Feasibility and social rights

Abstract

Social interactions and personal relationships are essential for a minimally good life, and rights to such things – social rights – have been increasingly acknowledged in the literature. The question as to what extent social rights are feasible – and properly qualify as rights – however, remains. Can individuals reliably provide each other with love and friendship after trying, for instance? At first glance, this claim seems counterintuitive. This paper argues, contrary to our pre-theoretic intuitions, that individuals can reliably provide each other with such relationships, rendering even “thick” social rights feasible. This conclusion challenges the assumption that such things cannot be reliably provided after trying, and suggests that a surprisingly wide class of social rights are feasible. Claiming relationships characterised by love or friendship as a matter of justice, therefore, is possible, and our theories of justice should appropriately widen to accommodate this fact where appropriate.

Feasibility and social rights ¹

It is typically thought that individuals have rights to (at least) that which is necessary for the living of a minimally good life (Nickel, 2007). We can make sense of rights to subsistence, shelter, safety, and, perhaps, even bodily organs (Fabre, 2003) through this notion. Recently, there has been a growing literature defending the idea that people require social interactions and personal relationships in order to live a minimally good life (Brownlee, 2013, 2020; Liao, 2006, 2015). This thought is an intuitive and empirically supported one (Cacioppo, 2002; Coyle and Dugan, 2021; Eames et al., 1993). It is also especially persuasive in the context of children given their developmental requirements. ² As a result, there is an increasingly popular view that if we think persons have rights to things that are necessary for a minimally decent life, they also have rights to social interactions and personal relationships: they also have social rights. ³ Examples of such social rights might include the right to love, the right against social deprivation, the right to friendship, or the right to a fair opportunity to form social bonds.

However, as has been made clear elsewhere (Gilabert, 2018), having a sufficiently weighty interest in something is not enough in and of itself to ground a right. This is because (claim) rights infer duties (Cook, 1919) and, if there is no duty that can feasibly satisfy the claim, there can be no right (Southwood, 2016). This fact is a particular issue for social rights because, in many places, they also infer rights to relationships: social interactions characterised by relational emotions such as friendship, love, and/or romance. Traditionally, relational emotions have not been seen as “commandable” in the way needed to be eligible for duties (Gheaus, 2017; Kant, 2017; Liao, 2006) – if this is true, then it seems as if there could be no duty to provide another with a personal relationship, given the unfeasibility of doing so at will. As a result, it might be that social rights are specially unfeasible – and therefore non-rights – given the claims they aim to realise, irrespective of how important such things are for individuals’ interests.

Surprisingly, there has yet to be a sustained treatment of whether social rights meet the feasibility condition, and qualify as rights (for some engagement, see Brownlee, 2013; Gheaus, 2017; Liao, 2006; Valentini, 2016). This is a problem given that coming to a view on this matter is crucial: not only to clarify the current potential of social rights qua rights, but also to engage with worries that social rights are especially unfeasible in a way that renders them non-existent. This paper seeks to provide an answer to this question, asking: are social rights feasible?

For the sake of argument, I take a couple of things for granted.

First, I remain agnostic on whether or not social rights as I characterise them here are attractive all things considered. There may be reasons – at the bar of freedom of association (Brownlee, 2016a; White, 1997) for instance – for why such rights should not be realised. The matter of attractiveness, however, is orthogonal to the question of whether they could be realised, which is the focus of the present discussion. I bracket that other set of problems.

Second, I do not wish to give a definitive account of why social rights are justified in the first place, and remain agnostic – as far as I can – about the precise set of interests that ground social rights. These matters, again, are not crucial for the present question and I set them to one side.

Third, I am not seeking to establish that social rights are generally feasible in all cases, but only that they are not especially unfeasible in a way other (accepted) rights are (presumably) not. For what it is worth, I understand the prior to be true if the latter is true. However, this view is not uncontroversial(Eddy, 2008; Gilabert and Lawford-Smith, 2012; Lawford-Smith, 2013) and, in light of this, I opt for the more limited claim, leaving readers to draw their own conclusions on the larger question of what we might call the “all things considered” feasibility of social rights. For this reason, I refer to a right as being “not specially unfeasible” and “feasible” throughout the paper, using these different phrases to better suit the sentences they find themselves in, even though their substantive meaning is assumed to be the same.

Section 1 defines feasibility in order to provide a sufficiently precise benchmark from which social rights can be analysed and tested against. Section 2 offers a conceptual taxonomy of social rights, distinguishing between thick and thin social rights. Section 3 pursues the question of whether thin social rights meet the feasibility condition, arguing that they do. Section 4 pursues the question of whether thick social rights meet the feasibility condition, concluding that the answer is complex, but thick social rights are ultimately feasible. In Section 5, I offer an additional reflection on the feasibility of thick social rights that seeks to show, definitively, that they are not specially unfeasible given the potential for collective social duties. ⁴ In Section 6, I respond to a number of important objections that could be pressed against the account. Finally, in Section 7, I briefly restate the overall position by reflecting on the sometimes enduring unfeasibility of social rights.

Section 1

Before we get into the business of determining whether social rights are feasible, it pays to get clear on what we mean by feasibility. For the purposes of the present discussion, I take it for granted that rights – in order to be such – must be feasible: there must exist a prospective set of plausible ⁵ duties that persons can be placed under to meet the claim in question (Southwood, 2016). However, there is controversy on what it means for a duty to be plausible in this sense. Indeed, as I understand it, defining plausibility in this context is the raison d’etre of the wider feasibility debate in the literature. ⁶ It is this central question that will organise our thoughts when delimiting our feasibility condition: what does it mean for a duty to be plausible, and, therefore, the basis of a feasible right? Answering this question is important because once we get clear on what a plausible (feasible) duty is, we will have a sufficiently clear test for whether or not social rights can be considered as such.

One popular view on how we should understand feasibility is simple possibility: a duty is plausible if it requires action on the part of the duty-bearer which holds some possibility of success in realising a correlative claim. The test of feasibility on this view, then, is to investigate whether there is some act open to the prospective duty-bearer that could realise the right in question. To illustrate: if someone has a claim on a sip of my water, and it is possible – i.e., not impossible – for me to pass it over to them, then it is feasible for me to do so. As a result of the act being feasible, I can then be under a duty to perform the action.

This possibility notion of feasibility, whilst holding some intuitive attraction, must be rejected. Specifically, possibility must be rejected because it would deem plenty of acts as feasible which do not square with our considered judgments on plausibility. To motivate this point, we can offer the following case(Southwood and Brennan, 2007; Stemplowska, 2020): it may be possible (by fluke) that a philosopher could successfully perform open-heart surgery, but this act is not plausible for her given the extremely low likelihood of success. As a result, we cannot say that the philosopher can have a feasible duty to perform the surgery. By corollary, there can be no right of the patient to have it performed by the philosopher. We can also use a modified version of our earlier “sip-of-water” case to motivate this point: if I can only give a claimant a sip of my water if I make a 147-point break on a snooker table, it might be possible for me to do so, but it is clearly not feasible for me (unless I am Ronnie O'Sullivan, which I am not). As a result, it cannot be feasible for me to provide the claimant with a sip-of-water despite it being possible, and there cannot be a bona fide right to it as a result.

If one wished to reject this view, they would have to claim that persons could be sometimes under a duty to perform miracles, but this seems absurd: how could one be under a plausible duty to do what it is possible-but-virtually-impossible to do? Therefore, as a result of possibility not being able to leave out possible-but-highly-unlikely acts – miracles – from the universe of the feasible, and thereby the scope of duty, we must reject the possibility account of feasibility, and look elsewhere for our testing criteria for social rights.

Another prominent view on feasibility is likelihood: a duty is plausible if it demands action with a sufficiently high likelihood of eventuating. To test if a right is feasible, on this view, is to investigate how likely it is under current conditions that the right in question could be met and then determine if that level of likelihood is sufficiently high.

One way of understanding this notion of feasibility is to flip it on its head. As the point is put by Estlund, an action is unfeasible if it is “impossible” (Estlund, 2014). This seems fairly non-controversial. However, this could mean one of two things: a) the act is impossible because it cannot be done (possibility); or b) there is zero chance of the act being performed given certain facts about the world (likelihood). Indeed, securing world peace might be unfeasible given unchangeable constraints on resources, or immutable human nature (possibility); or it could be unfeasible because there is no chance anything will be done about it given a lack of motivation on the part of the powerful (likelihood). The likelihood interpretation of feasibility is concerned with the second set of considerations: it says that an action is feasible – and therefore can qualify as a duty – if and only if there is a sufficiently high likelihood that its aims will be successful given the full consideration of current conditions ⁷ pertinent to an act's likelihood.

This view of feasibility escapes the pitfalls of possibility given that it would rule out the surgery and snooker cases, but it must still be rejected because it would deem unfeasible plenty of actions that are plausible for an agent to perform, even if unlikely. For instance, it is highly unlikely (impossible) that I or others would perform a chicken dance in front of my students, but it is not right to say that I could not plausibly perform such a dance (Estlund, 2014). Moreover, it is highly unlikely that the richest corporations in the world will start paying their fair share of tax, but to say it is implausible for them to do so is wrong given the relative ease at which they could perform the act if they so tried. As a result, likelihood also fails as a definition of plausible: it reconciles us too much to extant facts about our world that normative theorising and notions of rights and duties should not be sensitive to, such as mere reluctance. In other words, given that likelihood wrongfully treats as “forces of nature” ⁸ plenty of acts that are under our control as responsible moral agents, it makes the wrong cut between what is and is not plausible, and must be rejected accordingly.

Luckily for us, a superior view of feasibility has recently been elaborated which keeps the intuitive appeal of both prior accounts whilst avoiding their pitfalls: the incentives account of feasibility. The incentives account holds that: a duty is plausible if “there is an incentive I such that, given I, X is likely to [successfully perform the duty]”(Stemplowska, 2020).

Underneath the formal definition of the account, the “cut” made by testing for incentives is between feasible acts which both a) have a high likelihood of success and b) can be attempted; and unfeasible acts which c) have a (sufficiently) low likelihood of success if tried and/or d) cannot be tried at all given a lack of ability to try. ⁹ Testing for the existence of incentives that can induce successful trying does the primary work of differentiation between plausible and implausible acts on this account because such an incentive can only exist if a) one can attempt to perform the act, ¹⁰ and b) there is a sufficiently high likelihood of performing the act successfully. Indeed, you can throw all the money in the world at someone, but this will not enable them to turn water into wine; on the other hand, giving people money could get them into opera-houses. In short, as I understand it, the account uses the notion of incentives as something of a proxy for a more subtle test that can differentiate (by its own lights) what it is and is not plausible for an agent to do.

This account has great appeal. It can make sense of why we think it implausible for a philosopher to successfully perform open-heart surgery despite it being possible, and why it is plausible for a shy lecturer to do a chicken dance in front of their students despite it being incredibly unlikely. It gives us the correct result in the surgery case because there is no incentive that would increase the likelihood of successfully performing the surgery to such a point that would make it feasible. Indeed, we cannot have a duty to perform miracles, and the incentives view correctly agrees. It would also give us the right result in the chicken dance case because its unlikeliness is determined by mere reluctance, and it is, therefore, plausible to think that incentives exist which could change the willingness of the lecturer to successfully dance. Indeed, if he was given an ultimatum or offer: “be fired or do the dance!”, or “do the dance and be given £1m!” there is a high chance the act will be attempted successfully. ¹¹ As a result, the incentives view can also give us the correct result in those cases where the likelihood of a duty being performed is low because of mere reluctance, as opposed to general inability.

The cut made by incentives on the plausibility question, as exemplified by the prior examples, therefore, fits well with our wider judgments on what a normative theory should and should not be sensitive to at the level of duty. Moreover, it escapes the pitfalls of other major accounts whilst building on their appeal. For these two reasons, I understand it as the preferred account of plausible duty, and define feasibility here in line with it. It should be noted, however, that the incentives account of feasibility is very similar to the conditional account of feasibility offered by Estlund which, by all accounts, is more widely accepted. That account claims that v-v-ing is feasible if there is a sufficiently high likelihood that an agent will successfully v conditional on them trying to v. The main difference between these two accounts is that the conditional account does not provide a test for when it is feasible for an agent to bring themselves to try to v, whereas the incentives account does. Consequently, the incentives account will class a more narrow range of acts as feasible compared to the conditional account, but this will be a fairly minor thing. In any case, it seems that if we can show that social rights are feasible under this comparatively more stringent metric of feasibility, it will also show that they are feasible according to the conditional account a fortiori.

A final clarification: some have claimed that the incentives account provides a poor test of feasibility because, even if there were some incentive I that could motivate A to v, if it is not feasible to give I to A, and then v is still not feasible, even though the incentives account would deem it so. This objection fails because it relies on a conflation between motivational unwillingness to v and motivational inability to v (Stemplowska, 2020).

Motivational unwillingness to v is characterised by an agent being able to v successfully after trying, but not v-ing even though they could if they willed themselves to. An example of this is a case in which Andy could swim in the sea and save Byron from drowning if he tried, but he is lazy, and would prefer to stay lounging on the warm beach. In this case, it is clearly feasible for Andy to save Byron, even though it is unlikely.

Motivational inability to v is characterised by cases where an agent could successfully v after trying, but cannot bring themselves to try to v given some kind of hard impediment they face which blocks their will-to-v, despite their desire to. One clear example might be an intense arachnophobe who cannot bring themselves to flick off the deadly spider that is about to bite their child's hand. Clearly, it is not feasible in such cases for the individual to v, and this differentiates it in this regard from the previous set of cases.

The incentives account provides a test that helps us (among other things) differentiate between these two classes of case. The thought is that, if there were some incentive that could bring an agent to v, this would tell us that we are not dealing with a case where v-ing is impeded by motivational inability – like the arachnophobe case – but, instead, motivational unwillingness, given that incentives are things which change the appeal of a set of options across a given set. For this reason, it does not matter if the incentive can actually be provided or not – all that matters is that the agent could be brought to v if it were credibly posed. This would tell us that Andy can feasibly save Byron, given that threatening him with a bucket of cold water would get him to rescue, whereas the arachnophobe cannot feasibly slick the spider given that there is plausibly no incentive which could bring her to overcome her paralysing fear of spiders. This is why the above objection is not decisive.

Now we have our notion of feasibility in-hand, and a metric by which to test social rights, we can address our main question: are social rights feasible?

Section 2

Before getting into the meat of the question, we must first get comfort on what social rights are. As defined here, social rights are claim-rights to social interactions and/or personal relationships. By way of unified definition, we can say that social rights confer upon an agent the ability to place another under a duty to provide them (or another) with social labour.

Taking this definition forward, we can isolate two kinds of social labour that are – as will become clear later – substantively and meaningfully different from one another at the bar of feasibility: a) functional social labour and b) affective social labour. ¹²

Functional social labour refers to social interactions that do not require any particular affective disposition to be successfully performed. We can think of having a conversation as a form of functional social labour given that conversing is a) a form of social interaction and b) does not (necessarily) require any form of affective disposition to be successfully performed. Indeed, two people who are completely indifferent to one another in any affective or relational sense can still have a very successful conversation, as is evidenced by polite conversation at dinner, or the interaction between a teacher and a student.

Affective social labour, on the other hand, is the name given to social interaction that does require an affective disposition to be successfully performed. We can understand personal relationships – such as those characterised by love, friendship, romance etc. – in this way given that being in a personal relationship is a) a form of social interaction and b) does necessarily require a set of (appropriate) affective dispositions to be manifested. For example, to share a loving relationship with another individual, the emotion of love (and the valuing of the object of that affection as such) must be present: simply performing “loving” acts is not sufficient. Similarly, to have a genuine friendship with another individual, platonic attitudes and emotions towards another must be present – solely performing friendship-like acts without the appropriate affective dispositions and/or valuing intentions would mean that there simply is no friendship. ¹³

For ease of exposition, I will refer to social rights that lay claim to the functional social labour of others as thin social rights, and social rights that lay claim to affective social labour as thick social rights. To put this in the language of duties: thin social rights are understood as only requiring the functional social labour of a duty-bearer to be satisfied; thick social rights require (solely, or perhaps in addition to functional social labour) the affective social labour of a duty-bearer to be satisfied. A good illustration: the right against social deprivation is a thin social right, whereas the right to be loved is a thick social right.

Interestingly, these two kinds of social labour – and, consequently, social right – are often conflated in the literature, which has led to confusion about what particular social rights are claims to, and how feasible they are. For example, in Brownlee's “On the Right to Be a Social Contributor” (Brownlee, 2016b), social contribution is defined in a way that includes both functional social labour – “ordinary decent human interaction” – and affective social labour – “loving relationships marked by mutual investment, intimacy, and care”. As a result, the right Brownlee argues we have is necessarily mixed, but unacknowledged as such ¹⁴ : it requires the performance of both functional and affective social labour on the part of duty-bearers to be satisfied. Given the conflation of these different types of claim under a single right, however, the important differences between the constitutive claims in the right at the bar of feasibility can be missed, and lead to confusion.

Indeed, this lack of conceptual clarification elicited a response to Brownlee by Valentini (2016), who claimed – among other things – that Brownlee's right to be a social contributor fails in an important sense because claims to “meaningful social relationships” are unfeasible. Whilst it is not entirely clear what Valentini means by “meaningful” in this context, if we interpret “meaningful” as referring to relationships characterised by affective social labour– which we have reason to think given her focus on friendship as an example of such a relationship – this only reckons with one half of Brownlee's right, leaving it an open question as to how attractive and efficacious the right to be a social contributor is all things considered. By offering a clear conceptual distinction between thin and thick social rights – and their respective labour-claims – analytic ambiguity such as this can be avoided in the social rights debate, providing clarity to defenders and objectors of social rights at the bar of feasibility and beyond. ¹⁵

To get clear on just how feasible social rights are, then, it is not sufficient to test social rights sui generis. We must test to what extent each kind of social right can satisfy our incentives requirement: are social rights of each type feasible?

Section 3

Let us first deal with the question of feasibility as it pertains to thin social rights. As we have already stipulated, thin social rights are claim rights to the functional social labour of other persons. Therefore, the specific question we are faced with here is this: do incentives exist that could motivate people, generally speaking, to successfully provide others with functional social labour?

The strongest argument in favour of thinking such incentives exist is the argument from induction. Virtually all positive claim rights require functional labour to be realised ¹⁶ , given that positive rights place other people under a duty to do something. Examples of positive rights include the right to rescue, the right to subsistence, and the right to security, among others. We do not consider positive rights to be specially unfeasible given their inference of correlative (positive) duties to perform functional labour: if I have a duty to pass the salt, and I can easily reach it and slide it over to the claimant next to me, why might we think there is something specially unfeasible about me having to perform the act given that it requires functional labour to pass the shaker?

Furthermore, there are plenty of more complex or burdensome positive duties that require functional labour to be successfully realised, but are nevertheless not specially unfeasible: paying one's taxes, meeting realistic deadlines, caring for children, and doing one's fair share of the housework, for instance. As a result, it is not right to conclude that positive rights qua claims to functional labour are specially unfeasible. Consequently, there is good inductive support for the idea that thin social rights are not specially unfeasible, either: if functional labour sui generis does not make an act specially unfeasible, there is a presumptive case in favour of generalising this point to functional social labour, and thin social rights.

Whilst there might be some instances of social functional labour that are unlikely to be successful after trying – delivering a rousing speech if one has social anxiety, or intentionally making someone laugh hysterically without any comedic talent, for instance – there is no reason to think that these specific transgressions of feasibility within the realm of functional social labour indicate any general unfeasibility with functional social labour as a type of act. Indeed, there are plenty of acts of generic functional labour that are specially unfeasible, but we do not think these are indicative of a more general unfeasibility attached to acts of that broader (functional) type: performing open-heart surgery untrained; staying awake forever on sheer willpower; always rolling a 6 when you throw dice; or turning water into wine. Moreover, there are plenty of acts of functional social labour that are clearly feasible in this general sense: keeping a lonely person company; having a conversation with someone; going for dinner with someone; or making sure to visit someone regularly. Moreover, the existence of professional industries which offer functional labour (babysitters; social workers; teachers; servers) underlines the plausibility of this thought even further, and clarifies that incentives do exist that could move persons to perform such acts.

Together, this shows that likelihood after trying, whilst of course setting limits on what kinds of functional social labour can be claimed as a right, is not a specially salient issue for the validity of thin social rights. Just as many non-social rights are unfeasible because reliant upon miracles to be realised, so too are some thin social rights, but this gives us no reason to think feasibility concerns at the bar of likelihood renders thin social rights non-existent.

Some may respond that, whilst it might be true that there are some instances of thin social rights which are feasible on the grounds of likelihood after trying, we cannot know how successful this point is in defusing worries about the feasibility of thin social rights generally speaking until we know what sorts of functional social labour persons have presumptive claims to. Indeed, plenty of acts of functional social labour might be feasible, but if thin social rights only lay claim to acts that fall outside the scope of feasibility – given these unfeasible acts are the only kind that would satisfy a weighty interest – thin social rights will still be specially unfeasible as a class of positive rights. This is a genuine concern.

However, the force of such a worry – whilst legitimate – is severely limited given that all dominant accounts of thin social rights do not require any specially unfeasible acts at the bar of likelihood.

For instance, Brownlee's articulation of the right against social deprivation demands only that individuals have ongoing access to “minimally adequate opportunities for decent… human contact including interpersonal interaction, associative inclusion, and interdependent care”(Brownlee, 2013). As far as I can tell, this amounts to a claim on the social labour of others in the form of conversation, company, group activities, and mutual (functional) care. If correct, there is no obvious reason to think that the right against social deprivation is specially unfeasible qua thin social rights at the level of likelihood: what is so unfeasible about keeping someone company, speaking to them, including them in social activities, and/or taking a communicated interest in their wellbeing? Indeed, these actions are all properly part of the ambit of plausible action insofar as persons regularly successfully engage in them after trying in a range of contexts underpinned by various incentives. Moreover, this right is anything but trivial: without access to and enjoyment of these things, there is plenty of evidence that persons will live a less than minimally good life in the starkest possible terms. ¹⁷

As a result, the claim that only unfeasible acts of functional social labour are those which are claim-worthy is false: critically important thin social rights do not demand the impossible, and are properly understood as feasible in a comparable way to other types of positive right at the bar of likelihood as a result.

Together, the two criteria of feasibility being satisfied in the case of thin social rights defuse any objection to the argument from induction and, indeed, leaves it well-grounded and robust. The idea that thin social rights are specially unfeasible compared to other kinds of positive right are wrong and unfounded. As a result, we can conclude that thin social rights do, in fact, meet the feasibility condition, and they do so straightforwardly. While this may not come as a surprise, theorists prominent in the social rights literature have assumed that the preceding analysis (or something like it) is true without doing the careful work needed to know such acts are feasible. By doing this work, we now have comfort over the feasibility of thin social rights, and can understand why they are indeed feasible.

Section 4

Let us now deal now with the question of feasibility as it pertains to thick social rights. As previously stipulated, thick social rights are claim rights to the affective social labour of other persons. Therefore, the specific question we are concerned with here is this: do incentives exist that could motivate people, generally speaking, to successfully provide others with affective social labour?

The biggest challenge to thinking such incentives exist is the argument from spontaneity (inspired by Gheaus, 2017). This argument claims that emotional states are not subject to the will, and can only be produced spontaneously: involuntarily and unpredictably in response to some kind of quite specific stimulus typically unbeknownst to the relevant normative constituency. If it is true that emotional states are outside the scope of the will in this way, this would mean that plausible incentives which could bring about affective states in prospective duty-bearers could not exist, ¹⁸ rendering thick social rights unfeasible. This thought goes back to at least Kant (2017), when he suggested that: “Love is a matter of feeling, not of willing, and I cannot love because I will to, still less because I ought to”.

This conjecture also has considerable intuitive appeal across hypothetical and real-world cases. Indeed, we can imagine a case in which an individual is held up by a highwayman who declares: “your love or your life!” In this case, we do not think that the victim of the mugging could will themselves into a state of love (or friendship, romance, kinship, etc.) towards the attacker, regardless of the strength of the incentive. Further, there are plenty of real-world cases in which persons cannot bring themselves to love their children, regardless of their earnest attempts, and their common feelings of sadness and shame around the fact. ¹⁹ At the nicer end of the spectrum, giving an individual money to love might be able to induce love-like acts – given they are instances of functional social labour – but the suggestion that monetary incentive could regularly produce love-proper is a speculative and bizarre one that seems to cut against the “modally robust” (Pettit, 2015) nature of what we understand it to mean for someone to be in a loving relationship. Given the force of these cases, we have good initial reasons to think that incentives do not exist which can produce successful attempts to provide affective social labour, and, as a result, we have a strong initial case in favour of dismissing thick social rights as specially unfeasible non-rights.

We can articulate the argument from spontaneity, shot through with a feasibility condition framing, like so:

P1) For an act x to be feasible, it must be responsive to an incentive* that can produce (in all likelihood) a successful attempt of x.

P2) Thick social rights require the performance of affective social labour.

P3) Affective social labour is non-responsive to incentives* given its necessary spontaneity. ²⁰

C) Thick social rights are unfeasible.

Does the argument from spontaneity succeed?

The most valiant attempt currently found in the literature to refute the spontaneity objection is offered by Liao in his articulation and defence of the right (of children) to be loved.(Liao, 2006). He argues that, despite initial intuitive support for thinking that the duty to love is unfeasible, there are two ways for individuals to successfully induce emotional states towards others through trying: a) internal control; and b) external control. The internal control of emotions centres on actions which seek to either enhance or remove certain rationalisations that affect one's propensity to feel an emotion towards another. This can include a) reflecting on one's reasons for not having a particular emotion towards another, or b) reaffirming to oneself the reasons why one ought to have a given emotion towards another. An example of the former is seeking to untangle discriminatory attitudes and prejudices; an example of the latter is reaffirming to oneself the importance of a pro-tanto claim another has to your affective social labour, such as in the case of parents loving their children.

The external control of emotions centres on actions that exhibit selectivity concerning when, why, how, and where you interact with others. Two broad types of act under this category include a) placing oneself in situations and circumstances that are conducive to bond-forming; b) removing oneself from situations and circumstances which might alienate or otherwise lead to being “put-off” from a prospective claimant. An example of conducive environments might be those that are centred on a shared interest or hobby; an example of un-conducive environments might be those where individuals are stressed or tired, or where the disagreements and tensions between individuals are highlighted.

Liao argues that, taken together and applied repetitiously over a long period of time, it is possible for individuals to successfully “command” affective social labour through emotional control: by interrogating one's own affective preferences and beliefs on an ongoing basis, and consistently placing oneself in situations that maximise the potential for bond-forming, Liao states that it is feasible for persons to successfully will themselves into a loving state towards a specific other and other sorts of affective relational states. Therefore, despite initial thoughts, it is argued by Liao that the spontaneity objection fails to show that thick social rights are unfeasible: internal and external control, over a sufficiently long period of time, can (non-spontaneously) enable one to perform affective social labour successfully after trying.

If this argument goes through, it would show that we have good reason to reject the spontaneity objection, and deem thick social rights as meeting the feasibility condition: one could be incentivised to perform internal and external control and this, in turn, would amount to an eligible incentive to perform affective social labour, and, as a result, a feasible duty to do so.

Whilst promising on first glance, I think Liao's rebuttal to the spontaneity objection requires more argument. This is for two reasons.

First, Liao assumes that the feasibility of the right to love – and presumably other thick social rights – is secured so long as it is shown that it is possible for agents to successfully perform affective social labour. However, as was established in Section 1, this is the wrong metric of feasibility. For such labour to be feasible, it must not only be possible for one to perform, it must be a) sufficiently likely for one to successfully perform after trying, and b) pose no special problem at the bar of agents’ ability to make an attempt. As a result, even if we accept – for the sake of argument – that it is possible for some persons to perform affective social labour successfully at will, it has still not been established by Liao whether performing such labour is feasible. This leaves the feasibility of thick social rights in the context of the spontaneity argument ambiguous.

Second, whilst Liao's notions of internal and external emotional control are instructive and intuitively plausible, he does not corroborate his claims with any empirical evidence. This leaves his reliance on the efficaciousness of emotional control – and whether it, as a concept, has any action-guiding potential for prospective duty-bearers – in an insufficiently well-grounded position.

For Liao's rejection of spontaneity to succeed, then, both of these shortcomings must be overcome. Of course, these issues are also related: if it can be shown that robust empirical evidence does exist which proves emotional control as being sufficiently efficacious (to render the performance of affective social labour feasible), both points of concern will have been settled. Interestingly, such empirical support does exist, showing that Liao's account is effective in defusing the main thrust of spontaneity, even when we hold it to the feasibility-standard of incentives.

For example, there is a comprehensive dual-study that shows how the desirable “relationship traits” typically found within arranged marriages reliably induce strong feelings of love between spouses across various religious and cultural traditions (Epstein et al., 2013). Such traits are also understood as being predictively effective in other types of marriage and romantic relationship, too (Ibid.). This bolsters the plausibility of internal and external control as a bridge to the feasibility of thick social rights because it highlights the existence of cross-cultural patterns of choice – “commitment”, “sacrifice”, “communication”, to name three –which can effectively foster desired affective states towards specific others, even when starting from a place of indifference (or, at the very least, “non-love”). These mechanisms are also forms of thin social labour in the first instance, meaning that they are not specially unfeasible.

There are also a number of studies that highlight the existence of comparable forms of effective emotional control in the context of friendships (Hallinan, 1978; Marmoros and Sacerdote, 2006; Rusbult, 1980): individuals who spend frequent time with each other, share a close geographic proximity (Marmoros and Sacerdote, 2006), and reciprocate towards one another (Rusbult, 1980) are reliably likely to become friends, even when starting from a place of non-friendship. Whilst class and race are strong determinants of friendship formation (Marmoros and Sacerdote, 2006; Moody, 2001; Wright and Cho, 1992) – and may set certain constraints on the efficaciousness of emotional control across socioeconomic contexts – their impact can be importantly mitigated through various organisational and educational mechanisms (Moody, 2001). This shows that feasible emotional control is not merely limited to one kind of relationship, and plausibly extends to a wide range.

Taken together, these two sources of extant empirical work robustly ground the view offered by Liao that feasible mechanisms of internal and external control exist which can reliably enable the successful performance of affective social labour after trying. Moreover, these studies show not only that it is possible for agents to successfully influence their relational emotions in accordance with their will but that it is likely under specific circumstances. This fact seriously challenges the view that thick social rights are unfeasible on the grounds that individuals cannot attempt to perform affective social labour, or have a virtually impossible chance of succeeding after trying. As a result, we have very strong reasons to accept Liao's rebuttal of the spontaneity argument on the grounds that emotional control is efficacious in placing affective relational states under the will, and that emotional control can meet the incentives account of feasibility.

However, even though the empirical work corroborates Liao's claims about emotional control we must – at this stage – still only conclude that we have tentative (as opposed to conclusive) reason to reject the spontaneity objection. This is because there is a deeper, underlying feasibility concern grounded in spontaneity beyond effectiveness considerations (Gheaus, 2017): non-spontaneous performances of affective social labour – those performed under duty – might not be appropriately sensitive to the interests persons have in the affective social labour of others. Indeed, it could be that the right to be loved – or the right to friendship, or something else – reflects the interest individuals have in being valued spontaneously as a particular. If true, being loved (or befriended, or whatnot) by someone simply because they are under a duty to do so, and have acted accordingly, might not satisfy the interest we have in affective social labour after all, making emotional control a failure in validating the feasibility of thick social rights.

We can motivate some of the intuitive plausibility of this thought through the following example: imagine an individual has a right to be loved (or befriended, or whatever) and the corresponding duty-bearer, out of a sense of moral responsibility, decides to take a drug which will produce a lifetime effect of the appropriate affective disposition towards them. Given that the choice to take the drug was out of a sense of moral duty, and the love is decidedly non-spontaneous, there is a reasonable case to be made that the affective disposition is in some way tarnished, or ineligible to realise the prospective claim. Indeed, would the right of children to be loved to satisfied if their parents only loved them because they were chronically drugged? What if all of one's friends only cared for them because they decided to collectively, out of a sense of duty, receive neurological shock therapy to train themselves into friendship? The fact that one's performance of affective social labour does not have any concern in the first instance with who you are as a particular, and merely what general characteristic you have (Reiman, 2007), can very plausibly be seen as a perverse and disvaluable iteration of affective social labour. That is, it might be that the typical spontaneity and particularity of affective states are constitutive of their underlying importance to our interests and, as a result, the very notion of a duty to love – or imperative to perform affective social labour – is conceptually unfeasible (as opposed to empirically or contingently).

Precisely determining the importance of spontaneity to the value of affective social labour (and, consequently, thick social rights) is a difficult matter philosophically, with the answer surely being tied up with what the precise underlying interest in receiving affective social labour actually is. Whilst an initially worrying and difficult matter, however, I believe that there is a convincing response to the objection which at least retains the idea that thick social rights are feasible in particular, important cases. The main thought is that the spontaneity objection only plausibly has force over the class of cases in which a person dutifully wills themselves into an affective emotional state in a way that leaves no room for “spontaneous” or authentic valuing at any point. That is, their feelings are entirely “synthetic” in some way. This is most clearly seen in cases where someone takes a love potion, or puts themselves through intense neurological shock therapy. We observe strong intuitions that the resulting relationships are disvaluable in an important sense and this, I think, gives plausibility to the spontaneity objection. As a result, it is implausible that one could feasibly discharge their thick social duties in these “direct” ways.

However, there is a different class of case in which one does not directly produce an emotion in a way that bypasses any authentic valuing of the other, but, instead, takes steps to increase the likelihood that spontaneous valuing will actually occur. Importantly, it is this class of case that Liao and the empirical literature is concerned with. Examples of such cases might include an individual going on a blind date with a family friend out of a sense of duty to their insistent parent, and then falling in love after spending a period dating them. Clearly, we do not think that the “non-spontaneous”, counterfactual causal element of dutifulness in the eventual spontaneous bonding undermines or brings into question the value of the eventual relationship in this sort of case and, therefore, it seems that the spontaneity objection does not have force here. The reason for this difference in intuitions is most likely that the spontaneity objection is true when the bond itself has been created out of a sense of duty in a way that bypasses any instance of spontaneous valuing, but does not include cases when the circumstances or intention to form a bond is non-spontaneous and performed out of a sense of duty, but the production of the bond itself remains authentic and spontaneous in the relevant sense. Given this is the case, we can say that wholly non-spontaneous thick social duties are not feasible, whilst partially spontaneous thick social duties are. This, I think, appropriately responds to and defuses most of the force of the spontaneity objection, and shows that there is a class of thick social duties that are feasible.

This brings us to accept the following conclusion: the (plausible) conceptual tension between spontaneous valuing and duty provides us with sufficient reason to be agnostic towards the claim that thick social rights can meet the feasibility condition through emotional control in certain cases, however, it is clear that there is a sufficient wide class of cases involving emotional control that are feasible and do not face the same objection.

To be pithier: Liao's appeal to emotional control succeeds in proving thick social rights are feasible in spite of the spontaneity argument, even though the underlying justification for thick social rights might be spontaneous valuing itself.

Section 5

In this section, I wish to show that there is a way to definitively prove the feasibility of thick social rights without having to refute the (potential) necessity of spontaneity to the interest we have in affective social labour. Put another way, I try and show here that even if I am wrong, and the spontaneity objection goes through, we still need not conclude that thick social rights are unfeasible.

To begin, it is important to clarify that the potential problem of necessary spontaneity is only a problem on the level of duties placed on individuals (duties of the form “x has a duty to perform affective social labour for y”.) Indeed, once we widen our scope beyond this individual level – and, with it, beyond the realm of emotional control of the individual agent – we begin to see that we need not disagree with any of the claims made by the spontaneity argument in order to defuse its claim that all thick social rights are specially unfeasible. For, while it might be true that spontaneous acts of affective social labour are both a) uniquely equipped to meet our interests in (and therefore rights to) such labour and b) indifferent to incentives at the bar of the individual agent, this does not rule out the possibility that collective incentives exist that could plausibly bring about the realisation of more, directed spontaneous acts that can meet the claims in question. In other words, whilst it might be unfeasible for individual agents to dutifully will themselves into performing affective social labour in the right way, this says nothing about whether or not there can be feasibly willed changes to the background structure of our social interactions that could indirectly realise extant claims to spontaneous affective social labour.

For ease of exposition, we can call duties requiring the emotional control of individuals individual thick social duties (individual duties), and duties requiring the realisation of changes in the background structure of social interaction collective thick social duties (collective duties).

To illustrate the concept of collective duties further, we can differentiate between two different types (even though there may be more).

First, we have those potential collective duties which are concerned with the modification of the social norms that structure social interaction, and the propensity of individuals to engage in affective social labour with particular persons spontaneously. For instance, it might be the case that living in a culture in which heterosexual interactions are hegemonic and stigmatising to non-heterosexual interactions could form sexual identities and preferences that militate against – and make it less likely – that non-straight individuals have their thick social needs met, given a lack of spontaneously willed performances of affective social labour to those seen as “deviant”. In this case, it seems natural to think that we might have a collective duty to correct this pattern by taking steps to change socio-sexual norms. The same logic also applies to norms surrounding the perceived disvalue of disabled bodies, individuals of certain castes, and the conventionally unattractive (Mason, 2021). To be pithy: if we did not live in a world where individuals were so picky about who they fell in love with, wanted to be friends with, or raise children with, more individuals would have their claims to (even spontaneous) affective social labour fulfilled.

Second, we have those potential collective duties which are concerned with redistributing resources and opportunities in ways that increase the real freedom (Van Parijs, 1997) ²¹ individuals have to interact with one another. For instance, it may be difficult for the physically-impaired to enter public spaces, and this can make it more difficult for those individuals to find themselves in positions with a chance to be spontaneously valued as a friend, partner, or lover.

As a result, it is clear that by addressing the architecture of social interaction – inclusive and socially-positive architecture, inclusive social norms, and a sufficient right to leisure (among other things, such as widespread development of the capacity to be social (Nussbaum, 1997)) – we can indirectly increase the number of individuals who have their thick social claims respected, irrespective of whether or not persons can feasibly will themselves into successfully performing acts of affective social labour, or not.

Of course, for collective duties to prove the feasibility of thick social rights, we must be able to show that making these changes to the background structure is itself plausible and within our power. In other words, for collective duties to show thick social rights to be definitively feasible, we must show that incentives exist that could bring about successful attempts to discharge such duties.

Promisingly, there is plenty of support for the view that they are themselves feasible.

Firstly, as has been articulated by Cordelli (2014), sociologists already have a very clear grasp of how the “relational distributive structure” operates, and what measures we can undertake in order to more efficaciously distribute affective social labour (and, as a result, satisfy claims to such labour). The main factors that determine the formation of spontaneous bonds are: “(1) the opportunity to interact informally; (2) the opportunity to interact repeatedly, frequently and for a sustained amount of time; (3) a common purpose or project or interest connected to their interaction; and (4) the opportunity to interact in conditions of cooperation rather than pure competition (Small, 2009).” In theory then, by maximising these opportunities in organisational settings where interaction takes place – schools, voluntary associations, occupational settings, care homes etc. – we can discharge collective duties.

One real-world example that can illustrate the feasibility of such interventions is found in the sociological work on educational segregation, and the finding that “friendship segregation [along racial lines] is a common outcome in racially heterogenous schools, but it need not be”, granted that internal policies that structure interaction are sensitively designed (Moody, 2001). This gives support to the notion that we already have the epistemic resources necessary to discharge collective duties – a fact which makes them feasible (or, at least, not specially unfeasible).

Secondly, increasing opportunities for social interaction – as we have said – can depend on the amount of leisure, resources, and physical access persons have. Increasing the amount of these things persons enjoy does not seem a specially unfeasible task: the vast majority of welfare states already take steps to ensure persons have a sufficient amount of these goods – as is shown through various legal and economic interventions – so there is a strong inductive case that the level could be increased in light of social rights.

Thirdly, we have good empirical evidence which suggests that comprehensive sex and relationship education for adolescents in schools is highly efficacious at shaping the social norms surrounding intimate relationships (Goldfarb and Lieberman, 2021). Not only does this show that shaping sexual norms through educational intervention is feasible, but it also highlights more general possibilities to affect the shaping of relationship-norms more generally through the tool of civic education.

Overall, then, even if we have reason to doubt that individual duties can realise thick social rights given the potential necessity of spontaneity, it is absolutely clear that such doubt does not extend to the feasibility of collective duties. This shows that, despite reasons provided by the spontaneity objection to be sceptical of the feasibility of thick social rights, and agents can still have plausible duties to ensure that the “relational distributive structure” allocates affective social labour in accordance with just principles. ²²

Section 6

In this section, I wish to engage with a number of objections to the feasibility and plausibility of collective thick social rights, and make some clarifications and amendments to my account in light of them.

Cowden raises a number of points worth addressing.

First, Cowden rejects the idea that collective duties which aim at creating conditions conducive to the spontaneous performance of affective social labour – such as loving – count as thick social duties at all (Cowden, 2012). Instead, it is argued, any such duty would actually be a “rebadging” of other sorts of (non-affective) claim rights, and could be “better” described in those non affective-labour terms. The purported reason for this is that many of the acts required by a collective agent to ensure a conducive background for the development of spontaneous, thick social relationships would themselves not require affective social labour to be performed, nor (potentially) any intention on the part of the relevant agent to produce these outcomes, given that these acts could be justified by reference to others sorts of ends. As a result, Cowden concludes that:

“A child's right to be loved as a claim-right against the State may be the most convincing way of interpreting the right to be loved but it is better expressed without mention of love… I argue that the right to be loved does not properly exist in this context either, for really it is only a re-badging of a range of other claim-rights…”

If this is true, it would undermine my account, given that it would mean that my move to collective duties to definitively prove the feasibility of thick social rights would fail given that, conceptually, the collective duties in question would not be properly conceived as thick social duties at all. Instead, they would be best understood as another distinctly non-social kind of duty, meaning that they lack traction in proving the feasibility of thick social rights as such.

Second, Cowden maintains that even if collective duties to ensure favourable social background conditions could prove the feasibility of thick social rights in some cases, they would still, taken on their own, fail to explain what thick social duties individuals have (if any) in cases where persons have enjoyed these favourable social conditions and still require affective social labour to have their interest in a minimally decent life secured. ²³ This matter does not undermine my account as such, but it does highlight its incompleteness in dealing with the spontaneity objection: without clarifying where the move to collective duties leaves us in regards to individual obligations, and what this means for the landscape of social duties, we will be ill-equipped to assess what the account offered means for the social rights literature as a whole, and its future direction. Whilst not strictly speaking a concern about feasibility, given its wider relevance, I think it important to engage with this here.

On Cowden's first objection: I do not think her argument has force.

For one, it is not clear to me why a given collective duty only properly qualifies as being about “love”, or as properly constituting a thick social duty, if the required enabling action v exhibits core features of that which is being enabled. In other words, it does not seem like a duty is only really “about” something when what it demands is sufficiently similar to the end it serves, which is what Cowden seems to be suggesting. To see this, imagine a case in which I can only prevent Andy from starving to death by giving him some food. However, the only way that I can get Andy the food is by solving a quick and easy Sudoku puzzle, at which time a mechanism will be set off that provides Andy with a nourishing meal. If Cowden's view was correct, we could not say that my duty to solve the Sudoku puzzle was really about rescuing Andy from starvation, given that we could describe my act of puzzle-solving without any reference to food or starvation. This seems wrong. Clearly, in this case, what is important in constituting my solving the puzzle as a duty of rescue is that I have the duty because it serves the end of rescue. Similarly, even if the relevant collective duties only require functional, non-social labour to be achieved – and can be described without any reference to “love” – they still qualify as thick social duties given that they exist insofar as they serve the end of fulfilling the thick social needs of claimants. As a result, the conceptual point Cowden makes is, I think, misplaced, and my move to collective duties does have traction in proving the feasibility of thick social duties.

For two, it is not clear that describing thick collective social duties as social duties is mere “rebadging”, if by “rebadging” we mean that it is a mere semantic move away from understanding the issue as one understood through non-social rights and duties. For one, we have an interest in self-understanding, and this includes understanding what we owe to other people. As a result, understanding the reasons we have for pursuing certain ends, and what ends we are required to pursue, is important as a way of ensuring we understand our own lives, and our relationship with others. Given this, there is value in understanding that we have collective social duties to others, even if our performance of the relevant actions that ensure this outcome is overdetermined given overlap with other normative considerations. This highlights that the duty is better expressed with reference to “love”, and all other relevant instances of affective social labour. Moreover, I find it hard to believe that our collective social obligations would be overdetermined in this way – indeed, given the specific nature of the claim, it seems much more likely that clarifying the full range of ends for why we have a given duty will ensure that we live up to our duties. Cowden's suggestion risks keeping us in a normative ignorance which could have repercussions for those who are entitled to affective social labour whereas understanding our collective duties as distinctly social protects against this risk. Overall, then, it seems that collective thick social duties qualify conceptually as such, regardless of whether they are constituted by performances of affective social labour, and our rendering of them in this way is not threatened by whether or not our duties are overdetermined in any case. This leaves the force of the prior account, and its implications for the feasibility of social rights, intact.

On Cowden's second objection: I think it has force, and requires me to make some comment on how this picture of thick collective duties relates to thick individual duties. Assuming that I am right about the scope of the spontaneity objection and my claim that partially spontaneous thick social duties can feasibly satisfy social needs, I think individuals will still have thick social duties to others even when those others have enjoyed favourable background conditions brought about by performance of collective duty. That is, I do not think the move to collective social duties to definitively solve the feasibility problem necessarily renders individual thick social duties redundant. Indeed, there are at least two clear cases when this will not be the case.

The first case in which individuals will have thick social duties despite the successful discharging of collective social duties is in cases when persons can meet the thick social needs of another at a reasonable cost to themselves, and a failure to do so would mean the suffering of serious harm or the living of a less than minimally decent life on the claimant. In other words, one set of cases that cement the importance of individual thick social duties is cases of social rescue. To see this, imagine a standard case of rescue in which someone has enjoyed favourable background conditions to not find themselves in peril:

Safe Pond: Andy has accidentally fallen in a shallow pond, and is now drowning, despite there being robust safety measures in place that make such events highly unlikely. Byron sees Andy drowning, and can save him at the cost of one of his fingers.

It seems clear to me that Byron has an obligation to give up his finger in order to save Andy's life even though Andy enjoyed favourable background conditions in not falling into the pond. Indeed, my intuition even remains the same when I assume that Byron has taken on his fair share of the cost in installing the safety measures. What this tells us, I think, is that one's duty of rescue is not defused simply because an individual has been provided with a very good chance of having their needs met, and not having to experience serious harm. If this is true, I think it is true arguendo in the case where an individual can prevent another from living a less than minimally decent life by performing – or earnestly trying to perform – necessary affective social labour at a reasonable cost. If true, this would mean that even if the social architecture were changed so that it was made much more likely that people had their thick social needs satisfied, this would not mean that individuals would not still have obligations to meet these needs in cases where doing so is reasonably costly. Of course, providing such affective social labour can sometimes be unreasonably costly and, therefore, in these cases, it will be incumbent upon collective agents to enable these “social rescues” and ensure they can be reasonably performed where appropriate by coordinating the provision of social labour so burdens are shared widely between multiple well-placed agents, and, perhaps, even by compensating people for the costs they incur in the process. It is clear, however, that, in principle, these individual duties will remain independently of the existence and performance of collective social duties.

The second case where individuals will have thick social duties towards others beyond collective duties is in the context of associative obligations whereby the provision of affective social labour is required in the context of a special relationship between two or more agents.

One specific case where this is clear is in the context of promises and contracts. To see this, consider Safe Bodyguard:

Villain has shot a slow bullet at Andy. Bodyguard promised Andy that he would jump in front of any bullet fired at Andy if he could feasibly do so, and is able to jump in front of the slow bullet before it hits. The state Bodyguard and Andy live in is incredibly safe given effective collective coordination, making any attempt on Andy's life highly unlikely.

Here, it seems clear that Bodyguard has an obligation to keep his promise and jump in front of the bullet, given that he has promised to do so, and thereby entered into a particular kind of associative relationship with Andy that requires him to act in this way given his role as “Bodyguard”. This is true even though Andy (and Bodyguard) both enjoy favourable background conditions generated by collective coordination that make it highly unlikely that he would be shot at by Villain. This shows that when one agent has promised to v in a certain situation, they have an obligation to v where appropriate, even if the likelihood that they will have to v is low given the nature of the social world. If this is true, and then it also seems true that an individual who promises to perform affective social labour for another in a given circumstance – regardless of how favourable background conditions are – is required to do so when certain conditions are met. Concrete examples of this might include those who have contracted to be care workers, babysitters, professional family members, or even those who have entered into certain kinds of marriages. Of course, promises are not absolute in all cases, and the requirement of having to keep a promise can be superseded by factors such as costs to oneself or others in doing so, or the actions of the individual you have made the promise to, but there is a presumption they will be kept, making them fairly robust examples of sources of individual thick social duties.

Another case where associative duties ground thick social duties is when one shares a relationship of care with another, whereby they occupy a role that justly requires them to meet these affective social needs. The clearest example of this is probably the parent-child relationship: the associative bond of the family plausibly brings with it an obligation for members of that family to bear costs for each other in order to provide all with what they need to live a minimally good life. I remain agnostic on whether this is because parents have responsibly caused the needs of their children, or because there is something sui generis about parental obligation to provide affective social labour. In any case, the thought that this is owed is deeply plausible and, furthermore, its plausibility is resistant to the existence of favourable social conditions. Indeed, I do not think that whether or not a parent has a duty to (try to) love their child is dependent upon how likely it is that someone else will love them, nor do I think that a parent's duty to perform affective social labour for their children is defused once this has not happened spontaneously under favourable background conditions. It seems clear that a parent has a duty to try even in these cases and, therefore, it seems clear that the fulfilment of collective social duties will not displace the existence of individual thick social duties in cases where persons occupy a role in which they have specific social obligations to others, as is seen in the family or, perhaps, even in religious communities of certain kinds.

As a result, I conclude that Cowden is correct to highlight the question of how the move to collective social duties – in solving feasibility problems or others – leaves the status of our individual social duties, especially towards those who remain needy even after enjoying favourable social conditions. I have tried to show that collective duties will not be a replacement for individual duties, and will be present even after the social architecture works how it ought to.

Section 7

I wish to finish, finally, by making a brief note on the sometimes enduring unfeasibility of social rights. ²⁴

As I have noted in the introduction, I am interested in establishing the general feasibility of social rights. In this vein, however, is important to recognise that, sometimes, there will be cases where someone needs social labour, but it is not feasible to provide them with it. This might be because there is nobody with the skills or resources that can meet the person's needs, or it might be because the individual in question cannot access the goods of social labour themselves. To give one stylisation of the issue: there might be some persons who cannot bring themselves to love, and those who cannot feel the love of others. I do not rule out this possibility. There might also be informational constraints that make social rights unfeasible given that persons cannot reasonably know that there is someone out there who needs their aid. Again this is a difficult problem.

However, I maintain that none of these problems undermine the general feasibility of social rights as a class of entitlements that lay claim on the acts of others. That is, I maintain there is nothing specially unfeasible about performing social labour for and with others and, as a result, we should hold it to the same set of expectations we hold other classes of entitlements in our normative theorising. Just as the fact that some people cannot rescue others from ponds does not render the duty of rescue specially unfeasible, neither does the existence of those who cannot provide others with friendship undermine the feasibility of providing friendship to another in general. Similarly, just because we do not currently know who our duties are owed to specifically today, does not mean that collective agents cannot make progress on this question, and act accordingly.

Understanding social rights as a class of entitlement as feasible as any other does not mean there are never problems in discharging them: it means that we ought to fully consider them rights, and strive towards them accordingly, without refusing to deliberate about how we might satisfy their requirements ex ante (Southwood, 2022).

Conclusion

Social rights, despite initial intuitive concerns, are not specially unfeasible. Indeed, thin social rights seem to be straightforwardly feasible. Even thick social rights, whilst seeming unfeasible across a particular class of particularly non-spontaneous individual cases, are definitively feasible at the bar of collective duties and over a sizable range of more spontaneous individual cases. By clarifying that there is nothing inherently unfeasible about even thick social rights, this clearly places them – and social rights in general – into the running to be considered moral imperatives.

Of course, the matter of whether social rights are properly considered part of our normative landscape will depend on more than their feasibility alone. However, it has at least now been made clear that they cannot simply be brushed aside on the grounds that they cannot be rights, and stand only as inherently empty aspirations, because they require the performance of miracles.