Not by Contract Alone: The Contractarian Theory of the Corporation and the Paradox of Implied Terms

Contractarians view the corporation as a nexus of contracts, constituted by the express or implied consent of each party to or contracting with it. Strong-form contractarianism takes this claim literally and holds that a corporation can be created and sustained by contract alone, thanks notably to the courts’ supportive gap-filling role. We argue that this view is undermined by the way courts actually treat implied terms. While courts do attempt to fill gaps and hold parties to their bargains, they do not typically manufacture counterfactual consent by resorting to the hypothetical bargain logic of contractarianism. Even under the most flexible form of contract law, the common law contract, the capacity of courts to imply third-party obligations in multi-party contracts is highly limited. This makes the contractarian reliance on contract and the courts to construct the complex set of multi-party obligations that make up the corporate form implausible.


Introduction
The 'contractarian' or 'contractual' theory of the corporation holds that a business corporation is a nexus of (incomplete) contracts between directors, shareholders, employees, suppliers, customers and other parties. 1 This idea, which is rooted in the economic theory of the firm, draws attention to the express or implied consent of all these participants and suggests that the role of corporate law and the courts is to enable and support private ordering: corporate law supplies the transaction costreducing standard form terms the parties would have chosen and agreed to had they addressed them explicitly, and courts settle disagreements by filling the gaps in the incomplete contracts comprising the nexus using the same hypothetical bargain logic. 2 Much of the existing critique of contractarianism has focused on the mandatory or public nature of corporate law. 3 In this paper, by contrast, we articulate a critique that targets the role contractarians assign to courts.
Specifically, we show that this role does not sit comfortably with the 'strong-form contractarian' position, 4 which demands a literal interpretation of the claim that the corporation is a nexus of contracts. Strong-form contractarians (to whom we shall refer to simply as 'contractarians' in what follows) reject the suggestion that their use of the term 'contract' is too loose to correspond to the legal understanding of an actual, legally enforceable reciprocal promise or agreement, 5 and claim that the contracts involved in corporate ventures are 'real contracts' 6 consisting of express or implied terms struck by real parties. When courts are called upon to imply missing contractual terms, contractarians argue, they do so with a wealth-maximisation objective in mind. We show that a considerable extension of standard contract doctrine is necessary for this account to work and contend that courts will refuse to extend it as contractarians require, with the consequence that many gaps either cannot be filled or can only be filled by non-contractual doctrines.
The problem can be roughly stated as follows. Given that contracts are incomplete, whenever courts adjudicate contractual disputes, they attempt to fill the gaps using the doctrine of implied terms in order to enforce mutually agreed obligations. But giving effect to obligations by implication comes with the risk that the implied term may not have been truly consented to. This undermines the very basis of contractual liability, namely consent, and conflicts with the courts' practical imperatives to maintain legal certainty and deter excessive litigation. Courts, as a result, will more often than not refrain from implying terms, even when this runs the risk of not holding parties to their bargains. While the imperative to hold parties to their bargains is rooted in consent, so is the imperative to refuse to imply terms. We call this antinomy the 'paradox of implied terms'.
As the number of parties increases, the effects of this paradox multiply, inherently limiting contract law's reach beyond two parties. This constraint limits the courts' ability to build a nexus by implying third-party obligations by inferring, for example, a contract between its existing members and a putative new member. It follows that the vehicle of contract cannot create the complex set of multi-party obligations which constitute the corporate form. Conversely, in a true corporation, parties can be added to (or removed from) the nexus without the express or implied consent of all other parties in the nexus, and without affecting their liabilities. This is only possible because non-contractual doctrines, that is, doctrines that are not rooted in consent, are involved. This poses a significant challenge to contractarianism.
In order to substantiate this argument, we first explain how the contractarians' key claims rest on economic theories of firms and contracts and show that the portrayal of the courts' gap-filling role relies on an efficiency theory of contract that conflicts with the objective theory of contract employed by English and other common law courts. We then demonstrate how the paradox of implied terms prevents courts from manufacturing counterfactual consent. Finally, we study the limits of the courts' ability to imply third-party obligations in real-life multi-party contract cases involving few and many parties, which we call 'micro-' and 'macro-nexuses' respectively. We use recent shipping cases as straightforward illustrations of micronexuses and old cases involving the unincorporated joint-stock company, often cited as evidence of the contractual nature of the corporation, as illustrations of macronexuses. We also test the claim that one other feature of the corporation, the fiduciary duties of directors, can be created by consent. Overall, our analysis shows how the macro-nexus that is a fully functioning body corporate cannot plausibly be created by contract with the support of the courts.

The Contractarian Paradigm
Contractarianism holds that 'everything is negotiable '. 7 From this perspective, to say of a corporation that it is a nexus of contracts is to refer, in shorthand form, to 'the complex arrangements of many sorts that those who associate voluntarily in the corporation will work out among themselves'. 8 Whether the contracts involve suppliers of capital or suppliers of labour or any other stakeholder, all these arrangements are bound by express or implied agreements: 'Some may be negotiated over a bargaining table. Some may be a set of terms that are dictated by [a party] and accepted or not … Some may be implied by courts or legislatures trying to supply the terms that would have been negotiated had people addressed them explicitly.' 9 Corporate law, from this perspective, is a 'set of terms available off-the-rack so that participants in corporate ventures can save the costs of contracting', thereby 'enabling the venturers to concentrate on matters that are specific to their undertaking' 10 such that their joint wealth is maximised. The role of courts is viewed in exactly the same light: parties 'call on courts to duplicate the terms [they] would have selected, in their joint interest, if they had contracted explicitly '. 11 To show that corporations are essentially contractual and that courts fulfil their gap-filling role with an eye on the parties' wealth-maximisation objective, contractarians supply historical examples and appeal to the authority of economic science.

Two Explanatory Strategies
Contractarians reject the old grant theory of corporations. According to this theory, the key features of business corporations, most notably separate legal personality, perpetual existence and limited liability, are public or state concessions of privileges. 12 Corporations are not creatures of the state, contractarians argue, they are a species of contract, just like any other voluntary human association. Furthermore, all the legal features of corporations, and indeed, corporate law itself, are shaped not by legislation but by the evolutionary pressures exercised by competitive market forces. These claims are substantiated in one of two ways (and often in both).
The first strategy involves the use of selected historical examples that show that features of 'corporateness' have been, and therefore can be, 'created by contract, supplemented perhaps by other common law devices such as trusts or agency'. 13 For instance, the fact that in England joint-stock companies, which exhibited most of the features of corporations, were formed under partnership law, often supplemented by trusts, in the period between the Bubble Act 1720 and the Joint Stock Companies Act 1844, is taken to suggest that 'the benefits of treating a business as something separate from its owners are so obvious … that it has never required substantial governmental assistance to achieve'. 14 Parliament's choice to extend incorporation to joint-stock companies did not alter the fundamentally contractual essence of these firms, but it did reduce some of the contractual costs involved. 15 These considerations pave the way for the contractarians' second strategy, which is to argue that the claim that 'market forces rather than legal ones have dictated [the] organization and structure' of corporations 16 is effectively an 'application of modern economic analysis'. 17 Contractarians turn to the economic theory of the firm, which holds that firms are special kinds of centralised contractual arrangements that emerge to enhance efficiency by reducing the transaction costs that would have otherwise been incurred, that is, if production were organised entirely through decentralised market exchange. 18 The rhetoric of scientific legitimacy found in this appeal to economic analysis enabled contractarians to produce a paradigm shift in the legal theory of corporations. 19

The Economic Theory of the Firm
The economic theory of the firm holds that, just like market exchanges will only occur if there are gains from trade, individual resource owners will only join forces with others in the pursuit of some common objective if the expected joint surplus is greater than the sum of what could have been otherwise achieved individually. 20 To the extent that each party is free to take their business elsewhere, there is no coercion but only a meeting of the minds: relations between all and any of the parties involved are the result of voluntary agreements. If the value of the joint surplus is to be maximised, a set of agreements that no one will wish to change must somehow be reached. Moreover, since various transaction cost-reducing arrangements are therefore valuable, their comparative assessment is desirable and indeed necessary.
The fact that one observes several forms of association between resource owners suggests that there is competition between the various forms of contractual arrangements available. Consequently, the contractual forms best suited to a large variety of needs and circumstances tend to perform better than alternatives and, through a process of 'propagation by imitation', tend to be adopted. 21 In the long run, given the parties' ability to alter their mutually beneficial relationships, only the most innovative and improved contractual features of business associations tend to survive in this dynamic competitive selection process. Although the ultimate distribution of forms of contractual arrangements is the unintentional outcome of intentional individual actions, at the level of individual contracts a high degree of consent among all the parties involved is essential.
One form of contractual arrangement that has survived the market test against alternatives involves the delegation by resource owners 'to a central agent, for some period of time, specific rights to direct their assets in production in return for a payment'. 22 By greatly reducing the number of requisite contracts in the organisation of production, contractual centralisation-from which follows the famous 'nexus of contracts' 23 expression that contractarians have adopted-reduces transaction costs, and is thus a key source of mutually beneficial efficiency gains. The centralised contractual structure is common to both the sole proprietorship, where the central agent is the entrepreneur, and the corporation, where the central agent is a 'legal fiction which serves as a nexus for a set of contractual relationships among individuals'. 24

Incomplete Contracts and the Role of Courts
The next step of the contractarian logic is to demonstrate how the precise content of this contractual structure is determined or altered by either express or implied consent, when it is 'tested in the courts'. 25 Here, contractarians turn to the economic theory of contracts, which explains how the contractual allocation of costs and benefits among parties with conflicting interests motivates them to perform mutually beneficial actions. The problem is that contracts, particularly the 'relational contracts' governing long-term relations, will always be to some degree incomplete, given the transaction costs of drafting very detailed agreements, the limitations of language and the impossibility of foresight of all future contingencies. 26 Whether contractual incompleteness is due to imperfect foresight or the deliberate will of the parties, 27 an implication of the fact that parties to a long-term contract are typically unable to write what economists call 'complete state-contingent contracts' is that contracts will often need to be renegotiated as unforeseen contingencies arise. In such circumstances, parties may have an incentive to take advantage of the renegotiation to increase their share of the joint surplus. In anticipation of such opportunistic behaviour, parties have an incentive to agree expressly on protective contractual solutions. If, despite such express agreements, parties still fall into dispute, contractarians argue that it is up the courts to supply the missing contractual terms and thereby to prevent the evasion of the true agreement as concluded ex ante.
While one strand of the economics of contract does not consider court behaviour, thereby assuming that courts mechanically enforce the agreements brought before them, 28 another focuses explicitly on the courts' interpretation of contractual terms 29 and their reliance on 'default rules' which specify the parties' obligations in the absence of any explicit agreement to the contrary. 30 In this literature, a key transaction cost-reducing function of law, which serves its role of facilitating allocative efficiency, namely the movement of resources into their most valuable uses, is to supply a set of standard terms that the parties would have had to negotiate explicitly.
The adoption of such a 'hypothetical bargain' approach 32 -sometimes referred to as the 'market-mimicking' approach 33 -is essential if courts are to 'complete incomplete contracts' 34 in a manner that gives effect to the parties' intentions, the assumption being that parties will always have intended to make themselves better off.
Since court intervention is itself costly and may involve errors, in the sense that courts may frustrate rather than give effect to the parties' intentions, contracting parties may wish to incur the transaction costs of additional specifications of performance. On the other hand, when transaction costs are high and contracting parties have good reason to believe that the likelihood of courts committing errors is low, contracts may be incomplete by design. Indeed, even for foreseeable contingencies 'it may be cheaper for the court to "draft" the contractual terms necessary to deal with the contingency if and when it occurs'. 35 Ultimately, the division of labour between contracting parties and the courts is driven by efficiency considerations, as is the courts' behaviour. Even if the courts had little commitment to efficiency, contractarians argue, appeals from dissatisfied litigants would push them toward implying wealth-maximising terms. 36

The Paradox of Implied Terms
One can see why the claim that legislators promote and courts enforce market-mimicking rules is intuitively appealing. If one subscribes to the view that the actions of self-interested individuals in well-functioning markets is to ensure that resources are allocated to their most valued uses through voluntary exchange, and that aggregate social welfare is maximised as a result, it is what courts should do. But 'the intuitive appeal of the argument does not substitute for secure analytical foundations '. 37 In what follows, we further clarify the assumptions of the contractarian paradigm, and explain why the contractarian claim that courts fulfil their gap-filling role with an eye on the parties' wealth-maximisation objective is far from convincing, even in a common law system, the kind most compatible with contractarianism.
We show that the gap-filling role contractarians assign to common law courts is likely to conflict with their other imperatives, namely their need to maintain legal certainty and deter excessive litigation. Indeed, when a court is asked to supply missing contractual terms, it is effectively being asked to rewrite a bargain by manufacturing consent. But implying a term which may not have been truly consented to undermines the very basis of contractual liability, consent. Faced with this uncertainty, courts will typically refuse to imply terms. This presents a problem for the contractarian narrative, which does not take into account the possibility that courts 34 Masten (2000), p 32. 35 Posner (1986), p 82. 36 Posner (1990), p 360. 37 Kornhauser (2000), p 88. 32 Charny (1991). 33 Craswell (2000), p 1. will refuse to imply terms, and therefore weakens the claim that, with the exception of liability for tort and restitution, everything to do with corporations, including the competitive process of selecting contractual arrangements, rests on express or implied consent.

Will vs Efficiency in Contract
Although all forms of multi-party obligation can be reduced to a set of individual rights between persons, 38 it does not follow that all such obligations must be based on consent. However, if the contracts in the nexus of contracts included non-consensual liability, this would impede the parties' ability to choose the model of association they prefer and inhibit the competitive selection process described above. 39 For the grounding of contractarianism in economic theory to work, the 'contracts' in the nexus must be sufficiently close to contracts of consensual liability, i.e., to 'real contracts'. Not accepting this proposition opens the contractarian paradigm to the accusation that it wishes to cherry-pick the aspects of contract law that support it without taking the aspects that do not.
The claim that courts will give effect to the private ordering the parties desire, and will second-guess it when that desire is not expressly spelled out, is based on two underlying philosophies of contract. 40 The first, which can be traced back to the formative cases of English contract law towards the end of the Industrial Revolution, 41 is that the state, and its courts, value individual liberty, and have a moral commitment to the enforcement of reciprocal promises made by autonomous individuals. This idea is behind the 'will theory of contract' and its variants, 42 which sit well with the general philosophical commitments of contractarians.
But since courts, from this perspective, hold parties to their promises retrospectively, being brought into the picture to adjudicate disputes after these have arisen, will theories of contract cannot explain the content and function of default rules. 43 By contrast, the 'efficiency theory of contract', derived from economic theory, holds that ex ante wealth-maximising default rules of contract law would have been chosen by the parties had they been able or willing to do so. 44 This is the theory of contract underpinning the hypothetical bargain approach, which is thus able to overcome the deficiency of the will theory while retaining, in appearance at least, the general philosophical commitment to individual will.
A conflict between will and efficiency theories of contract appears in situations where the parties have written a bad or incomplete bargain, in the sense that welfare 38 Hohfeld (1919). 39 See Sect. 2.2. 40 See Kraus (2004). 41 Baker (2002), Atiyah (1979) and Oman (2016). 42 Or 'autonomy theory of contract'. Classic statements include: Cohen (1933) and Radin (1943). The most influential defences today are: Fried (1981) and Barnett (1986). 43 Craswell (1989). 44 Long (1984), Craswell (1992), Riley (2000), Farber (2000) and Schwartz and Scott (2003). improvements remain possible. On the efficiency theory, a 'liberal court' should 'freely' imply the missing terms that maximise the parties' joint wealth. 45 A consistent efficiency theorist might further argue that even the written terms of the contract should not be enforced if courts are able to improve on the outcome that the parties would achieve without their intervention. 46 The will theorist would object to this form of ex post counterfactual consent and argue that, given the principle that contractual liability must arise from true ex ante consent, the courts' role is to hold the parties to their expressed bargain, good or bad.

The Potential for Reconciliation in Doctrine
Although the tensions between will and efficiency theories of contracts continue to preoccupy philosophers of law, 47 a potential reconciliation may be found in doctrine. If the legal system were able to accommodate both will and efficiency theories, then instead of requiring legislation and the registration of the body corporate, a corporation might conceivably arise from the brute fact of the joint intentions of the parties, forming a nexus.
Some civil law systems are based on a subjective theory of contract. For instance, the French system interprets the intention of the parties in forming terms as stemming from their subjective volonté psychologique (psychological will). When interpreting contracts, the Code Civil provides that '[o]ne must ... seek to ascertain the [actual] intention of the contracting parties'. 48 For instance, if an offeror changes her mind about contracting but has not communicated the revocation to the offeree, this is nonetheless effective to cancel the offer. Other doctrines must step in to deal with any detriment suffered by the offeree in this situation and others.
Hence, the subjective theory of contract precludes the effective functioning of contractarianism. It is too brittle. On the one hand, it subordinates any hypothetical bargain to the strictly subjective intentions of the parties. On the other, it is ready to impose non-consensual doctrines such as delict or unjust enrichment. Its anti-evasion strategy, while extant, is not one that tends towards upholding the ostensible contract on terms. It does the opposite, unwinding the contract and raising another money claim for money expended in reliance of it. This can be contrasted with the common law approach.
In England, the law of contract took on the laissez-faire principle that the state should not intervene save to uphold the arrangements the parties entered into out of their own free will. 49 Nonetheless, despite ostensible adherence to true subjective agreement, English common law developed a looser adherence to the will theory known as the 'objective theory of contract', 50 which leads to a different response to 48 Art. 1156. See Valcke (2009), p 72. 49 Dickinson v Dodds (1876) 2 Ch D 463 (CA) 472. See Simpson (1987). 50 Gordley (1993). 45 Posner et al. (2000), p 127. 46 Anderlini et al. (2011). 47 Kraus (2001) and Klass et al. (2014). the above example of offer and acceptance: 'If whatever a man's real intention may be, he so conducts himself that a reasonable man would take the representation to be true [he] would be … precluded from contesting its truth.' 51 The US followed the same path. 52 Consequently, such a contract would be effective and there would be no need for nakedly non-consensual legal rules to step in. The requirement of actual consent was supplanted by the need only for a reasonable appearance of it, which is said to be enough for its moral justification. 53 This approach has proven to be extremely successful. Not only that, common law jurisdictions such as England, the US 54 and Australia 55 extended this analysis to their gap-filling doctrines, characterising the implication of terms in fact as simply interpreting, using the objective method, what the parties would have recorded had they thought about it. 56 It is thus clear that a common law system of contract is more likely to uphold an uncertain contract than a civil law system. The flexibility of the objective method to construe, through appearances, unexpressed consent as agreement to certain contractual terms, is what the contractarian paradigm needs. Contractarianism requires a commitment to individual will-but not such a brittle adherence that the remedies for less than a total meeting of the minds are so drastic as to destroy the contractand a doctrine of implied terms that fills gaps along the same lines. There is scope in a common law system for the requisite balance between efficiency and will theories. But the question of whether courts actually reconcile the two in the manner that fits the contractarian narrative remains.

The Constraint of Adjectival Law
Contractarians maintain that the division of labour between contracting parties and courts is, and should always be, governed by an efficiency or a transaction costreducing imperative. By downplaying the possibility that the legal process may be governed by its own norms and imperatives-which need not coincide with, or be related to, the parties' wealth-maximisation objective, or correspond to what the parties would have contracted for-contractarians open the door to a number of 51 Smith v Hughes (1871)  objections. 57 One important objection, stemming from the constraints of adjectival law, has hitherto received little attention.
There is a sense in which the contractarian's tendency to take the legal process for granted is justified. Indeed, one of the features of a mature legal system in a liberal democracy is that it is possible to do just that; one can get on with business without worrying about whether the legal system is capable of supporting such an endeavour. Parties can be confident that their counterparties' obligations will be enforced, and that general principles of law will govern this process. A key contribution of the rule of law both in general and for contract law specifically is that it provides a degree of certainty. 58 Legal certainty benefits wider society, in that it reduces costs and delays. It also benefits the courts themselves, because it reduces the number of cases and appeals. 59 Excessive discretion or variability is anathema to this legal certainty.
Given that the roles of adjudication and enforcement are assigned solely to courts, 60 they are in a position of dominance. The truism that greater power relative to a counterparty usually enables one to obtain one's desired terms applies, somewhat unsurprisingly, to courts as well. Their word is final, save for legislative revision. Since courts do have their own imperatives, even if they strive to give effect to the parties' intentions and their bargains-as indeed they do-if that conflicts with their imperatives, courts are in a position to prioritise those.
These imperatives arise from considerations of procedure, that is, from adjectival law, which influences substantive law by limiting the range of practicable substantive rules. 61 Some adjectival law is made for convenience. For instance, the rule in English law is that pre-contractual negotiations are not admissible in a claim for the interpretation of a contract, which is justified only because of the additional time and trouble the extra work would bring. 62 More often than not, however, adjectival law springs from necessity. The old maxim that determines the burden of proof-'he who avers must prove'-is a response to the near-impossibility of proving a negative. One important implication of such constraints on substantive law is that there are limits to the courts' use of the objective method. 57 We do not wish to argue that contractarians ignore completely the matter of the courts' own agenda but only that contractarians rely on the same hypothetical bargain logic to explain it away. For instance, although Macey (1993) acknowleged that 'when judges and lawmakers formulate non-contractual rules of corporate governance they inevitably import their own norms', he argued that this was still done under the belief that it is 'what entrepreneurs and investors … would have contracted for' (p 32). 58 In the absence of the rule of law, the lack of certainty implies high transaction costs and is endemic of underdeveloped economies. See North (2005) and Dam (2006). 59 Medcalf v Hall (1782) 3 Doug KB 112, 99 ER 566. See also Thomas (2016). 60 The parties may agree to having disputes settled by arbitration. While arbitral tribunals may have different objectives to state courts, such as the willingness to preserve the parties' anonymity and keep the hearing and evidence private, the matters with which we are concerned ought not to be received differently. This is because there is typically a right to appeal to a state court on a point of law: e.g., in the UK, Arbitration Act 1996, ss. 45, 69. 61 Baker (2002) and Main (2010). 62 ICS (n. 56) 912.

The Limits of the Objective Method
One undoubted utility of the objective method is that, in the absence of telepathy, it is the most reliable proxy for the parties' true intentions. 63 Equally undoubtedly, it can be, and has been, deployed to loosen contract's adherence to actual consent. To illustrate, consider a company seeking to profit from a government green grant and special feed-in tariff, which contracts with a supplier to install a wind turbine and connect it to the national grid. Given industry regulations, the national grid will not connect the turbine unless it is up to standard. The contract is silent as to which party is responsible for ensuring this. It takes only a moment to work out that the supplier is responsible. The supplier is providing the turbine; only the supplier is in the position to ensure it is up to standard. This is what the parties would have reduced to writing had they thought about it. If a dispute were to arise, the court would easily complete the otherwise incomplete contract.
Consider now a more complicated real case, Baird Textiles Holdings Ltd v Marks & Spencer plc. 64 Baird Textiles had a long-term relational contract with M&S: it was one of M&S's four trusted clothing suppliers and for some 30 years had supplied 15-18% of M&S's clothing range. In their bid to return to profitability following some financial difficulties, M&S terminated the contract with Baird and sought other suppliers. Baird sued, claiming there was an implied term providing for a reasonable notice period that had not been complied with. The evidence of cooperation and thus the existence of a notice period was considerable, but the fatal flaw of the claim was the lack of certainty regarding price and quantity. There were no objective criteria available to assess such things, and this led the court to refuse to impute ex post any agreement to such a term based on what the parties would have consented to, had they thought about it.
Although this example is taken from English law, the principle it illustrates, namely that courts operate under a requirement of certainty, is universal. 65 In England and Wales, 66 New Zealand, 67 Australia 68 and Canada 69 there is also a requirement of necessity. 70 The US jurisprudence is relatively underdeveloped and its leading texts refer to the analysis in the English cases as developing the fundamental principles of the doctrine, no doubt due to their shared heritage. 71 However, where economic arguments about what parties would have agreed to have been considered, they have been subordinated to this search for intention, justified by the principle that it is not the courts' place to put words in the parties' mouths 72 nor to imply terms in the face of express intent to the contrary. 73 It therefore seems that the legal principles are universal in common law systems and English law may be ahead of the game.
In the case of the wind turbine, it is beyond doubt that the supplier is responsible for ensuring the relevant safety standards are met. But for more complicated issues, the matter may not be so clear. Who is responsible for obtaining the necessary planning consents does not 'go without saying' (a proxy for the necessity requirement) 74 and the contract can function without it being the supplier. If a necessity standard is also applied, then such a term will certainly not be implied. Instead, there will be no liability for either party to perform that obligation, and any loss will lie where it falls.
The necessity standard operates as a useful proxy for certainty 75 and confidence. 76 If it is necessary for the contract to function, it is likely that, even if they had not thought about it, the parties would have assented to such a term. If so, even under the traditional will theory, liability would be justified. But the necessity standard goes further and helps soften an uncomfortable conclusion. Suppose the supplier positively did not intend to ensure the turbine was up to standard. The objective process holds that a reasonable observer, dispassionately examining the facts, would conclude that the supplier ought to be responsible because it is the only practicable outcome. In such circumstances, insisting that the court is merely implying what was intended helps disguise the conclusion that it is imposing its own rules in the absence of actual consent to liability, meaning that the usual justification of consent underpinning liability falls away. 77 It also rationalises as a form of consent the previous justification for imposing liability in the teeth of positive intention not to be bound in these circumstances: estoppel, on the basis that the party conducted herself in such a way that liability is justified on the basis of fault. 78 The question posed is how far from actual ex ante consent the objective method can go. One may argue that even if the courts have little commitment to efficiency, they will be subject to the pressures of increased appeals from dissatisfied litigants, at which point they will have the opportunity to reconsider their inefficient rulings. 79 Consequently, their judgements will tend toward efficiency. The ostensible requirement for consent is satisfied because greater efficiency results in greater wealth overall, something which most parties would indeed consent to. 80 At first blush, it appears that the anti-evasion function of the objective method is well placed to accommodate this view. But to answer the question posed, one must reduce the discussion to first principles.

The Paradox
The preceding considerations hint at a more general problem inherent in contract doctrine, which can be usefully expressed in the following manner. The courts wish to give effect to the parties' bargain. To do so, they must imply terms to fill the gaps in the express agreement with the parties' mutually intended obligations (Imperative 1). 81 But courts must not do so where the risk of implying terms contrary to the parties' intentions, i.e., imposing non-consensual liability, is too great (Imperative 2). 82 The courts' internal objectives are to avoid uncertainty in the law and to deter excessive litigation (Imperative 3). 83 In examining how these imperatives direct the court, consider the basic function of the trial process, the search for facts. The imperatives import the empirical matter of the standard of proof, which, in civil trials, is a low one-only whether the veracity of the contended fact is more likely than not. Facts are thus 'made' by the court as much as found in the process of weighing up the evidence. 84 The low standard 79 Posner (1990). 80 Unanimous acceptance of wealth-maximising solutions is not required. See Posner (1980) Posner's (2003) analysis of how judges judge (pp 59ff), arguing for the importance of predictability in the context of contractual interpretation (see particularly p 62). 84 Roberts and Zuckerman (2010), p 135. of proof carries the risk that terms will be implied in the absence of true or even merely objectively apparent consent.
A benefit of the necessity requirement is that it is a way of mollifying this problem by asking a question that demands more accessible and easily verifiable evidence, rather than an enquiry into the states of mind of the parties. In sum, there is clearly a potential difference between what the court decides and what the parties agreed to. In the context of making a finding related to consent, the risks are twofold: finding consent where there is no true consent; and, conversely, finding an absence of consent where there is in fact true consent.
The courts must balance these imperatives. To give effect to the parties' intentions, the courts would need to cast off the restrictive tests of necessity and certainty for implication. This would achieve Imperative 1. But rather than achieving the second, it would do the opposite. The courts would start rewriting bargains, implying terms the parties might not have consented to, contrary to Imperative 2. This would undermine the very basis of contractual liability, consent. Conversely, keeping restrictive tests for implication runs the risk of failing to give legal force to the parties' actual intentions for the bargain. Yet both imperatives are rooted in the same thing-the need for consent. This antinomy, which produces 'a self-contradiction by accepted ways of reasoning', 85 is what we call the paradox of implied terms.
Simply put, when faced with the conflict between the two imperatives, courts are unable to adjudicate on the basis of just one imperative, to the total exclusion of at least some elements of the other. Given that it is impossible to achieve both imperatives, courts have applied their own, Imperative 3, and concluded that refusing to intervene when there is insufficiently clear assent to the term is the lesser evil. 86 Imperative 3 thus tips the balance between Imperatives 1 and 2 towards Imperative 2.
As summarised in Fig. 1 below, courts err strongly on the side of caution and will refrain from implying terms, unless they have a high degree of confidence they can reconcile these issues on the individual facts of the case. They may also affirm a necessity standard in order to ameliorate the time, trouble and risk from having a less restrictive test. In England, following A-G of Belize v Belize Telecom Ltd, 87 it was thought by some, including some courts, that the restrictive tests in English law had been relaxed. 88 However, the Supreme Court soon moved to reverse this and affirm the traditional requirements in Marks and Spencer plc v BNP Paribas Securities Services Trust Co (Jersey) Ltd, where it signalled concern over the dangers of over-implication. 89 This state of affairs contradicts the claim that even if the courts had little commitment to efficiency, the greater number of appeals from dissatisfied litigants would push them toward implying wealth-maximising terms. 90 Courts have every incentive, and the power, to resist pressure from dissatisfied litigants, and in consequence prefer to maintain legal certainty rather than rewriting contracts efficiently.

The Nexus in the Shadow of the Paradox
We now appraise, in the shadow of the paradox of implied terms, the fundamental contractarian proposition that the corporate form was, and therefore can be, created by contract alone, supplemented perhaps by other common law devices such as trusts or agency. 91 To carry out the evaluation, we consider how courts have behaved in real-life multi-party contract cases, which we call 'micro-' and 'macro-nexuses', involving few and many parties respectively. We also consider one other significant feature of the corporation, the fiduciary duties of directors. We take the unincorporated joint-stock company of the seventeenth and eighteenth centuries, often cited as evidence of the contractual nature of the corporation, 92 as an example of a macronexus, and use recent cases found in shipping as illustrations of a micro-nexus.
The older cases show the obstacles implication faces in creating a corporation and how the legal system fared in working with them. The recent cases, decided after the shedding of Victorian-era formalism and procedural restrictions which inhibited flexibility, show the limits of the modern courts' ability to imply third-party obligations. Despite the nexuses' different factual contexts, from the perspective of  Posner (1990). 91 Mahoney (2000). 92 E.g., Anderson and Tollinson (1983), Butler (1986), Blumberg (1986) and Mahoney (2000). upholding consensual obligations and being wary of the paradox, the material differences only go to the size of the nexus. Only by considering the best of both can the contractarian claim that the courts are able to supply sufficient implied terms to create a corporation be fairly tested. We show that the paradox bites in both classes of nexus cases; that it also does in the matter of fiduciary duties is an added bonus.
The cases considered show three things: (1) how lack of the parties' knowledge of the terms means there can be no implication; (2) how narrow the obligations that can be implied must be in order to avoid imposing non-consensual liability or the risk of doing so; and (3) how the courts have tended away from contract and therefore consent-based law when alternatives were available. Even under the pressure to develop solutions, the courts were prevented by the paradox of implied terms from basing those solutions in consent, which supports our conclusion: the corporation cannot be created by contract alone.

Micro-Nexuses: Bills of Lading
Micro-nexuses are particularly apparent in cases concerning shipping bills of lading, which emerged from commercial custom. 93 Unlike bills of exchange, which are fully negotiable-meaning the entirety of the benefits and burdens of the contract are transferred to the indorsee by operation of law-until as recently as 1992 such broad rules did not fully affect bills of lading in England. After indorsement of a bill of lading, the indorsee took the title to the cargo but not, without more, the other benefits or burdens of the contract of carriage. 94 Nonetheless, the indorsee could still deal on the strength of the bill. This enabled the use of bills of lading to obtain secured credit, because the bill could be pledged or the goods charged. Where further obligations were contended, in the absence of the later legislation, it was for the parties to argue for an implied contract, which would in effect create a micro-nexus.
The courts have overcome the technical doctrinal problems such as the doctrine of privity, and the need for offer, acceptance, consideration and intention to create legal relations, when there are 'very powerful grounds' to 'give business reality to the transaction' and imply a contract. 95 In many cases, including shipping cases, the inevitable need to deal with multiple parties also supplies the necessity requirement. 96 Moreover, the presence of a written 'seed' contract-evidenced in the bill 93 There are also the 'horizontal contract' cases, but these are less instructive and space constraints preclude their discussion. of lading-overcomes the uncertainty problem seen in Baird. 97 Consequently,in Brandt v Liverpool,Brazil and River Plate Steam Navigation Co Ltd, all the terms of the bill of lading were incorporated into the contract between the indorsee of the bill and the carrier. It would have been unreal, from a business perspective, for payment to have been for release of the lien without importing any terms about the description and condition of the cargo. 98 Here, Imperative 1 is dominant. The issues are then concerned with reach-with whom a contract will be implied, and whether it will include every term, or if Imperatives 2 and 3 will outweigh it.
In Scruttons Ltd v Midland Silicones Ltd 99 implication was refused. In this case, the liability of the carrier to the consignees was limited by the bill of lading. The stevedores, contracted to the carrier, negligently damaged the cargo. Absent a contract between stevedores and consignees, the stevedores were liable in tort to the consignees for the whole of the damage, which was in excess of the liability limit. The stevedores sought a declaration of contractual relations with the consignees that would include the limitation clause. The House of Lords refused the claim on the grounds that the consignees 'knew nothing of the relations between the carrier and the stevedores'. 100 They had not agreed to bringing third parties into their nexus. The solution mooted was agency. The stevedores might have been brought into the nexus if the carrier was understood to have contracted with the stevedores as their agent, incorporating any limitation clause and other terms. This approach was successful in New Zealand Shipping Co Ltd v AM Satterthwaite & Co Ltd (The Eurymedon). 101 However, the so-called 'Himalaya clause' that makes this possible requires the creation of a tightly defined agency relationship between third party (as principal) and carrier for the purposes of protecting the third party. 102  [1975] AC 154 (PC). The clause in this case was outright exemption after one year had elapsed. The dispute in the final court of appeal concerned the existence of consideration. The majority held that an additional contract (in the nexus) arose through the offer in the bill of lading accepted by the performance of the stevedores. The consideration for the shipper was the benefit of unloading in return for the stevedores taking the benefit of the exclusion clause. This shows how the courts are unhappy with an overly technical approach to the law of contract and will strive to find ways to overcome such difficulties. 102 'Himalaya clause' comes from the name of the ship in Adler v Dickson (No 1) [1955] 1 QB 158 (CA). The requirements are: (1) it must be clear that the third party is to be protected; (2) the carrier acts as agent for the third party and this protection is passed through the agency to the third party; (3) the carrier has authority or ratification for the agency from the third party; and (4) consideration moves from the third party.
Without such explicit reliance on the principles of agency, the notion of objective consent would be stretched too far, offending Imperatives 2 and 3. On initial inspection, the consignees in Scruttons appear to have been given a windfall: they agreed to limit their remedies and escaped the stipulation they had consented to. But a closer look reveals some material differences to The Eurymedon, where the conditions were such that the rule of agency applied. The consignee can choose a seller, has the opportunity to negotiate over a carrier and limits its remedies with those known parties. However, the consignee has no control over any subsequent parties, who may be less reliable. The carrier is free to choose these, increasing the risk to the consignee. Without the clearest consent given by the consignee to the carrier to engage unknown third parties, it is difficult to say that it (or an equivalent party) has consented to the arrangement. It does not go without saying; the consignee might consider that if she is to take on this additional risk, she should buy insurance or negotiate the price down. The stringent requirements of the Himalaya clause go some way to ameliorating this problem. 103 Thus any obligations implied into a contract with a party joining the nexus will be narrow and only what the courts can be confident the relevant parties, old and new, would have accepted. They would rather defer to ex ante consent and leave the loss to lie where it falls than impose ex post such nakedly non-consensual liability. If there is to be private ordering in this situation-where a party prefers to join the nexus on more advantageous terms-the courts in effect insist on explicit prior authorisation in accordance with the strictures of the appropriate, non-contractual doctrine.

A Macro-Nexus: The Unincorporated Company
The proposition that a corporation can be created, in practice, from a nexus of contracts alone is not new. By upholding many facets of the unincorporated company, also known as the deed of settlement company, English courts had already travelled a fair distance down this path by the mid-nineteenth century before the process was cut short by the Companies Acts. However, if one examines the issues more closely, the paradox of implied terms bites: it is impossible to propagate reliably the features of corporateness, specifically strong entity shielding (liquidation protection) and owner shielding (limited liability), 104 to all parties involved.
The unincorporated or deed of settlement company constituted a body of sorts, with the facility for 'shareholders' to subscribe and unsubscribe from it. This provided some degree of alienability of shares and thus the ability to raise capital. 105 Technically, such bodies were partnerships, subject to all the features and problems of partnerships. Entity shielding was weak. 106 There was only the judge-made 'jingle rule' which provided that partnership creditors had to execute against partnership assets first and personal creditors had to execute against personal assets first. 107 The solution was to adopt the trust. Third parties were taken to have intended to contract with the trustees. 108 Vesting company property in trustees provided a somewhat stronger form of weak entity shielding because the courts of equity would not allow the execution of trustees' personal debts against trust property at all. 109 But it certainly did not constitute strong entity shielding. 110 Company assets were not locked in, 111 meaning a 'director', qua partner, could, in most circumstances, dissolve the partnership at will 112 or threaten to do so, causing disruption. 113 That left the matter of limiting 'shareholder' liability to the share value (owner shielding). By 1852 this could be done, to some extent, via contractual allocation. 114 But using express terms increased transaction costs, and there was the clear possibility of inadvertently omitting them. Under the trust system, the trust's terms, recorded in the deed of settlement, provided for limited liability. The company would then attempt to fix creditors with notice of those terms by including 'limited' in the company name and documentation, whereupon the courts of equity would hold creditors to those terms. 115 The doctrine marshalled in support of limited liability was thus not that of implied terms, but the equitable doctrine of notice. 116 106 Hansmann et al. (2006). 107 Craven v Knight (1683) Hansmann et al. (2006). 111 Blair (2003). 112 Featherstonhaugh v Fenwick (1810Blisset v Daniel (1853) 10 Hare 493, 68 ER 1022; the court would go no further than delaying dissolution. A limited form of lockin was created by the decision in Van Sandau v Moore (1826) 1 Russ 441, 38 ER 171. By requiring the joining of all the partners to be joined to the action at law, dissolution was made potentially very difficult in practice for large concerns. See Morley (2006), p 2174. 113 Lamoreaux and Rosenthal (2006). 114 Hallett v Dowdall (1852) 18 QB 2, 118 ER 1; see D'Angelo (2014), p 327. 115 See Bubb (2015), pp 345-347. 116 It is not wholly clear how this worked doctrinally. Ordinarily, if one acquires an interest in trust property, as a creditor would, for value and in good faith and without notice of a breach of trust or fiduciary duty, one takes (as 'bona fide purchaser') the property absolutely, free of any trust obligations. Notice includes constructive notice, meaning knowledge the creditor ought to have known (implied from suspicion or from investigations that ought to have been done) and is thereby deemed to know. While the creditor then takes the debt subject to its equities, the clear difference is that here there has been no breach, and if the creditor tries to execute against more than the terms declare he should, it is he who commits the breach, not his counterparty. Nonetheless, some comfort for the workability of this doctrine can be drawn from the well-known cases of Ernest v Nicholls (1857) 6 HLC 401, 10 ER 1351 (HL) and Royal British Bank v Turquand (1856) 6 El & Bl 327, 119 ER 886, where the courts decided how much knowledge of the company's constitution could be imputed to a creditor. Moreover, in recent times the House Authoritative secondary sources do not go into any depth on the matter, 117 nor do they cite much case law, as there are, it seems, very few reported cases. 118 Occasional reference to the possibility of such a term was made, such as in Cape's Executor's Case, where its absence meant that 'shareholder' liability was not limited and thus the reach of the doctrine of notice did not have to be considered. 119 Another rare example is Re Worcester Corn Exchange Company. 120 Here, the creditor bank was fixed with notice of the limited liability provisions because one of its partners was a shareholder in the unincorporated company, not because of the use of 'limited'. The matter was raised cursorily and there is nothing in the court's judgement on whether this is a stronger or weaker case than merely having 'limited' in the company name. In Re Sea Fire and Life Assurance Co the court considered that the limited liability provision in the deed of settlement was ineffective to limit liability, but on the grounds that the wording was insufficiently strong. 121 We therefore have authority that this strategy worked, but little information as to its limits and no theorisation of it. Developing these points is a task we undertake here.
While the paradox of implied terms has been explained in the setting of contractual terms implied in fact, there is a clear, although incomplete, equivalence to the doctrine of notice. There is a firm need for a sufficient level of actual knowledge (matters the party subjectively knew) upon which to fix constructive notice (matters the party can be taken to have known or 'objective knowledge'). In both cases, liability must be sufficiently rooted in actual consent to liability, or limiting it, based on the party's knowledge, per Imperative 1. Beyond some point, the party will not be bound by the terms limiting liability because it did not know them or their extent and could not reasonably have discovered them and therefore cannot be taken to have consented to them, per Imperative 2. Of course, Imperative 3 applies in all cases. The three imperatives, and thus the paradox, therefore also applies to the doctrine of notice insofar as it applies to the starting point of determining what the party can be taken to have known and agreed to. of Lords has taken the view that the doctrine of notice could be used in a similar way in Barclays Bank Plc v O'Brien [1994] Maitland (1911). See also Harris (2000) and Morley (2006). Atiyah (1979), p 566, suggests briefly that had the Limited Liability Act 1855 not been passed, the courts might have been willing to imply terms limiting liability but this would have been tantamount to being the doctrine of constructive notice. Atiyah cites Ridley v The Plymouth, Stonehouse, and Davenport Grinding & Baking Co (1848) 2 Ex 711, 154 ER 676, as supporting this hypothesis. 118 We used Westlaw to execute the following searches in the English Reports but found no other cases: Court = Chancery, dates between 01/01/1700 to 01/01/1860, "limited liability"; Court = Chancery, dates between 01/01/1700 to 01/01/1860, "limited liability" "deed of settlement"; Court = Chancery, dates between 01/01/1700 to 01/01/1860, "restricted liability" "deed of settlement"; Court = Chancery, dates between 01/01/1700 to 01/01/1860, "deed of settlement" "notice" "company". 119  There are four core situations where the paradox may well prevent the full joining of another party to the nexus. The first is that there may be opportunities for a creditor to escape the limited liability clauses through insufficient knowledge and thus consent. The creditor may not know even the rough extent of the practical limits on his remedies, which will depend largely on the financial health of the company. It takes 'sufficiently explicit words' before a court would accept that a person would do 'anything so foolish' as limit his liability. 122 While in the prevailing commercial environment the use of the word 'limited' in the company name may be sufficient to take the creditor to have consented to limited liability, that would not have been so in the Victorian period. Even after there was easy access to limited liability following the Limited Liability Act 1855, it was not until Salomon v A Salomon and Co, Ltd in 1896 that it was established as universal, applying to all bodies corporate, rather than only economically large ones. 123 In early-to mid-Victorian times, the use of 'limited' might not always have been enough without more. Even today, it would be a big step to take a creditor to have intended to limit its liability where there was an option not to, especially for a big supplier dealing with a small close held company. The habitual use of secured credit, particularly the floating charge, suggests quite the opposite.
Second, consider the possibility of additional parties joining the nexus via the creditor, for instance, if the creditor assigns or charges the debt. Even if the creditor is bound by the company's limited liability clauses, this obligation, being contractual, is only personal and will not bind third parties as a true proprietary right (or statutory limited liability) would. While no doubt it would be made contractually incumbent upon the creditor to make sure the assignee has notice of the liabilitylimiting terms or consents to them, this simply may not be done. The assignee would then take clear of notice and the shareholders' and directors' liability would not be limited vis-à-vis the assignee of the debt. This is analogous to the case of Scruttons, where the new party was not joined to the nexus for similar reasons. Consequently, this matter is now regulated by legislation. 124 The third is deliberate refusal. Suppose a creditor stipulates that it will not accept limited liability and contracts with an unincorporated company with 'limited' in its name. 125 This 'battle of the forms' places the courts on the horns of a dilemma. With notice of the limiting terms, the courts have two options. One option is for the courts to uphold the party's intention not to be so limited, in which case the corporate form is too weak. The other option-for both the second and third situations-is to uphold limited liability in the teeth of ignorance of the stipulations or positive intention not to be bound, in the interests of efficiency and anti-evasion. However, 122 Maitland (1911), p 392. 123 [1897] AC 22 (HL); Ireland (1996) and Worthington (2001). The very fact that Salomon was litigated all the way to the House of Lords suggests it was generally thought that limited liability did not apply in all circumstances, particularly close held corporations. 124  this would have the devastating consequence that the process simply would not be contractual. There would be mandatory rules overriding consent.
The fourth situation concerns involuntary creditors, who will usually be claimants in tort or unjust enrichment. Such creditors, by definition, do not join the nexus of contracts consensually. Hence, any terms limiting liability and the like within the nexus will not bind such parties. While contractarians may see their number as too small and their position too remote to offer anything more than a limited exception to the theory, we beg to differ. The stevedores in Scruttons were liable in tort and can hardly be said to be a remote party. On the contrary, theirs was a core case.
It seems unlikely that these issues could be reliably resolved in all circumstances. They indicate the limits of relying on consent to create and extend nexuses. When pushed, the consent basis needed for purely private ordering comes up against the paradox of implied terms, where Imperatives 2 and 3 dominate on the facts. The solution has been to use non-consensual doctrines. But Imperative 2 is modified when liability is not rooted in consent. Rather than struggle to imply sufficiently certain terms, the courts can impose non-consensual rules of law without creating legal uncertainty. Non-contractual devices such as agency or the trust certainly do have more flexibility to strengthen entity shielding beyond what is available under ordinary partnership law, and the courts have found these routes preferable. Their strictures mean Imperative 3 falls away. In any event, the utility of consent-based doctrine ceases just when the contractarian paradigm needs it the most. Consequently, the corporation cannot be created by contract alone.

A Non-Consensual Doctrine: Fiduciary Duties
Having seen how the paradox applies to the corporate form somewhat more generally, it is instructive to consider the bearing that the paradox of implied terms has on an additional feature of the corporation, namely fiduciary duties, and more specifically the duty of loyalty. What is significant about fiduciary duties, on the present line of analysis, is that they are at least in part totally non-consensual, meaning that some of their special obligations are imposed in direct opposition to the intentions of the parties. 126 This makes their analysis a somewhat curious application of the paradox. Here, Imperative 2 is strongly in play and Imperative 1 is not engaged at all. Thus, the outcome is to refuse implication even before Imperative 3 steps in to assist Imperative 2.
In one sense, this is not an application of the paradox at all, since we are going beyond difficult cases of implication to impossible ones; applying it to such clear cases runs the risk of obscuring its analytic value, which comes out where there is more nuance. On the other hand, these are doctrines cited in the contractarian literature as evidence of the contractual-and thus consensually constituted-nature of the corporation. 127 It is therefore useful to apply the same critique.
The directors' (and potentially other employees') fiduciary duties, which include putting the company's interests ahead of one's own and acting in good faith, are typically not specified in contracts of employment, so if they were to exist, they would have to arise through implication. 128 With a sufficiently flexible doctrine of implied terms, where the approach to consent leans sufficiently in the direction of Imperative 1, it might be possible for fiduciary duties to arise as contractual terms implied in fact. 129 Certainly, contractual terms can be used to shape the duty of loyalty to some extent. 130 In Ross River Ltd v Waveley Commercial Ltd, it was made clear that the fiduciary Ross River Ltd could have paid itself a profit provided it was compliant with the conditions of the contract. Had it adhered to those conditions, it would not have been a breach of fiduciary duty. 131 But there is one aspect of the duty of loyalty that is non-consensual, namely its 'irreducible core' that cannot be excluded no matter how hard the parties try.
One aspect of this is that one may exclude compensatory liability for negligence, even gross negligence, but one cannot exclude liability for acting in bad faith or dishonestly. 132 Another aspect is that there must be fully informed authorisation before what would otherwise be breach of fiduciary duty might be excused. Consider the quotidian case where a fiduciary takes a commission on management activities. An undisclosed commission is an unlawful secret commission, contrary to the no-profit rule (benefiting from one's fiduciary office without authorisation) of the duty of loyalty. While in a non-fiduciary relation it is possible to stipulate for 'a reasonable commission', in a fiduciary relation the precise details must be disclosed; there must be 'full and frank disclosure of all material facts'. 133 Recent case law on 'half-secret commissions', where the existence of the commission is declared but the amount is not, has affirmed this requirement. It will be a breach unless the amount is immaterial. The most that can be said is that if the rate is a known trade usage, this may be sufficient. 134 But once there is a fiduciary duty, it is impossible to contract out of this requirement of fully informed authorisation.
Hence, fiduciary duties do not always respond to consent, much less the degree of consent that courts are willing to manufacture through the objective method. It is scarcely possible to stretch the doctrine of implied terms to contentious cases of possible objective consent because of the paradox. The propositions of fiduciary law stated and the authorities for them show that it is fully impossible to stretch it even further to cases where facets of the duty of loyalty are imposed positively against the intentions of the parties.
A doctrine of implied terms that could do this, i.e., create the duty of loyalty as the courts see it, would have to be so detached from actual consent so as to allow an implied term to displace an express one in the teeth of the obvious obstacle that express terms are the primary exposition of intent. 135 The fiduciary case law, in accordance with the paradox, shows this is not the law, a proposition amplified by the case of Marks and Spencer plc v BNP Paribas Securities Services Trust Co (Jersey) Ltd reining back the doctrine of implied terms. 136 Fiduciary duties are then imposed in opposition to the will theory and are justified not by consent, nor by efficiency considerations, but by the need to protect the vulnerable in specified relationships or where there is a particular need for it on the peculiar facts of the relationship. 137

Conclusion
The claim that the corporate form cannot be created by contract alone is not new. The distinctive feature of this article is that it adds a new argument to support this position: there is only so much that can be done by contract and the courts. Even under the most flexible form of contract law, the common law contract, courts will run up against the paradox of implied terms before the process of reliably creating the corporate form is complete.
Contractarianism requires that courts fill the gaps in incomplete contracts by implying the missing terms and that they do so according to a wealth-maximisation criterion. We have shown that while the courts want to hold parties to their bargains and imply terms to uphold their consensually agreed liability, uncertainty about the content of the missing terms means that courts will often refuse to imply them, lest they undermine the very thing that justifies liability in the first place-consent. Moreover, we have argued that the constraints imposed by adjectival law, which have to do with the imperatives of maintaining legal certainty and deterring litigation, further drive the courts to refuse to imply terms in fact. When the matter at hand involves adding a new party to a contractual nexus by implying the consent of all existing parties, the paradox bites harder. Rather than attempting the impossible task of resolving the paradox, the courts resorted to non-contractual devices such as agency or trusts, but even these were limited by the selfsame phenomenon.
While courts might have adopted the hypothetical bargain logic as a basis of contractual liability, they have not done so. This abundance of caution is not the result of them taking an ideologically motivated anti-contractarian position. There are no statements to this effect in the judgements, and instead there are a great many statements that point to the imperatives that lead to the paradox. The paradox of implied terms cannot be avoided because its imperatives cannot be eluded, except by admitting non-consensual and mandatory rules. This means accepting corporate law's conventional role as a necessary tool to constitute a corporation. The issues we have discussed are then simply not seen in the statutory corporation in everyday use. We conclude that the strong-form contractarian claim that a body corporate can be created by contract, supplemented by the gap-filling role of courts, is implausible.