1.1. A problem of seeing, not solving

This article does not propose a solution to any doctrinal problem. Rather, it proposes changing the unit of comparison. Comparative trust scholarship has long acknowledged that ‘ownership’ travels poorly between civilian and common law traditions: the ‘dual ownership’ of the common law trust has repeatedly been exposed as metaphor rather than description, and the drafters of the Hague Trusts Convention, faced with the task of defining the trust for civilian and common law systems at once, conspicuously declined to define it in proprietary terms at all.Footnote ¹ What the literature has not yet supplied is the affirmative replacement that this acknowledgment invites: a unit of comparison that is independent of any single system’s doctrinal vocabulary. This article proposes one. The argument is that doctrinal difference is itself a symptom of divergent functional allocation along analytically independent axes. Once those axes are made explicit, certain features of civilian and common law systems become visible that neither tradition’s own doctrinal vocabulary can adequately name. The proposed unit is a three-axis framework — control, administration and benefit — whose variables can be mapped independently of one another without first deciding where title vests. Title is examined separately as a question of doctrinal implementation (see Section 3.5). The methodological claim stands independently of any doctrinal application and is addressed to comparative private-law scholarship more generally. Where the framework exposes a misalignment between the structural allocation a system has made and the doctrinal apparatus it has deployed to administer that allocation, recurrent doctrinal difficulties are reconstituted as the predictable consequences of a structural choice rather than as isolated anomalies.

The methodological difficulty that prompts this reorientation lies not in the accuracy of the doctrinal analysis that comparative trust scholarship has undertaken but in a vocabulary on which that analysis has continued to rely even where its limitations have been acknowledged.Footnote ² The vocabulary is that of ‘ownership’ — whether the civilian dominium, the common law legal estate or equity’s beneficial interest. Treated as a comparator, the category is not stable: its content is not sufficiently determinate to anchor cross-system comparison. ‘Ownership’ in each tradition denotes a bundle of functional capacities that the respective legal system has chosen to aggregate under a single doctrinal label. The aggregation differs across systems, and the differences are not merely nominal. They are constitutive of what each system recognises and, correspondingly, of what each system is liable to render invisible. Where comparison nonetheless proceeds through the category — and, in the absence of an articulated alternative unit, there is no obvious other route — the result is, in the argument advanced here, not peripheral imprecision but structural misdescription.

The trust has furnished comparative private law with its most revealing test case for this difficulty. The reception of the trust into mixed legal systems,Footnote ³ the relation between the trust and the civilian concepts of patrimoine and fiducie,Footnote ⁴ the emergence of ‘trusts without equity’Footnote ⁵ and the structural features of the trust in Quebec, Scotland, South Africa, Louisiana and the NetherlandsFootnote ⁶ have each been treated with doctrinal depth, and each had to negotiate, in different ways, the instability of ‘ownership’ as a comparator. Within common law jurisdictions, the difficulties posed by trading and commercial trusts have given rise to a sustained body of scholarship.Footnote ⁷

1.2. The three-axis framework in preview

The positive proposal advanced in this article is that comparative analysis of proprietary and fiduciary structures should proceed by reference to three functional axes: control, administration and benefit. These axes are not synonyms for the incidents of ownership parcelled into thirds.Footnote ⁸ They are structurally independent variables whose separate allocation across different actors distinguishes a proprietary arrangement from simple unitary ownership. The diagnostic power of the framework lies precisely in the capacity of the axes to diverge: in the civilian bewind, they diverge cleanly; in the common law trust, they converge in a manner that produces the structural pathologies for which that form is routinely reformed but rarely redesigned.

1.3. The bewind as diagnostic, not as model

The second methodological commitment of this article is that the civilian bewind — and, specifically, the bewind as codified in Section 1 of the South African Trust Property Control Act 57 of 1988 and in Book 4 of the Dutch Burgerlijk Wetboek (BW)Footnote ⁹ — is used as a diagnostic contrast, not as an analogy and not as a normative model. An analogy claims similarity: it proposes that one structure resembles another sufficiently to justify inference from one to the other. A normative model claims desirability: it proposes that the admired structure should be emulated. A diagnostic contrast makes no claim to similarity or desirability. It uses the structural difference between two comparators to make visible a feature of one comparator that its own doctrinal resources cannot adequately bring into view. The claim is not that the bewind solves the trust’s problems or that it should be adopted. The claim is that, by occupying a position in the axis-space that the trust does not occupy, the bewind reveals what the trust has structurally assumed. The revelation is the comparative payoff.

1.4. Argumentative architecture

The argument proceeds in seven further sections. Section 2 develops the instability thesis. It demonstrates that the existing taxonomies of ownership do not supply a cross-system framework because they operate on the incidents of a single ownership relation rather than on the functional variables along which different systems allocate entitlement. Section 3 develops the three-axis framework. It defines each variable precisely, and distinguishes the axes from the incidents of ownership. Section 4 applies the framework to the bewind. Drawing on the Roman-Dutch material in Grotius and Voet, the Dutch material in Book 4 of the Burgerlijk Wetboek and the Belgian material introduced by the 2013 reform,Footnote ¹⁰ it shows that the three jurisdictions converge on a structural allocation in which the axes are separately and identifiably distributed. Section 5 applies the framework to the common law trust. It shows that the trustee holds not only the legal title but also the axes of control and administration simultaneously, and examines how the trustee’s personal liability and indemnity rights operate within that concentration of functions and legal title. Section 6 generalises the diagnosis. The pathologies identified in common law commercial trust law — the priority contest over the equitable lien, the Barnes v Addy architecture, the insolvency uncertainties documented in Carter Holt Harvey Woodproducts Australia Pty Ltd v Commonwealth Footnote ¹¹ and Naaman v Jaken Properties Australia Pty Ltd Footnote ¹² — are reconceptualised as outputs of a structural misalignment rather than as isolated doctrinal difficulties. Section 7 situates the method within the broader comparative-law literature. It argues that structural diagnosis is a distinctive comparative technique that complements, rather than replaces, the functional method of Zweigert and Kötz, the legal formants of Sacco and the transplant methodology of Watson. Section 8 concludes.

1.5. A preliminary observation about what is at stake

It may be objected that the argument of this article, although presented at the level of method, has obvious doctrinal implications and that it is therefore disingenuous to claim a purely epistemic purpose. The objection is anticipated and, to an extent, conceded. Any adequate methodology will have doctrinal consequences, because methodology determines what counts as a doctrinal problem. The claim made here, however, is more limited: that the doctrinal consequences do not follow from the comparison until the comparison has been correctly framed, and that the acknowledged instability of ‘ownership’ as a comparator puts the correctness of the inherited framing in question. The bulk of the difficulty to which comparative trust scholarship has addressed itself — the question whether the trust is ‘really’ a fragmented ownership, a relational obligation or a patrimony of destination — is a question that the framework developed here renders tractable by reconstituting. What is reconstituted is not the trust’s ontological nature. It is the functional allocation of control, administration and benefit across actors, and the doctrinal apparatus that each system deploys to give that allocation effect. The reconstitution is a methodological contribution, not a doctrinal one. That is the sense in which this article proposes to redefine how comparative trust problems are seen, rather than to solve them.

2.1. The assumption and its sources

The assumption that ‘ownership’ is a stable cross-system comparator has two sources. The first is internal to each legal tradition: both civilian and common law scholarship, when operating within a single system, treat ownership as an intuitively recognisable concept whose internal structure requires only occasional analytical excavation.Footnote ¹³ The second is comparatist: the mid-twentieth-century comparative movement, in the search for functional equivalents, frequently identified ‘ownership’ as a functional category capable of anchoring comparison between the civilian dominium and the common law fee simple or the beneficial interest.Footnote ¹⁴ Each source breaks down, however, when the comparison moves from ownership arrangements that are structurally unitary to arrangements — trusts, fiduciary structures, fideicommissary substitutions, civilian bewind and its cognates — in which the entitlements and powers conventionally associated with ownership are deliberately and durably distributed across more than one actor. In those contexts, taking ‘ownership’ as the comparator is not so much wrong as question-begging. It assumes the stability of the category whose stability is precisely at issue.

Two clarifications should be made at once, because the argument of this section is easily overstated. First, the claim is not that comparative trust scholarship has naively believed dominium and the fee simple to be the same thing, or that any identifiable comparatist has expressly adopted ‘ownership’ as a sufficient comparator: the literature has repeatedly registered the difficulty, as Section 2.6 explains.Footnote ¹⁵ The claim is structural rather than biographical: in the absence of an articulated alternative unit, comparison conducted in the vocabulary of ownership inherits the instability of the category, however self-aware the comparatist who deploys it. Second, and correspondingly, the burden of this section is not to convict the existing literature of error but to convert an acknowledged difficulty into an analysed one — to explain precisely why the category is unstable, so that the replacement developed in Section 3 answers to the diagnosis rather than to an intuition.

2.2. Honoré and the incidents of ownership

The most sophisticated Anglophone treatment of the internal structure of ownership is Honoré’s.Footnote ¹⁶ His work is discussed here in detail, not because it addresses comparative trust law directly, but because his taxonomy demonstrates both the strengths and the limitations of analysing ownership by decomposing its internal incidents. Honoré identified 11 incidents — the rights to possess, use, manage, the income and the capital; the right to security; the power of transmissibility; the absence of term; the duty to prevent harm; liability to execution; and the residuary character — and argued that the full complement of these incidents is present in the ‘liberal’ concept of ownership.Footnote ¹⁷ The incidents are not, on Honoré’s analysis, a checklist for determining ownership in any given case; they are a conceptual reconstruction that exposes the functional complexity internal to the category. Subsequent writers, including Penner and Harris, have reworked Honoré’s framework along different lines,Footnote ¹⁸ but the analytical strategy has remained substantially the same: ownership is examined from within by decomposing a unitary relation into its functional incidents.

For the present argument, the limits of Honoré’s framework must be identified with care. What it does not supply is a cross-system unit of comparison. The reason is structural. Honoré’s 11 incidents are inherent in a single owner; they describe the bundle of capacities that the full liberal owner holds. They do not describe how a legal system allocates those capacities when the system has chosen — as every trust-recognising system has — not to vest the full complement in a single actor. Honoré’s framework locates the incidents within an owner; it does not locate the owner within a pattern of distribution. The latter, not the former, is the comparative question. Honoré’s own later engagement with the trust confirms the point from within his framework. In ‘Trusts: The Inessentials’, he argued that much of what the common lawyer treats as constitutive of the trust — including the divided-ownership schema itself — is inessential to the institution, which he characterised instead by reference to the office of trusteeship and its obligations.Footnote ¹⁹ The argument of this article travels in the same direction from the opposite side: where Honoré subtracted inessentials from within the common law concept, the three-axis framework identifies, from outside any single system’s vocabulary, the variables whose allocation the surviving essentials organise.

2.3. Rudden and the wealth/possession distinction

Rudden’s contribution is of a different order, and it helps clarify what is missing from the Honoré framework for comparative purposes.Footnote ²⁰ Rudden distinguished ‘things as things’ — the physical, possessory objects of immediate enjoyment — from ‘things as wealth’ — intangible holdings whose value is realised through administrative intermediation rather than through direct possession.Footnote ²¹ The distinction is important because it identifies, within the category of ownership, a structural bifurcation that Honoré’s incidents do not isolate: the ‘wealth’ dimension of property, unlike the ‘possessory’ dimension, is intrinsically administered. A shareholding, a bank balance, a unit in a collective investment vehicle — each is a thing as wealth in Rudden’s sense, and each presupposes a structural apparatus of administrative management separate from the shareholder’s, depositor’s or unit holder’s enjoyment.

Rudden’s distinction helps, but it also sharpens the difficulty. If ‘things as wealth’ are intrinsically administered — if administration is a structural feature, not a contingent adjunct, of that form of property — then comparison between legal systems on the question of ‘ownership’ of such wealth will inevitably founder. The civilian system, the mixed system and the common law system each organise the administration of wealth through different doctrinal apparatuses; the apparatuses are not functionally interchangeable; and the functional allocation of administrative capacity differs in ways that ‘ownership’ as a comparator cannot capture.

2.4. The civilian ownership literature

Within the civilian tradition, the analytical treatment of ownership has followed a different trajectory. The Roman-law inheritance of dominium as the paradigmatic unitary right has been preserved in most modern civil codes, with the German Bürgerliches Gesetzbuch Section 903 and the French Code civil Article 544 providing canonical statements.Footnote ²² Within this tradition, the incidents of ownership have conventionally been less important than the doctrine of numerus clausus: the closed list of real rights that a civilian system recognises, and outside which proprietary entitlements cannot be created.Footnote ²³ Gretton, writing within the mixed-system perspective of Scotland, has shown that the numerus clausus is not a peripheral feature of civilian property law but its architectural principle: proprietary entitlements in a civilian system have the force they have precisely because they issue from a closed list of recognised forms, and departures from that list are not merely uncommon but conceptually inadmissible.Footnote ²⁴ Van Erp and Akkermans, in their treatment of European property law, develop the same point on a pan-European footing, showing that the numerus clausus continues to operate, even under European harmonisation pressures, as the defining structural feature of civilian property thinking.Footnote ²⁵

The point for the present argument is that the civilian dominium, understood against the background of the numerus clausus, is as system-specific a concept as the common law fee simple. When civilian scholars speak of dominium, they refer to a specific proprietary form that their system has chosen to recognise as the paradigm of the real right; when common law scholars speak of ownership, they refer to a legal estate constituted by a different set of historical and doctrinal commitments. The two are not, in any strong sense, the same thing. Each functions as its system’s conventional benchmark of proprietary entitlement, but that shared functional role does not make them a common comparator. A functional role is not a functional category.

2.5. Gordley’s genealogy of ownership

Gordley has developed an influential genealogical account of the Western concept of ownership, tracing its structure to the late-Scholastic and early-modern natural law writers who adapted Roman-law concepts to novel economic conditions.Footnote ²⁶ On Gordley’s account, the category ‘ownership’ acquired its modern form through a specific doctrinal trajectory that was neither inevitable nor one that other traditions could not have taken. The trajectory produced a bundle of entitlements that, in any given legal system, is held together by doctrinal convention rather than by conceptual necessity.

Gordley’s genealogy, like Honoré’s taxonomy and the civilian numerus clausus literature, is illuminating but, again, not sufficient for the comparative purpose pursued here. The genealogical account explains how ‘ownership’ in the Western legal tradition came to have its present contours; it does not explain how to compare arrangements in which those contours are deliberately redrawn. If the concept was constructed over centuries by specific doctrinal choices, then any system that has made different choices — including every trust-recognising system — is not fruitfully analysed through a category whose content those choices deliberately reject.

2.6. The residual intuition and why it must be resisted

The instability thesis nevertheless encounters a residual intuition that deserves acknowledgment. Scholars who work with trust arrangements often observe that, at some level of abstraction, ‘ownership’ does seem to be divided in a trust. Is this not simply a correct description of a structurally divided ownership?

The residual intuition must be resisted because its apparent descriptive adequacy is illusory. ‘Dual ownership’ is not a description; it is a metaphor that masks the doctrinal complexity it covers. What the trustee holds is not half an ownership; it is a specific set of powers and capacities. What the beneficiary holds is not the other half of ownership; it is a specific entitlement to the economic surplus, coupled with specific enforcement mechanisms. To call both ‘ownership’ is not to have described a divided unity; it is to have submerged two structurally distinct bundles of functional capacities under a single nominal label. The submergence is the source of the comparative difficulty. It is also the source of the doctrinal difficulties within the common law trust, but that is a consequence, not a cause, of the methodological problem.

The point is illustrated by reference to the literature, which has not taken the residual intuition at face value. Smith and others in The Worlds of the Trust,Footnote ²⁷ Gretton in ‘Trusts without Equity’,Footnote ²⁸ de Waal on the reception of the trust into South African lawFootnote ²⁹ and du Toit on the taxonomic difficulties that the South African trust raises for civilian thoughtFootnote ³⁰ have all identified, in different ways, that the metaphor of ‘dual ownership’ fails to travel across systems. The recognition is not confined to scholarship; it has been institutionally embodied. The 1985 Hague Convention on the Law Applicable to Trusts and on their Recognition was drafted, with the decisive participation of leading comparatists, precisely to accommodate both traditions within a single instrument, and its drafters conspicuously declined to define the trust in proprietary terms. Article 2 characterises the trust through a neutral vocabulary of segregation, power and duty: the assets constitute a separate fund and are not part of the trustee’s own estate; title stands in the name of the trustee or of another person on the trustee’s behalf; and the trustee holds powers and duties, in respect of which the trustee is accountable, to manage, employ or dispose of the assets.Footnote ³¹ The Convention’s drafting choice is, in effect, a practitioner’s anticipation of the argument developed in this section: where a definition must hold across systems, ‘ownership’ is the vocabulary that must be given up first. What the Convention achieved by studied avoidance, however, comparative scholarship has yet to convert into an affirmative analytical unit. Section 3 turns to the construction of that replacement.

3.1. The three axes defined

The framework proposed here identifies three functional variables whose allocation across actors is the proper unit of comparative analysis for structured proprietary arrangements. Each variable is defined by reference to the functional capacity it denotes, not by reference to any system’s doctrinal vocabulary.

Control denotes the power to direct dealings with the res. It comprises the faculty of determining how the subject-matter will be deployed, invested, disposed of or retained. Control is prospective and discretionary: it concerns the shaping of future dealings, not the execution of past ones. In its pure form, control is identifiable by asking who, as against the world and as against competing actors within the arrangement, has the faculty of choosing what shall be done with the res. Control may be shared, conditional or subject to consent requirements; the variable does not presuppose exclusive vesting.

Administration denotes the legally recognised capacity to manage, transact and effect legal dealings on behalf of the structure. It is the executive rather than the directive function. Administration concerns the implementation of decisions, the interface with third parties in the legal system, the discharge of duties of care, the maintenance of records and the conduct of the mechanical operations through which the res is handled day-to-day. Administration is backward- and present-looking where control is forward-looking.

Benefit denotes the entitlement to the economic surplus generated by the res or by its deployment. It comprises the right to the residual value after liabilities, the right to the fruits whether income or capital in character and the right, on termination of the arrangement, to the residue of the res itself. Benefit is the entitlement whose accrual is the point of the arrangement from the perspective of the actor who holds it; the axis is reliably identified by asking for whose ultimate enrichment the structure is being operated.

The three variables are related but not reducible to one another. Control can be exercised by an actor who has no benefit; administration can be undertaken by an actor who has neither control nor benefit; benefit can accrue to an actor who has neither control nor administration. In a trading trust, control and administration ordinarily vest in the trustee, whereas benefit accrues to the beneficiaries. A bewindvoerder exercises two. A shareholder exercises benefit with attenuated control. A bare trustee exercises administration alone.

Three questions may nevertheless be anticipated about the choice of three axes. Why not two? Why not four? Why these three rather than another set? The case for three, rather than two, rests on the structural independence of control from administration. Collapsing control into administration, or administration into control, obscures real structural variation. Control and administration routinely come apart. In the Quebec fiducie, the fiduciaire exercises administration while control is, to a greater extent than in the common law trust, constrained by the terms of the fiducie instrument and by the beneficiary’s supervisory position.Footnote ³² In the bewind, the bewindvoerder exercises administration while control is shared with the rechthebbende through the latter’s proprietary position.

The case against four — against adding title, or risk-bearing, or some further variable as a fourth axis — rests on a different analytical ground. Title, as Section 3.5 explains, is a legal designation with significant consequences rather than an independently defined functional capacity. Different systems vest title in different actors and attach different rules to that vesting; these effects are examined after mapping the allocation of control, administration and benefit. Adding title as an axis would smuggle the civilian/common-law discriminator back into the comparator at the very stage of analysis at which the comparator needs to be free of it. The risk-bearing candidate fails on different grounds: risk is an incident of benefit, because whoever receives the surplus bears the economic risk of loss on the underlying deployment. Treating risk-bearing as an independent axis would double-count benefit. Other candidates — duty, enforcement, access — are similarly derivative. The claim is not that three axes are the final word for all conceivable comparative purposes; the claim is that, for the class of proprietary arrangements within which trust and trust-analogous structures fall, three are the minimum independent variables and no fourth is independently necessary.

3.2. Why not the incidents of ownership?

A natural objection to the three-axis framework is that its variables are simply coarse-grained restatements of certain of Honoré’s incidents. Is not ‘administration’ roughly Honoré’s ‘right to manage’? Is not ‘benefit’ roughly his ‘right to income and capital’? Why introduce a new framework at all?

The objection misapprehends the analytical function of the three axes. Honoré’s incidents are features of ownership. They are identified by examining the internal structure of the owner’s position. The three axes are features of allocation. They are identified by examining the pattern of distribution of functional capacities across the set of actors within an arrangement. The two frameworks do different work, and neither is a subset or a restatement of the other.

This is more than a formal distinction. The practical consequence is that the three axes do not require commitment to, or even identification of, an ‘owner’ in any of the arrangements to which they are applied. That is essential to their comparative utility. In the civilian bewind, the question ‘who is the owner?’ has an answer — the rechthebbende — but the answer is laden with doctrinal assumptions that do not readily translate to the common law trust. In the common law trust, the question has multiple competing answers — the trustee at law, the beneficiary in equity, the ‘trust’ (on some modern renderings) as a form of patrimony. Applying Honoré’s framework to these arrangements produces the question: whose incidents are these? The question is unanswerable at the comparative level, because the two systems ascribe the incidents differently as a matter of doctrinal choice. The three axes bypass the question altogether. They do not ask who owns; they ask who controls, who administers and who benefits. Those questions have determinate answers in both systems without requiring commitment to any underlying characterisation of title.

3.3. The axes and the subjective/real/personal distinction

A second anticipated objection is that civilian comparative law already operates with a well-developed taxonomy of rights — the subjective right, the real right and the personal right — that performs much the same work as the three axes. Is the framework not reinventing the civilian taxonomy under different names?

The answer is that the frameworks operate at different levels. The civilian taxonomy classifies the juridical character of entitlements. A real right binds the world and follows the res; a personal right binds a specific debtor and does not follow the res; a subjective right is the generic category to which both belong.Footnote ³³ These distinctions are important, and the three-axis framework does not displace them. What the three-axis framework provides is a level of analysis prior to the juridical characterisation of any particular entitlement. The question ‘who controls?’ can be asked without settling whether the control-holder’s right is real or personal; the question ‘who benefits?’ can be asked without settling whether the benefit-holder’s entitlement is proprietary or in personam. The axes are about functional allocation. The civilian taxonomy is about juridical character. Only the first, however, is available without first committing to the doctrinal vocabulary of one tradition, and only the first is therefore available as an instrument of cross-system comparison at the level of structure.

The distinction has consequences. Under the civilian taxonomy, the South African bewind beneficiary’s entitlement is classified as a real right; the common law beneficiary’s equitable interest is often classified, with varying confidence, as something between a real right and a personal right, but the classification itself has been a persistent point of doctrinal controversy.Footnote ³⁴ The three-axis framework does not require the classification to be settled before comparison can proceed. The question is not whether the beneficiary’s right is real or personal; the question is whether, as a functional matter, the beneficiary is the locus of benefit. Both systems return the same answer to that question, and the comparison can proceed on the answer without the classificatory controversy being first resolved.

3.4. The axes and the patrimony literature

A third anticipated objection concerns the recent comparative law literature on the trust as a form of patrimony. Lionel Smith and others have argued that the trust is best analysed as a ‘patrimony by appropriation’ — a fund separated from the trustee’s personal patrimony and committed to a specific purpose.Footnote ³⁵ The analysis is elegant and has substantial traction, particularly in the Quebec fiducie and Scottish ‘two-patrimonies’ contexts.Footnote ³⁶ Is it not the case that the patrimony analysis already performs the analytical work that the three-axis framework purports to supply?

The patrimony analysis is a complementary rather than a competing framework. Its focus is on the arrangement’s object — the fund and its separateness from other funds. The focus of the three-axis framework is on the functional roles of the actors involved in that fund. The patrimony analysis explains why, in each system, the trust fund is not available to the trustee’s personal creditors; the three-axis framework explains how control, administration and benefit over that fund are allocated to different actors. A complete comparative analysis of a trust-like arrangement will draw on both. The arguments of this article are confined to the second, because that is where the methodological difficulty addressed here is located.

3.5. The axes and the question of title

It might finally be asked why ‘title’ is not itself one of the three axes. Legal title, after all, is often treated as the central variable in trust analysis — the variable around which the common law/equity distinction has historically been organised.

Title is not a fourth functional axis; it is a system-specific legal characterisation that the comparison must examine at a second stage. The systems compared here all have some concept of title, but the concept does different work in each. In the civilian bewind, title vests in the rechthebbende and carries with it the locus of benefit; title and benefit are joined. In the common law trust, legal title vests in the trustee and is separated from beneficial entitlement; title and benefit are disjoined. The structural significance of the disjunction is that it brings with it a series of further consequences — the trustee’s personal liability on trust contracts, the indemnity right, the equitable lien — that arise by operation of law and not by structural necessity. The disjunction is a doctrinal choice. Its consequences are doctrinal consequences. Neither the vesting of title nor the liability rules associated with it should be treated as a fourth functional axis. The axes identify who controls, administers and benefits; the subsequent doctrinal inquiry identifies who holds title and how that designation interacts with those functions, including in relation to personal liability and recourse. Title is therefore legally consequential, but its significance is examined rather than assumed at the comparative starting point.

The remaining sections are driven by this point. Replacing ‘ownership’ with the three axes as the initial comparator makes it possible to distinguish functional allocation from the rules through which each system implements it. The vesting of title, together with the attendant rules on liability and recourse, may materially affect the operation of that allocation and remains an essential part of the subsequent doctrinal inquiry.

4.1. The bewind as functional allocation

The South African bewind trust and its civilian cognates furnish a particularly clear instance of axis-separation and, for that reason, provide a useful diagnostic comparator. Where the common law trust concentrates title, administration and control in a single actor, the bewind separates administration from benefit without disturbing title. The consequences of each arrangement depend both on the allocation of functions and on the rules through which its legal system gives that allocation effect. Identifying the functional distribution is the first step in comparing those consequences.

This section maps the bewind onto the three axes. It does so based on three closely related civilian variants: the bewind as codified in South African statute, the testamentair bewind of Dutch law and the protective bewind as restructured in Belgium by the 2013 reform.Footnote ³⁷ The three variants differ in the detail of their statutory articulation but converge on the same underlying allocation pattern. That convergence is itself significant: it shows that the structural principle is transplantable within the civilian tradition and is not a peculiarity of any one jurisdiction’s drafting choices.

4.2. Roman-Dutch foundations

The bewind is not a creation of the modern civilian codifier. Its structural principle — the separation of administrative authority from beneficial ownership — is traceable to the Roman-Dutch ius commune, where administrative offices were routinely vested in one party without displacing the dominium of another. Grotius, in the Inleidinge tot de Hollandsche Rechts-geleerdheid, treats voogdij (guardianship) and curatele (curatorship) as administrative offices in which the voogd or curator exercises control over the ward’s assets for the ward’s benefit while the ward retains dominium.Footnote ³⁸ Voet, in the Commentarius ad Pandectas, develops the principle with greater doctrinal density, treating tutela and cura as offices whose functions are circumscribed by the beneficial ownership of the ward and whose administrators’ personal estates remain wholly segregated from the administered property.Footnote ³⁹

The structural pattern is readily mapped onto the three-axis framework. The voogd or tutor held the axes of administration and (within the limits of the office) control; the ward retained the axis of benefit together with the locus of title, which in the civilian tradition vests with the beneficiary rather than with the administrator. The segregation of the administrator’s personal estate from the administered property — the protection against the administrator’s insolvency — was a structural consequence of the allocation, not a subsequent doctrinal adjunct. The Roman-Dutch material accordingly supplies not merely a historical antecedent of the modern bewind but an indication that the underlying structural principle is older than, and analytically independent of, any statutory instantiation.

4.3. The Dutch testamentair bewind

The modern Dutch expression of the principle is found in Book 4, Title 4.5, Section 4.5.7 of the Burgerlijk Wetboek, which governs testamentair bewind — testamentary administration.Footnote ⁴⁰ Article 4:153 provides that a testator may establish a bewind over one or more assets left to heirs or legatees, with the bewind taking effect on the testator’s death unless otherwise provided.Footnote ⁴¹ Article 4:166 vests the bewindvoerder (administrator) with exclusive management powers over the administered assets, while the rechthebbende (the proprietor — the heir or legatee) retains beneficial ownership.Footnote ⁴² The allocation is unambiguous on the three axes. Administration vests in the bewindvoerder. Control, in the sense of the faculty of directing dealings with the res, is substantially vested in the bewindvoerder but subject to the rechthebbende’s proprietary position and to the detailed regime of duties, accounts and transparency that Articles 4:160, 4:161, 4:163 and 4:178 impose upon the administrator.Footnote ⁴³ Benefit vests in the rechthebbende. Title — the variable that, in the common law trust, is collapsed into administration — here vests in the rechthebbende: the administrator acquires no proprietary interest in the administered assets and holds no title from which personal liability on dealings could flow.Footnote ⁴⁴

The structural consequence of this allocation is that the administrator’s insolvency is without effect on the rechthebbende’s proprietary position. The administrator is not personally liable for contracts concluded within authority, because the contracts are concluded with authority over property belonging to the beneficiary, not to the administrator. No right of indemnity arises over the administered property, because no personal liability has been incurred from which indemnity would need to be sought. The administrator’s personal creditors have no recourse against the administered assets because those assets are segregated by operation of the structural allocation. All these features are consequences of the axis-separation. None is an independent doctrinal contribution, and none could survive a reallocation that joined administration to title.

4.4. The Belgian unified bewind

Belgium has, since 2014, given effect to the same structural principle through a differently drafted statutory vehicle. The Act of 17 March 2013, which entered into force on 1 September 2014, replaced the four earlier protective regimes (verlengde minderjarigheid, onbekwaamverklaring, gerechtelijk raadsman and voorlopig bewind) with a unified bewind measure governed by Articles 488/1 to 502 of the oud Burgerlijk Wetboek.Footnote ⁴⁵ Article 494 defines the bewindvoerder (administrator) and establishes the administrator’s twofold possible function of bijstand (assistance with legal acts the protected person is judicially declared unable to perform autonomously) and vertegenwoordiging (representation for acts the protected person is judicially declared incapable of performing).Footnote ⁴⁶ Article 492 formalises the subsidiarity principle, and Articles 488/1 and 488/2 codify the principles of necessity and proportionality.Footnote ⁴⁷

The Belgian position is significant for the present argument: it demonstrates that the structural principle is not bound to any particular legislative vehicle even within civilian law. The Belgian legislature abolished that plurality of regimes precisely because the graduated measures introduced doctrinal complexity that the underlying allocation principle did not require. A single unified administrative measure, calibrated by judicial discretion to the circumstances of each case, proved sufficient to give effect to the structural axis-separation without the elaborate doctrinal apparatus retained by the French Code civil.Footnote ⁴⁸

4.5. The South African codification

The South African Trust Property Control Act 57 of 1988 provides the analytically cleanest statutory expression of the structural principle and, correspondingly, its most developed judicial elaboration. Section 1 of the Act defines ‘trust’ in two forms: the Ownership Trust, in which property is transferred to the trustee to be administered for the benefit of beneficiaries; and the Bewind Trust, in which property is transferred to beneficiaries to be administered by a designated administrator.Footnote ⁴⁹ The distinction is constitutive: it determines the locus of title, and all other incidents follow from that determination.

The South African courts have repeatedly confirmed the operative character of the structural allocation. In Braun v Blann and Botha NNO,Footnote ⁵⁰ Joubert JA drew a sharp distinction between the ownership trust and the bewind trust, holding that the nature of the trust determines the proprietary incidents that attach to the trustee’s role. In Badenhorst v Bekker,Footnote ⁵¹ the Natal Provincial Division confirmed that the beneficiary of a bewind trust holds proprietary rights directly enforceable against third parties. In Bafokeng Tribe v Impala Platinum Ltd, Friedman JP stated the principle in terms that go directly to the three-axis allocation: ‘the ownership of the assets of the trust vests in the beneficiary, but the administration of the trust vests in the trustee or bewindhebber’.Footnote ⁵² In Conze v Masterbond Participation Trust Managers (Pty) Ltd,Footnote ⁵³ the Cape Provincial Division applied the three-party structure to a pooled investment vehicle with active commercial management, confirming that the statutory architecture automatically determines the legal incidents: the administrator’s financial difficulties did not affect the assets of the investors, and the investors’ proprietary rights were directly enforceable against the administrator and third parties. In Land and Agricultural Bank of South Africa v Parker,Footnote ⁵⁴ the Supreme Court of Appeal identified the functional separation of enjoyment from control as ‘the essential notion of trust law’; the corresponding insolvency segregation is supplied by Section 12 of the Act, under which trust property does not form part of the trustee’s personal estate except in so far as the trustee is a beneficiary entitled to it.

For the three-axis framework, the South African material shows that axis-separation can be codified with sufficient precision to support commercial application without requiring case-by-case judicial construction of individual trust instruments. The axes are settled by the statutory classification itself; the incidents follow from the axes. The comparative point is not that the South African statute is well-drafted (though it is) but that the clarity of the doctrinal outcome is a consequence of the structural allocation rather than of any doctrinal resource that the South African legal tradition does not share with the common law. Both traditions could, in principle, effect the same allocation. The fact that only one has done so is a product of historical and doctrinal choice, not of analytical necessity.

4.6. The three variants converge

The three civilian variants — Dutch, Belgian and South African — converge on the same three-axis allocation. In each, administration vests in a designated administrator; control is structured around that administration with varying degrees of subsidiarity and proportionality; benefit vests in the rechthebbende or bewindgever or equivalent; and title is aligned with benefit rather than with administration. The particular statutory vocabulary differs. The supervising institutional machinery — testamentary instrument, judicial order or trust deed — differs. The protective calibrations differ. But the structural allocation is the same.

The convergence has methodological significance. Three statutory instruments, produced by three different legislatures at three different historical moments in response to three different practical problems, have converged on the same allocation. The allocation is therefore identifiable cross-systemically as a stable structural form.

4.7. Beyond the bewind: the Quebec and French fiducies as further axis-configurations

The three-axis framework yields further analytical traction when applied to other trust-analogous civilian structures. Two contemporary civilian implementations — the Quebec fiducie codified in Articles 1260 to 1298 of the Civil Code of Québec, and the French fiducie introduced by the 2007 reform of the Code civil — exhibit axis-allocation patterns that differ both from the bewind and from the common law trust. Examining them confirms that the framework identifies a genuine typology of structural forms rather than a binary civilian/common-law contrast.

The Quebec fiducie presents an axis of allocation that neither the bewind nor the common law trust replicates. Articles 1260 to 1262 of the Civil Code of Québec establish the fiducie as a patrimoine d’affectation — a patrimony by appropriation — autonomous and distinct from the patrimonies of the settlor, the fiduciaire (trustee) and the beneficiary.Footnote ⁵⁵ Title vests neither in the beneficiary (as in the bewind) nor in the trustee (as in the common law trust) but in the autonomous patrimony itself. The fiduciaire, under Article 1278, exercises control and administration with all the rights necessary for the carrying out of the fiducie; the settlor and beneficiary retain supervisory rights without holding title over the trust assets.Footnote ⁵⁶ In the three-axis framework, the Quebec allocation is as follows: control in the fiduciaire (subject to the fiducie instrument and the beneficiary’s supervisory position); administration in the fiduciaire; benefit in the beneficiary. Title is vested in the patrimony as an autonomous legal construct — a doctrinal move that neither civilian traditions of personality-based ownership nor common law traditions of dual ownership supplied naturally, and which Quebec law had to construct.

The French fiducie, introduced by Loi n° 2007-211 of 19 February 2007, presents a fourth axis configuration.Footnote ⁵⁷ Under Article 2011 of the Code civil, the fiducie is an operation by which one or more settlors transfer present or future property, rights or security interests to one or more fiduciaires who, keeping them separated from their personal patrimony, act with a determined purpose for the benefit of one or more beneficiaries.Footnote ⁵⁸ Title vests in the fiduciaire in a fiduciary mode, within a patrimoine d’affectation that is kept separated from the fiduciaire’s personal patrimony by operation of law.Footnote ⁵⁹ Article 2025 preserves the fiducie assets from the personal creditors of the fiduciaire and of the settlor, save in limited circumstances.Footnote ⁶⁰ The French allocation on the three axes is: control and administration in the fiduciaire; benefit in the beneficiary; title in the fiduciaire but within a ring-fenced patrimony.

The French fiducie is, in axis-terms, closer to the common law trust than the Quebec fiducie is, because title vests with the administrator. But it differs from the common law trust in a structurally significant respect: the segregation of the fiducie assets is not achieved through the equitable apparatus of the indemnity-backed lien but through the civilian concept of the patrimoine d’affectation — a dedicated patrimony whose separation from the fiduciaire’s personal patrimony is a structural feature, not a consequence of equity’s response to personal liability. The French reform thus achieves something that the common law trust achieves only through a parallel doctrinal apparatus: it concentrates title, administration and control in the administrator while structurally preserving the beneficiary’s position against the administrator’s insolvency.

The comparative synthesis across the four trust-analogous structures examined in this article — the bewind, the Quebec fiducie, the French fiducie and the common law trust — yields a four-pattern typology that no single tradition’s doctrinal vocabulary captures. The bewind separates title from administration by structural allocation, with title aligned with benefit. The Quebec fiducie disaggregates title from any natural person through the autonomous patrimony. The French fiducie concentrates title with the administrator but ring-fences the administered property through the patrimoine d’affectation. The common law trust concentrates title, administration and control in the trustee while distributing benefits through equitable enforcement, using an indemnity-backed lien to manage the structural exposure that the concentration generates. Each is a distinct structural form; each produces a distinct set of doctrinal consequences; and each is analysable through the three-axis framework without prior commitment to any tradition’s ownership vocabulary. The framework is accordingly not a binary discriminator between civilian and common-law traditions. It is a typology-generating instrument. That is the sense in which structural contrast, as a comparative technique, is distinct — in the qualified sense developed in Section 7.3 — from the classical functional, formant and transplant approaches.

5.1. The common law allocation

The three-axis framework, applied to the common law trust, exposes a concentration that civilian structures do not share. The trustee of a common law trust holds legal title; exercises the axis of administration; holds the axis of control (subject to the trust instrument and to fiduciary constraints); and is personally liable on contracts concluded while administering the trust. The beneficiary holds the axis of benefit, enforceable in equity against the trustee and (subject to the bona fide purchaser rule) against third parties. Title, administration and control vest together. Benefit stands apart.Footnote ⁶¹

The disjunction that the civilian bewind places at the level of structure (administration versus benefit) is placed, in the common law trust, at the level of doctrine (legal title versus equitable interest). Both systems achieve something like the separation of administrative function from beneficial entitlement. The difference — and it is a difference that the vocabulary of ownership is ill-equipped to register — is that the civilian system achieves the separation structurally, by allocating the axes to different actors from the outset, whereas the common law system achieves it doctrinally, by superposing two different systems of entitlement over the same physical arrangement. The doctrinal route is not a structural route. It imports a set of consequences that the structural route does not.Footnote ⁶²

A brief explanation of the common law comparator is needed. Because the doctrinal material deployed in this section is predominantly Australian (references in this section to the ‘High Court’ are to the High Court of Australia), a generalist reader might reasonably ask whether the analysis concerns Australian law specifically, rather than the common law of trusts more generally. The axis-allocation just described is not jurisdiction-specific: the concentration of title, administration and control in the trustee, the trustee’s personal liability on trust dealings, and the indemnity secured by an equitable lien are features of the trust throughout the common law world, in England as in Australia.Footnote ⁶³ What differs is the intensity with which the structural consequences have been litigated. Australia is the jurisdiction in which the trading trust — the trust deployed as the operating vehicle of an active business — has achieved the deepest commercial penetration, and in which the interaction between the trust structure and corporate insolvency law has therefore been forced, repeatedly, to apex-court level.Footnote ⁶⁴ English law shares the structural allocation but has not been required to administer it under equivalent commercial stress, and the leading modern refinements have accordingly been Australian. The Australian material is used here, not because the diagnosis is Australian but because Australia supplies a comprehensive litigated record of a common law system administering the consequences of the allocation.

5.2. The indemnity right as hybrid artefact

The principal consequence is the trustee’s right of indemnity. The indemnity right is the mechanism by which the common law trust addresses a structural fact that the bewind never encounters. Because the common law trustee holds title and bears personal liability on dealings with the property, the trustee requires a remedy that matches the personal liability to the assets from which the liability was incurred. Equity responds with the right of indemnity: a right vested in the trustee to be reimbursed from the trust fund for liabilities properly incurred, secured by an equitable lien over the trust assets.Footnote ⁶⁵

A common law trust is not a legal person. When the trustee of a trading trust enters a contract with a supplier or lender, the trustee contracts in its own capacity and is personally liable. Unless the contract effectively limits that liability, it is unlimited and not confined to the value of the trust assets. Equity’s response is the right of indemnity out of the trust fund for liabilities properly incurred, which takes two forms: recoupment, where the trustee has discharged the liability from personal funds, and exoneration, where the trustee applies trust assets directly in discharge. The right is secured by an equitable lien — a non-possessory proprietary security over the trust assets that arises by operation of law and ranks ahead of the beneficiaries’ interests. The lien is the hinge on which the insolvency litigation turns. Because it is a proprietary right held by the trustee in the trustee’s own capacity, it is capable of constituting ‘property’ of an insolvent trustee available, in a defined measure, to satisfy the trustee’s creditors. The disputes examined in this section — concerning priority, successor trustees and the scope of the statutory insolvency regime — are, at their core, contests over that lien.Footnote ⁶⁶

The indemnity right within the three-axis framework is a hybrid. It is neither pure administration nor pure benefit. It is administration backed by a proprietary claim; it is the trustee’s personal exposure to the world translated into a lien on assets to which the trustee is not beneficially entitled. It exists, in other words, as the doctrinal residue of the concentration of axes that the common law trust effects.

The Carter Holt Harvey decision is revealing on this point.Footnote ⁶⁷ The High Court of Australia confirmed that trust assets constitute ‘property of the company’ to the extent of the trustee’s right of indemnity. The trust assets are, in that carefully calibrated qualification, partly property of the corporate trustee. The qualification is extraordinary from any system-neutral perspective. How can assets beneficially owned by one actor simultaneously be ‘property of’ another actor, even to a limited extent? The answer is that the trustee’s title and personal liability on dealings have generated a proprietary claim over the trust assets — the indemnity right — that the insolvency regime then treats as ‘property of the company’ for the purposes of Section 433 of the Corporations Act 2001 (Cth). The qualification follows from the concentration of axes. It would have no foothold in a bewind, because there the administrator has no title, incurs no personal liability, has no indemnity right and therefore has no proprietary claim over the administered property that an insolvency regime could be asked to recognise.

5.3. The Naaman decision and the trajectory of uncertainty

The Naaman v Jaken Properties Australia Pty Ltd decision illustrates the same point from a different angle.Footnote ⁶⁸ The question before the High Court of Australia was whether a successor trustee owes a fiduciary duty to preserve a former trustee’s indemnity against the trust fund. A 4:3 majority held that it did not. The narrowness of the division — on a question that affects the insolvency outcomes of hundreds of thousands of commercial structures — is significant. It signals that the doctrinal refinement of the indemnity right has not converged on a stable state. That is not a failing of any bench of judges; it is the predictable output of a doctrinal apparatus attempting to administer a structural misalignment.

The Naaman 4:3 split is precisely the kind of datum that the three-axis framework renders intelligible. The successor-trustee-duty question concerns the status, upon a change of trustee, of the indemnity-backed proprietary claim of the former trustee. Under the common law trust, that claim is a proprietary interest in the trust fund — partial, derivative, contingent, but real — and its fate on the change of trustee is not structurally determined. The bewind does not face this question, because no indemnity-backed proprietary claim is generated by the administrator’s position in the first place. If the administrator is replaced, the administered property remains the property of the rechthebbende; the former administrator has no claim over it; the successor administrator assumes the office and continues the function. The simplicity of the bewind’s treatment of administrator replacement is not a superior doctrinal solution. It is the absence of a problem that the common law trust, by virtue of its concentrated allocation, has been constrained to generate.

5.4. The control axis and the governance question

A further feature of the common law trust that the three-axis framework exposes concerns the allocation of control. In a common law trust, control over the res is substantially vested in the trustee. The beneficiary’s remedies are equitable: a right to compel due administration, a right to proprietary tracing against third parties and the exceptional right under Saunders v Vautier Footnote ⁶⁹ to collapse the trust and direct the trustee on distributions if the beneficiaries together hold the entire beneficial interest and are sui juris. Control and enforcement are not the same axis.

The concentration of control in the trustee is the counterpart of the concentration of title and administration. Equity’s response to the concentration — the battery of fiduciary constraints on the trustee’s exercise of control, enforceable by the beneficiary — is a doctrinal compensation for a structural choice. It is not an alternative to the structural choice; it operates within it. That is why the compensations, while often effective, are reactive. They engage after the trustee has made, or failed to make, a decision that affects the beneficiary’s economic position, and the beneficiary’s role is to seek remedies ex post.

5.5. The common law trust’s structural coherence within its own terms

It is important to observe that the common law trust is not structurally incoherent within its own terms. The concentration of axes in the trustee, coupled with the doctrinal response of equitable enforcement and the indemnity-backed lien, is a coherent apparatus that has served Anglo-Australian commerce for centuries. The apparatus is not defective relative to any internal criterion that the common law system itself applies. What the three-axis framework shows is not defect but structure: that the doctrinal complexity of the common law trust — the elaborate interlocking of legal title, equitable interest, fiduciary duty, equitable lien, priority contests and the insolvency machinery that the lien triggers — is the architectural cost of a decision taken at the level of axis-allocation.

What the framework also shows, however, is that the doctrinal resources cost something. They cost judicial time in cases like Carter Holt Harvey and Naaman, in which the apex court is asked to refine the consequences of the structural concentration without being positioned to alter the concentration itself. They cost commercial certainty, as creditors, beneficiaries and trustees navigate a liability architecture whose outcomes are not knowable ex ante. They cost legislative attention in the form of repeated reform consultations that do not produce legislation.Footnote ⁷⁰ The structural concentration is a feature, not a bug. But it is a feature whose costs become apparent when the arrangement is measured against an alternative allocation that does not incur them.

6.1. Generalising the insight

The argument developed in Sections 4 and 5 may be generalised. The common law commercial trust’s recurrent difficulties — the priority contest over the indemnity-backed lien, the Barnes v Addy architecture for third-party accessory liability (the equitable regime by which third parties who receive trust property with notice of a breach, or who knowingly assist in a dishonest and fraudulent design, are exposed to liability),Footnote ⁷¹ the creditor-subrogation mechanism (by which trust creditors, whose claims lie against the trustee personally, stand in the trustee’s place to enforce the indemnity against the fund), the insolvency uncertainties confirmed by Carter Holt Harvey and Naaman — are not isolated doctrinal anomalies. They are the predictable output of a structure in which axis-separation was never formally achieved. The three-axis framework generates a unified diagnosis of what has otherwise appeared as a family of unrelated puzzles.

This diagnostic function is the methodological contribution the article has been building toward. The contribution is that the difficulties are structural. They arise from a single source: the concentration, in the trustee, of title, administration and control, coupled with the doctrinal response (the indemnity-backed lien) by which the common law has sought to match the trustee’s personal exposure to the assets from which the exposure is incurred. Identifying the source as structural, rather than as doctrinal, has consequences for what counts as progress in the reform debate. Doctrinal reforms leave the source unaddressed; structural reforms do not; and the choice between them is not a choice that doctrine itself can adjudicate.

6.2. What the diagnosis does not claim

Two qualifications to the diagnostic claim should be stated explicitly. First, the diagnostic claim does not entail the conclusion that the common law trust ought to be replaced, restructured or modified. It is entirely consistent with the three-axis framework to hold that the structural concentration in the common law trust is, on balance, worth the doctrinal costs it imposes; that the commercial flexibility, tax treatment, historical embeddedness and private-law familiarity of the form outweigh the difficulties the framework makes visible. What the diagnostic claim does not permit is the continued treatment of those difficulties as isolated doctrinal puzzles amenable to local solution. The difficulties are not local; they are systemic.

Second, the diagnostic claim does not entail that the bewind, or any other civilian structure, is normatively superior to the common law trust. The bewind has its own costs: the narrower range of commercial applications that it has historically served; the civilian doctrinal architecture on which it depends and which common law systems do not share; the fact that its insolvency segregation is a structural feature rather than a flexible commercial choice. Whether those costs are worth the benefits is, again, not a question the diagnostic method answers.

6.3. Structural misalignment as a comparative law category

The broader proposal of this section is that ‘structural misalignment’ should be recognised as a diagnostic category in comparative law. It is a descriptive category that identifies, in each legal structure, the presence of doctrinal machinery whose function is to administer a structural choice the system has made but not articulated. Where the machinery is elaborate, highly developed, recurrent in the reform literature and persistently unstable under stress, the underlying structural choice is a candidate for analytical excavation.

The category is useful for comparative-law research precisely because it redirects attention from the machinery to the choice. Much comparative-law work on difficult trust questions has proceeded by surveying the machinery: the varieties of fiduciary duty, the architecture of tracing, the nuances of accessory liability and the transitions of creditor priority. The surveys are valuable, but they operate on the surface of the structural allocation. The three-axis framework and the diagnostic use of structural misalignment invite the comparatist to look beneath the machinery for the allocation.

6.4. A brief extension: other candidate misalignments

Within trust-analogous structures beyond the four analysed in Sections 4 and 5, further candidates for three-axis diagnosis include the Scottish trust, the Liechtenstein and Swiss Treuhand, and the various ‘international trust’ regimes of offshore jurisdictions.Footnote ⁷² Each can be mapped onto the framework. The comparative payoff is that these varieties cease to be a series of unrelated doctrinal idioms and become points on a structural map that the typology developed in Section 4.7 already begins to populate.

7.1. Three main methodological influences

The comparative-law literature has been shaped by three methodological influences. What follows is an overview, and a deliberately stylised one: each tradition is internally diverse, and the purpose of this section is not to survey the traditions but to locate the technique proposed in this article relative to their central commitments. The functional method, associated most closely with Zweigert and Kötz, proposes that comparative law should identify the practical problem to which a rule responds and then compare rules across jurisdictions by reference to how they address that problem.Footnote ⁷³ The legal formants approach, developed by Sacco, identifies the multiple and sometimes conflicting sources (statutory, doctrinal, judicial, scholarly) that together constitute a legal rule, and proposes that comparative analysis should work at the level of the constituent formants rather than at the level of the rule as conventionally stated.Footnote ⁷⁴ The legal transplants approach, developed by Watson, proposes that legal rules can, and frequently do, migrate between systems, and that comparative analysis should attend to the mechanics and consequences of such migrations.Footnote ⁷⁵ Each tradition has generated a substantial body of work; each has been subjected to its own line of critique, most prominently Legrand’s critique of transplants as culturally impossible;Footnote ⁷⁶ and each continues to shape comparative scholarship.

The three-axis framework developed in this article does not displace any of these traditions. It complements them. Its relation to the first, however, requires more careful statement than the language of complementarity alone supplies, which Section 7.3 provides.

7.2. The post-2010 methodological renewal

Since approximately 2010, a significant renewal of comparative methodology has responded to the limits of the three classical traditions. Husa’s two syntheses, A New Introduction to Comparative Law and Interdisciplinary Comparative Law, articulate a methodological pluralism in which the functional, formants and transplants approaches are treated as complementary analytical lenses within a broader reflective practice, rather than as competing paradigms between which the comparatist must choose.Footnote ⁷⁷ Siems, in Comparative Law, extends the methodological discussion to include quantitative and numerical comparative methods, arguing that doctrinal and empirical approaches are mutually informing rather than rival traditions.Footnote ⁷⁸ Valcke, in Comparing Law: Comparative Law as Reconstruction of Collective Commitments, reconceives comparative law as the reconstruction of the ‘collective commitments’ that constitute each legal system’s self-understanding, situating the comparative exercise as an interpretive rather than a strictly descriptive undertaking.Footnote ⁷⁹ Kischel’s Comparative Law develops a contextual method that integrates cultural, economic, political and historical framings into the comparative exercise, reviving in more analytically careful form the cultural-embedding concerns that Legrand had pressed in his critique of transplants.Footnote ⁸⁰

The three-axis framework proposed in this article sits comfortably within the pluralist methodological register the post-2010 literature has established. Structural contrast is at home within Husa’s pluralist register, adds a structural-allocation dimension to the empirical/quantitative/qualitative integration Siems defends, and is consistent with Valcke’s reconstruction thesis. The relation to Kischel’s contextualism is somewhat more complex: the three-axis framework deliberately abstracts from the cultural and historical embedding of structural allocations, whereas Kischel deliberately attends to that embedding. It should, however, be said precisely what is and is not being claimed when the framework ‘abstracts’ from cultural embedding, because the claim is easily overread. The framework brackets the cultural and doctrinal embedding of structural allocations for a defined analytical purpose; it does not assert that structure subsists prior to culture, logically or otherwise. Whether even that bracketing is possible is itself contested. A comparatist of Legrand’s persuasion would deny that the structural can be isolated, even provisionally, from the cultural formations through which any lawyer encounters it; Sacco’s formants analysis implies that what this article calls ‘structure’ is itself constituted by formants and cannot stand outside them; and Kischel’s contextualism presses in the same direction.Footnote ⁸¹ The defence offered here against that family of objections is pragmatic rather than metaphysical. The bracketing earns its keep by its output: it generates the four-pattern typology of Section 4.7, which the culturalist positions, whatever their other merits, have not generated; and the results of the bracketed analysis can then be re-embedded in precisely the contextual and cultural analysis those positions supply. The claim, restated in its defensible form, is that the structural allocation is analytically separable for the duration of the comparison — not that it is ontologically or logically prior to the culture in which it is embedded. Identifying structural contrast as a distinctive technique accordingly adds a focused instrument to the broader pluralist toolkit the post-2010 literature has assembled.

7.3. Structural contrast and the functional method

The functional method, in its classical Zweigert-Kötz form, operates by identifying equivalent problems and comparing the rules adopted by different systems to address them. The method has been criticised for its assumption of shared problems across systems and for its tendency to elide system-specific differences in problem-framing.Footnote ⁸² Michaels in his influential account has shown that the method has always presupposed a theoretical commitment to functional equivalence that is itself open to challenge.Footnote ⁸³ What the method does well is compare rules.

Two objections must be confronted at this point, because together they threaten the claim that structural contrast is distinct at all. The first is that functional comparison has not in fact confined itself to rules. Braun’s survey of comparative trust research documents functional comparisons pitched at the level of institutional structure rather than of individual rules,Footnote ⁸⁴ and Graziadei has shown that the functionalist aspiration was from the outset to cut across municipal legal categories and to separate the objects of comparison from their doctrinal vocabulary.Footnote ⁸⁵ The second objection arises from this article’s own terms. The framework developed here is avowedly functional: the axes are defined as functional variables, and the comparison proceeds by asking how functions are allocated. If the aspiration is functionalist and the variables are functional, is structural contrast anything more than functional analysis done with unusual care?

The answer given here concedes the premise of both objections. Structural contrast belongs to the functional family. It is not presented as a rival to functionalism, and nothing in this article should be read as claiming that the functional method is inherently incapable of comparing structures; Braun’s survey shows that it has compared them. What distinguishes structural contrast within the family is a conjunction of three specifications that the classical formulations leave open. The first concerns the unit of comparison: not the practical problem and the rules or institutions responding to it, but the allocation pattern of independently defined functional variables. The second concerns the operation and its output: classical functional comparison asks whether different systems reach functionally equivalent solutions, and its characteristic product is an equivalence claim; structural contrast asks how the space of possible allocations is partitioned, and its characteristic product is a typology. It requires no presumption of similarity, because difference rather than equivalence carries the diagnostic yield. The third concerns the direction of illumination: functional equivalence uses the comparator to show that another system solves the same problem; structural contrast uses the comparator to make visible what one system has structurally assumed and cannot see through its own doctrinal resources.

On this account, the shortcomings of earlier comparative work on trusts derive not from an inherent limitation of the functional method but from the absence, within its practice, of these specifications; and the contribution of this article is accordingly to refine and to deepen the functional analysis of trust structures by supplying them. Whether the result is described as a distinct technique within the functional tradition or as functionalism made precise is ultimately a question of labelling. What matters is that the specifications are articulable, that they discipline the comparison in identifiable ways, and that they produce a result — the four-pattern typology of Section 4.7 — which the tradition, across several decades of sophisticated engagement with the trust, had not produced without them. The remainder of this section uses ‘distinct technique’ in that qualified sense.

7.4. What the legal formants approach does not capture

Sacco’s legal formants approach identifies the composite character of any legal rule: the statutory, doctrinal, judicial and scholarly components that together constitute what the rule ‘is’ in any given system. The approach is subtle and has been applied with great analytical benefit to a wide range of problems.Footnote ⁸⁶ It is, however, oriented to the internal composition of a rule rather than to the structural allocation underlying a class of arrangements. Applied to the trust, the formants approach would yield a detailed map of the statutory, doctrinal and judicial components of trust law in each system compared. Applied to the bewind, it would yield a parallel map for the civilian variants. What the formants approach does not yield, because it is not designed to yield, is the structural allocation that determines why the maps are organised the way they are. The three-axis framework supplies exactly that. It is not a replacement for the formants approach; it is analytically prior to it. Once the axis-allocation is identified, the formants can be understood as the system-specific apparatus by which the allocation is given doctrinal expression.

7.5. What the legal transplants approach does not capture

Watson’s legal transplant approach has been central to comparative-law discussion of trust-like arrangements because the question of whether, and how, the common law trust can be ‘transplanted’ into civilian systems (or vice versa) is a recurrent one.Footnote ⁸⁷ Legrand’s critique — that legal rules are culturally embedded and cannot genuinely be transplanted — has generated an important alternative literature, although Legrand’s stronger claim (that transplants are impossible) is, on most readings, best understood as a challenge to the transplant approach rather than as its refutation.Footnote ⁸⁸

The three-axis framework supplies a partial resolution of the transplant-versus-impossibility debate. What can be transplanted, on the framework, is the structural allocation: the decision to separate administration from title and benefit, or not to do so. What cannot be transplanted is the doctrinal apparatus by which a system administers its chosen allocation, because that apparatus is constituted by the system’s doctrinal tradition and will not function in another tradition without substantial re-expression. The transplant debate, understood in these terms, is not a debate about whether civilian and common law systems can share rules; it is a debate about whether they can share structural allocations. The framework makes the question tractable by distinguishing the two levels. The Belgian 2013 reform, examined in Section 4, is a concrete instance: the Belgian legislature transplanted the structural allocation of the unified bewind from an internal precedent while deliberately discarding the doctrinal architecture of the earlier plural regimes. The transplant succeeded because it operated at the structural level; the doctrinal architecture was discarded because it was not what needed to be transplanted. That analysis, which the Belgian material directly supports, is not available without the framework’s separation of allocation from apparatus.

7.6. Structural contrast as a distinctive technique

The proposal of this section is that ‘structural contrast’ should be recognised as a distinctive technique within the comparative-law toolkit. It stands within the functional tradition, in the qualified sense defended in Section 7.3. The technique is particularly well-suited to comparison between civilian and common law property arrangements because these systems allocate functional variables differently while each presents itself internally as a coherent self-contained apparatus.

The technique has three features that set it apart. First, it operates without commitment to the doctrinal vocabulary of any single system. The three axes can be stated without using the categories of ownership, dominium, legal estate or beneficial interest, and the axes are stable across translations. Second, it produces structural rather than doctrinal comparanda. What is compared is the allocation pattern, not the rule. Third, it is diagnostic rather than normative. The output is the identification of the structural choice each system has made; it is silent on whether the choice should be revisited, preserved or changed.

7.7. A concluding observation on the relation between method and doctrine

The development of the argument in Sections 2 to 6 has drawn on a substantial body of doctrinal material, and it is fair to ask whether the methodological claim has been sustained or whether, despite the article’s disclaimers, the method has collapsed back into doctrinal argument.

The answer is that the method has been sustained because the doctrinal material has been employed as data for the framework rather than as conclusions drawn from it. The argument has not been that any particular rule of South African, Dutch, Belgian, English or Australian law is correct or mistaken. The argument has been that the pattern of allocation across the three axes is constant within each tradition and that the pattern is different between the civilian and common law traditions. The difference is a structural fact. The consequences that each tradition’s doctrinal apparatus derives from that fact are doctrinal facts. The framework separates the two levels. What the framework forecloses is treating structural allocation as a doctrinal fact rather than a structural choice. That foreclosure is the methodological contribution.