The argument from the ‘special goods of childhood’

The first argument that we want to consider takes as its starting point the claim – recently made by a number of philosophers of childhood (e.g. Brennan Reference Brennan, Baylis and Macleod2014; Brighouse and Swift Reference Brighouse and Swift2014; Gheaus Reference Gheaus, Bagattini and Macleod2015a, Reference Gheaus2015b; Macleod Reference Macleod, Brighouse and Robeyns2010, Reference Macleod, Bagattini and Macleod2015; Weinstock Reference Weinstock2018) – that there exist ‘special’ (or ‘intrinsic’) goods of childhood, that is, goods that make a special contribution to children’s wellbeing. The claim that there exist special goods of childhood has been defended for a variety of purposes in the literature, Footnote ⁷ including in the context of discussions about what is owed to children as a matter of justice (see Brennan Reference Brennan, Baylis and Macleod2014; Macleod Reference Macleod, Brighouse and Robeyns2010) and about the overall value of childhood as a stage of life (see Brennan Reference Brennan, Baylis and Macleod2014; Weinstock Reference Weinstock2018). As far as the issue of generalism is concerned, however, this claim is relevant because it may be argued (see, e.g. Tomlin Reference Tomlin2018a) that if the items that make children’s lives go well for them, as children, are different from the items that make adults’ lives go well for them, as adults, Footnote ⁸ then children’s and adults’ wellbeing are fundamentally different.

Examples of special goods of childhood mentioned in the literature include: unstructured play, freedom of experimenting with different selves, sexual innocence, an ability to love and trust without apprehension, purposeless imagination and a sense of being care-free. It must be noted that different authors concerned with the existence of special goods characterise this notion slightly differently. For instance, Harry Brighouse and Adam Swift refer to such goods as items that ‘may have value only, or much more readily, in childhood’ (Brighouse and Swift Reference Brighouse and Swift2014, 65). By contrast, Colin Macleod talks about goods ‘to which children as juvenile agents have privileged and perhaps unique access’ (Macleod Reference Macleod, Bagattini and Macleod2015, 59). It is reasonably clear, however, that in order to support a plausible argument against generalism, the claim that there exist special goods of childhood must be specifically understood as the claim that there exist some items that are good only for children, but not for adults.Footnote ⁹

The resulting argument can be summarised as follows.

• (1) There exist some items that are good only for children, but not for adults.

• (2) If (1) is true, then children’s and adults’ wellbeing is fundamentally different.

• (3) Therefore, children’s and adults’ wellbeing is fundamentally different.Footnote ¹⁰

We want to raise two objections against this argument. The first targets premise (1). In order to support the rejection of generalism, this premise should be understood as a claim about final goods, rather than merely instrumental goods. That is, premise (1) should be interpreted as saying that there exist some items that are finally good (i.e. good for their own sake or as ends) only for children, but not for adults. Final goods are constituents of wellbeing. They contrast with instrumental goods, which are just sources of (or means to) wellbeing. The problem is that the goods that are deemed to be special to childhood appear to be just sources, rather than constituents, of wellbeing. Indeed, goods such as unstructured play or a sense of being care-free seem to be just means to obtain other things that are (supposedly) finally good for the individual, e.g. pleasure, happiness, excellence, etc.Footnote ¹¹ In fact, it is difficult to see how (e.g.) unstructured play could be non-instrumentally good for children, independently of whether it is (e.g.) pleasurable, conducive to happiness or conducive to the development of higher-level capacities. If this is true, then premise (1) can be challenged. First, it is possible for goods such as unstructured play and purposeless imagination to be occasional sources also of adults’ wellbeing. Second, and most importantly, it is possible for the sources of children’s and adults’ wellbeing to differ, without this implying a difference at the level of the items that are finally good for them.

Our second objection targets premise (2). Suppose, for the sake of argument, that there exist special goods of childhood in the sense of premise (1), that is, things that are finally good only for children. Does it follow that children’s and adults’ wellbeing are fundamentally different? Once again, we are sceptical. Two familiar distinctions in wellbeing theory can help us articulate the reasons for our scepticism. The first is the distinction between the items that are finally good for an individual and the properties that make these items finally good for the individual (see Crisp Reference Crisp2006). This distinction corresponds to two ways in which the notion of ‘wellbeing constituents’ can be specified. It also underlies two distinct projects in wellbeing theory. The first is to offer a list of items that are finally good for the individual. This is the goal of enumerative theories of wellbeing. The second is to offer an explanation of why these items are finally good for the individual. This is the goal of explanatory theories of wellbeing. In order to appreciate the distinction, let us consider, as an example, the desire-fulfilment theory of wellbeing. According to this theory, wellbeing consists in the satisfaction of the individual’s non-instrumental desires. Below the surface of this definition lies an important division of theoretical labour. In fact, the theory combines two claims. The first is that any item that the individual desires non-instrumentally (i.e. as an end) is finally good for her (i.e. good for her as an end). The second is that what makes that item finally good for the individual is its being desired by her.Footnote ¹² We can use this example to illustrate the significance of the distinction between enumerative and explanatory theories of wellbeing for the present discussion. First, two individuals’ wellbeing may differ at the enumerative level, but not at the explanatory level. That is, although the items that are finally good for these individuals may be different, the explanation of why they are finally good may still be the same. Second, it is plausible to think that the project of providing an explanation of why certain items are finally good for an individual is the most fundamental project in wellbeing theory. If the latter point is true, it follows that what makes two accounts of wellbeing fundamentally different is the fact that they differ at the explanatory level.

In light of these remarks, we can raise the following objection against the argument from the special goods of childhood. In order for the argument to succeed, it is not enough to show that children’s and adults’ wellbeing differ at the enumerative level. One needs to show, most fundamentally, that they differ at the explanatory level. Thus, even if we grant that there exist items that are finally good only for children, we cannot infer that children’s and adults’ wellbeing is fundamentally different unless we have shown that their status as final goods is determined by different good-for-making properties. No such demonstration has been given in the literature. Thus, the argument from the special goods of childhood remains inconclusive.Footnote ¹³, Footnote ¹⁴

At this stage, one may counter that the explanatory project should not be regarded as the most fundamental project in wellbeing theory. What matters most is to provide a list of kinds of items that are finally good for the individuals. There is no further explanation of why they, and only they, are finally good for those individuals. As a matter of fact, this position is typically held by defenders of objective list theories of wellbeing (see Fletcher Reference Fletcher2016). It is especially relevant for the present discussion because it implies that, if it is true that there exist some kinds of items that are finally good only for children, then children’s and adults’ wellbeing is indeed fundamentally different, since the list of wellbeing constituents is different in the two cases.

Once again, however, we doubt that this picture is correct. In order to show why, we will appeal to a second distinction in wellbeing theory, namely, the distinction between items that are derivatively finally good for an individual and items that are non-derivatively finally good for her.Footnote ¹⁵ An item is non-derivatively finally good for an individual if and only if it is good for her for its own sake and in its own right. This means that its value as an end for the individual does not derive from the value as an end of any other item. It has not been sufficiently recognised that it is possible for an item to be finally good for an individual also derivatively. There are various ways in which this may happen. For instance, an item may stand in a part-to-whole relation with another item that is non-derivatively finally good for the individual. Alternatively, an item may be a specification of a more general item that has non-derivative final value for the individual.Footnote ¹⁶ None of these relations is equivalent to the means-end relation. (To give a non-evaluative example, red is a specification of colour, but clearly not a means to colour.) This shows that it is possible for an item to be both finally good for an individual and for its final goodness to be derivative.

This distinction provides the basis for the following response. It seems plausible to say that the ultimate goal of an objective list theory of wellbeing is to identify the kinds of items that are non-derivatively finally good for (a class of) individuals. If this is the case, then two objective list theories of wellbeing cannot be said to be fundamentally different unless they differ with respect to the kinds of items that are non-derivatively finally good for the individuals. In support of this claim, consider the following: if children’s and adults’ list of wellbeing goods did not differ at the level of the non-derivative final goods, then their wellbeing would still be comparable, even if there were other ‘mid-level’ differences between the two. This is a sign that the accounts of children’s and adults’ wellbeing would not be fundamentally different.

The problem for the argument under consideration is that, to the extent that it is plausible to regard them as non-instrumental goods, the items that are typically mentioned as being special to childhood do not seem to be non-derivatively finally good for children. Rather, they seem to stand in one of the relations described above with items that are non-derivatively finally good. Crucially, however, the latter items seem to be non-derivatively finally good also for adults. To see this, let us examine the main candidate special goods of childhood that have been discussed in the literature. Consider, first, unstructured play. Arguably, being unstructured is a way of being free, whereas play is a kind of enjoying activity. However, freedom and enjoyment are goods that presumably figure in any objective list of wellbeing goods – for children as well as for adults. So, it seems that the final value for children of unstructured play derives from the final value of goods that are also non-derivatively finally good for adults. The same diagnosis applies to the sense of being care-free, which can be interpreted as a particular way of realising freedom and avoiding unpleasant burdens. As for experimentation with different selves, it can be seen as a way for children to exercise their developing capacity for practical reasoning, to achieve particular goals and to acquire knowledge about themselves. Very plausibly, however, if practical reasoning, achievement and knowledge are non-derivative final goods for children, they also are for adults. Finally, insofar as sexual innocence – understood as a certain lightness about sexuality and one’s sexual desiresFootnote ¹⁷ – and the ability to trust and love without apprehension are final goods (which there is reason to be sceptical about), Footnote ¹⁸ then they can be seen as ways of avoiding unpleasant burdens and of realising the goods of enjoyment, happiness, and (perhaps) authenticity. If these are non-derivative final goods for children, they are, of course, also for adults.

The lesson to draw is that, in all these cases, the final value of the goods that are typically deemed ‘special’ to childhood seems to be grounded in, and derivable from, the final value of other goods, which are likely to figure also in an objective list of adults’ goods. If this is true, then children’s and adults’ wellbeing are not fundamentally different, but differ at most at an intermediate level, i.e. at the level of the derivative final goods. We believe that this is actually the real locus of the debate between those who affirm and those who deny that there exist items that are finally good only for children, but not for adults. That is, we think that it is most plausible to interpret them as disagreeing about which items are derivatively finally goods for children and for adults, rather than about the ultimate constituents of children’s and adults’ wellbeing.Footnote ¹⁹ Be that as it may, the upshot of our argument is that children’s and adults’ wellbeing do not differ fundamentally.

To sum up, in this section we have offered various reasons to be sceptical of the argument from the special goods of childhood. If we are right, the argument is either unsound or inconclusive.

The argument from the role of higher-level capacities

Several authors have noticed that the main contemporary theories of wellbeing seem to be ill-suited to characterise children’s wellbeing, for they link wellbeing to the possession and exercise of higher-level capacities, which young children do not possess (see Raghavan and Alexandrova Reference Raghavan and Alexandrova2014; Skelton Reference Skelton, Bagattini and Macleod2015; Tomlin Reference Tomlin2018a). One may take this to be a sufficient ground to reject generalism and to claim that a distinctive account of children’s wellbeing is needed. The argument underlying this position can be summarised as follows.

• (1) Some prominent views hold that wellbeing requires the possession and exercise of higher-level (cognitive and emotional) capacities.

• (2) Either one of these views is the correct account of adults’ wellbeing.

• (3) Children do not possess such higher-level (cognitive and emotional) capacities.

• (4) Therefore, insofar as children are wellbeing subjects, children’s and adults’ wellbeing is fundamentally different.

Here are a few examples of the kind of views to which (1) refers.

• (i) Wellbeing consists in the achievement of rational and sufficiently informed plans (see Rawls Reference Rawls1971; Raz Reference Raz1986). This view presupposes the possession of a capacity for planning, some executive control skills, a capacity to understand and process information about the object of one’s plans, and a minimal level of (instrumental) rationality.

• (ii) Wellbeing consists in attitudes of satisfaction towards one’s life considered as a whole (see Sumner Reference Sumner1996; Suikkanen Reference Suikkanen2011). This view presupposes a sufficient level of self-consciousness and a capacity to conceptualise one’s life as a whole, as well as to form attitudes towards it, presumably by synthesising one’s overall affective experiences.

• (iii) Wellbeing consists, either fully or in part, in virtuous activity, where the latter includes morally virtuous activity (see Annas Reference Annas1993; Badhwar Reference Badhwar2014). This view presupposes the possession of a capacity to act morally, which is grounded in a capacity to distinguish moral considerations from considerations of a non-moral kind, e.g. conventional, aesthetic, etc. Insofar as virtue involves a disposition to act for the right reasons, it also presupposes a capacity to recognise reasons as such and to take them as grounds for one’s behaviour.

These views are paradigmatic representatives of the kind of capacities that wellbeing is supposed to require. What is especially striking is that all these theories make wellbeing depend on the possession of capacities that younger children seem to lack or possess to a particularly low degree.

We call this ‘the argument from the role of higher-level capacities’. Before presenting our main objection against it, we want to say a few preliminary words about premise (3). Recall that (3) claims that children do not possess higher-level (cognitive and emotional) capacities. Surely, however, this claim must be qualified in some ways. For a start, children develop the relevant capacities gradually, to various degrees, and at different stages of their development. To give a few examples, children begin to develop various types of planning-related skills starting from their second year of age (see McCormack and Atance Reference McCormack and Atance2011) and begin to display inhibitory control by age 3 (see Posner and Rothbart Reference Posner and Rothbart1998). Self-representational and other meta-representational skills, such as those involved in the understanding of other people’s minds, emerge between 15 and 24months of age (see Lewis Reference Lewis, Overton, Molenaar and Lerner2015; Meltzoff Reference Meltzoff1995). Children show signs of concerned affect for others as early as 10–12months of age (see Zahn-Waxler et al. Reference Zahn-Waxler, Radke-Yarrow, Wagner and Chapman1992; Roth-Hanania, Davidov, and Zahn-Waxler Reference Roth-Hanania, Davidov and Zahn-Waxler2011) and engage in comforting behaviour between 12 and 18months of age (see Vaish, Carpenter, and Tomasello Reference Vaish, Carpenter and Tomasello2010). By age 3, they can identify and intervene in transgressions against other individuals (see Vaish, Carpenter, and Tomasello Reference Vaish, Carpenter and Tomasello2010, Reference Vaish, Carpenter and Tomasello2011). Shortly after, they begin to develop a rudimentary sense of fairness and reciprocity (see Warneken and Tomasello Reference Warneken and Tomasello2013; Killen and Smetana Reference Killen, Smetana, Lamb and Lerner2015).

These findings in developmental psychology and neuroscience show that infants and young children are much less hopeless than it is commonly thought.Footnote ²⁰ They also suggest an argument of the following kind. Either the accounts of wellbeing according to which the latter consists in the exercise of higher-level capacities apply to an individual as soon as she begins to develop such capacities or they apply to an individual only when she has met a given threshold. In the first case, the accounts of wellbeing under consideration are applicable to children basically from the beginning of their lives, since children start to develop the relevant capacities at very early stages. This shows that, even if premise (3) is true and children do not possess (in the threshold sense) the relevant higher-level capacities, the accounts of wellbeing under consideration are still applicable to them, because their applicability does not require the possession (in the threshold sense) of those capacities. By contrast, in the second case, their applicability to children depends on where the threshold is set. As we have seen, infants might be capable of meeting minimal thresholds in the exercise of at least some of the relevant capacities earlier than commonly thought. Yet, this is not enough to conclude that premise (3) is false. It merely serves to limit the scope of the argument to very young children or babies.

Be that as it may, our objection against the argument from the role of higher-level capacities is a different one. Our main claim is that all the accounts of wellbeing to which premise (1) refers face some independent serious problems, so that they are either unsuitable even as accounts of adults’ wellbeing or, if they are reformulated so as to correct for their shortcomings, they become applicable also to very young children and babies. While these views constitute only a sub-set of all the possible higher-level capacity accounts of wellbeing, their paradigmatic status casts more than one doubt on the very plausibility of this general approach to wellbeing.

Consider (i). While this view faces a number of problems (see Fletcher Reference Fletcher2016; Heathwood Reference Heathwood and Fletcher2016), the most significant for the present purpose is that it is too narrow. Some items appear to be wellbeing-enhancing independently of whether they are targets of any plans and beyond their instrumental contribution to such plans. For instance, an individual may accidentally find success and happiness in one career, despite having planned to devote her life to a completely different career. Taking this objection into account, one may reformulate the view by replacing the notion of ‘plans’ with the broader notion of ‘desires’. One can then say that something is good for an individual if and only if it satisfies her desires, whether or not the latter are part of the individual’s ‘plans’. Independently of the merit of this move, what matters for the present purpose is that such a reformulation makes the view applicable to young children as well, since they (including babies) do possess a capacity for desires and goal-directed behaviour.

Consider (ii). The whole-life-satisfaction view of wellbeing seems inadequate as an account of wellbeing even for adults. First, it is part of the ordinary understanding of wellbeing that the latter can be assessed at any arbitrary moment in time. However, few adults experience actual attitudes of satisfaction towards their life as a whole with the frequency necessary to satisfy this condition (see Haybron Reference Haybron2008; Feldman Reference Feldman2010). A more plausible formulation of the view, then, is that wellbeing consists in dispositional attitudes, rather than occurent attitudes. Even so, it appears that some well-functioning adults may lack dispositional attitudes that have ‘life as a whole’ as their intentional object. So, perhaps, the view should be reformulated further by saying that what matters for wellbeing is the disposition to experience feelings of satisfaction towards one’s life circumstances, which do not require conceptualising one’s life ‘as a whole’. While this variant of the theory seems more plausible, for the present purpose it is enough to notice that, once we go this route, the theory becomes applicable also to younger children, for even babies are capable of experiencing feelings of satisfaction, e.g. when they are fed and held by a loving parent.

Finally, consider (iii). There are two versions of this theory. The stronger holds that virtuous activity is sufficient for wellbeing. This view has few supporters nowadays.Footnote ²¹ It appears indeed that other things must be present in order for an individual’s life to go well, e.g. external goods, psychological appreciation, etc. The weaker version of the theory holds that virtuous activity is necessary for wellbeing, but not sufficient (see Darwall Reference Darwall2002; Badhwar Reference Badhwar2014). However, even this version is subject to powerful counter-examples, which are meant to show that one can flourish without virtue (see Hooker Reference Hooker and Crisp1996; Dorsey Reference Dorsey2010). It is often said that these counter-examples show that (iii) is not really a theory about wellbeing, but rather a theory about the good life tout court, i.e. the life that is all-things-considered worth-living, taking into account also non-prudential considerations, such as moral and aesthetic considerations (see Haybron Reference Haybron2008, chapter 8). These problems are compounded by the difficulty in determining exactly which capacities one must virtuously exercise for her life to go well. Within standard Aristotelian versions of the theory, the answer is: the capacities that are constitutive of human nature. However, it has traditionally proven difficult to arrive at a view of human nature that is sufficiently broad to include (e.g.) individuals with various disabilities, and yet sufficiently narrow to avoid the inclusion of capacities whose exercise is often intuitively detrimental to one’s wellbeing, e.g. the capacity for gossiping. All in all, these considerations suggest rejecting this view of wellbeing, at least in its most common versions.

The upshot of this section is the following. The accounts according to which wellbeing involves the exercise of higher-level capacities are either seriously problematic or stand in need of a reformulation. When the latter option is chosen, these accounts become typically applicable to young children and babies as well. Either way, the argument from the role of higher-level capacities fails.

The argument from the change in prudential standards

The first argument is based on two claims. The first is that the gradual and non-linear character of children’s development makes it implausible to think that there is a sudden change in their prudential standards while moving from childhood to adulthood, as the rejection of generalism seems to imply. The second claim is that the more an opponent of generalism tries to avoid such an implication, the more difficult it becomes for her to account for the existence of a fundamental change in children’s and adults’ prudential standards.

Here is the idea. Prima facie, it is plausible to think that if children’s wellbeing were fundamentally different from adults’ wellbeing, then there would exist a point where the standards of an individual’s wellbeing change fundamentally. Suppose, for instance, that hedonism were the true account of children’s wellbeing, whereas a higher-level capacity view were the true account of adults’ wellbeing. According to this picture, there would exist a time t such that, before t, the individual’s wellbeing is exclusively determined by the balance of pleasures and displeasures in her life, whereas, after t, her wellbeing is exclusively determined by the exercise of higher-level capacities.

One problem with this picture is that it is in tension with the gradual character of the transition from childhood to adulthood. In fact, the present picture establishes an implausible discontinuity in the individual’s evaluative and normative prudential standards. Within this picture, a sudden change occurs both with respect to what makes the individual’s life go well and with respect to the individual’s prudential reasons. Before t, for instance, the individual has no final reasons to exercise and develop higher-level capacities and, in some circumstances (i.e. when her hedonic balance is unaffected or negatively affected), not even instrumental reasons to do so, which have to do purely with her wellbeing as a child, rather than with her wellbeing as a future adult. By contrast, after t, she acquires both final and instrumental prudential reasons of this kind. This seems to presuppose that a radical change occurs in the individual herself.Footnote ²² The problem is that the individual at time t ₋₁ is not that different from the individual at time t. Crucially, there seems to be no ‘tipping point’ in an individual’s development, which turns her into an adult. The reality is most likely one characterised by progresses and setbacks. Adding to this, there is evidence that a single higher-level capacity may develop at a different pace in different areas of children’s life. For instance, decision-making autonomy develops much earlier in personal domains, e.g. one’s appearance, than in the prudential domain (see Wray-Lake, Crouter, and McHale Reference Wray-Lake, Crouter and McHale2010). These considerations make the idea of a radical discontinuity in an individual’s prudential standards counter-intuitive.

An opponent of generalism might reply that the rejection of generalism is perfectly compatible with the recognition of the gradual, and non-necessarily linear, character of the transition from childhood to adulthood. It is indeed open to her to adopt a different picture, according to which, as the individual gradually develops from child to adult, the theory of wellbeing that applies to her gradually changes. In this case, the child’s gradual development is accompanied by a gradual change in her prudential standards. Thus, within this picture, the rejection of generalism does not imply any ‘sudden’ change in the individual’s prudential standards.

We think that this reply does not succeed. To see why, notice that, intuitively, there will be areas of indeterminacy in an individual’s development, i.e. areas where the individual is no longer a child but not quite an adult yet. It seems that, in those areas, the individual’s wellbeing will be determined both by the constituents of children’s wellbeing and by the constituents of adults’ wellbeing. This will be true, for instance, when children’s higher-level capacity development proceeds differently in different domains of their life. This makes the following problem especially salient. There seems to be no non-ad hoc rationale for holding that some constituents of children’s wellbeing cease to determine the individual’s wellbeing when the individual moves from an area of indeterminate adulthood to an area of determinate adulthood. Suppose, once again, that hedonism is the correct account of children’s wellbeing, whereas a higher-level capacity view is the correct account of adults’ wellbeing. Ex hypothesi, in moving from indeterminate to determinate adulthood, the individual’s only relevant change is that she now determinately possesses some additional higher-level capacity. The question is: why should this fact alter the prudential status of the capacity to experience pleasure and pain or of the experiences themselves? After all, the individual has not lost this capacity, nor has shebecome incapable of having those experiences. If pleasures and pains were determinants of the individual’s wellbeing while she was in a state of indeterminate adulthood, why should the passage to a state of determinate adulthood drastically change their role as determinants of the individual’s wellbeing? An opponent of generalism owes us an explanation, but there seems to be no obvious explanation available.Footnote ²³

At this point, an opponent of generalism might concede that, if a capacity, or a kind of experience, determined the individual’s wellbeing when she was a child, then it remains a determinant of her wellbeing when she becomes an adult (unless the individual loses that capacity or becomes unable to have those experiences). It is just that, once the individual reaches a particular developmental threshold, her wellbeing is also determined by the exercise of her now fully developed higher-level capacities. We may think, for instance, that while children’s wellbeing is determined only by the balance of pleasures and pains, adults’ wellbeing is jointly determined by such experiences and by the realisation of their autonomous plans. This would be enough to make children’s and adults’ wellbeing fundamentally different.

It seems to us, however, that this move is bound to fail, for it is inconsistent with a principle that we have independent reasons to adopt.Footnote ²⁴ The principle in question states that if the exercise of a fully developed capacity is intrinsically good for an individual (i.e. good in itself, in virtue of its internal properties), then the exercise of the same capacity is intrinsically good for that individual even when such a capacity is only partially developed.Footnote ²⁵ Before proceeding, note that so far, for simplicity, we have taken adults to be individuals that possess some higher-level capacities to a sufficient degree. Surely, however, this is not a necessary condition for being an adult, for many individuals would otherwise implausibly not qualify as adults. In fact, in everyday contexts, we classify as adults many individuals of a certain age that possess the relevant capacities to a much more limited degree. Let us use the term ‘adults*’ to denote this broader understanding of adults. With this in mind, let us go back to the previous principle. When adults* are concerned, this principle seems to enjoy widespread acceptance. Consider the capacity for autonomy. It is often said that the exercise of a developed capacity for autonomy is intrinsically good for an individual (see Rice Reference Rice2013; Fletcher Reference Fletcher2013; Hooker Reference Hooker2015). However, some adults* possess such a capacity only to a limited degree. It seems hard to deny that the exercise of those individuals’ capacity for autonomy is still intrinsically good for them, though perhaps to a lesser degree. The alternative is to say that the exercise of a partially developed capacity for autonomy is good for those individuals only as a means for the full development of such capacity. However, this is problematic. It implies that an individual who will (perhaps, knowingly) never be able to fully develop her capacity for autonomy has no prudential reason to exercise her capacity for autonomy.Footnote ²⁶ If we think that this implication is counter-intuitive, then we are forced to reject the alternative explanation. Generalising from this example, we arrive at the principle stated above. The next step consists in noticing that if this principle applies to adults*, then it applies also to children that are relevantly similar to them. To deny this, one would need to show that the mere difference in terms of age between them is sufficient to make the principle valid for adults*, but not for children. Contra this, it is more plausible to say that the principle applies only to adults* but not to children only if the difference in age is accompanied by other differences that are prudentially significant. As far as autonomy is concerned, however, some children are comparable to some relatively well-functioning adults*. Thus, if the previous principle applies to the latter, it must also apply to the former.

These considerations cast significant doubts on the anti-generalist view according to which the exercise of a higher-level capacity becomes a determinant of an individual’s wellbeing only when she has fully developed that capacity. Contrary to this view, it seems more plausible to think that, for any individual, if the exercise of a higher-level capacity is intrinsically good for her when she has fully developed that capacity, then it is also intrinsically good for her when she has only partially developed that capacity. This applies also to children. Perhaps, as stated above, the exercise of a higher-level capacity contributes less to an individual’s wellbeing when that capacity is only partially developed, and up to zero when it is not developed at all.Footnote ²⁷ Yet, if the previous considerations are on the right track, the point remains that we have strong reason to reject the view according to which new determinants of wellbeing appear at different stages of an individual’s development and, more generally, a picture that presupposes the existence of a fundamental difference in children’s and adults’ prudential standards.

The argument from normative utility

There are two main criteria that are typically used to assess competing theories of wellbeing: the criterion of descriptive adequacy and the criterion of normative utility (see Sumner Reference Sumner1996; Haybron Reference Haybron2008). According to the former, one should favour the theory that best fits our considered intuitions about wellbeing. According to the latter, one should favour the theory that best vindicates our practical and theoretical interests in the subject. The argument in favour of generalism offered in the previous section makes implicit appeal to the criterion of descriptive adequacy. In this section, we consider instead the criterion of normative utility and argue that the latter favours generalism over its denial.

Here is the idea. It is common to think that a plausible theory of wellbeing must help us make sense of all our wellbeing judgements, including judgements of wellbeing levels and differences (see Sumner Reference Sumner1996, 13). The latter requirement is especially important from the point of view of normative utility. Judgements of wellbeing levels and differences play indeed a particularly important role both in everyday deliberation and normative theory. Crucially, however, these judgements may involve comparisons of children’s and adults’ wellbeing. For instance, a parent may try and convince her child to perform an undesired action by telling her that a little sacrifice now may bring about significant benefits for her in the future.Footnote ²⁸ In so doing, the parent appeals to an intra-personal, inter-temporal comparison of wellbeing, according to which the future adult’s wellbeing gains outweigh the present child’s wellbeing losses. From a more theoretical point of view, many think that the prudential value of an individual’s life as a whole is an additive function of that individual’s wellbeing at different moments in her life. Yet, this position is tenable only if it is possible to make intra-personal, inter-temporal comparisons of wellbeing differences. Similarly, the very meaningfulness of many moral theories (most notably, aggregative welfarist moral theories) depends on the possibility of making inter-personal comparisons of wellbeing (see Hausman Reference Hausman1995), including children’s and adults’ wellbeing. Such comparisons may be equally important for determining what we owe children versus adults, whenever wellbeing is a relevant consideration of justice or morality.

These remarks show that it is important for a theory of wellbeing to make room for comparisons of children’s and adults’ wellbeing. More specifically, in order to be normatively useful, a plausible theory of wellbeing must be such as not to construe comparisons of children’s and adults’ wellbeing as indeterminate or impossible. The latter notions can be defined as follows. Comparisons of children’s and adults’ wellbeing are indeterminate if and only if there is no fact of the matter about them, so that it is indeterminate whether any comparison between children’s and adults’ wellbeing holds. Comparisons of children’s and adults’ wellbeing are impossible if and only if the two are incomparable, that is, if it is determinately false that any positive comparison between children’s and adults’ wellbeing holds.Footnote ²⁹ The problem is that, depending on how they are formulated, accounts of wellbeing that reject generalism make comparisons of children’s and adults’ wellbeing vulnerable to either one of these problems. Indeed, if children’s and adults’ wellbeing are conceived of as fundamentally different, it appears that comparisons between them cannot be made in either of these two senses, because there is no common denominator to which one can appeal to make such comparisons.Footnote ³⁰ The upshot is that, if the criterion of normative utility has some weight in the construction of a plausible theory of wellbeing – as it is normally believed – then it favours an account of wellbeing according to which children’s and adults’ wellbeing is the same, at least at the most fundamental explanatory level.

An opponent of generalism might reply that the rejection of generalism does not entail the incomparability of children’s and adults’ wellbeing. Even when they are fundamentally different, the theories of children’s and adults’ wellbeing remain theories of the same thing, namely, prudential value. It is true that they offer different criteria (or principles) for determining what is good for children and adults. But in the end, they are still meant to provide us with information about the extent to which children’s and adults’ lives possess the same kind of value. If this is the case, then children’s and adults’ wellbeing can be meaningfully compared, even if they are determined by fundamentally different theories.

We believe that this reply does not succeed. To see why, consider the following case, analogous to the one under consideration. Suppose we want to determine which of two societies – let us call them S₁ and S₂ – is more just. Suppose also that justice is determined in accordance with standards that are relative to each of these societies. (That is, suppose that normative moral relativism is true). Thus, for instance, justice in S₁ is determined in accordance with utilitarian standards, whereas justice in S₂ is determined in accordance with Rawlsian standards. Finally, suppose that, relative to its utilitarian standards, S₁ is a relatively just society. By contrast, relative to its Rawlsian standards, S₂ is a very unjust society. Now, one may argue that, although utilitarianism and Rawlsianism offer different criteria of justice, they are still theories of the same thing, namely, justice. As such, they can still lead us to make meaningful judgements of the extent to which different societies are just. If this line is correct, then we should conclude that S₁ is more just than S₂. But this is a mistake, for moral relativism excludes the possibility of making this sort of non-relative judgement. The only thing that moral relativism allows us to say is that S₁ occupies a higher position in its utilitarian justice scale than S₂ does in its Rawlsian justice scale. Yet, this is not equivalent to the claim that S₁ is more just than S₂ tout court.

One might insist that, if given the choice of where to live and if justice were the only relevant criterion, then one should surely choose to live in S₁ rather than S₂. But if this choice makes sense, this can only be because some form of intersocietal comparison of justice is possible. The problem is that choosing S₁ over S₂ on the basis of justice does not actually make much sense. For as far as we know, if utilitarian standards were applicable to S₂, then S₂ might turn out to be more just than S₁ relative to such (common) standards. This shows that the fact that S₁ ranks higher in its utilitarian scale than S₂ does in its Rawlsian scale does not give any reason to prefer S₁ to S₂ in terms of justice, Footnote ³¹ nor does it offer any basis for meaningfully comparing them in terms of justice.

Our claim is that, when the theories of children’s and adults’ wellbeing are fundamentally different, then comparing their wellbeing is as problematic as comparing how just two societies are, when justice is determined in accordance with different societal standards. In both cases, such comparisons are either indeterminate or impossible. This has some important implications for the present purpose. If generalism is false, then one cannot make meaningful comparisons of children’s and adults’ wellbeing. It seems, however, that such comparisons are relevant in a variety of ordinary and theoretical contexts. Thus, a position that does not vindicate the normative utility of such comparisons is at a disadvantage over a position that does. This is to say that the criterion of normative utility gives us a reason to favour generalism over its denial.