Recordability and the Limits of Juridical Translation - A Preliminary Discussion of Generative Conditions in Legal Form 【（Preliminary）Draft】

Posted on 2026-07-29 In Governance , Law Views: Disqus: Word count in article: 14k Reading time ≈ 51 mins.

oindent Legal and institutional programmes are increasingly asked to secure something described as a capacity: an organization that can recognize those within it, a system that can correct itself, an order that can take up experience it has not yet accommodated.

Abstract

\noindent
Legal and institutional programmes are increasingly asked to secure something described as a capacity: an organization that can recognize those within it, a system that can correct itself, an order that can take up experience it has not yet accommodated. In practice such programmes arrive as training counts, representation ratios, published procedures, documented consultations and filed reports. The received readings treat this as implementation failure, symbolic compliance or decoupling. This paper proposes that the pattern has a structural explanation and that its direction can be stated as a selection principle: juridical translation preserves the recordable aspects of an aim and discards the rest, after which the preserved aspects operate as the aim itself. The paper offers three further proposals, each at declining confidence. A boundary condition distinguishes the translations that succeed, whose objects are conduct occurring at an identifiable time with a party whose position it alters, from those that degrade, whose objects are capacities whose loss produces no event and no claimant. A claim about the source of the pressure holds that the movement toward recordability tracks answerability to a party positioned outside the relation, so that relocating a generative aim from a court to an agency, a mediator or an internal governance process does not escape it. A hypothesis, offered with the observations that would refute it, holds that where an instrument is installed at recordable sites and the underlying conduct persists, the conduct is reorganized toward sites the instrument cannot reach, so that measured compliance and the incidence of the failure may rise together. The paper concedes at the outset that the position it occupies is largely occupied already, by reflexive law, by the juridification literature and by the sociology of audit, and it states its contribution as a narrow one: a positive characterization of what translation fails to carry, drawn from Generative Relational Being, together with the limits that characterization inherits.

\vspace{0.8em}
\noindentKeywords: juridical translation; recordability; reflexive law; juridification; audit and indicators; adjudication; polycentricity; generative relational being.

Keywords: juridical translation; recordability; reflexive law; juridification; audit and indicators; adjudication; polycentricity; generative relational being.

A note on the standing of this paper. This is a preliminary discussion and is offered as one. Its central proposals are conjectures at differing degrees of support, and they are stated in a form that permits them to be attacked. The paper concedes early and at length that the ground it stands on is occupied: the claim that legal form transforms what it takes up has been made in stronger versions by others, and a reader who holds that nothing survives those versions will find the concessions of Section 2 congenial. The paper advances no doctrinal proposal, offers no theory of attribution, and supplies no procedure by which the capacity it discusses could be measured in a particular case. Its conclusion is negative in form, and the negative form is deliberate: the argument is that a certain class of aims resists legal instrumentation for reasons internal to the requirements of legal determination, and that the resistance is a property of the aims and of the form together. The sections are severable. A reader may accept the selection principle of Section 3 and the boundary of Section 4 while rejecting the migration hypothesis of Section 6, which is the weakest thing here and is marked as such. Objection, correction and counter-evidence are welcome at huangwanhong@serendip.ngo.

1. Introduction

A recurrent pattern in contemporary institutional life invites explanation. A programme is set up with an aim described in terms of a capacity. An organization is to become one in which those within it can be recognized. A regulatory regime is to secure an industry that corrects itself before harm occurs. A university is to sustain a community in which a member’s experience can enter the shared understanding of the place. A supply chain is to be governed by an enterprise that identifies and addresses the human consequences of its relationships. The aims are stated in this way because those who state them mean something by it, and what they mean concerns a capacity of a set of relations, which is a different object from the occurrence of a particular act.

The programmes then arrive. They arrive as a count of training sessions completed, a proportion of positions held, a published grievance procedure, a record of consultations conducted, a policy adopted by a board, an annual report filed and, where the regime has teeth, an audit of whether the policy was adopted and the report filed. The distance between the aim as stated and the programme as realized is familiar to everyone who has worked inside such a regime, and it is ordinarily explained in one of three ways. The first treats the distance as implementation failure, remediable by better design, stronger sanctions or more resources. The second treats it as capture, in which the regulated construct the meaning of the regulation (Edelman, 1992; Edelman, 2016). The third treats it as decoupling, in which formal structure is adopted for legitimacy while practice proceeds otherwise (Meyer & Rowan, 1977; Bromley & Powell, 2012).

Each of these explanations is supported by evidence, and none of them accounts for the direction of the distance. In every case the aim moves the same way. It moves toward what can be written down and checked by someone who was not present. The uniformity of that movement across regimes with different sanctions, different designers, different degrees of capture and different levels of resourcing suggests that something in the passage from an aim to a legal or institutional instrument selects, and selects on a particular property. Identifying that property, and saying what follows from it, is the object of this paper.

The proposal is that the property is recordability, and that the passage from an aim to an instrument preserves the recordable aspects of the aim and discards the others. The proposal is not that legal abstraction loses information, which is true of all abstraction and is a condition of its usefulness. It is that the loss has a direction fixed by the requirements of legal determination, and that the surviving fragment then takes over the operative role of the aim, so that the loss propagates from the representation of the aim into the conduct of those who pursue it. On the second point the sociology of measurement has already assembled the mechanism: measures that are used to allocate consequences alter what they measure, and the alteration runs toward the measure (Campbell, 1979; Espeland & Sauder, 2007).

The ground here is occupied, and the concession is made at the outset because the paper’s contribution can only be stated once the concession is in place. That legal form transforms what it takes up, in place of transporting it, is the thesis of the reflexive-law tradition and of its self-criticism (Teubner, 1983; Teubner, 1987; Teubner, 1998). That the transformation damages a domain of social life whose reproduction proceeds by other means is Habermas’s account of juridification and of the colonization of the lifeworld (Habermas, 1987). That verification constructs the objects it verifies, so that the demand for audit produces auditability where none existed, is Power’s account of the audit society (Power, 1997). That the requirements of adjudication exclude a class of tasks by their structure is Fuller’s account of polycentricity (Fuller, 1978). This paper adds nothing to those results and depends on all of them.

What it proposes is narrower, and consists of four things at declining confidence. The first is the selection principle just stated, offered as an explanation of a direction that the surveyed accounts describe. The second is a boundary condition separating translations that preserve their object from translations that degrade it, since an account condemning every passage into legal form would be both unfalsifiable and false; anti-discrimination law in its eventful applications is a considerable achievement, and any account that fails to say why it works has failed. The third is a claim about the source of the pressure: it tracks answerability to a party positioned outside the relation, which entails that relocating a generative aim from a court to an administrative agency, a mediator or an internal governance process does not escape the transformation and in some settings intensifies it. The fourth is a hypothesis about what happens after an instrument is installed, namely that conduct which the instrument reaches at recordable sites is reorganized toward sites the instrument cannot reach, with the consequence that measured compliance and the incidence of the failure may rise together.

The framework from which the positive characterization is drawn is stated here once, in the form its author has given it.

Generative Relational BeingGenerative Relational Being (GRB): A theory of how subject, meaning, value, creation, and normativity co-emerge through generative relational processes.

Two features of the framework govern its use below. It treats subjects, meanings and norms as products of relational processes, in place of treating them as the inputs of such processes, which is why it can pose a question about the conditions of generation as a question distinct from questions about validity, entitlement and attribution. And it is conducted under a keynote of historical dialectics: claims are advanced as provisional stabilizations open to revision, the finitude of knowledge and of practice within a historical frame is acknowledged rather than transcended, and philosophical argument is joined to whatever operable and testable formulation can be had. The keynote governs the register of this paper as well as its content, and it is the reason the central hypothesis is stated together with the observations that would tell against it.

The contribution the framework makes here is a single one and is deliberately modest. The accounts surveyed in Section 2 converge on the existence of a residue that legal form fails to carry, and they characterize the residue negatively: as what is not auditable, as what the medium of law displaces, as what the transplanted rule fails to transplant. Generative Relational Being permits a positive characterization, namely the capacity of a set of relations to generate a further determination of a matter that some determination has settled. That characterization is examinable, it connects to a distinction the law already draws between the finality of a case and the closure of a question, and it identifies one legal form in which the aim is carried without being converted into a stock of documents. It also inherits a difficulty severe enough to be stated in the same breath: the capacity so characterized has no observer, and a tribunal is precisely an observer positioned outside the relation.

The paper proceeds in eight further steps. Section 2 sets out what five bodies of work have established about the transformation and states what they leave open. Section 3 proposes recordability as the principle by which the transformation selects, defines the term, and separates the proposal from the trivial claim that abstraction loses information. Section 4 states the conditions under which translation preserves its object and the conditions under which it degrades it, and gives the comparative prediction the boundary yields. Section 5 argues that the pressure toward recordability follows answerability rather than forum, and prices the consequence for the proposals that would house generative aims elsewhere. Section 6 states the migration hypothesis, the bound on installed instruments from which it follows, its relation to decoupling, and the evidence that would refute it. Section 7 examines the one case in which legal form has installed an instrument reaching a site that the preceding section held to be beyond installation, and fixes the extent of the exception. Section 8 states what Generative Relational Being contributes and what it must concede. Section 9 sets out the consequences for legal theory, the risk of quietism, and the limits of the account, and Section 10 closes with open questions and a note on method.

2. The Reflexive Turn and Its Settled Results

This section assembles the prior work on which the paper depends and states what that work has established. Its objective is to fix the concessions before any proposal is made, so that the proposals of later sections can be assessed for what they add. It proceeds through five bodies of work in the order in which they bear on the problem: the programme of reflexive and responsive law; the self-criticism of that programme in the regulatory trilemma and the theory of legal irritants; the juridification literature and its strongest thesis; the sociology of audit, commensuration and indicators; and the organizational literature on symbolic structure and decoupling. It closes with a statement of what the five leave open. The method is a review of the arguments at the points where they bear on the direction of transformation, and the review is selective by design; the omissions are many.

2.1 The Programme of Reflexive and Responsive Law

The proposition that law should secure the conditions under which a social sphere regulates itself, in place of prescribing the outcomes that sphere is to reach, was stated in a developed form four decades ago. Nonet and Selznick distinguished a responsive type of law oriented to purpose and to the competence of institutions to serve it (Nonet & Selznick, 1978). Teubner, drawing on the systems theory of social differentiation, proposed reflexive law as a stage in which legal rationality restricts itself to the regulation of self-regulation, providing procedural norms, organizational competences and structures of discourse within which other subsystems conduct their own coordination (Teubner, 1983; Luhmann, 2004). Responsive regulation added an enforcement architecture in which escalation is graduated and cooperation is the default (Ayres & Braithwaite, 1992), and decentred accounts extended the picture to a field of state and non-state regulators (Black, 2001). Adjacent to this tradition, the study of common-pool institutions established that arrangements crafted from within by those affected by them can outperform arrangements imposed from outside, and that participation in the modification of operational rules belongs among the design principles of durable institutions (Ostrom, 1990). The experimentalist programme in constitutional law drew the procedural consequence, proposing rolling standards, benchmarking and a right whose content is the destabilization of an institutional settlement that has ceased to correct itself (Dorf & Sabel, 1998; Sabel & Simon, 2004).

The relevance of this tradition to the present paper is direct, and it stands here as a difficulty for the present argument. Any claim that legal theory has overlooked the conditions of self-correction is unavailable. The position has been occupied since 1983, it has an enforcement theory, a constitutional theory and an institutional-design literature, and its central concern is the same concern.

2.2 The Regulatory Trilemma and the Legal Irritant

The reflexive programme produced its own criticism, and the criticism is closer to the present argument than the programme is. Teubner’s account of juridification states a trilemma: where legal regulation and the internal logic of a regulated sphere are incongruent, one of three outcomes follows. The regulation remains without effect in the sphere it addresses. The regulation disintegrates the sphere by displacing the mechanisms through which the sphere reproduced itself. Or the sphere disintegrates the regulation, absorbing the legal programme into its own operations and leaving the legal form standing while the operative content becomes the sphere’s own (Teubner, 1987). The trilemma is a general statement of translation failure, and its second and third limbs anticipate a good deal of what follows below.

The theory of legal irritants states the mechanism at a smaller scale. A rule taken from one legal order into another is not transferred; it triggers a series of events in the receiving order whose outcome resembles neither the rule in its origin nor the receiving order before it arrived (Teubner, 1998). Translation is a perturbation whose effects are determined by the receiving structure. That formulation removes any suggestion that the problem discussed in this paper is peculiar to the passage from social aim to legal form, and locates it in transfers between structured domains generally.

2.3 Juridification and the Colonization Thesis

The strongest available version of the damage thesis is Habermas’s. Legal-bureaucratic media, in securing the entitlements of the welfare state, restructure the relations they secure. The pupil, the client and the family member acquire enforceable claims, and in acquiring them the relation is reconstituted as one administered through money and power, displacing the communicative processes by which the relation had reproduced its own understandings. The gain in protection is real and the transformation is real, and Habermas holds the two to be inseparable in that class of case (Habermas, 1987).

Two features of the thesis bear on what follows. It is diagnostic at the level of media, holding that the legal medium can carry certain contents and not others, which is a stronger claim than any made here. And its normative conclusion is genuinely double-edged, since juridification in the same account also emancipates; the four waves are not a decline narrative. A paper proposing that generative conditions resist legal form must state that this ground is held by a more powerful argument than its own, and must say what remains for it to do. The answer given here is in Section 8: the colonization thesis characterizes the displaced content by the medium that fails to carry it, and the characterization offered here is of the content itself.

2.4 Audit, Commensuration and Indicators

A fourth body of work supplies the mechanism by which a recordable proxy displaces the aim it was introduced to serve. Power’s account of the audit explosion holds that audit does not simply inspect a pre-existing object; the demand for verification produces auditability, reorganizing the audited body so that it presents a surface an auditor can check, and the resulting assurance concerns that surface (Power, 1997). Commensuration, understood as the transformation of qualitative differences into differences of quantity along a common metric, is a social process with consequences for what its objects become (Espeland & Stevens, 1998). The study of rankings supplies the empirical form: measures used consequentially generate reactivity, through self-fulfilling prophecy and through the reorganization of the measured body around the categories of the measure (Espeland & Sauder, 2007). The general statement is older, and Campbell’s formulation records that indicators used for social decision-making are subject to corruption pressures and tend to distort the processes they monitor (Campbell, 1979). Related work traces the political effects of indicators in global governance (Merry, 2016), the costs of metric fixation (Muller, 2018), the ritual dimension of verification in institutional life (Strathern, 2000), and the requirement of legibility in administrative schemes, which reduces a complex practice to the aspects a central authority can read (Scott, 1998).

This literature establishes the second-order effect on which Section 3 relies, and it establishes it empirically. What it does not do is state a principle governing which aspect of an aim survives the passage; it observes that what survives is measured and that measurement distorts.

2.5 Symbolic Structure and Decoupling

The fifth body of work concerns what organizations do with legal requirements. Formal structures are adopted in conformity with institutionalized expectations, and their adoption is loosely coupled to operational activity (Meyer & Rowan, 1977). Where a legal requirement is ambiguous, organizations construct compliance through visible structures whose existence comes to be treated, by managers and eventually by courts, as evidence of compliance; the meaning of the law is thereby constructed within the field it regulates and returned to the legal system, which defers to it (Edelman, 1992; Edelman, 2016). The later literature distinguishes policy-practice decoupling from means-ends decoupling, in which the practice is genuinely performed and its connection to the intended outcome is unestablished (Bromley & Powell, 2012).

The second form is the more serious for present purposes, since it describes a case in which nobody is dissembling. The training is delivered, the procedure is followed, the report is accurate, and the relation between all of this and the aim is unknown.

2.6 Findings and Residue

Four results are treated as settled in what follows. Legal form transforms what it takes up, and and the transformation is structural. The transformed content tends toward the documentary. Consequential measurement reorganizes the measured body around the measure. Courts and regulators come to treat the resulting structures as the thing they were introduced to secure.

Three questions are left open by the five bodies of work, and the remainder of the paper addresses them. The direction of the transformation is described in each account and explained by none of them in terms of a property that decides what survives. No account states the conditions under which translation preserves its object, with the consequence that the literature is read, in practice, as a general suspicion of legal instruments; that reading is not supported by the evidence and is unhelpful to anyone designing one. And no account says what becomes of the conduct at issue after the instrument is installed and the recordable surface is secured.

3. The Selection Principle of Juridical Translation

This section proposes an explanation for the direction of the transformation described in Section 2. Its objective is to state, in a form open to counter-example, the property on which the passage from an aim to a legal or institutional instrument selects. It proceeds in four steps: the requirements a determination imposes on its object; a definition of recordability; the selection principle itself, together with the second-order effect that carries the loss from the representation into the world; and a statement of what the principle does not assert, which is needed because a stronger and emptier claim lies close by. The method is conceptual analysis of the form of legal determination, joined to the empirical results recorded in Section 2.

3.1 The Requirements Imposed by the Form of Determination

Adjudication is distinguished among modes of social ordering by the mode of participation it affords to the affected party, which is the presentation of proofs and reasoned arguments (Fuller, 1978). Everything that follows about the form of a legal determination follows from that. An object susceptible of determination in this mode must admit of proof, which requires that it be the kind of thing about which evidence can be led. It must permit the allocation of a burden, which requires that its presence or absence be capable of standing as a proposition that one party asserts and another denies. It must admit of a reasoned justification addressed to parties and to reviewing bodies, which requires that the ground of the decision be statable in advance of the case and applicable to others. And the determination must terminate, since a decision that settles nothing is not a decision.

Fuller drew from the same premise the conclusion about polycentric tasks. Where the elements of a problem are so interdependent that a change at one point redistributes the entire configuration, the affected parties cannot be identified in advance, the effect of any proposed disposition cannot be presented as proof, and the mode of participation characteristic of adjudication has no purchase. The task remains tractable in another form, and Fuller assigns it to managerial direction or to contract (Fuller, 1978). The doctrine of justiciability records the same limit from inside the law, in the requirement that a controversy admit judicially discoverable and manageable standards, and comparable devices are found in other systems where courts decline to review the substance of an evaluative judgment.

These requirements are achievements. Each of them protects a party against a determination made on grounds that were never stated, against an attribution of responsibility supported by nothing that can be examined, and against a decision by an authority answerable to no one for its reasons. Nothing in this paper is to be read as a complaint about them, and Section 9 returns to the point.

3.2 The Recordability Condition

The requirements just set out have a common presupposition, and naming it is the first step of the argument.

RecordabilityA feature of a state of affairs is recordable when it can be fixed in a description such that a party who did not observe the state of affairs can determine, from the description together with materials available to that party, whether the feature obtains. Recordability is relative to the position of the checking party and to the materials that position commands, and it therefore admits of degrees.

The relativity in the definition is deliberate and does work later. A feature may be recordable for a participant in a relation and unrecordable for a tribunal, and the difference is not a matter of the tribunal’s diligence. Whether a colleague’s silence in a meeting was a withdrawal of standing or an ordinary economy of speech is a question the room may be able to answer and a question the record cannot, because the materials that settle it are the accumulated history of that room and are not available in the form of a description.

Three classes of feature are worth separating. A feature is directly recordable when it consists in an event with a time, a place and participants: a letter sent, a payment withheld, a post advertised and filled. A feature is recordable through a proxy when its presence is inferred from a directly recordable feature standing in a stable relation to it, which is the form of nearly every institutional indicator. A feature is unrecordable at a given position when neither of these holds, which may be so because the feature consists in an absence, because it consists in a disposition of interpretation held by participants, or because it consists in a rule of distinction that governs conduct and is never pronounced.

3.3 The Principle

The principle can now be stated.

Selection by recordabilityIn the passage from an aim to a legal or institutional instrument, those aspects of the aim that are recordable at the position of the party to whom the instrument is answerable are preserved in the instrument, and those aspects that are unrecordable at that position are discarded. The preserved aspects thereafter carry the operative content of the aim, in the conduct of the bodies subject to the instrument and in the determinations of the bodies applying it.

The first sentence concerns the instrument and is close to analytic, given the requirements set out in Section 3. An instrument whose application turned on a feature that could not be checked by the applying body would be inapplicable, and a legislature or regulator drafting such an instrument would produce a text that courts would decline to enforce or would reduce to whatever in it could be checked. The interest of the principle lies in the second sentence, which asserts a substitution. What the instrument preserves does not remain a partial representation of the aim; it becomes the aim’s operative content, so that the question whether the aim has been served is answered, by everyone with a stake in the answer, by reference to the fragment.

The mechanism of the substitution is documented in the literature of Section 2 and needs no fresh argument. Where a measure carries consequences, the measured body reorganizes toward the measure (Campbell, 1979; Espeland & Sauder, 2007), and where verification is demanded, the verified body produces the surface that verification requires (Power, 1997). The contribution here is to say which fragment the reorganization will centre on, and why it is that fragment in every case: it is the one the answerable position can check.

A consequence follows that is worth stating separately, since it is the point at which the account acquires teeth.

Displacement of the residueThe aspects of an aim discarded in translation are not merely absent from the instrument. Where the instrument governs the allocation of attention, budget, personnel and justification within the bodies subject to it, the discarded aspects lose those resources in the bodies as well. The loss therefore propagates from the representation of the aim into the conduct of those pursuing it.

A university that must demonstrate the operation of an equality programme allocates its effort where demonstration is possible. This is not cynicism on the part of anyone in the university. It is what any body does when it must answer for itself with the materials that answering admits.

3.4 The Distance from the Trivial Claim

A stronger and emptier claim lies close to the principle and must be separated from it. The empty claim is that legal translation loses information, or, in a formal idiom, that the map from a social process to a space of legal determinations has no inverse. This is true of every abstraction and is the reason abstractions are useful. The concept of negligence discards the whole history of the defendant’s morning; the concept of consideration discards what the parties felt about their bargain. No legal system has ever claimed to be a lossless encoding of the conduct it governs, and a paper demonstrating that it fails to be one would have demonstrated nothing.

The principle proposed here differs in three respects. It asserts a direction, and the direction is fixed by a property that can be identified independently of the case. It asserts a substitution, so that the surviving fragment acquires the operative role of the whole. And it asserts a feedback, so that the discarded aspects are diminished in the world and not only in the description. The first is a proposal about legal form, the second and third are supported by the empirical literature already cited, and all three are contestable in a way that the invertibility claim is not.

3.5 Findings and Residue

This section has proposed a property, recordability, that decides which aspects of an aim survive the passage into an instrument, and has asserted that the surviving aspects thereafter operate as the aim. It leaves two things unsettled, which the next two sections take up. It gives no account of when the surviving fragment is nearly the whole aim, and it therefore reads, on its own, as a general suspicion of legal instruments that Section 4 withdraws. And it locates the answerable position without asking what fixes it, which Section 5 takes to be the more important question.

4. The Boundary Between Benign and Degrading Translation

This section states the conditions under which the passage into legal form preserves its object. Its objective is to prevent the selection principle from condemning every legal instrument, which would be both unfalsifiable and contrary to the evidence. It proceeds in five steps: a comparison of two objects that differ in the relevant respect; the conditions of a preserving translation, stated as a proposition; the characterization of the objects that fail those conditions; the intermediate cases that show the boundary to be graded; and the comparative prediction the boundary yields, together with the observation that would refute it. The method is analysis of the two objects against the requirements of determination set out in Section 3, checked against the doctrinal treatment each has received.

4.1 Two Objects Compared

Consider first a refusal to appoint a qualified applicant on a prohibited ground. The object has a time and a place. It has an author. It admits a counterfactual comparison, since the treatment of comparably situated applicants is available and statistical evidence of a pattern is admissible in many systems. It has a party whose position it alters, and that party has a reason to bring it forward. The elements of proof are contested in practice, the burden is difficult to discharge, and evidentiary devices such as shifting burdens have been developed precisely because of the difficulty. The point is that the difficulty is evidentiary. The object is the kind of thing about which evidence can be led.

Consider second the condition of an organization in which a member’s account of his situation can no longer be taken up, because every description he offers is received under a settled interpretation of him and returns confirming it. Silence is read as sullenness, protest as confirmation of difficulty, courtesy as calculation. No act on his part alters the interpretation, and no rule has been broken by anyone. There is no moment at which the condition began. There is no counterfactual comparison, since the comparison would have to be with the same organization holding a capacity it no longer holds. The party whose position is altered has no proposition to assert other than the condition itself, and the condition is the very thing whose statement the organization has ceased to be able to receive.

The two objects differ in kind and not in difficulty. The second is not a harder instance of the first.

4.2 The Conditions of a Preserving Translation

The comparison supports a proposition, offered as a hypothesis and open to counter-example.

Conditions of preserving translationThe passage of an aim into a legal instrument preserves the aim to the extent that the aim’s object satisfies three conditions: the object occurs at an identifiable time, so that a determination has something to attach to; the object admits a comparison with an alternative course that might have been taken, so that its wrongfulness can be argued; and the object alters the position of an identifiable party, so that a claim exists and someone holds it. Where all three obtain, the recordable fragment of the aim approximates the aim, and the substitution asserted in Claim 3.2 is harmless.

The proposition explains what the literature of Section 2 leaves unexplained, namely why some legal programmes are among the more successful instruments of modern governance while others reliably produce paper. Safety regulation of the eventful kind, wage and hours law, the law of dismissal, and anti-discrimination law in its individual applications concern objects satisfying the three conditions. Their difficulties are real and are difficulties of enforcement, of access and of proof.

4.3 Objects Lacking an Event

The objects on the other side of the boundary share a structure, and naming it is the substantive claim of this section.

Objects of capacity typeAn object is of capacity type when what has gone wrong consists in the reduced ability of a set of relations to produce a further determination of some matter, and when the reduction has occurred through the accumulation of acts none of which is separately impermissible. Such an object has no moment of occurrence, no available counterfactual, and no party who holds a claim at the time the reduction occurs. In consequence its recordable fragment stands in no stable relation to it, and the instruments that address it record the presence of procedures, the volume of activity and the adoption of policies.

Three features of the characterization deserve emphasis. The absence of an impermissible constituent act is essential and not incidental; where an impermissible act exists, the case has an eventful component and belongs in the graded region discussed below. The absence of a claimant at the time is a structural feature and not an artefact of standing rules, since the person affected may hold no complaint that the legal system recognizes until a decision goes against him, by which time the relevant capacity was already gone. And the absence of a counterfactual is what defeats causal argument, since the alternative course is the persistence of a capacity whose past extent is undocumented.

4.4 The Graded Region

The boundary is a gradient. Harassment law is the clearest intermediate case: the object is a course of conduct, its constituents are often individually trivial, and legal systems have responded by constructing an object of the eventful kind out of the course, through concepts such as the hostile environment, through the aggregation of incidents, and through duties on the employer to maintain the working environment. The construction is instructive. It succeeds to the degree that the constituent conduct can be described, and it is criticized precisely where the conduct is exclusion, omission and silence, which are recordable only as absences.

Duties of due diligence occupy the same region. An obligation to identify, prevent and account for adverse impacts through business relationships (UN, 2011) concerns an object of capacity type, and its implementation has moved toward reporting almost everywhere. Reasonable-accommodation duties sit closer to the eventful side, since a request refused is an event with a claimant. Preventive measures in criminal justice occupy an instructive position: they operate prospectively, and the conditions under which legal systems tolerate them include a present-fact predicate that has occurred and is disputable, a term stated in advance, and a standing review at which the measure lapses unless it is re-justified (Ashworth & Zedner, 2014). The third condition is the one that matters here, and Section 8 returns to it.

4.5 The Prediction and Its Refutation

The boundary yields a comparative prediction, and the prediction is what makes the account of Section 3 more than a suspicion. Legal programmes whose objects satisfy the three conditions should exhibit a stable relation between the recorded measure and the aim, so that improvement in the measure is accompanied by improvement in the aim, and the ordinary difficulties should be difficulties of enforcement and access. Programmes whose objects are of capacity type should exhibit divergence, so that the measure improves while the aim does not, and the ordinary difficulty should be the one the decoupling literature reports.

The observation that would refute the account is available in principle. If a programme whose object is of capacity type were shown to secure its aim stably through documentary instruments, in a setting where the aim was independently assessed, the account would be wrong. A weaker version of the same refutation would be a demonstration that the divergence in such programmes is fully explained by sanction strength, resourcing or design quality, which would return the phenomenon to the implementation-failure reading and leave the selection principle without work to do.

4.6 Findings and Residue

This section has proposed three conditions under which the passage into legal form preserves its object, has characterized the objects that fail those conditions, and has stated the comparative prediction that follows. It leaves open the question that Section 5 takes up. Nothing in the argument so far has established that the transformation is a property of the legal forum in particular, and the natural response to everything said here is to propose that objects of capacity type be housed elsewhere.

5. Answerability and the Independence of the Transformation from Forum

This section examines the proposal that objects of capacity type be pursued outside adjudication. Its objective is to determine whether the transformation described in Sections 3 and 4 is a property of the judicial forum or of something the judicial forum instantiates in a sharp form. It proceeds in four steps: the proposal and the reasons in its favour; the evidence that the same transformation occurs in the settings the proposal recommends; a claim locating the source of the pressure in answerability to a party positioned outside the relation; and the consequences of that claim, including a trade-off the paper declines to resolve. The method is comparison of the transformation across four settings, joined to analysis of what each setting requires of the body that must answer.

5.1 The Proposal to Relocate Generative Aims

The proposal has evident attractions. If the difficulty lies in the requirements of a determination, then bodies that make no determinations should be free of it. Administrative agencies act continuously, revise their positions, and stand in an ongoing relation to those they regulate. Mediation addresses the future relation of the parties, leaving aside the wrongfulness of a past act. Internal governance in a university, a firm or an international organization operates on the relations themselves and is subject to no requirement of proof. Deliberation before legislation determines nothing and closes nothing. On the face of it, each of these settings could hold an aim of capacity type without converting it into a stock of documents.

The proposal also has a respectable pedigree, since the reflexive programme reached the same conclusion by a different route, and the study of the semi-autonomous social field established that the settings in which conduct is actually ordered generate their own rules and their own means of inducing compliance, whether or not the state’s rules reach them (Moore, 1973).

5.2 The Recurrence of the Transformation Outside Adjudication

The evidence is against the proposal. Administrative governance under performance management is the setting in which the reorganization of a body around its indicators has been documented most thoroughly, and the documentation covers exactly the reforms that were introduced to make administration responsive (Hood, 1991; Power, 1997; Muller, 2018). Mediation, where it has been institutionalized at scale, has acquired accreditation schemes, session records, settlement rates and quality frameworks, since a state that refers cases to mediation must answer for the referral. University governance has produced quality assurance, and quality assurance has produced documentation on a scale that no adjudicative body has ever demanded. Reporting under human rights due diligence and under environmental and social disclosure regimes concerns objects of capacity type and consists, in its realized form, almost entirely of records (UN, 2011).

The processes by which law is made and applied furnish a further instance, and it is the one closest to home. Where the openness of a law-forming process is itself made a requirement, the requirement is discharged through consultation periods held, comments received and responded to, impact assessments conducted, registers of interests maintained and evidence sessions convened. Each of these is a record, each is checkable by a party who was absent, and none of them establishes that any participant’s understanding of the matter altered the framing of the question. A process required to demonstrate its openness will demonstrate the aspects of openness that admit of demonstration.

The pattern holds across settings that differ in every respect the proposal relies on. It holds where nothing is determined, where the relation is continuous, where no burden of proof is allocated, and where the body is internal to the relations it governs.

5.3 The Source of the Pressure

The common element in these settings is not their procedural form. It is that in each of them a body must justify its conduct to a party who was not present when the conduct occurred.

Answerability and recordabilityThe pressure toward recordability is generated by answerability to a party positioned outside the relation whose conduct is in question, and its intensity varies with the intensity of that answerability. Adjudication is the sharpest form of such answerability and is not its only form. Any body required to demonstrate to an external constituency that it has pursued an aim will pursue the aim in the aspects that admit of demonstration to that constituency, whatever the procedural character of the setting.

The claim explains the pattern just described without appeal to procedure. An agency answerable to a legislature, an auditor and a press produces indicators. A mediation service answerable to a ministry that funds it produces settlement rates. A university answerable to a regulator and a ranking produces evidence of process. And a body answerable to no one produces nothing, which is the reason answerability is demanded in the first place.

Two corollaries follow, and the second is the one that matters.

Displacement of the audienceThe constituency to which a body is answerable is ordinarily distinct from the party affected by the conduct in question. Instruments therefore operate on the body’s standing with a third party, and they protect the affected party only through the coincidence of that third party’s interest with his. Where the coincidence fails, the instruments continue to function as designed and the affected party is unprotected.

The affected party is in a poor position with respect to every element of the demonstration. He is a single case where the constituency reads aggregates. He is present once where the body appears repeatedly. And his account concerns the aspects of the matter that are unrecordable at the constituency’s position, which are the aspects the instrument discarded.

5.4 The Trade-off, Declined

An apparent implication should be resisted. It does not follow from the claim that answerability should be reduced, and this paper takes no position of that kind. Answerability is what stands between a body and the arbitrary exercise of its power, and settings of low answerability preserve more of an aim of capacity type while offering the affected party nothing at all when they turn against him. The two considerations run in opposite directions, and no ordering between them is proposed here. A framework that supplied one would be supplying an objective function over institutional states, and the commitments set out in Section 1 exclude that.

What the analysis supports is narrower and is a matter of expectation. Where answerability is intense, the aims that can be pursued are those whose objects are recordable at the answerable position, and an aim of capacity type placed in such a setting will be realized as its recordable fragment. This is a prediction about what will happen and not a recommendation about what to prefer.

5.5 Findings and Residue

This section has argued that the transformation follows answerability, with the procedural character of the forum playing no separate part, that relocation therefore offers no escape, and that the constituency receiving the demonstration is ordinarily distinct from the party the instrument was meant to protect. It leaves open what becomes of the conduct once the recordable surface has been secured, which is the subject of the next section and the weakest part of this paper.

6. The Downward Migration of Failure and the Bound on Installed Instruments

This section states the consequence that distinguishes the present account from those surveyed in Section 2. Its objective is to say what happens to conduct of the kind an instrument addresses after the instrument is installed and the recordable aspects of the aim are secured. It proceeds in five steps: an ordering of the sites at which recognition failure can occur, arranged by recordability; a bound on what an installed instrument can reach, stated as a proposition; the migration hypothesis itself; its relation to decoupling and to the audit literature, from which it must be distinguished; and the evidence that would refute it, together with a statement of why such evidence is hard to obtain. The method is conceptual derivation from the definition of Section 3, followed by an assessment of the resulting hypothesis against the standard the paper has set for its own claims.

6.1 An Ordering of Sites by Recordability

The failures that generative aims are introduced to address occur at sites that differ in how far they can be described to an outside party. The ordering in Table 1 is offered as a working device and is neither exhaustive nor sharply partitioned; a further site, concerning the terms on which a person is admitted to the class of those with whom relations are held at all, lies below the last row and is left aside here.

Table. Sites of recognition failure, ordered by recordability at the position of an external body

Site	What occurs there	Recordability at an external position
Institutional	rules, decisions, allocations, appointments	full; the acts leave documents by design
Interpretive	the readings under which rules are applied to a person	partial; reaches the record through stated reasons
Categorial	the set of descriptions under which a person can appear at all	present only as absence; an unavailable category leaves no trace
Cognitive	private attributions held by participants	unavailable; accessible through self-report of known limits
Symbolic	an unpronounced rule of distinction governing conduct toward a person	unavailable by construction; the rule is never stated by anyone

The categorial site in the third row is where testimonial and hermeneutical injustice have been located, and the hermeneutical case consists by definition in the absence of a resource for describing an experience (Fricker, 2007), which is why it leaves nothing behind. The last row is the one that governs the argument. Where a community holds a settled distinction concerning a person which no member articulates, every constituent act remains permissible and the distinction operates through the pattern of who is addressed, who is consulted, whose account is taken up and whose absence is unremarked. No instrument can require that such a rule be stated, since no one is in a position to state it, and its principal manifestation is a silence indistinguishable, in any record, from the ordinary silences of institutional life.

6.2 The Bound on Installed Instruments

A bound follows from the definition of Section 3 and from what installation requires.

The installation boundAn instrument that is installed must be specifiable in advance, since a body cannot be required to operate a mechanism whose operation cannot be described. What is specifiable in advance is recordable, since the specification supplies the description by which an outside party checks whether the mechanism operated. Installed instruments therefore reach the sites at which failure is recordable, and no further. The limit is a property of installation and is independent of the diligence, resources and good faith of those who install and operate the instrument.

The proposition explains why the instruments listed in the first row of Table 1 concentrate there. A consultation channel, a grievance procedure, an ombudsman, a mandatory training programme, a reporting duty and an audit are all mechanisms that must be described before they can be required. Each is therefore fitted to what a description can carry. The proposition also explains a difficulty that the practitioners of such mechanisms report and that the design literature reads as a defect: a grievance procedure requires the person to hold an account and a complaint, and the failure at the symbolic site is precisely the condition of holding neither, since nothing was done to him that he can state.

6.3 The Migration Hypothesis

The hypothesis can now be given, and it is offered as the weakest claim in this paper.

Downward migrationWhere an instrument is installed at the recordable sites and the disposition that produced the conduct persists, the conduct is reorganized toward sites the instrument does not reach. In consequence, measures of compliance and the incidence of the failure may rise together, and an improvement in the measured surface is compatible with an unchanged or increased incidence of the failure at the unrecordable sites.

Two supports are available for the hypothesis and neither is strong. The first is the general result about reactivity: a consequential measure reorganizes the measured body, and the reorganization is toward the measure and away from what the measure omits (Campbell, 1979; Espeland & Sauder, 2007). The second is the observation, in the literature on covert and subtle forms of discrimination, that overt expressions decline in settings where they carry costs while measures of exclusion, incivility and withheld consideration do not decline correspondingly. The second support is contested and the underlying measurements are difficult, and nothing here turns on any particular study.

The hypothesis is stated because it is the only thing in this paper that a legal theorist would not already grant, and because its practical consequence is serious enough to warrant stating a conjecture. If it is correct, then compliance data are uninformative about the aims that compliance regimes were introduced to serve, and a regime may be improving on every reported measure while the condition it addresses is unchanged.

6.4 The Distance from Decoupling and from Constructed Auditability

The hypothesis must be separated from two neighbouring accounts, since a reader may reasonably take it for a restatement of either.

Decoupling contrasts formal structure with operational practice, and in both of its forms the practice is in principle observable; the difficulty is that the observer is outside and the practice is inside (Meyer & Rowan, 1977; Bromley & Powell, 2012). Migration asserts something else, namely a relocation of conduct into a site that is unobservable by construction, where the conduct consists in a pattern that no participant articulates. The difference has a consequence for method: decoupling can be established by an investigator who gains access, and migration cannot be established that way, because there is nothing at the site for an investigator to gain access to.

The audit literature holds that verification constructs the object it verifies (Power, 1997). That claim concerns what happens to the audited surface. Migration concerns what happens to everything else, and it asserts a movement, where the audit claim asserts a construction. The two are compatible and are logically independent: an organization could produce an auditable surface while the underlying conduct remained where it was.

6.5 Refutation and the Evidentiary Difficulty

Two observations would refute the hypothesis. The first is a demonstration, in matched settings over time, that measures of covert exclusion fall alongside measures of formal compliance, which would show that installation reaches the underlying disposition. The second is a demonstration that the sites called unrecordable in Table 1 are recordable by an instrument that could be installed, which would dissolve the bound of Proposition 6.1 and with it the hypothesis that depends on it.

The evidentiary difficulty must be stated plainly, since it is severe and is not remediable by effort. The hypothesis asserts a movement into what cannot be recorded, and direct evidence of such a movement is barred by the content of the hypothesis itself. What remains available are proxies: self-report instruments for ostracism and exclusion, longitudinal audit studies, measures of exit and of voice, and comparisons between settings that differ in the intensity of their compliance regimes. Each proxy carries a defect. Self-report depends on the person’s own recognition of what has happened to him, which the symbolic site interferes with. Audit studies address the eventful sites well and the others poorly. And any comparison between regimes must contend with the possibility that the settings differ in the underlying disposition and not in its expression.

The honest summary is that the hypothesis is weakly supported, that the support available to it is indirect, and that its principal claim to attention is the seriousness of its consequence if it holds.

6.6 Findings and Residue

This section has proposed a bound on what installed instruments can reach, derived it from the requirement that an installed mechanism be specifiable, and offered a hypothesis about the reorganization of conduct that follows installation. It leaves the hypothesis in the state in which it found it, weakly supported and severable from everything preceding. What remains is to say what the framework of this paper contributes to the description of the residue, and what that contribution costs.

7. The Interpretive Site and the Partial Exception in Legal Reasoning

This section examines a case that tells against the bound proposed in Section 6. Its objective is to state the exception accurately and to fix its extent, since an account that concealed its clearest counter-example would be worth little. It proceeds in four steps: the instrument through which legal systems reach the interpretive site; the respect in which the instrument constitutes a genuine exception to Proposition 6.1; the three limits that confine it; and the condition under which an exception of this form is available at all. The method is examination of one institutional device against the definition of Section 3.

7.1 The Device of Reason-Giving

Legal systems require the bodies that determine matters to exhibit the reasoning by which they determined them. Judgments are published, grounds are stated, dissenting and concurring opinions record the disagreements that the disposition conceals, doctrinal commentary works over the reasoning afterward, and later bodies may depart from it or overrule it. The requirement is old, it is enforced through appeal, and its rationale is understood: a reason offered for a decision commits the body offering it beyond the case at hand, since a reason has a generality that a disposition lacks, and the commitment can be held against the body later (Schauer, 1995).

The device is remarkable in the terms of this paper. The interpretive site, in the ordering of Table 1, is where the readings under which a rule is applied to a person are formed, and such readings are ordinarily unavailable to an outside party. Legal systems have installed an instrument that reaches it.

7.2 The Extent of the Exception

The instrument does something that Proposition 6.1 appeared to exclude. It leaves aside any site that was already recordable, and requires the occupant of the site to externalize what occurs there, and by requiring the externalization it converts the site into one that a party who was absent can examine. The proposition therefore stands in need of qualification, and the qualification is stated here rather than left to a reader.

ExternalizationAn installed instrument may reach an otherwise unrecordable site where it requires the occupant of that site to state what occurs there, and where the statement can be held against the occupant afterward. The instrument does not record the site; it produces a record by imposing a duty of exhibition, and the bound of Proposition 6.1 applies to what an instrument can observe, and leaves open what it can require to be said.

The qualification strengthens the account, since it identifies the only route by which an instrument reaches beyond the first row of Table 1, and the conditions of that route turn out to be demanding.

7.3 The Limits of the Device

Three limits confine the exception, and together they explain why the device has not been generalized.

The first is the realist objection and it is old. A stated reason need not be the operative one. The device secures that a justification is offered and open to examination, and it does not secure that the justification is what moved the body. The point recurs in the finding that courts come to accept an organization’s own compliance structures as evidence of the thing those structures were introduced to secure (Edelman, 2016): the reasoning is exhibited, examinable and deferential to a construction made elsewhere.

The second concerns who can be required to speak. The duty falls on an office, in respect of a decision, at a time. The interpretive dispositions of a community are held by no office, attach to no decision and can be assigned no time, and there is accordingly nobody on whom a corresponding duty could fall. This is why the device stops at the second row of Table 1 and why nothing analogous is available at the cognitive and symbolic sites.

The third concerns what a reason can be given for. Reasons are given for a disposition reached under a description. The description under which a person was considered can be exhibited; the description under which he was never considered leaves nothing for anyone to state, since the body was unaware of an absence. The categorial site is therefore beyond the device by its structure, and the point has been made in adjacent literatures: courts confronted with a plurality of normative worlds select among them and suppress the others in the act of deciding (Cover, 1983), and a person whose experience has no available description cannot bring the experience forward for a body to consider (Fricker, 2007). Interpretation understood as an ongoing collective work (Dworkin, 1986) is compatible with all of this, since the work proceeds within the descriptions the practice makes available.

7.4 The Condition of Availability

The exception is available where three things hold together: an occupant of the site who is an office, a decision that the office must justify, and a duty of exhibition enforceable by a body positioned above it. Legal adjudication satisfies all three, which is why it is the setting in which the device developed. Organizations satisfy the first two intermittently and the third seldom. Communities holding an unpronounced distinction satisfy none of them, since there is no office, no decision and nobody who could be required to state what is not stated by anyone.

7.5 Findings and Residue

This section has identified an instrument by which legal systems reach a site that Section 6 held to be beyond installation, has restated the bound to accommodate it, and has set out the three limits that confine it to the reasoning of offices about decisions. What it leaves open is whether the device can be extended downward by any means, and the question is listed in Section 10.

8. The Contribution of Generative Relational Being, Stated at Its Weakest

This section states what the framework introduced in Section 1 adds to the accounts surveyed in Section 2, and states what it must concede. Its objective is to price the contribution accurately, since a framework whose contribution is overstated in a preliminary paper is difficult to correct later. It proceeds in four steps: the positive characterization of the residue and its distance from the negative characterizations already available; a distinction the characterization supports, between the finality of a case and the closure of a question; one legal form in which an aim of capacity type is carried without conversion into a stock of documents; and the concessions, of which the most serious concerns the absence of an observer. The method is comparison of characterizations, followed by examination of two legal forms against the account.

8.1 The Residue Characterized Positively

Each account surveyed in Section 2 identifies something that legal form fails to carry, and each characterizes it by what fails. The reflexive tradition speaks of the internal logic of a sphere with which regulation is incongruent (Teubner, 1987). The colonization thesis speaks of communicative processes displaced by a medium (Habermas, 1987). The audit literature speaks of what falls outside auditability (Power, 1997). In each case the residue is identified through the operation that misses it.

Generative Relational Being permits a different characterization, and it is the only contribution this section claims.

The residue characterizedWhat juridical translation fails to carry is the capacity of a set of relations to generate a further determination of a matter that some determination has settled. The instruments record the determinations. The capacity to produce another one, on the same matter, in the light of what has since been understood, is what fails the recordability condition, since its presence at a given moment consists in what could be done and not in what was done.

The characterization has three properties that recommend it over the negative forms. It is stated in terms of the object, where the negative forms are stated in terms of the operation that misses it, so it can be examined by someone who rejects the whole apparatus of this paper. It applies uniformly across the sites of Table 1, since the capacity in question can be extinguished at the level of rules, of interpretations, of available categories or of an unpronounced distinction, and the resulting condition is the same in each case. And it explains why the residue is unrecordable, which the negative characterizations assert: a capacity is a modal property, its presence is compatible with any particular record, and its absence is visible only when a case arises that requires it, at which point the absence is discovered, no record of it having been made beforehand.

8.2 The Finality of a Case and the Closure of a Question

The characterization supports a distinction that legal systems already draw and do not treat as a single subject. The finality of a case is a legal achievement: a matter once determined between parties is settled, appeals are exhausted, and the parties may order their affairs. The closure of a question is a different condition, in which the matter can no longer be raised at all, by anyone, on any occasion.

Legal orders separate these in several places. A judgment binds the parties while the rule it applied remains open to reconsideration in a later case; systems of precedent provide for departure and overruling; extraordinary review permits a concluded matter to be reopened on defined grounds; and legislative revision remains available throughout. The separation is one of the more impressive achievements of legal form, and it shows that the maintenance of a capacity to determine a matter again is not foreign to law.

The difficulty is the one this paper has been describing. The mechanisms just listed operate on recordable objects: a judgment, a rule, a ground of review, a statute. Closure at the interpretive, categorial and symbolic sites has no such object. A question closes there when no one can any longer be heard on it, and nothing is enacted, published or filed when that happens.

8.3 A Form That Carries a Continuing Burden

One legal form is worth examining because it comes closer than the others. Where a state is responsible for an internationally wrongful act, the obligations arising include, beyond cessation and reparation, the offering of appropriate assurances and guarantees of non-repetition (ILC, 2001), and the corresponding principles on remedy for gross violations treat guarantees of non-repetition as a distinct element of reparation, with content extending to institutional change (UN, 2005). Structural remedies in domestic public law have a similar shape, where a court retains jurisdiction over an institution and supervises a course of reform (Chayes, 1976), and the experimentalist account of such remedies describes their function as the destabilization of a settlement that has ceased to correct itself (Sabel & Simon, 2004).

Three features of these forms are pertinent. The burden is continuing, and a payment does not discharge it. It rests on the party that holds the capacity in question, which places the assessment inside the relation. And in the strongest versions the measure lapses unless it is re-justified, which places the burden of maintaining it on the party that maintains it (Ashworth & Zedner, 2014). That third feature is the institutional counterpart of the condition this paper’s characterization identifies, since a regime under which a settlement must be re-justified is a regime that has kept a question open.

The reservation is immediate and should be stated in the same place. Supervision of this kind is itself administered through reports, benchmarks and compliance findings, and the literature on structural remedies records the familiar movement toward the documentary. The form comes closer, and it does not escape.

8.4 The Concessions

Four concessions are required, and the first is decisive.

The characterization supplies no observer. Whether a set of relations retains the capacity to generate a further determination is a question about what could occur, and the parties best placed to judge it are the parties within the relation, whose judgment is itself formed under the conditions in question. A tribunal is by construction an observer positioned outside, which is what makes it impartial and what makes it unable to assess this. The difficulty here is the same one this paper has attributed to the instruments, arriving now at its own door, and better evidence does not reach it.

The characterization supplies no procedure. Nothing in this paper indicates how the capacity would be assessed in a particular organization at a particular time, and the absence is not an oversight to be repaired in a later section.

The characterization supplies no remedy. Even where the condition is agreed to obtain, nothing follows about what anyone is obliged to do, and nothing in the framework generates such an obligation.

And the framework contributes nothing to the theory of attribution. The question of who is to bear a legal consequence for a harm remains where the law has left it, and the difficulties of attributing responsibility for conditions produced by many hands are addressed by existing bodies of work, including the distinction between a liability model and responsibility grounded in participation in structural processes (Young, 2011), and the correlative structure that private law imposes on any claim it recognizes (Weinrib, 1995). Nothing said here disturbs either.

8.5 Findings and Residue

This section has offered a positive characterization of what translation fails to carry, has connected it to a distinction the law already draws, has identified one form in which a continuing burden is carried, and has conceded that the characterization has no observer, no procedure, no remedy and no bearing on attribution. What follows is an account of what this implies for legal theory and of the limits within which the whole is to be read.

9. Consequences for Legal Theory and the Limits of the Account

This section states what follows from the argument for the assessment of legal instruments, and states the limits within which the argument holds. Its objective is to prevent two misreadings, one that takes the account for a criticism of legal form and one that takes it for a reason to relax the enforcement of the instruments it discusses. It proceeds in four steps: the standing of the requirements that generate the transformation; what follows for expectation and design; the risk of quietism and the answer to it; and the limits, including the conditions under which the account would be unfalsifiable and the severability of its parts. The method is a statement of consequences, followed by a self-assessment against the standard set in Section 1.

9.1 The Standing of the Requirements

The requirements that generate the transformation are achievements and should be described as such. That a determination must rest on grounds that were stated in advance, that a party must be able to contest the evidence against him, that a decision must be justified in terms applicable to others, and that a body must answer to someone for what it has done, are among the more valuable things that human societies have constructed. They exist because determinations made otherwise are made arbitrarily, and the historical record of arbitrary determination is available to anyone who wants it. The argument of this paper is that these achievements fix the class of objects a legal determination can take, and that a certain class of aims falls outside it. The requirements are constraints on the form of legal ordering and are silent about the content ordered (Fuller, 1964), and the standing objection to any account proceeding by constraints alone, that such constraints are compatible with substantive iniquity (Hart, 1965), applies to the present account as much as to its target. The exclusion is a consequence of the achievement and is a property of the pair.

The same holds for the second condition. Answerability to an external party is what prevents a body from becoming the sole judge of its own conduct, and the settings in which generative aims survive best are the settings in which that protection is weakest. Section 5 declined to order these considerations, and the refusal is maintained here.

9.2 Consequences for Expectation and for Design

Three consequences follow, and all three concern expectation.

An instrument addressed to an aim of capacity type should be assessed by what it can secure, which is the recordable fragment of the aim, and its documented success at that should not be read as evidence about the aim. Where a programme of this kind fails to alter the condition it addressed, the failure is predicted by the account and supplies no evidence that the aim was mistaken, that those who pursued it were naive, or that the condition does not exist. This consequence is worth stating because the opposite inference is common, and because programmes are frequently abandoned on the strength of it.

A designer choosing between an eventful and a capacity-type formulation of an aim should expect the eventful formulation to be secured and the other to be recorded. Where an aim admits of an eventful formulation without loss, that formulation is to be preferred on the account given here, and where it does not, the residue should be identified and stated, so that it does not disappear into the instrument.

And the assessment of a body’s condition should not be conducted through the materials the body produces for its constituency, since those materials are selected by the property this paper has been discussing. What might be used in their place is an open question and is listed as such in Section 10.

9.3 The Risk of Quietism

An account holding that legal instruments cannot secure a class of aims can be enlisted by anyone who would prefer that no instrument be introduced, and the paper should say what it does and does not support.

It supports no relaxation of any existing instrument. The instruments addressed to eventful objects are, on the argument of Section 4, the ones that work, and nothing here bears on the case for enforcing them vigorously. It supports no inference from the difficulty of securing an aim to the conclusion that the aim is not owed. And it supports no claim that organizations which produce records are thereby acting in bad faith; the argument has been throughout that the movement toward the recordable occurs among people acting in good faith under a requirement to answer.

What the account does support is a reallocation of expectation, and a corresponding refusal to treat the record as the thing the record was introduced to stand for.

9.4 Limits

The limits are stated in the order of their seriousness.

The account has no procedure and no observer, as Section 8 conceded, and it therefore cannot be applied to a case. It is an account of why a class of instruments behaves as it does, and it is not a method for anyone.

The migration hypothesis of Section 6 is weakly supported, and its evidentiary situation is structurally poor. A reader who rejects it may keep Sections 3 to 5, which stand independently. The reverse severance is unavailable, since the hypothesis depends on the bound and the bound depends on the definition.

The account risks unfalsifiability in one specific way. If every observed failure of a compliance regime is attributed to migration into the unrecordable, the account explains everything and predicts nothing. Two things prevent this, and both are load-bearing: the boundary of Section 4, which identifies the class of cases in which the account predicts that instruments will succeed; and the refutation conditions of Sections 4 and 6, which are stated in terms of observations that could be made. A reader who finds those conditions too weak has identified the paper’s principal vulnerability.

The account gives no ground for choosing between institutional arrangements, since it declines to order the considerations that bear on the choice, and a reader looking for a recommendation will not find one. The refusal is a commitment of the framework and not a gap in the argument.

Finally, the treatment of prior work is selective, and the selection has been made at the points where the argument required it. The literatures on juridification, on audit, on decoupling and on reflexive law are each larger than the use made of them here, and the summary in Section 2 is not a survey.

10. Open Questions and a Note on Method

This section records the questions the argument leaves open, grouped by the part of the argument they bear on, and closes with a statement of what has been done and what has been left to further work. Its objective is to make the paper’s incompleteness usable by others.

10.1 Questions on the Selection Principle

Whether recordability is the single property on which translation selects, or whether it is one of several, of which repeatability across cases and attributability to an author may be independent members.
Whether the relativity of recordability to a checking position can be given an ordering, such that positions could be compared by what they can check, and whether such an ordering would be partial.
Whether legal systems differ systematically in the severity of the selection, and whether systems that place greater weight on the reasons given for a determination preserve more of the interpretive site than systems that place greater weight on the disposition reached.

10.2 Questions on the Boundary

Whether the three conditions of Proposition 4.1 are independent, and whether the third can be satisfied by a party whose position is altered only in respect of a capacity.
Whether the graded region can be characterized more precisely than by the presence of an eventful component, and whether the devices by which harassment law constructs an eventful object out of a course of conduct can be extended without limit.
Whether an aim of capacity type can be decomposed into eventful components without residue in any case, and what the residue consists in where it cannot.

10.3 Questions on Answerability and Migration

Whether the intensity of answerability can be measured independently of the transformation it is invoked to explain, without which Claim 5.1 risks circularity.
Whether any institutional arrangement makes the affected party the constituency to which a body answers, and what such an arrangement would require.
Whether migration, if it occurs, is bounded, or whether conduct continues to relocate as instruments extend to further sites.
What could be used in place of self-produced records to assess the condition of a body, given that the parties within the relation are the only available observers and are formed under the conditions to be assessed.

10.4 Questions on Generative Legality and the Law-Forming Process

Whether the requirement that a legal system retain the capacity to take up experience it has not yet accommodated can be stated as a formal condition of legality, alongside the conditions concerning generality, publicity, prospectivity, clarity, consistency, possibility, constancy and congruence.
Whether such a condition can be stated in terms that resist proceduralization, given that a system may provide for revision while the standing interpretation absorbs every attempt to invoke the provision, and whether the appropriate test is therefore whether the standing interpretation of a party can be defeated by any act available to that party.
Whether a condition of that kind could be installed at all, or whether the argument of Section 6 entails that a formal condition of legality may be statable and unenforceable at once.
Whether the account given here bears on the legitimacy of law-forming processes, and how it stands to accounts that ground legitimacy in the democratic genesis of legal norms.

10.5 A Note on Method

The argument is conceptual and has been left in prose. It rests on a definition, a claim about what legal determination requires, a proposition about the conditions of a preserving translation, a claim about the source of the pressure, a bound derived from what installation requires, and a hypothesis about what follows installation. Of these, the definition and the bound are close to analytic, the two claims are supported by the empirical literatures cited, the proposition is offered for counter-example, and the hypothesis is offered for refutation.

Two lines of further work are indicated and neither is undertaken here. The first is empirical and would test the comparative prediction of Section 4 directly, by assembling programmes of both kinds and examining the relation between their recorded measures and independent assessments of their aims. The second would formalize the ordering of sites in Table 1, together with the operation of an installed instrument upon them, which would give the migration hypothesis a statement precise enough to be modelled and would expose whatever the prose formulation is concealing. Both would test claims that are here argued alone.

References

Ashworth, Andrew, and Lucia Zedner (2014). Preventive Justice. Oxford: Oxford University Press.

Ayres, Ian, and John Braithwaite (1992). Responsive Regulation: Transcending the Deregulation Debate. New York: Oxford University Press.

Black, Julia (2001). Decentring Regulation: Understanding the Role of Regulation and Self-Regulation in a `Post-Regulatory’ World. Current Legal Problems, 54(1), 103–146.

Bromley, Patricia, and Walter W. Powell (2012). From Smoke and Mirrors to Walking the Talk: Decoupling in the Contemporary World. Academy of Management Annals, 6(1), 483–530.

Campbell, Donald T. (1979). Assessing the Impact of Planned Social Change. Evaluation and Program Planning, 2(1), 67–90.

Chayes, Abram (1976). The Role of the Judge in Public Law Litigation. Harvard Law Review, 89(7), 1281–1316.

Cover, Robert M. (1983). The Supreme Court, 1982 TermForeword: Nomos and Narrative. Harvard Law Review, 97(1), 4–68.

Dorf, Michael C., and Charles F. Sabel (1998). A Constitution of Democratic Experimentalism. Columbia Law Review, 98(2), 267–473.

Dworkin, Ronald (1986). Law’s Empire. Cambridge, MA: Harvard University Press.

Edelman, Lauren B. (1992). Legal Ambiguity and Symbolic Structures: Organizational Mediation of Civil Rights Law. American Journal of Sociology, 97(6), 1531–1576.

Edelman, Lauren B. (2016). Working Law: Courts, Corporations, and Symbolic Civil Rights. Chicago: University of Chicago Press.

Espeland, Wendy Nelson, and Michael Sauder (2007). Rankings and Reactivity: How Public Measures Recreate Social Worlds. American Journal of Sociology, 113(1), 1–40.

Espeland, Wendy Nelson, and Mitchell L. Stevens (1998). Commensuration as a Social Process. Annual Review of Sociology, 24, 313–343.

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

Fuller, Lon L. (1964). The Morality of Law. New Haven: Yale University Press.

Fuller, Lon L. (1978). The Forms and Limits of Adjudication. Harvard Law Review, 92(2), 353–409.

Habermas, Jürgen (1987). The Theory of Communicative Action, Volume 2: Lifeworld and System. Trans. Thomas McCarthy. Boston: Beacon Press.

Hart, H. L. A. (1965). Review of Lon L. Fuller, The Morality of Law. Harvard Law Review, 78(6), 1281–1296.

Hood, Christopher (1991). A Public Management for All Seasons? Public Administration, 69(1), 3–19.

International Law Commission (2001). Draft Articles on Responsibility of States for Internationally Wrongful Acts, with Commentaries. Yearbook of the International Law Commission, vol. II, Part Two.

Luhmann, Niklas (2004). Law as a Social System. Trans. Klaus A. Ziegert. Oxford: Oxford University Press.

Merry, Sally Engle (2016). The Seductions of Quantification: Measuring Human Rights, Gender Violence, and Sex Trafficking. Chicago: University of Chicago Press.

Meyer, John W., and Brian Rowan (1977). Institutionalized Organizations: Formal Structure as Myth and Ceremony. American Journal of Sociology, 83(2), 340–363.

Moore, Sally Falk (1973). Law and Social Change: The Semi-Autonomous Social Field as an Appropriate Subject of Study. Law & Society Review, 7(4), 719–746.

Muller, Jerry Z. (2018). The Tyranny of Metrics. Princeton: Princeton University Press.

Nonet, Philippe, and Philip Selznick (1978). Law and Society in Transition: Toward Responsive Law. New York: Harper & Row.

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

Power, Michael (1997). The Audit Society: Rituals of Verification. Oxford: Oxford University Press.

Sabel, Charles F., and William H. Simon (2004). Destabilization Rights: How Public Law Litigation Succeeds. Harvard Law Review, 117(4), 1015–1101.

Schauer, Frederick (1995). Giving Reasons. Stanford Law Review, 47(4), 633–659.

Scott, James C. (1998). Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed. New Haven: Yale University Press.

Strathern, Marilyn, ed. (2000). Audit Cultures: Anthropological Studies in Accountability, Ethics and the Academy. London: Routledge.

Teubner, Gunther (1983). Substantive and Reflexive Elements in Modern Law. Law & Society Review, 17(2), 239–285.

Teubner, Gunther (1987). Juridification: Concepts, Aspects, Limits, Solutions. In Gunther Teubner (ed.), Juridification of Social Spheres. Berlin: de Gruyter, 3–48.

Teubner, Gunther (1998). Legal Irritants: Good Faith in British Law or How Unifying Law Ends Up in New Divergences. Modern Law Review, 61(1), 11–32.

United Nations General Assembly (2005). Basic Principles and Guidelines on the Right to a Remedy and Reparation for Victims of Gross Violations of International Human Rights Law and Serious Violations of International Humanitarian Law. A/RES/60/147.

United Nations (2011). Guiding Principles on Business and Human Rights: Implementing the United Nations `Protect, Respect and Remedy’ Framework. HR/PUB/11/04.

Weinrib, Ernest J. (1995). The Idea of Private Law. Cambridge, MA: Harvard University Press.

Young, Iris Marion (2011). Responsibility for Justice. New York: Oxford University Press.