Fiduciary Loyalty, Inside and Out

Stephen R. Galoob & Ethan J. Leib[*]

Introduction

A fiduciary is someone with a certain form of discretion, power, or authority over the legal and practical interests of a beneficiary.[1] As a result of this arrangement, the beneficiary is vulnerable to predation by the fiduciary. Fiduciary relationships trigger a suite of duties, at the core of which is the duty of loyalty. In a sense, the fiduciary relationship is oriented around the possibilities of trust and betrayal. One point of fiduciary duties is to prevent betrayal or, failing that, to assure that betrayals are rectified insofar as possible. What constitutes loyalty or betrayal in fiduciary law, however, is not always clear.

Consider Item Software (UK) Ltd. v. Fassihi.[2] Messrs Fassihi and Dehghani were corporate directors of a small software distribution company called Item Software, whose main business was selling software developed by Isograph. Dehghani was the managing director, and Fassihi was the sales marketing director. In November 1998, Dehghani decided to renegotiate the terms on which Item sold Isograph’s products. Fassihi urged Dehghani to drive a hard bargain with Isograph, so Deghani negotiated aggressively. Ultimately, the negotiations between Item and Isograph broke down, and Isograph terminated its contract with Item.

Fassihi’s advice to Dehghani, although plausibly in Item’s best interest, had the air of duplicity. Unbeknownst to Dehghani, during the negotiations Fassihi had approached Isograph with a proposal to establish an independent company to market Isograph’s products. At the same time Fassihi counseled Dehghani to engage in brinksmanship, he also urged Isograph to terminate its relationship with Item. In a subsequent lawsuit, Item alleged that Fassihi only urged Dehghani to negotiate aggressively in order to increase the prospects of undermining the negotiations and subsequently obtaining Isograph’s business for himself.

Was Fassihi disloyal? The answer, of course, depends on what loyalty means. It seems clear that Fassihi was disloyal to his partner in the ordinary sense of that term. Fassihi’s conduct bears a striking resemblance to that of Iago in Shakespeare’s Othello and of Littlefinger in George R.R. Martin’s A Song of Ice and Fire novels, arguably the preeminent historical and contemporary literary examples of treachery. Yet as a legal matter, whether Fassihi violated his fiduciary duty of loyalty is not as obvious. This discrepancy might be explained on the grounds that the notion of loyalty applicable in life (let alone Elizabethan tragedy and genre fiction) differs from the standard that leads to legal liability against fiduciaries like corporate directors, trustees, and attorneys.

Cases like Fassihi implicate a lively scholarly debate concerning how the legal notion of loyalty relates to the notion applicable outside the law. Although the terms of this debate are not always clear, some see a deep connection between the legal and non-legal notions of loyalty. Call this position “moralism.” For the moralist, this connection to the moral or ordinary notion of loyalty informs the legal requirements that apply to fiduciaries, as well as the determination of whether a fiduciary has violated duties to a beneficiary in any particular case. By contrast, a position we can call “amoralism” denies that there is any meaningful connection between the loyalty that applies to fiduciaries and its moral counterpart. To be sure, the amoralist does not (and cannot) deny that judges sometimes invoke moralized language to describe fiduciary concepts. However, for the amoralist, any such connection is rhetorical flourish rather than real law. We more fully describe the parameters of the debate between moralists and amoralists in Part I.

The debate between moralists and amoralists is a species of a much broader dispute about the comparative importance of legal materials and broader normative principles in theorizing the private law. Consider the connections between the morality of promise and the law of contract, or the role the concepts of “wrong” and “duty” play in tort law. Much private law litigation and scholarship concerns whether legal concepts resemble and operationalize concepts from ordinary morality. Within fiduciary law, this debate about loyalty has particularly important practical implications, since it bears on how to elaborate standards in areas where fiduciary norms already apply and on whether fiduciary norms should apply to a particular legal domain in the first place.

The debate between moralists and amoralists is long running and perhaps intractable. We propose to finesse, if not resolve, this impasse by focusing on what we term the cognitive dimension of fiduciary loyalty. On this view, whether someone satisfies the requirements of fiduciary loyalty depends, at least in part, on how she deliberates and how her deliberation is connected with her actions. Fiduciary loyalty also imposes demands on a person’s commitments: a fiduciary does not satisfy her duty of loyalty toward a person or cause if her commitments to that person or cause prove themselves too flimsy. These standards apply to loyalty both inside and outside of law. For the most part, they apply irrespective of whether the best understanding of fiduciary loyalty is moralist or amoralist. We elaborate and defend these claims in Part II, drawing on doctrines in corporate law, trust law, agency law, bankruptcy law, and the law governing lawyers that are, we argue, best explained by the cognitive dimension of fiduciary loyalty.

Part III then clarifies our conclusions regarding cognitivism and fiduciary loyalty, highlighting some of the implications of our analysis for fiduciary law. Cognitivist accounts can catalyze both doctrinal and policy innovations. Regardless of whether loyalty has an identical meaning inside and outside of legal institutions, fiduciary loyalty, like ordinary loyalty, imposes important standards on a fiduciary’s cognition. Appreciating this structural feature of loyalty will enable judges and scholars to transcend many debates about moralism and to resolve practical questions that do not turn on whether moralism is true.

I.  Loyalty In Law

What is the best way to understand the fiduciary duty of loyalty? Is the loyalty demanded of fiduciaries identical to or deeply connected with the notion of loyalty applicable in the real world? Or is fiduciary loyalty a purely juridical concept, one with no important connection to the concept as it applies outside the law? Sections I.A and I.B describe an ongoing debate between two positions that can (somewhat misleadingly) be called “moralism” and “amoralism.” Section I.C explains why breaking the impasse between moralism and amoralism is especially difficult and identifies some practical consequences of this stalemate.

A.  “Moralism” About Fiduciary Loyalty

The moralist position is that fiduciary loyalty references the notion of loyalty that applies outside of legal institutions.[3] A moralist view need not contend that the legal and non-legal conceptions of loyalty are identical, only that they have some deep connection. Moralism appears in debates about the nature of fiduciary law,[4] as well as in debates about substantive areas of law that are oriented around fiduciary duties.[5] We follow the convention in describing this position as moralist, although the term itself is something of a misnomer.[6] A more accurate description would be that the legal notion of loyalty is substantially connected to the non-institutional (that is, the “ordinary” or “genuine”)[7] notion of loyalty.

Moralism raises a number of important questions about the nature of loyalty. A moralist need not resolve all of these general questions in order to analyze fiduciary loyalty. Nor need the moralist accept that every aspect of “ordinary” or “genuine” morality is automatically incorporated into fiduciary loyalty. Variants of moralism might well differ over which dimensions of ordinary morality are implicated in fiduciary law.

Moralism has a long history in fiduciary law. In the United States, the most prominent example of, and citation for, moralism is Justice Cardozo’s opinion in Meinhard v. Salmon.[8] The court there affirmed a lower court’s holding that Salmon breached a fiduciary duty to Meinhard, his co-participant in a joint venture for managing and leasing a building. Salmon failed to disclose to Meinhard a re-leasing opportunity that rightfully belonged to the joint venture. In finding that Salmon had breached his fiduciary duty, Justice Cardozo wrote that

[j]oint adventurers, like copartners, owe to one another, while the enterprise continues, the duty of the finest loyalty. Many forms of conduct permissible in a workaday world for those acting at arm’s length, are forbidden to those bound by fiduciary ties. A trustee is held to something stricter than the morals of the market place. Not honesty alone, but the punctilio of an honor the most sensitive, is then the standard of behavior. As to this there has developed a tradition that is unbending and inveterate. Uncompromising rigidity has been the attitude of courts of equity when petitioned to undermine the rule of undivided loyalty by the “disintegrating erosion” of particular exceptions. Only thus has the level of conduct for fiduciaries been kept at a level higher than that trodden by the crowd.[9]

The language of Meinhard is the high-water mark of moralism about fiduciary loyalty. The case contemplates a standard for fiduciaries that invokes (and may exceed) the requirements of moral loyalty and honor. To be sure, the standard that Cardozo’s rhetoric describes was not and is not the general legal definition of fiduciary loyalty, nor has it been incorporated into partnership law in particular.[10] However, a moralist might contend that the rhetoric of Meinhard and similar cases indicates that fiduciary loyalty tracks ordinary or genuine loyalty.

Moralism is also implicated, either expressly or implicitly, in many other judicial opinions in a variety of jurisdictions and legal contexts. Courts sometimes resemble Meinhard in embracing moralism explicitly.[11] Other courts embrace moralism implicitly by invoking moralized concepts like “utmost good faith” and “fidelity”[12] or appealing to commonsense moral precepts[13] to resolve questions of fiduciary law.

Many academic commentators deploy moralism as a descriptive or interpretive claim, a way of explaining how moral precepts matter for understanding what fiduciary law is. As part of this interpretive task, ordinary moral notions might be held to illuminate a number of features of fiduciary law. For example, a moralist might see moral precepts as elucidating the point of fiduciary duties—that is, why non-fiduciary mechanisms of accountability are insufficient for fiduciary relationships, and what special features explain the grounds of fiduciary duties. Ordinary moral precepts might also be thought to illuminate the content of fiduciary duties, not only explaining widely recognized dimensions of fiduciary duties (such as the prohibitions on a fiduciary’s conflicts of interest or profiting from the relationship), but also explaining the standards of conduct that apply to fiduciaries. A moralist could also invoke ordinary moral notions in order to explain the standards of liability that apply to fiduciaries, including the criteria for determining when fiduciary duties have been violated. Furthermore, moralism might be thought to explain the remedies for violating fiduciary duties, including the determination of which remedies are justified for fiduciary breaches as such and which remedies are appropriate in any particular case. To be sure, a moralist need not contend that ordinary morality supplies answers to all of these questions or answers any of them completely. Rather, if moralism is correct, then legal actors can (or should be able to) appeal directly to at least some aspects of ordinary morality in debates about at least some legal questions related to fiduciary law.

However, moralism may extend beyond descriptive and interpretive claims into purely normative terrain. Moralists have offered a variety of justifications for the incorporation of ordinary moral precepts into fiduciary law. For example, Tamar Frankel sees moralism as “exert[ing] pressure on the fiduciary to fulfill his obligations once he has agreed to enter into the [fiduciary] relation” and “elevating the purpose for which the fiduciary’s power is granted to a position of priority over other values which may guide the fiduciary.”[14] Another prominent justification for moralism is more instrumental: fiduciary law’s incorporation of moral notions might be defended as promoting better behavior by fiduciaries.[15] Or, perhaps, morality supplies the grounding or justification for the juridical relational duties within fiduciary law, exerting a gravitational pull on fiduciary law in marginal cases even if most of the development of fiduciary law is mainly driven by institutional, rather than moral, considerations.

Moralism, then, can be framed as both a descriptive and a normative thesis. The descriptive thesis is that moral (or other non-institutional) considerations are deeply connected to understanding what fiduciary law is, and especially to understanding fiduciary loyalty. The normative thesis is that there are good reasons for this arrangement.

B.  “Amoralism” About Fiduciary Loyalty

Many judges and commentators on fiduciary law reject moralism. The common element of such amoralist views is a denial that fiduciary law makes reference to ordinary moral precepts or tracks in any deep way concepts from moral life. In particular, the amoralist contends that the ordinary understanding of loyalty cannot explain fiduciary relationships or the content of fiduciary duties in more than a superficial way. As Frank Easterbrook and Daniel Fischel put it, “[f]iduciary duties are not special duties; they have no moral footing . . . .”[16] The amoralist might also deny that ordinary moral notions are relevant to important legal questions, for example when a fiduciary duty has been violated or what remedies should be available for such a violation. The amoralist would contend that the standards applicable to fiduciaries in, say, Delaware Chancery Court or a bar disciplinary hearing can differ wildly from (and contradict) the standards used in the evaluation of soldiers, friends, or literary characters.

Judges and commentators have espoused many varieties of amoralism. Some commentators deny moralism on empirical grounds. This empiricist position bases its conclusions on evidence that courts eschew or reject moralized notions in actually deciding fiduciary law cases. For example, the empiricist might point to Lord Herschell’s opinion in Bray v. Ford, which explicitly rejects a moralized understanding of fiduciary loyalty:

It is an inflexible rule of the Court of Equity that a person in a fiduciary position . . . is not, unless otherwise expressly provided, entitled to make a profit; he is not allowed to put himself in a position where his interest and duty conflict. It does not appear to me that this rule is, as has been said, founded upon principles of morality. I regard it rather as based on the consideration that, human nature being what it is, there is danger, in such circumstances, of the person holding a fiduciary position being swayed by interest rather than by duty, and thus prejudicing those whom he was bound to protect. It has, therefore, been deemed expedient to lay down this positive rule.[17]

This empirical strategy might deploy an inductive methodology of legal theorizing.[18] Furthermore, because decisions in fiduciary law routinely utilize moralized language and invoke “ordinary” understandings of loyalty, the empiricist needs to provide an explanation for why this language does not capture anything essential about fiduciary law. Perhaps, an empiricist might argue, the moralized rhetoric of fiduciary law opinions should be subordinated to assessments of how judges actually resolve particular cases, rather than the language in which they announce their resolutions.[19]

Few, if any, argue for amoralism solely on empirical grounds. Rather, most amoralists offer principled justifications for an amoralist understanding of fiduciary loyalty. One such principled version of amoralism is “pluralism.”[20] Many commentators have noted the wide variation in how fiduciary duties are formulated across legal contexts.[21] For example, the standards for determining whether an attorney has lived up to her fiduciary duties are different from the standards applicable to trustees or corporate directors.[22] Some pluralist approaches extend this insight to deny that any non-trivial propositions about fiduciary loyalty generalize across fiduciary contexts.[23] Others are willing to specify a core of fiduciary loyalty, but insist that this core must be very thin in order to apply across types of fiduciary relationships.[24]

Pluralist positions about fiduciary loyalty often import a dose of empiricism. Gold, for example, argues that the non-generalizability of fiduciary norms turns on the empirical proposition that at least some courts in some fiduciary contexts reject a moralized understanding of loyalty. Gold focuses on the case of Jordan v. Duff & Phelps,[25] in which a majority opinion (written by Judge Frank Easterbrook) deployed a hypothetical bargaining methodology to conclude that fiduciary duties were owed to an employee of a closely held corporation who owned shares in the firm. The Court found that these duties were violated when the corporation failed to disclose information about an upcoming merger prior to the employee’s resigning to take a position at another firm and selling back his shares. A dissenting opinion by Judge Richard Posner deployed the same methodology to reach the opposite result.[26] Gold argues that, while moral precepts might play some explanatory or justificatory role in some legal contexts, other jurisdictions and areas of fiduciary law do not invoke moralized explanations of fiduciary concepts.[27] Moreover, pluralism suggests that insights from one fiduciary context have no intrinsic force across contexts (for example, to different domains of fiduciary law within the same jurisdiction, as well as to the same domain of fiduciary law across jurisdictions). Pluralism might also rule out criticizing particular areas of fiduciary law jurisprudence (for example, Delaware corporate law) on the grounds that it deviates from the “essence” or “core” of fiduciary law. An extreme version of pluralism holds that that there is no essence or core of fiduciary loyalty.[28]

Another amoralist approach is an argument from incompatibilism. The incompatibilist contends that certain aspects of “moral” or “ordinary” loyalty render these ideas inapplicable to fiduciary law and prevent them from being legitimately imposed by legal institutions.[29] In light of these aberrant elements, the legal definition of fiduciary loyalty does not and should not incorporate the moral definition of loyalty. Incompatibilism is rooted in both a specific understanding of the purposes of fiduciary law and a liberal account of the limits of state power. For the incompatibilist, many aspects of “ordinary” loyalty are irrelevant to the law and, in any case, inappropriate to enforce. For example, an incompatibilist might contend that “ordinary” loyalty has an affective dimension: someone is only loyal to another person or a cause if she is disposed to have appropriate emotional responses based on how the object of her loyalty fares.[30] However, legal obligations (including those imposed as part of fiduciary law) do not and should not turn on considerations related to the manifestation of emotion or affect. Determining whether someone has lived up to a legal duty should be a question of how he acted, not what he felt. Indeed, it would be illegitimate for legal institutions to impose duties on people to feel certain ways.[31] How someone feels while discharging his fiduciary responsibilities is irrelevant to the most important normative question of fiduciary law, namely whether he actually abused his power while doing so. Thus, fiduciary loyalty diverges from “ordinary” loyalty in that the affective dimension that ex hypothesi characterizes the latter is inapplicable to the former. The incompatibilist argues for amoralism on the grounds that the moral notion of loyalty does not fit well within legal institutions generally, and within the various domains of fiduciary law specifically.

Another strain of amoralism, which we have previously called “proscriptivism,” asserts that fiduciary loyalty can be fully described in terms of the prohibitions (in particular, the no-profit and no-conflict rules) that apply to fiduciaries.[32] For the proscriptivist, fiduciary obligations do not require or prescribe any particular way of acting or deliberating. So long as the fiduciary avoids behaving in ways that violate these proscriptions, she has lived up to her fiduciary duties. Proscriptivism might be defended on empirical grounds (for example, that courts sometimes formulate fiduciary duties entirely in terms of proscriptions),[33] on normative grounds,[34] or on both.[35] The proscriptivist typically draws a clear distinction between fiduciary duties (which are solely a function of the fiduciary relationship) and “nominate” duties (which apply to both fiduciary and non-fiduciary relationships).[36] Although these nominate responsibilities might apply to the fiduciary in virtue of her fiduciary relationship, they are not core aspects of fiduciary loyalty because they do not apply uniquely to fiduciary relationships.

Several principled arguments might be offered on behalf of the proscriptivist position. For example, proscriptivism’s narrow reading of fiduciary loyalty ensures non-interference by the state, where interference might undermine the development of mutual trust between the fiduciary and beneficiary. Another argument for proscriptivism could be based on effectiveness: imposing narrow constraints on the fiduciary’s actions empowers her to exercise her discretionary power on the beneficiary’s behalf and avoid risk-aversion that would be counterproductive to advancing the interests of the beneficiary. Proscriptivism might also be justified non-instrumentally, on grounds of autonomy: limiting fiduciary duties to the no-profit and no-conflict duties prevents the state from violating the sanctity of an ongoing fiduciary relationship.

Proscriptivism’s narrow notion of fiduciary loyalty evokes Oliver Wendell Holmes, Jr.’s figure of the “bad man.”[37] Holmes’s bad man is rational, self-interested, and cynical. He is only motivated to conform to the law in order to avoid negative consequences that might arise from non-conformity.[38] As such, the Holmesian bad man sees potential legal sanctions as the most important aspect of law; only when the costs of non-conformity with a legal directive are higher than those associated with conformity can the bad man be expected to conform. Many theorists contend that a primary goal of law is to speak to the bad man.[39] On this approach, the clarity of a legal standard and the certainty with which it will be enforced are important questions because they would matter to the bad man’s calculations. Fiduciary law, with its moralistic rhetoric and open-ended responsibilities, might seem a difficult language in which to communicate with the bad man. However, this problem is avoided if the vagueness of fiduciary jurisprudence can be formulated more precisely. The proscriptivist offers exactly this precisification: the bad man can know exactly what behavior is expected of him (for example, “do not engage in a conflicted transaction” and “do not derive inappropriate profits from your work as a fiduciary”) and what will happen if his behavior deviates from this expectation. Proscriptivism, then, is fiduciary loyalty for the bad man.

A final type of argument for amoralism, prominent among legal economists, is “contractarianism.”[40] Contractarians contend that fiduciary loyalty should be understood as a species of contractual obligations, since both types of obligations typically arise from voluntary interactions. The version of contractarianism most relevant to amoralism sees fiduciary duties in terms of hypothetical bargaining: the standards that should apply to fiduciaries are those that would arise as part of an ideal negotiation between the fiduciary and the beneficiary.[41] Many in the law-and-economics tradition who advocate for contractarianism also contend that behavior is the fundamental unit of analysis relevant to legal norms[42] and that contractual performance is ultimately a function of behavior.[43] Therefore, many contractarians also embrace the idea that fiduciary obligations ultimately concern a fiduciary’s behavior: the duty of loyalty requires a fiduciary either to behave in specific ways or else to pay damages.[44]

As noted above, the contractarian position might be defended on empirical grounds, as some courts rationalize their fiduciary law decisions in contractarian language.[45] However, the most powerful normative argument for the contractarian approach is based on efficiency. For the contractarian, fiduciary duties arise as a response to legal relationships in which the agent is vested with discretionary power to act on behalf of the principal. This arrangement creates high agency costs and raises the prospect that (1) the agent might exercise power to benefit the agent and not the principal; and (2) alternative mechanisms of monitoring the agent’s actions will be unavailable.[46] In these contexts, fiduciary duties deter the principal from misconduct and opportunism. On the contractarian paradigm, the fiduciary duty of loyalty (which prohibits the agent’s opportunistic behavior and requires action in the interest of the principal) also has a disclosure effect: it “induce[s] the fiduciary to avoid . . . conflict[s of interest] or to disclose the material facts of how [such conflicts] might compromise the fiduciary’s judgment.”[47] The fiduciary duty of care, by contrast, applies a loose but objective standard for assessing the agent’s work on behalf of the principal.[48] The duties of loyalty and care together provide broad standards for the functioning of an efficient fiduciary relationship. Gaps in the application of the standards can be filled in by subsidiary doctrinal rules[49] and, if necessary, judicial interpretation, which should be oriented around the question of what the fiduciary and beneficiary would have agreed to ex ante as part of a hypothetical bargaining process.[50]

The contractarian position, then, provides a unified explanation of why fiduciary duties matter (namely, to reduce agency costs by deterring agential misconduct, while enabling agents to exercise discretionary authority), the content of fiduciary duties of loyalty and care, and the character of remedies for breach of fiduciary duties (which are broad and prophylactic in order to effectively deter misconduct in the wake of monitoring problems). The position is amoralist because the moral or “ordinary” notion of loyalty plays no meaningful role in any part in this contractarian story.

The above list does not exhaust the possible versions of amoralism. Nor are these types of amoralism mutually exclusive. For example, some commentators combine the contractarian and proscriptivist positions.[51] The unifying theme is that some amoralists would deny that moral precepts are generally or uniformly relevant to understanding fiduciary loyalty, while others would deny that they should be relevant, and still others would deny both propositions.

C.  Breaking the Impasse?

How does and should fiduciary law incorporate the “moral” or “ordinary” understanding of loyalty? Is fiduciary loyalty merely a technical term, applicable only within legal institutions? The divide between moralists and amoralists regarding fiduciary loyalty resembles similar divides in other areas of law over the connection between moral and legal concepts. For example, scholars of tort and criminal law debate whether the notion of causation at issue in law is identical to the metaphysical notion of causation.[52] Similarly, debates about the interpretation of the Eighth Amendment’s prohibition on “cruel and unusual” punishment turn on whether “cruelty” should be understood commonsensically or, alternatively, in a more technical way.[53] Perhaps the closest cousin to the debate between moralists and amoralists occurs in contract law, where some scholars contend that there is a deep connection between contract law and promissory morality[54] and others see no necessary connection.[55] Thus, the debate between moralists and amoralists about fiduciary law might be understood as a species of a broader genus of discussions about whether legal concepts with apparent referents in moral life have a specialized institutional meaning.[56]

As such, the debate between moralists and amoralists implicates deeper questions about legal theorizing. One barrier to resolving this debate about fiduciary loyalty is that, in many respects, the moralist and amoralist positions are so capacious that they can reach the same conclusions in most cases. It is also difficult to provide an empirical resolution to the debate, since moralists and amoralists might disagree about the relevant unit of empirical assessment. To be sure, many courts analyzing fiduciary duties employ moralist language. However, some courts reject it outright, and others deploy moralized language differently across domains of fiduciary law. It is possible that a thorough-going moralism applies in some pockets of fiduciary law in some relational contexts but not trans-substantively. Moreover, the moralized rhetoric of fiduciary law opinions strikes some amoralists as mere rhetoric, and therefore, an unreliable predictor of what courts and relevant legal actors will do in actual cases.[57]

The empirical disagreements between moralists and amoralists in turn implicate broader jurisprudential disagreements. Resolving the debate between moralists and amoralists might be simple with a relatively well-developed theory about what law is and how past cases figure into the existence of legal standards. Yet such theories of law are notoriously difficult to formulate and defend.

The impasse between moralists and amoralists also turns on a variety of contested substantive and normative issues. For example, moralists and amoralists might disagree about the fundamental point of fiduciary relationships. There is no uniformity on these issues, even among proponents of amoralist views. Although nearly all who study fiduciary law worry about predation and opportunism, these concerns do not resolve most core differences between moralism and amoralism. Breaking the impasse about the proper understanding of fiduciary loyalty would not only require overcoming the empirical, jurisprudential, and substantive disagreements between moralists and amoralists, but also transcending these same disagreements among different versions of moralism and amoralism.

Our goal in what follows, then, is not to resolve the debate between moralists and amoralists. Instead, in Part II, we expose a cognitive dimension of fiduciary loyalty that we argue is part of positive fiduciary law. Therefore, it can be appreciated from a variety of amoralist perspectives. And this cognitive dimension can provide common ground from which moralist and amoralist theories can debate substantive questions of fiduciary law, since most moralists should also be able to see the roots of a cognitive dimension of loyalty in ordinary morality.

II.  Cognitivism About Loyalty[58]

Our main goal in this Part is to demonstrate that there is a cognitive dimension to fiduciary loyalty. A cognitivist account of fiduciary loyalty posits that a fiduciary’s cognition bears on whether she satisfies her fiduciary duties. By contrast, non-cognitivist accounts (for example, the existing versions of proscriptivism and contractarianism that we discussed in Part I) interpret fiduciary loyalty entirely in terms of how a fiduciary behaves or, perhaps, a combination of the fiduciary’s behavior and the results of its action or inaction. Cognitivism is neither moralist nor amoralist per se. Rather, it provides a way to traverse the impasse between these positions.

Our case for cognitivism about fiduciary loyalty is largely indirect. In Section II.B, we identify several propositions that generalize across a wide range of fiduciary law contexts. Each of these propositions is consistent with a cognitivist account of fiduciary loyalty. Further, cognitivism provides a unified and powerful way to explain these propositions. If our reasoning is sound, then our conclusions regarding fiduciary loyalty’s cognitive dimension can be accepted by any moralist or amoralist who aims to describe the law as it is. Section II.A briefly explores the cognitive dimension of “ordinary” loyalty. Section II.B shows how a cognitivist account of loyalty explains fiduciary law as well.

A.  Cognitivism About “Ordinary” Loyalty

What is the nature of “ordinary” loyalty? What makes a person loyal? What follows from identifying a person as loyal? Centuries of reflection on these questions have not produced a consensus. Nearly every candidate for the essential consideration constitutive of loyalty has been disputed.[59] Moreover, the very moral significance of loyalty is also sharply contested among philosophers.[60]

We cannot here resolve these longstanding philosophical debates. However, some requirements relating to the cognitive dimension of loyalty appear to generalize across accounts of “ordinary” loyalty. In other words, the conclusion that a person is loyal or acted loyally is, in part, based on an assessment of her cognition. Satisfying the cognitive aspects of loyalty vis-à-vis the object of one’s loyalty is necessary, but not sufficient, to be loyal to that object. Although this cognitive dimension might not appear explicitly in accounts of loyalty that focus on affect, perceptual bias, habit, or unthinking partiality, the best candidates for a theory of loyalty (rather than, say, devotion or allegiance or fealty) have a cognitive dimension. In previous work,[61] we have identified three different cognitive aspects of loyalty: deliberation, conscientiousness, and robustness. Rather than reiterate our arguments here, a trio of examples might better illustrate them.

Consider Shakespeare’s Iago, the quintessential literary example of treachery. Iago is a longtime adviser to Othello. Jealous at Othello’s promotion of Cassio,[62] another soldier in the company, to the rank of lieutenant, Iago plots to bring about Othello’s downfall. Iago fabricates evidence and manipulates conditions in order to lead Othello to believe that Cassio is having an affair with Desdemona, Othello’s wife. At the same time, Iago purports to provide Othello and Cassio unvarnished advice about what to do, some of which is even credible. As Iago puts it, he leads each man down a “parallel course” that is “directly to his good.”[63] Iago’s handiwork is so skillful that Othello seems to convince himself of each step down the road of murdering Desdemona.[64]

One insight from Iago’s perfidy is that “ordinary” loyalty imposes requirements on deliberation. If your deliberation regarding the object of your loyalty does not satisfy these requirements, then you are not loyal. At least two aspects of Iago’s disloyalty bear mentioning. First, although Iago’s behavior toward Othello (lying, manipulating evidence) is objectionable in its own right, Iago would not be so reviled if he had merely given bad advice or misled his captain. There is something about the way that Iago uses Desdemona, Cassio, and Othello that strikes many readers as particularly objectionable. Second, Iago’s disloyalty is evident long before he gives his fateful advice to Othello—indeed, long before Iago announces his treacherous orientation in an opening soliloquy.

The cognitive dimension of loyalty explains both of these aspects of Iago’s treachery. Independently of the quality of his advice to Othello, Iago’s deliberation was defective because his plan to bring about Othello’s downfall prioritized his own lust for revenge and disregarded the interests of Othello, to whom he owed duties of loyalty. This conclusion implies that loyalty imposes requirements on how someone deliberates. It also suggests that these requirements can be violated regardless of how someone actually behaves. Furthermore, Iago betrayed Othello by formulating his treacherous plan. Counseling Desdemona’s murder and engineering Cassio’s imprisonment were downstream effects of Iago’s prior betrayal.

A second insight is that “ordinary” loyalty requires conscientiousness—that is, a certain type of connection between deliberation and action—regarding the putative object of loyalty.[65] You are not loyal to a person or idea if you lack this conscientiousness toward them. Consider the saboteur, another paradigmatic example of disloyalty. The saboteur feigns allegiance to an ostensible object of loyalty while simultaneously working to undermine it. A saboteur might have another object of loyalty (in which case she is a double agent), or she might simply work to frustrate the success of the ostensible object. The saboteur has an interest in behaving and deliberating in ways that mimic loyalty, as did Iago. Doing so might help the saboteur or the Machiavellian lieutenant evade detection and provide greater opportunities for subversion. If the saboteur is disloyal even though she behaves and deliberates exactly as a loyal person would, then loyalty must implicate more than loyal behavior and deliberation. This additional element is conscientiousness: given certain motivational profiles, someone can act disloyally despite deliberating and behaving exactly as a loyal person would have done.

Third, loyalty is a function of the robustness of one’s commitments to the object of loyalty.[66] If you are committed to a particular object, but your commitment is too flimsy, then you are not loyal to that object. Consider Petyr Baelish, nicknamed “Littlefinger,” from George R.R. Martin’s A Song of Ice and Fire novels and the television show Game of Thrones. Littlefinger is an ambitious nobleman from an insignificant family. In order to transcend his station, he employs techniques such as ingratiating himself with members of more powerful houses, catering to other nobles in a set of brothels that he owns, and forming secret alliances with still other nobles. Many of these efforts sow civic unrest and lead to the deaths of innocents (and, to be fair, some not-so-innocents).[67] However, as the character notes in an Iago-esque soliloquy, all are in the service of acquiring power:

Chaos isn’t a pit. Chaos is a ladder. Many who try to climb it fail and never get to try again. The fall breaks them. And some are given a chance to climb, but they refuse. They cling to the realm, or the gods, or love—illusions. Only the ladder is real. The climb is all there is.[68]

As an example of his climb, Littlefinger facilitates an alliance between the warring Tyrell and Lannister families that would be secured by intermarriage between members of both families. Later, he plots with the matriarch of the Tyrell family to arrange for the murder of Joffrey Lannister before the consummation of his marriage to Margaery Tyrell, one of the very marriages that Littlefinger brokered.

What makes Littlefinger such a powerful example of disloyalty? Although like Iago he is fueled in part by jealousy, few, if any, of Littlefinger’s plans involve the type of sustained fixation on and malice toward a specific other that characterize Shakespeare’s villain. Rather, Littlefinger is distinguished by his contingency: regardless of how committed he is to a person or cause now, he will abandon them without compunction in order to pursue a future, aggrandizing option. A strategically contingent commitment is not genuine loyalty. If the flimsiness of Littlefinger’s commitment is evidence of disloyalty, then it follows that robustness is a hallmark of “ordinary” loyalty. A loyal commitment must be capable of withstanding at least some changes in circumstances.

B.  Cognitivism in Fiduciary Law

Even if there is a cognitive dimension to ordinary loyalty, it does not follow that this dimension also applies to fiduciary loyalty. Although a number of fiduciary law concepts or doctrines (for example, “good faith,” “fidelity,” the “duty to disclose,” and even certain ways of construing the “duty of care”) seem to implicate cognitive considerations, these concepts and doctrines are not explicitly a part of every jurisdiction’s fiduciary law. Nor do they necessarily apply to every domain of fiduciary law within any particular jurisdiction. Furthermore, an amoralist might well contend (following Easterbrook and Fischel) that these doctrines are little more than window dressing.

In the remainder of this Part, we provide an indirect argument for a cognitive dimension of fiduciary loyalty. We first identify a series of propositions that describe fiduciary loyalty in a variety of legal contexts. A cognitivist account of fiduciary loyalty provides a straightforward explanation of each proposition and a unified explanation of all of the propositions together. To be sure, some of these propositions might also be explained by non-cognitivist accounts of fiduciary loyalty, especially if they are revised to account for the propositions of fiduciary law that we identify below. The explanation provided by non-cognitivist accounts is, by comparison, less complete, less parsimonious, and less persuasive.

Our analysis here, if correct, does not definitively resolve the debate between moralists and amoralists. However, it reduces the scope of disagreement between these positions. Varieties of both amoralism and moralism that aim to explain existing legal materials should accept that fiduciary loyalty has a cognitive dimension. So-called “pluralist” accounts should also embrace cognitivism to the extent that fiduciary loyalty has a cognitive dimension across jurisdictions and fiduciary contexts. Finally, our case for the cognitivist dimension of fiduciary loyalty should allay many of the most significant concerns of incompatibilists.

In what follows, we show how the law routinely acknowledges the relevance of cognitive dimensions to fiduciary obligation. Along the way, we demonstrate the explanatory power of a cognitivist account by contrasting it with two prominent versions of amoralism described in Part I—proscriptivism and contractarianism—that cannot satisfactorily explain all of the legal propositions we identify. These versions of amoralism are the most popular among jurists and commentators, and proponents of these views tend to formulate fiduciary loyalty exclusively in terms of behavior.

P1. Someone can breach her obligation of fiduciary loyalty even if the results of her actions are consistent with those of a loyal course of action and she behaves in exactly the way that a loyal fiduciary would have behaved.

Whether a fiduciary has lived up to her duty of loyalty depends on more than the results of her actions. A fiduciary’s actions can violate her fiduciary duty of loyalty even though the beneficiary is not harmed by them—that is, even when the beneficiary’s position is not made worse off by the fiduciary’s action. Likewise, a fiduciary can violate her duty of loyalty even though she behaves in exactly the way that a loyal fiduciary would have behaved.

Some might doubt that P1 actually describes U.S. law. For example, in many U.S. jurisdictions and the Restatement (Second) of Torts, the tort of breach of fiduciary duty imposes a harm requirement: the fiduciary is only liable to the beneficiary for “harm resulting from a breach of [fiduciary] duty . . . .”[69] However, it does not follow that loss causation is part of the fiduciary duty itself, such that a fiduciary whose treachery is harmless has lived up to the duty of loyalty. Rather, harm is best seen as an independent, redress-related inquiry that arises in specific legal domains. In other words, the question of harm might be relevant to whether the fiduciary is liable in tort or for professional malpractice, but it is not integral to the notion of fiduciary loyalty itself.[70]

A fiduciary’s harmless disloyalty violates her fiduciary duty. Put differently, harmless disloyalty can have the same legal significance as harmful disloyalty. This principle of fiduciary law applies at least within the law governing lawyers,[71] agency law,[72] and trust law.[73] To illustrate, consider an example from the law governing lawyers. Elliot Friedman was an attorney acting for a group of plaintiffs in a tort suit that sought compensation for victims of a sunken vessel near Kodiak Island, Alaska.[74] Friedman put $81,000 of settlement money from one defendant into a trust account for his plaintiff-client while awaiting a payment of $662,400 from other defendants. Although Friedman had authority to put that first contribution into his client’s trust account, the rest of the plaintiffs could not agree how the remaining contributions should be distributed within the plaintiff group. Indeed, it took approximately a year to come to terms about the distribution of the settlement monies and to collect the whole sum. Four days after the first contribution was made, and without permission from the other plaintiffs or their attorneys, however, Friedman wrote himself a check for $15,000 to cover some of his fees and outlaid costs. Subsequently, he wrote a check for $10,000 to his client and paid himself another $3,000 in fees out of the trust account.

In disciplining Friedman, the hearing committee “noted that Friedman correctly believed [his client] would be entitled to substantially more than the amount he had advanced to it and that [Friedman’s] total fees . . . exceeded the fees he paid himself” out of the original $81,000 deposit.[75] The Alaska Supreme Court noted that “no client of Friedman’s suffered actual injury [and] [t]he potential for financial injury . . . was minimal, because Friedman had sufficient personal funds to cover any demand for money any client or [plaintiff] might have made.”[76] Yet even though the results for Friedman’s clients were the same as if he had not misappropriated funds from them, the Friedman court found that Friedman had breached his fiduciary obligations to his clients.[77] There is nothing remarkable about this conclusion. Friedman took money from a trust account too soon, and this action was no less a breach of trust because it (fortuitously) did not harm his clients.

The fiduciary’s course of action can also violate fiduciary loyalty even when it is identical to the behavior that a loyal fiduciary would have undertaken. Consider the case of In re Nine Systems Corp. Shareholders Litigation, in which the Delaware Chancery Court was asked to evaluate a board’s discharge of its fiduciary duties in connection with the recapitalization of a media start-up company.[78] The 2002 recapitalization—which valued the company at $4 million—enabled controlling shareholders to increase their equity in the company, diluting the plaintiffs’ ownership interest from 26% of the company to 2% of it. When the company was sold in 2006 for $175 million, plaintiffs sought $130 million in damages on the grounds that, because the 2002 recapitalization was unfair, the board’s approval of this transaction constituted a breach of fiduciary duties.

The Nine Systems court held that, although the price paid during the 2002 recapitalization was fair, the directors still breached their fiduciary duties because they did not follow a fair process. According to the court, the board and independent sources were insufficiently involved in the valuation, which was done as a “back of the envelope” calculation with “handwritten scribbles.”[79] The plaintiffs were not fully aware of the recapitalization until after it was implemented. As such, they were unable to participate in the valuation process. The only director who sought to dissent from the recapitalization (who was also the sole independent director) was deliberately excluded from deliberations: the controlling group chose to hold meetings on the Sabbath or other Jewish holidays when they knew he could not attend because of his religious commitments.[80]

The court found that the value of the company in 2002 was actually close to $0.[81] Thus, the 2006 transaction was even more generous to plaintiffs than a loyal fiduciary would have provided to them. However, the directors were still found to have breached their fiduciary duties by engaging in an inadequate procedure.[82] An animating idea in Vice Chancellor Noble’s opinion in Nine Systems is that behavior otherwise identical to that of a careful and loyal fiduciary can be deficient due to defects in “process.”[83]

A cognitivist account of fiduciary loyalty can explain why both aspects of P1 make sense. Harmless treachery violates fiduciary loyalty, not merely because the beneficiary is made worse off by it (as in Friedman) or because a loyal fiduciary would have behaved differently (as in Nine Systems). Rather, some aspects of fiduciary loyalty cannot be reduced to results or behavior. Cognitivism posits an irreducibly cognitive aspect of fiduciary loyalty: fiduciary loyalty, like ordinary loyalty, imposes standards on how the fiduciary should deliberate about what to do. These deliberative requirements are freestanding. Failure to satisfy them is sufficient to constitute a violation of fiduciary loyalty. The cognitivist can thus provide a unified explanation why Friedman violated his fiduciary duty to his clients and why the Nine Systems board violated its fiduciary duties. Both Friedman’s behavior and deliberation were defective, even though the results of his behavior did not harm his clients. The Nine Systems board’s deliberation was defective, even though the results of their actions did not harm the corporation and a loyal board might have reached exactly the same decision.

Proscriptivism cannot explain both aspects of P1. To be sure, a proscriptivist can straightforwardly account for the possibility of a harmless breach of fiduciary duties by invoking the prophylactic nature of fiduciary duties.[84] On proscriptivism, Friedman’s performance was deficient (even if his clients were not harmed) because he violated the no-conflict and no-profit rules. However, the proscriptivist cannot easily explain how fiduciary loyalty could be violated when a fiduciary behaves in exactly the way that a loyal fiduciary would. The Nine Systems board had no conflict of interest and did not extract any direct profit from the 2002 recapitalization. Nevertheless, it violated its duty of loyalty because its deliberative process was defective. The proscriptivist cannot explain a violation that does not involve a transgression of either the no-profit or no-conflict rules. Even if a fiduciary’s decisional process could be reframed as a kind of conduct, proscriptivism could not capture the significance of a deliberative command as cognitivist accounts can.

The contractarian position also has difficulty explaining both aspects of P1. For the contractarian, gaps in the fiduciary relationship should be filled ex post by inquiring what the parties would have bargained for ex ante. This hypothetical bargaining methodology might explain why a fiduciary might violate the duty of loyalty even when she behaves in the way that a loyal fiduciary would. In Nine Systems, for example, the board’s decision procedure regarding the 2002 recapitalization might have been judged unnecessarily risky, since it imperiled the interests of minority shareholders in a way that increased the possibility that the value of their shares would be diminished in some future transaction. This exposure to risk could be appreciated from the contractarian’s ex ante perspective.[85] However, the contractarian cannot easily explain the possibility of a harmless violation of fiduciary loyalty in general, or the Friedman case in particular. Friedman’s clients were not harmed by his temporary misappropriation; moreover, because Friedman’s personal wealth exceeded the amount of the settlement funds, they could not have been harmed by his pre-payment (so long as Friedman would have been liable to cover any losses from his personal account). Under the contractarian picture of hypothetical bargaining, if there is no prospect of harm from an action, then that action cannot be part of the content of fiduciary duties. Cognitivist accounts, then, can more straightforwardly explain both aspects of P1 than either proscriptivism or contractarianism.

P2. Ceteris paribus, a fiduciary whose deliberation and behavior are both deficient more seriously violates fiduciary loyalty than one whose behavior or deliberation alone is deficient.

At first, the cognitivist position appears difficult to reconcile with the seemingly “strict” nature of fiduciary liability. In criminal law, violations usually require both a prohibited course of action and a specified mental state. By contrast, breach of fiduciary duty is sometimes said to be a matter of “per se” or strict liability: the fiduciary’s behavior is sufficient to support the conclusion that she has breached her duty.[86] Given the prophylactic nature of fiduciary rules, establishing a fiduciary’s culpable mental state is generally not required to show that she is subject to liability. Likewise, a fiduciary who behaves in a prohibited way cannot generally evade liability on the grounds that the fiduciary’s mental state was innocent or that the fiduciary made an honest mistake.

Despite this apparent tension, however, the strict nature of fiduciary liability is consistent with the cognitivist picture of fiduciary loyalty. That deviant behavior is sufficient for fiduciary liability does not imply that it is also necessary. Even if good intentions don’t exculpate otherwise deviant behavior, bad intentions (of a certain type, at least) might inculpate otherwise innocuous behavior.[87] To draw a parallel with criminal law, the existence of some strict liability crimes does not rule out the possibility that other, more serious crimes might have a mental state component.

Across a range of legal contexts, courts differentiate breaches of fiduciary liability in which the fiduciary lacks a culpable mental state from those in which the fiduciary has a mens rea. Both types of case involve a breach of fiduciary duty. However, courts deem the latter type of breach to be more serious and assign more substantial collateral consequences to it than the former. This logic is captured in P2.

As a first example, take the Friedman case considered above. Although the Friedman court noted that “fiduciary violations do not require subjective awareness of wrongdoing,”[88] Friedman’s “mental state,”[89] his “conscious intention,”[90] his “state of mind,”[91] and his “dishonest” and “selfish motive”[92] were relevant to determining how serious his breach of fiduciary duty was. This framework is consistent with both ABA standards and the Restatement (Third) of the Law Governing Lawyers. Indeed, some analysts conclude that considerations of fault generally explain how courts and other bodies enforce the rules of professional responsibility that govern lawyers, even if many rules of professional conduct can be violated without regard to a lawyer’s culpability.[93] Thus, although “fault” is not required to find a breach of fiduciary obligation as a matter of fiduciary law, it is relevant to considerations about how to classify specific violations of fiduciary duty.

Courts provide a similar (and indirect) kind of scrutiny to the intentional stance of fiduciaries in U.S. bankruptcy law. The U.S. Supreme Court has interpreted § 523(a)(4) of the Federal Bankruptcy Code[94]—which lists “defalcation while acting in a fiduciary capacity” as a non-dischargeable debt—to include a “culpable state of mind requirement.”[95] Notwithstanding that the underlying debt may have been imposed with indifference to the state of mind of the fiduciary, the dischargeability of that debt in bankruptcy turns on whether the fiduciary’s actions “involve[d] knowledge of, or gross recklessness in respect to, the improper nature of the relevant fiduciary behavior.”[96] As far as the Court is concerned, “conscious[] disregard[]” of a risk of violating a fiduciary duty is enough to trigger non-dischargeability.[97] This conclusion supports P2, in that a breach of fiduciary obligation with mens rea is more serious and has a different legal status than a negligent or unwitting breach.

In both the United States[98] and other jurisdictions,[99] an assessment of the fiduciary’s state of mind bears on important questions, such as the availability of defenses and the seriousness of collateral consequences of the fiduciary’s breach, even if a fiduciary can violate the duty of loyalty without a culpable mental state. As the attorney discipline and fiduciary defalcation standards illustrate, the more culpable the fiduciary’s mental state, the more serious the breach of her fiduciary duty will be.

P2 is clearly consistent with cognitivism. The cognitivist can contend that behavior and cognition are both aspects of fiduciary duty. Deficient behavior is sufficient to constitute a breach of fiduciary duty, regardless of whether it is the product of deficient cognition. Thus, the cognitivist can explain the “strict” character of fiduciary liability in terms of institutional considerations—for example, that the discretion fiduciaries have in paradigmatic fiduciary relationships complicates any effort to distinguish malicious actions from benign ones. Fiduciary rules primarily implicate assessments of a fiduciary’s behavior because legal institutions are ill-equipped to conduct routine retrospective assessments of the complex judgments and mental states that constitute fiduciary loyalty. Yet despite the ostensibly behaviorist orientation of many fiduciary rules, actions involving both prohibited behavior and a discernable mens rea are treated as more serious than actions involving only prohibited behavior.

Non-cognitivist accounts of fiduciary loyalty have greater difficulty explaining P2. Existing versions of proscriptivism not only define fiduciary liability entirely in terms of the no-profit and no-conflict rules, but also define both of these prohibitions largely in terms of behavior. As we explained in Part I, construing fiduciary loyalty in terms of behavior is part of what makes proscriptivism akin to fiduciary law for the Holmesian bad man. In light of its implicit behaviorism, proscriptivism cannot attribute any significance to mental states in answering questions related to fiduciary law. Therefore, proscriptivism rules out in advance the possibility of the mens rea-based standards described by P2.[100] P2 also seems inconsistent with existing versions of contractarianism. Contractarians usually see issues of fault and mens rea as irrelevant to questions regarding breach of contract.[101] To the extent that P2 identifies ways that mental states matter to specific questions of fiduciary law, it is inconsistent with existing versions of proscriptivism and contractarianism.[102] Any path to explaining P2 would seem to invoke cognitivism then.

P3. Efforts toward betrayal can violate fiduciary loyalty, regardless of whether these efforts are completed or successful.

In western criminal law, an “inchoate” or incomplete crime is one for which a person can be liable, despite not engaging in prohibited behavior or bringing about a prohibited result. Inchoate crimes include attempt, solicitation, and conspiracy. The most powerful justifications for inchoate crimes presuppose that the criminal law is concerned with a person’s cognition: attempting, soliciting, or conspiring to engage in prohibited behavior or bring about a prohibited result indicates a defendant’s culpably defective deliberation.[103]

As in criminal law, fiduciary law prohibits inchoate efforts to betray a principal. A fiduciary’s attempt to betray a principal—for example, where the fiduciary, with the goal of betraying the principal, takes some significant action towards realizing that goal—constitutes a violation of her fiduciary duty. Solicitations of betrayal—for example, where the fiduciary has the goal of betraying the principal and seeks the help of another in order to realize that goal—also violate the fiduciary’s duty of loyalty. The same conclusion also applies to conspiracies to betray, in which the fiduciary and another agree on a course of action to betray the principal. P3 captures each of these inchoate ways of violating fiduciary loyalty.

Existing fiduciary law supports the idea that attempts, solicitations, or conspiracies to betray the principal can constitute a violation of fiduciary duty. For example, the Restatement (Third) of Law Governing Lawyers announces that “a lawyer is . . . subject to professional discipline . . . for attempting to commit a violation” of the lawyer’s obligations to her client[104] and the American Bar Association’s Model Rule of Professional Conduct 8.4(a) deems it professional misconduct for a lawyer to “violate or attempt to violate the Rules of Professional Conduct.”[105] Likewise, inchoate efforts at betrayal (including solicitations and conspiracies) have been found to violate the duty of loyalty in a variety of other domains of U.S. fiduciary law, for example in agency law,[106] corporate law,[107] and elsewhere.[108] The same conclusions also apply to the law governing fiduciaries in other jurisdictions.[109]

The criminal law of many jurisdictions distinguishes between attempts to commit a crime and “mere preparation” towards committing a crime. The former are criminally prohibited, while the latter are not. In both, a defendant has a plan or conscious objective to act in a prohibited way or bring about a prohibited result. However, “merely preparatory” plans are remote and therefore not appropriately criminalized.[110] Here, fiduciary law diverges from criminal law: some “merely preparatory” efforts of a fiduciary to betray her principal can still constitute a violation of fiduciary loyalty.[111]

To summarize, P3 indicates that a broader variety of activity that does not involve prohibited behavior can nevertheless violate fiduciary loyalty. As in criminal law, fiduciary loyalty can be violated by inchoate attempts, solicitations, or conspiracies. However, unlike criminal law, fiduciary loyalty can be violated by a fiduciary’s “merely preparatory” actions toward betraying the principal.

A cognitivist account of fiduciary loyalty can straightforwardly explain P3. If fiduciary norms concern the fiduciary’s cognition (in particular, her deliberation and conscientiousness), then certain patterns of deficient or deviant cognition can constitute violations of fiduciary norms.[112] Cognitivism can explain why a fiduciary’s deviant behavior is unnecessary to violate fiduciary norms, even if it is sufficient to do so. Trying to betray someone is a way of betraying them, even if the treachery does not succeed and, indeed, even if it does not proceed very far down the path toward success.

Proscriptivism, by contrast, cannot easily explain P3. Proscriptivism interprets the no-conflict and no-profit rules in terms of results. Proscriptivism thus reduces the requirements of fiduciary loyalty to two rules: the fiduciary cannot “act[] with a conflict between duty and interest” and she cannot “mak[e] a profit off of the fiduciary position.”[113] If the fiduciary does not have a conflict of interest, then she does not violate the no-conflict rule. Likewise, if the fiduciary does not make an inappropriate profit from a relationship, then she does not violate the no-profit rule. The clarity of these standards is another basis for our assertion that proscriptivism is fiduciary law for the Holmesian bad man: the fiduciary and others can easily determine whether there has been a violation of fiduciary duty at any time. Yet requiring prohibited results in order to find a violation of fiduciary loyalty is inconsistent with P3, since this requirement rules out the possibility that fiduciary loyalty can be violated by attempted, solicited, conspired, or prepared treachery.

To be sure, proscriptivism might be revised to capture P3. In particular, a proscriptivist might expand the understanding of the no-profit and no-conflict rules to prohibit both successful and inchoate violations. Such a cognitivist revision to proscriptivism seems more defensible than existing versions. After all, whatever makes it wrong for a fiduciary to take profits from a beneficiary would also make it wrong to try or conspire to take these profits. Indeed, we conjecture that courts in jurisdictions that explicitly endorse proscriptivism (“even in Australia,” to paraphrase a popular children’s book from our youth)[114] would adopt something like this cognitivist formulation if confronted with cases of inchoate treachery. But these revisions would reduce the clarity that makes proscriptivism so appealing as fiduciary law for the bad man.

Contractarianism also has difficulty explaining P3. There is no cause of action for an attempted breach of contract; an unsuccessful effort to breach a contract is not a breach.[115] If fiduciary loyalty were merely a species of contract law, then we would expect that fiduciary loyalty could not be breached through attempts. Yet P3 denies exactly this implication. A contractarian might respond to this criticism by construing a prohibition on attempted betrayals as an implicit term that would be agreed upon as part of a hypothetical bargaining process. But this move would concede that fiduciary norms are not fully explicable in terms of contractual norms, that a fiduciary’s deliberation and cognition matter more than a typical contractual counterparty’s do. Moreover, recognizing a cognitive dimension of fiduciary duties would violate one of the basic tenets of the law-and-economics tradition out of which contractarianism arises—namely, that behavior is the fundamental unit of analysis. In sum, both proscriptivism and contractarianism have difficulty explaining P3 because they are not formulated in cognitivist terms.

P4. At least some connections between a fiduciary’s deliberation and action can constitute violations of fiduciary loyalty.

P1, P2, and P3 address some of the deliberative demands the law imposes on fiduciaries. Aside from these deliberative demands, fiduciary loyalty can also implicate the fiduciary’s motivations.[116] In other words, it is possible for a fiduciary to behave and deliberate in appropriate ways, yet to violate her fiduciary duty of loyalty because of inappropriate motivations. Which motivations can violate the duty of loyalty? It is difficult to answer this question in the abstract. However, the point can be illustrated by focusing on some paradigmatic examples.

One type of inappropriate motivation is that of the double agent. A double agent pledges her allegiance to a principal, while simultaneously operating on behalf of another (adverse) principal.[117] Because of the prophylactic nature of the conflicts of interest rules applicable to fiduciaries, the double agent obviously violates fiduciary loyalty regardless of how she deliberates and behaves. That said, perhaps the best explanation for the prophylactic nature of the conflict of interest rule rests on the subtle influence of motivation. The fiduciary’s conflicted interest raises the possibility of changes in her actions, changes so subtle that others (and even the fiduciary herself) might not appreciate them.

Like the double agent, the classic self-dealer clearly violates the fiduciary duty of loyalty. Given the prophylactic character of the no-profit rule, no further assessment of the fiduciary’s motivations or behavior is needed to reach this conclusion. Here too, the prophylactic character of the rules might be justified by a worry about the difficulty of discerning what, exactly, motivates any particular course of action by a fiduciary.

Yet the double agent and the self-dealer do not exhaust the parameters of disloyalty. Consider the saboteur, who acts in order to undermine the projects of her principal rather than to promote her own interests or those of a third party. The saboteur seems at least as treacherous as the double agent. As the U.S. Supreme Court once recognized, “[h]e who is loyal is by definition not a spy or a saboteur.”[118] The main difference between the cases concerns the structure of the treacherous motivation: the double agent subverts in order to advance the interests of a third party, while the saboteur might lack such an ulterior inspiration. There is no good reason, however, for fiduciary law to treat the saboteur differently from the double agent. To be sure, the prophylactic nature of conflict of interest rules makes proving someone to be a double agent easier than proving her to be a saboteur. Yet across a range of fiduciary law domains[119] and jurisdictions,[120] sabotage constitutes a violation of fiduciary duty.

Consider Commonwealth v. Washington,[121] a case concerning effective assistance of counsel under the Sixth Amendment to the U.S. Constitution. Vinson Washington was convicted of robbery and first-degree murder and sentenced to death. On appeal, Washington argued that his counsel at trial was constitutionally ineffective, a claim that requires showing that counsel’s performance was deficient and also that this deficient performance prejudiced the defendant—in particular, that there was a “reasonable probability that the outcome of the trial would have been different” if the defendant had “been represented by adequate counsel.”[122] Washington alleged that his trial counsel hated him.[123] This antipathy, Washington contended, prompted the trial counsel to sabotage Washington’s defense, which (according to Washington) violated counsel’s duty of loyalty toward Washington.[124] Washington argued first that “prejudice” should be presumed “upon proof of animosity between an attorney and client”[125] and (alternatively) that Washington actually suffered prejudice because trial counsel’s animosity led him to perform deficiently on Washington’s behalf.[126] The state countered that Washington’s trial counsel vigorously represented his interests at trial, for example, by extensively cross-examining key witnesses and arguing that Washington’s confession was invalid.[127]

The Washington Court accepted Washington’s argument in part. The animosity of Washington’s trial counsel was not directly relevant to his ineffective assistance claim, as “[t]he Sixth Amendment does not govern the feelings that flow between an attorney and his client.”[128] However, proof of such animosity was relevant to the extent that it indicated trial counsel’s sabotage: “[i]f counsel abrogates his obligation to his client because of personal animosity that the accused will be deprived of his right to a fair trial, such action will support a claim for a violation of the Sixth Amendment,” so long as a client can establish both a direct connection between “the animosity expressed by counsel and the actions of counsel taken on behalf of [the client]” and that “but for the actions of counsel, the outcome of the matter would have been different.”[129] The Washington Court rejected the state’s argument that an entirely objective standard should determine whether the attorney has met his duty of loyalty to the client. Rather, the sabotage alleged in this case (if true) would indicate “a total disintegration of the function of trial counsel, implying a violation of the ethical standards of the profession, a dereliction of counsel’s duty to the court, and a profound failure to the client.”[130] The lesson of Washington is that sabotage violates fiduciary loyalty, while hatred alone is not necessarily inconsistent with fiduciary loyalty. Washington, then, not only illustrates the (legitimate) relevance of cognition to fiduciary obligation, but also shows how cognitive inquiries might devolve into the (illegitimate) policing of the fiduciary’s affect and emotions.

If the saboteur violates a fiduciary duty, then fiduciary law imposes standards regarding the fiduciary’s conscientiousness. However, courts could disagree in how to classify the problem that such a lack of conscientiousness presents. On a more direct analytic route, sabotage might be seen as a violation of fiduciary loyalty itself. This is consistent with the famous Delaware case of Stone v. Ritter, in which the court found the duty of good faith applicable to fiduciaries to be part of the classic fiduciary duty of loyalty.[131]

There are, however, more circuitous routes to ground the legal liability of the saboteur. One such indirect route evaluates the saboteur in light of the rules that implement fiduciary loyalty by “elaborat[ing] the application of loyalty . . . to recurring circumstances.”[132] The saboteur’s lack of conscientiousness could be considered to violate one or more of these implementing rules. Recall Item Software (UK) Ltd. v. Fassihi,[133] discussed in the Introduction. Fassihi, while a director of Item Software, encouraged Dehghani, Item Software’s managing director, to negotiate very aggressively with Isograph. Like Iago, Fassihi had an ulterior motive: to sabotage these negotiations in order to swoop in and obtain Isograph’s business for Fassihi’s own (future) company. The trial court found that Fassihi’s advice was colorable; Fassihi, like Iago, counseled his compatriot down a “parallel course, directly to his good.” In any event, Fassihi’s advice did not influence Dehghani’s conduct. Yet the Fassihi Court upheld the trial court’s finding that Fassihi’s efforts did not violate his duty of loyalty toward Item Software. Rather, Fassihi was liable for violating his fiduciary “duty to confess” by “failing to disclose his intention and attempt to compete with” Item Software.[134] The Fassihi Court took a circuitous strategy because the fiduciary’s conscientiousness was evaluated as part of an implementing rule (namely, the “duty to confess”), rather than as part of the duty of loyalty itself.

The postulation of a fiduciary “duty to confess” has made the Fassihi decision controversial.[135] The decision in Fassihi might have been justified more straightforwardly on the grounds that Fassihi’s efforts violated the “fundamental duty to which a director is subject, that is the duty to act in what he in good faith considers to be the best interests of the company.”[136] In other words, the Fassihi Court could have ruled (consistent with P3) that Fassihi’s unsuccessful attempt to usurp business from Item Software by recommending harsh negotiating tactics against Isograph was a breach of loyalty, even though (consistent with P4) Fassihi’s actual recommendation for this course of advice would not have violated his fiduciary duty if it had been motivated to advance the interests of Item Software.

We do not mean to champion either the direct or circuitous routes for evaluating the conscientiousness of fiduciaries. Under either strategy, whether the fiduciary has lived up to her duty depends not only on how she behaves and deliberates, but also on how that behavior and deliberation fit together. On either strategy, P4 is best explained in terms of a fiduciary’s conscientiousness. The double agent and the saboteur violate their fiduciary duty because the structure of their motivation is inappropriate. The former acts to serve another master, while the latter acts to subvert the beneficiary.

This argument leaves open an important question. If a fiduciary lacks conscientiousness when she acts for inappropriate reasons, then what are appropriate reasons for a fiduciary to act? An unsatisfying (Potter Stewart-esque) response is that one needn’t articulate the parameters of the right kind of reasons for action in order to prove that actions for the wrong kind of reasons are problematic. In any event, courts typically define the right kinds of reasons for action broadly, for example in terms of “good faith” or “honest efforts to advance the beneficiary’s interest.” This broad definition might be justified by the institutional limits to discerning a fiduciary’s motives, as well as from concern with inhibiting the discretion or flexibility of the fiduciary. The conscientiousness norm operates whatever shape the implementing rules take.

What of a fiduciary who has animosity towards his beneficiary? As discussed above in connection with Commonwealth v. Washington, hating a beneficiary seems reconcilable with acting in good faith to advance their interests. Animosity would matter only when it exerts motivational force. That said, evidence of sabotage is difficult to detect, especially after the fact. Many discretionary actions are open to multiple interpretations, some of which are consistent with fiduciary loyalty and others of which are not. A lawyer’s failure to investigate a potential alibi witness might be a strategic decision, or it might be a way for the lawyer to ensure that a reviled client gets what he deserves. Therefore, even though a fiduciary’s hatred of the beneficiary is insufficient to establish a violation of fiduciary loyalty, it is a good place to look for evidence of sabotage, which does violate fiduciary loyalty.

What of a fiduciary whose posture resembles what Harry Frankfurt called wantonness, a complete ambivalence for the interests or ends of the beneficiary?[137] Although genuine wantonness is difficult to find in the world, such a thoroughgoing disregard would seem to violate fiduciary loyalty. Consider the Restatement (Third) of Agency’s discussion of the “General Fiduciary Principle” in section 8.01.[138] Comment (b) posits that “[w]hen an agent’s agreement with a principal confers discretion on the agent to take action in the agent’s sole discretion, the agent has a duty to exercise the discretion in good faith,” regardless of whether the agreement specifies such a duty.[139] As an example, the Restatement contemplates a scenario in which a fiduciary exercises discretionary authority by “flipping a coin,” rather than seeking out information most relevant to the decision. Such an action would violate the fiduciary’s duty to act in good faith, even though the agent would have satisfied this obligation by acting in the same way after “making an honest attempt to make an informed decision” about how to act.[140]

Another example of wantonness as a violation of fiduciary loyalty arises in Delaware corporate law. As Delaware judge Leo Strine and his co-authors put it, “good faith has long been used as the key element in defining the state of mind that must motivate a loyal fiduciary.”[141] A Delaware corporation may not waive damages for actions that are taken “not in good faith.”[142] And as the Delaware Supreme Court has held, the “failure to act in good faith may result in liability because the requirement to act in good faith ‘is a’ . . . condition ‘of the . . . duty of loyalty.’”[143] The notion of good faith under Delaware law works not only to target those (like the saboteur) who are “motivated by a dishonest purpose or ill will towards the corporation and/or its shareholders,” but also two other defective motivational arrangements: fiduciaries acting for “(1) a purpose other than advancing the best interests of the corporation; or [their] (2) intentional disregard of [their] fiduciary duties to the corporation.”[144]

Some of the legal standards for assessing the duty of good faith arose from litigation surrounding the severance pay that the Walt Disney Company negotiated with its outgoing CEO Michael Ovitz. Shareholders argued that, when deciding on Ovitz’s contract, the board was not acting in the best interests of the corporation so much as doing a favor for Michael Eisner, the CEO who hired Ovitz.[145] In Disney, the relevant directors were not accused of having proscribed conflicts of interest, but were instead alleged to have been bad faith stewards of the corporation’s best interests.[146] At issue, however, was not “classic” or “subjective bad faith” requiring ill will or intent to do harm.[147] Rather, the bad faith involved was the directors’ “intentionally act[ing] with a purpose other than that of advancing the best interest of the corporation.”[148] The court found it actionable as a breach of fiduciary duty when directors “adopt[] a ‘we don’t care about the risks’ attitude.”[149] When directors make “material decisions without adequate information and without adequate deliberation,”[150] they are in derogation of their fiduciary obligation. To be sure, the legal elaboration of the duty of good faith under Delaware corporate law is highly complex.[151] Yet the Disney Court’s “we don’t care about the risks” standard identifies the same wantonness that is also condemned in the Restatement (Third) of Agency’s example of coin flipping. In both cases, the (hypothesized) motivational posture violates the conscientiousness that fiduciary loyalty requires.

To summarize, a cognitivist account of fiduciary loyalty can explain why a fiduciary’s motivation matters to the performance of her fiduciary duties, either directly through appraisal in terms of a duty of loyalty or more circuitously through appraisal in terms of some implementing rule or doctrine, such as the duty to confess or—perhaps—the duty of good faith. Cognitivism can also begin to explain which motivations are inappropriate: sabotage definitely, wantonness probably, and hatred only insofar as it motivates actions that are inconsistent with the beneficiary’s interests or ends.

On the other hand, P4 is difficult to reconcile with either the contractarian or proscriptivist positions. Extant versions of both views define loyalty solely in terms of how a fiduciary behaves. Thus, both views would deny that the sorts of motivational considerations described in P4 are relevant to assessing whether a fiduciary has violated the duty of loyalty.

The contractarian might further respond to our critique by denying that P4 is analytically distinct—that is, by asserting that each of our examples of conscientiousness can be accounted for by positing that a distinctive motivational component to fiduciary loyalty would have been bargained for ex ante. For example, the contractarian rules out the fiduciary saboteur from a hypothetical bargaining perspective. The hidden saboteur who plans from the outset of the fiduciary relationship to undermine the beneficiary’s interests could be said to commit promissory fraud.[152] Yet the saboteur whose effort to undermine arises only after the commencement of the fiduciary relationship could not be ruled out from this ex ante perspective. Thus, the contractarian position could conclude that the hidden saboteur violates his fiduciary duties only at the cost of denying that the subsequent saboteur does so as well. Yet there seems to be no meaningful difference between the loyalty of the hidden saboteur and the loyalty of the subsequent saboteur. Nor is there any principled basis for a legal system to treat these cases differently.

Likewise, contractarianism might rule out the wanton fiduciary on the grounds that some sort of performance standard would be agreed on as part of a hypothetical bargain between the fiduciary and beneficiary. To the extent that the wanton falls short of this performance standard, he violates his fiduciary duty toward the beneficiary. Yet this contractarian strategy would have difficulty distinguishing wanton inaction—for example, inaction as a matter of course—from prudent inaction—for example, inaction based on a strategic calculation that it is the best option for advancing the beneficiary’s interests. In contrast to the hidden and subsequent saboteurs, the difference between wanton and prudent inaction does seem meaningful. There is very good reason for a legal system to treat these two cases differently. Yet due to its behaviorist assumptions, contractarianism could not readily explain this disparity.

Proscriptivism has similar difficulties explaining the defectiveness of both the saboteur and the wanton fiduciary, although an advocate of this position might concede that both cases violate fiduciary loyalty.[153] Neither the saboteur nor the wanton serve another master, so neither runs afoul of the no-conflict rule. Further, by hypothesis, neither the saboteur nor the wanton derives any inappropriate profit from the fiduciary relationship. If these rules exhaust the content of fiduciary loyalty, then proscriptivism must deny that either case constitutes a violation of fiduciary loyalty. This conclusion is not only normatively indefensible, but also contrary to existing fiduciary law.[154]

P5. The insensitivity or flimsiness of a fiduciary’s commitment to the beneficiary’s interests or ends can constitute a violation of fiduciary loyalty.

The fiduciary’s duty of loyalty encompasses more than how she happens to behave and deliberate, and more than whether she happens to act for the right kinds of reasons. Fiduciary duties automatically impose demands that are more robust than the default duties imposed by other types of legal norms. A fiduciary must exhibit a special sensitivity towards and a sturdy commitment to the interests or ends of the beneficiary. Fiduciary loyalty can be violated when a fiduciary is insensitive to the beneficiary’s interests or ends or when the fiduciary’s commitment to these interests or ends is too contingent.

First, fiduciary loyalty requires the fiduciary to be sensitive to the interests or ends of the beneficiary.[155] The sensitivity we have in mind here resembles an epistemic notion of sensitivity, according to which someone’s belief in a proposition is sensitive only if he would not have believed that proposition if the proposition had been false.[156] Consider the example of a non-functioning grandfather clock whose hands are stopped at 8:00. The clock’s reading is correct twice a day. It does not follow that someone observing the broken clock at 8:00 PM knows what time it is. One explanation for why not is that the clock is an insensitive mechanism. The clock’s reading does not ground knowledge, even when it is correct, because if the time had been different, the clock still would have read 8:00. Thus, sensitivity acts as a counterfactual constraint on true belief: the clock’s epistemic defects exist at both 7:00 PM and at 8:00 PM.

Courts often invoke the notion of insensitivity to explain why a fiduciary’s actions breach his or her duty of loyalty.[157] However, the parallel with epistemic sensitivity allows for a more precise articulation of the term in legal settings. In the same way that the requirements for accurate time telling vary based on changes in the external world, so too do the requirements on fiduciaries change based on changes to the situation of the beneficiary.

Some characterize these changes in terms of the open-endedness of fiduciary duties.[158] We have described this phenomenon in terms of “morphing”—the requirements applicable to a fiduciary change automatically based on changes to the interests or ends of the beneficiary.[159] On either formulation, fiduciary loyalty can be violated by a fiduciary’s failure to monitor relevant changes to the interests or ends of the beneficiary.[160] For example, in In re Caremark International Inc. Derivative Litigation, the Delaware Court of Chancery found that director-fiduciaries must monitor the ongoing operations of a company and must “exercise appropriate attention” to what is going on within the corporation in order to fulfill their duty of loyalty.[161] Thus, the duty of loyalty can also be violated by a fiduciary’s failure to update his understanding of his responsibilities that apply to him based on such changes, or by a failure to revise his course of action based on the changes to the beneficiary’s interests or ends.[162] Moreover, as with epistemic insensitivity, there can be purely counterfactual violations of fiduciary duties—that is, violations of fiduciary loyalty based on defects of performance that are based on features that are not realized in the actual world but that would have been realized in close possible worlds.[163] This possibility is contemplated in the U.S. Supreme Court’s interpretation of a conflict-of-interest statute as being “more concerned with what might have happened in a given situation than with what actually happened.”[164] These requirements of sensitivity are best seen as implications of fiduciary loyalty, even if courts do not always classify all of them in this language.

A second aspect of the robustness of fiduciary duties is that they can be violated when a fiduciary’s commitment to the beneficiary is too flimsy or contingent. There is a modal constraint on the fiduciary duty of loyalty. For example, while the fiduciary need not stick with the beneficiary through thick and thin, he may not simply leave the beneficiary in order to pursue a more lucrative relationship. The flimsiness of the fiduciary’s commitment to the beneficiary can therefore constitute a violation of the duty of loyalty.[165]

The so-called “hot potato” doctrine in the field of professional responsibility[166] illustrates how sturdiness is an aspect of fiduciary loyalty. Suppose that Alice Attorney, a general practitioner, represents Claude Client on a routine estate planning matter. Pete Patron retains Alice for a potentially lucrative tort suit against number of defendants, including Claude. The ABA’s Model Rules of Professional Conduct (“Model Rules”) prevent an attorney from representing one client when that representation will be “directly adverse to another client,” even if (as in our scenario) the representation of one client is substantially unrelated to the representation of the other, unless both clients consent.[167] On the current-client conflict rule, then, Alice could not represent Pete in the tort suit unless Claude consented to the representation. However, a more lenient rule applies to suits against a former client: so long as the lawyer’s representation of the current client is not substantially related to the lawyer’s representation of the former client, the lawyer may represent the current client even if the former client does not consent to the arrangement.[168] On this former-client conflict rule, then, Alice could represent Pete in the tort suit without needing Claude’s consent if Claude were Alice’s former (rather than current) client. Finally, the Model Rules provide wide latitude for terminating the representation of a client: a lawyer may conclude a representation for any reason, so long as the termination of a representation does not have a “material adverse effect” on the interests of a client.[169] The intersection of all three of these rules suggests that Alice could obviate the need for Claude’s consent to her representation of Pete in the tort suit by dropping Claude “like a hot potato”[170]—that is, by terminating her representation of Claude and thereby converting a current-client conflict into a former-client conflict.

Many courts resolve these “hot potato” scenarios by prohibiting the lawyer from “firing” the current client in order to represent the more lucrative client, even though this move (however unseemly) is licit under the professional conduct rules. This restriction is typically justified in terms of the lawyer’s overarching duty of loyalty to the client. Although the language of many opinions in these cases is phrased in terms of “undivided loyalty,” the most relevant factors cited by courts concern the flimsiness of the lawyer’s commitment to the less-remunerative client. For example, one court argued that a lawyer’s termination of a relationship only triggers the former-client conflict rule if “the lawyer’s primary motivation for terminating” the representation “was not his desire to represent” a new, more lucrative client.[171]

A cognitivist account can straightforwardly explain why both insensitivity and flimsiness violate fiduciary loyalty. Insensitivity is problematic, regardless of whether it leads to problematic results, because it violates an independent requirement of fiduciary loyalty. The flimsiness of a fiduciary’s commitment is also problematic, independent of performance, because it suggests a similar defect regarding these considerations.

On the other hand, contractarian accounts cannot explain why insensitivity is problematic. As noted above, extant versions of contractarianism define loyalty in terms of behavior and expected results. Although contractarianism might deem insensitivity to be problematic insofar as it leads a fiduciary to engage in problematic behavior or produce substandard results, it cannot constitute a violation of fiduciary loyalty on its own. Yet P5 contemplates that insensitivity is sufficient to violate fiduciary loyalty. Furthermore, it is difficult to reconcile a modal constraint like insensitivity within the contractarian framework. In contract law, unrealized contingencies do not typically constitute breaches of contractual duty. From the perspective of an ex ante bargain, such contingencies might be seen as priced into the bargain. Therefore, a contractarian couldn’t easily explain why purely counterfactual considerations (like unrealized but “possible” conflicts of interest) could constitute a breach of a contractual duty. Yet the insight behind insensitivity is that purely counterfactual considerations can be sufficient to constitute a breach of the fiduciary duty of loyalty.

Proscriptivism faces similar challenges accounting for P5. Mere insensitivity cannot violate a fiduciary duty if no explicit fiduciary rule requires sensitivity or prohibits insensitivity. Proscriptivism cannot explain why the flimsiness of a commitment might run afoul of fiduciary duties, since the relevant proscriptions are framed in terms of actual events rather than in terms of modal operators.

* * *

Our case for cognitivism about fiduciary loyalty has been both empirical and interpretive, albeit largely indirect. A cognitivist account of fiduciary loyalty provides a unified explanation for all of the legal propositions we have identified in this Part. Accounts of fiduciary law that deny cognitivism—for example, contractarian and proscriptivist approaches—might be able to explain some of these legal propositions, but they fail to provide a unified explanation for all of them. Moreover, the explanations offered by contractarianism and proscriptivism are ad hoc, perhaps due to the implicit commitment of these views to assess fiduciary loyalty entirely in terms of behavior. The best interpretation of fiduciary law, then, is that fiduciary loyalty has a cognitive dimension.

III.  Objections and Replies

This Part articulates and responds to several concerns about cognitivism regarding fiduciary loyalty. These responses both summarize our conclusions and highlight some important implications of our analysis.

A.  Does Cognitivism Actually Resolve the Dispute Between Moralism and Amoralism?

Our initial inquiry was framed in terms of the seeming impasse between moralist and amoralist accounts of fiduciary loyalty. Appreciating the cognitive dimension of fiduciary loyalty does not fully resolve the debate between moralism and amoralism. Indeed, many of the most pressing questions in this debate remain disputable. For example, cognitivism has little to say about the disclaimability of fiduciary obligation. Nor does it yield any clean insights about important questions such as whether ordinary moral considerations are relevant to understanding fiduciary liability or remedies for breach of fiduciary duties. Rather, we see cognitivism as a potential common ground for moralist and amoralist positions.

To be sure, we are not wholly agnostic between the moralist and amoralist positions. On our definition, a moralist account holds that fiduciary loyalty references at least some aspects of the ordinary notion of loyalty. If, as we argued in Section II.A, ordinary loyalty has a cognitive dimension, then cognitivism about fiduciary loyalty could entail a weak form of moralism: that at least one aspect of ordinary loyalty (namely, its cognitive dimension) applies directly to fiduciary loyalty. (We leave open whether other structural features of the ordinary concept of loyalty might illuminate fiduciary loyalty.) No stronger form of moralism follows from cognitivism, however. It is possible to accept that fiduciary loyalty has a cognitive dimension while denying, for example, that the ultimate justification for fiduciary and ordinary loyalty are identical, or that ordinary moral considerations inform the appropriate remedies for violations of fiduciary loyalty. Furthermore, the kinds of deliberation, conscientiousness, and sensitivity appropriate for fiduciaries almost certainly diverge from the requirements as they apply to loyal friends, lovers, and compatriots. Indeed, the minimal standards of deliberation, conscientiousness, and robustness are likely to vary significantly across fiduciary contexts. If so, then the structural requirements of loyalty are unlikely to be identical across moral life and fiduciary types. Ultimately, amoralists can embrace cognitivism about fiduciary loyalty as an explanation for the legal propositions identified in Section II.B, since none of these propositions directly invokes the “ordinary” notion of loyalty.

Appreciating the cognitive dimension of fiduciary loyalty, then, does not fully resolve the debate between moralism and amoralism. And both sides in this debate might find our analysis unsatisfying. The amoralist might find our conclusion distasteful because it points to some substantial structural similarity between fiduciary loyalty and ordinary loyalty. On the other hand, the moralist might see our conclusion as too concessive, since it admits the possibility of explicating large swaths of fiduciary law in purely conventional or institutional terms. We see this lack of resolution as appealing, rather than problematic. Accepting that fiduciary loyalty has a cognitive dimension still leaves open many interesting theoretical questions about fiduciary law. Yet the common ground provided by cognitivism provides a new basis for addressing policy and design questions regarding fiduciary law generally, and corporate law, trust law, and the law governing lawyers, in particular.

For example, cognitivism has implications for debates about the applicability of fiduciary duties to investment advisers and brokers. An investment adviser engages “primarily in advisory activities, including portfolio selection, asset allocation, portfolio management, selecting and monitoring other advisers, and financial planning.”[172] A broker provides investment advice to an investor, while also “executing trades, selling securities, lending money to investors to invest on margin, [and] maintaining custody of funds and securities.”[173] As a matter of U.S. law, investment advisers have long been held to operate under “federal fiduciary standards,” including affirmative duties of good faith and care.[174] Recent legal developments have also applied a “fiduciary standard” to actions by brokers, although this application is more contested.[175] The controversy surrounding the Department of Labor’s (“DOL”) Fiduciary Rule[176] is, in the main, about whether both investment advisers and broker-dealers should be subject to fiduciary standards.[177]

Cognitivism can support those who argue for the application of fiduciary duties to investment managers and brokers, as well as those who support the DOL Fiduciary Rule. Investment advice invites the possibility of betrayal generally, as well as concerns with each of the cognitive elements of loyalty that we have identified above. The client of an investment manager or broker has reason to be concerned with potential defects in the deliberation, conscientiousness, or commitment of an investment adviser, regardless of how the adviser actually behaves (that is, independently of the content of the advice that she provides).[178] Broker-client relationships implicate all of the same concerns with cognition that arise in the advice-giving context, while also inviting the possibility of betrayal in the context of a broker’s acting on behalf of clients in the purchase of securities for their accounts. All of these concerns with betrayal apply regardless of whether the broker has a recognized conflict of interest and independently of an objective assessment of the broker’s advice or behavior. Therefore, a cognitivist understanding of fiduciary loyalty can be marshaled to hold advisers and brokers to fiduciary standards.

A further question concerns whether fiduciary duties should apply to so-called “robo-advisers,” automated services that provide financial advice (usually concerning investment, but sometimes about insurance and banking) to clients on the basis of algorithms.[179] Here, too, cognitivism provides support to those who favor applying fiduciary duties. However, this support is more contingent. A cognitivist might contend that fiduciary duties should apply to the extent that algorithms invite the possibility of betrayal—in particular, betrayal by the humans who formulate, market, or implement the algorithm. For example, an algorithm programmed to favor investment vehicles that promote the adviser’s products over superior products from other firms would seem to violate the duty of loyalty. As Baker and Dellaert note, “the human/machine handoff provides significant opportunities to take advantage of consumers . . . .”[180]

However, cognitivism suggests that different standards might apply to so-called “black-box” algorithms, which utilize “opaque computational models to make decisions” about what to advise or implement.[181] Such models are “non-transparent,” in that the “relationships they capture cannot be explicitly understood” by either customers or programmers “and sometimes cannot even be explicitly stated.”[182] While “black-box” robo-advisers might be evaluated in terms of the quality of their advice, they do not seem to invite the same possibilities of betrayal as human-based (or human-algorithm coordinated) investment management. As such, cognitivism does not clearly provide additional support for applying fiduciary duties to “black box” robo-advisers, although other kinds of considerations might support the application of fiduciary duties in connection with these products.[183]

Our goal here is not to definitively resolve these ongoing debates about fiduciary duties. However, our analysis might be utilized in these debates. Cognitivism not only helps to explicate what responsibilities fiduciaries have, but also supports arguments about when specific legal relationships are or should be governed by fiduciary duties.

B.  Is Cognitivism a Conceptual Claim About the Nature of Fiduciary Loyalty?

Our conclusions about cognitivism are not only empirical and interpretive. They are also normative—that is, they offer a specific vision about what fiduciary law ought to be. However, our conclusions are not conceptual. We do not claim that cognitivism is a necessary truth about fiduciary law.

As demonstrated in Section II.B, the cognitive dimension of fiduciary loyalty is part of fiduciary law across legal contexts in a variety of jurisdictions. The normative costs of denying cognitivism would likely be steep. The tenets of cognitivism are consistent with the intuitive understanding of ordinary loyalty, and they are part of many fiduciaries’ self-understanding. Moreover, the costs of rejecting cognitivism exceed the normal drawbacks of defining legal terms in technical ways that deviate from common sense and positive law. A non-cognitivist fiduciary loyalty would have difficulty explaining why the saboteur or the schemer has breached her fiduciary duty. More broadly, a non-cognitivist understanding of fiduciary loyalty would subject beneficiaries to the kinds of predation and exploitation that fiduciary duties generally are supposed to prevent. Whatever the ultimate goals of fiduciary law are, they are likely better served by attributing at least some cognitive dimension to fiduciary loyalty.

However, just because it would be unwise for a legal system to deny the cognitive dimension of fiduciary loyalty does not mean that doing so is impossible. Although cognitivism is part of fiduciary law in the jurisdictions and legal contexts we have examined, it is logically possible for a jurisdiction to define fiduciary loyalty entirely in non-cognitive terms. It is an intelligible position that fiduciary loyalty is entirely a matter of how the fiduciary behaves and/or what results from his actions. Our case for cognitivism does not rise to the level of a Fullerian claim about the internal morality of law. In a legal system genuinely concerned with limiting opportunism and predation, it would be unwise, but not incoherent, to reject cognitivism. Nor do we deny that there might be some advantages to such a non-cognitivist understanding of fiduciary loyalty.[184] These benefits, however, would likely not outweigh the costs of rejecting cognitivism. In any event, such a legal system would be very different from any that we have observed.[185]

C.  Can Cognitivism Be Reconciled with the Systematic Underenforcement of Fiduciary Duties? (Or: If Fiduciary Law Is Cognitivist, How Come It’s so Lax?)

In many legal contexts, the enforcement of fiduciary loyalty is relatively rare.[186] This laxity of enforcement is often thought to arise from divergence between the more exacting standards of conduct and the more forgiving standards of review. If fiduciaries are only occasionally held to any standard of loyalty (let alone the rarefied, cognition-based standard that we have identified here), then is cognitivism a meaningful part of fiduciary law?

To be sure, implementing rules diverge from primary norms in many other areas besides fiduciary law.[187] However, if fiduciary law ignored the cognitive dimensions of fiduciary loyalty, then saboteurs would be treated differently than double agents and unsuccessful betrayals would fulfill the duty of loyalty. Such an arrangement would contradict the propositions of fiduciary law that we have identified here and run afoul of the norms against opportunism and predation that seem fundamental to fiduciary law on both moralist and amoralist views. That said, the systematic underenforcement of the deliberative, motivational, and robustness aspects of fiduciary loyalty might be explained in at least two principled ways. Both of these explanations concede that non-cognitivist accounts of fiduciary loyalty have some power to describe and shape the institution of fiduciary law. However, both ultimately vindicate cognitivism about fiduciary loyalty.

One kind of explanation is a version of Meir Dan-Cohen’s “acoustic separation” thesis, which posits that law speaks simultaneously to multiple audiences.[188] On Dan-Cohen’s formulation, conduct rules speak to those who are governed by the positive law, while decision rules communicate information to legal officials. A slightly different separation seems applicable to fiduciary law. To wit, on Henry Smith’s understanding of equity law, fiduciary law might be seen to speak simultaneously to the Holmesian “bad man” and to the “good person.” If law is construed as dealing with the good person (one who internalizes a legal norm and its cognitive implications), then the ambiguity of a legal standard can actually increase the likelihood that people will adhere to that standard.[189] In these cases, ambiguity about the dictates of a legal standard regarding enforcement serves to “crowd in” fiduciaries whose prior orientation is toward norm-adherence.[190] For this group, the vague standards of conduct applicable within extant fiduciary law can be expected to reinforce and promote deliberation about what to do and about what their fiduciary responsibilities require.[191]

For the Holmesian bad man, however, fiduciary law aims to police opportunism. On the acoustic separation story, “the same literal message” regarding fiduciary law’s clear proscriptions “can serve as an antievasion device for bad-faith actors while not interfering with (or even while promoting) the intrinsic [orientation] [] of the good–faith actors.”[192] This acoustic separation helps reconcile the cognitive dimension of fiduciary loyalty with the systematic underenforcement of fiduciary duties regarding deliberation, motivation, and commitment: more aggressive policing of these dimensions would undercut much of the value of fiduciary loyalty for the good person. On this strategy, then, we would expect judicial relief only in the most egregious cases of deliberative, motivational, and commitment failure.

As discussed in Section II.B, this is the pattern that the cases reveal. For example, in Commonwealth v. Washington, the contempt exhibited by Washington’s lawyer might have indicated disloyalty, even though the more prosaic antipathy that many lawyers sometimes feel toward their clients would not have provided any evidence of disloyalty. The systematic underenforcement of cognitive aspects of fiduciary loyalty, then, might allow fiduciary law to speak to both the good person and the bad man at the same time.

A second, and related, explanatory strategy for lax enforcement is rooted in what one might call a paradox of loyalty at the center of the incompatibilist’s insight. Loyalty in life (as in law) requires someone to act for the right kinds of reasons. If someone is motivated to act solely because of fear of legal enforcement, then (depending on the context) her motivational structure might constitute acting for the wrong kind of reason. Too tightly specifying the rules governing the cooperation and trust between fiduciary and beneficiary might inhibit the formation of genuine trust.[193] Thus, combining lax enforcement policies with the uncertainty generated by hortatory language (as in Meinhard v. Salmon) establishes legal incentives that can facilitate the development of intrinsic motivations, even for the bad man. On this explanation, then, the case for underenforcement is itself prophylactic: extensive enforcement of the cognitive aspects of fiduciary loyalty would be self-defeating.

In sum, if cognitivism is true, then the standards for living up to fiduciary loyalty are often far more demanding than the legal standards of accountability that courts and officials apply to fiduciaries. However, both the “acoustic separation” and “self-defeating” stories can explain this divergence between legal norms and legal enforcement.

Conclusion

Loyalty is, at least in part, a matter of cognition: how someone deliberates, what motivates her, the sturdiness of her commitments. This cognitive structure to loyalty in the world is as intuitive as it is important. To disregard the thoughts and plans of Iago and Littlefinger is to understate their treachery.

The same is true about fiduciary loyalty. Behavior is, of course, the most important concern of legal rules. However, a purely behavioral understanding of fiduciary duties is incomplete. Many widely-held legal propositions are difficult to explain without referencing a fiduciary’s cognition. Cognitivism about fiduciary loyalty challenges the most prominent amoralist accounts of fiduciary duties, especially proscriptivism and contractarianism, which define fiduciary duties solely in terms of how a fiduciary behaves.

Cognitivism does not necessarily resolve the longstanding debate between moralists and amoralists about fiduciary loyalty, let alone in other areas of law where similar debates arise.[194] Rather, it provides a common ground from which proponents of both positions might better understand the central norms of fiduciary law. Both inside the law and out, loyalty makes demands on the inside of you.

[*] *. Stephen R. Galoob is an Associate Professor of Law at the University of Tulsa College of Law; Ethan J. Leib is the John D. Calamari Distinguished Professor of Law at Fordham University School of Law. For comments and questions on the draft, we thank Aditi Bagchi, Alan Brudner, Hanoch Dagan, Deborah DeMott, Chris Essert, Mark Gergen, Andrew Gold, John Goldberg, Jeff Gordon, Rob Kar, Dmitry Karshtedt, Larissa Katz, Greg Keating, Sung Hui Kim, Daniel Markovits, Lauren Scholz, Henry Smith, Steve Smith, Ben Zipursky, and, especially, Paul Miller and Zoë Sinel. We presented an earlier version of this paper at the North American Workshop on Private Law Theory at the University of Southern California Gould School of Law.

 [1]. See, e.g., Frame v. Smith, [1987] 2 S.C.R. 99, 136 (Can.) (“Relationships in which a fiduciary obligation ha[s] been imposed seem to possess three general characteristics: (1) The fiduciary has scope for the exercise of some discretion or power. (2) The fiduciary can unilaterally exercise that power or discretion so as to affect the beneficiary’s legal or practical interests. (3) The beneficiary is peculiarly vulnerable to or at the mercy of the fiduciary holding the discretion or power.”); United States v. Chestman, 947 F.2d 551, 569 (2d Cir. 1991) (“A fiduciary relationship involves discretionary authority and dependency. . . . [T]he beneficiary of the relation may entrust the fiduciary with custody over property of one sort or another. Because the fiduciary obtains access to this property to serve the ends of the fiduciary relationship, he becomes duty-bound not to appropriate the property for his own use.”); Paul Miller, A Theory of Fiduciary Liability, 56 McGill L.J. 235, 261–62 (2011) (“In my view, a sound definition of the fiduciary relationship [is that] a fiduciary relationship is one in which one party (the fiduciary) enjoys discretionary power over the significant practical interests of another (the beneficiary).”) (emphasis omitted).

 [2]. Item Software (UK) Ltd. V. Fassihi [2004] EWCA (Civ) 1244 (Eng.).

 [3].               See, e.g., Paul B. Miller, Dimensions of Fiduciary Loyalty, in Research Handbook on Fiduciary Law 180, 181 (Andrew S. Gold & D. Gordon Smith, eds., 2018) (noting that “moralist” interpretations of fiduciary loyalty suggest that “fiduciary loyalty incorporates moral standards of loyalty or that it expresses moral ideals of selflessness”); Henry E. Smith, Why Fiduciary Law Is Equitable, in Philosophical Foundations of Fiduciary Law 261, 261 (Andrew S. Gold & Paul B. Miller, eds., 2014) (“Fiduciary duties are expressed in moral language of exacting honor, which is treasured by some as the essence of the area and as a flowery show by others. The moralists see in fiduciary law a fixed and mandatory system . . . .”); Hanoch Dagan & Sharon Hannes, Managing our Money: The Law of Financial Fiduciaries as a Private Law Institution, in Philosophical Foundations of Fiduciary Law, supra, at 91, 103 (describing as “moralist” accounts of fiduciary loyalty that beneficiaries “are owed the highest degree of altruism” and/or that fiduciary law “should be particularly concerned with motives in the context of fiduciary relationships”).

 [4]. Examples of moralism about fiduciary law might include Peter Birks, The Content of Fiduciary Obligation, 34 Isr. L. Rev. 3, 15–17 (2000); Scott FitzGibbon, Fiduciary Relationships Are Not Contracts, 82 Marq. L. Rev. 303, 338 (1999) (“Fiduciary relationships are not creatures only of law and lawyers. Fiduciary relationships and fiduciary duties reflect the precepts of social morality and practice.”); Tamar Frankel, Fiduciary Law, 71 Calif. L. Rev. 795, 829–30 (1983) (“Courts regulate fiduciaries by imposing a high standard of morality upon them. This moral theme is an important part of fiduciary law. Loyalty, fidelity, faith, and honor form its basic vocabulary.”); Matthew Harding, Trust and Fiduciary Law, 33 Oxford. J. Legal Studs. 81, 82 (2013) (arguing that “moral duties referring to trust play a role in the justification of fiduciary duties”); Irit Samet, Fiduciary Loyalty as Kantian Virtue, in Philosophical Foundations of Fiduciary Law, supra note 3, at 126 (“By importing the loaded concept of loyalty from ethics and sociology, equity endeavors to encapsulate a subtle and complex aspect of the fiduciary relationship.”); Lionel D. Smith, Can We Be Obliged To Be Selfless?, in Philosophical Foundations of Fiduciary Law, supra note 3, at 142 (“[T]here is a legal requirement of loyalty, which is found in the positive law of fiduciary obligations and which is consistent with the requirements of the law’s philosophical foundations.”).

 [5].               Examples of scholarship that incorporate direct connections to moral concepts include Lyman Johnson, After Enron: Remembering Loyalty Discourse in Corporate Law, 28 Del. J. Corp. L. 27, 47–55 (2003) and Leo E. Strine Jr. et al., Loyalty’s Core Demand: The Defining Role of Good Faith in Corporation Law, 98 Geo. L.J. 629, 655 (2010). The most prominent example in the field of legal ethics is Charles Fried, The Lawyer as Friend: The Moral Foundations of the Lawyer-Client Relation, 85 Yale L.J. 1060 (1975). Recent examples of trust law moralism include Melanie Leslie, Trusting Trustees: Fiduciary Duties and the Limits of Default Rules, 94 Geo. L.J. 67 (2005) and Lee-Ford Tritt, The Limitations of an Economic Agency Cost Theory of Trust Law, 32 Cardozo L. Rev. 2579 (2010).

 [6]. To call this position moralist is to beg important questions about the moral or normative significance of loyalty in the first place. See Simon Keller, The Limits of Loyalty (2007). As noted below, someone might adopt the moralist position about fiduciary loyalty while denying that loyalty has specific hallmarks of moral significance—for example, the capacity to change what someone is morally permitted or required to do.

 [7]. Andrew Gold, Interpreting Fiduciary Law, in Research Handbook on Fiduciary Law, supra note 3, at 37–38 [hereinafter Gold, Interpreting].

 [8]. Meinhard v. Salmon, 164 N.E. 545 (N.Y. 1928).

 [9]. Id. at 546 (internal citations omitted).

 [10]. For example, sections 103(b) and 404(e) of the Uniform Partnership Act (“UPA”) indicate that conflicted transactions can be authorized by a vote of partners and that a partner’s benefitting from conduct is not, per se, a violation of fiduciary duty. Elsewhere, the UPA specifically denies that the standards applicable to trustees are applicable to partners. See Uniform Partnership Act
§ 404, cmt. 5.

 [11]. See, e.g., In re Cooperman, 633 N.E.2d 1069, 1073 (N.Y. 1994) (“The conduct of attorneys is not measured by how close to the edge of thin ice they skate. The measure of an attorney’s conduct is not how much clarity can be squeezed out of the strict letter of the law, but how much honor can be poured into the generous spirit of lawyer-client relationships.”); Maritrans GP, Inc. v. Pepper, Hamilton & Scheetz, 602 A.2d 1277, 1283 (Pa. 1992) (“There are few of the business relations of life involving a higher trust and confidence than those of attorney and client . . . [and] few governed by sterner principles of morality and justice.” (quoting Stockton v. Ford, 52 U.S. (11 How.) 232, 247 (1850))); Darby v. Furman Co., 513 S.E.2d 848, 849 (S.C. 1999) (“Real estate agents occupy a fiduciary relationship with their clients and are under a legal obligation as well as a high moral duty to give loyal service to the principal.”); Owen v. Shelton, 277 S.E.2d 189, 192 (Va. 1981) (finding that fiduciary duties “illustrate[] the high regard the law holds for the fiduciary relationship, founded as it is upon one man’s trust in the integrity and fidelity of another”); Zastrow v. Journal Commc’ns, Inc., 718 N.W.2d 51, 60 (Wis. 2006) (“A breach of the duty of loyalty imports something different from mere incompetence; it connotes disloyalty or infidelity.” (internal citations omitted)).

 [12]. E.g., Wis. Stat. § 112.01(c) (2018) (“A thing is done ‘in good faith’ within the meaning of this section, when it is in fact done honestly, whether it be done negligently or not.”); Daugherty v. Runner, 581 S.W.2d 12, 16 (Ky. 1978) (“Since the relationship of attorney-client is one fiduciary in nature, the attorney has the duty to exercise in all his relationships with this client-principal the most scrupulous honor, good faith and fidelity to his client’s interest.”). Lusina Ho summarizes developments in English company law as elevating “good faith” from “an exonerating circumstance in specific contexts to a general duty.” Lusina Ho, Good Faith and Fiduciary Duty in English Law, 4 J. Equity 19, 21 (2010).

 [13]. United States v. White Mountain Apache Tribe, 537 U.S. 465, 475 (2003) (“[E]lementary trust law . . . confirms the commonsense assumption that a fiduciary actually administering trust property may not allow it to fall into ruin on his watch.”); United States v. Mett, 65 F.3d 1531, 1538 (9th Cir. 1995) (Kleinfeld, J., concurring) (“A lawyer’s duty to avoid conflict of interests arises from agency law. The ancient principle is that ‘no man can serve two masters.’ Matthew 6:24.”).

 [14]. Frankel, supra note 4, at 831.

 [15].               See, e.g., Stephen M. Bainbridge, Must Salmon Love Meinhard? Agape and Partnership Fiduciary Duties, 17 Green Bag 2D 257, 270 (2014) (contending that moralized rhetoric of fiduciary law has instrumental value because, “[p]artners who love one another can trust one another. In turn, partners who trust one another will expend considerably less time and effort – and thus incur much lower costs – monitoring one another.”); Lawrence E. Mitchell, The Death of Fiduciary Duty in Close Corporations, 138 U. Pa. L. Rev. 1675, 1696 (1990) (“The language expressing [fiduciary] norms is aspirational and studiously imprecise. The very ambiguity of the language conveys its moral content as the court’s refusal to set lines is designed to discourage marginal conduct by making it difficult for a fiduciary to determine the point at which self-serving conduct will be prohibited, and thus to encourage conduct well within the borders.”).

 [16]. Frank H. Easterbrook & Daniel R. Fischel, Contract and Fiduciary Duty, 36 J.L. & Econ. 425, 427 (1993).

 [17]. Bray v. Ford [1896] AC 44 (HL) 51 (Lord Herschell) (appeal taken from Eng.); see also Market St. Assocs. Ltd. P’ship v. Frey, 941 F.2d 588, 595 (7th Cir. 1991) (finding the duty of good faith in carrying out a contract “is, as it were, halfway between a fiduciary duty (the duty of utmost good faith) and the duty merely to refrain from active fraud. Despite its moralistic overtones it is no more the injection of moral principles into contract law than the fiduciary concept itself is”).

 [18]. Gold, Interpreting, supra note 7, at 38–42 (citing, inter alia, William Lucy, Method and Fit: Two Problems for Contemporary Philosophies of Tort Law, 52 McGill L.J. 605 (2007)).

 [19]. See, e.g., Easterbrook & Fischel, supra note 16, at 429 (“[W]e seek knowledge of when fiduciary duties arise and what form they take, not a theory of rhetoric—a theory of what judges do, not of explanations they give.”).

 [20]. See, e.g., Avihay Dorfman, On Trust and Transubstantiation: Mitigating the Excesses of Ownership, in Philosophical Foundations of Fiduciary Law, supra note 3, at 339, 341; Andrew Gold, The Loyalties of Fiduciary Law, in Philosophical Foundations of Fiduciary Law, supra note 3, at 176, 192 [hereinafter Gold, Loyalties].

 [21]. See, e.g., Rob Atkinson, Obedience as the Foundation of Fiduciary Duty, 34 J. Corp. L. 43, 49n.16 (2008) (explaining that fiduciary duties of loyalty, care, and obedience “vary in their particular content and contours among various kinds of fiduciary relationships”); Deborah A. DeMott, Beyond Metaphor: An Analysis of Fiduciary Obligation, 1988 Duke L.J. 879, 879 (“Although one can identify common core principles of fiduciary obligation, these principles apply with greater or lesser force in different contexts involving different types of parties and relationships.”).

 [22]. See, e.g., Easterbrook & Fischel, supra note 16, at 432.

 [23]. See, e.g., Gold, Loyalties, supra note 20, at 193–94 (“Loyalty duties are an essential feature of fiduciary relationships, but no specific conception of loyalty appears to carry the day. . . . [P]erhaps part of what is special about fiduciary law is that it requires different kinds of loyalty for different kinds of relationship[s]. Loyalty varies in our social experiences—it also varies in the law.”). Stronger and weaker versions of the pluralist position are possible. On the weaker version, certain aspects of fiduciary norms (for example, the standards for living up to the duty of loyalty or the implications of violating the duty of loyalty) differ across jurisdictions or domains of law, while others (for example, the overall justification for fiduciary duties and the remedies available to beneficiaries for violations by fiduciaries) might overlap. On a stronger version of pluralism, no aspects of fiduciary norms necessarily generalize across legal contexts.

 [24]. E.g., Hanoch Dagan, Fiduciary Law and Pluralism, in Oxford Handbook of Fiduciary Law (Evan J. Criddle, Paul B. Miller & Robert H. Sitkoff eds., forthcoming 2019).

 [25]. Jordon v. Duff & Phelps, Inc., 815 F.2d 429 (7th Cir. 1987).

 [26]. Gold, Interpreting, supra note 7, at 39–42. Gold observes that a variety of other judicial opinions, “perhaps most notably . . . Delaware corporate law,” also employ hypothetical bargaining analysis. Id. at 41.

 [27]. See, e.g., id. at 2 (“For every case in fiduciary law that applies an efficiency-focused, hypothetical bargain approach to ascertaining fiduciary duties, there is another case that applies a morally resonant, deontological alternative.”).

 [28]. A weaker form of pluralism—which acknowledges a thin core—might allow that core to cross-reference moral norms. Since our definition of moralism requires a deeper connection of fiduciary law to ordinary loyalty, it is likely that even such a weak pluralism would be amoralist on our terms. Not much, however, turns on this classification.

 [29]. See, e.g., Stephen A. Smith, The Deed, Not the Motive: Fiduciary Law Without Loyalty, in Contract, Status, and Fiduciary Law 213 (Andrew S. Gold & Paul B. Miller eds., 2017); J.E. Penner, Is Loyalty a Virtue, and Even If It Is, Does It Really Help Explain Fiduciary Liability?, in Philosophical Foundations of Fiduciary Law, 159 (Andrew S. Gold & Paul B. Miller ed., 2014); Paul B. Miller, Justifying Fiduciary Remedies, 63 U. Toronto L.J. 570 (2013); Miller, supra note 3. One common element among all of these views is that crucial structural features inherent in fiduciary relationships are not necessarily realized in non-legal relationships, and vice versa.

 [30]. See, e.g., Penner, supra note 29, at 161–62. This incompatibilist logic might be extended to a variety of other dimensions of “moral” or “ordinary” loyalty, such as the concern with devotion or motivation. See Deborah A. DeMott, Breach of Fiduciary Duty: On Justifiable Expectations of Loyalty and Their Consequences, 48 Ariz. L. Rev. 925, 926 (2006) (“Loyalty for the law’s purposes, unlike Josiah Royce’s, does not mandate an all-embracing “thoroughgoing devotion” to the beneficiary of a fiduciary duty. Its demands neither disregard for the autonomy of an actor subject to fiduciary duties nor require an all-encompassing subordination of the actor’s interests to those of the beneficiary. Instead, within the scope of their relationship, the fiduciary duty of loyalty proscribes self-dealing by the actor and other forms of self-advantaging conduct without the beneficiary’s consent.”); Miller, supra note 3, at 20 (“Notwithstanding the moralizing rhetoric of judges in select fiduciary cases, the law rarely, if ever, requires of fiduciaries the kind of devotion that we expect of persons with moral loyalties.”).

 [31]. Alternatively, the incompatibilist might contend that these requirements are self-defeating when they are backed by the threat of legal sanction. See discussion infra Section III.C (exploring this argument from self-defeat at greater length).

 [32]. See Stephen R. Galoob & Ethan J. Leib, Intentions, Compliance, and Fiduciary Obligations, 20 Legal Theory 106, 129 (2014). Examples of proscriptivism include Matthew Conaglen, The Nature and Function of Fiduciary Loyalty, 121 Law Q. Rev. 452 (2005); Robert Flannigan, The Core Nature of Fiduciary Accountability, 2009 N.Z. L. Rev. 375; Darryn Jensen, Prescription and Proscription in Fiduciary Obligations, 21 King’s L.J. 333 (2010); and Larry E. Ribstein, Fencing Fiduciary Duties, 91 B.U. L. Rev. 899 (2011). Other views adopt a proscriptivist understanding of the fiduciary duty of loyalty, but not of the (arguably) fiduciary duty of care. See generally, e.g., Robert H. Sitkoff, An Economic Theory of Fiduciary Law, in Philosophical Foundations of Fiduciary Law 197 (Andrew S. Gold & Paul B. Miller eds., 2014).

 [33]. See generally, e.g., Matthew Conaglen, Fiduciary Loyalty: Protecting the Due Performance of Non-Fiduciary Duties (2011); Jensen, supra note 32.

 [34]. See, e.g., Ribstein, supra note 32, at 903–06.

 [35]. See, e.g., Flannigan, supra note 32 passim.

 [36]. See, e.g., Conaglen, supra note 33. Darryn Jensen draws a similar distinction between fiduciary obligations and “logically prior” responsibilities, which are not essentially fiduciary. Jensen, supra note 32, at 336.

 [37]. Oliver Wendell Holmes, Jr., The Path of the Law, 10 Harv. L. Rev. 457, 459 (1897).

 [38]. Id. (construing the bad man as caring “only for the material consequences which . . .  knowledge [of the law] enables him to predict”).

 [39]. See Yuval Feldman, The Law of Good People: Challenging States’ Ability to Regulate Human Behavior (2018); Hanoch Dagan, Between Rationality and Benevolence: The Happy Ambivalence of Law and Legal Theory, 62 Ala. L. Rev. 191, 191 (2010); Rebecca Stone, Legal Design for the “Good Man”, 102 Va. L. Rev. 1767, 1767 (2016).

 [40]. See, e.g., Henry N. Butler & Larry E. Ribstein, Opting Out of Fiduciary Duties: A Response to the Anti-Contractarians, 65 Wash. L. Rev. 1 passim (1990); Easterbrook & Fischel, supra note 16, at 424; James Edelman, When Do Fiduciary Duties Arise?, 126 L.Q. Rev. 324, 326 n.173 (2010); John H. Langbein, The Contractarian Basis of the Law of Trusts, 105 Yale L.J. 625, 627 (1995).

 [41]. See Kelli Alces, The Fiduciary Gap, 40 Iowa J. Corp. L. 351, 375 (2015) (“To determine the hypothetical bargain, we must reach a conclusion about what the parties would have agreed given their requirements and expectations of the relationship.”); Easterbrook & Fischel, supra note 16, at 431; Larry Ribstein, Fiduciary Contracts in Unincorporated Firms, 54 Wash. & Lee L. Rev. 537, 541 (1997).

 [42]. See Stephen R. Galoob & Adam Hill, Norms, Attitudes and Compliance, 50 Tulsa L. Rev. 613, 622–26 (2015) (summarizing “reductive behaviorism” that characterizes law-and-economics understanding of social and legal norms).

 [43]. Richard McAdams & Eric B. Rasmusen, Norms and the Law, in Handbook of Law and Economics 1573, 1576 (A. Mitchell Polinsky & Steven Shavell eds., 2007) (defining norms as “behavioral regularities supported at least in part by normative attitudes” and conventions as “behavioral regularities that lack such normative attitudes”); Lewis A. Kornhauser, A World Apart? An Essay on the Autonomy of the Law, 78 B.U. L. Rev. 747, 751 (1998) (“[A]n operative legal norm characterizes the behavior to which legal sanctions will attach.”).

 [44]. See, e.g., Alces, supra note 41, at 353 (“That duty of loyalty to the other party guides courts in deciding what the parties would have agreed about how the fiduciary should behave in unanticipated circumstances.”); Langbein, supra note 40, at 658 (“Loyalty and prudence, the norms of trust fiduciary law, embody the default regime that the parties to the trust deal would choose as the criteria for regulating the trustee’s behavior in these settings in which it is impractical to foresee precise circumstances and to specify more exact terms.”).

 [45]. See Jordan v. Duff & Phelps, 815 F.2d 429, 438 (7th Cir. 1987).

 [46]. Robert Cooter & Bradley J. Freedman, The Fiduciary Relationship: Its Economic Character and Legal Consequences, 66 N.Y.U. L. Rev. 1045, 1046–49 (1991).

 [47]. Sitkoff, supra note 32, at 201.

 [48]. Id. at 202. To assess the fiduciary’s satisfaction of the duty of care solely in “objective” terms is to presuppose behaviorism in several ways. Due to their relative ease of monitoring, behavioral considerations are the most relevant to reducing agency costs. Further, the modal beneficiary can be expected to care ex ante about the fiduciary’s behavior, especially given a prior methodological assumption that the beneficiary is agnostic between the fiduciary’s performing or paying damages. Moreover, given the possibility of contracting around default terms, a contractarian account can deny any inherent problems with presuming an objective standard. If a beneficiary has strong preferences regarding the fiduciary’s cognition, then the parties could bargain for adoption of a more subjective standard.

 [49]. Id. at 202–03.

 [50]. Alces, supra note 41, at 356.

 [51]. See, e.g., Butler & Ribstein, supra note 40, at 71–72.

 [52]. Compare Michael Moore, Causation and Responsibility: An Essay in Law, Morals and Metaphysics 5 (2010) (“It is better to think that ‘cause’ is univocal; it means the same thing in contexts of attributing responsibility as in contexts of explanation: it refers to a natural relation that holds between events or states of affairs. . . . [W]hat criminal law and the law of torts mean by ‘cause’ is what we ordinarily mean by ‘cause’ as we explain the world, viz some kind of natural relation.”), with Jane Stapleton, An ‘Extended But-For’ Test for the Causal Relation in the Law of Obligations, 35 Oxford J. Legal Stud. 697, 697 (2015) (“Legal enquiries . . . differ from philosophical ones in terms of the character of factors [that] law is prepared to recognize as appropriate candidates for being a ‘cause’ of the existence of a particular phenomenon. . . . [L]aw is not constrained by . . . metaphysical commitments and so it seamlessly accommodates as ‘causes’ factors variously characterised as specific events, facts, states of affairs, aspects of conduct events or things, absences, omissions, information, and
reasons . . . .”).

 [53].               Compare John F. Stinneford, The Original Meaning of “Cruel”, 105 Geo. L.J. 441, 450–51 (2016) (positing that the legal definition of “cruel” punishment should be assessed entirely in terms of a punishment’s effects, in contrast to ordinary notion according on which the punisher’s intention is also relevant to assessing cruelty), with Meghan Ryan, Judging Cruelty, 44 U.C. Davis L. Rev. 81, 124 (2010) (positing that legal definition of “cruelty” overlaps with ordinary notion of term, according to which cruel punishment is “exceptionally brutal” and involves “inflicting pain for a purpose other than punishment”).

 [54]. See, e.g., Seana Valentine Shiffrin, The Divergence of Contract and Promise, 120 Harv. L. Rev. 708, 708 (2007); Charles Fried, Contract as Promise: A Theory of Contractual Obligation 1 (1980).

 [55]. See Michael Pratt, Contract: Not Promise, 35 Fla. St. U. L. Rev. 801, 802 (2008) (“[T]he law of contract is not concerned with promises as such and . . . contract and promise do not diverge in a way that calls for justification.”).

 [56]. See Frederick Schauer, Is Law a Technical Term?, 52 San Diego L. Rev. 501, 502–03 (2015).

 [57]. See, e.g., Easterbrook & Fischel, supra note 16, at 429.

 [58]. Our use of the term “cognitivism” differs from Gregory Alexander’s “cognitive theory of fiduciary relationships.” Gregory S. Alexander, A Cognitive Theory of Fiduciary Relationships, 85 Cornell L. Rev. 767 (2000). Alexander’s approach focuses on the likely cognitive biases of judges and fact-finders interpreting fiduciary law. Our interest is with more fundamental questions about the nature of fiduciary relationships and the cognitive requirements of fiduciary loyalty. Nor does our use of the term implicate the metaethical position known as “cognitivism,” which holds that moral statements express genuinely cognitive phenomena and are truth-apt. See Alexander Miller, Contemporary Metaethics: An Introduction 3 (2d ed., 2013). Apologies to Ben Zipursky, who advised us to jettison “cognitivism” because of its connotation in metaethics. We almost certainly did not attribute the requisite significance to his advice in our practical deliberations.

 [59]. For example, some commentators argue that loyalty has an inextricably affective dimension: if someone does not feel specific feelings on behalf of a person or cause, then she is not loyal to that person or cause. See Keller, supra note 6, at 21. Others deny that loyalty has such an affective component, contending that someone can be loyal to a person or cause without feeling specific emotions (or even being disposed to). See John Kleinig, Loyalty, in Stanford Encyclopedia of Philosophy (rev. ed. 2017) (“[F]eelings of loyalty are probably not constitutive of loyalty, even if it is unusual to find loyalty that is affectless.”).

 [60]. On these accounts, someone’s loyalty to a person or cause might require her to act in ways that are otherwise optional, or even permit her to act in ways that are otherwise forbidden. See, e.g., David Owens, Shaping the Normative Landscape 251 (2012) (“[T]he special value of friendship derives in part from the bonds of loyalty that it entails. The involvement of obligation in friendship is part of what makes friendship the great good that it is and the same is true of many other valuable relationships. Thus our interest in friendship evinces a normative interest, an interest in obligation for its own sake.”). Other accounts dispute the normative significance of loyalty either in part (for example, by denying that loyalty can generate permissions) or in whole (for example, by denying that loyalty can generate reasons or obligations). See, e.g., Keller, supra note 6, at 144–45. A similar divide arises in debates over the normative significance of patriotism: for example, whether loyalty provides a basis for political obligation.

 [61]. See Ethan J. Leib & Stephen R. Galoob, Fiduciary Political Theory: A Critique, 125 Yale L.J. 1820, 1828 (2016); Stephen R. Galoob & Ethan J. Leib, The Core of Fiduciary Political Theory, in Research Handbook of Fiduciary Law 401, 407–12 (2018); Ethan J. Leib & Stephen R. Galoob, Fiduciary Principles and Public Office, in Oxford Handbook of Fiduciary Law, supra note 24, at 20–25.

 [62]. And perhaps, enraged by the rumor that Othello had conducted an affair with Emilia, Iago’s wife. William Shakespeare, Othello act 1, sc. 3, 1. 380–83 (Burton Raffel ed., Yale Univ. Press 2005) (1622). (“[I]t is thought abroad that ‘twixt my sheets/He’s done my office. I know not if’t be true/But I, for mere suspicion in that kind/Will do as if for surety.”). One disturbing interpretation of Iago’s shifting rationalizations is that they reflect “motive-hunting,” an “attempt to bring intellectual speculation to bear on a malignity that [Iago] does not really understand himself.” Laurence Lerner, The Machiavel and the Moor, 9 Essays in Crit. 339, 342–44 (1959).

 [63]. Shakespeare, supra note 62, act 2, sc. 3, 1. 325–26.

 [64]. For example, Iago leads Othello to notice the most compelling evidence of Desdemona’s (supposed) treachery in the following passage, during which he feigns to argue for her fidelity:

Iago: Her honour is an essence that’s not seen:

They have it very oft that have it not;

But for the handkerchiefs—

Othello: By heaven, I would most gladly have forgot it:

Thou said’st O, it comes o’er my memory,

As doth the raven o’er the infected house,

Boding to all he had my handkerchief.

Shakespeare, supra note 62, act 4, sc. 1, 1. 19–22. The “he” in the last sentence refers to Cassio, who has Desdemona’s handkerchief only because Iago had given it to him.

 [65]. See Keller, supra note 6, at 2–3.

Just because someone deliberatively follows a principled pattern of behavior, or is committed—perhaps fiercely—to a cause, does not mean that she is loyal. . . . [T]he facts about how a person acts are not in themselves enough to tell us whether, or to what, she is loyal. If I reliably keep my promises to you, then that might be because you are someone to whom I am loyal. Or, it might be because I believe that people should always keep their promises, . . . or because I promised your father that if I made you any promises I would keep them.

Id.

 [66]. See Philip Pettit, The Robust Demands of the Good: Ethics with Attachment, Virtue, and Respect 84–85 (2015); Henry S. Richardson, Moral Entanglements: The Ancillary-Care Obligations of Medical Researchers 65, 85, 95–96 (2012) (arguing that moral duties can morph and arise incidentally based on new information generated from types of interactions); Nicholas Southwood, Democracy as a Modally Demanding Value, 49 Noûs 504, 505 (2015).

 [67]. William P. MacNeil, Machiavellian Fantasy and the Game of Laws, 57 Critical Q. 34, 41 (2015) (For Littlefinger, “[s]urvival entails being ready, willing and able to advance one’s cause through any and all means, including . . . taking the life of his not-so-fair lady wife, . . . all the while further plotting to exterminate his pathetic little stepson, Lord Robert, the heir to the Eyrie and the guardianship of the West.”).

 [68]. Game of Thrones: The Climb (HBO television broadcast May 5, 2013).

 [69]. Restatement (Second) of Torts § 874 (Am. Law Inst. 1979); see also Longaker v. Evans, 32 S.W.3d 725, 733 (Tex. App. 2000).

 [70]. For example, the “prejudice” requirements in “ineffective assistance of counsel” or collateral relief contexts do not indicate that non-prejudicial betrayals are not themselves juridical or legal betrayals.

 [71]. See, e.g., Burrow v. Arce, 997 S.W.2d 229, 238 (Tex. 1999) (“It is the agent’s disloyalty, not any resulting harm, that violates the fiduciary relationship . . . .”).

 [72]. See, e.g., Kinzbach Tool Co. v. Corbett-Wallace Corp., 160 S.W.2d 509, 514 (Tex. 1942); Restatement (Second) of Agency § 469 (Am. Law Inst. 1958).

 [73]. See Restatement (Second) of Trusts § 243 cmt. a (Am. Law Inst. 1959); see also In re Bradish’s Estate, 8 Pa. D. 38, 42 (Orphans’ Ct. 1898) (“It is no answer to say, that after all there may be no loss upon the judgments; this is little more than a plea, that a negligent trustee should be allowed a further opportunity to take risks, because the inexcusable risks he has already taken have not yet fallen out against him.”).

 [74]. In re Friedman, 23 P.3d 620, 622–23 (Alaska 2001).

 [75]. Id. at 624.

 [76]. Id. at 631.

 [77]. It does not follow that issues of harm are irrelevant to the assessment of Friedman’s misconduct. Then-applicable disciplinary rules (which tracked the American Bar Association’s Standards for Imposing Lawyer Sanctions) directed the court to assess any injury caused by Friedman’s misconduct, and the Friedman Court found that Friedman’s actions inflicted “actual injury to the public, because ‘the public suffers injury whenever a lawyer fails to maintain personal integrity by improperly handling funds held in trust.’” Id. However, the issue of harm was relevant to the secondary question of the kind discipline (suspension or disbarment) that Friedman should have been subjected to for having violated his fiduciary duties to clients. It was not relevant to the primary question of whether Friedman violated his fiduciary duties in the first place.

 [78]. In re Nine Sys. Corp. S’holders Litig., Consol. C.A. No. 3940-VCN, 2014 Del. Ch. LEXIS 171 (Sept. 4, 2014), aff’d sub nom. Fuchs v. Wren Holdings, LLC, 129 A.3d 882 (Del. 2015).

 [79].               Id. at *5, *25.

 [80]. Id. at *13, *21, *42.

 [81]. Id. at *140.

 [82]. Although the Nine Systems court did not award disgorgement or rescission on account of the finding of a fair price, it granted attorney’s fees to plaintiffs for successfully establishing that defendants had breached their fiduciary duties. See id. at *160. For another case—this time a trust law example—in which a court found a breach of fiduciary duty in the absence of damage to the plaintiff and then remediated through awarding of attorney’s fees, see In re Wilson, 930 N.E.2d 646, 652 (Ind. Ct. App. 2010).

 [83]. See In re Nine Sys., 2014 Del. Ch. LEXIS 171, at *145–46. A companion opinion in Ross Holding & Management Co. v. Advance Realty Group, C.A. No. 4113-VCN, 2014 Del. Ch. LEXIS 173, at *60 (Sept. 4, 2014) similarly reinforces that fair dealing and fair process is a part of fiduciary obligation, even when the relevant transaction produces a price that loyal fiduciaries would have reached.

 [84]. See Conaglen, supra note 33, at 70–71, 74.

 [85]. This approach would require the contractarian to abandon the methodological commitment to behaviorism, which (as noted above) many legal economists posit as a core tenet of the contractarian position. Or it would require thinking about process as just another kind of behavior.

 [86]. See, e.g., Robert Flannigan, The Strict Character of Fiduciary Liability, 2006 N.Z. L. Rev. 209 passim (2006); Penner, supra note 29, at 175; Smith, supra note 29 passim. In some jurisdictions, the character of fiduciary liability is less-than-strict, in that it is subject to exceptions (for example, “entire fairness” review under Delaware corporate law). Yet even these types of exceptions are consistent with the “strict” character of fiduciary liability in that questions of neither liability nor exception turn on aspects of the fiduciary’s cognition.

 [87]. An example of this dynamic in the private law might be found in Hollywood Silver Fox Farm, Ltd. v. Emmett [1936] 2 KB 468 (Eng.), in which the court found that although a neighbor had a right to shoot a gun off on his own property, it would be actionable if the neighbor intended this shooting to disturb the breeding activities of the fox farm next door.

 [88]. In re Friedman, 23 P.3d 620, 630 n.37 (Alaska 2001).

 [89]. Id. at 625–26, 630, 632.

 [90]. Id. at 632.

 [91]. Id. at 630, 633.

 [92].               Id. at 632–33.

 [93].               See Nancy J. Moore, Mens Rea Standards in Lawyer Disciplinary Codes, 23 Geo. J. Legal Ethics 1, 20 (2010). Although Moore’s conclusions about the presumption against strict liability in these contexts seem motivated by policy considerations pertaining to the consequences of lawyer discipline on the lawyer, one might see these cognitive requirements as themselves flowing from fiduciary obligation, thus requiring no explicit codification. See Charles Wolfram, A Cautionary Tale: Fiduciary Breach as Legal Malpractice, 34 Hofstra L. Rev. 689 (2006). But see Stephen Gillers, Regulation of Lawyers: Problems of Law and Ethics, 186–87 (7th ed. 2005) (finding many fiduciary obligations to trigger “absolute liability” with no mens rea requirement).

 [94]. 11 U.S.C. § 523(a)(4) (2012).

 [95]. Bullock v. BankChampaign, N.A., 569 U.S. 267, 269 (2013).

 [96]. Id.

 [97]. Id. at 273–74.

 [98]. E.g., Restatement (Second) of Trusts, § 243, cmt. d (Am. Law Inst. 1959) (compensation ordinarily denied to trustee who misappropriates trust property or “intentionally or negligently mismanages the whole trust”); DeMott, supra note 30, at 929 (“[A] fiduciary may forfeit commissions or other compensation paid or otherwise due during a period of disloyal service, although, at least in the agency context, courts qualify the availability of forfeiture by requiring that the breach have had a deliberate character, often that it have been ‘wilful’ or ‘egregious.’”).

 [99]. For example, the UK’s law controlling trustees, Trustee Act 2000 § 31(1) (UK), protects trustees, indemnifying them for expenses—even if those expenses are unauthorized—when the trustee acts in “good faith.” Evidence regarding the fiduciary’s deliberation therefore even provides immunity against liability in certain circumstances. See also Trustee Act 1925 § 15 (Gr. Brit.); In re Smith’s Estate [1937] Ch. 636 (Eng.); Vyse v. Foster (1872) LR 8 Ch. App. 309, 336–67 (Eng.). This feature of UK trustee law is discussed in Ho, supra note 12, at 17–18.

 [100]. The proscriptivist might respond by defining the dynamics identified in P2 as deviant or, alternatively, by contending that these are “nominate,” rather than distinctively “fiduciary,” aspects of fiduciary duties. This strategy is not a promising way to identify what is distinctive about fiduciary duties, since it pre-commits to defining the distinctiveness of fiduciary responsibilities in terms of their unique aspects. By conflating distinctiveness and uniqueness, this strategy rules out in advance the possibility that what is distinctive about fiduciary responsibilities is a unique combination of “nominate” duties, none of which are themselves unique to fiduciary relationships. To analogize, other hockey players might be able to pass, shoot, skate, evade checks, score goals, and set up assists as well as Wayne Gretzky could. Gretzky, however, was a unique hockey player because of his combination of these (non-unique) talents. Yet if distinctiveness were defined entirely in terms of the uniqueness of specific components, as many proscriptivists suggest we should do to establish the fiduciarity of a duty, then one could reach the risible conclusion that Wayne Gretzky was not a distinctive hockey player.

 [101]. On the role of fault in contract law, see Omri Ben-Shahar & Ariel Porat, Foreword: Fault in American Contract Law, 107 Mich. L. Rev. 1341, 1346–47 (2009).

 [102].               Neither of these incompatibilities is based on a conceptual commitment. One possible version of proscriptivism might see a fiduciary’s mental state as relevant to the formulation of how to provide relief for violations of the no-profit and no-conflict rules, either of which could turn on cognitive considerations. Likewise, one could imagine a version of contractarianism on which mental states mattered to performance—that is, in which both the behavior and cognition of a counterparty were modeled as relevant to a hypothetical bargainer. However, most theorists who embrace proscriptivism and contractarianism would likely reject these modifications, since advocates of these views see their implicit behaviorism as a strength, rather than as a weakness.

 [103].               See R.A. Duff, Criminal Attempts 133–34 (1997); Gideon Yaffe, Attempts 21–25 (2010).

 [104]. Restatement (Third) of Law Governing Lawyers § 5(2) (Am. Law Inst. 2000) (emphasis added). The comments explain that the same standards for assessing the charge of attempt in criminal law also apply in the lawyer disciplinary context—namely, whether the lawyer had the requisite intent to violate the provision and took a substantial step in the course of conduct that was planned to culminate in the commission of the offense (and that, on the whole, strongly corroborates this purpose). Id. § 5(2) cmt. e.

 [105]. Model Rules of Prof’l Conduct r. 8.4(a) (Am. Bar Ass’n 1983); see also Financial Gen. Bankshares v. Metzger, 523 F. Supp. 744, 746, 772 (D.D.C. 1981) (stating attorney’s unsuccessful attempts to take over company constituted violations of fiduciary duty); Cal Pak Delivery, Inc. v. United Parcel Serv., Inc., 60 Cal. Rptr. 2d 207, 212–13 (Ct. App. 1997) (stating attorney’s attempted and unsuccessful betrayal of client interests by selling out client for a personal payment was a breach of fiduciary duty).

 [106]. Union Miniere, S.A. v. Parday Corp., 521 N.E.2d 700, 703 (Ind. Ct. App. 1988) (stating an agent’s unsuccessful attempt to harm the principal was a breach of fiduciary duty); Town & Country House & Homes Serv., Inc. v. Evans, 189 A.2d 390, 392–94 (Conn. 1963) (stating an agent’s solicitation of principal’s clients is a breach of fiduciary obligation); Town & Country House & Home Serv., Inc. v. Newbery, 147 N.E.2d 724, 726 (N.Y. 1958) (same). But see Restatement (Third) of Agency § 8.04 (Am. Law Inst. 2006) (disallowing solicitation of clients but providing narrow window for “preparation” to compete with the principal).

 [107]. E.g., Dower v. Mosser Indus., Inc., 648 F.2d 183, 188–89 (3rd Cir. 1981) (“A merger intended solely to ‘freeze out’ or ‘cash out’ minority equity holders from their positions may be enjoined as an attempted breach of that fiduciary duty”); Orchard v. Covelli, 590 F. Supp. 1548, 1557 (W.D. Pa. 1984) (“[A]ny attempt to ‘squeeze out’ a minority shareholder must be viewed as a breach of . . . fiduciary duty.”).

 [108]. Navigant Consulting, Inc. v. Wilkinson, 508 F.3d 277, 284–85 (5th Cir. 2010) (finding employees who attempted to sell their employers business can be found to have violated “the most basic norms”).

 [109]. E.g., Item Software (UK) Ltd. v. Fassihi [2004] EWCA (Civ) 1244, para 41 (Eng.) (holding that unsuccessful attempt to usurp a business opportunity violated director’s duty to confess, which was entailed by the “fundamental” fiduciary duty to “act in what he in good faith considers to be the best interests of his company”); Shepherds Invs. Ltd. v. Walters [2006] EWHC 836 (Ch).

 [110]. The different standards for determining remoteness are reflected in different ways of specifying the actus reus required for attempt liability. See, e.g., Duff, supra note 103, at 33–75; Yaffe, supra note 103, at 255–83.

 [111].               E.g., Mattern & Assocs., L.L.C. v. Seidel, 678 F. Supp. 2d 256, 268 n.10 (D. Del. 2010) (recognizing an action for attempted breach of fiduciary duty in a “preparations to compete” case); Shepherds Invs. Ltd. v. Walters [2006] EWHC 836 (Ch) (suggesting that preparatory steps to breach fiduciary obligation is itself a betrayal). But see Restatement (Third) of Agency § 8.04 (Am. Law Inst. 2006) (suggesting that some room for permissible preparation is part of the law of agency).

 [112]. To be sure, cognitivists might be concerned with imposing too onerous standards on fiduciaries. The concern might arise out of fairness (the notion that controlling one’s own motivations and deliberation, if not impossible, is far more difficult than controlling one’s behavior) or as a potential defeater to the ostensible benefits of discretion. To allay this concern, the cognitivist might draw a further parallel to the criminal law distinction between liability for completed and attempted crimes. For result or status crimes, liability can be grounded on satisfying the behavioral and/or result elements (for example, by bringing about a prohibited result or behaving in a prohibited way), even though the defendant does not have a highly culpable mental state. Yet a highly culpable mental state is required in order to establish liability for an attempt of the same crime. For example, someone who has a mental state of negligence or recklessness can be liable for battery, but liability for attempted battery requires a purposive mental state. On this logic, the cognitivist might concede that liability for a successful breach of fiduciary duty can be based on low-grade mental states like negligence or even without a mens rea, while still maintaining that a higher-order mental state is required to ground liability for inchoate violations of fiduciary duty.

 [113]. Conaglen, supra note 33, at 459–60.

 [114]. Judith Viorst, Alexander and the Terrible, Horrible, No Good, Very Bad Day (1972).

 [115].               To be sure, an attempted breach of contract might give a promisee a justification for demanding an assurance from her promisor—but it would not be itself treated as a breach of the underlying contractual commitment. That said, some courts might see an announced intent to breach or anticipatory breach as a kind of “attempted breach of contract.” See, e.g., First Nat’l Bank of Louisville v. Cont’l Ill. Nat’l Bank & Tr. Co. of Chi., 933 F.2d 466, 469 (7th Cir. 1991).

 [116]. Lionel Smith concludes that fiduciary duties require the fiduciary to act with an appropriate motivation, even though (he argues) this requirement is not and should not be legally enforceable. See Lionel Smith, The Motive, Not the Deed, in Rationalizing Property, Equity, and Trusts: Essays in Honour of Edward Burn 53 passim (J. Getzler ed., 2003). But see Smith, supra note 4, at 152; Strine et al., supra note 5, at 633 (“[Under Delaware law] good faith has long been used as the key element in defining the state of mind that must motivate a loyal fiduciary.”).

 [117]. See, for example, Council on American-Islamic Relations Action Network, Inc. v. Gaubatz, 31 F. Supp. 3d 237, 259–60 (D.D.C. 2014) and the discussion of this case in Deborah A. DeMott, The Poseur as Agent, in Agency Law in Commercial Practice 35, 44–45 (Danny Busch et al. eds., 2016).

 [118]. Ex parte Endo, 323 U.S. 283, 302 (1944).

 [119]. See MiMedx Grp., Inc. v. Fox, 2017 U.S. Dist. LEXIS 121801, at *15 (N.D. Ill. Aug. 2, 2017) (“A corporate officer who sabotages the company may do so for no personal gain, but nevertheless breach a fiduciary duty owed to firm.”); Gantler v. Stephens, 965 A.2d 695, 709 (Del. 2008) (holding that allegations of sabotage is a violation of fiduciary duty).

 [120]. See Ho, supra note 12, at 50; Lusina Ho & Pey Woan Lee, A Director’s Duty to Confess: A Matter of Good Faith?, 66 Cambridge L.J. 347, 360–61 (2007).

 [121]. Commonwealth v. Washington, 880 A.2d 536 (Pa. 2005).

 [122]. Id. at 544.

 [123]. In support of this conclusion, Washington noted a letter to a potential defense psychiatric expert in which trial counsel stated that Washington “may epitomize the banality of evil.” Id. at 541.

 [124]. Id. at 541–42.

 [125].               Id. at 544.

 [126].               Id. at 541.

 [127].               Id. at 542.

 [128].               Id. at 543. In support of this point, the Washington Court cited Fisher v. Gibson, 282 F.3d 1283, 1294–98, 1307 (10th Cir. 2002) (finding a cognizable, but not dispositive, Sixth Amendment claim where attorney failed to investigate potential alibis or relevant information, did nearly no preparation for case, and admitted afterward in affidavit that he abhorred homosexuals—like his client—and believed that these feelings affected his representation) and Frazer v. United States, 18 F.3d 778, 780 (9th Cir. 1994) (cognizable, but not dispositive, ineffective assistance claim where attorney called client “a stupid . . . son of a bitch” and opined that client should be imprisoned for life).

 [129].               Washington, 880 A.2d at 545; see also Rickman v. Bell, 131 F.3d 1150, 1157 (6th Cir. 1997) (lawyer violated the duty of loyalty through “a total failure to actively advocate his client’s cause” and “repeated expressions of contempt for his client for his alleged actions”).

 [130]. Washington, 880 A.2d at 545.

 [131]. Stone v. Ritter, 911 A.2d 362, 369–70 (Del. 2006).

 [132]. Sitkoff, supra note 32, at 202. Under this standard, the duty of good faith, the duty of honesty, the duty of fidelity, the duty to confess, the “fraud on a power” doctrine, and the “proper purposes” doctrine might all be classified as implementing rules.

 [133]. Item Software (UK) Ltd. V. Fassihi [2004] EWCA (Civ) 1244 (Eng.).

 [134]. Ho & Lee, supra note 120, at 350.

 [135].               Brandeaux Advisers (UK) Ltd. v Chadwick [2010] EWHC (QB) 3241 [47]; P & V Indus. Pty., Ltd. v. Porto, [2006] V.S.C. 131 [23], [42].

 [136]. Fassihi, [2004] EWCA (Civ), at [41].

 [137]. See Harry G. Frankfurt, Freedom of Will and the Concept of a Person, 68 J. Phil. 5, 11 (1971).

 [138].               Restatement (Third) of Agency § 8.01 (Am. Law Inst. 2006).

 [139].               Id. § 8.01 cmt. b.

 [140].               Id. (citing Greenwood v. Koven, 880 F. Supp. 186 (S.D.N.Y. 1995), which hypothesizes that an auctioneer’s decision about whether to rescind a sale made by a coin-toss was paradigmatic case of an action taken in bad faith). Of course, many fiduciary duties arise in relationships that also involve contractual duties. Wantonness might violate not only a fiduciary duty, but also specific contractual duties (like the duty of good faith). See Daniel Markovits, Sharing Ex Ante and Sharing Ex Post: The Non-Contractual Basis of Fiduciary Relations, in Philosophical Foundations of Fiduciary Law 209, 209–10 (2014). Yet in such relationships, fiduciary loyalty is not superfluous to a contractual duty of good faith; rather, fiduciary loyalty picks out a range of behaviors and motivations that would not (obviously or directly) be condemned by contract law alone. We discuss this issue more extensively in Galoob & Leib, supra note 32, at 128–29.

 [141]. Strine et al., supra note 5, at 633.

 [142]. Del. Code Ann. tit. 8, § 102(b)(7) (2001).

 [143]. Stone v. Ritter, 911 A.2d 362, 369–70 (Del. 2006).

 [144]. Joseph K. Leahy, A Decade After Disney: A Primer on Good and Bad Faith, 83 U. Cin. L. Rev. 859, 863 (2015) (emphasis removed).

 [145]. See In re Walt Disney Co. Deriv. Litig., 907 A.2d 693, 697, 760 (Del. Ch. 2005) (Disney III).

 [146]. See Claire A. Hill & Brett H. McDonnell, Stone v. Ritter and the Expanding Duty of Loyalty, 76 Fordham L. Rev. 1769, 1781 (2007) (emphasizing that the claim in Disney was not that the board was conflicted but that it was “simply rubber-stamping” Eisner’s proposals).

 [147]. See In re Walt Disney Co. Deriv. Litig., 906 A.2d 27, 64 (Del. 2006) (Disney IV); see also Christopher M. Bruner, Is the Corporate Director’s Duty of Care a “Fiduciary” Duty? Does It Mattter?, 48 Wake Forest L. Rev. 1027, 1047 (2013) (“[G]ood faith is fundamentally about state of mind—the quality of a director’s intentions vis-à-vis the company.”).

 [148]. Leahy, supra note 144, at 867. Fiduciary loyalty also prohibits a director from acting for “greed, hatred, lust, envy, revenge . . . shame or pride.” Disney III, 907 A.2d at 754 (internal quotations and citations omitted).

 [149]. In re Walt Disney Co. Deriv. Litig., 825 A.2d 275, 289 (Del. Ch. 2003) (Disney II).

 [150]. Id.

 [151]. For the purposes of our argument, it does not matter how much “good faith” or “bad faith” in corporate law is meant to be evaluated objectively or subjectively. Compare Strine et al., supra note 4, at 655 (arguing for a subjective understanding), with Melvin E. Eisenberg, The Duty of Good Faith in Corporate Law, 31 Del. J. Corp. L. 1 (2006) (advocating a more objective approach). Nor does it matter whether the duty of good faith should be subsumed under the duty of loyalty. Nor does it matter whether the duty of loyalty is disclaimable or how infrequently Delaware courts have upheld claims of bad faith. See Leahy, supra note 144, at 875–76, 880–82. Rather, Delaware courts’ explanations of the standards for fiduciary faithfulness indicate, consistent with P4, that some motivational structures are inconsistent with fiduciary obligation, which is enough to show that P4 is part of fiduciary law.

 [152]. See Ian Ayres & Gregory Klass, Insincere Promises: The Law of Misrepresented Intent 96–97 (2005) (contending that “blank promises,” in which one party intends at the time of agreement not to act in the way that he or she promises, lack mutual value).

 [153]. See, e.g., Robert Flannigan, The Economics of Fiduciary Accountability, 32 Del. J. Corp. L. 393, 397 n.15 (2007) (“[W]hat may appear to be low effort may actually be the result of sabotage by another agent (e.g., at a bottleneck point) attempting to mask relative effort. In that instance, the sabotaging agent commits the fiduciary breach.”). For the reasons described above, Flannigan’s conclusion cannot be reconciled with his embrace of proscriptivism.

 [154]. Perhaps, the proscriptivist might contend, the saboteur and the wanton violate a duty to the beneficiary, but not a fiduciary duty. On an alternative reading of the proscriptivist position, one might see the saboteur and the wanton as violating (nominate) duties of care, rather than duties of fiduciary loyalty. However, this solution does not resolve the normative or descriptive gaps in proscriptivism. For one thing, it is not clear that the saboteur or the wanton violates an objective duty of care; in the Fassihi case, for example, Fassihi’s advice would satisfy an objective duty of care, and yet it still violated a fiduciary duty owed to the entity.

 [155]. Elsewhere, we have argued that sensitivity is a component of ordinary loyalty as well. See Leib & Galoob, supra note 61, at 1843.

 [156]. Robert Nozick, Philosophical Explanations 179 (1981); Fred I. Dretske, Epistemic Operators, 67 J. Phil. 1007 passim (1970). See generally Sherrilyn Roush, Tracking Truth: Knowledge, Evidence, and Science (2005); Keith DeRose, Insensitivity Is Back, Baby!, 24 Phil. Perspectives 161 (2010); David Enoch & Talia Fisher, Sense and “Sensitivity”: Epistemic and Instrumental Approaches to Statistical Evidence, 67 Stan. L. Rev. 557 (2015).

 [157]. For example, the Alabama Supreme Court found that a trustee bank violated its duty of loyalty based on “insensitivity . . . in the performance of its duty of loyalty to the trust’s beneficiaries” by retaining the trustee bank’s own stock in the trust. First Ala. Bank of Huntsville, N.A. v. Spragins, 515 So. 2d 962, 964 (Ala. 1987). Similarly, in cases involving professional discipline of a lawyer for violation of applicable rules of professional conduct, courts often distinguish the lawyer’s insensitivity to fiduciary duty as a basis for finding a breach of fiduciary duty. See, e.g., In re Evans, 578 A.2d 1141, 1151 (D.C. 1990); In re Lupo, 851 N.E.2d 404, 414 (Mass. 2006); Chiles v. Robertson, 767 P.2d 903, 926 (Or. Ct. App. 1989).

 [158]. E.g., Robert C. Clark, Agency Costs Versus Fiduciary Duties, in Principals and Agents 55, 75–76 (John W. Pratt & Richard J. Zeckhauser eds., 1985). See generally Markovits, supra note 140.

 [159]. See Leib & Galoob, supra note 61, at 1843–44.

 [160]. See, e.g., Miller v. McDonald (In re World Health Alts. Inc.), 385 B.R. 586, 591–92 (Bankr. D. Del. 2008) (finding a failure to monitor claim cognizable as a breach of fiduciary duty).

 [161]. In re Caremark Int’l, 698 A.2d 959, 967, 970 (Del. Ch. 1996). A critic might note the extreme burdens to grounding liability for corporate directors solely on Caremark factors. Id. at 967 (“The theory here advanced is possibly the most difficult theory in corporation law upon which a plaintiff might hope to win a judgment.”). Our main contention here, however, is that Caremark’s monitoring duty is a component of the duty of loyalty, regardless of whether or how the duty to monitor figures into litigation strategies or standards of review.

 [162]. See FDIC ex rel. Wheatland Bank v. Spangler, 836 F. Supp.2d 778, 792 (N.D. Ill. 2011) (refusing to allow the business judgment rule to insulate defendants who “disregarded regulatory warnings of unsafe lending practices and monthly reports reflecting dangerous loan concentration and excessive growth . . . and took no action to reform underwriting practices in response to criticism.”).

 [163]. An example here might be possible conflicts of interest in trust law—not actual conflicts but the mere “possibility of conflict.” See Penner, supra note 29, at 168 (citing Keech v. Sandford, 2 Eq. Cas. Ab. 741, Cas. Ch. 61, 25 Eng. Rep. 223 (1726)). A similar set of considerations apply to positional conflicts of interest of lawyers, which “occur[] when a law firm adopts a legal position for one client seeking a particular legal result that is directly contrary to the position taken on behalf of another present or former client, seeking an opposite legal result, in a completely unrelated matter.” John S. Dzienkowski, Positional Conflicts of Interest, 71 Tex. L. Rev. 457, 460 (1993). Positional conflicts are problematic when

there is a significant risk that a lawyer’s action on behalf of one client will materially limit the lawyer’s effectiveness in representing another client in a different case; for example, when a decision favoring one client will create a precedent likely to seriously weaken the position taken on behalf of the other client. . . . If there is significant risk of material limitation, then absent informed consent of the affected clients, the lawyer must refuse one of the representations or withdraw from one or both matters.

Model Rules of Prof’l Conduct r. 1.7 cmt 24 (Am. Bar Ass’n 1983). In other words, a positional conflict can violate a lawyer’s duty of loyalty on the grounds that it might harm the client’s interest or inhibit the lawyer’s representation of the client, regardless of whether the conflict has or actually does have these effects.

 [164]. United States v. Miss. Valley Generating Co., 364 U.S. 520, 549–50 (1961) (emphasis added).

 [165]. See, e.g., Dana A. Remus, Reconstructing Professionalism, 51 Ga. L. Rev. 807, 856 (2017) (“The value of a lawyer’s loyalty is often expressed from the perspective of the client: [It]  . . . ensures that a lawyer will not abandon a client as soon as a more lucrative opportunity arises.”). The possibility that a flimsy commitment can constitute a breach of fiduciary duty does not imply that fiduciary loyalty is inevitably “maximal,” in the sense that George Fletcher uses that term, see George P. Fletcher, Loyalty: An Essay on the Morality of Relationships 61–77 (1995), or what Matthew Harding calls a “thick” form of interpersonal trust, see Harding, supra note 4, at 83. Allowing some form of contingency in the fiduciary’s commitment to the beneficiary is necessary to make sense of the possibility that, for example, employees might have fiduciary duties toward employers in an age of at-will employment. See Matthew I. Bodie, Employment as Fiduciary Relationship, 105 Geo. L.J. 819, 847–54 (2016); Marian Riedy & Kim Sperduto, At-Will Fiduciaries: The Anomalies of a Duty of Loyalty in the Twenty-First Century, 93 Neb. L. Rev. 267, 272–79 (2014). Our point is not that the requisite sturdiness is uniform across fiduciary relationships; rather, in contexts where fiduciary duties arise, some kinds of flimsy commitments would be sufficient to constitute a violation of fiduciary duty. For example, even among at-will employees, an untriggered, secret plan to betray would constitute a breach of fiduciary loyalty. Such a violation of fiduciary loyalty might be appraised either in terms of planned treachery (which runs afoul of P3) or in terms of the flimsiness of commitment that is evinced by such a plan (which runs afoul of P5).

 [166]. See generally John Leubsdorf, Conflicts of Interest: Slicing the Hot Potato Doctrine, 48 San Diego L. Rev. 251 (2011).

 [167]. Model Rules of Prof’l Conduct r. 1.7(a)(1) (Am. Bar Ass’n 1983).

 [168]. Id. at r. 1.9(a).

 [169]. Id. at r. 1.16(b)(1).

 [170]. Picker Int’l., Inc. v. Varian Assoc., Inc., 670 F. Supp. 1363, 1365 (N.D. Ohio 1987); see also Ex parte AmSouth Bank, N.A., 589 So. 2d 715, 721–22 (Ala. 1991) (“[A] law firm should not be allowed to abandon its absolute duty of loyalty to one of its clients so that it can benefit from a conflict of interest that it has created.”).

 [171]. ValuePart, Inc. v. Clements, No. 06 C 2709, 2006 U.S. Dist. LEXIS 98167, at *5 (N.D. Ill. Aug. 2, 2006).

 [172]. Arthur R. Laby, Selling Advice and Creating Expectations: Why Brokers Should Be Fiduciaries, 87 Wash. L. Rev. 707, 710 (2012).

 [173]. Id. at 709–10.

 [174]. Several U.S. Supreme Court opinions have found that the Investment Advisers Act of 1940 creates fiduciary duty for advisers. See, e.g., Santa Fe Indus. v. Green, 430 U.S. 462, 478–80 (1977); SEC v. Capital Gains Research Bureau, Inc., 375 U.S. 180, 181–82 (1963). Although this legal question is settled, there is a dispute among scholars of fiduciary law about whether fiduciary duties should apply to financial advisers. For example, Paul Miller, one of the preeminent theorists of fiduciary law, contends that financial advisers do not have the kind of discretionary authority that grounds fiduciary duties. On Miller’s “fiduciary powers” view, advisers “are not fiduciaries . . . by virtue of giving advice. Instead, they are fiduciaries only where they exercise discretionary power over the practical interests of their clients. In such cases, provision of advice is incidental to the exercise of discretionary power.” Paul Miller, The Fiduciary Relationship, in Philosophical Foundations of Fiduciary Law, supra note 3, at 63, 84.

 [175]. See, e.g., United States v. Skelly, 442 F.3d 94, 98 (2d Cir. 2006); United States v. Szur, 289 F.3d 200, 211 (2d Cir. 2002); Associated Randall Bank v. Griffin, Kubik, Stephens & Thompson, Inc., 3 F.3d 208, 212 (7th Cir. 1993); MidAmerica Fed. Sav. & Loan Ass’n v. Shearson/Am. Express Inc., 886 F.2d 1249, 1257 (10th Cir. 1989).

 [176]. 29 C.F.R. § 2510.3–.21 (218). On May 22, 2017, the Department of Labor (“DOL”) issued Field Assistance Bulletin No. 2017-02, providing temporary non-enforcement of the new fiduciary rules.

 [177]. See generally The DOL Fiduciary Rule Explained,  Investopedia (Oct. 16, 2018), https://www.investopedia.com/updates/dol-fiduciary-rule  (explaining the history and status of the DOL’s rule). The courts have not been univocal on whether the rule is valid. Compare Market Synergy Grp., Inc. v. U.S. Dep’t of Labor, 885 F.3d 676, 685–86 (10th Cir. 2018) (upholding the rule), with Chamber of Commerce v. U.S. Dep’t of Labor, 885 F.3d 360, 363 (5th Cir. 2018) (invalidating the rule).

 [178]. E.g., SEC v. Blavin, 760 F.2d 706, 711–12 (6th Cir. 1985) (holding that advisers to a duty of investigation about the advice they peddle).

 [179]. See, e.g., Scott MacKillop, First Ascent Asset Management, LLC, Can a Robot Be a Fiduciary? 3–5 (2016), http://firstascentam.com/wp-content/uploads/2016/12/FA-Position-Paper-Can-Robot-Be-Fiduciary_FINAL.pdf; Nicole G. Iannarone, Computer as Confidant: Digital Investment Advice and the Fiduciary Standard, Chi.-Kent L. Rev. 141, 156–60 (2018); Megan Ji, Note, Are Robots Good Fiduciaries? Regulating Robo-Advisers Under the Investment Advisers Act of 1940, 117 Colum. L. Rev. 1543, 1563–71 (2018); Melanie L. Fein, Robo-Advisers: A Closer Look, SSRN (June 30, 2015), https://ssrn.com/abstract=2658701; Robert Powell, Can a Robo Adviser Really Act in Your Best Interest?, MarketWatch (Apr. 28, 2016),  http://www.marketwatch.com/story/can-a-robo-adviser-really-act-in-your-best-interest-2016-04-28.

 [180]. Tom Baker & Benedict Dellaert, Regulating Robo Advisers Across the Financial Services Industry, 103 Iowa L. Rev. 713, 740 (2018).

 [181]. W. Nicholson Price, Black-Box Medicine, 28 Harv. J.L. & Tech. 419, 421 (2015).

 [182]. Id.

 [183]. To wit, since we might want the programmer writing the code for the “black box”—or the salesperson selling the code—to have a proper commitment to the interests of the end-user, there still may be sense in applying fiduciary obligations upstream from the “black box.” We also cannot rule out the possibility that the “black box” itself could engage in undetectable betrayal.

 [184]. A non-cognitivist interpretation of fiduciary loyalty would be much easier to administer than one that inquired about the fiduciary’s deliberation and/or motivation, not to mention the robustness of her commitment to the beneficiary. Furthermore, a purely behavioral definition of loyalty would provide the kind of clear guidelines that appeal to Holmes’s “bad man.” Therefore, to the extent that fiduciaries in a legal domain more closely resemble the Holmesian bad man, a non-cognitivist interpretation of fiduciary loyalty might be justified in terms of efficiency (for example, as reducing transaction, information, enforcement, and error costs).

 [185]. A version of amoralism about fiduciary loyalty based solely on normative arguments (for example, the one advocated by Easterbrook and Fischel) might dispute whether a cognitivist account of fiduciary loyalty is ultimately justified. Our empirical claim that extant fiduciary law implicates cognitive concerns might be appraised as normatively misguided. Perhaps courts lack institutional competence to appraise cognitive considerations. Perhaps behaviorist and cognitivist approaches can be expected to produce similar results, but the behaviorist understanding imposes lower expected costs. Our case for cognitivism does not rule out the possibility of such a normative argument against cognitivism. However, abandoning cognitivism would enable several forms of abuse and exploitation that fiduciary law is designed (and uniquely suited) to protect against. Moreover, such a purely normative argument would be highly revisionist, since it would require denying many (and perhaps all) of the legal propositions identified in Section II.B.

 [186]. See Julian Velasco, The Role of Aspiration in Corporate Fiduciary Duties, 54 Wm. & Mary L. Rev. 519, 521–22, 526–27 (2012); see also Melvin Aron Eisenberg, The Divergence of Standards of Conduct and Standards of Review in Corporate Law, 62 Fordham L. Rev. 437 passim (1993).

 [187].               See, e.g., Shiffrin, supra note 54, at 722–27.

 [188]. Meir Dan-Cohen, Decision Rules and Conduct Rules: On Acoustic Separation in Criminal Law, 97 Harv. L. Rev. 625, 630–34 (1984).

 [189]. Yuval Feldman & Henry E. Smith, Behavioral Equity, 170 J. Institutional & Theoretical Econ. 137, 145 (2014).

 [190]. Id. at 148.

 [191]. Id. at 147 (citing Seana Valentine Shiffrin, Inducing Moral Deliberation: On the Occasional Virtues of Fog, 123 Harv. L. Rev. 1214, 1222–25 (2010)). This story is, in broad strokes, consistent with the explanation for the combination of lax enforcement and widespread compliance in Lynn A. Stout, On the Export of U.S.-Style Corporate Fiduciary Duties to Other Cultures: Can a Transplant Take?, in Global Markets, Domestic Institutions: Corporate Law And Government in a New Era of Cross-Border Deals 46, 55 (Curtis J. Milhaupt ed., 2003) and Margaret M. Blair & Lynn A. Stout, Trust, Trustworthiness, and the Behavioral Foundations of Corporate Law, 149 U. Pa. L. Rev. 1735, 1754–55, 1782–83, 1785 (2001).

 [192]. Feldman & Smith, supra note 189, at 150.

 [193].               See generally Ernst Fehr & Simon Gächter, Fairness and Retaliation: The Economics of Reciprocity, 14 J. Econ. Perspectives 159 (2000) (arguing that explicit incentives can “crowd out” voluntary cooperation and intrinsic motivation).

 [194]. For example, a cognitivist approach to contract theory might resemble Daniel Markovits’s work, which focuses less on the divergence between contract and promise and more on the structural morphology between the two. See Daniel Markovits, Contract and Collaboration, 113 Yale L.J. 1417, 1473 (2004); see also Ethan J. Leib, On Collaboration, Organizations, and Conciliation in the General Theory of Contract, 24 Quinnipiac L. Rev. 1, 5–8, 18–20 (2005).