The Value of Contract and the Politics of Personal Detachment

1. Introduction

What kind of contribution do contracts make to human life? In the usual case, we enter into contractual relationships with one another to do what for each of us is in our best interest, when that happens to be something we cannot do all by ourselves. It is hard to deny or overstate the importance of being able to do this. Contracts are, if nothing else, clearly of instrumental value to our lives.

Outside of economics departments, it seems fair to say that the prevailing opinion is that contracts are, in addition to being of instrumental value, also of noninstrumental value to human life. To say that something is noninstrumentally good or of value is to say that its benefits accrue directly from engagement with it for its own sake, rather than indirectly, that is, as a means to something else. I will call the view that contracts are of noninstrumental value inflationism about contract because it tends to puff up the nature of the contribution contracts make to our lives, at least relative to the deflationist view that their contribution is “merely” instrumental.

I believe that inflationist contract theorists are quite right to push back against the excesses of economism in their field. But I am doubtful of their way of doing so, as I believe that deflationism about contract is true. The tendency towards inflationism typically expresses liberal political commitments in the broadest sense. Such theories can be grouped in two. The bulk of them are helpfully regarded as so many versions of what historians of ideas call the doux commerce thesis, the Enlightenment view that the commercial way of life—and the contractual relationship that is its foundation—is a civilising force in society, advancing the cause of freedom, pluralism, trust, and other such characteristically optimistic values.¹ The remainder seem to valorize the mores of personal detachment that arise at a rather later stage of economic development, and may be thought to express a more pessimistic take on the contemporary liberal polity. This paper targets pessimistic inflationism, leaving discussion of its more optimistic varieties for another occasion.

The paper is structured as follows. In §2, I present and strengthen Dori Kimel’s (2003) contract theory, which is the most developed version of pessimistic inflationism of which I am aware. In §§3–4, I consider various interpretations of his view, finding that on its most promising version, the value of personal detachment lies in how having it as an option deepens the meaning of our existing relationships in a sui generis way. In §5, I argue that, even so interpreted, Kimel’s theory faces political challenges of a sort that Nancy Fraser (2013) has probably done the most to illuminate. In §6, I offer, in a conciliatory spirit, a story about the relationship between values and technology that may allow us to salvage what is attractive in Kimel’s theory. Section 7 concludes that, whether or not the story is persuasive, there is serious pressure in favour of contractual deflationism.

Before starting, a quick point on method. We defined as inflationist any theory of contract that claims that the contractual relationship is of noninstrumental value to our lives. But not every contract theorist who makes appeal to some noninstrumental value or other thereby advances an inflationist theory, even if the rhetoric has an inflationary feel to it. For their theories to count as genuinely inflationist, it must be the case that when contracts are said to ‘promote,’ ‘foster,’ ‘enable,’ or ‘enhance’ the preferred noninstrumental value, these words pick out constitutive and not contingent relations. To see why, consider the humble cheesegrater. We can all agree that the cheesegrater contingently promotes all sorts of noninstrumentally valuable effects, both by design and by accident, such as, most obviously, the existence of certain culinary delights and experiences. But this does not imply or even tend to show that cheesegraters themselves are of noninstrumental value, directly benefitting us by engagement with them for their own sake. By the same token, a theory that says that contracts promote some preferred value only contingently does not automatically count as genuinely inflationist. To count as genuinely inflationist, contracts would have to be said to be an essential constituent of the relevant value; something without which the value could not be realized. In what follows, I thus assume that a theory of contract is genuinely inflationist if and only if the value it is said to promote: (a) is noninstrumental, and (b) cannot be realized in the absence of contract.

2. Introducing Kimel’s Theory

In the contract literature, inflationism has tended to come in one of two forms. The first, we said, is animated by the spirit of doux commerce. The second, we said, reflects a move beyond Enlightenment optimism about the liberal social order, and seems, if anything, to represent something like the obverse of the doux commerce thesis. Our focus will be on this second, more pessimistic form of inflationism.

The language of the ‘obverse’ of doux commerce is not my own. This is how Hirschman describes what he calls the “self-destruction thesis,” according to which, far from promoting the genteel good manners and civility necessary for peace and social order, the commercial organization of society in fact “carries within itself the seed of its own destruction” (1986: 111). Though, as Hirschman makes clear, the self-destruction thesis takes many forms, the basic criticism of doux commerce that runs throughout is that it mistakenly appraises the whole historical dynamic of commercialization on the basis of what are in fact merely the effects of its honeymoon stages (1986: 109–117). So even if it is true, as doux commerce predicts, that the spread of commerce may initially work to erode the old bonds and hierarchies of an ungovernable feudal system, it is also true that, inevitably, the same forces tend to hollow out the social bonds that lie at the foundations of the new market society. In the end, the logic of commercialization is thus said to leave the atomized denizen of late modernity without even the minimal morality needed to prop up its own way of life. This critique has recognizable echoes of both conservative and Marxian forms of discontent with the liberal socioeconomic order, though, as Hirschman notes, it has been espoused by thinkers whose discontent took neither form (1986: 110).

It may thus come as a surprise that a contract theory that represents the “obverse” of doux commerce could also, in its own way, be expressive of values of an undeniably liberal flavour. To see how this could be, we need only recognize that it is always possible to lean into criticism and to own it. Liberals who adopt this strategy are not afraid to look their opponents in the eye whilst spinning the target of their criticism into a virtue. The idea would be to find something in the commercial dissolution of our social bonds which is, on closer inspection, in fact a good thing to be welcomed for its own sake. To my mind, a liberal contract theory built around some such value would be more robust than any alternative based in doux commerce, since it would express a more sober and nuanced take on just what it is about how we have organized ourselves today that we all, no matter our political persuasion, can agree we hold dear. Quite possibly it will convey a more pessimistic image of the merits of the late modern liberal-capitalist polity, but it will, in my view, be the sounder a liberal contract theory for it. In the contemporary literature, the most advanced proponent of this strategy is probably Dori Kimel. Let us briefly introduce Kimel’s theory and lay aside some well-represented objections to it.

In dialectical terms, Kimel is concerned to offer a liberal corrective to what he sees as a tendency in liberal contract theory to overegg the continuity between contracts and promises. He pursues this aim through a careful comparison of the two. At the most fundamental level, promises and contracts are seen as distinct though related practices that recognize our ability to voluntarily undertake obligations to other people. There are important similarities and differences between the values of each practice. One key similarity lies in the instrumental function of facilitating a certain kind of reliance, which enables us to do things we happen not to be able to do on our own (2003: 65–66). Another is that, in each case, the practice embodies the value of respect for our personal autonomy (2003: 133; 2010: 220). However this latter value, Kimel suggests, finds itself embodied in contracts and promises in radically different ways, due to differences in the mechanism by which they secure their common instrumental function (2003: 78–80; 2010: 220). Roughly speaking, he thinks that promises typically operate via a thick form of trust native to more involved or intimate kinds of relationships, whilst contracts proceed instead via the more impersonal incentives associated with legal enforceability (2003: 74–77; 2014: 113). As a result, the two practices are said to realize quite distinct aspects of the value of personal autonomy (2003: 80; 2010: fn 10). Whereas the trust-based practice of promising realizes autonomy under the aspect of enhancing our more involved relationships, contract does so under the aspect of what Kimel calls personal detachment, the noninstrumental significance of which is “in a sense, the diametrically opposed value” (2003: 78).

We could say, simplifying somewhat, that for Kimel, contracts foster detachment whereas promises foster attachment. Put this way, the view may seem confused to those, like Seana Shiffrin (2012), who would agree that promises foster attachment, but who would add that contracts are promises. Perhaps Kimel could agree with this latter statement if all it means, as per the above, is that contract essentially involves voluntarily undertaking obligations to other people, making them promises only in this thin sense.² Nonetheless, for Kimel, as we have seen, there remains a thicker and more salient sense in which contracts are not promises. The legal enforceability of contract “casts a thick and all-encompassing veil over the motives and … attitudes” (2003: 74) of the parties, making it a “singularly inadequate arena for revealing character traits and expressing attitudes of the kind on which personal relationships thrive” (76), such as “trust and respect” (77). By contrast, promises can be “understood as designed to promote special relations or special bonds” (66) precisely because, in the absence of the threat of state coercion, “There is no equivalent problem of transparency” (76). To reiterate: for Kimel, the shadow of legal enforceability is what prevents contracts from fostering attachment; contracts are not promises in this sense.

The view may yet be considered confused, as it were, in the opposite direction. For example, Daniel Markovits (2011) could agree with someone like Shiffrin that contracts are promises, though not because promises foster attachment, but because they foster relations at arm’s length that are in a clear sense detached. On his view, “Promising … involves a form of recognition: to promise to someone is to respect her as a person—to acknowledge her moral personality—in a particular way” (2011: 296). Unlike our more involved relationships, however, this way is relatively detached: it only requires that we appreciate the “generic perspectival capacities in virtue of which promisees are like all other persons,” rather than “the peculiar way [it] receives expression in the particular person who he is” (2011: 308). But even promises so conceived will not foster relations of detachment in the sense that Kimel thinks contracts do. That is because the relatively detached forms of recognition and respect that Markovits thinks characterize promises are essentially attitudinal, requiring deference to the “expressed intentions” of the promisee (308). Of course, if, as Kimel says, legal enforceability makes the presence or absence of such attitudes interpersonally opaque, he can maintain that contracts foster detachment in a way promises do not, whatever truth there may be in Markovits’s view of the latter.

As we can see, quite a lot comes to turn on the role Kimel ascribes to the legal enforceability of contract. There is something intuitively right in the idea that enforceability crowds out attitudes of respect and trust, or at least renders them interpersonally opaque. But equally, we do not ordinarily suppose that common knowledge of the fact that our interactions are governed by legal norms enforceable by a violent state apparatus suffices to justify a generalized suspicion about their attitudinal basis. As Shiffrin (2012: 253) points out,

In the case of tort, there is little serious concern that legal regulations on bodily contact have come to dominate the motives of citizens or that citizens believe their safe passage across the streets is generally a matter attributable to law and not to basic civic decency. Why should we worry more about the case of contract law’s infiltrating and tainting moral citizens’ primary motives of promissory fidelity and trustworthiness, rather than its working as a form of official recognition and source of backstop assurance?

Shiffrin’s point is well-taken, though it may be addressed as follows. For any given interaction between parties, it is governed by tort law norms whether they like it or not.³ Thus the mere fact that an interaction is governed by tort law norms does not thereby give parties any presumptive reason to think that they prefer their interaction to be so governed, as they might, for example, if they did not antecedently trust each other to interact out of respectful motives. But things are different where it is up to the parties to choose whether or not their interaction is governed by legal norms, as is the case, by default, in contractual contexts. Where this is so, the mere fact that an interaction is governed by legal norms does plausibly justify the parties in presuming that they do not antecedently trust each other to perform as agreed. After all, if they did, they would not have chosen to invoke state violence when they had the option not to. So even if the omnipresence of a violent state need not be inimical to trust-based relations, it surely may be, contra Shiffrin, when the parties actively choose for that violence to underwrite their interaction.

One last well-trodden objection we should consider is an empirical one. Ethan Leib (2010: 654–655) has pointed out that it is by now widely accepted that some contracts are created against the background of more involved relationships between the parties, and may even serve as a starting point for them. If some such observation is to be accepted, that may seem to refute Kimel’s theory on empirical grounds, since his theory predicts that the legal enforceability of contract would prevent the flourishing of the background attachments. But since the theory is not offered as an exceptionless law of nature, it is only empirically refuted if the observation can be taken to reflect the paradigm case of contract, which in turn, as Kimel argues, is both statistically unlikely (2003: 81, fn 50) and potentially theoretically undesirable (2007: 235). Moreover, even if we do take such contracts as paradigmatic, Kimel’s theory is in any case primarily offered as a normative theory. So, as Leib (2010: 656–659) acknowledges, Kimel may take the observation to bear on the theory in different ways: he may adjust the theory to accommodate the data, or he may hold fast to the theory, using it to critique the relevant contractual arrangements. To the extent that Kimel can be read as taking the first option, we shall consider that in more detail in later sections.⁴ But it is also open to him to take the second option, and all the more so given the tendency, as he sees it, to “vastly exaggerate” the pervasiveness of the relevant contractual arrangements (2003: 81, fn 50). Maybe, as Leib suggests (2010: 719), the critique of those arrangements will trade on “a conception of intimate relations that is altogether too pure or Pollyannaish.” But then again, maybe not: how our intimate relations ought to be falls squarely in the realm of reasonable disagreement.

So concludes my brief introduction to Kimel’s theory. In the complex and hotly contested terrain of contract law, we should not expect for a given theory to command anything like universal acceptance. That being said, I hope it is clear that Kimel’s theory represents a distinct option in the contemporary contract literature, one that is capable, at least prima facie, of holding its own against various better-known competitors. There is much to admire in Kimel’s work; it is dense with kernels of thought that repay turning over again and again. In what follows, we will have to artificially narrow our focus to what he says about personal detachment. Making plausible the noninstrumentality of the value of personal detachment seems to be the key to establishing the liberal credentials of Kimel’s theory at a depth sufficient for it to play the desired dialectical role. I must admit, therefore, to finding it odd that he devotes so few words to this task. Though what little he does say is characteristically replete with possibilities, I do not think, in the end, that there is enough in there to get the job done.

3. Instrumental Benefits of Personal Detachment

Contract is said to make us able to do things by agreement with others without having to get personally involved. But what exactly does Kimel have in mind? He begins with the following gloss. The value of personal detachment, he says:

… consists in the very framework contracts provide for doing certain things with others not only outside the context of already-existing relationships, but also without a commitment to the future prospect of such relationships, without being required to know much or form opinions about the personal attributes of others, and without having to allow others to know much and form opinions about oneself. (2003: 78)

No doubt the tacit sentiment here will be all too familiar to the twenty-first century metropolitans that can safely be assumed to make up the bulk of readers. For my part, after a busy day, I like that I can immediately put Spotify in my ears, buy a soup with a minimum of niceties, and get off the bus without need of a “thanks.” I like that these transactions do not have to be interactions.⁵ I do not want to deny that such detachment can be a good thing. But when we reflect on what might lead us to want such detachment from others, it becomes clear that the familiar sentiment harbours all the pessimism—some of it quite possibly justified—of how we think of the mass of strangers out there in late modern society. We basically think of them as unlikely to really warrant a personal investment of time, effort and other such limited resources, not least because it would be better allocated to the people and projects we already have in our busy lives. So understood, the value of personal detachment looks plainly instrumental.⁶

To make all of this more concrete, I suggest that we unpack the pessimism of the sentiment and sharpen it as follows. What sorts of generic worries might be intelligible when we are faced with the prospect of sustained interaction with a complete stranger? Here is an indicative list: (a) they might pry or meddle in our affairs, (b) they may form unwanted expectations of us going forwards, (c) we may simply not get along, and (d) the whole exercise will consume time, effort and other such limited resources. The list aims to be exhaustive, and the intention is for each item to be interpreted such that they are ordered in increasing strength, in the sense that each item necessarily instantiates the subsequent one but not vice versa. I submit that for each of the worries on the list, if it really is good for us to be able to be rid of them, then that value will be instrumental.

We can group (a)-(c) together on the basis that they are worries about possibilities, rather than certainties, as in (d). Note, first of all, that what we worry about with (a)–(c) is not always unwelcome in our relations with others. After all, the possibility of these things happening, at least on occasion, seems to be part and parcel of what makes our most intimate relationships worthwhile: our being friends, family, lovers, and the like. These relationships could not have the meaning for us they do if they had no latitude for the sorts of annoyances that stem ultimately from our caring about the other (who they are; how they live; how they see us), as are captured under (a)–(c). This lends our being meddled with, held to account and not getting along a special meaning—even if it really might not feel that way at the time.

It seems to me that if we should want rid of these worries at all, we need to interpret them more specifically. Their object cannot be so much that (a)–(c) could occur per se as that people who we are not interested in could subject us to these things. Put this way, we would be rid of the worries if we could control for ourselves who it is that we become personally involved with. The thought would be that we have noninstrumental reason to want the relationships best suited for us, and that, since we have privileged access to who we are and what makes us tick, our having control over who we do and do not get to know more intimately is a good way of getting to have the relationships that are best suited for us. I emphatically agree that this sort of control that comes with the option of detachment can be a good thing. But the value, so explained, is surely instrumental, i.e. to having relationships that are better suited to us.

Now let us move onto (d), the worry that our interaction with the complete stranger will take up time, effort, and other scanty resources. Since we humans are finite beings, any interaction whatsoever—strangers, acquaintances and intimates alike—is certain to have this consequence. For this reason, it is harder to accuse someone with worry (d) of the sort of pathological risk-aversion that worries (a)–(c) may express. And yet, as with (a)–(c), it is not that what we worry about under heading (d) is always unwelcome. For it is, again, part and parcel of what makes our more intimate relationships worthwhile not only that we can be expected to find the time, effort and resources needed to keep them vibrant, but that, at least on occasion, we actually relish in the opportunity to do so. In this context, someone who claims to have found the “solution” to the expenditure of time, effort and resources betrays a fundamental confusion as to the noninstrumental significance of their intimate relationships.

So once again, if we should want rid of worry (d), then its object cannot be that we will consume our limited resources per se, so much as that we waste them in so doing. In the first instance they would be wasted on people who do not matter to us personally, or to whom we do not personally matter.⁷ More broadly, and derivatively, our resources would be wasted on people with whom the prospects of a deep relationship are doomed from the start. In other words, in addition to helping us have relationships that are better suited to us, our control over our involvements also helps us to efficiently allocate time, effort and other such scanty resources to the people who really merit their consumption, as adjudged by those best positioned to do so: ourselves. Therein lies the value of personal detachment, our ability to remain at a cool distance from those on whom we think our resources will be wasted. This seems to be yet more instrumental value: it helps us get the most out of the sources of noninstrumental value we already have in our lives.

4. Are There Noninstrumental Benefits?

Kimel’s initial gloss therefore does not obviously help his inflationist contention that relations of personal detachment are noninstrumentally valuable. He does however go on to say more about what he has in mind:

Not only is it easy to see that, when examined in isolation, both personal relationships and personal detachment can be, in the right circumstances, a good thing, but in fact when the two co-exist as options in people’s lives, their respective values tend to be mutually reinforcing. (2003: 79)

There are at least two thoughts in this passage that may be taken to have a bearing on the alleged noninstrumental value of personal detachment. One seems to me to be somewhat confused, whereas the other is, I think, quite profound. Let us first tackle the confused thought, so as to avoid any distractions it may cause.

When Kimel says that both personal relationships and personal detachment can be a good thing, this will, of course, be both trivial and unhelpful to his cause if ‘good’ is to be read instrumentally. But on a noninstrumental reading there arise difficulties as soon as we factor in for the suggestion, as Kimel says, that both relationships and detachment are a good thing “when examined in isolation.” The very idea of noninstrumental value seems to be such that if something has it, then it would have it if examined in isolation, that is, independently of the noninstrumental value of anything else.⁸ Thus, even if it happens to be the sole source of noninstrumental value in our life, our dependence on it, in that sense, would not threaten its status as noninstrumentally valuable. Suppose, for example, that we were to get stranded on a desert island with just one friend, so that we are dependent on that one friend as a source of the value of personal relationships in our life. There is no doubt that our social life on the island would be impoverished, at least as compared to the vast majority back in civilization. Indeed, it is likely, on some reasonable measure, to count as absolutely impoverished. But our dependence, I take it, does not necessarily abnegate or even tend to diminish the noninstrumental significance of the friendship. As a conceptual matter, it seems that what is of noninstrumental value remains so when examined in isolation, and is insensitive to our dependence on it in that sense.

The confusion becomes apparent when we recall Kimel’s claim, in the above, that both personal relationships and personal detachment can be of noninstrumental value when examined in isolation. Though he says nothing that obviously contradicts the first conjunct,⁹ he appears, in light of what we have just said, to go back on himself regarding the second. As he says in several places, what is of noninstrumental value is not so much our standing in relations of personal detachment simpliciter, but rather our having the option of proceeding via such relations. “Detachment,” he says, “is valuable as an option, not a predicament” (2003: 79). The idea here is attractive and can be spelled out as follows. Personal detachment can be a good thing, but if, in our dealings with others, it is not open to us to proceed in a more involved way—i.e. if detachment is not an option but a necessity—then that makes it seem no good. For instance, socially marginalized people may have no choice but to invoke the machinery of law because the others with whom they have to deal tend to prove untrusting, untrustworthy, or indeed both.¹⁰ It would be, I think, a grave error of moral and political judgment to maintain that this is a blessing and not a curse. Kimel shows acuity of judgment when he says that “dependence on personal detachment is itself not so much a thing of value but a predicament” (2003: 142). But this, alas, contradicts his claim that personal detachment remains a good thing when examined in isolation, that is, even in a world where we are dependent upon it—as would have to be the case if the “good” claimed is noninstrumental. As we have seen, the value of personal detachment differs from the value of personal relationships in this respect, which reflects that the latter but not the former has the status of noninstrumental.

This is why I believe the first thought in the above passage is confused, and can be safely disregarded. The emphasis on the value of the option of personal detachment, as opposed to relations of personal detachment simpliciter does, however, bring us closer to the second thought. Now, of course, if all this means is that personal detachment is noninstrumentally valuable, but somehow only when we are not dependent on it, then, as we have just seen, it will be a nonstarter too. But this is not the only way to unpack the value of the option. There is also the possibility that when personal detachment and personal relationships coexist side by side as options, their values mutually reinforce each other. Once we recognize this possibility, another route to the noninstrumentality of the value of personal detachment presents itself, namely, that having it as an option is constitutive of the enhancement of the value of our personal relationships, making them deeper or more meaningful in some sui generis way. This is the second thought in the Kimel passage, and it is, to my mind, far more penetrating than the first. I believe it puts a finger on something primordial in the doctrine of liberalization.

To be clear: the proposal is that the option of personal detachment constitutively enhances, deepens, or otherwise makes more meaningful our relationships, and in that way receives a noninstrumental significance. What exactly does Kimel have in mind? His example (2003: 79) of the liberalization of family relations is indicative.

[E]ven where examples such as family relations are concerned, the relative freedom to draw lines, to exert control over the scope, the depth and the intensity of the relationship usually tends to improve the quality of the relationship and enhance its value for the parties, rendering it more meaningful—indeed, it is tempting to say more viable—than the kind of all-encompassing family ties that deny the parties the opportunity to maintain relative privacy or detachment in any way or with regard to any aspect of their lives.

The general idea here seems to be that our relationships become more meaningful to the extent that we gain control over the scope, depth and intensity of what is required of us within the sphere of the relationship. Control comes, fundamentally, from having the option to keep certain of our thoughts, feelings and actions private, thus cordoning off those aspects of our lives from the demands of the relationship. These aspects of our lives are often important, and so we do not have this control if we are simply unable to realize the relevant thoughts, feelings and actions altogether. Instead, we have to be able to do this in a manner that suits us qua individuals, and not just as required under the jurisdiction of the relationship, where the other can always interfere. With thought and feeling, whether we have this option basically depends on the dispositions of the parties: how pushy or violent they are; how poker-faced and resilient we are. But with action, it basically depends on how things external to the parties are set up. This is in part a function of there being favourable social and cultural norms that permit us to detach ourselves from our relationships insofar as they require us to do things in a certain way. But it is, in equal measure, a function of there being concrete ways to do these things in a manner that suits our individuality, which there would not be but for the existence of markets for services in these things.¹¹

It is this last idea that will concern us more specifically. Our relationships are said to be more meaningful when we have the option to detach ourselves from them in the manner just described, giving us control over how we do the important things we have to do within their sphere. As there comes to be a market for more and more human activity traditionally thought to fall within the realm of our personal relationships, our options for outsourcing them to service-providers become more and more extensive, running the whole gamut of human activities, from the most to the least tangible. Our having these options, whether we use them or not, is supposed to deepen the meaning of our relationships in that it becomes a constituent of their noninstrumental value that their shape is sensitive to our individuality. That we could, if we pleased, opt to proceed in a more detached way in any aspect of a given relationship seems to be part and parcel of the depth of its value to who we are as individuals.

I believe this proposal has a lot going for it. The idea that our personal relationships are deepened by having options over the extent of their jurisdiction in our lives chimes keenly with our modern sensibilities. It is also plausible that the enhanced sensitivity of our relationships to our individuality is constituted by the existence of the option of personal detachment, and that this option, in turn, cannot be had without markets in services and their contractual foundations. Does this interpretation of Kimel’s appeal to personal detachment secure the inflationist ambitions of his theory? I want to argue that, despite appearances, it does not. Even on this promising reading, I submit, Kimel faces political challenges that thwart the inflationist ambitions of his theory.

5. The Politics of Personal Detachment

Think of Kimel’s proposal in the following terms. There are, first of all, what we could call old-style relationships bound up with norms and expectations—quite possibly both asymmetrical and systematically so across morally arbitrary groupings of parties—that are expressive of traditional values and authority. The potential meaning for us of these old-style relationships is genuine.¹² But it is, in a second step, deepened by our having the option of personal detachment, which turns them into new-style relationships. Since the relevant deepening on display—viz. their sensitivity to our individuality—cannot be achieved without our having the option of detachment, the option is constitutive of the deepened meaning of the new-style relationships.

Before anything, we should register the clear emancipatory impulse behind Kimel’s proposal. Whatever else we think of it, the proposal begins as a reflex against traditional authority, and should be commended as such. With this firmly in place, we can start to apply some pressure. Notice that, on Kimel’s proposal, new-style relationships are going to turn out noninstrumentally worthwhile just so long as the degree of detachment they exhibit (which could be nil) reflects a sensitivity to the individuality of the relevant party. We shall problematize this implication. In particular, we will consider a counterexample to it that, fully spelled out, reflects a certain tension in the politics of the wider proposal.

Here is the counterexample. Imagine two people for whom career is the organising principle of life, who are intensely aspirational, ambitious and status-driven, and who fundamentally identify with the values these traits express. They have heard that there are certain advantages that come with romantic involvements, and they find in each other precisely the degree of involvement their respective career goals will tolerate. At some point, they hear that kids, too, have their advantages, and so they decide to become parents. When they do, of course, it turns out that being parents is rather more work than their careers will allow. And so, slowly but surely, they outsource every aspect of the material, social and emotional labour of care for their children to a rotating circle of third-party providers. In the parents’ eyes, the relationship with their kids is ideal: its non-existent demands are exactly sensitive to who they are as high-powered careerists. The thought is that, having outsourced so extensively, they surely cannot honestly count a meaningful relationship to their children amongst the sources of noninstrumental value in their lives.¹³

There is a certain kind of conservative that could, I think, agree with a certain kind of progressive about this counterexample. For example, some social conservatives will cite the sanctity of the family and its social bonds, which they regard as the foundation of society and social order. More often than not, this comes bundled with the claim that it is unnatural for parents not to care for children themselves, perhaps with approving gestures towards a gendered division of labour alleged to reflect biological differences. Of course, such grounds for suspicion of the noninstrumental value of the relationship need not worry Kimel, for, as we saw, his proposal is intended as an emancipatory reflex against just such a conservatism. A fundamental disagreement with traditionalist values is baked in to the proposal from the start.

More worryingly for Kimel, there are serious progressives who will regard this as a counterexample, and indeed on emancipatory grounds. The idea would be that at least some of what is outsourced here is the stuff of the good life, with the result that we are alienated from the relational sources of our own good.¹⁴ Our relationships, which should be a source of deep meaning and significance, instead come to confront us as something alien and external. The activity of satisfying their demands is reduced to the status of services, which, by contracting out, facilitates our pursuit of something yet further that is the true source of meaning in our lives: work, wages, and the freedoms that come with socio-economic security. When our social relations become alienated in this way, the thought goes, we suffer a form of domination or oppression.¹⁵ And this is something that should worry Kimel. Though his proposal sets out to resist oppressive structures of traditional authority, it inadvertently ends up valorising oppression in another guise: the alienation of our social lives under the pervasive imperatives of the market.

Nancy Fraser would say that the counterexample shows that Kimel’s theory suffers the same political fate as did many emancipatory projects of the latter 20^th century. The starting point for Fraser is The Great Transformation, Karl Polanyi’s seminal 1944 study of market society. In Polanyi’s analysis, market forces tend to dissolve social bonds, so that, as the economy gets increasingly marketized, ordinary people instinctively resist the attendant threat to social life as they know it. As a result, societies whose economies assume a market form are constitutively politically unstable. Polanyi (2001: 80–81) calls their powerful political dynamics the “double movement,” using it to explain apparently heterogeneous historical outcomes ranging from the establishment of social protections (e.g. welfare systems) to the rise of fascism. For Fraser, however, the analysis conceals as much as it illuminates, and will continue to do so until we recognize—opposed both to the forces of social protection and marketization—the distinct forces of emancipation. We must extend the analysis from a “double” to a “triple movement” (2013: 230).

To see why, take Fraser’s analysis of the fate of feminist critique of social insurance schemes based on the male-breadwinner model of the “family wage.” The original intent behind the critique is clearly emancipatory: the family wage model is patriarchal in that it entrenches the second-class status of women in the economic and social spheres. But without a clear view of the triple movement, Fraser suggests (2013: 220–221), subsequent feminists struggled to frame their emancipatory opposition to social protectionism in ways that did not unwittingly play into the forces of marketization:

Our critique of the family wage now supplies a good part of the romance that invests flexible capitalism with a higher meaning and a moral point. Endowing their daily struggles with an ethical meaning, the feminist romance attracts women at both ends of the social spectrum: at one end, the female cadres of the professional middle classes, determined to crack the glass ceiling; at the other end, the female temps, part-timers, low-wage service workers, domestics, sex workers, migrants, EPZ workers, and micro-credit borrowers, seeking not only income and material security, but also dignity, self-betterment, and liberation from traditional authority. At both ends, the dream of women’s emancipation is harnessed to the engine of capitalist accumulation. Thus, second-wave feminism’s critique of the family wage has enjoyed a perverse afterlife.

Just as feminism falls prey, in this way, to what Fraser elsewhere (2013: 211) calls the “cunning of history,” the worry for Kimel is that what starts life as an emancipatory proposal against traditional authority ends up being co-opted by the oppressive forces of marketization. That is the force of the counterexample. Is there any way for Kimel to respond?

We may imagine that sensible liberals must have things to say about the value of autonomy that can put legal limits on the freedom to contract sufficient to rule out the counterexample. But on reflection, we see that sensible liberals are unlikely to find this solution attractive. Here is how that works in Kimel’s case (2003: 129–134). On his view (131), “An autonomous life is valuable when spent in the … pursuit of valuable activities and relationships, but not otherwise.”¹⁶ Personal autonomy is thus not just a matter of having a range of any old options, but specifically valuable ones. To protect autonomy, a liberal legal order may thus eliminate options that are reasonably regarded as bad,¹⁷ including options that are bad because they are such as to essentially threaten autonomy: they are exploitative, unfair, or otherwise unduly restrictive. But the contractual options in the counterexample are not eliminable by this route. Contracts for childcare services are unobjectionable, at least taken one by one. Of course, it may be that they threaten autonomy merely accidentally, for example, if their wanton use makes it harder to participate in meaningful parent-child relationships. In that case, it may be possible to limit these contractual freedoms. But this threat to autonomy only obtains if the availability of such contracts tends to hamper meaningful parent-child relationships in general, which need not be the case: the careerist parents, we can assume, are quite idiosyncratic in their outsourcing practices. So, for sensible liberals, as Kimel says (2010: 230–231), the focus is always on “what kind of transactions the law of contract should facilitate,” so as not to actively “facilitate immoral conduct.” For unobjectionable kinds of transactions, a sensible liberal cannot limit our freedom to make them simply because we could opt to squeeze them into our lives in silly ways.

An initial response for Kimel might be as follows. In general, it may be that there is value in our having certain powers even as particular exercises of those powers are not themselves valuable in that way. Thus, having the option of detachment may really be worthwhile in that it potentially makes for more meaningful relationships, even as the use of that option fails to do so in a given case. In our preferred language, this suggests a type-token ambiguity in the new-style relationships Kimel proposes. In saying that the option of detachment makes for more meaningful relationships, what he means, he could claim, is that it makes for a more meaningful type of relationship: the new-style type of relationship. But as with any worthwhile type of relationship—friends, family, lovers, you name it—its tokens may not themselves be worthwhile, because they are in some way pathological tokens of the type, i.e. the result of bad or pathological exercises of an otherwise worthwhile option of detachment. Perhaps this explains Kimel’s point (2003: 79) from before that having this option is a “good thing” only “in the right circumstances.” The suggestion would be that the counterexample is cherrypicked from the defective tokens of a generally worthwhile type of relationship, so that, although the token parent-child relationship in question really is sensitive to the individuality of the parents, Kimel is nonetheless not forced to concede that it is in fact worthwhile.

But if this is the idea, then Kimel needs to tell us what exactly he thinks it is about the parent-child relationship in question that plausibly suffices to make it a pathological token of the type, and so not in fact worthwhile. It cannot be that the degree of detachment it involves fails to reflect the individuality of the parties, since it does not fail to do this ex hypothesi. Another suggestion could be that the extent of detachment in the relationship amounts to a form of abuse or neglect—a suggestion that becomes more plausible when we focus our attention on the kids. But is it really so clear that the kids suffer abuse or neglect in the relationship? The kids are, after all, being cared for by someone, and that care is ultimately paid for by the parents. Indeed, the kids may, having been brought up the way they have, feel that their relationship to their parents is just as they wish. At this point, perhaps, we may want to attribute something like false consciousness to some or all of the characters involved. Yet by Kimel’s own lights, this solution is unavailable: it would presumably be anathema to the liberal emphasis on individuality within his wider proposal.¹⁸

This train of thought suggests a more promising line for a liberal like Kimel to take. He could reply that the intuition behind the counterexample—i.e. that the relationship is alienating and oppressive despite the fact that it reflects the parties’ individuality—is based in a comprehensive doctrine about the good life. He can then say that, though he respects this doctrine, which seems reasonable enough, he objects to the tacit assertion that it is ultimately correct. This may be on grounds of liberal neutrality: that trying to elevate this doctrine above those of the characters in the counterexample is patronising, paternalistic, or otherwise fails to treat them with equal respect and concern. But it need not be. The objection could just as well be on the epistemic grounds that the superiority of one comprehensive doctrine over any other is unknowable, which may be, in turn, for the metaphysical reason that there is no supreme value against which conceptions of the good can be meaningfully compared. An objection on any of these grounds could, it seems, be enough to get Kimel off the hook.

No matter the ground of the objection, however, the promise of this approach turns out to be illusory. The point of the counterexample is to present a new-style relationship that is intuitively not, as Kimel would have it, noninstrumentally worthwhile. The reply at issue objects to the counterexample’s tacit invocation of a substantive conception of the good. But if Kimel really can nullify the counterexample in this way, he must also, by the same token, nullify his own proposal that new-style relationships are just the old-style ones but with a deepened meaning or noninstrumental value. This is because it is far from clear how Kimel can justify his claim to deeper meaning—where that is a way of being more worthwhile—without invoking some comprehensive doctrine himself, thereby falling foul of the same objection, whatever its ground.

Perhaps at this point Kimel will concede that he misspoke in using the language of “deepening.” What he meant, he might say, is that the axiological upshot of our gaining the option of detachment is not so much a deepening of the old-style relationships, as their creeping replacement by mutually exclusive new-style ones. It is not that the new-style relationships are arrived at by any sort of improvement of the old-style ones, so that the latter come to have a greater noninstrumental significance for us. It is rather that the values embodied in the old- and new-style relationships are just fundamentally different, such that it is simply not possible to meaningfully compare the ways our participation in them can directly enrich our lives. There is no sensible answer to the question: which is the better, richer, or deeper style of relationship? It is just, as they say, apples and oranges.

Of course, if this is right, then the often fraught political contestation around these issues in recent times will turn out to be basically pointless. Since the traditional values embodied in old-style relationships are just some amongst many incomparably diverse values, each worthwhile in their own way, it turns out to be little more than a matter of taste whether, for instance, the place of women in economic and social life should be to furnish the background conditions for the productive activity of men. Similarly, since the values of liberal individuality embodied in each and every new-style relationship are just some amongst many incomparably diverse and equally legitimate values, it turns out, once again, to be no more than a matter of taste whether we should aspire to a form of life that does not cultivate social alienation. Perhaps the voices of social conservatism, laissez-faire and emancipation are all, in the end, just talking past each other. There is a type of liberal who could find comfort in this thought, though I am not convinced Kimel would regard himself as one of them. But this, it would appear, is the price he has to pay to salvage his appeal to the distinctive value of personal detachment.

6. Contract as a Form of Technology

To briefly recap. According to Kimel’s sophisticated inflationism, the noninstrumental significance of contract lies in the value of personal detachment. Specifically, the idea is that, without it, we would not have the option to proceed in our special relationships in a way that reflects our individuality, or the deepened meaning said to come with it. This view, though attractive, faces a counterexample that can only be avoided by retreat to a form of value pluralism that empties it of its emancipatory potential.

Of course, this is bad news for Kimel’s inflationism. But it may strike us as bad news quite apart from that debate. After all, it is plausible that contracts constitute our having the option of personal detachment, and that this really can be worthwhile in the manner proposed, i.e. that it can make our relationships more meaningful by giving us control over the extent of their jurisdiction in our lives. We may reasonably not want to give up on these ideas. Can we salvage Kimel’s appeal to personal detachment in contract theory in its original emancipatory spirit?

The discussion of §5 brought out several value-theoretic desiderata that any such attempt would have to meet. To start with, for old-style relationships to be such as to have their meaning deepened by the option of personal detachment, (i) they must genuinely have had a meaning to deepen, at least once upon a time. But here and now, in the twenty-first century, where we are lucky enough to have the option of detachment, old-style relationships can no longer be genuinely worthwhile: they smack of traditional authority, are outdated, and are—in a word—old. As such, (ii) it must be that only new-style relationships can be genuinely worthwhile. Nonetheless, (iii) it must also be that not just any old exercise of the option of detachment necessarily makes for a genuinely valuable new-style relationship. Otherwise we will have to admit that where detachment proceeds in an unthinking or excessive manner, as in the counterexample, the alienated relationships that result are worthwhile in their own way too. Desideratum (i) falls out as a presupposition of how personal detachment is said to receive its significance for us, whereas (ii) and (iii) are needed to avoid the value pluralism discussed before, and the political impasses therein contained.

I would like, in closing, to explain how we might get results (i)–(iii) out of an appeal to personal detachment in contract theory. Admittedly, the story will be impressionistic. In my defence, it is not really my job to supply it. I offer it in a conciliatory spirit, a way to preserve what is attractive in a view I have otherwise been concerned to argue against. We will see that it makes no difference to the deflationary ambition of the paper whether my story is persuasive or not. The present context is one in which something—even something fairly sketchy—is better than nothing.

My story takes shape via reflection on the relations between values and technology. We have been understanding noninstrumental values as the sorts of things that we are able to benefit from directly by engagement with them for their own sake. Suppose, with Joseph Raz, that specific noninstrumental values are (a) partly constituted by attendant forms of excellence and the standards of successful engagement implicit in these (2003: 31–32), which (b) are as a rule ultimately dependent for their existence on social practices that do or did once sustain them (19). These assumptions about value allow us to make the following structured observations about the potential impacts of technology on values and our engagement with them.

The introduction of a new technology into society will have many effects, some of which touch on values and our engagement with them in various ways. Most obviously, there will be intended effects, at least if it is a good technology that actually achieves the purpose it was designed to fulfil. What this looks like for specific technologies will vary from case to case, but it seems plausible that, in general, the purpose of technology is to enable us to better do things that facilitate our engagement with what is of value. When a good new technology becomes available, it makes excellence in our engagement with valuable things more attainable, at least initially. On top of these intended effects, there will be unintended effects too. Some of these relate to the technology’s achievement of its purpose, and others not. We will focus on just those effects of new technology insofar as it is good at what it was designed to do, which, as we have just seen, will include both intended and unintended effects.

My claim is that the intended effects of introducing good new technology will have unintended effects on the very value that it succeeds in helping us engage with. Consider the impact of artificial intelligence (AI) on chess. Chess has been a great game ever since it was invented. But it appears that the introduction of neural net AI engines is changing the value of chess at the elite “super” grandmaster (GM) level. Just like any technology, the purpose of AI chess engines is to enable us to better do things that help us engage with what matters—in this case chess. They do this very well: they offer deep evaluations, calculate the best lines, show us what (not) to do in a given position, giving a rationale, and so on. All of this helps us to play excellent chess. AI engines have become ubiquitous in the chess world for precisely this reason: every chess player these days has access to one. At the super GM level, high-end engines are available to help gain an advantage in classical chess, a format that remains prestigious within the chess world. As developers compete to write better engines, super GMs stay up-to-date to maintain their edge. As the limits of AI engines—or at least human players’ ability to assimilate their insights—become clearer over time, super GMs and their teams have begun to experiment with how to integrate the technology into training for the best results in classical chess. These are some of the unintended effects of the success of chess AI engines at fulfilling their purpose. They are effects on the social practices that sustain the value of chess.

As the social practices that sustain the forms of excellence in an activity change, so too, I claim, does the value that those forms partly constitute. In particular, what counts as excellence in how we engage with the relevant value comes to include standards of propriety in the application of the new technology. To see this, consider the practices of super GMs and their teams today. Having the players use AI as much as possible—for example by trying to memorize as many top AI lines for as many moves of as many openings as possible—has not emerged as a promising way to gain an advantage over-the-board. What works better is for the players’ teams to research themes and concepts based on AI analysis, to be humanly explored—including for theoretically weak lines that harbour opportunities for a human opponent to make mistakes—by players who themselves studiously avoid contact with AI engines.¹⁹ In other words, what it means to play classical chess at the super GM level has come to include the sophisticated use of AI; and that, I submit, makes the value of chess today fundamentally different from the chess of old. Now that we have AI, there is a clear sense in which a super GM who pines after the old chess would be missing the point of chess as it exists today.²⁰ This is not to deny that chess back in the Golden Era was a great game. It is to say that super GMs (or their teams) who systematically shun AI simply fail to grasp what is good about the only chess that exists any more. The value of the chess of old has not just been replaced but superseded by that of the new chess.²¹

We can understand how the new availability of the option of personal detachment interacts with the value of our personal relationships in an analogous fashion. Perhaps once upon a time it was possible for us to find meaning in old-style relationships, just as chess really was great in the Golden Era. But when we acquired the option of personal detachment and the associated control over the reach of our old-style relationships, this really did enable us to better have lives fit for ourselves as individuals, as was the express purpose of our having the option. As more and more people came to exercise the option, the social practices sustaining the value of old-style relationships changed, transforming that value. What counts as doing well in this new style of relationship—the attendant forms of excellence—has come to include standards of propriety in how we exercise the option of detachment.

Notice how this analogy satisfies desiderata (i)–(iii). It straightforwardly satisfies (i), since it allows that old-style relationships were genuinely meaningful once upon a time. Moreover, just as in chess, the value of new-style relationships is said not only to replace but to supersede that of the old. So the analogy also satisfies (ii). Now that we have the option of personal detachment, only the new-style relationships can be truly worthwhile. There is really something tragic about those social conservatives who are nostalgic for personal relationships expressive of traditional values: in refusing to get with the times, they are condemned to miss the point of the only style of meaningful relationship that exists for them to participate in here and now. And finally, (iii) is satisfied, because the analogy suggests that figuring out how to integrate the option of personal detachment into our relationships to secure excellence in the new style is likely to involve sensitivity, thoughtfulness and experimentation. Just as the wanton use of AI is not a way to excel at the new chess, we should not expect that wanton detachment will be a way to excel in our personal relationships as they exist today. Those high-powered professionals who choose to hollow out every aspect of their relationship to their kids need not be doing a good job as parents in the new style.

Accepting a story along these lines allows us to salvage the emancipatory appeal to the value of personal detachment in contract theory. We can say that, without contract, we would not have the option of personal detachment, and we would thus be deprived of the deepened noninstrumental significance of personal relationships as they figure in our lives today. And we get to say this without conceding to social conservatives or proponents of laissez-faire that whether we prefer our style of relationship over theirs is, in the end, no more than a matter of taste. But to get the story going—and this is the point—we had to analogize the option of personal detachment with AI engines in chess. Of course, this presupposes that, like AI, the option of personal detachment is a form of technology and is, as such, first and foremost of instrumental value.

The consequences for the contract-theoretic appeal to personal detachment should be clear. Recall how that appeal is supposed to work. Contracts are said to be an essential constituent of the value of the option of detachment, deriving their own value from this source. If that is so, then, on the present story, the value of contracts will of course turn out to be “merely” instrumental. To reinvigorate the emancipatory potential of Kimel’s contract theory, we ended up telling a story that deflates it.

7. Conclusion

The broadly liberal tendency towards inflationism in contract theory says that, without contracts, we would be deprived of certain noninstrumental values, so the contribution contracts make to human life is itself noninstrumental. This paper focussed on Kimel’s inflationism, which appeals to the somewhat pessimistic but decidedly liberal value of personal detachment. I unpacked several possible readings of his view, arriving at its most promising version: the value of our having the option of personal detachment lies in how it deepens the meaning of our existing personal relationships in a sui generis way. I argued that, even on this reading, the view faces a Fraserian counterexample, which it can only avoid by forsaking whatever emancipatory potential it may have had. I offered an impressionistic story about the relations between values and technology that may be used in a last ditch effort to rescue what is attractive in Kimel’s view. But it is a story that makes contract a form of technology, casting its value in a “merely” instrumental role. It seems fair to say, then, that the contract-theoretic appeal to personal detachment faces significant deflationary pressures.

Acknowledgements

I am grateful to the Philosophy, Law & Politics Graduate Forum for the opportunity to present this material, and to Veronica Rodriguez-Blanco and Anna Stelle for prepared responses on that occasion. Thanks also to Showkat Ali, Daniel Guillery, Rowan Mellor, Filippa Ronquist and Nikhil Venkatesh for feedback at a meeting of UCL MOPWiP, and to two referees at Ergo for excellent comments. Special thanks to Ulrike Heuer and Prince Saprai for their support and guidance, and to Nina Bang David for valuable conversations about politics. This work was financially supported by the AHRC (grant reference number: 2387877).