Article

Being a Child: A Social Constructivist Account

Authors
  • Andrée-Anne Cormier (York University, Glendon College)
  • Mauro Rossi (Université du Québec à Montréal)

Abstract

In recent years, many scholars have offered innovative accounts of social categories such as gender, race, and disability. By contrast, comparatively little work has been done on the category of children. The goal of our paper is to offer a new account of what children are. We start by discussing the two main accounts that have been put forward so far in the literature: naturalistic accounts and normative accounts. According to the former, to be a child is a matter of possessing, or lacking, some ‘natural’ and ‘objective’ properties. According to the latter, to be a child is to possess a particular normative status. We argue that both naturalistic and normative accounts are subject to a variety of objections. We then propose our own social constructivist account. According to it, to be a child in any given society is to be regarded by the dominant ideology of that society as the target of a set of broadly paternalistic norms, in virtue of the possession of a cluster of natural properties such that, according to the dominant ideology of that society, it is justified to subject that individual to such norms. We conclude our paper by showing that our account meets the criteria of success for an account of what it is to be a child and by addressing some objections.

Keywords: children, adults, childness, childhood, social construction, children’s moral status, social categories

How to Cite:

Cormier, A. & Rossi, M., (2023) “Being a Child: A Social Constructivist Account”, Ergo an Open Access Journal of Philosophy 9: 39. doi: https://doi.org/10.3998/ergo.2908

609 Views

310 Downloads

Published on
17 Jul 2023
Peer Reviewed

1. Introduction

In the past few decades, childhood studies have emerged as a new multidisciplinary academic field, on the model of gender and disability studies. Participating in this trend, the philosophy of childhood has attracted significant interest and is now recognized as a distinct area of philosophy, rather than just a branch of philosophy of education (see Matthews & Mullin 2015). An increasing number of philosophers have started investigating issues concerning children’s moral status, children’s autonomy and development, children’s rights, parents’ and society’s obligations towards children, children’s wellbeing, and the value of childhood.

As a prolegomenon to these inquiries, one would expect to find a sizable philosophical literature devoted to the question of what a child is. However, there are very few works directly addressing this question. In fact, the latter has so far mostly received a cursory treatment, confined to a few scattered remarks in articles exploring other issues about children. At least at first sight, however, both the framing and the outcome of many of the most central debates in the philosophy of childhood are likely to depend on the answer to the question of what children are. For instance, this seems relevant for determining what is owed to children as a matter of justice, and how they ought to be treated by adults. The same is true for other key questions currently debated in the literature:1 Is being a child bad for children? Is childhood less valuable than adulthood? Are there any special goods of childhood, that is, things that are especially or uniquely good for children, but not for adults? Is children’s wellbeing the same as adults’ wellbeing?

Some scholars have explicitly noted the importance of the issue. For instance, David Archard and Colin Macleod stated that “[. . .] any discussion of the moral and political status of the child should be grounded in a logically prior agreement on who should be counted a child” (Archard & Macleod 2002: 14). Similarly, Gareth Matthews and Amy Mullin claimed that “[h]ow childhood is conceived is crucial for almost all the philosophically interesting questions about children. It is also crucial for questions about what should be the legal status of children in society, as well as for the study of children in psychology, anthropology, sociology, and many other fields” (Matthews & Mullin 2015: 3). The problem is that while most philosophers treat the category of children as unified, they fail to explain what, if anything, children have in common.2

The goal of our paper is to offer an account of what it is to be a child. Our project is metaphysical in nature. We are not interested (if not indirectly) in the meaning of our folk concept of child or how our conceptions of children have changed across history.3 Rather, we are interested in understanding what makes an individual a child. This is often presented as the issue of the ‘nature’ of childhood. However, the notion of childhood can be understood in two different ways, as referring either to the state of being a child or to the period of life when one is a child (see Hannan 2018, for this distinction). To avoid confusion, we will use the term ‘childness’4 to refer to the state of being a child and reserve the term ‘childhood’ for the time period when one is a child. In this sense, we can say that our paper is about childness, rather than childhood.

What is childness, then? It is a platitude that being a child is distinct from being an adult and that the former must be characterized, at least in part, by contrast with the latter. However, this general idea can be developed in a number of ways. In this paper, we discuss and reject the two main accounts that have been put forward so far in the philosophical literature: naturalistic accounts (Section 3.1) and normative accounts (Section 3.2). Our account belongs to the family of social constructivist theories (Section 4). Sociologists have been exploring accounts of this sort for a while. By contrast, while philosophers have offered social constructivist accounts of other important social categories, such as gender and race, they have not, to our knowledge, offered a social constructivist account of childness. After formulating a brief hypothesis as to why this is the case, we provide an in-depth discussion of what we take childness to be. Roughly, our idea is that to be a child in any given society is to be regarded by the dominant ideology of that society as the fitting object of a set of broadly paternalistic norms and practices. Before doing all this, we provide a series of criteria for evaluating competing accounts of childness (Section 2).

2. Methodological Criteria

As the feminist literature on the metaphysics of gender reveals,5 the project of providing an account of a social category may split into many, often very different projects, depending on the criteria of success that one adopts. The same applies to the project of giving an account of childness. Here, we propose to adopt three criteria: descriptive adequacy, explanatory adequacy, and practical and theoretical utility.6

The criterion of descriptive adequacy holds that a plausible account of childness must fit our paradigmatic judgments about which individuals count as children. In other words, a plausible account must classify clear-cut cases of children as children and, conversely, exclude clear-cut cases of individuals that are not children.

In addition to being extensionally adequate, a good account of childness must also be explanatorily adequate. That is, it must explain why some individuals count as children and others do not. The category of children typically includes individuals inhabiting rather different stages of life, namely, infancy and toddlerhood (from birth to 2 years old), early childhood (from 2 to 6 years old), middle childhood (from 6 to 11–12 years old), teenhood and adolescence (from 11 to 18 years old7) (Berk 2017). An account of childness must explain what individuals ranging from newborns to adolescents (or at least teens) have in common, in virtue of which they count as children. Put differently, a plausible account must be unifying: it must tell us what unites such a heterogeneous group of individuals.

The third criterion states that an account of childness must be practically and theoretically useful. In other words, an account of childness must be able to satisfy our interests in having the category of ‘children’. This requires some clarification. It should be noted from the start that we need the category of ‘children’ for a variety of different purposes, for example, scientific, social, moral, political, and so on. It is highly unlikely that a single account of childness will satisfy all these purposes at once. Thus, it must be recognized that there might exist different accounts of childness, all of which may be useful for the specific (e.g., legal, medical, etc.) purposes they are meant to serve. How should the criterion of practical and theoretical utility be interpreted, then? In this paper, we are looking for an account of childness that can play a role within social theory alongside categories such as gender, race, and disability. More specifically, we are looking for an account that can, on the one hand, be used to explain a range of social phenomena in which children are implicated and, on the other hand, serve as an effective practical tool for identifying, understanding, and changing social practices involving children that raise moral and political issues.

In particular, an account of childness must be sensitive to, and help explain, three related things. The first is that children typically occupy particular social roles, for example, within the family or in school. The second is that childness is typically associated with a distinct set of moral and legal norms, although these norms are contestable and have indeed been challenged, for instance by child liberationists (Farson 1974; Holt 1975; Cohen 1980), who have questioned the rigid demarcation between children and adults that characterizes contemporary Western societies. The third is that children stand in a relation with adults that typically presents some morally significant features. These include, for instance, (i) the power imbalance existing between children and adults, (ii) the fact that children are generally considered less competent and autonomous than adults, (iii) children’s subjection to various forms of subordination, silencing, and marginalization, but also (iv) the important role that adults play or should play in children’s development.

One important caveat is in order before proceeding. While we have included descriptive adequacy as one of our criteria of success, we do not mean to suggest that our ordinary judgments about who counts as a child—not even our judgments about paradigmatic cases—are sacrosanct and cannot be questioned. Descriptive adequacy is indeed only one of the relevant criteria, which must be weighed against the others. There is no way to decide a priori exactly how weighty each criterion should be. Methodology is not independent from substantive theorizing in this case. The only thing that we can say from the start is that our paradigmatic judgments provide some evidence in favor or against alternative accounts. Yet, this evidence is defeasible. If there is good reason to do so—for instance, if the criterion of practical and theoretical utility is given most prominence and if it turns out that the most useful account of childness is somewhat revisionary—then our paradigmatic judgments about children should be rejected.

3. Standard Accounts

In this section, we examine the two most prominent families of accounts of childness currently on offer: naturalistic accounts and normative accounts.

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. Social Constructivist Accounts

The most significant remaining possibility is that childness is constituted by social facts.17 According to social constructivist views, childness should be conceived of as a social construction, the product of social relations, positions, or hierarchies. Sociologists of childhood and childhood studies scholars overwhelmingly, if not consensually, believe that an account of this sort is indeed true of childness.18

Support for this perspective typically comes from a large body of studies showing a significant historical and cross-cultural variability in the conceptions and treatments of children. One does not need to accept Philippe Ariès’s (1962) claim that the concept of childhood is a modern invention to recognize that this is the case. It suffices to think about how laws and practices have changed in Western societies alone in domains such as child labor, schooling, rearing, and punishment. Consider an example of historical variability, taking the UK as a benchmark. We normally think that children should not be employed in industries and businesses, at least not if labor harms their physical, cognitive, and psychological development. However, laws regulating or forbidding child labor are quite recent. They appeared in the UK only at the onset of the Industrial Revolution, at the beginning of the nineteenth century. This is not because in previous centuries child labor did not exist. Quite the contrary, in the early modern period, large numbers of children were used as labor force. However, their labor simply raised few or no moral concerns (see Hendrick 2011: 108). Consider, now, an example of cultural variability. In Western societies, it is customary for parents to adopt a hands-on approach to their children’s development. This contrasts with the attitude that Mary Martini observed in the 1970s in Ua Pou, an island in French Polynesia, where “as soon as a child learned to walk, his mother turned him over to the care of other children, a group that adults distantly oversaw but little ones governed” (Hopgood 2012: 319).

Despite the appeal of social constructivism in sociology and childhood studies, the idea that childness is socially constructed has not been philosophically investigated in detail. This is surprising, especially considering that several philosophers have offered rich and innovative social constructivist accounts of other social categories, such as gender, race, and disability (Haslanger 2012; Mills 1998; Barnes 2016). Part of the explanation for this neglect may be that providing an account of the latter categories is often thought to be more pressing from the point of view of social justice.19 Another reason may be that the main social constructivist accounts that have been proposed for other categories are not so easily translatable into accounts of childness. As an example, consider Sally Haslanger’s (2000; 2012) influential account of gender and race. Modeling an account of childness along Haslangerian lines would involve saying that to be a child is to occupy a subordinate position in the social hierarchy, systematically disadvantaging children along some dimensions, in virtue of some features that they are perceived or imagined to have. More specifically, a direct transposition of Haslanger’s account of gender to childness would look something like this:

C is a child if and only if:

  1. C is regularly and for the most part observed or imagined to have features presumed to be evidence of biological immaturity;

  2. That C has these features marks C within the dominant ideology of C’s society as someone who ought to occupy certain kinds of social position that are in fact subordinate (and so motivates or justifies C’s occupying such position); and

  3. The fact that C satisfies (i) and (ii) plays a role in C’s systematic subordination, that is, along some dimension, C’s social position is oppressive, and C’s satisfying (i) and (ii) plays a role in that dimension of subordination.

At first sight, this view has some plausibility. Individuals we observe, or imagine, to have features presumed to be evidence of biological immaturity often do occupy subordinated positions in society, along with other groups such as women and racialized minorities. At least this is true if we understand subordination as involving such things as having little control over one’s life and, more generally, being disempowered. As a matter of fact, children as a group have significantly less political, social, and economic power than adults. Adults have authority over them. Adult-child relations are clearly hierarchical, with adults on top of that hierarchy. Children do not even have the legal right to choose or exit their families at will, should they wish to do so. Thus, it is hard to deny that children vis-à-vis adults, like women vis-à-vis men, occupy subordinated social positions.

Nevertheless, the view under consideration has two problematic aspects. First, it is implausible to say that, in order to count as a child, it is sufficient that one be perceived or imagined to be biologically immature and subjected to systematic subordinating social roles on this basis. Consider a fully autonomous and biologically mature agent, for example, a 40-year-old man, who, because of some unusual physical appearance, is perceived to be a child, and thus treated like a child. Is this person a child? It seems doubtful. It is more plausible to say that such an individual is mistaken for a child, or mistakenly treated like a child, not that he is in fact a child. It seems that to be a child has at least something to do with chronological age or how one’s mind and body really are.20

The second problem is that, although children arguably are in many ways oppressed21 as a group, being a child is not reducible to occupying a subordinated and oppressive social position. This marks another important difference between childness and gender (or race). According to Haslanger, to be a woman is nothing but to be subject to systemic subordination and oppression. So, the category ‘women’ (at least as we know it) will disappear once we achieve gender justice. But it seems that there are good reasons to preserve the category ‘children’ even after we have eliminated their oppression.22 That is, the category of children may still serve a useful purpose even after children’s oppression is terminated—as we will show in Subsection 4.3.

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.

5. Conclusion

In this paper, we have considered the oft-neglected question of what it is to be a child. We have examined and rejected the main accounts offered in the literature: naturalistic and normative accounts. We have proposed a new social constructivist account, according to which to be a child is to be regarded as the fitting object of a set of paternalistic treatments and measures, in virtue of the possession of a cluster of relevant natural properties. We have argued that this account presents several advantages over its competitors and addressed some objections that it faces.

Acknowledgements

We are very grateful to Paul Bou-Habib, Amandine Catala, Aron Edidin, Keith Frankish, Anca Gheaus, Angela Hobbs, Colin Macleod, Sebastiano Moruzzi, Serena Olsaretti, Joshua Rust, Jordan Walters, Daniel Weinstock, Vanessa Wills, the participants to the Family Justice Work-in-progress Workshop at the Universitat Pompeu Fabra in Barcelona, the participants to the 2nd Critical Social Ontology Workshop at Saint Louis University, and the participants to the Philosophy and Childhood conference at the University of Bologna for their generous feedback on previous drafts or presentations of this paper. We would also like to thank two anonymous referees for this journal for their very helpful comments and suggestions. Finally, many thanks to Ashton Black and Robert Mason for their editing assistance. This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 Research and Innovation programme [Grant Agreement Number: 648610].

Notes

  1. See, for instance, Brennan (2014), Brighouse and Swift (2014), Cormier and Rossi (2019), Gheaus (2015a; 2015b), Hannan (2018), Skelton (2015), Tomlin (2018), Weinstock (2018).
  2. In the literature on gender, this is sometimes called ‘the problem of commonality’. See Haslanger (2000).
  3. For an analysis of how children have been portrayed and conceived in the history of political philosophy, see Kane (2016).
  4. This term is sometimes also used by sociologists of childhood. See, e.g., Alanen (2011).
  5. See, for instance, Haslanger and Saul (2006).
  6. We follow quite closely (mutatis mutandis), in letter (for the criteria of descriptive and explanatory adequacy) and in spirit (for the criterion of practical and theoretical utility), the criteria offered by Barnes for assessing different accounts of physical disability (see Barnes 2016: esp. 10–13 and 39–43).
  7. More than for all the other periods, the end of adolescence is somehow arbitrary, being mostly fixed so as to match the legal age. In fact, in industrialized countries, the transition from childhood to adulthood has recently become longer than what this classification suggests. In order to keep this into account, many scholars now distinguish a period of ‘emerging adulthood’, which goes roughly from 18 to 25–30 years old.
  8. This account is also descriptively inadequate. Suppose, for instance, that scientists become one day able to create a fully developed individual from the laboratory, possessing all the capacities commonly associated with adults. Since their chronological age is that of a newborn, the present account implies, counterintuitively, that such an individual is a child.
  9. The qualification ‘in statistically normal circumstances’ is meant to take into account the existence of children who will never develop into adults due to death, disease, or other preventing circumstances.
  10. The Aristotelian account is vulnerable to the same objections that we raise against normative accounts in Subsection 3.2.
  11. It is, of course, open to a defender of a naturalistic account of childness to deny that childness is inherently normative and, therefore, that this is a feature that an account of childness needs to explain.
  12. It is important to understand the difference between Schapiro’s account and a naturalistic account according to which childness consists in lack of autonomy. Both accounts share the idea that lack of autonomy plays a crucial role in determining who counts as a child. (Incidentally, this explains why some of the objections against Schapiro’s account—most notably, the objection from descriptive adequacy—mirror the objections raised against corresponding naturalistic accounts.) The difference, however, is that, according to Schapiro’s account, the fundamental childness-making property, i.e., the property that makes one a child, is a normative property, namely, the property of possessing a particular moral status, whereas according to the naturalistic account under consideration, the childness-making property is a natural property, namely, the property of lacking autonomy. Of course, according to Schapiro, children’s moral status is grounded in their lack of autonomy. Yet, the difference remains, for unlike the naturalistic account under consideration, Schapiro’s account construes childness as constitutively normative. Because of this, arguably, only Schapiro’s account can explain childness’ alleged inherent normativity.
  13. Dekker was finally able to begin her global circumnavigation shortly before her 15th birthday and eventually became the youngest person to complete this challenge. Anderson and Claassen (2012) discuss this case in detail.
  14. At one point in her text, Schapiro claims indeed that “many of the people that we conventionally call children [. . .] have adult status with respect to some domains of discretion, but not others” (1999: 734). As such, they have authority over the former domains, but not over the latter. This suggests that Schapiro’s account is a domain-relative account of childness. According to it, one and the same individual may be a child in some domains, but an adult in other domains. Thanks to an anonymous referee for drawing our attention to this point.
  15. Understood as a domain-relative account of childness, Schapiro’s account is also subject to an objection concerning its normative adequacy. As mentioned in the main text, in order to obtain permission to begin her journey, Dekker had to convince the court not just that she possessed the relevant sailing skills, but also that she was competent to make her own choices. If this is true, it is not entirely clear why it is correct to categorize Dekker as a child in domains such as education or health care, where she could also have exercised her decision-making competence. While ‘technical competence’ may vary in different domains and may thus greatly influence the outcome of an individual’s choices in these domains, ‘decision-making competence’ seems to refer to a relatively general capacity. So, the difference in Dekker’s differential moral status in the sailing and the health domains cannot be explained merely in terms of technical competence, e.g., by saying that she possessed robust sailing skills, but very limited medical knowledge. In addition, differences in knowledge or skills are typically considered irrelevant when adults are concerned, e.g., we do not make health care decisions on behalf of mature individuals simply because they have limited medical knowledge. This brings out the following problem. If Schapiro claims that Dekker was an adult in the sailing domain, but not in domains such as education and health care, then she must provide a justification for Dekker’s differential moral status, which appeals to Dekker’s differential degree of autonomy in these domains. As we have seen, however, this may be more difficult to do than what appears at first sight, for if Dekker is categorized as an adult in the sailing domain in virtue of her decision-making competence, then she should also be categorized as an adult in other domains where decision-making competence is the relevant ground of moral status.
  16. See also Schapiro (1999: 723).
  17. Of course, social facts may have normative significance. However, the idea is that, by themselves, they do not entail any normative facts.
  18. At least, they endorse the basic idea that there is nothing ‘universal’ or ‘natural’ about the category, and that “the biological facts of life, birth and infancy [do not, or do not fully,] explain the social facts of childhood” (James & Prout 1997: 14). See also James and James (2004), Kassem, Murphy and Taylor (2010).
  19. Notice, however, that this is contested by some feminists, who claim that children are as oppressed as women. See, e.g., Firestone (1970).
  20. Elizabeth Barnes (2016) makes a similar claim with respect to physical disability.
  21. Oppression is a fully normative concept. It names an injustice. As Ann Cudd convincingly writes: “to make a claim of oppression is to show that the harms involved are unjustified, or correlatively, to show that some harms are justified is to show that they are not oppressive” (Cudd 2006: 23). It may well be the case that children are currently oppressed, and have been oppressed across history, because the type and the scope of the social and legal power that is conferred to adults (especially to parents or guardians) over them is partly unjustified, and thus unjust.
  22. Child liberationists would certainly welcome the implication that we should get rid of the category. In this sense, a Haslangerian account of childness fits well with their position. Child liberationism, however, faces several important objections. See Archard (2015; 2016), for an overview.
  23. The extent of this commonality is subject to debate amongst different authors. For instance, in her critique of Ariès, Linda Pollock (1983) argues that care and protection are relatively stable parental attitudes across history. John Gillis (2011) agrees that in the early modern period “adults were no less concerned with the well-being of children” than contemporary adults, but adds that the specific forms that such concern took across history are variable.
  24. Underlying our account is the idea that whether an individual is the fitting holder of a differential moral status is not determined by whether or not the dominant ideology of a particular society regards the individual as such. Rather, it is a more objective matter, which depends on whether the individual ought, objectively, to be subject to the paternalistic treatments and measures that constitute that particular society’s moral regime of childhood. Notice that this position is perfectly consistent with our account. Indeed, relativism about whether one counts as a child does not entail relativism about whether one is the fitting holder of a particular moral status. (See also our remarks below, on p. 1073.)
  25. What counts as possessing ‘enough’ of the relevant properties? The answer is that this depends on what the dominant ideology in a given society considers as ‘enough’. It is important to emphasize that having ‘enough’ of the relevant natural properties may not be simply a numerical matter, i.e., a matter of counting how many of these properties an individual possesses. For certain properties may be regarded by the dominant ideology of a given society as more important than others for child status. These properties will thus have more weight for determining whether an individual has ‘enough’ of the relevant properties. This is to say that the notion of ‘enough’ relevant properties may not be exclusively descriptive, but it may be partly evaluative. This implies, for example, that if the dominant ideology in a given society considers age as significantly more important than other relevant properties, then it may be the case that an individual counts as a child in that society even if they do not possess most of the other relevant properties.
  26. We borrow the term ‘regime of childhood’ from Anderson and Claassen (2012). It must be kept in mind, however, that our characterization of the moral regime of childhood is different from theirs.
  27. For the distinction between ‘welfare paternalism’ and ‘moral paternalism’, see Dworkin (2005; 2020) and de Marneffe (2018) (whose understanding of moral paternalism we follow here).
  28. To illustrate, suppose that the dominant ideology of a given society regards subjecting individuals ranging from 0 to 18 years old to a specific set of paternalistic norms as justified in virtue of the fact that this maximizes overall social utility. These individuals count as children in that society. Still, it may be that social utility is maximized only if these individuals are subject to paternalistic norms in some domains (e.g., education), but not in others (e.g., labor). In this case, the dominant ideology will regard those individuals as fitting objects of paternalistic norms only in the former domains.
  29. See Corsaro (2018: 38–43).
  30. The basic moral standards on the basis of which the dominant ideology of a given society justifies the adoption of a given moral regime of childhood can also be assessed as correct or incorrect. Note that this opens the possibility that two societies may adopt the same moral regime of childhood, that this moral regime of childhood is correct (i.e., that the status-conferring properties are genuinely morally relevant and that the status-constituting norms are the norms that ought, objectively, to apply to the individuals with those properties), yet the moral regime of childhood is adopted for the wrong reasons in one or both of these societies, e.g., if it is justified by reference to incorrect moral standards.
  31. Suppose that, as a result of a classification that is mistaken by the society’s own standards, the individual is subject to paternalistic norms and treatments that they ought, objectively, to be subject to. This may still be morally problematic in some respects, for instance, if the mistaken classification is the result of unfair treatment or prejudice.
  32. It is also one that is adopted for the right reasons, i.e., that is justified on the basis of correct moral standards.
  33. We owe these examples to Colin Macleod.
  34. See also footnote 24.
  35. Thanks to Colin Macleod for inviting us to clarify this point.
  36. This objection was raised both by Vanessa Wills and by Aron Edidin.
  37. Recall from Section 4 that, according to our account, it is not sufficient to be perceived or imagined to be biologically immature, and subjected to systematic subordinating social roles on this basis, in order to count as a child in a given society. One must genuinely possess (enough of) the properties that are regarded by the dominant ideology of that society as grounding a relevant set of paternalistic treatments and measures.
  38. This objection was raised by an anonymous referee.

References

1 Alanen, Leena (2011). Generational Order. In Jens Qvortrup, William A. Corsaro, and Michael-Sebastian Honig (Eds.), The Palgrave Handbook of Childhood Studies (159–74). Palgrave Macmillan.

2 Anderson, Joel and Rutger Claassen (2012). Sailing Alone: Teenage Autonomy and Regimes of Childhood. Law & Philosophy, 31(5), 495–522.

3 Archard, David (2015). Children, Rights and Childhood (3rd ed.). Routledge.

4 Archard, David (2016). Children’s Rights. In Edward N. Zalta (Ed.), The Stanford Encyclopedia of Philosophy (Summer 2016 Edition). Retrieved from https://plato.stanford.edu/archives/sum2016/entries/rights-children/

5 Archard, David and Colin Macleod (Eds.) (2002). The Moral and Political Status of Children. Oxford University Press.

6 Ariès, Philippe (1962). Centuries of Childhood. Random House.

7 Barnes, Elizabeth (2016). The Minority Body: A Theory of Disability. Oxford University Press.

8 Berk, Laura E. (2017). Child Development (9th ed.). Pearson.

9 Brennan, Samantha (2014). The Goods of Childhood and Children’s Rights. In Françoise Baylis and Carolyn McLeod (Eds.), Family Making: Contemporary Ethical Challenges (29–48). Oxford University Press.

10 Brighouse, Harry and Adam Swift (2014). Family Values: The Ethics of Parent-Child Relationships. Princeton University Press.

11 Cohen, Howard (1980). Equal Rights for Children. Littlefield, Adams, and Co.

12 Cormier, Andrée-Anne and Mauro Rossi (2019). Is Children’s Well-Being Different from Adults’ Well-Being? Canadian Journal of Philosophy, 49(8), 1146–68.

13 Corsaro, William A. (2018). The Sociology of Childhood (5th ed.). Sage.

14 Cudd, Ann E. (2006). Analyzing Oppression. Oxford University Press.

15 de Marneffe, Peter (2018). Moralism and Moral Paternalism. In Kalle Grill and Jason Hanna (Eds.), The Routledge Handbook of the Philosophy of Paternalism (35–45). Routledge.

16 Doris, John M. (1998). Persons, Situations, and Virtue Ethics. Noûs, 32(4), 504–30.

17 Doris, John M. (2002). Lack of Character: Personality and Moral Behavior. Cambridge University Press.

18 Dworkin, Gerald (2005). Moral Paternalism. Law and Philosophy, 24(3), 305–19.

19 Dworkin, Gerald (2020). Paternalism. In Edward N. Zalta (Ed.), The Stanford Encyclopedia of Philosophy (Spring 2020 Edition). Retrieved from https://plato.stanford.edu/archives/spr2020/entries/paternalism/

20 Farson, Richard E. (1974). Birthrights. Collier Macmillan.

21 Firestone, Shulamith (1970). The Dialectic of Sex. Bantam Books.

22 Franklin-Hall, Andrew (2013). On Becoming an Adult: Autonomy and the Moral Relevance of Life’s Stages. The Philosophical Quarterly, 63(251), 223–47.

23 Gheaus, Anca (2015a). The ‘Intrinsic Goods of Childhood’ and the Just Society. In Alexander Bagattini and Colin Macleod (Eds.), The Nature of Children’s Well-Being (35–52). Springer.

24 Gheaus, Anca (2015b). Unfinished Adults and Defective Children: On the Nature and Value of Childhood. Journal of Ethics and Social Philosophy, 9(1), 1–21.

25 Gillis, John (2011). Transitions to Modernity. Jens Qvortrup, William A. Corsaro, and Michael-Sebastian Honig (Eds.), The Palgrave Handbook of Childhood Studies (114–26). Palgrave Macmillan.

28 Haidt, Jonathan (2001). The Emotional Dog and Its Rational Tail: A Social Intuitionist Approach to Moral Judgment. Psychological Review, 108(4), 814–34.

29 Hannan, Sarah (2018). Why Childhood Is Bad for Children. Journal of Applied Philosophy, 35(1), 11–28.

30 Haslanger, Sally (2000). Gender and Race: (What) Are They? (What) Do We Want Them to Be? Noûs, 34(1), 31–55.

31 Haslanger, Sally (2012). Resisting Reality: Social Construction and Social Critique. Oxford University Press.

32 Haslanger, Sally and Jennifer Saul (2006). Philosophical Analysis and Social Kinds. Aristotelian Society Supplementary Volume, 80(1), 89–118.

33 Hendrick, Harry (2011). The Evolution of Childhood in Western Europe c.1400–c.1750. In Jens Qvortrup, William A. Corsaro, and Michael-Sebastian Honig (Eds.), The Palgrave Handbook of Childhood Studies (99–113). Palgrave Macmillan.

34 Holt, John C. (1975). Escape from Childhood: The Needs and Rights of Children. Penguin.

35 Hopgood, Mei-Ling (2012). How Eskimos Keep Their Babies Warm. Algonquin Books.

36 James, Allison (2011). Agency. In Jens Qvortrup, William A. Corsaro, and Michael-Sebastian Honig (Eds.), The Palgrave Handbook of Childhood Studies (34–45). Palgrave Macmillan.

37 James, Allison and Adrian James (Eds.) (2004). Constructing Childhood: Theory, Policy, and Social Practice. Palgrave Macmillan.

38 James, Allison and Alan Prout (Eds.) (1997). Constructing and Reconstructing Childhood (2nd ed.). Routledge.

39 Jenks, Chris (1992). Constituting the Child. In Chris Jenks (Ed.), The Sociology of Childhood. Essential Readings (9–24). Gregg Revivals.

40 Kane Wildemann, Laura (2016). Childhood, Growth, and Dependency in Liberal Political Philosophy. Hypatia: A Journal of Feminist Philosophy, 31(1), 156–70.

41 Kassem, Derek, Lisa Murphy, and Elizabeth Taylor (Eds.) (2010). Key Issues in Childhood and Youth Studies. Routledge.

43 Matthews, Gareth and Amy Mullin (2015). The Philosophy of Childhood. In Edward N. Zalta (Ed.), The Stanford Encyclopedia of Philosophy (Spring 2015 Edition). Retrieved from https://plato.stanford.edu/archives/spr2015/entries/childhood/

44 Mayall, Berry (2002). Towards a Sociology of Childhood. Open University Press.

45 McMillan Cottom, Tressie (2019). Black Girlhood, Interrupted. In Thick: And Other Essays (171–94). The New Press.

46 Mills, Charles W. (1998). Blackness Visible: Essays on Philosophy and Race. Cornell University Press.

47 Pollock, Linda A. (1983). Forgotten Children: Parent-Child Relations from 1500 to 1900. Cambridge University Press.

48 Qvortrup, Jens (1987). Introduction: The Sociology of Childhood. International Journal of Sociology, 17(3), 3–37.

49 Qvortrup, Jens (2011). Childhood as a Structural Form. In Jens Qvortrup, William A. Corsaro, and Michael-Sebastian Honig (Eds.), The Palgrave Handbook of Childhood Studies (21–33). Palgrave Macmillan.

50 Schapiro, Tamar (1999). What Is a Child? Ethics, 109(4) 715–38.

51 Schapiro, Tamar (2003). Childhood and Personhood. Arizona Law Review, 45(575), 575–94.

52 Skelton, Anthony (2015). Utilitarianism, Welfare, Children. In Alexander Bagattini and Colin Macleod (Eds.), The Nature of Children’s Well-Being (85–103). Springer.

53 Tomlin, Patrick (2018). Saplings or Caterpillars? Trying to Understand Children’s Wellbeing. Journal of Applied Philosophy, 35(1), 29–46.

54 Weinstock, Daniel (2018). On the Complementarity of the Ages of Life. Journal of Applied Philosophy, 35(1), 47–59.

55 Woodhead, Martin (2011). Child Development and the Development of Childhood. In Jens Qvortrup, William A. Corsaro, and Michael-Sebastian Honig (Eds.), The Palgrave Handbook of Childhood Studies (46–61). Palgrave Macmillan.