The Governance of Co-Experience Transmission - A Generative Relational Theory of Institutional Design under Interpretive Asymmetry

Transcript

Abstract

An account of a shared experience is composed by one participant in a relation and used by a third party who took no part in it. An earlier paper in this series established how positive law allocates authority over such an account and stated the conditions any defensible allocation would satisfy, proposing no regime. This paper asks whether an apparatus could administer those conditions, at what cost, and with what failure modes. It argues that only process requirements are available, since an output standard would demand a comparison between an account and the experience it reports that the ontology of the object denies, and that this is the limiting case of a condition under which regulatory theory independently recommends prescribing process rather than output. Five mechanisms are examined against the inherited specification: a duty to solicit and consider the other participant’s rendering before issuing, a forum operating before an account is relied upon, preservation of the materials a later reopening would require, expiry of the entitlement to rely, and restriction of exchange. The last is disqualified by the paper’s own definition of a mechanism, since the non-occurrence of a private exchange cannot be established; two lines of the specification are reached by nothing available. Three findings run against the proposals. The feature that makes these mechanisms resistant to ceremonial adoption, namely that discharge requires an act by the party they protect, is the feature that places the work upon that party; the cost falls on every occasion of issuance while the benefit falls on the few contested; and raising what is required of an issuer may contract the practice, in a direction that is documented and by a magnitude that is not, falling hardest on those with fewest relations from which an account might be drawn. That result exposes an inconsistency internal to the framework rather than a conflict between the paper’s two routes, and two of the framework’s criteria are accordingly restated as matters of degree rather than as thresholds.

Keywords: institutional design; co-experience; process regulation; contestation; compliance cost; ceremonial compliance; references.

Discussion Paper Note

This paper is a preliminary discussion paper intended to share an evolving idea and invite further dialogue, criticism, revision, and independent development. Its definitions, distinctions, and constructions remain provisional. Circulation across scholarly and practical communities is part of the purpose of releasing the manuscript at this stage.

The author treats the viewpoints, concepts, and lines of reasoning presented here as contributions to a shared field of inquiry. Similar or related ideas may have appeared in other intellectual, cultural, and disciplinary traditions. The manuscript therefore states its known antecedents, separates the researcher-origin proposal from later formal reconstruction, and leaves historical priority open pending a systematic originality review.

The arguments should be understood as provisional and historically situated. Readers are encouraged to question, test, revise, extend, reinterpret, or independently develop the ideas presented here. Where appropriate, acknowledgment of this paper as one point of encounter in the development of a related idea is appreciated. Such acknowledgment records an intellectual route; the ideas themselves remain available for criticism, revision, and independent development.

Responsible Use and Rights Reservation

This section separates requested scholarly conduct from the legal permissions stated on the following page. It records an ethical request for responsible use and then defines the narrower scope of retained legal rights.

The author encourages good-faith discussion, criticism, independent inquiry, and responsible use of the material in this work. Separately from the licence’s terms, the author asks users to consider foreseeable harms when adapting or applying the arguments made here. Two misuses are foreseeable and both invert the paper’s purpose.

The first concerns the finding, reported in Section 7, that issuers may withdraw from the practice when exposure to liability rises. That finding is offered as a constraint on the design of a remedy. It is not an argument that the present arrangement should be left undisturbed, and it should not be cited in support of that conclusion. A design constraint identifies what a remedy must avoid; it establishes nothing about whether a remedy is owed.

The second concerns the finding, reported in Section 6, that internal review structures are frequently adopted ceremonially and may attract deference on the strength of their existence alone. That finding is offered as a condition any forum must satisfy. It is not an argument that no forum should be established, and a reader who takes it that way leaves the person described with less recourse than the paper’s analysis says they are owed.

The author reserves all rights not granted by the licence stated in the notices below.

Notices

Status.

Preliminary discussion paper. Not peer reviewed. Not submitted for publication at the time of circulation.

Licence.

This work is made available under the Creative Commons Attribution–NonCommercial 4.0 International licence (CC BY-NC 4.0).

Language-model statement.

The author used a large language model as a drafting and verification aid. All argumentative claims, all doctrinal readings, and all decisions about scope and concession are the author’s own. Every source cited was verified against a publisher, journal, statutory, judicial, or institutional record before it was admitted to the manuscript.

Declared interest.

The author has been the subject of institutional judgments of the kind this paper examines. The paper’s object is general and its arguments are stated for that general object. The author’s own case is not a source of evidence in the paper and is not discussed in it.

Related research programme.

This paper belongs to a series on trust, neutrality, and the transmission of shared experience. It takes as settled the ontology of the judged subject, the duties of the issuing party, and the allocation of rights over an account of a shared experience, each established earlier in that series, and it takes from the last of these the specification set out in Section 5. It cedes distribution and the unequal conditions under which accounts are obtained to the injustice paper, and circulation, accumulation, and exchange to the political-economy volume.

Suggested citation.

Huang, Wanhong. The Governance of Co-Experience Transmission: A Generative Relational Theory of Institutional Design under Interpretive Asymmetry. Preliminary discussion paper, 2026.

Introduction

An earlier paper in this series established how positive law allocates authority over an account of a shared experience, and argued that the allocation is consistent across provisions enacted in different systems for different purposes: the person described is granted some measure of access to an account of themselves and is withheld authority over its content. That paper stated the conditions any defensible allocation would have to satisfy and proposed no regime. This paper takes up what it left: whether an apparatus could administer those conditions, what it would cost, and how it would fail.

The object is unchanged and is restated in Section 2. One participant in a working, instructional, or therapeutic relation writes an account of the other for a third party who did not participate in it and who will decide something about the person described. The experience characterised was produced by both parties; the characterisation is produced by one; and the person described is ordinarily absent from the transaction. The governing difficulty for institutional design follows from that structure and gives this paper its subtitle. Where two parties produced an experience and one of them renders it, authority over what the experience amounted to is held asymmetrically, and an apparatus governing such instruments cannot appeal to a standard of correctness by which the rendering could be checked, since no position among the parties yields a view from outside.

This constraint is more than an inconvenience, and one body of regulatory theory has already identified it as a criterion of institutional form. Management-based regulation, which directs regulated parties to conduct a specified process rather than to achieve a specified output, is argued to suit conditions in which regulated entities are heterogeneous and regulatory outputs are difficult to monitor (Coglianese and Lazer 2003). Both conditions hold here in an acute form: issuers differ from one another in every respect that matters, and the output in question is an interpretation whose correctness cannot be monitored at all. The mechanisms examined in Section 6 are accordingly process mechanisms, and Section 5 states why an outcome standard is unavailable rather than merely difficult.

Five mechanisms are examined: duties attaching at the occasion of issuance, forums operating before an account is relied upon, preservation requirements on those who hold records, expiry and reopening of recorded interpretations, and restriction of exchange. They are assessed in Section 6 against the specification set out in Section 5, and Section 7 reports what each would cost and how each would fail. The paper proposes no regime and reaches no recommendation about adoption. Its claim is about what an apparatus meeting the specification would have to include, what it would demand of the parties, and which of its failure modes are documented rather than conjectured.

Much of the ground is owned, and the owners are conceded at the point each claim arises rather than in a closing note. Five concessions are large enough to state here, and the first two bear directly on whether the paper’s proposals are worth making at all.

First, a reform of this exact practice has already been attempted and has already failed. More than forty jurisdictions in the United States enacted statutes conferring immunity on employers who supply references in good faith, in order to restore candid issuance where liability had discouraged it. The leading study of those statutes finds them prevalent and irrelevant, reporting that they left the common law in place and had no effect on reference practices, which continue to consist largely of confirming dates and positions (Cooper 2001). An experimental test of the premise on which such statutes rest, assigning supervisors to conditions in which the protection of qualified privilege was made salient, returned largely null results (Walsh, Heighton, and Dingens 2023). Any proposal advanced here therefore stands in a domain where the obvious legislative instrument was tried at scale and did not work, and Section 3 states what follows for the mechanisms examined in Section 6.

Second, the most novel mechanism examined has a strong prior objection. Organisations adopt formal structures ceremonially, for the legitimacy the structures confer, and decouple them from what is actually done (Meyer and Rowan 1977). Grievance procedures in particular are constructed as symbolic responses to ambiguous legal obligation (Edelman 1992) and become institutionalised as the appropriate form of compliance irrespective of what they deliver (Edelman, Uggen, and Erlanger 1999); and courts have been shown to defer to such structures on the strength of their existence, with analysis of federal employment decisions finding that the mere presence of a grievance or evaluation procedure is increasingly taken as evidence that no wrong occurred (Edelman et al. 2011). A forum established to answer the difficulty identified in this series may therefore leave the person described worse placed than no forum, by supplying the institution with evidence of its own fairness. Section 6 meets this objection before proposing the mechanism, and states the conditions under which a forum escapes it.

Third, the regulatory strategy the paper works within is not its own. That enforcement should begin with the least interventionist instrument and escalate only on failure is the responsive-regulation argument (Ayres and Braithwaite 1992); that enforcement must be calibrated to the regulated party’s operating framework and to the institutional environment rather than to attitude alone (Baldwin and Black 2008); and that instruments work as complements rather than alternatives is the case for regulatory pluralism (Gunningham, Grabosky, and Sinclair 1998). The design principles associated with enduring self-governed institutions, including graduated sanctions, arenas for conflict resolution that are cheap to reach, and monitoring by parties accountable to those monitored (Ostrom 1990), together with their empirical reassessment (Cox, Arnold, and Villamayor-Tomás 2010), supply the general form of several mechanisms in Section 6. None of this literature addresses an instrument of unilateral interpretation, and the paper adds nothing to any of it.

Fourth, the position of the issuer has an established analysis. That liability may be placed on a private party who can prevent a harm by withholding cooperation is the gatekeeping strategy (Kraakman 1986), and that such a party operates by pledging reputational capital accumulated across many occasions to vouch for one is the account of the reputational intermediary (Coffee 2006). The distinction between a licensing regime, which bars practice, and a certification regime, which supplies a signal without barring it, is likewise established (Kleiner 2000). What the present paper takes from this is a vocabulary rather than a result.

Fifth, reform of this instrument has been proposed and in one domain implemented, and the proposals concern the first of the five mechanisms. Standardised reference forms achieve materially higher predictive validity than narrative letters (McCarthy and Goffin 2001); narrative letters in their current form correlate weakly with subsequent performance and add little beyond other material (Kuncel, Kochevar, and Ones 2014); requiring evaluators to elaborate and justify their ratings reduces the operation of bias (Morgan, Elder, and King 2013); and a standardised letter of evaluation has been in use in one medical specialty since the 1990s (Hegarty et al. 2014). Duties running to the quality of what is written are likewise established in the law of references (House of Lords 1994; Court of Appeal of England and Wales 1998). The mechanism examined in Section 6 is narrower than any of these and is what remains once they are conceded: a duty to seek and consider the other participant’s rendering before issuing.

Two features of the paper’s position follow from these concessions and are stated here because they govern how its claims should be read.

The paper argues along two routes and keeps them apart, as Section 4 sets out. The institutional route reports what existing regimes do and what has been observed of them. The framework route derives requirements from principles of relational reproduction stated in the earlier work. Section 8 reports where the routes converge, where a result rests on one alone, and where they conflict. The conflict expected here differs from the one reported in the preceding paper: there, positive law denied what the framework asserted. Here, the framework states requirements and the institutional record indicates that meeting them may produce effects the framework itself prohibits. Section 9 states what the framework must concede if that is so.

And the paper’s most consequential finding is one that runs against its own proposals. Where exposure rises, those who issue such accounts may cease to issue them, and the practice contracts. The evidence for this is stated at the strength it will bear in Section 7: the direction of the effect and the resistance of the practice to legislative correction are documented, while its magnitude is not, since no study isolates it. The nearest quantified analogue lies in professional practice under liability exposure, where a large majority of surveyed specialists in a high-exposure environment reported practising defensively and a substantial minority reported restricting their practice by declining complex cases (Studdert et al. 2005). If that pattern holds here, a governance regime that raises the cost of issuing will reduce the number of accounts written, and the parties least able to absorb that reduction are those with fewest relations from which an account might be drawn. The framework in which this paper works prohibits foreclosing the field of what a person may become, so a remedy with that effect is condemned by the same principles that motivate the remedy. Section 9 takes up the consequence, and this paper does not resolve it in favour of its own proposals.

Sections 2 and 3 fix the object and state the existing treatments. Section 4 states the routes and the conditions of disconfirmation. Section 5 assembles the inherited specification and argues that only process may be regulated here. Section 6 examines the five mechanisms and assesses them against the specification. Section 7 reports costs, burdens, and failure modes. Sections 8 and 9 report the standing of each result and what follows for the framework, and Sections 10 and 11 state the limits and conclude.

Two general limitations qualify every claim that follows. The institutional route draws on a bounded set of regimes and jurisdictions, so a finding that no regime does something is a finding about that set. And the paper assesses mechanisms without proposing one, which means it establishes what an apparatus would have to include and establishes nothing about whether such an apparatus should be built.

Background and Preliminaries

This section fixes what is to be governed and what would count as governing it, records the results inherited from earlier work in this series together with the sections that depend on each, and marks the questions reserved to later work. Its method is stipulative and expository: it defines and delimits, and defers every contested claim to the sections that argue for it. The section proceeds in four parts, describing the transaction and defining a mechanism of governance, bounding the domains, jurisdictions, and implemented regimes examined, listing the inherited results, and stating the reservations.

Transmission of Co-Experience in Institutional Selection

The transaction governed here has three parties and two moments. Two parties participate together in a working, instructional, or therapeutic relation, and what either does within it is conditioned by the other. At a later moment a third party, which took no part in that relation, must decide something about one of the first two and requests an account from the other. The first moment produces the experience; the second produces a characterisation of it and moves that characterisation outside the relation that generated it.

Definition 1 (The Parties To A Transmitted Account). The issuer is the party who composes the account. The person described is the party the account characterises and who participated in the relation from which it is drawn. The receiving institution is the party who requests, holds, or acts upon the account and who took no part in that relation.

Transmission is the second moment, and it is what an apparatus would govern. The first moment lies outside reach: no regime can require that a supervisory or instructional relation take one shape rather than another, and none is proposed here. What can be required is something of the occasion on which an account is composed, of the terms on which it moves, of the parties who hold it afterwards, and of the period for which it may be relied upon.

The form such requirements may take is constrained at the outset, and the constraint is what distinguishes this domain from most others in which institutional design is discussed. An apparatus governing the manufacture of a product may specify the product; an apparatus governing an emission may specify a threshold. Here the thing produced is an interpretation of an experience that two parties made, and no position among them yields a standpoint from which its correctness could be established. A requirement that an account be accurate is therefore unenforceable, not because verification would be costly but because the standard it invokes is unavailable. What remains available is the occasion, the procedure, and the record.

Definition 2 (A Mechanism Of Governance). A requirement attaching to a determinate party at a determinate occasion, whose discharge can be established without assessing the correctness of any interpretation.

Every mechanism examined in Section 6 satisfies this definition, and the definition is what excludes the mechanisms a reader might otherwise expect. A requirement that issuers write accurately, a review of accounts for soundness, and a penalty for a judgment later shown to be mistaken each presuppose the standard that Section 5 argues is unavailable. Their exclusion follows from the object rather than from a preference for light regulation.

Scope of the Domains and Jurisdictions Considered

The instruments examined are those in which the three features above hold: the letter of recommendation in academic and professional selection, the employment reference, the performance appraisal used in promotion and retention, the peer review report, the clinical formulation transmitted to a party other than the patient, the teacher evaluation entering a pupil’s record, and the supervisory assessment entering a licensing or disciplinary process. Assessments composed from records by a party who never encountered the person fall outside, since no relation was jointly produced and the question of divided authority does not arise.

Evidence is drawn from three kinds of source and the kinds are kept distinct. Legal provisions are read for what they require and permit, across the federal law of the United States and the law of its states so far as references and records are concerned, the law of England and Wales, and the law of the European Union. Scholarly findings about how institutions behave are treated as empirical claims with the samples and methods that produced them, and are reported with those samples. And implemented regimes are treated as cases: the standardised letter of evaluation adopted in one medical specialty in the 1990s and twice revised since (Hegarty et al. 2014) is the principal one, because it is the only regime located that governs an instrument of this kind, has operated for three decades, and has been studied.

Statements about what regimes require are statements about this set. A provision or a regime outside it that imposed a requirement absent here would qualify the findings of Section 6 without contradicting anything reported.

Results Carried from the Preceding Papers

Six results established earlier in this series are taken as settled and are not re-argued. Each is stated with the sections of this paper that rely upon it, so that a reader rejecting an inherited result can identify what falls with it.

The first is the ontology of the judged subject: an attribute exhibited within a relation depends for its exhibition upon that relation, so an observation is valid for the relation in which it was made and its extension beyond is an error of scope. Sections 5 and 6 rely on this wherever the reach of an account is distinguished from its correctness.

The second is the set of duties borne by the issuing party, which are duties appropriate to an exercise of discretionary power rather than norms of assertion, run primarily to the person described rather than to the party who asked, and make refusal the act requiring the stronger justification. Section 6 takes these as given and adds nothing to them.

The third is the allocation of rights over an account, and in particular the finding that positive law grants the person described access while withholding authority over content, and that access and interpretive authority are separately allocable. Section 6 relies on the separability at every point where a mechanism affects one without affecting the other.

The fourth is the specification the preceding paper produced and did not implement: criteria derived from the principles of relational reproduction, and conditions any forum would have to meet. Section 5 assembles both and this paper builds to them.

The fifth is the set of principles of relational reproduction themselves, together with two extensions: that the conditions making later revision possible must be preserved rather than merely permitted, and that where an act confers interpretive authority the capacity to contest it must be generated by the same act. Section 5 states these for the present object and Section 9 reports what must be surrendered among them.

The sixth is a pair of problems the preceding paper raised and left open. A forum reaching evaluative content requires a mode of participation in which reasons are answerable by reasons rather than a standard of correctness (Fuller and Winston 1978), and what such a forum should conclude was left undetermined. And the operative force of these instruments was found to have no description in positive law, with three conditions stated for any description that might be supplied. Section 5 carries both forward as constraints on design rather than as settled matters.

Boundaries Observed toward the Later Papers in the Series

Three bodies of question adjacent to this one are reserved, and the reservations are stated so that their absence is read as a boundary rather than an oversight.

Questions of distribution belong to the injustice paper: how unequally the capacity to obtain a favourable account is distributed, how differences in the conditions under which people generate observable performances produce differences in what may be said of them, and how the standing of those who issue such accounts functions where it is held unevenly. Section 7 enters this territory only to report on whom the cost of a mechanism would fall, and stops there.

Questions of circulation and accumulation belong to the political-economy volume: what accrues to those who issue, what is exchanged, and how a capacity to issue operates as a holding. Section 6 considers the restriction of exchange as a mechanism and takes no position on what exchange yields to whom.

Questions of conversion between heterogeneous systems of value, and of justice across domains, belong to the two later papers named for them. This paper governs one practice in one form and generalises to neither.

Existing Treatments of Institutional Design for Contested Accounts

This section identifies the existing owners of each claim the paper will make and states how far each owner goes, so that the assessment in Section 6 can be measured against what is already held. Its objective is concession rather than survey: each body of work is presented for the claim it forecloses to this paper, with its limit stated in the same paragraph. The section treats in turn the strategy of enforcement, the design of self-governed institutions, gatekeeping regimes, the analysis of reputational intermediaries, the recorded behaviour of internal review procedures, and the statutory reform already attempted upon this very instrument, closing with the boundary that remains. The method is textual and doctrinal for provisions, and for empirical findings reports the samples and methods that produced them.

Regulatory Strategy and the Ordering of Enforcement Responses

That a regulator should begin with the least interventionist instrument and escalate only upon failure, so that persuasion precedes deterrence and deterrence precedes incapacitation, is the responsive-regulation argument (Ayres and Braithwaite 1992). That the response must be calibrated to the regulated party’s operating framework, to the institutional environment, and to the differing logics of the available tools rather than to the party’s attitude alone is its later refinement (Baldwin and Black 2008). And that instruments function as complements rather than as exclusive alternatives, so that a regime combines command, economic incentive, information, and self-regulation, is the case for regulatory pluralism (Gunningham, Grabosky, and Sinclair 1998). Section 6 adopts the ordering and adds nothing to it.

One further account supplies something the others do not, and it bears on the form of every mechanism examined here. Management-based regulation directs a regulated party to conduct a specified process rather than to reach a specified output, and is argued to suit conditions in which the regulated entities are heterogeneous and the regulatory outputs are difficult to monitor (Coglianese and Lazer 2003). The account was developed for food safety, industrial safety, and environmental protection, where monitoring outputs is costly rather than impossible. The present domain is the limiting case of the same condition: what an account concludes about a person cannot be monitored for correctness at all, on the ontology inherited in Section 2. The criterion therefore selects process regulation here more strongly than in the cases for which it was formulated, and Section 5 states why.

Design Principles for Self-Governed Institutions

Institutions that govern shared resources durably tend to exhibit a set of recurring features, among them sanctions that escalate with the seriousness and repetition of a violation, arenas for resolving conflict that are cheap and quick to reach, and monitoring conducted by parties who are accountable to those whose conduct is monitored (Ostrom 1990). A synthesis of ninety-one subsequent studies finds the principles well supported while proposing to subdivide several of them, and records the standing observation that they describe structure rather than the processes of trust and legitimation that animate it (Cox, Arnold, and Villamayor-Tomás 2010).

Three of the mechanisms in Section 6 take their general form from these principles, and the paper claims no part of that form. What the principles were derived from is a class of cases in which the parties share an interest in the resource, interact repeatedly, and can observe each other. The transaction examined here has none of these features: the issuer and the person described may never interact again, the receiving institution is a stranger to both, and the whole difficulty is that one party cannot observe what the other has written.

Gatekeeping through Licensing, Accreditation, and Certification

Regimes that control entry to a practice are analysed along a distinction the paper uses throughout. Licensing bars practice to those who lack authorisation and is defended as quality assurance to those who receive the service, its principal cost being restricted supply with consequent effects on wages and prices; certification supplies a signal without barring practice and is the less restrictive alternative (Kleiner 2000).

The distinction transfers to the present object and the transfer settles one question in advance. A regime that barred the unqualified from issuing accounts of persons would restrict the practice at exactly the point where Section 7 finds it most fragile, and no such regime is examined in Section 6. The mechanisms considered attach requirements to the act of issuing rather than to the standing of the issuer.

Reputational Intermediaries and Third-Party Enforcement

That liability may be placed upon a private party who is positioned to prevent a harm by withholding cooperation, and that the strategy succeeds only where such a party can and will withhold reliably, is the gatekeeping analysis (Kraakman 1986). That a party of this kind operates by pledging reputational capital accumulated across many occasions in order to vouch for one of them is the account of the reputational intermediary (Coffee 2006).

Both describe the issuer’s position accurately, and the second explains why the instrument works at all. Their limit is the reform each proposes. The gatekeeping strategy operates by making the intermediary liable for what passes; the reputational analysis proposes reorienting the intermediary toward the party who relies. Neither addresses the party this series is concerned with, since in both the person assessed is the object of the transaction rather than a participant in it. The paper takes a vocabulary from this work and no result.

Internal Review Procedure and the Recorded Tendency toward Symbolic Compliance

The most serious objection to the mechanisms proposed in Section 6 comes from this literature, and it is stated here in full rather than answered.

Organisations adopt formal structures for the legitimacy those structures confer and decouple them from what is actually done, substituting confidence and good faith for inspection (Meyer and Rowan 1977). Where a legal obligation is ambiguous, organisations construct symbolic structures that come to stand for compliance independently of what they deliver, a finding drawn from a national survey of 346 organisations (Edelman 1992); grievance procedures in particular become institutionalised as the appropriate response to legal uncertainty, modelled on the public legal order and adopted because they are recognised rather than because they work (Edelman, Uggen, and Erlanger 1999). And courts have been shown to accept such structures as evidence of compliance: analysis of a random sample of 1,024 federal employment-discrimination opinions between 1965 and 1999 finds judicial inference from the presence of a grievance or evaluation procedure to the absence of discrimination, with that deference increasing over time (Edelman et al. 2011).

The consequence for a paper proposing a forum is direct and adverse. A mandated contestation procedure may be adopted, decoupled, and then relied upon by the institution that hosts it as evidence that the person described was treated fairly, leaving that person worse placed than before the procedure existed. Section 6 meets this objection at the point the mechanism is proposed, and states the conditions under which a forum escapes it.

Statutory Reform of Reference Liability and the Recorded Outcome

A reform of this exact practice has already been legislated across most of one country, and its recorded outcome is the most directly relevant evidence available to this paper.

The problem the legislation addressed was two-sided liability. An issuer who writes adversely faces exposure to the person described, whose remedies in defamation and negligence are well established and include a duty of care in the preparation of a reference (House of Lords 1994) and a requirement that the whole not mislead (Court of Appeal of England and Wales 1998). An issuer who writes favourably while omitting known misconduct faces exposure to the party who relies, third-party liability for a misleadingly positive reference having been imposed (Supreme Court of California 1997). Caught between the two, issuers withdrew, and practice converged on confirming dates and positions.

More than forty states enacted statutes conferring immunity on employers supplying references in good faith, in order to remove the disincentive. The leading study of those statutes reports that they left the surrounding common law in place and had no effect on practice, which continued to consist of neutral and no-comment responses (Cooper 2001). An experiment testing the premise on which the statutes rest, assigning 420 supervisors across conditions in which consent, immunity, and confidentiality were made salient, returned largely null results (Walsh, Heighton, and Dingens 2023). A legislature elsewhere reached the opposite conclusion and withdrew the subject’s access instead, exempting the confidential reference from data-subject rights in the hands of giver and recipient alike (Parliament of the United Kingdom 2018).

Two things follow, and they pull against each other.

The first is unfavourable to this paper. A targeted statutory intervention on this practice failed because it operated at a distance from the decision, leaving in place the surrounding exposure that actually governed conduct. Any mechanism proposed here is exposed to the same failure, and Section 6 states for each whether it operates on the occasion itself or at a distance from it.

The second distinguishes the cases and is the paper’s answer, stated here so that it may be tested against what follows. The immunity statutes were permissive: they removed a reason for silence and relied upon issuers to choose speech. A permissive reform fails whenever the parties decline the invitation, which is what the null experimental result records. The mechanisms examined in Section 6 are mandatory, attaching a requirement to an occasion rather than removing a disincentive from a choice, and a mandatory requirement does not depend on the issuer preferring to comply. The distinction is real and it is not free: raising what is required of an issuer raises the exposure that produced withdrawal in the first place. The escape from the first difficulty therefore feeds the second, and Section 7 reports the consequence.

Boundary of the Present Contribution

Three things follow from the survey in this section.

First, the general form of every mechanism examined here is owned. Escalation, graduated sanction, cheap conflict arenas, accountable monitoring, process requirements in place of output standards, and the analysis of the issuer as a reputational intermediary are all conceded above and again at the point each is used.

Second, one of the five mechanisms has direct prior owners in this domain. Standardised forms in place of narrative letters, elaboration requirements that reduce the operation of bias, and duties running to the fairness and accuracy of what is written are established, and are conceded in Section 6 at the point the mechanism is proposed.

Third, a search of the regulatory, organisational, and legal literatures conducted for this paper located no proposal for a forum operating before an account is relied upon, for preservation duties on those who hold such accounts, for the expiry of a recorded interpretation, or for restriction of exchange in such instruments. On the fourth of these the located practice runs the other way, since services that hold and forward such accounts store them without expiry. These four mechanisms are accordingly advanced as the paper’s own, subject to the qualification stated in Section 1 that a bounded search establishes the absence of a located owner rather than the absence of one.

Method, Routes, and Conditions of Disconfirmation

This section states how the paper argues, over what material, and what would show it to be wrong. Its objective is to make the standing of each later result inspectable, so that a reader can tell for any claim whether it rests on the institutional record, on the framework, or on both, and what finding would defeat it. The section describes the two routes and the reason for keeping them apart, records the cases and provisions examined and the criteria that selected them, explains how route dependence is marked, and sets out the conditions under which each principal claim would fail. Table [tab:routes] collects the last two together.

Institutional Evidence Alongside Derivation from the Framework

The paper argues along two routes and keeps them separate to the end.

The institutional route reports what existing regimes require and what has been observed of their operation. It reads provisions for what they demand of which party, treats a legislature’s considered choice as evidence of what a system has decided rather than as authority on what is right, and treats an empirical finding as a claim about the sample and method that produced it. Where a study is relied upon, its design is stated in the text, so that a reader may judge how far the finding travels to the present object. The route is descriptive in what it reports and evaluative only in assessing whether a stated legislative rationale succeeded on its own terms.

The framework route derives requirements from the principles of relational reproduction inherited in Section 2, applied to the object defined there. It asks what an apparatus would have to do if a shared experience is jointly produced in the sense the framework gives that phrase. The route is deductive relative to its premises and inherits whatever standing they have.

Separation matters more in this paper than in its predecessor, because the two routes are expected to conflict in an unusual way. The predecessor reported a contradiction: positive law denied what the framework asserted. Here the framework states requirements and the institutional record indicates that meeting them may produce an outcome the framework itself prohibits. A merged argument would obscure this, since the requirement and the evidence against satisfying it would appear as one undifferentiated body of reasoning rather than as a demand and a constraint that defeat one another.

Cases and Instruments Selected

The instruments are those defined in Section 2. Four bodies of material are examined and each was selected on a stated ground.

The standardised letter of evaluation adopted in one medical specialty is taken as the principal implemented case (Hegarty et al. 2014), on the ground that it is the only located regime that governs an instrument of this kind, has operated for three decades, and has been studied rather than merely described. It supplies Section 6 with something to assess besides criteria.

The statutes conferring immunity on those who supply references, together with the study of their effect (Cooper 2001) and the experimental test of their premise (Walsh, Heighton, and Dingens 2023), are taken as the principal case of attempted reform. They were selected because they address this practice directly and because their outcome is recorded, which is rare.

The law of references in England and Wales (House of Lords 1994; Court of Appeal of England and Wales 1998) and the treatment of the confidential reference in data-protection law (Parliament of the United Kingdom 2018) are examined as the two settled positions a legal system has reached on the subject’s standing, one imposing duties on the issuer and the other withdrawing access from the subject.

Findings about organisational behaviour are drawn from work on symbolic structures and their judicial reception (Edelman 1992; Edelman, Uggen, and Erlanger 1999; Edelman et al. 2011), on gaming under measurement, and on practice under liability exposure. These are findings about other domains and are used as evidence of mechanism rather than as evidence about this practice; Section 7 states in each case what would have to hold for the transfer to succeed.

Dependence of Each Result on Its Originating Route

Each principal claim is marked in the text with the route that produces it, and Section 8 collects the markings. Three conventions apply. A claim derived from the framework alone is stated as a consequence of the principles and identified as such where it is made. A claim resting on the institutional record alone is stated with the provisions or findings that carry it, and with the samples where the carrier is empirical. A claim reached by both is stated once, with the convergence noted, and is relied upon more heavily in later sections than claims of the first two kinds.

Inherited results carry the same discipline. The six results taken as settled in Section 2 were established elsewhere, and the sections depending on each are named there rather than here.

P3.7cmP2.3cmY Claim & Originating route & Finding that would defeat the claim
Availability of process requirements alone, an outcome standard being unavailable & Both & A standard for assessing an interpretation that does not require a position outside the parties
Exhaustiveness of the five mechanisms under the definition adopted & Framework & A requirement satisfying the definition in Section 2 that is none of the five
Escape of a forum from ceremonial adoption under stated conditions & Both & A forum satisfying those conditions that was nonetheless decoupled, or a decoupled forum that improved the position of subjects
Distinction in effect between mandatory requirement and permissive immunity & Institutional & A mandatory requirement upon issuance that failed for the reason the immunity statutes failed
Contraction of issuance where exposure rises & Institutional & A study isolating the effect of exposure upon rates of issuance and finding it absent or negligible
Incidence of compliance cost upon the party a mechanism protects & Institutional & Evidence that the burden of such a requirement rests upon the issuer or the receiving institution instead

Conditions of Disconfirmation

Table [tab:routes] states, for each principal claim, the finding that would defeat it. Three features of that list deserve comment, and the third distinguishes this paper from its predecessors in the series.

The first concerns the exhaustiveness claim, which is the most fragile of the six and the easiest to defeat. It asserts only that the five mechanisms exhaust the forms available under the definition adopted in Section 2, so a reader who rejects that definition is not answered by anything in Section 6. The definition is stated early for this reason.

The second concerns the two claims about forums and about mandatory requirements. Both would be defeated by a case rather than by an argument, and both accordingly depend on the bounded character of the survey recorded in Section 1.

The third has no counterpart in the earlier papers of this series. The claim about contraction of issuance would be defeated neither by discovery nor by argument but by measurement. What is documented is the direction of the effect and the resistance of the practice to a legislative correction; what is undocumented is the magnitude, since no study isolates the effect of exposure upon rates of issuance. The claim is therefore stated in Section 7 as a directional finding with an unmeasured magnitude, and a reader should treat it as a reason for caution in design rather than as a quantity that could be traded against the benefits of a mechanism. A paper that treated it as a quantity would be asserting something no available evidence supports.

Three kinds of claim are outside the paper’s scope and are therefore outside the table. The paper makes no estimate of how often the instruments examined are issued or contested, since it argues about the form of a requirement rather than about incidence. It reaches no recommendation about whether any mechanism should be adopted, and identifies the conditions under which each would operate rather than the case for operating it. And it proposes no complete regime, so questions of institutional architecture, staffing, and jurisdiction are left open.

The Specification Inherited from the Preceding Papers

This section assembles what an apparatus would have to satisfy. Its objective is to convert the criteria and conditions produced by earlier work in this series into a specification against which Section 6 can be assessed, and to establish what form such an apparatus may take. It states the criteria derived from the principles of relational reproduction, the conditions any forum would have to meet, the two problems the earlier work raised and left open, and the argument that only process may be regulated in this domain. Its method is restatement for the first three parts and derivation for the fourth. Table [tab:spec] collects the specification.

One point of continuity governs the whole section. Two of the principles were amended by the paper that produced this specification, and the criteria in this section are stated in the amended form rather than the original. Where the amendment matters to a mechanism, Section 6 says so.

Criteria Derived from the Principles of Relational Reproduction

Six criteria follow from the principles, one from each.

An apparatus satisfying the principle of co-reproduction gives each producer of a shared experience standing to participate in settling its interpretation, or supplies some substitute where a single author is unavoidable. The amendment matters here: the principle as first stated spoke of shared authorship, which in law is an ownership concept and would confer a power to exclude. As amended, the standing carries no power to license, block, or take proceeds, and a mechanism satisfying the criterion therefore admits participation without creating any entitlement over the account produced.

An apparatus satisfying the principle of counter-power attaches the capacity to contest at the moment an instrument issues, rather than supplying it afterwards through a separate proceeding, since the authority is conferred by the act of issuing and what answers it must arise from the same act.

An apparatus satisfying the principle of historical revisability preserves the materials that a later reopening would require: the account, the identity of the party who composed it, the period it covered, and any divergence between the participants known at the time. The amendment matters here also. Revision of an evaluative interpretation cannot mean correction against what the experience really amounted to, since no such standard is available; it means supersession, in which a later interpretation displaces an earlier one. The institutional form the criterion calls for is accordingly the expiry of reliance rather than a right of correction.

An apparatus satisfying the principle of subject preservation keeps the person described within the relation whose product is being reported, rather than treating them as its topic while excluding them as a party.

An apparatus satisfying the principle of generativity leaves open the field of what the person described may still become, which bears upon the duration for which an account may be relied upon as much as upon its contents.

An apparatus satisfying the principle of non-exploitation refrains from treating either party as a means to the receiving institution’s convenience.

Conditions a Forum Would Require

Four conditions attach specifically to any body before which an account might be contested, and they are narrower and more demanding than the criteria above.

Such a forum operates before the account is relied upon rather than after a determination has been made, which follows from the counter-power criterion and is the condition most difficult to institutionalise. It reaches the evaluative content of an account and not merely the fidelity of the record, since a body able to confirm only that an assessor said what the file records leaves the interpretation untouched. It preserves the materials of reopening, in the sense just stated. And it admits the person described as a participant in the proceeding rather than as its subject.

P3.4cmY Source of the requirement & Requirement upon an apparatus
Co-reproduction, as amended & Standing for each producer to participate in settling the interpretation, carrying no power of exclusion
Counter-power, with the simultaneity extension & Capacity to contest generated by the act of issuance itself
Historical revisability, as amended & Preservation of account, author, period, and known divergence; supersession and expiry in place of correction
Subject preservation & Retention of the person described as a party rather than as a topic
Generativity & Limitation upon the period for which an account may be relied upon
Non-exploitation & Refusal to treat either party as a means to the receiving institution’s convenience
Forum condition, timing & Operation before reliance rather than after determination
Forum condition, reach & Access to evaluative content rather than to the record alone
Forum condition, materials & Preservation of what a reopening would require
Forum condition, standing & Participation by the person described rather than description of them

Problems Left Unresolved by the Jurisprudential Account

Two problems arrive with the specification and constrain what may be built to it. Neither was solved by the work that produced them, and neither is solved here.

The first concerns what a forum reaching evaluative content would decide. An account of adjudication is available on which what distinguishes it is the mode of participation afforded the affected party, namely participation through proofs and reasoned arguments, and on which a task may be adjudicated where the meaning of that participation survives being assigned to it (Fuller and Winston 1978). On that account a forum requires no external standard of correctness, and a party before it has no right to a particular outcome; what is required is that reasons offered be answerable by reasons. This removes the objection that an evaluative dispute cannot be adjudicated at all. It leaves open what the forum should conclude, and the residue is sharp: a body empowered only to let a later interpretation displace an earlier one, with nothing by which either may be assessed, registers whichever account was offered most recently rather than adjudicating. Section 6 states what a forum meeting the conditions above would therefore have to do, and Section 10 reports what remains unsettled.

The second concerns the operative force of these instruments, which was found to have no description in positive law, on the ground that the doctrines naming the force of a document name a force the legal system itself constituted. Three conditions were stated for any description that might be supplied: the force would have to be individuable, so that it could be said when two instruments have the same effect and when an effect has occurred; attributable to an act, so that a party could be identified as having produced it; and answerable by a remedy that is not terminal, since a remedy consisting only in prohibition removes the instrument rather than governs it. Every mechanism in Section 6 operates without such a description, and the third condition in particular constrains the mechanism examined last.

The Governing Constraint of Interpretive Asymmetry

The specification could in principle be met by requirements of two kinds. A requirement upon the output would specify what an account must be like, and would be enforced by comparing accounts against that specification. A requirement upon the process would specify what must be done on the occasion of composing, transmitting, or holding an account, and would be enforced by establishing whether it was done. Only the second is available here, and the reason is a feature of the object rather than a limitation of enforcement capacity.

An output requirement of the relevant kind would have to demand that an account be accurate, or sound, or fair to what occurred. Each formulation invokes a standard by which the account could be compared with the experience it reports. On the ontology inherited in Section 2, an attribute exhibited within a relation depends upon that relation, and the account is an interpretation of an experience both parties produced. Neither party occupies a position from which what the experience amounted to could be settled independently of the positions occupied. There is accordingly nothing for the account to be compared against, and a requirement of accuracy is unenforceable because its standard is unavailable rather than because verification would cost too much.

This is the condition under which one body of regulatory theory recommends prescribing process instead of output, namely that the regulated parties are heterogeneous and the outputs difficult to monitor (Coglianese and Lazer 2003). The present domain occupies the limit of that condition, and the recommendation therefore applies here more strongly than in the cases from which it was derived. Section 2 defined a mechanism accordingly, as a requirement whose discharge can be established without assessing the correctness of any interpretation, and Section 6 examines only requirements of that kind. What the restriction excludes is worth naming: a duty to write accurately, a review of accounts for soundness, and a sanction for a judgment later shown to be mistaken are all outside the available range.

A concession is owed before that consequence is drawn, and it is owed to a claim stronger than the one Section 3 recorded. Where verification is demanded, activity is reorganised into a form that can be verified, so that performance is formalised and made auditable rather than merely reported, and the apparatus of checking imposes its own values upon what it checks (Power 1997). This goes beyond the observation that adopted structures decouple from practice. It holds that the requirement of demonstrability alters the thing demonstrated. Applied to the mechanisms of Section 6, a duty whose discharge must be evidenced produces an artefact shaped to be filed, and the paper claims no part of this analysis.

One consequence of the restriction runs against the paper and is stated here because it shapes everything that follows. A process requirement is discharged by doing the specified thing, and whether the specified thing was done is observable, which is what makes such a requirement enforceable. It is also what makes it separable from any effect upon how the person described is actually treated, since an occasion may be conducted and a form completed while the disposition that produced the account remains what it was. The literature recorded in Section 3 describes precisely this separation and finds it to be the normal fate of adopted procedure. The constraint that makes governance possible in this domain is therefore the same constraint that makes it fragile, and Section 6 must show for each mechanism why it would not decouple.

Mechanisms Available for Governing Transmission

This section examines five mechanisms and assesses them against the specification assembled in Section 5. Its objective is to establish what an apparatus meeting that specification would have to include, and to report which lines of the specification no available mechanism reaches. Each of the first five subsections follows one order: the mechanism as it would operate, the prior owners conceded at the point the claim arises, whether the requirement attaches to the occasion itself or operates at a distance from it, and why the requirement would resist the decoupling that Section 5 identified as the characteristic fate of process requirements. The sixth subsection reports the assessment. The method combines the two routes of Section 4 and marks which carries each result.

Two obligations imposed earlier govern every subsection. The failure recorded in Section 3 is attributed to a reform that operated at a distance from the decision it meant to affect, so each mechanism below states where it attaches. And Section 5 established that the observability which makes a process requirement enforceable is the same feature that makes it separable from any effect upon the person described, so each mechanism below states what would resist that separation.

Duties Attaching at the Occasion of Issuance

Standardized Forms and Elaboration Requirements in Existing Practice

Requirements attaching to the composing of an account are the most developed of the five, and the development is not this paper’s.

Standardised reference forms achieve materially higher predictive validity than narrative letters, a comparison of three such forms in a military applicant sample reporting validity substantially above that ordinarily obtained from narrative material (McCarthy and Goffin 2001). Narrative letters in their current form correlate weakly with later performance and add little beyond other material already available, on a meta-analysis of admissions data (Kuncel, Kochevar, and Ones 2014). Requiring an evaluator to elaborate and justify a rating reduces the operation of identity-based bias, on an experimental study of evaluation (Morgan, Elder, and King 2013). And a standardised letter of evaluation has operated in one medical specialty since the 1990s, been revised twice, and been studied, with reported gains in comparability across writers and persistent difficulties of inflation and of bias in the narrative portions (Hegarty et al. 2014).

All of this is conceded. A requirement that an account be composed on a specified form, with specified fields and a requirement to state reasons, is an existing proposal with an existing evidence base and an implemented instance, and nothing in this paper strengthens the case for it.

The implemented case is nonetheless instructive for what it shows about decoupling, and Section 7 returns to it. Standardisation constrains the form of an account without constraining its content, and the studied regime accordingly reports improvement in comparability alongside the persistence of the difficulties that motivated it. A form is completed; a disposition is not altered.

Duties of Care and Fair Overall Impression in Existing Doctrine

Duties running to the quality of what is written are likewise established. An employer owes the subject a duty to take reasonable care in preparing a reference (House of Lords 1994), and a reference must not convey a misleading impression overall even where its discrete statements are accurate (Court of Appeal of England and Wales 1998).

These duties are conceded and their direction is the point. Each binds the issuer to a standard concerning what is written, and neither gives the person described any part in the writing. The duty in negligence is discharged by taking care, and the requirement of fair overall impression is a requirement upon the composer. A regime consisting of these duties alone leaves the allocation reported by the preceding paper undisturbed: the person described may obtain a remedy after the fact for a badly made account, and has no standing on the occasion of its making.

Solicitation of the Other Participant’s Rendering before Issuance

What remains once the above is conceded is narrower than a general duty of fairness and is the mechanism this paper advances. Before issuing an account of a shared experience, the issuer would be required to invite the person described to supply their own rendering of that experience, to record what was supplied, and to state that it was considered.

Three features fix what the requirement is not. It confers no agreement requirement: the issuer composes the account and reaches whatever conclusion they reach. It confers no veto: the person described cannot prevent issuance. It confers no entitlement over the account, which follows from the amended principle of co-reproduction stated in Section 5, on which standing to participate in settling an interpretation carries no power to license, block, or take proceeds. What the requirement establishes is participation in the occasion, and nothing beyond it.

The requirement attaches to the occasion itself rather than operating at a distance from it, which distinguishes it from the reform whose failure Section 3 recorded. That reform altered the consequences attaching to a choice and left the choice where it was; this alters what must be done before an account may issue.

Its resistance to decoupling rests on a structural feature that the paper states as a general point, since it recurs in the subsections that follow. Procedures decouple when the institution that hosts them also performs them, so that discharge is a matter of the institution’s own conduct and may be conducted as ceremony. Here discharge requires an act by the party the requirement protects: the rendering is composed by the person described, and the record of it exists because they made it. An issuer may read that rendering and disregard it, and often will; what the issuer cannot do is produce the artefact alone. The requirement therefore fails in a visible way rather than an invisible one, since an absent rendering is absent on the face of the record.

The limit is stated with the mechanism. The requirement does not change what an account says, and on the argument of Section 5 no process requirement could. It changes what exists alongside the account, which is what the criteria of counter-power and historical revisability ask for, and Section 6 assesses below whether that is sufficient.

Forums Operating before Reliance

The Objection from Ceremonial Adoption and Judicial Deference

The mechanism examined in this subsection faces an objection strong enough that it is stated before the mechanism is proposed.

Organisations adopt formal structures for the legitimacy those structures confer, and decouple them from what is done (Meyer and Rowan 1977). Grievance procedures are the paradigm case, constructed as symbolic responses to ambiguous obligation (Edelman 1992) and institutionalised as the recognised form of compliance irrespective of what they deliver (Edelman, Uggen, and Erlanger 1999). And the adoption is rewarded: analysis of a random sample of 1,024 federal employment-discrimination opinions decided between 1965 and 1999 finds courts inferring from the presence of a grievance or evaluation procedure that no wrong occurred, with the inference strengthening over the period (Edelman et al. 2011).

Transferred to the present object the objection runs as follows. A forum established so that a person may contest an account of themselves would be hosted by the institution holding the account. Its existence would be a fact about the institution, available to be cited. Its outputs would be internal and mostly unobserved. A later decision-maker asked whether the person was treated fairly would find a procedure in place and could reason from its presence. The person described would then face an account they still cannot alter, together with an institutional record indicating that they had their say. That is a worse position than the one they occupy without the forum.

The transfer is not automatic and the conditions of its success should be stated, as Section 4 requires. The findings above concern grievance procedures within employing organisations, addressed to ambiguous statutory obligations, and reviewed by courts asking whether a statutory duty was met. A forum of the kind proposed here would be addressed to a determinate private account rather than to a statutory standard. What carries across is the mechanism rather than the setting: wherever a procedure’s existence rather than its output is what a later observer can see, existence is what the procedure will be adopted to supply.

Conditions Distinguishing a Forum from a Symbolic Structure

The objection identifies what a forum would have to do to escape it, and the conditions follow from the mechanism the objection names rather than from general aspiration. Four are stated, and the paper claims no more for them than that they are necessary.

The first is that the forum’s output rather than its existence must be what travels. Where a procedure produces a record that accompanies the account wherever the account goes, a later observer sees what was decided and on what reasons, and cannot reason from presence alone. This converts the forum from a fact about an institution into a document about a case.

The second is that the person described must initiate it. A procedure the institution convenes on its own motion is conducted at the institution’s discretion and on its schedule; one that exists only when invoked cannot be adopted as a standing display, because it produces nothing until someone uses it.

The third is that its output must be answerable in the sense Section 5 took from the account of adjudication: reasons offered against an interpretation must be met by reasons rather than by restatement. A body permitted to record disagreement without engaging it satisfies the form of participation and not its meaning.

The fourth is that no deference may attach to its existence. This condition differs from the other three because the paper cannot supply it. Whether a later decision-maker treats the presence of a procedure as evidence of fair treatment is a matter for that decision-maker, and the evidence recorded above indicates what happens when the question is left open. Section 7 reports the consequence and Section 10 states it as a limit on the mechanism.

Operation before Reliance rather than after Determination

The remaining condition on a forum is the one the specification insists upon and the one hardest to institutionalise. A forum satisfying the criterion of counter-power operates before the account is relied upon, since by the time a decision has been taken the alteration the instrument effected has already occurred and what a forum can restore is a record rather than a position.

Two arrangements would satisfy the condition and they differ in cost. Under the first, an account does not become available to the receiving institution until a stated interval has passed, during which the person described may invoke the forum. Under the second, the account becomes available immediately and travels marked as contested until the forum has reported, with the receiving institution required to await the report before relying. The first delays every account in order to protect the few that are contested; the second delays only contested accounts and requires the receiving institution to observe a constraint on its own decision-making.

Both arrangements impose the burden the preceding paper identified as the demanding one, since each converts issuance from an act one party performs alone into an exchange requiring two, and the second additionally binds a party that took no part in the relation and receives no benefit from the constraint. Section 7 reports what this would cost and upon whom it would fall, and Section 8 records that the requirement rests on the framework route alone, no surveyed regime having imposed anything comparable.

Preservation Requirements on Holders of Records

A requirement of this kind binds whoever holds an account, for as long as it is held, to retain the materials that a later reopening would need: the account itself, the identity of the party who composed it, the period of the relation it draws upon, the rendering solicited under the mechanism of Section 6, and any divergence between the participants known at the time.

The requirement is partly owned and the convergence is worth stating precisely. Where an account is held as personal data, the accuracy principle already demands that a record represent what was expressed, by whom, and when (European Parliament and Council 2016). Two of the four materials are therefore preserved already, for reasons a legal system arrived at independently of anything in this series. What is unowned is the extension to the period covered and to known divergence, and the requirement that preservation persist for as long as reliance is possible rather than for as long as processing continues.

The requirement operates at a distance from the occasion of issuance, and Section 3 recorded that distance as the diagnosis of a failed reform. The cases differ in a way that matters. The reform that failed operated at a distance from a choice, altering the consequences attaching to a decision that the issuer remained free to take either way. This operates continuously upon a state of affairs rather than upon a choice: a party either holds the materials or does not, and whether they do is establishable at any moment without reconstructing anyone’s reasoning.

Its exposure to decoupling is greater than that of the solicitation duty, and the reason follows from the principle stated there. Discharge here is the holding institution’s own conduct, so nothing in the structure of the requirement obliges any act by the party it protects, and a regime of this kind may be maintained nominally while the materials are allowed to lapse. What limits the exposure is the moment of discovery: the materials are called for when a person invokes the forum, so an absence is found at the point where it does the most damage and by the party with the strongest reason to report it. This is weaker than the resistance available to the preceding mechanism and the paper does not claim otherwise.

Expiry and Reopening of Recorded Interpretations

Under this requirement an account may be relied upon for a stated period, after which reliance requires renewal by the issuer or a fresh account from someone else. The interpretation is neither corrected nor withdrawn; what lapses is the entitlement to act upon it.

The requirement follows from two lines of the specification rather than from independent motivation. The criterion of historical revisability, as amended, calls for supersession rather than correction, and expiry is the institutional form supersession takes where nothing is available by which a later interpretation could be shown superior to an earlier one. The criterion of generativity bears on the period for which an account may be relied upon, since an interpretation of a relation that ended long ago forecloses what a person may become without any continuing warrant for doing so.

No prior owner was located, and the located practice runs the other way: services that hold and forward such accounts store them without expiry, so the default arrangement is indefinite reliance.

The requirement binds the receiving institution rather than the issuer, which distinguishes it from the other four and places it at the point of reliance rather than of composition. And its resistance to decoupling is the strongest of the five, for a reason that is structural rather than institutional: expiry requires no act by anyone. A date passes and an entitlement lapses. A requirement that no party performs cannot be performed ceremonially, and the only way to defeat it is to disregard it openly.

Two limits are stated with the mechanism. Expiry is blunt, and discards accounts that remain informative along with those that do not, since nothing distinguishes them without the standard Section 5 showed to be unavailable. And it does nothing within the period, so a person burdened by an account newly issued has no help from it.

Restriction and Prohibition of Exchange

The last mechanism would restrict the exchange of such instruments for consideration, and in particular the arrangement in which one party supplies the account and another supplies the signature that gives it force.

No prior owner was located. Rules governing paid endorsement in commercial advertising address representations made to consumers about products and reach neither the instruments examined here nor the parties to them.

Two difficulties attend the mechanism and together they are severe enough that the paper reports it as the weakest of the five.

The first was inherited. Section 5 recorded a condition on any legal treatment of the operative force of these instruments, namely that the remedy must not be terminal, since a remedy consisting only in prohibition removes the instrument rather than governing it. A prohibition on exchange is terminal in exactly this sense with respect to the transaction it addresses.

The second is more serious and arises from the paper’s own definition. A mechanism was defined in Section 2 as a requirement whose discharge can be established without assessing the correctness of any interpretation. A prohibition is discharged by non-occurrence, and the non-occurrence of a private exchange between two willing parties is not establishable at all. The requirement therefore satisfies the definition in form and fails it in substance: there is nothing to observe, and a regime containing it would consist of a rule that could be complied with or breached without either being visible.

The mechanism is retained in the concluding assessment of this section because the practice it addresses is real and because its exclusion should be argued rather than assumed. Section 8 records the result as one reached by the framework route and defeated on the institutional one.

Assessment of the Mechanisms against the Specification

Table [tab:assess] sets the five mechanisms against the ten lines of the specification assembled in Section 5. Three findings follow, and the third is the section’s principal result.

The first is that four of the five mechanisms reach lines of the specification and that no single mechanism reaches more than three. An apparatus meeting the specification would therefore be composite, which is what the literature on instrument mixes would predict (Gunningham, Grabosky, and Sinclair 1998) and which raises the cost question Section 7 takes up.

The second is that the mechanisms divide by the party they bind, and the division is uneven. Three bind the issuer, one binds whoever holds the account, and one binds the party who would rely upon it. No mechanism binds the receiving institution at the point where it requests an account, which is the occasion at which the practice is set in motion.

The third is that two lines of the specification are reached by nothing available, and one further line is reached only conditionally.

The criterion of non-exploitation is reached by no mechanism examined. It forbids treating either party as a means to the receiving institution’s convenience, and every mechanism in this section regulates how an account is made, held, or relied upon rather than why it was requested. A requirement addressed to the purpose of a request would have to distinguish legitimate from convenient reasons for seeking an account of a person, which returns to the assessment of reasons that Section 5 placed outside reach.

The condition that no deference attach to a forum’s existence is likewise reached by nothing, for the reason given when the condition was stated: it governs the conduct of later decision-makers who are not parties to the apparatus and cannot be bound by it.

And the forum condition of reach, requiring access to evaluative content rather than to the record alone, is met only conditionally. A forum meeting the conditions of Section 6 affords participation in which reasons are answerable by reasons, which is what the account of adjudication inherited in Section 5 requires. What such a forum should conclude remains undetermined, and until it is, the mechanism supplies a procedure whose outcome is unspecified. Section 10 states this as the paper’s principal unresolved difficulty.

P3.3cmP2.6cmY Mechanism & Party bound & Lines of the specification reached
Solicitation of the other rendering & Issuer & Co-reproduction standing; counter-power, in part; subject preservation, in part
Forum before reliance & Issuer and receiving institution & Counter-power; forum timing; forum standing; forum reach, conditionally
Preservation of materials & Holder of the account & Revisability, in the preservation limb; forum materials
Expiry of reliance & Receiving institution & Revisability, in the supersession limb; generativity
Restriction of exchange & Parties to the exchange & None established; the requirement is unobservable in discharge
Reached by no mechanism & — & Non-exploitation; absence of deference to a forum’s existence

Costs, Burdens, and Failure Modes

This section reports what the mechanisms of Section 6 would cost, upon whom the cost would fall, and how each would fail. Its objective is to supply the assessment the preceding paper in this series deferred to this one, and to state the evidence for each failure mode at the strength that evidence will bear. It proceeds by examining the burden created by discharge at the occasion of issuance, the incidence of compliance cost, the two failure modes already visible in adopted procedure, the gaming and capture of a contestation procedure, and the contraction of issuance under raised exposure. Its method is to treat each finding as a claim about the sample and setting that produced it, and to state in every case what would have to hold for the finding to transfer to the present object. Table [tab:failure] collects the mechanisms against the failure modes to which each is exposed.

The Burden Imposed by Discharge at Issuance

Two of the mechanisms discharge at the occasion of issuance, and both convert an act that one party performs alone into an exchange requiring two. The preceding paper identified this as the demanding requirement and left its weight to be assessed here.

The weight has a shape that matters more than its size. A duty to solicit and consider another rendering falls upon every occasion of issuance, whether or not the account would have been contested. A forum operating before reliance is invoked only where someone invokes it, but the arrangement that makes invocation possible constrains every account, either by delaying all of them or by requiring receiving institutions to observe a marker on any of them. The cost is therefore distributed across the whole practice while the benefit is concentrated in the small proportion of cases where an account is disputed.

This shape is what makes the burden difficult rather than merely large. A requirement whose cost and benefit fall on the same occasions can be justified occasion by occasion. A requirement whose cost falls on all occasions and whose benefit falls on few must be justified in the aggregate, and the aggregate case depends on how often accounts are contested, which Section 4 placed outside this paper’s scope.

Distribution of Compliance Cost among the Parties

Procedural requirements impose costs of learning that a procedure exists, of complying with what it demands, and of bearing the experience of doing so; these costs are choices rather than accidents; and they fall disproportionately upon those with fewest resources for navigating them (Herd and Moynihan 2018).

The transfer condition should be stated, since the framework was developed for citizens seeking public benefits rather than for parties to a private account. What has to hold is that the person protected by a requirement is the party who must learn of it, invoke it, and sustain the invocation. That holds here, and it holds for the mechanisms this paper regards as strongest.

One owner must be conceded before the consequence is drawn, and it holds the claim rather than merely resembling it. A party who comes before a procedure repeatedly enjoys advantages over one who comes once: advance intelligence and the ability to plan for the encounter, standing access to those who specialise in it, start-up costs already borne, and informal relations with those who administer it (Galanter 1974). The same analysis observes that a procedure which acts only when invoked favours the party equipped to invoke it, being the party with the information, the capacity to bear the cost, and the skill to navigate what the procedure requires. Both points bear directly here. The issuer and the receiving institution encounter accounts of persons repeatedly; the person described encounters one. And the forum proposed in Section 6 was made invocable by the person described precisely so that it could not be adopted as a standing display, which places it in the class this analysis identifies as favouring the repeat participant. The paper adds nothing to that analysis and inherits its consequence.

The consequence runs directly against a result reported in Section 6. The solicitation duty was argued to resist decoupling because its discharge requires an act by the party it protects, so that failure is visible on the record. The same feature places the work on that party: a rendering must be composed by someone who may be unaware that an account is being written, may lack the vocabulary the occasion rewards, and bears whatever the composition costs them. The forum is worse in this respect, since invocation requires knowing that a forum exists, judging that an account is adverse, and pressing the point against a party with more practice at it. The property that makes these mechanisms robust is the property that makes them costly to the person they protect, and the paper reports this as a finding rather than resolving it.

Ritual Compliance and Defensive Practice

Two failure modes are already visible in the material assembled in Section 3 and need only be applied.

The first is ritual compliance. A procedure adopted for the legitimacy it confers may be decoupled from what is done (Meyer and Rowan 1977), grievance procedures being the paradigm (Edelman 1992; Edelman, Uggen, and Erlanger 1999). Applied here, the solicitation duty becomes a template message sent without expectation of reply, the preservation duty becomes a policy that no one audits, and the forum becomes a body that meets and records. Section 6 argued that solicitation and expiry resist this and that preservation resists it less; the present section adds that the resistance is a matter of degree in every case.

The second is defensive practice. Where a professional’s decisions expose them to liability, practice adjusts toward what is defensible rather than toward what is judged best: in a survey of specialists in a high-exposure environment, a large majority reported practising defensively and a substantial minority reported restricting practice by declining complex cases (Studdert et al. 2005). The transfer condition is that the party faces exposure over discrete decisions and can reduce exposure by altering what they produce. That holds for an issuer, who may write blandly, omit adverse material, or decline to write. Applied to the mechanisms of Section 6, a solicitation duty gives an issuer a further reason to write an account that no rendering could usefully answer, since an account containing little cannot be contradicted.

Capture and Gaming of a Contestation Procedure

Where a procedure’s operation is measured, the measure changes the behaviour measured. An observed regularity tends to collapse once pressure is placed upon it for control purposes (Goodhart 1975), a point later given the formulation that a measure ceases to be a good measure once it becomes a target (Strathern 1997); a quantitative indicator used for decision-making is subject to corruption pressures and distorts the process it monitors (Campbell 1979); and a regime of targets rests on the assumption that a measured part stands for the whole and that gaming is immaterial, both of which fail under examination of a public health system governed by targets (Bevan and Hood 2006).

The transfer condition is that some quantity derived from the procedure becomes the object of attention. That is likely here rather than speculative: a forum generates counts of invocations and of outcomes, and an institution reporting on its own fairness has both the material and the motive. The predictable distortions follow the pattern the literature records. Where invocations are counted, they are discouraged before they are made. Where outcomes are counted, the proportion resolved in the institution’s favour becomes something to manage. Where solicitation is audited, the audit records that renderings were requested rather than what became of them.

Capture is the adjacent risk and the paper states it with the caution the literature now requires. That regulation is often acquired by the industry regulated and operated for its benefit is the classical claim (Stigler 1971); the modern reassessment holds that capture is frequently diagnosed where it has not been shown, and asks for a definition against which the diagnosis could fail (Carpenter and Moss 2014). Applied here, a forum hosted, staffed, and funded by the institution holding the account is structurally exposed, since the party whose conduct is examined controls the examining body. The paper reports the structural exposure and declines to assert that capture would occur, which is the position the reassessment supports.

Withdrawal from Issuance under a Regulated Regime, and the Strength of the Evidence for It

The most consequential failure mode is the one that would defeat the paper’s own proposals, and this subsection states it at the strength the evidence bears and no further.

The finding is that raising what is required of an issuer may cause issuers to issue less. Three pieces of evidence bear on it and they establish different things.

The direction of the effect is documented. Practice in one jurisdiction converged on confirming dates and positions under two-sided liability exposure, and continued to do so after more than forty legislatures acted to relieve it (Cooper 2001). What this record supports is that the practice responds to exposure, and that the observed response is contraction rather than adjustment. It does not establish that exposure is the sole cause of the convergence, which the record does not isolate.

The resistance of the practice to legislative correction is documented. The statutes left the surrounding common law in place and had no measured effect, and an experiment assigning 420 supervisors across conditions making consent, immunity, and confidentiality salient returned largely null results (Walsh, Heighton, and Dingens 2023). What this supports is that an intervention addressed to the issuer’s incentives, without alteration of the surrounding exposure, has not been shown to reverse the contraction. The experiment measured willingness to disclose rather than rates of issuance, so it bears upon the premise of the statutes rather than directly upon their effect.

The magnitude of the effect is not documented. No study isolates the effect of exposure upon rates of issuance, and the nearest quantified analogue comes from a different profession, where two in five surveyed specialists reported restricting practice under liability pressure (Studdert et al. 2005). That figure is evidence about physicians and is not evidence about referees.

Section 4 recorded the consequence for how this claim may be used, and it is repeated because the temptation to exceed it is strong. The finding is a reason for caution in design. It is not a quantity that may be set against the benefit of a mechanism, and an argument of the form that a requirement should be rejected because contraction would outweigh its benefit asserts a comparison no available evidence supports.

Two further points complete the analysis and both are unfavourable.

The contraction, if it occurs, falls unequally. A person with many relations from which an account might be drawn is inconvenienced when one issuer declines; a person with few is not. The parties least able to absorb a contraction of the practice are those the mechanisms were formulated to assist, and this is the same incidence result reported earlier in this section, arrived at by a different route.

And the framework condemns the outcome. The criterion of generativity requires that an apparatus leave open the field of what a person may still become. An apparatus that reduces the number of accounts written forecloses that field for those who cannot obtain one, so a remedy with this effect is condemned by the same principle that motivates the remedy. Section 9 takes up what follows for the framework.

P3.3cmP3.4cmY Failure mode & Mechanisms exposed & Standing of the evidence
Ritual compliance & Solicitation, preservation, forum & Documented in adjacent settings; transfer argued
Defensive composition & Solicitation, forum & Documented for a different profession; transfer argued
Gaming of a measured procedure & Forum, solicitation & Documented across measured regimes; transfer likely
Capture of the hosting body & Forum & Structural exposure stated; occurrence not asserted
Contraction of issuance & All mechanisms raising what is required & Direction and legislative resistance documented; magnitude unmeasured
Unequal incidence of cost & Solicitation, forum & Documented for procedural requirements generally; transfer argued

Convergence and Divergence of the Institutional and Framework Routes

This section collects the route markings made in Sections 6 and 7 and reports the standing of each result, so that a reader who accepts one route and rejects the other can determine what survives. It lists the results carried by both routes, those carried by one, and the points of conflict, and distinguishes a conflict between the routes from a conflict the routes jointly reveal. Nothing new is argued.

Results Reached by Both Routes

Three results are reached from the framework and matched on the institutional record.

That only process requirements are available in this domain follows from the inherited ontology, since an output standard would require a comparison the ontology denies; and it is the condition under which one regulatory literature independently recommends prescribing process instead of output (Coglianese and Lazer 2003). The convergence is close: the framework supplies the reason the standard is unavailable, and the regulatory account supplies the form that follows.

That a forum must afford participation in which reasons are answerable by reasons follows from the criterion of counter-power and from an account of adjudication reached on other grounds (Fuller and Winston 1978).

That the materials of reopening should be preserved follows from the criterion of historical revisability and is partly required already, since the accuracy principle attaching to records of opinion demands that a record represent what was expressed, by whom, and when (European Parliament and Council 2016). Two of the four materials are preserved for reasons a legal system arrived at independently.

Results Reached by a Single Route

Five results rest on the framework alone. The derivation of the five mechanisms from the specification; the requirement that a forum operate before reliance, no surveyed regime having imposed anything comparable; the solicitation duty; the expiry of reliance; and the claim that the five mechanisms exhaust the forms available under the definition adopted. Nothing in the surveyed record supports any of these, and the last holds only relative to that definition.

Six results rest on the institutional record alone. The contraction of issuance under raised exposure, in direction and in resistance to correction. The failure of a permissive reform addressed to this practice. The decoupling of adopted procedure from what is done. The gaming of a measured procedure. The unequal incidence of procedural cost. And the defeat of the fifth mechanism, which the framework proposes and which the definition adopted in Section 2 disqualifies, since the non-occurrence of a private exchange is not establishable.

The asymmetry between the lists repeats the pattern reported in the preceding paper. The framework contributes requirements; the record contributes what happens to requirements. Neither half would carry the argument alone, and the paper’s contribution lies in setting one against the other.

Points at Which the Routes Contradict Each Other

One conflict is between the routes and one is of a different kind, and the distinction matters for what Section 9 must do.

The conflict between the routes concerns the restriction of exchange. The framework’s criteria support restricting a transaction in which the force of an instrument is supplied separately from its content. The institutional route defeats the mechanism, since a prohibition is discharged by non-occurrence and the non-occurrence of a private exchange is unobservable. Here one route proposes and the other disqualifies, and the disqualification stands because the definition it applies was adopted before the mechanism was reached.

The second conflict is not between the routes at all, and describing it as one would misplace it. The framework requires an apparatus that generates the capacity to contest at the occasion of issuance and preserves the field of what a person may become. The institutional record indicates that raising what is required of an issuer contracts the practice, and that a contraction falls hardest on those with fewest relations from which an account might be drawn. Foreclosing the field of those persons is what the framework’s own criterion of generativity forbids. The two routes therefore agree about what would happen; what they jointly reveal is that two of the framework’s criteria cannot both be satisfied in this domain by any apparatus this paper has identified. The inconsistency is internal to the framework and is exposed rather than created by the institutional evidence.

Implications for the Generative Relational Framework

This section states what the framework must give up in consequence of the findings of Sections 7 and 8. Its objective is to settle the inconsistency identified in Section 8 by amending the framework rather than by defending it, and to record what the paper hands forward. It proceeds by stating the revision required to the principle of counter-power, the revision required to the principle of generativity, and the results transmitted to later work. The method is application of the paper’s own findings to its own premises.

Revision Required to the Principle of Counter-Power

The principle requires that the capacity to contest a reproduction be generated by the act conferring interpretive authority rather than supplied afterwards. The extension is sound and the finding of Section 7 shows it to be stated at the wrong grain.

As stated the requirement is a threshold: an instrument either carries the capacity or does not, and one that does not is defective when issued. Applied uniformly it attaches the full apparatus to every occasion of issuance, which produces the shape of burden reported in Section 7, where cost falls on all occasions and benefit on few. A requirement of that shape is defensible only in the aggregate, and the aggregate case rests on a frequency this paper does not establish.

The revision states the requirement comparatively rather than absolutely. What an occasion of issuance must generate is a capacity to contest proportionate to the foreclosure the instrument effects, so that an account entering a decision of consequence carries the full apparatus and an account of slight effect carries less. The general form is familiar from the graduated sanction of the self-governance literature (Ostrom 1990), and the revision consists in importing that grain into a principle previously stated without it.

One objection must be met, since the revision appears to reintroduce an assessment the paper excluded. Grading the requirement requires an estimate of what an instrument forecloses. That is an assessment of the instrument’s effect in the receiving system, and not an assessment of whether the interpretation it contains is correct. Section 5 excluded the second and said nothing about the first, so the revision survives the constraint that governs the rest of the paper.

Revision Required to the Principle of Generativity

The principle requires that a reproduction leave open the field of what those it describes may still become. Section 8 reported that an apparatus satisfying the framework’s other criteria may contract the practice and thereby foreclose that field for those who cannot obtain an account at all. Stated as an unconditional prohibition, the principle condemns the apparatus and, applied consistently, condemns any intervention whatever, since every governance act forecloses something.

The revision states the principle comparatively. An apparatus is condemned where it forecloses more than the arrangement it would replace, rather than where it forecloses at all. The present arrangement is not neutral ground: it forecloses the field for anyone characterised adversely without recourse, which is the condition this series was formulated to address. The question is therefore which of two foreclosing arrangements forecloses less.

The revision does not rescue the mechanisms and the paper says so plainly. A comparative criterion requires a comparison, and the comparison requires the magnitude that Section 7 reports as unmeasured. The amended principle is therefore applicable in form and inapplicable in fact to the question this paper raises. What the revision achieves is to identify what would have to be measured before the question could be answered, which is a smaller result than a decision and a more honest one.

Results Transmitted to the Later Papers

Four results are handed forward.

To the injustice paper: a procedural remedy loads its cost on the party it protects, because the feature that makes such a remedy resistant to decoupling is that its discharge requires an act by that party. Any proposal to address unequal treatment through procedure inherits this, and a remedy shaped as transparency alone leaves the underlying allocation untouched.

To the political-economy volume: the restriction of exchange fails as a governance mechanism because non-occurrence is unobservable, so the practice of exchanging such instruments is reachable, if at all, by something other than prohibition. And the practice responds to raised exposure by contracting rather than by adjusting, which bears on any account of what circulates and why.

To the later papers on justice and on conversion: the comparative form given to the principle of generativity in this section, and the reason for it, namely that an unconditional prohibition on foreclosure condemns every intervention including the ones the framework requires.

To the framework generally: two of its six principles required amendment on contact with the institutional record, as two required amendment on contact with doctrine in the preceding paper. The pattern is consistent and worth stating. The principles are formulated as thresholds, and each has survived contact by being restated as a matter of degree. A framework whose principles are stated absolutely will keep producing requirements that are unsatisfiable rather than demanding, and the remedy is not to weaken the principles but to state at the outset the grain at which each is meant to apply.

Limits of the Account

This section states what the paper leaves unresolved and what would be needed to resolve it. Four questions were forwarded here from earlier sections and are taken first; the remaining limits concern the scope of the survey and the range of the object.

The first forwarded question is what a forum reaching evaluative content should conclude. Section 5 established that adjudication requires a mode of participation in which reasons are answerable by reasons rather than a standard of correctness, which removes the objection that such a dispute cannot be adjudicated at all. Section 6 reported that the mechanism therefore supplies a procedure whose outcome is unspecified.

One clarification is available and it reduces the difficulty without removing it. A forum operating under the constraint of Section 5 cannot conclude that one interpretation of a shared experience is correct, since nothing is available by which correctness could be established. It could conclude something narrower and procedural, namely whether the reasons offered against an interpretation were answered by reasons or merely met with restatement, and its output would then be a record of the exchange together with that finding. This is a determination a forum can make, it is of the kind the account of adjudication permits, and it satisfies the requirement that the forum’s output rather than its existence be what travels.

What it does not do should be stated plainly. A finding that reasons went unanswered leaves the account standing and leaves the receiving institution free to act upon it. The forum would improve the record and would not alter the allocation of authority over content. That is a consequence of the constraint governing the whole apparatus rather than a defect peculiar to the forum, and the paper reports it as such: no mechanism examined here can give the person described authority over what an account concludes, because Section 5 established that no process mechanism could. Whether a procedural finding of that kind is worth its cost is undetermined here.

The second forwarded question is whether a forum can be prevented from attracting deference on the strength of its existence. Section 6 stated this as a condition and conceded that the paper cannot supply it, since it governs the conduct of later decision-makers who are not parties to the apparatus. The evidence assembled in Section 3 indicates what happens when the question is left open. A regime containing such a forum would therefore have to be accompanied by something addressed to those decision-makers, and what that would be is outside the scope of this paper.

The third forwarded question is the magnitude of the contraction reported in Section 7. What is documented is the direction of the effect and the resistance of the practice to a legislative correction. What is undocumented is how large the effect is, and the comparison the amended principle of generativity requires cannot be made without it. A study isolating the effect of exposure upon rates of issuance would settle a question this paper can only frame, and until one exists the assessment of any apparatus proposed here remains incomplete in a way that argument cannot repair.

The fourth concerns a gap in the specification rather than in the mechanisms. Section 6 reported that no mechanism binds the receiving institution at the occasion on which it requests an account, which is where the practice is set in motion. The specification inherited from the preceding paper contains no line addressed to that occasion, so the absence is a limit of the specification and not a failure against it. Whether a requirement attaching to the request would be available, and what it would demand, is left open.

Five further limits follow from how the paper was conducted.

The institutional route surveyed a bounded set of regimes, provisions, and findings. Every claim that no regime does something is a claim about that set. Findings drawn from other domains were used as evidence of mechanism, and Section 7 stated in each case what would have to hold for the transfer to succeed; a reader who rejects a transfer condition should treat the corresponding failure mode as unsupported here.

The claim that five mechanisms exhaust the available forms holds only relative to the definition adopted in Section 2. A reader who rejects that definition is not answered by anything in Section 6.

The premises of the framework route are stated in Section 5 rather than defended there, and results resting on that route inherit whatever standing those premises have.

The paper assesses mechanisms without proposing a regime, so questions of architecture, staffing, jurisdiction, and sequencing are untouched, and the composite character reported in Section 6 means that a regime would face integration problems this paper has not examined.

And the paper reaches no recommendation about adoption. It establishes what an apparatus meeting the inherited specification would have to include, what it would demand of the parties, and which of its failure modes are documented. It does not establish that such an apparatus should be built, and Section 9 reports why that question cannot be settled on present evidence.

Conclusion

An earlier paper in this series established that the person described in an account of a shared experience is granted access to it and withheld authority over its content, and stated the conditions any defensible arrangement would have to satisfy. This paper asked whether an apparatus could meet them.

The answer is partly affirmative and its limits are precise. Only process requirements are available, since an output standard would demand a comparison that the ontology of the object denies, and this is the limiting case of a condition under which regulatory theory independently recommends prescribing process. Four mechanisms satisfying that constraint reach eight of the ten lines of the inherited specification: a duty to solicit and consider the other participant’s rendering before issuing, a forum operating before an account is relied upon, preservation of the materials a reopening would require, and expiry of the entitlement to rely. A fifth, the restriction of exchange, is disqualified by the paper’s own definition, since the non-occurrence of a private exchange cannot be established. Two lines are reached by nothing available, and one is reached only conditionally.

Three findings run against the proposals and are reported rather than resolved. The feature that makes these mechanisms resistant to ceremonial adoption, namely that discharge requires an act by the party they protect, is the feature that places the work on that party. The cost of the strongest mechanisms falls on every occasion of issuance while the benefit falls on the few that are contested. And raising what is required of an issuer may contract the practice, in a direction that is documented and by a magnitude that is not, with the contraction falling hardest on those with fewest relations from which an account might be drawn.

The last of these exposes an inconsistency within the framework rather than between the paper’s two routes, since both routes agree about what would happen. Two of the framework’s criteria cannot both be satisfied here by any apparatus identified. Both have accordingly been restated as matters of degree rather than as thresholds, which is the same amendment the preceding paper made to two others on contact with doctrine. The revision to the principle of generativity does not rescue the mechanisms; it identifies the measurement that would have to exist before the question they raise could be answered.

Ayres, Ian, and John Braithwaite. 1992. Responsive Regulation: Transcending the Deregulation Debate. Oxford Socio-Legal Studies. New York: Oxford University Press.

Baldwin, Robert, and Julia Black. 2008. “Really Responsive Regulation.” The Modern Law Review 71 (1): 59–94. https://doi.org/10.1111/j.1468-2230.2008.00681.x.

Bevan, Gwyn, and Christopher Hood. 2006. “What’s Measured Is What Matters: Targets and Gaming in the English Public Health Care System.” Public Administration 84 (3): 517–38. https://doi.org/10.1111/j.1467-9299.2006.00600.x.

Campbell, Donald T. 1979. “Assessing the Impact of Planned Social Change.” Evaluation and Program Planning 2 (1): 67–90. https://doi.org/10.1016/0149-7189(79)90048-X.

Carpenter, Daniel, and David A. Moss, eds. 2014. Preventing Regulatory Capture: Special Interest Influence and How to Limit It. New York: Cambridge University Press.

Coffee, Jr., John C. 2006. Gatekeepers: The Professions and Corporate Governance. Clarendon Lectures in Management Studies. Oxford: Oxford University Press.

Coglianese, Cary, and David Lazer. 2003. “Management-Based Regulation: Prescribing Private Management to Achieve Public Goals.” Law and Society Review 37 (4): 691–730. https://doi.org/10.1046/j.0023-9216.2003.03703001.x.

Cooper, Markita D. 2001. “Job Reference Immunity Statutes: Prevalent but Irrelevant.” Cornell Journal of Law and Public Policy 11 (1): 1–68.

Court of Appeal of England and Wales. 1998. “Bartholomew v London Borough of Hackney [1999] IRLR 246.”

Cox, Michael, Gwen Arnold, and Sergio Villamayor-Tomás. 2010. “A Review of Design Principles for Community-Based Natural Resource Management.” Ecology and Society 15 (4): 38. https://doi.org/10.5751/ES-03704-150438.

Edelman, Lauren B. 1992. “Legal Ambiguity and Symbolic Structures: Organizational Mediation of Civil Rights Law.” American Journal of Sociology 97 (6): 1531–76. https://doi.org/10.1086/229939.

Edelman, Lauren B., Linda H. Krieger, Scott R. Eliason, Catherine R. Albiston, and Virginia Mellema. 2011. “When Organizations Rule: Judicial Deference to Institutionalized Employment Structures.” American Journal of Sociology 117 (3): 888–954. https://doi.org/10.1086/661984.

Edelman, Lauren B., Christopher Uggen, and Howard S. Erlanger. 1999. “The Endogeneity of Legal Regulation: Grievance Procedures as Rational Myth.” American Journal of Sociology 105 (2): 406–54. https://doi.org/10.1086/210316.

European Parliament and Council. 2016. “Regulation (EU) 2016/679 (General Data Protection Regulation).” Official Journal of the European Union.

Fuller, Lon L., and Kenneth I. Winston. 1978. “The Forms and Limits of Adjudication.” Harvard Law Review 92 (2): 353–409.

Galanter, Marc. 1974. “Why the ‘Haves’ Come Out Ahead: Speculations on the Limits of Legal Change.” Law and Society Review 9 (1): 95–160.

Goodhart, Charles A. E. 1975. “Problems of Monetary Management: The u.k. Experience.” In Papers in Monetary Economics. Vol. I. Sydney: Reserve Bank of Australia.

Gunningham, Neil, Peter Grabosky, and Darren Sinclair. 1998. Smart Regulation: Designing Environmental Policy. Oxford: Clarendon Press.

Hegarty, Cullen B., David R. Lane, Jeffrey N. Love, Christopher I. Doty, Douglas S. Ander, Wendy C. Coates, and Michael S. Beeson. 2014. “Council of Emergency Medicine Residency Directors Standardized Letter of Recommendation Writers’ Questionnaire.” Journal of Graduate Medical Education 6 (2): 301–6.

Herd, Pamela, and Donald P. Moynihan. 2018. Administrative Burden: Policymaking by Other Means. New York: Russell Sage Foundation.

House of Lords. 1994. “Spring v Guardian Assurance Plc [1995] 2 AC 296.”

Kleiner, Morris M. 2000. “Occupational Licensing.” Journal of Economic Perspectives 14 (4): 189–202. https://doi.org/10.1257/jep.14.4.189.

Kraakman, Reinier H. 1986. “Gatekeepers: The Anatomy of a Third-Party Enforcement Strategy.” Journal of Law, Economics, and Organization 2 (1): 53–104. https://doi.org/10.1093/oxfordjournals.jleo.a036906.

Kuncel, Nathan R., Reagan J. Kochevar, and Deniz S. Ones. 2014. “A Meta-Analysis of Letters of Recommendation in College and Graduate Admissions: Reasons for Hope.” International Journal of Selection and Assessment 22 (1): 101–7. https://doi.org/10.1111/ijsa.12060.

McCarthy, Julie M., and Richard D. Goffin. 2001. “Improving the Validity of Letters of Recommendation: An Investigation of Three Standardized Reference Forms.” Military Psychology 13 (4): 199–222. https://doi.org/10.1207/S15327876MP1304_2.

Meyer, John W., and Brian Rowan. 1977. “Institutionalized Organizations: Formal Structure as Myth and Ceremony.” American Journal of Sociology 83 (2): 340–63. https://doi.org/10.1086/226550.

Morgan, Whitney B., Kristen B. Elder, and Eden B. King. 2013. “The Emergence and Reduction of Bias in Letters of Recommendation.” Journal of Applied Social Psychology 43 (11): 2297–2306. https://doi.org/10.1111/jasp.12179.

Ostrom, Elinor. 1990. Governing the Commons: The Evolution of Institutions for Collective Action. The Political Economy of Institutions and Decisions. Cambridge: Cambridge University Press.

Parliament of the United Kingdom. 2018. “Data Protection Act 2018, Schedule 2, Paragraph 24: References.”

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

Stigler, George J. 1971. “The Theory of Economic Regulation.” The Bell Journal of Economics and Management Science 2 (1): 3–21. https://doi.org/10.2307/3003160.

Strathern, Marilyn. 1997. “‘Improving Ratings’: Audit in the British University System.” European Review 5 (3): 305–21. https://doi.org/10.1017/S1062798700002660.

Studdert, David M., Michelle M. Mello, William M. Sage, Catherine M. DesRoches, Jennifer Peugh, Kinga Zapert, and Troyen A. Brennan. 2005. “Defensive Medicine Among High-Risk Specialist Physicians in a Volatile Malpractice Environment.” JAMA 293 (21): 2609–17. https://doi.org/10.1001/jama.293.21.2609.

Supreme Court of California. 1997. “Randi w. V. Muroc Joint Unified School District, 929 p.2d 582.”

Walsh, Benjamin M., Brittany Heighton, and Chloe Dingens. 2023. “What Should i Say? Testing Ways to Reduce Fear and Increase Disclosure of Incivility in Reference Checks.” PLOS ONE 18 (8): e0290011. https://doi.org/10.1371/journal.pone.0290011.