Attachment Theory. London: Routledge.

Coleman, James S. (1988). Social Capital in the Creation of Human Capital. American Journal of Sociology, 94, S95–S120.

Dépelteau, François (2008). Relational Thinking: A Critique of Co-Deterministic Theories of Structure and Agency. Sociological Theory, 26(1), 51–73.

Donati, Pierpaolo (2011). Relational Sociology: A New Paradigm for the Social Sciences. London: Routledge.

Drigotas, Stephen M., Caryl E. Rusbult, Jennifer Wieselquist, and Sarah W. Whitton (1999). Close Partner as Sculptor of the Ideal Self: Behavioral Affirmation and the Michelangelo Phenomenon. Journal of Personality and Social Psychology, 77(2), 293–323.

Duflo, Esther (2003). Grandmothers and Granddaughters: Old-Age Pensions and Intrahousehold Allocation in South Africa. World Bank Economic Review, 17(1), 1–25.

Eglash, Ron (2016). An Introduction to Generative Justice. Teknokultura, 13(2), 369–404.

Emirbayer, Mustafa (1997). Manifesto for a Relational Sociology. American Journal of Sociology, 103(2), 281–317.

Engels, Friedrich (1884). The Origin of the Family, Private Property and the State. Hottingen-Zurich.

Erikson, Erik H. (1950). Childhood and Society. New York: Norton.

Feeney, Brooke C. (2004). A Secure Base: Responsive Support of Goal Strivings and Exploration in Adult Intimate Relationships. Journal of Personality and Social Psychology, 87(5), 631–648.

Folbre, Nancy (1994). Who Pays for the Kids? Gender and the Structures of Constraint. London: Routledge.

Folbre, Nancy (2001). The Invisible Heart: Economics and Family Values. New York: New Press.

Fricker, Miranda (2007). Epistemic Injustice: Power and the Ethics of Knowing. Oxford: Oxford University Press.

Giddens, Anthony (1992). The Transformation of Intimacy: Sexuality, Love and Eroticism in Modern Societies. Cambridge: Polity Press.

Himmelweit, Susan (1995). The Discovery of ``Unpaid Work’’: The Social Consequences of the Expansion of Work. Feminist Economics, 1(2), 1–19.

Hirschman, Albert O. (1958). The Strategy of Economic Development. New Haven: Yale University Press.

Hirschman, Albert O. (1967). Development Projects Observed. Washington, DC: Brookings Institution.

Hodgson, Geoffrey M. (1993). Economics and Evolution: Bringing Life Back into Economics. Cambridge: Polity Press.

Huang, Wanhong (2026a). A Metamodel for Ontologically Heterogeneous Social Dynamics — Part I: The Highest Abstract Layer. Draft preprint.

Huang, Wanhong (2026b). Emptiness and Value Generation: A GR-Eco Study of Luxury Space. Draft preprint.

Illouz, Eva (2007). Cold Intimacies: The Making of Emotional Capitalism. Cambridge: Polity Press.

Illouz, Eva (2012). Why Love Hurts: A Sociological Explanation. Cambridge: Polity Press.

Lundberg, Shelly, and Robert A. Pollak (1993). Separate Spheres Bargaining and the Marriage Market. Journal of Political Economy, 101(6), 988–1010.

Manser, Marilyn, and Murray Brown (1980). Marriage and Household Decision-Making: A Bargaining Analysis. International Economic Review, 21(1), 31–44.

Marx, Karl (1857–1858). Grundrisse: Foundations of the Critique of Political Economy. Trans. M. Nicolaus, Harmondsworth: Penguin, 1973.

McElroy, Marjorie B., and Mary Jean Horney (1981). Nash-Bargained Household Decisions: Toward a Generalization of the Theory of Demand. International Economic Review, 22(2), 333–349.

Nelson, Richard R., and Sidney G. Winter (1982). An Evolutionary Theory of Economic Change. Cambridge, MA: Belknap Press of Harvard University Press.

Nussbaum, Martha C. (2000). Women and Human Development: The Capabilities Approach. Cambridge: Cambridge University Press.

Ostrom, Elinor (1990). Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge: Cambridge University Press.

Ricardo, David (1817). On the Principles of Political Economy and Taxation. London: John Murray.

Robbins, Lionel (1932). An Essay on the Nature and Significance of Economic Science. London: Macmillan.

Romer, Paul M. (1986). Increasing Returns and Long-Run Growth. Journal of Political Economy, 94(5), 1002–1037.

Romer, Paul M. (1990). Endogenous Technological Change. Journal of Political Economy, 98(5), S71–S102.

Schumpeter, Joseph A. (1934). The Theory of Economic Development. Cambridge, MA: Harvard University Press.

Sen, Amartya (1985). Commodities and Capabilities. Amsterdam: North-Holland.

Sen, Amartya (1990). Gender and Cooperative Conflicts. In I. Tinker (ed.), Persistent Inequalities: Women and World Development. Oxford: Oxford University Press, 123–149.

Sen, Amartya (1999). Development as Freedom. Oxford: Oxford University Press.

Spence, Michael (1973). Job Market Signaling. Quarterly Journal of Economics, 87(3), 355–374.

Stark, Evan (2007). Coercive Control: How Men Entrap Women in Personal Life. Oxford: Oxford University Press.

Stevenson, Betsey, and Justin Wolfers (2006). Bargaining in the Shadow of the Law: Divorce Laws and Family Distress. Quarterly Journal of Economics, 121(1), 267–288.

Stiglitz, Joseph E., and Andrew Weiss (1981). Credit Rationing in Markets with Imperfect Information. American Economic Review, 71(3), 393–410.

Thomas, Duncan (1990). Intra-Household Resource Allocation: An Inferential Approach. Journal of Human Resources, 25(4), 635–664.

Veblen, Thorstein (1898). Why Is Economics Not an Evolutionary Science? Quarterly Journal of Economics, 12(4), 373–397.

Waring, Marilyn (1988). If Women Counted: A New Feminist Economics. San Francisco: Harper and Row.