Being a Child: A Social Constructivist Account

3.1. Naturalistic Accounts

The central idea underlying naturalistic accounts is that being a child is a matter of possessing, or lacking, some ‘natural’ and ‘objective’ properties. Arguably, the folk account of childness is naturalistic in this sense. According to it, to be a child is to be in a particular chronological age range, going from birth to 18 years old (or more) if adolescence is considered part of childhood, or from birth to approximately 12 years old if adolescence is considered a separate stage of life. This account counts as naturalistic because chronological age is a natural property of individuals. The problem with this account is that it is explanatorily and normatively arbitrary. It does not offer a plausible explanation of what childness is, nor of why childness is supposed to be normatively important.⁸ On the first count, there seems to be no non-arbitrary explanation available to a defender of the chronological age account as to why childness ends precisely at 18 years of age, rather than, for example, at 14, 16, 17, or even 17 and 364 days of age. On the second count, it is hard to explain why having a particular age warrants, by itself, being subject to specific moral and legal norms.

It is plausible to say that, to the extent that age matters for childness, it is only because it is correlated with the lack of some capacities that are characteristic of adultness. This idea is at the heart of the account of children as unfinished adults. According to it, to be a child is to be in a developmental stage that will lead, in statistically normal circumstances and in due course, to the acquisition of the capacities of normal adults (see Archard 2015, for an overview).⁹

This account is more promising. However, it faces one powerful objection: none of the typical ways to specify what counts as a ‘normal adult’ delivers plausible results. Broadly speaking, there are four main options: (1) statistical, (2) teleological, (3) essentialist, and (4) normative. According to (1), a normal adult is an individual that possesses the statistically normal capacities of individuals in a mature age range. According to (2), a normal adult is an individual that possesses the capacities that adults are ‘designed’ (in some sense) to have. According to (3), a normal adult is an individual that possesses the capacities that are essential for adultness. According to (4), a normal adult is an individual that possesses the capacities that one ought, practically or morally, to have as an adult. Of these options, (4) turns childness into a fully normative category. For this reason, we will not discuss it in this section.

Option (1) is clearly problematic. Some capacities that statistically normal adults possess, for example, the capacity to drive a car, seem irrelevant for determining whether one is a child, from the point of view of both descriptive adequacy and practical and theoretical utility. The same is true of (2) when interpreted in an evolutionary sense. Some capacities that evolution has designed adults to have, for example, the capacity for gossiping, do not seem to be descriptively or normatively useful to distinguish children from adults. Option (2) must be interpreted in a non-evolutionary way to remain in the contest. In fact, a popular account of childness—the so-called Aristotelian account—combines insights from both (2) and (4). According to it, to be a child is to lack the capacities that adults are ‘designed’ to have to be mature specimens of their kind (see Matthews & Mullin 2015), where the relevant capacities are individuated in a normative way, for example, as the capacities that matter for a good life.¹⁰

Option (3) appears more promising than (1) and (2). The capacities that are typically mentioned as essential to adultness include emotional regulation, a minimally stable sense of self, a capacity for planning, a capacity for rational agency, etc. These are indeed intuitively relevant capacities, which individuals that we count as children generally lack, at least to some extent and especially at a very young age. Nonetheless, this option is problematic too. It classifies young individuals who have developed faster or earlier than statistically normal children, that is, so-called ‘early developers’, as adults. Think about a 10-year-old with relatively stable preferences and sufficient autonomy. The account under discussion implies that this individual is an adult. This is in contrast with our ordinary judgments. Option (3) is problematic also on the adults’ side. In fact, some individuals that we paradigmatically count as adults may lack one or more of the capacities that are considered essential to adultness.

Before moving on and considering normative accounts of childness, we want to explore one last possibility. The upshot of the previous discussion is that there seems to be no single natural property, or conjunction of natural properties, that is both necessary and sufficient for being a child. This leaves the possibility open that childness is a cluster concept, that is, a concept defined by a cluster of natural properties. Suppose, for instance, that the cluster of properties that define childness includes chronological age, limited physical development and strength, limited emotional regulation, a limited capacity for practical reasoning, and a limited capacity for autonomy. One could maintain that, although none of these properties is necessary or sufficient for childness, to be a child is to have enough of these properties. As a matter of fact, these properties tend to be highly correlated, in the sense that they tend to be instantiated together, because of how human development typically proceeds.

As we will show below, this account represents a step in the right direction. However, in the present form it is still unsatisfactory, for it does not satisfy the criterion of explanatory adequacy. What we need is an account that explains what unifies the category of children, that is, an account that explains why individuals possessing a particular subset of natural properties count as members of one and the same category. To appreciate why this is important, notice that 12- to 14-year-old individuals share more natural similarities with adults than with infants. Yet, they are typically categorized as children alongside infants. We need an explanation of why some subset of natural properties is more important than others for our categorization system, for example, why, for the categorization of 12- to 14-year-old individuals, their similarities with infants are more important than their similarities with adults. In the present form, the cluster account of childness can be used, at most, as an epistemic guide to individuate the extension of the concept of childness, but not to identify what makes an individual a child.

3.2. Normative Accounts

The general idea underlying normative accounts is that to be a child is to possess some particular normative property. It is useful to say a few words about why it may be tempting to adopt a normative account of childness. In examining naturalistic accounts in the previous subsection, we have primarily insisted on the fact that they are descriptively and explanatorily inadequate. Perhaps, however, the most serious problem for all naturalistic accounts is that they are inadequate from the point of view of practical and theoretical utility. As we have seen in the introduction, children occupy important social roles, which are typically associated with various norms and expectations. For example, we think that children should undergo an appropriate socialization process and that they should develop certain capacities. We also expect them to behave in specific ways, appropriate to their age. This shows that childness is ‘infused’ with normativity, so to speak. A bit more strongly, it seems that the category of children has a constitutive normative dimension. As we have mentioned, some of the norms and expectations to which children are subject are sometimes harmful to them and unjust, and part of the interest in having an account of childness is to help us shed light on these morally important facts. The problem for naturalistic accounts is that, on the one hand, they have trouble accounting for childness’ inherent normativity,¹¹ for no normative conclusions follow from the possession of purely natural properties; and, on the other hand, they are not especially helpful for shedding light on the features that are most relevant for the purpose of advancing justice for children. These considerations offer powerful reasons to abandon a naturalistic account of childness and point in the direction of normative accounts.

The most common normative account conceives of childness as a matter of possessing a special moral status, one that sets children apart from adults (and other categories of individuals). It is understood that to have a moral status is to fall under the scope of a particular set of moral norms, which specify what the individual is permitted, obligated, or forbidden to do, as well as what the individual is entitled to or owed from other individuals. To be well-defined, any ‘moral status account’ of childness must specify two things. First, it must specify the content of the moral norms that constitute children’s moral status. Second, it must specify what grounds the possession of such a status, that is, which properties of children make them fitting holders of that status or, more simply, which properties are status-conferring. In what follows, we will consider the most influential version of this account in the literature, namely, the one put forward by Tamar Schapiro (1999; 2003).

According to Schapiro, to be a child is to have a ‘nonideal’ moral status. Schapiro mentions three respects in which children differ from adults from a moral point of view. First, they are fitting objects of paternalistic treatment, in the sense that paternalism is morally justified towards them, but not towards adults. Second, children’s words and actions have less moral significance than adults’, in the sense that they either carry less moral weight or warrant ascriptions of reduced moral responsibility. Third, children have a duty to pull themselves out of their nonideal condition. Related to this, adults have a negative duty not to hinder children’s moral development and not to treat them as belonging to a “permanent underclass” (1999: 735), as well as a positive duty to help children overcome their nonideal condition through appropriate forms of discipline and education.

What confers children this special moral status is their lack of independence or autonomy. In Schapiro’s words, children lack “a voice which counts as [theirs]”. In less metaphorical terms, they lack “an established constitution, that is, a principled perspective which would count as the law of [their] will” (1999: 729) and on the basis of which they can resolve and adjudicate between conflicting motivational impulses. In more explicitly Kantian terms, children lack stable maxims, expressing a principle of choice. Schapiro concludes that, because of their underdeveloped capacity for autonomy, children are “morally immature” and childness is a moral predicament, “an obstacle to morality” (1999: 735).¹²

One of the virtues of Schapiro’s account of childness is that it vindicates some of our normative beliefs about children, most notably, the belief that they ought to be protected, educated, and supported in their development from childness to adultness. The account is also action-guiding. Children are individuals who can justifiably be paternalized, that is, who ought to be treated differentially (in various ways) for their own good.

Normative accounts of this sort are nevertheless subject to two objections: one concerns their descriptive adequacy, the other their practical and theoretical utility. Consider the case of Laura Dekker, a Dutch girl who announced her decision to sail solo around the world when she was only 13 years old. In order to be allowed to pursue her journey, Dekker had to fight a lengthy legal battle, at the end of which she showed the court not just that she possessed the requisite sailing skills, but also that she was competent to make her own choices.¹³ If this is true, then Laura Dekker was not a fitting object of paternalistic treatment, at least in some areas of her life. Schapiro seems thereby committed to saying one of these two things: either Dekker was an adult simpliciter or she was an adult in the sailing domain, though perhaps a child in other domains, such as education, finance, and health care.¹⁴ Arguably, however, a more plausible option is to say that Laura Dekker was a relatively autonomous teen, vis-à-vis whom paternalistic treatment was not fitting, once again at least in certain domains.¹⁵

A related problem for Schapiro’s account is that it does not seem to match even our paradigmatic judgments about which individuals count as adults. For a start, it seems that not all individuals that we typically classify as adults possess the robust independence that Schapiro deems constitutive of adultness. In fact, some relatively well-functioning adults may never act out of ‘stable’ maxims (most of us, indeed, if the situationist critique is correct; see Doris 1998; 2002). But the problem for Schapiro gets worse if it is true—as Daniel Weinstock (2018) suggested—that Schapiro’s account of adultness must be further expanded to be normatively plausible. Weinstock argues that for an individual to count as an adult in Schapiro’s normative sense, they must not just possess stable maxims, understood as a stable set of dispositions (or rules or policies) on the basis of which they reliably act, but they must also reflectively endorse them (the ‘endorsement condition’) and be capable of reasoning with such maxims and using them reflectively to guide their actions (the ‘articulacy condition’). Only if the individual satisfies these two further conditions can the attribution of moral responsibility to them be warranted (Weinstock 2018: 53–54). The problem for Schapiro is that if her account of adultness is tweaked in this way to be normatively attractive, then it becomes descriptively inadequate. In fact, for many adults the adoption of maxims governing their actions is seldom the result of reflective endorsement. Likewise, as various empirical studies have shown (e.g., Haidt 2001), few adults decide how to act on the basis of conscious deliberation and reasons weighing. If endorsement and articulacy are necessary conditions for a normatively attractive conception of adultness, it follows that all these individuals count as children. This is deeply counterintuitive.¹⁶

It is reasonably clear from her texts that Schapiro would be unfazed by issues of descriptive adequacy. According to her, ‘child’ is primarily a moral category (1999: 717). As a moral category, ‘child’ admits of a pragmatic, rather than a metaphysically precise, application to the world (1999: 717, fn. 3). This means that, in applying the category ‘child’, we will be confronted with indeterminate, borderline cases. To make these cases determinate, we need to draw a line somewhere. Insofar as ‘child’ is primarily a moral category, the line will be set where it is more convenient for moral purposes. If this is true, then issues of descriptive adequacy are to be expected, but they are not very worrisome. In some cases, we may use the category ‘child’ in a way that does not match our intuitions. But this is ok: cases where intuitions conflict are precisely the indeterminate, borderline cases which require some “degree of arbitrary ‘line drawing’” (1999: 717, fn. 3). Ultimately, what matters for assessing whether we have drawn the line at an acceptable point is whether the category ‘child’ that we have thus configured is useful for moral theorizing.

One problem with this reply is that ‘child’ seems to be as much a descriptive category as a moral category. This is relevant because in deciding how to adjudicate cases where our intuitions conflict, that is, in deciding where to ‘draw the line’, we need to keep in mind that the category ‘child’ is used not just for moral purposes, but also for descriptive and explanatory purposes (e.g., within social theory). Because of this, issues of descriptive adequacy are more important than what Schapiro seems to think and can legitimately be used as an objection against her account.

Schapiro might insist that she is interested in the category ‘child’ only as a moral category, that is, as a category that we need for exclusively moral purposes. If this is Schapiro’s ultimate motivation for her account, then our disagreement ceases to be substantial. We are indeed happy to concede that, for the purpose with which Schapiro is concerned, a normative account of childness might be adequate. That said, we also want to stress that our purpose is different. As we have mentioned in the Introduction, our goal is metaphysical, not moral. We are interested in offering an account of the nature of childness, which can both explain social phenomena involving children and be helpful to identify the wrongs and injustices they are victims of and guide us in addressing those wrongs and injustices. Schapiro’s account is not especially useful for these purposes. It cannot explain the full range of social phenomena in which children are implicated because, as we have seen, it is not extensionally adequate. To put it differently, since Schapiro’s account does not correctly pick out which individuals count as children and which count as adults, or does not identify all the individuals that count as children, it cannot explain the full range of phenomena in which children are involved. For essentially the same reasons, it cannot identify the full range of torts to which children are subject, such as, for example, being treated paternalistically when this is not appropriate while one is still a child. We conclude that Schapiro’s account still leaves us without an answer to the metaphysical question of what childness is.

4.1. An Alternative Socially Constructed Account

In the previous subsection, we suggested that, in order to be a child, it is not sufficient to be treated like a child. To be a child is, in part, to possess certain natural properties. In Section 2, however, we argued that there is no single natural property, or conjunction of natural properties, which is both necessary and sufficient to be a child. This seems to leave us in an unstable situation. How exactly do natural properties contribute to making one a child? In what sense is childness still a socially constructed category if being a child partly depends on natural properties?

We propose the following definition of childness: To be a child in society x is to be regarded by society x’s dominant ideology as the fitting object of a set of paternalistic treatments and measures, in virtue of the fact that one possesses a cluster of natural properties such that, according to society x’s dominant ideology, it is justified to impose such treatments and measures on them.

Some clarifications are in order before proceeding. We follow Haslanger and take the term ‘ideology’ to refer to the cognitive and affective framework we use to navigate our social world, where this framework includes not just conscious beliefs, but also “habits of thought, unconscious patterns of response, and inarticulate background assumptions” (Haslanger 2012: 18; see also 447–48). Second, we take ‘x’ to be a variable that ranges over all possible societies. One implication is that our account of childness is relativistic. According to it, one and the same individual can count as a child in one society (e.g., society x₁), if they are regarded by the dominant ideology of that society as the fitting object of a set of paternalistic treatments and measures, but not in a different society (e.g., society x₂), if the dominant ideology of that society sees their status differently. We will come back to and further discuss this point below.

Our definition combines three threads of the previous discussion. The first concerns the idea that childness is associated with a special moral status. As we have seen in Section 3, there is reason to think that accounts that define children as the fitting holders of a special moral status are likely to be inadequate. Even if we grant this, however, we may still take the association between childness and moral status to be a historical and sociological fact that is relevant for shedding light on the nature of childness. In particular, there is a lesson from history and sociology that should be taken seriously. Despite the historical and cultural differences about which individuals are considered to be children, children have been systematically treated as a normatively homogeneous group, in the sense that they have been subject to the same kind of moral norms. Indeed, beyond their historical and cultural differences, such norms seem to preserve some common features: they are norms prescribing that children be treated differently from adults for their own good, that is, they are, in a broad sense, paternalistic norms.²³ Together, these considerations suggest a more fundamental idea: to be a child is not a matter of being the fitting holder of a differential moral status—one that is constituted by broadly paternalistic norms—but rather a matter of being regarded as the fitting holder of such a status.²⁴

The second thread concerns the idea that, in order to be a child, it is necessary that one genuinely possesses some relevant natural properties. To elaborate on this, it is useful to distinguish between the signs that an individual possesses certain properties and the properties themselves. Our suggestion is that to be a child, one must be regarded as the fitting holder of a special moral status in virtue of the fact that they genuinely possess certain natural properties, not in virtue of the fact that there are signs that they possess such properties. Which natural properties are relevant for one to count as a child in a given society? The short answer is: the natural properties that are regarded by the dominant ideology of that society as conferring a differential moral status to that individual. There is likely to be historical and cultural variation about what these properties are. Most commonly, they include age as well as physical, psychological, and cognitive capacities and traits, which children are supposed to have or to lack. Must an individual possess all the relevant natural properties in order to count as a child? We do not think so. Take Western societies as an example. In these societies, we typically regard lack of autonomy as a legitimate ground for paternalistic treatment, yet we still consider young individuals that have a significant degree of autonomy to be children. This suggests that the relation between childness, natural properties, and moral status is different. Our view is that, in order to be a child in a given society, one must possess enough of the properties which, according to the dominant ideology of that society, make their possessor the fitting holder of a differential moral status.²⁵

This brings us back to a third thread of our discussion. Earlier on, we considered the possibility that childness is a cluster concept, that is, a matter of possessing a subset of relevant natural properties. We saw that this account was explanatorily inadequate because it did not really explain what unifies the properties in the cluster and, thereby, the category of children. We now have a solution to this problem, one that is compatible with a social constructivist account of childness. We can say that what provides unity is the fact that the properties in the cluster are regarded by the dominant ideology of a given society as grounding children’s moral status in that society. According to our account, then, to be a child is to possess enough of the properties in this cluster.

We can explain the general idea underlying our account as follows. A group of individuals shares some similarities with respect to their physical, cognitive, and psychological capacities and traits. Such capacities and traits are morally significant according to the standards prevalent in a given society. They motivate and justify the attribution of a special moral status to those individuals, which distinguishes them from other categories of individuals in that society. Accordingly, childness is a matter of possessing a special moral status, which is socially constructed out of similarities that are thought to be normatively significant.

We can formulate our account in a simpler way. As we have seen, having a moral status involves two things: being subject to status-constituting norms and possessing particular status-conferring properties. Let us call a given set of status-constituting norms and status-conferring properties ‘a moral regime of childhood’.²⁶ Then, we can say that to be a child in a given society is to be regarded by the dominant ideology of that society as falling under a particular moral regime of childhood.

4.2. Features of Our Account

Our account presents a few characteristics that are worth discussing in more detail. The first thing to note is that which particular moral regime of childhood is adopted in a given society depends on the moral standards that are prevalent in that society. To illustrate, consider as a toy example a utilitarian society that imposes a series of paternalistic measures, including mandatory schooling, to individuals ranging from 0 to 18 years old, on the ground that this maximizes the overall social utility. According to our account, these individuals count as children in this society. These individuals are indeed subject to some specific (status-constituting) moral norms (e.g., mandatory schooling) in virtue of the fact that they possess some specific (status-conferring) natural properties (i.e., a certain age). Importantly for present purpose, what justifies, according to the dominant ideology of this society, subjecting individuals with these natural properties to those particular norms is the fact that this is required by their most fundamental moral standards (i.e., the utilitarian standards).

Second, as we have seen, the norms that constitute a moral regime of childhood are paternalistic norms, that is, norms prescribing that children be treated differently from adults for (what is regarded by the dominant ideology of a given society as) their own good. Keep in mind, however, that we use the expressions ‘for their own good’ and ‘paternalistic norms’ in a broad sense. In particular, we take a moral regime of childhood to typically include both welfare paternalistic norms, that is, norms having to do with the individual’s prudential good (as understood by the dominant ideology of a given society), narrowly considered, and moral paternalistic norms, that is, norms having to do with the individual’s moral good (as understood by the dominant ideology of a given society).²⁷

The fact that some individuals are regarded by the dominant ideology of a given society as fitting objects of a set of broadly paternalistic norms, in virtue of the possession of a cluster of natural properties, is what makes these individuals ‘children’. It is what unifies the category of children. That said, it is important to recognize that a moral regime of childhood typically includes several heterogeneous elements. For one, the norms in the regime may be of different types. They may include norms of permissions, obligations, prohibitions, as well as entitlements and rights. A moral regime of childhood typically includes norms that apply to children (e.g., the norms about what children are permitted, obligated, or forbidden to do), but also norms that apply to other individuals, such as parents, educators, and other citizens (e.g., the norms about what these individuals are permitted, obligated, or forbidden to do in relation to children). These norms may concern different domains, for example, children’s subjective wellbeing, development, education, labor, moral responsibility, and so on. They may include, for instance, parental obligations of care as well as a duty to meet children’s physical and psychological needs; children’s rights to education and health care; the recognition of their limited moral and legal responsibility, etc. Finally, and importantly, the relevant paternalistic norms may be different for different sub-categories of children, so that (e.g.) the norms concerning teenagers differ from the norms concerning toddlers (see Subsection 4.3 for more details).

Similar considerations apply to the status-conferring properties. Moral and political philosophers often point to lack of autonomy as the main defining characteristic of childness. However, there may be other properties that ground specific paternalistic norms and that constitute a moral regime of childhood in a given society. For instance, status-conferring properties of this kind may include age, limited physical development and strength, limited emotional regulation, and a limited capacity for practical reasoning.

Third, and related to the previous point, it is important to emphasize that, according to our account, being a child in a given society does not entail being regarded by the dominant ideology of that society as the fitting object of paternalistic norms in all areas of life or in all situations. For one, the dominant ideology of that society may regard paternalizing individuals that possess certain relevant natural properties as justified only in some domains of life, but not in all.²⁸ Or the dominant ideology of a given society may regard some individuals as the fitting objects of paternalistic norms in a given domain in general, but also hold that these individuals should not be subject to paternalistic treatments in that domain in some specific circumstances. The dominant ideology may indeed include conditions of applicability or appropriateness of paternalistic treatment, which specify under which conditions a particular individual that counts as the fitting object of paternalistic treatment in a given domain ought or ought not to be paternalized. Finally, in addition to conditions of applicability concerning the domains and circumstances in which paternalistic treatment is appropriate, the dominant ideology of a given society may also specify who should paternalize children in particular domains and contexts, who should have primary paternalistic authority over them, and how much authority they should have.

Fourth, as we mentioned before, the norms constituting a moral regime of childhood vary both historically and culturally. For example, John Gillis claims that in early modern periods “the idea that [children] deserved a time and a space all their own was as unthinkable as it was impracticable” (Gillis 2011: 114). Likewise, the properties that are thought to justify the attribution of a special moral status also vary historically and culturally. For instance, around the seventeenth century, children’s “sinful souls” and “potential for evil” were thought to be reasons for harsh discipline and punishment (Hendrick 2011). While some might still harbor these views nowadays, it is fair to say that they are no longer mainstream. Finally, it is important to notice that which norms apply to which individuals typically also depends on features such as their gender, race, class, and so on. These norms as well are historically and culturally variable.

Fifth, the norms that constitute a given moral regime of childhood, which children are subject to in a given society, may be incorrect moral norms. That is, they may not be the norms that really ought to apply to those individuals. For instance, corporal punishment was for a long time a daily experience for children enrolled in schools, being thought of as a perfectly apt instrument to teach students obedience. To give another example, the current trend towards the institutionalization and hyper-structuring of play might not be beneficial to children, as it appears to reduce their spontaneity and enjoyment.²⁹ It may also be the case that there are other norms that ought to apply to those individuals beyond those in the prevalent regime.

Likewise, the properties that are taken to be status-conferring in a given society may not be actually morally significant or may not include all the properties that are actually morally significant. For instance, age is often considered to be a status-conferring property, particularly in contemporary Western societies, where the fact that an individual has a certain age is often regarded as a morally relevant consideration for subjecting that individual to a range of paternalistic norms and treatments. It might be, however, that age is just a sign that the individual possesses some other properties that are morally significant.³⁰

Sixth, while our account appeals to status-conferring properties to characterize childness, it recognizes that attributions of childness are typically based on signs that an individual possesses such properties. For instance, in addition to being considered a morally relevant property, chronological age is also typically regarded as a sign of childness (although not a universal one; see Woodhead 2011: 52), and used as evidence to classify one as a child. Importantly, however, attributions of childness may be mistaken. That is, one may judge that an individual is a child by observing that she has features, for example, a very high-pitched voice, which are normally correlated with properties that are regarded as status-conferring in their society, for example, psychological immaturity; yet, their judgment may be mistaken because that individual is simply an adult with an unusual voice.

Seventh, although they are often not regarded as such (especially in contemporary Western societies), children are as much contributors to an ideology as recipients of it. As many scholars in childhood studies insisted upon, children have and exercise agency. They are social actors that can actively participate to the construction and change of societal norms and practices (see James & Prout 1997; James 2011; Corsaro 2018). For instance, the Gillick case, in which a 15-year-old English girl sought contraceptive advice from a local doctor without informing her parents, offers an example of how a teenager’s behavior contributed to changing both people’s perception of the relationship between age and competence and the legal system itself (see James & James 2004: ch. 6, for a discussion).

This brings us to another observation. The dominant ideology in a given society is typically sustained by social, political, and legal institutions. This explains the sense in which to be a child can also be thought of as a matter of possessing a particular institutional status (Qvortrup 1987; Jenks 1992; Mayall 2002). Relatedly, insofar as childhood is shaped by institutions such as the family, schools, the economic and legal systems, which (amongst others) constitute what we may call the (basic) structure of society, then it makes sense to talk about ‘childhood’ not just as a period of life, but also as a ‘structural form’ (Qvortrup 2011), that is, as a social institutional configuration (with both stable and changing elements) through which different generations of children pass over the years.

4.3. Advantages

In this subsection, we want to show that our account satisfies all the criteria of success listed in Section 2. Let us start with descriptive adequacy. Our account seems to match well our ordinary judgments about which individuals count as children. It preserves the common idea that the category of ‘children’ is infused with normativity and, more specifically, that it is tied to paternalistic norms. Our account is also compatible with the claim that early developers are children, provided that they possess enough of the properties that the dominant ideology takes to ground paternalistic treatment. Let us reconsider, for instance, the case of Laura Dekker. At the time when she started her battle to sail solo around the world, Dekker was 13 years old and undergoing some physical and psychological development characteristic of this period of life. In contemporary Western societies, these features are typically regarded as justifying a wide range of paternalistic norms and treatments. Age, in particular, is often taken to be a prima facie sufficient reason for imposing such norms and treatments. If this is true, then, despite displaying a considerable degree of personal autonomy, in addition to competent sailing skills, Laura Dekker still possessed enough of the properties that, according to the dominant ideology of her society, grounded the attribution to her of the differential moral status proper of children. In other words, according to the standards of her society, Laura Dekker was a child that happened to display a surprising, and rather rare for her age, degree of autonomy and independence.

As we have mentioned earlier on, our account entails that, to the extent that our ordinary judgments reflect the standards prevalent in our society, we typically classify as children even individuals that are not, or were not, considered children in their own societies. For instance, we count as children even 10-year-old individuals living in the Middle Ages, despite the (alleged) fact that they were not regarded as such at the time, given the prevalence of different societal standards. Our account of childness is thus relativistic. Whether one counts as a child in a given society depends on the dominant ideology prevalent in that society, so that one and the same individual may count as a child in one society, but not in another. This seems to be a welcome implication. We will consider some possible objections in Subsection 4.4.

Consider now the criterion of explanatory adequacy. Our account explains what children have in common, despite their biological differences and the differences in their historical and cultural contexts, and what makes them ‘children’: to be a child is to be regarded as falling under a differential moral regime according to a society’s dominant ideology. Despite its relativism, our account is still unifying, for it points to a common element that all children share: they are all subject to norms of a broadly paternalistic kind, in virtue of the possession of a cluster of relevant properties. At the same time, our account can recognize the heterogeneity of children of different ages. Infants, young children, and teens are similar in some respects, but not in others. However, all these individuals count as children insofar as they possess enough of the properties that ground specific forms of paternalistic treatment according to the dominant societal standards.

Finally, consider the criterion of practical and theoretical utility. Our account is useful insofar as it provides a framework for submitting to rational scrutiny the moral regime of childhood prevalent in our society. By emphasizing the contingent association between childness and a given moral regime of childhood, it invites us to critically assess its main components: the moral norms that we take children to be subject to, the features that we consider as grounds for such norms, and the moral standards on the basis of which the adoption of a moral regime of childhood is justified.

More specifically, our account can explain various phenomena involving children which are relevant from a moral and political point of view. First, it explains why the power imbalance existing between children and adults seems to be a constitutive feature of childness. The paternalistic norms that constitute a regime of childhood in a given society are norms issued and enforced by adults. Because of this, children are effectively under adults’ authority. Moreover, to the extent that it conforms to the society’s standards, adults’ authority is regarded as legitimate by the dominant ideology.

Second, our account explains why children are generally regarded as incompetent by adults, although they are not necessarily so. The reason is that the paternalistic norms and treatments that apply to them in a given society are supposed to be justified by children’s presumed physical, cognitive, emotional, or psychological incompetency. Yet, in many cases, children are not incompetent in the way envisaged.

Third, our account helps identify and explain various forms of injustice to which children are subject. We have already mentioned two (non-mutually exclusive) salient forms of injustice. First, an individual may be classified as a child in virtue of the fact that they possess certain properties that are regarded as status-conferring in their society, although these are not morally significant properties. Second, an individual may be correctly classified as a child given the standards prevalent in their society, but the norms to which that individual is subject are incorrect moral norms (or else, the society has failed to apply the correct moral norms at the appropriate time). In these cases, the individual is subject to adults’ authority in a way that is morally illegitimate.

There are other problematic cases. Consider the case in which an individual is mistakenly classified as a child because they are erroneously taken to possess the properties that are thought to ground paternalistic norms and treatments according to the dominant ideology in their society. An injustice occurs whenever the individual is subject, as a result, to paternalistic norms that they, objectively, ought not to be subject to.³¹ Consider also the symmetrical case in which an individual is mistakenly classified as an adult. Incorrect ascriptions of this sort are morally problematic whenever they exclude from special forms of protection and care individuals that, objectively, ought to be subject to paternalistic norms and treatments. In her essay “Black Girlhood, Interrupted” (2019), Tressie McMillan Cottom vividly describes the latter form of injustice, to which African American girls are subjected.

This brings us to another important point. Above, we have argued against a Haslangerian account of childness on the ground that the category of children may still serve a useful purpose even once children’s oppression is terminated. We can now explain why. There are individuals that really ought to be subject to broadly paternalistic norms, treatments, and protections. Put differently, there is such a thing as a correct moral regime of childhood. It is one in which a differential moral status is ascribed to some individuals on the basis of properties that are genuinely morally relevant, and which involves paternalistic norms and treatments that are correct and that genuinely ought to apply to such individuals.³² In a society where all of the injustices mentioned above have been eliminated, then the category of children can serve a useful purpose, namely, the purpose of grouping together individuals that really ought to be treated differently from adults. In such a society, all the individuals that are classified as children will be individuals that fall under a correct moral regime of childhood.

Note that if Schapiro is right in maintaining that the correct moral regime of childhood is one in which we ought to subject individuals that lack autonomy or independence to paternalistic norms for Kantian reasons, then, in a non-oppressive society, the individuals that count as children according to our account will coincide with the individuals that count as children according to Schapiro’s. That is, our account will be extensionally equivalent to Schapiro’s normative account. This does not impugn the practical and theoretical utility of our account. While the ultimate goal is to produce a society where all the individuals that are regarded as children are those, and only those, that really ought to be subject to the norms of the correct moral regime of childhood, and while Schapiro might be right that the correct moral framework is a Kantian one, in order to realize a society of this sort we need to start from an account which can help us identify the individuals that are currently wronged by our system and which can guide us towards correcting the injustices to which they are subject.

4.4. Objections and Replies

In this subsection, we discuss and reply to some objections that can be raised against our account. The first objection targets its relativistic character. As stated above, our account implies that one and the same individual may count as a child in one society, but not in another, depending on the moral regime of childhood that is deemed to be justified by the dominant ideology of those societies. Critics may argue that this feature of our account generates counterintuitive implications. To take two extreme examples, our account entails that there would be no children in a society whose dominant ideology held that no one should be subject to paternalistic treatments and norms (imagine a society dominated by radical child liberationists); or, conversely, that if a society’s dominant ideology held that everyone should be subjected to the same kind of paternalistic treatments and norms, then there would only be children in that society.³³ This seems counterintuitive.

We are willing to bite this bullet. To paraphrase what Haslanger says about gender (2000: 38), our proposal is to treat the category of children “as a genus that is realized in different ways in different contexts”. We are then willing to accept that if, for instance, Philippe Ariès is right in maintaining that individuals in what we now consider middle childhood were not considered fitting objects of differential moral norms in the Middle Ages, then there were no ‘middle children’ within that society, according to their standards; or that in the societies described in the previous examples, there were, respectively, no children and only children, according to their standards. Two things should be noticed, however. First, this is compatible (and actually an implication of our account) with saying that there were ‘middle children’ in the Middle Ages, according to our own standards; and that there were, respectively, some children and some adults, according to our own standards, in the two previous examples. Second, as we have hinted at in the previous subsection, it is compatible with our account to maintain that a society’s moral norms and standards may be mistaken, and objectively so. Thus, for instance, a society whose dominant ideology holds that no one should be subject to paternalistic treatments is likely to have mistaken standards, for in a typical society there are, in fact, individuals who ought, objectively, to be treated paternalistically, in virtue of the (physical and psychological) properties that they possess. This claim is consistent with our account, since relativism about childness does not entail relativism about morality.³⁴

A second objection is that, to the extent that our account of childness is relativistic, it implies that intersocietal disagreement about who counts as a child is not a substantive disagreement. However, intersocietal disagreement is typically experienced as substantive by the members of these societies. But then, our account cannot make sense of their experience. This reduces its practical usefulness; more specifically, its explanatory power.

Our response is the following. It is indeed an implication of our account that, if two societies adopt distinct moral regimes of childhood, then intersocietal disagreement about who counts as a child turns out not to be a substantive disagreement. We can nevertheless still explain why members of these societies typically experience such a disagreement as substantive. As stated above, ours is a metaphysical account of childness, that is, an account of the nature of childness. This account may not coincide with how the members of different societies themselves understand childness, that is, it may not correspond to their concept of childness. In fact, it seems to us that this is true in Western societies, where many people simply assume the truth of a naturalistic view of childness according to which to be a child is to be in a particular age range. Were a member of a different society to argue that an individual in that age range is not a child—for instance on the ground that that individual is sufficiently autonomous—members of Western societies would most likely react by imputing to their interlocutor a mistake about what it is to be a child. All this is compatible with our account. In this scenario, the members of both societies understand childness as a naturalistic category. For the reasons offered in the paper, however, we think that both are mistaken. Their understanding does not correspond to what childness really is (as specified by our account). Because of this, they experience their disagreement about who counts as a child as substantive, even though, as a matter of fact, it is not.

It is worth reiterating that, although our account implies that intersocietal disagreement about who counts as a child is not a substantive disagreement, it does not imply that intersocietal disagreement about the constituents of a moral regime of childhood is also not a substantive disagreement. In fact, our account leaves ample room for substantive disagreement of this kind. Members of these societies may substantively disagree about which properties constitute proper grounds of child status, or about which norms the individuals that possess these properties should be subject to, or about the justification for subjecting individuals with the relevant properties to norms of a paternalistic kind.

A third objection is that, even if we leave its relativistic character aside, our account delivers the wrong verdicts about who counts as a child, for it categorizes as children individuals that should not be thus categorized. One version of the objection is the following. In contemporary Western societies, we think that mature individuals too are subject to a variety of paternalistic norms and laws, such as (e.g.) wearing seat belts in a car or a helmet while cycling. It seems that our account categorizes those individuals as children (in those societies). This is counterintuitive.

However, this is not what our account implies. Whether or not one is a child in a given society is determined by whether one possesses enough of the properties that are regarded as grounding the particular set of paternalistic norms and treatments that constitutes the regime of childhood of that society. We have claimed earlier on that in contemporary Western societies, the cluster of relevant properties typically includes age, limited physical development and strength, limited emotional regulation, a limited capacity for practical reasoning, and so on, and that these properties are regarded as grounding a specific set of paternalistic norms in domains such as education, health, and legal responsibility. Mature individuals typically do not possess enough of these properties and are therefore not subject to the specific set of paternalistic norms that such properties are taken to ground. Consequently, our account implies that they do not count as children (in these societies). Simply put, being paternalized or being regarded as a fitting object of paternalistic treatment in some contexts (justifiably or not) is not sufficient for one to fall under the moral regime of childhood of a given society and thus to count as a child.³⁵

The objection can, however, be pushed further.³⁶ What if the dominant ideology of a given society construes the moral regime of childhood broadly enough so that entire groups of mature individuals end up falling under its scope? For example, imagine a society whose dominant ideology does not assign chronological age a lot of weight within its moral regime of childhood, but includes lack of some psychological capacities, for example, limited autonomous decision-making, amongst its child-making properties. Imagine, furthermore, that as a result of that society’s oppressive standards, women are prevented from developing their psychological capacities in such a way that they actually end up possessing the child-making properties that are considered relevant in that society (that is, they are not just perceived to possess these properties, as is often the case in cultures of oppression³⁷). Our account implies that these women count as children in that society. Once again, the objector holds, this is counterintuitive.

It is true that our account implies that in the profoundly oppressive and unjust society described above, women would count as children, according to the standards of that society. However, we do not see this as a counterintuitive implication of our account. In fact, our account allows us to recognize the type of profound injustice to which women are subject in that society. These women are wrongfully kept in a state that the dominant ideology of their society regards as grounding a specific set of paternalistic norms and treatments, by being prevented from developing capacities that they have a moral and political right to develop. Promoting (or at least not hindering) the development of other people’s autonomy is indeed an important obligation that we have individually and collectively, according to pretty much all the main ethical theories and theories of justice. More generally, it is important to keep in mind that paternalistic treatment is morally appropriate not just when it is exercised for the right reasons and with respect to fitting targets, but also in the right way, that is, in a way that truly serves the present and future interests of the paternalized individual. Women in our example are wronged in all these respects. The paternalism they are subject to radically fails to serve their interests. Although they count as children according to the dominant ideology of their society, their falling under their society’s regime of childhood is thus the result of deep injustices. This implies that the regime of childhood in question is profoundly wrong.

One might object that our explanation of the wrong suffered by women in the previous example does not get to the bottom of things. The most plausible explanation of why women are wronged is that they are treated like children, even though they are adults. The idea is that, because the women in our example are adults, they should not be subject to paternalistic norms and treatments, even though they have an underdeveloped capacity for autonomy. By being subject to these paternalistic norms and treatments, they are wronged by the other members of society.

Let us note, for a start, that this explanation presupposes a non-relativistic account of childness (and adultness), for example, one according to which to be a child (an adult) is to be in a particular age range, such that one can (cannot) be legitimately subject to a broad range of paternalistic treatments. As such, this explanation is not available to us. As we have argued in the paper, we think that there is no single natural property, or conjunction of natural properties, that all children have in common, in virtue of which they count as children. Still, one may insist that any alternative explanation of the wrong suffered by women in our example that does not presuppose a non-relativistic account of what children are and of how they should be treated risks being unintelligible.³⁸

To address this objection, we need to show that there exists a plausible explanation of the wrong suffered by women in the example that is both intelligible and that does not presuppose a non-relativistic account of childness. Here is one of the possible explanations available to us. (While we think that the following explanation is independently plausible, we offer it here simply as one way to illustrate how the previous objection can be addressed.) It is plausible to think that respect for other people’s autonomy requires, amongst other things, securing the conditions and opportunities necessary for meaningful life authorship, that is, the conditions and opportunities that are necessary for an individual to live a globally autonomous life (see Franklin-Hall 2013). What these conditions and opportunities are may, of course, vary depending on the context and on the specific (physical, cognitive, psychological) characteristics of each individual and the stage of development they are in. Yet, broadly speaking, we can say that an individual’s autonomy rights are violated whenever they are deprived of the conditions and opportunities required for meaningful life authorship.

With this is mind, let us go back to the previous case. On the account under consideration, the wrong suffered by women is not that they are treated like children even though they are adults (according to a non-relativistic account of childness and adultness). Rather, the wrong is that because of their society’s oppressive norms and culture, they have been prevented from developing the capacities, and were not given the opportunities, that are necessary for having meaningful life authorship. Note that this explanation does not presuppose any non-relativistic account of childness. In fact, it is compatible with our account of childness. We can indeed say both that women in our example count as children in their deeply unjust society and that they have been wronged on the ground that they have been prevented from having meaningful life authorship. If this is true, it follows that to understand the wrong suffered by women in our example we do not need to adopt a non-relativistic account of childness or adultness.