A Journal of Philosophy, Applied to the Real World

Parents’ Rights, Children’s Religion: A Familial Relationship Goods Approach

University College London


The article presents a theory of the basis and nature of parents’ rights that appeals to the goods distinctively produced by intimate-but-authoritative relationships between adults and the children they parent. It explores the implications of that theory for questions about parents’ rights to raise their children as members of a religion, with particular attention to the issue of religious schooling. Even if not obstructing the development of their children’s capacity for autonomy, parents exceed the bounds of their legitimate authority in so far as they aim deliberately to influence their children’s religious views. Healthy familial relationships involve some identification of child with parent and require a sphere of spontaneous interaction between parent and child that are in any case likely to influence those views and constitute a standing threat to autonomy. Correcting over-deferential understandings of parents’ rights enables schools better to promote not only children’s autonomy but also other legitimate civic goals





When they are born, children have no religious views and they cannot identify as members of a religion. They become adherents of a religion—they come to engage in certain practices or endorse certain beliefs—just as they become atheists or agnostics: through a process of socialization and upbringing and education. Religious freedom is hugely important; people must be at liberty to act on their religious beliefs (or lack of them). But that leaves a question about parents’ rights to raise other people—their children—in ways that reflect their own beliefs.

Liberal political philosophy struggles with that question. Its core thought is that individuals should exercise their own judgment about how they are to live. One of the things that many choose to do with their lives, and one that many regard as among their most important life projects, is to raise children—and to exercise their own judgment about how to raise them. But parental rights are rights over others, others who have no realistic exit option and whose interest in making their own judgments about how they are to live their lives is no less important than that of the adults raising them.

It is difficult to know how to strike the right balance between the interests of parents and children, but that task is all the more challenging because children’s healthy development itself depends on their experiencing a very special kind of relationship with their parents. I will argue that, to become autonomous individuals, children need to grow up identifying with, and developing strong attachments towards, those who exercise authority over them. If that is right, then the kind of parent-child relationship that serves children’s interest in developing autonomy can itself pose a threat to that interest. Any plausible theory of parents’ rights must attend to the complex ways in which the agency of parents and children, and their wellbeing, are bound together.

Most political discussion of parents’ rights with respect to their children’s religious upbringing tends to focus on questions about schooling. It is taken for granted that parents must be free to raise their children as members of a religion at home. Indeed, that freedom is usually held to extend to homeschooling and to parents’ use of private schools. Only when public resources are involved—only in state or public schools—do questions about the regulation of religious schooling, and about the extent to which the state should cater to parents’ religious views, typically arise. Different countries take very different positions on that, all the way from the American insistence that public schools be religion-free zones to the Dutch view that any sufficiently large group of co-religionists is entitled to public funds to set up a school consonant with their religious beliefs.

My argument will indeed deliver conclusions about religious schooling—both where the state is involved and where it is not—but it will do so by digging deeper, querying background assumptions that unduly constrain our thinking and themselves require critical attention. Properly to address issues about the regulation of religious schools we must consider the general question of when the state may step in to limit parents’ authority over their children’s religious upbringing.

I start from a basic commitment to liberal equality. The “liberal” element means that it is important that people are free to form, critically reflect on, and act in accordance with their own judgments about how to live. (Cf. Rawls’ (1971) identification of a highest-order interest in our capacity to “frame, revise and pursue a conception of the good”.) They should be authors of their own lives, allowed and encouraged to develop the capacity for autonomy, without coercion or undue interference by others; it is part of the state’s function to protect and nurture that capacity. The “equality” element means that each person counts equally—nobody’s freedom is more important than anybody else’s and, from the state’s point of view, each person’s living the life of her choice matters equally.

This vision, originally developed to liberate citizens from an overbearing state founded on hierarchy, has been extended to the sphere formerly regarded as “private” with transformative implications for relations between men and women, but it has not yet had the same impact on relations between parents and children. Many children today are a bit like women were in the not-so-distant past, victims of an ideology that failed to recognize their claims to liberty and equality. We continue to grant parents extensive rights to control their children—including the right deliberately to inculcate the parents’ own views in ways that inhibit the development of children’s capacity to exercise or act on their own judgment about how they are to live their lives when they reach adulthood. And we fail to treat children as equal in importance with their parents: although we intervene to prevent abuse or neglect, we routinely allow children to be subordinated to their parents’ projects, to be treated as vehicles through which parents may seek to realize their own ideas about how to live. Advocates of school choice, for example, typically invoke the importance of parents’ getting to choose their children’s schools as if what mattered was the satisfaction of parents’ preferences rather than the interests of those subject to their choices.

Tackling questions about religious schooling forces us back to philosophical basics. To decide whether parents have the right to send their children to a religious school, or to one that teaches that all religion is a delusion, we need to know what rights they have with respect to their children’s upbringing in general. We need to think about why parents should get to exercise any authority in that—and indeed in any other—domain. Nobody thinks that it would be permissible for me to raise your child in my religion. Why should I get to do that to mine? Should I get to do that to mine? (Clayton 2006 and forthcoming).

My answer will appeal to the goods—“familial relationship goods”—that can be realized in parent-child relationships (Brighouse and Swift 2014). These are not goods like cars or washing machines; they are goods simply in being good for people—aspects of people’s lives that make those lives go well rather than badly. Only by understanding why it’s valuable that children should have parents at all can we think seriously about what rights parents should have to do things to, with and for their children. Only then can we assess claims about the extent of their right to influence their children’s emerging values, and only with that assessment in place can we work through the implications for religious schooling.

To develop the very capacities that liberals value, children need an intimate-but-authoritative relationship with particular adults. So those committed to liberal equality do not deny that parents should have the right to exercise considerable control over their children’s lives. Some discretionary authority—unmonitored by others, and to some extent unmonitored even by themselves—is essential if parents are to play their role in their children’s healthy development. As standardly framed, the philosophical challenge is to identify the proper limits on parents’ authority over their children; to strike the right balance between respecting parents’ freedom to live their lives according to their own beliefs, which for many include beliefs about how their children should be raised, and children’s interest in developing the capacity for autonomy. Focusing on familial relationship goods, and on the quality of the parent-child relationship, offers a distinctive way of approaching that challenge. In order to develop autonomy, and for many other reasons, children themselves need a close emotional bond which permits, and indeed requires, parents to be spontaneous and open in their relationships with their children. Making that kind of relationship central opens the door to more specific and difficult questions. How much, or what kinds of, authority must parents have for the relationship to work its magic? To what extent does the valuable intimacy between parent and child depend on parents’ being free to share themselves—including their religious views—with their children? How, if at all, might the kind of parent-child relationship that is in children’s interests depend on parents being free to influence the content of their children’s schooling?

The view to be presented invokes claims with which many religious parents will disagree, as will some atheists. Those who regard their children’s following the true path as more important than those children’s capacity to judge for themselves whether the path in question is indeed true will reject the autonomy condition. So will those who care more that their children see the falsity of all religious doctrines than that they come to see it for themselves. Both will deny my more specific claim that one of the duties that parents owe their children is overseeing the development of their capacity to decide for themselves how they want to live their lives. Building the facilitation of children’s autonomy into the parental role is, I believe, the right approach, and another way in which the familial relationship goods approach reconceives the more conventional opposition between parents’ and children’s interests. But it ups the stakes in the disagreement with the parent who simply denies that autonomy matters: on the view to be presented here, that parent has not only failed to recognize what she in fact owes her children, she has not understood what it is that grounds her claim to be a parent in the first place.

Of course, a parent’s freedom to raise her child as a member of a religion, and a child’s interest in developing the capacity for autonomy, are just two of many considerations relevant to the assessment of policy on religious schooling. In other, collaborative and interdisciplinary, work (Brighouse et al 2018), my colleagues and I have attempted to provide a framework for thinking about educational decisions quite generally. That framework emphasizes both the variety of different “educational goods” that schools should seek to produce and the range of other “independent values” that must be taken into account when forming judgments about policy all things considered. With regard to the former, the capacity for autonomy sits alongside the capacities for economic productivity, for democratic competence, for healthy personal relationships, for treating others as equals, and for personal fulfilment; these are all desirable educational outputs the production of which can pull decisions in different directions. It matters both what mix of those we should be aiming to achieve for any given individual and how they should be distributed among the population. As for the latter, parents’ interests rank with childhood goods, respect for democratic processes, and freedom of residence and occupation, as considerations that must be weighed in the balance when deciding how schools, and indeed other institutions with educational significance (such as the mass media), should be regulated. Given this multiplicity of relevant factors, one might wonder whether parents’ rights warrant the extensive attention they will receive here.

To see why they do, we need to distinguish between parents’ interests and parents’ rights. While this more general framework presents parents’ interests as just one consideration to be weighed in the consequentialist balance alongside others, claims about parents’ rights typically assert something stronger. The thought is precisely that, properly identified, parents’ rights—including their right to raise their children as members of a religion—have a normative status such that they should be respected even where doing so may be inimical to the optimal production and/or distribution of educational and other goods. They are non-consequentialist considerations that morally constrain the pursuit of good outcomes. If parents really do have the moral right to send their children to a religious school, then that is something that states must respect. Not to do so would be to treat those parents unjustly.

Correcting mistaken views about parents’ rights is a crucial part of the overall picture because excessive deference to parents drastically constrains efforts to achieve a better balance of educational goods, and to distribute those goods more fairly. Misjudging the proper extent of parents’ rights to control their children’s schooling affects all children, not only those having the control exercised over them. And it has the potential to deprive children of more than the capacity of autonomy. That is partly because many of the outcomes we rely on schools to achieve depend—in part—on the kinds of children who go to them (Clayton et al 2019). If parents are allowed to opt for segregated schools, then all children, not just their own, miss out on the educational benefits of mixed school composition. This applies most obviously to the other-regarding or “civic” goals of education, such as the capacity to treat others as moral equals. If parents have the right to protect their children from knowledge of other religions, or to deny them the kind of school environment that is most likely to foster tolerant attitudes, then we may all be condemned to share a society, and a democracy, with intolerant others who know little about their fellow citizens. (For the role that schools can play in promoting intergroup contact, see Wölfer et al 2018.) And since spending time with others from different home cultures promotes critical reflection on one’s own, and provides a sense of the different ways of living one’s life, mixed schools are also conducive to autonomy. If members of a religious faith school their children in particular ways that exclude, or are simply not attractive to, others, all children are deprived of the educational benefits they would get from being schooled alongside such children.

Correctly identifying the scope and content of parents’ rights with respect to children’s religious schooling is important, then, not only for the sake of their children’s autonomy—though that would be reason enough—but because it affects the education of other people’s children, and indeed the whole character of the society. By offering an alternative understanding of the nature and basis of parents’ rights, my aim is primarily to discredit the prevalent view, thereby removing—at least at the level of theory—an obstacle to education policies that can achieve better, and fairer, outcomes for all our children.

Our society currently gets parents’ rights not just marginally but massively wrong. We have, of course, moved well beyond the classical Roman view that a child was the father’s property and could be killed at his behest, but much of our thinking remains “proprietarian”. We accord parents rights over their children as if children are quasi-property—as if they belong to their parents in some sense. In many countries, parents who wish to raise their children in ways inimical to their developing the capacity for autonomy are free to “school” them at home, thereby avoiding almost all regulation, or to send them to private schools which, though regulated to some extent, can in practice reinforce the home culture so thoroughly that it becomes very hard for children to gain critical distance. But that is only the far end of the spectrum. Many readers who share the liberal commitment to autonomy and object to that kind of upbringing nonetheless see no problem with parents’ subjecting their children to the parents’ own views—religious or otherwise—in ways that exceed the proper limits of the parental role, as I understand it. Parents’ rights need more than incremental revision; they have to be fundamentally reconceived.

The article proceeds in three sections. Section one presents the familial relationship goods account of parents’ rights. Section two explores the implications of that account for questions about the right to raise one’s child as a member of one’s religion. Section three focuses on the issue of religious schooling.

Familial Relationship Goods and Parents’ Rights

To understand what rights parents should have over the children they parent, we have to start with what may seem like an odd question: why should children be raised by parents at all? The biological sense of “parent”—parent as procreator—is so common that it is natural to hear that question as asking why biological parents should get to raise the children they have produced. That question is indeed unsettling, but I mean to address a more fundamental issue. Properly to address the question of parents’ rights we need to take a step further back and consider the reasons for wanting children to be raised in families, by parents—whether biological or adoptive—in the first place. Children’s upbringing could be handed over to state-run quasi-orphanages. Doubtless readers will baulk at that suggestion, but other forms of collective or communal child-raising, such as the kibbutz, are less dystopian. What, if anything, would be lost if we were to get rid of the family and go with these alternatives? Only with a clear view of what is so valuable about the family, and the parent-child relationship in particular, can we think systematically about what rights parents need to have for that value to be realized.

Parent-child relationships make possible distinctive and weighty goods in people’s lives. I will argue that a very particular kind of relationship with one or more (but not too many) adults is valuable for children: it is essential both for their emotional, cognitive and moral development and it contributes to their wellbeing during childhood itself. Since children are—at least initially—dependent, vulnerable and involuntary participants in the process of being raised, and since childhood experiences have formative influence on our lives as a whole, children’s interests matter most. So the case for the family depends primarily on parent-child relationships being good for children and the adults they become. But the wellbeing of adults is also affected in a different way by how children are raised, and many have an interest in getting to play the role of parent. In the world of state-run quasi-orphanages, children would lack the kind of relationships with particular adults that they need both to flourish as children and to develop into flourishing adults, while adults would be deprived of the special—distinctive and weighty—goods that many achieve through parenting a child.

Let me start with children. Children’s interests—the things that make their lives go well—can be categorized in various ways. They have current interests, which contribute to their wellbeing during their childhoods, but they also have developmental interests—interests in developing the physical, cognitive, emotional and moral capacities that will enable their lives to go well as adults. Raising children well is a matter of getting something like the right balance between these various considerations. Some of the things that children need could indeed be provided by impersonal state functionaries: healthy nutrition, adequate clothing and protection from physical danger presumably fall into that category. But one does not have to be an expert in psychology or child development to know that healthy emotional development depends on children forming deep attachments to particular adults who are emotionally attuned to them, whom they experience as loving them, as having a special duty of care towards them, and with whom they can enjoy long-term intimate relationships. More interestingly, perhaps, children also need to experience their attentive carers as their central disciplinary models. To learn and internalize self-control, empathy, deferred gratification, and other modes of self-regulation, the child needs to see these traits modeled in people with whom she identifies; for children, identification comes through love and admiration, which are themselves responses to loving warmth in the carer (Gerhardt 2004). Healthy emotional development, in other words, depends on children experiencing intimate-yet-authoritative relationships with particular adults. That is the core of the child-centred case for parents and the family as the best way of raising children.

It is important, further, that those adults with authority over the child are experienced by her as acting at least somewhat spontaneously, as expressing their own individuality and sharing themselves with their children, and as having the discretion to act on their own judgments. Someone who, when deciding what to cook for supper, or what stories to read at bedtime, robotically executes the detailed instructions contained in an official state-approved child-raising manual will hardly be providing the kind of emotional responsiveness that tends to induce loving identification with the authority figure, nor herself experiencing the parent-child relationship as a source of joy and satisfaction in the way most helpful to the child’s emotional development. Some degree of external direction will doubtless not spoil the relationship. Parents’ administering medicines to their children according to doctors’ instructions are likely to be experienced as acting on an entirely healthy loving motivation. And, for some parents, following a more extensive “instruction manual”—such as a religious text—may be precisely how they manifest and share their own sense of what matters with their children. In such cases, however, it is significant that the parent is acting on her own judgment rather than experiencing the manual as an external imposition.

But the importance of that kind of relationship goes far beyond the child’s interest in emotional wellbeing. Emotional, cognitive and moral development are so intertwined that parent-child relationships play a crucial role in these other dimensions too. According to developmental psychologists, even basic mental processes like representational thinking, which is the precursor for symbolization and conceptual thought, and the capacity to imagine, which allows one to take the other’s point of view, depend on internalization of the caregiving relationship. As Anne C. Dailey (2006, p.130) puts it: “the capacity for reasoned thinking represents a developmental line, or maturational sequence, beginning in the earliest physical interactions with an emotionally responsive caregiver and ending in a mature complex capacity to lead an independent, autonomous, self-directed life”. The capacity for autonomy is complex partly in that it combines cognitive, emotional and moral aspects: the autonomous person can not only reflect on the options available to her, process information, and identify means to her ends; she can also trust and cooperate with others, defer gratification, and contain disruptive and destructive feelings. That complex capacity is best fostered when children are raised in intimate-yet-authoritative relationships with particular adults; i.e. when they are raised in families, by parents.

What about adults? What ways of raising children would be best for them? That might seem like a trick question. After all, all adults started out as children, and the last two paragraphs were mainly about the kind of upbringing that would be good for children not in the sense of being good for them during their childhoods but in the developmental sense of being good for them as the adults they will become. One response to the question, then, not falling for the trick, would be: “Adults are just developed children; they are the same people! So the way of raising children that is best for adults must be the same as the way that best serves children’s developmental interests.” Which is what we have just been talking about.

That way of formulating the question is salutary. In highlighting the fundamental continuity between children and adults it reminds us that the issue of how children should be raised just is the question of how people should be raised and makes it less likely that we will misjudge the balance of interests between children and the adults who raise them. But that doesn’t make it a trick question; there is indeed a residual issue about how different ways of organizing children’s upbringing might be good or bad for people-as-adults. The kind of developmental interests appealed to in the previous paragraphs were quite general—and their very generality helps to explain their importance. But there remain questions about the specific ways in which childrearing arrangements might contribute to the well-being of adults, and whether that contribution might affect the overall judgment about how childrearing should be arranged. Suppose my story about why children need parents turns out to be false: robots can do the job just as well—maybe even a bit better—and it is proposed that we hand the task over to them. Parenting would be banned. Even though robot-raised people would suffer no other loss—and even if they enjoyed a gain in other ways—many would resist the proposal. Some might do so selfishly—just because it suited them to get to be a parent and the cost would be borne by others—but that need not be part of the argument. Looking at the issue impartially, one could reasonably think that parenting a child makes enough of a contribution to enough people’s lives that our collectively retaining that option would be worth missing out on whatever other benefits were produced by the robot scenario.

Nearly everybody agrees that the adult interest in parenting is weighty enough to ground a right to be a parent, that we would be failing in our duties to one another if we denied people the opportunity to engage in that activity. And certainly many people make great efforts to become parents, regard raising a child as one of their most important life projects, and sacrifice lots of other valued opportunities in the process. But adults become parents for many different reasons and there is considerable dispute about how best to understand the value of parenting and the basis of the right to be a parent (Overall 2012). Some point to the significance of raising children as a way of extending or continuing oneself into the future, passing on some aspects of oneself to a child; those who take this kind of view can disagree about the specific nature of the extension or continuity that is valuable — perhaps it is one’s genes, perhaps one’s property, perhaps one’s religious identity (Reshef 2013). Others see raising a child primarily as a creative activity, or an opportunity for self-expression, a bit like an artist shaping a piece of marble into a sculpture (Page 1984, Macleod 2002). These different grounds for holding that there is a right to parent (i.e. the right to become a parent) will suggest different views about the rights of parents (i.e. the rights that parents have over their children)—including views about parents’ rights to influence their children’s values and beliefs.

The common problem with such accounts is that they make children means to their parents’ ends; the child is regarded, and treated, as a vehicle through which parents can achieve their own purposes. True, the child may not only be a means. If the interest in developing autonomy is satisfied, then a parent may plausibly claim that she has treated her child also as an end. Nonetheless, there is something inappropriately self-serving about this kind of attempt to justify the claim to parent a child. Of course, what we are looking for is precisely a way in which parenting contributes to the well-being of the adults doing it, so some element of adult self-interest is unavoidable. But the interest in question must be distinctive and weighty enough to provide a plausible answer to the question of why adults should get to parent children even if robots would do a better job. There are many ways in which people can extend or continue themselves into the future, or express themselves creatively, without claiming the right to control another human being. And there are different ways in which they can enjoy fulfilling and intimate relationships with others, so blanket appeals to the value of intimate affective relationships (such as that offered by Schoemann 1980) cannot succeed either. In my view, what is special about parenting, and important enough to count for something in the balance when weighed against our other interests, is the parents’ role as their children’s fiduciary. The parent has a special duty to protect and promote the child’s interests including the interest most children have in developing the capacity for autonomy and becoming someone who has no need of a parent’s special duty of care. The idea that parents have fiduciary duties toward their children is familiar from Locke (Locke 1689/1988). The additional claim here is that adults have a non-fiduciary interest in being able to play a fiduciary role; it is valuable for their children that they play it well, but it is also a distinctive source of their own flourishing that they play it.

Some elements in what is special about being a fiduciary for a child concern the fact that what we’re talking about here is a child: relevant here are the distinctive properties and moral standing of the person for whom one is acting as fiduciary: her possessing the capacity to develop into an autonomous adult, her degree of vulnerability to one’s responses and judgments, her involuntary dependence on one, her natural tendency to develop a deep attachment to one. Failing adequately to discharge your fiduciary duties to a child would be different from failing to discharge those owed to a client or patient, or even to an ageing parent, even if what was involved in fulfilling the duties were the same. But of course they are not the same. Other elements concern what it is that children need from their fiduciaries. As we have seen, they need a special kind of relationship—a relationship in which the adult offers love and authority, a complex and emotionally challenging combination of openness and restraint, of spontaneity and self-monitoring, of sharing and withholding. It’s that kind of relationship that many adults have an interest in too.

Imagine a world in which human children didn’t need much more looking after than guinea pigs, or Tamagotchi toys. Imagine that they could fully develop into autonomous, emotionally adjusted adults, and enjoy the intrinsic goods of childhood, with that kind and level of input from adults. Even in that hypothetical world, there would be some value to being the person responsible for ensuring that children’s interests were met. But what’s really valuable in the case of parenting is not being the fiduciary per se but having the kind of relationship that is, in fact, the kind that children need to develop into healthy adults. It’s that kind of relationship which presents a distinctive challenge, and distinctive sources of fulfilment, which together give adults unique opportunities for flourishing.

Parents’ rights, on this account, are precisely the rights that parents need in order to have the kind of relationship that justifies children being raised by parents—rather than robots, state functionaries or interchangeable members of a commune—in the first place. We can assess on a case by case basis whether an appeal to familial relationship goods justifies parents’ claims to control—to exercise authority over—children by looking at the role those rights play in realizing familial relationship goods. Roughly, familial relationship goods give us strong reason to grant parents the rights they require to fulfil the fiduciary role, to create or sustain the kind of intimate-and-authoritative relationship that children need and that is also valuable to the parent. The relationship goods approach helps us work out what room is necessary for the “free and flourishing internal life appropriate” to the family (Rawls 2001:165). Parents have the right to engage in those activities and interactions with their children that facilitate the realization of the extremely valuable goods that justify the family in the first place.

Parental claims to the right to do things to, with and for their children that cannot be justified by appeal to those goods may perhaps be defended in other ways; parents are not only parents and they may have other interests or prerogatives that can properly influence their dealings with their children. But those other considerations will not ground parents’ rights as that category is understood here; they will not be rights that they have in virtue of being parents. As we will see, the question of whether such non- or extra-parental considerations mean that parents have the right to act on their religious views when deciding matters for their children, and in what ways, will be important when thinking about religious schooling.

This account of the nature and basis of parents’ rights is deliberately quite general. The next section will consider its implications for the issue of parents’ rights to shape their children’s emerging beliefs and values, and of religious beliefs and values in particular. Before moving on, however, it may be helpful to highlight a few features of the general approach proposed so far. In my view these are attractive, but readers may suspect some sleight of hand in the way the various elements of the picture fit together, so it is better to have them out in the open.

First, the thrust of the approach is to start from children’s interests and identify what it is that children need from adults. That starting point does much to generate the conclusion that parents’ rights are just those rights that it is in children’s interests for parents to have, and some may regard it as biased or prejudicial. Certainly it has profound and transformative implications for the way we think about familial relationships. But children’s vulnerability, their involuntary subject to control by another person, and the far-reaching implications of how they are raised for their lives as a whole suggest that any approach that gives greater weight to the interests of parents is indefensible. As Shelley Burtt (2002, p.17) puts it: “authority over other human beings should extend only so far as making up the deficits that legitimate their subordination… the way we think of children and their needs determines the sort of authority we think it is appropriate to exercise over them”.

Second, the claim that the special value of parenting, for adults, consists in playing the fiduciary role—providing the kind of relationship that children need—is certainly controversial. Its effect is to change the picture of the relationship between parents and children in a way that massively reduces the scope for genuine conflicts of interest—and of rights—between the two. That may seem too good to be true. Of course, adults may want to be parents for a variety of reasons, and some may regard the proposed view about the distinctive value of parenthood as misidentifying what is at stake. Those who see their mission as populating the world with followers of a particular religious faith, or who view their children as means for them to express and realise their powers of creative self-expression, will think I neglect their true interests qua parents, and will leave greater scope for genuine conflicts between their own and their children’s interests. But such alternative specifications of the value of parenting do not provide plausible answers to the question of why people should have the right to parent children even when others—whether robots, state functionaries or simply other parents—would do a better job.

Third, the claim that parents’ rights are derivative of children’s interests means, in effect, that the adult interest in being a parent—the basis of the right to parent—plays no further role in the argument. We will need to consider how their other interests may properly influence parents’ relationships with their children, but from here on the adult interest in parenting drops out of the picture. My defence of a particular specification of that interest is important to the position as a whole, of course, because it challenges competing views that grant parents more extensive rights to shape their children’s values. But the effect of that defence was precisely to direct attention to what children need from their parents, and what rights parents must have in order to give them what they need. That is where our attention will indeed be directed from here on.

Fourth, autonomy plays a central role in the account. It is important that children develop the capacity for autonomy and parent-child relationships with a particular character are important for its development. As their fiduciaries, parents are charged with the task of facilitating the process whereby their children become capable of making independent and reflective decisions about how they are to live their lives. That process begins with a mixture of affective connection and clear and consistent regulation—the “downloading” from the parent of the capacity to contain and regulate her own desires and emotions—but as children begin to develop their own preferences and perspectives, it also requires parents to exercise the self-restraint that allows their children to begin to trust and act on their own judgments. That is challenging even for those parents who accept the importance of autonomy—it is hard to resist the temptation to use one’s authoritative position paternalistically, to attempt to shape one’s children’s beliefs, values and choices according to one’s own judgments about what it is to live well. But many reject that element of the fiduciary picture and conceive their role precisely as deploying their parental authority to guide their children towards (their own views about) how best to live. Here, as I have said, we hit bedrock disagreement: those who deny the importance of autonomy will reject all liberal views, including standard approaches that treat the facilitation of children’s autonomy as an external constraint on parents. For them, my account—on which facilitating autonomy is internal to the parental role and helps to explain why there is a right to parent, even when others would do a better job—will only make matters worse.

Parents’ Rights, Children’s Religion

We can now explore what this view about the basis and nature of parents’ rights means for their right to raise their children as members of a particular religion. Where other theorists see those rights as part of parents’ own expressive liberty (Galston 2002, p.102) or as an implication of their own religious freedom as individuals (Fried 1976, p.152), or as deriving from their own interest in a relationship with their children based on shared identity (Reshef 2013, p.132) the approach proposed here frames the issue very differently. Suppose that an intimate-but-authoritative relationship is indeed important for children’s development, including the development of their autonomy, and that parents’ rights should be derived entirely from the fiduciary aspect of the relationship. What are the implications for rights to control, or even to influence, the development of their children’s religious beliefs and identity?

Before addressing that question, we should be clear that, on the liberal picture, there is a big difference between people’s views about how they should live their own lives and their views about how others should be treated. The values that children need to acquire as part of their moral development—the liberal virtues such as tolerance, respect for others, and what Rawls (1962) calls “a sense of justice”—have a different status from the kind that individuals may choose to endorse as a matter of private conscience, such as those attaching to full-blown religious systems or ethical doctrines. To put it crudely, it is much more important that people get to form and act on their own views about the former than about the latter: the fact that we owe duties to others—that we are morally required to treat them in certain ways—means that people’s views about how to treat others do not demand the same kind of respect as their judgments about how to live their own lives more generally. This gives parents a very different role in the formation of their children’s moral views from that which is appropriate in the case of their religious beliefs. I said earlier that the parent-child relationship itself is justified partly because of the role that parents play in children’s moral development; the important point here is that some deliberate shaping of their children’s emerging values is itself part of that job. There will be other influences, of course, but instilling in children the virtue of honesty, the ability to distinguish right from wrong, and the sense that others are moral equals irrespective of their skin colour, religion, or gender, is a task primarily charged to parents and part of their fiduciary duty to their children (as well as being in the interest of third parties).

Returning to the particular issue of religion, parents’ role is to serve as loving authorities—to exercise the kind of discretion and induce the kind of identification that children need for their own development, including their moral development—while helping their children acquire the capacity to judge religious matters for themselves. That is a challenging job description, especially for parents with deep religious convictions (including atheism). To develop the kind of attachment that children need, they must gradually get to know their parents—who they are, what they care about—and parents must be free to be spontaneous, and to share themselves with their children. To conceal their religious views, or not to allow those views in any way to inform their exercise of parental authority, would require a kind of withholding, and a degree of self-monitoring, that is inimical to an intimate loving relationship. Parents whose religious convictions require them to say prayers before eating, for example, or prescribe and proscribe particular kinds of food, must have some discretion to act on those beliefs in the way they conduct family life. Apart from anything else, loving parents are naturally motivated to benefit their children, and religious views affect what is regarded as “benefit”. A parent who believes that her child will be condemned to eternal damnation unless she comes to endorse a particular doctrine cannot entirely bracket that belief in her relationship with the child without depriving the child of at least some of those very expressions of parental love that the child needs. But precisely because, when all goes well, children love and identify with their parents—want to please them, want to be like them—that same relationship inevitably threatens the autonomy that it is also the parent’s task to develop. So not only must children be exposed to other beliefs at appropriate ages, and in such a way that other ways of life become genuine options for them, but also those very processes of attachment and identification with the parent that are needed on developmental grounds have to be carefully managed so that those alternatives are not unthinkable, or adoptable only at excessive emotional cost.

On the proposed account of parents’ rights, then, parents must have the right to act in ways that will tend to influence their children’s religious views. In a healthy parent-child relationship, a parent’s religious views are bound to shape those of her children. Some influence will arise simply as a result of parents being themselves in their relationships with their children, and exercising in a more or less unreflective and personal way their sphere of discretion over the particular ways in which they interact with them. Some will arise from parents’ acting on their natural motivation to help their children’s lives go better. A loving parent who thinks that her child will benefit if she loves God is bound to find herself nudging her in that direction simply because of her automatic and natural tendency to relate to her child in ways that she thinks will be good for her. The same applies to the loving parent who decries belief in God as the opiate of the masses, or regards it as an irrational projection of human psychological needs. Parents will naturally tend to denigrate, and unthinkingly steer their children away from, what they take to be bad influences. The idea that parents should constantly monitor themselves in order to screen out anything that might influence their children’s views about religion would risk distancing them, creating artifice in the relationship, and depriving their children of the possibility of the warm, spontaneous, genuine relationship that they need. Most of us cannot simultaneously shield our children from those values and commitments that are central to our identities and spontaneously share ourselves with them in the way that the healthy parent-child relationship demands.

I have emphasized the significance of unmonitored discretion and spontaneity, but what does the account imply for parents’ right deliberately to act in ways likely to influence to their children’s religious beliefs? To answer that we must distinguish two things that might be going on under that description. On the one hand, a parent might be deliberately directing her child towards a particular faith—or towards a rejection of any—in the sense that she intends that the child come to endorse her own view. That, of course, is a right that parents conventionally claim—and that they are everywhere granted. On the other hand, the parent may simply be trying to give her child the right kind of relationship. Even if she does not herself consciously frame things in such terms, her aim, in this second case, is to provide her child with the “familial relationship goods” that, if I am right, explain why she has any rights at all over the child. Here the influencing of her children’s religious views, though deliberate, is ultimately motivated by a concern that her relationship with the child should go well.

Parents who see themselves as justified in deliberately guiding their children towards their own religious views, in the first sense, have misunderstood their role and the moral character of the parent-child relationship. If the relationship goods approach is right, children are of course helping parents to realize familial relationship goods in their lives, and their parents may have chosen to be parents for that very reason. Still, that approach gives parents no permission to treat their children as means by which they may permissibly seek to realize their own values in other ways, or to pursue their own, controversial, conception of how one should live. Asked “Why should adults get to exercise authority over children?”, we are not tempted to answer: “So that they can direct children towards their preferred religious doctrine”. Typically, then, when parents deliberately direct their children on religious matters, they are deploying their power improperly. This is certainly true of those parents who try to direct their children’s religious views without regard to the development of their children’s autonomy—i.e. their capacity to judge such matters for themselves. But it can be true also of parents who do take their children’s moral separateness seriously in that way. Recall Burtt’s (2002:17) nice articulation of the principle at stake: “authority over other human beings should extend only so far as making up the deficits that legitimate their subordination”. Even where their concern to guide their children towards the true path is motivated entirely by a loving concern for their children’s wellbeing, it is not their proper role, as parents, to exercise parental authority in that way. This is consistent with the view that parents may be distinctively well placed to discern their children’s particular developing talents and emerging interests, and on that basis may legitimately exercise their authority in ways conducive to their wellbeing (Richards 2016).

But parents may deliberately introduce their children to their religious views, or their views about religion, in ways that can be understood as part of the sharing with, or revealing to, the child that is itself conducive to the kind of relationship I have described. Here there need be no intention that the child should come to endorse the views in question. I have emphasized the extent to which parents’ revealing and sharing their religious commitments would naturally result from the spontaneous, unmonitored, quality of the relationship, but a parent may be right to think that the relationship will go better if she also acts in a considered way to show her child who she is and what she cares about. When a Christian parent takes his daughter to church, that is not usually an unthinking and automatic sharing of self between parent and child. It is more likely a deliberate decision to introduce the child to a world of belief and practice that the parent judges valuable. To be sure, that introduction is typically motivated by the desire that the child will come to share that judgment, but it need not be. The same applies to deliberate decisions to say prayers at certain times, such as before meals. The parent might require her child to engage in such practices as a way of introducing the child to the parent’s religious views, believing that her relationship with the child will be closer—the child will know her parent better—if the child experiences those practices for herself.

This justification extends to deliberate and considered parental exercises of authority the same claim that underpinned the case for spontaneous, unmonitored interactions between parent and child. The thought is that, for the relationship properly to serve children, there needs to be a kind of emotional bond and mutual identification that is incompatible with the requirement that parents withhold and conceal their own views on religious matters. To confine parents to the spontaneous expression of those views—to deny them the freedom deliberately to reveal their views to their children, including by means that involve controlling their children’s behavior, such as taking them to church—would be to do children a disservice.

Perhaps paradoxically, the possibility of children’s rejecting their parents’ religious commitments is another reason why exposure—including deliberate exposure—to those commitments can be in children’s interests. Everything in the last few paragraphs has assumed that, whatever else they are doing to and with their children, parents are not obstructing the development of their children’s autonomy. Autonomous individuals can make up their own minds about what to believe, and they may well end up believing different things from their parents. If we think not only about the developmental benefits of parent-child relationships but also about their value when people reach adulthood, it seems that, where the child does break from the parents’ values, the parent-child relationship will probably be sustained in a more meaningful way, and has a better chance of being sustained, if child and parent are in a position at least to appreciate the other’s point of view, to understand where the other is coming from. (For views that emphasize the value of familial relationships’ continuing into adulthood see Reshef (2013) and, especially, Ferracioli (2015).) In the child’s case, that can only happen if the parent has indeed made sure that the child has a real appreciation of how she lives her life and how she sees the world.

This view about the permissibility of parents’ deliberately introducing their children to their religious views when it is important for their relationship raises a number of complexities. One concerns how much of an “introduction” is really needed for the child to relate to her parent in the way that the account requires. One visit to a Church, temple, mosque or synagogue is hardly going to do the job, but it is implausible to regard this justification as permitting parents to require weekly visits for many years, or to demand that their children acquire a level of familiarity with the doctrine in the way that might justify requiring a high level of religious instruction. The thought is that children should have a sense of who their parents are, and what matters to them, not that they should develop an advanced understanding of the views to which their parents subscribe. The same applies to practices like praying. Even if it is permissible for parents to introduce their children to such a practice, that will not justify its becoming part of their daily routine, or at least not for long.

The example of praying raises a distinctive concern about the permissibility of requiring children themselves to engage in particular practices rather than merely observing their parents, and other adults, do so. One can imagine a parent maintaining that it is only by, say, actually praying to a divine being that one understands what it means to pray to that divine being, or to any divine being, and that without that experience a child will not really have been “introduced” to those things that matter to the parent. To lessen the oddity of that claim, consider that a parent who worships Beethoven is presumably permitted to take her child to a concert, and make her listen to the music, not just to observe others doing so. Of course, praying to a divine being seems to presuppose beliefs—such as that the divine being exists—of a kind that some regard as distinctively objectionable but the familial relationship goods account nonetheless has scope for that level and kind of deliberate introduction.

Third, a parent who exercises parental authority in a way that is likely to influence her child’s religious views, albeit without any intention of doing so and motivated only by relationship considerations, may nonetheless hope that the exposure will result in the child’s coming to endorse the parent’s own views. It is demanding enough to expect a parent committed to particular religious views to respect her child’s moral separateness not only by facilitating the development of her autonomy but also by abjuring any action intended to guide her towards those views. It would be psychologically impossible for such a parent not even to wish that the child will autonomously come to share them. Indeed, a parent who did not even want her child to choose what, ex hypothesi, she regards as the best way to live would surely be failing to provide the kind of loving relationship the child needs.

We will soon move on from this general discussion of parents’ rights to raise their children as members of a religion to the issue of schooling in particular. To lay the ground, a few more features of the proposed approach are worth bringing out. Most important is the point that parents constitute a standing threat to their children’s independence, understood as their capacity to choose their own lives. Close, intimate-but-authoritative, relationships between parents and children are vital, but such relationships can easily be too close. Even those influences that arise spontaneously in a loving parent-child relationship can threaten autonomy; that threat is all the greater where influence arises as a result of the parent’s deliberate decision to share herself with her child. In all cases, parents mindful of their duties to their children will take care not to engage in the kinds of revealing and sharing that will impede the development of the child’s capacity to make her own judgments about whatever is being shared and revealed. What kinds will in fact do that partly depends, of course, on the child’s age or stage of development. For young children, the emphasis can be on fostering the processes of identification and attachment. As they mature and become able to question their parents’ views, the balance needs to shift toward facilitating the process of separation and individuation.

So far I have talked about “religious views” in the abstract, paying no attention to their content or the differences between them. But different religious doctrines can have very different implications for children’s developing autonomy, and the extent to which parents may share their religious views with their children will vary accordingly. In a healthy parent-child relationship, children are naturally inclined to identify with their parents, and to seek their parents’ approval. So it makes a difference what it is that parents do and do not approve. Other things equal, children whose parents reveal their belief that all who do not subscribe to their own religion are wicked, or condemned to eternal damnation, are likely to find it harder to break with their parents’ religious views than those whose parents hold more moderate views. Given the variety in people’s constitutions and characters, children who know that their parents regard homosexuality as wicked, or girls who learn that their parents see motherhood and homemaking as divinely ordained for women, may find it harder to live a life that is right for them than those whose parents take a more tolerant line. On the other hand, where a religious doctrine itself puts great weight on the importance of people critically judging its validity for themselves and living according to their consciences, or simply regards what is at stake in such judgments as less weighty, there will be fewer obstacles to children’s developing their own views on religious matters; it may even be positively encouraged.

My view about parental rights, then, is not “neutral” between different religious views. Indeed, those who reject the claim that parents are under a fiduciary duty to facilitate their children’s autonomy will already have found that conception of the parental role to be biased against religious doctrines that deny the importance of autonomy. So it is no surprise that the view permits parents whose religious views themselves endorse liberal values to share their views with their children in ways that may be impermissible for those who do not. This is not an embarrassment—but it does point to a paradox. On my account, it is parents whose religious convictions do not put much, or any, weight on autonomy who must be most careful about sharing those convictions with children—for the sake of their children’s autonomy. While the argument is indeed unlikely to have much motivational traction with such parents—since they will simply reject the initial account of their parental duties—that is no objection. Nor is the fact that my view yields differential implications for believers in different religions. Both simply show how much rests on the validity of that account.

But what about parents whose religious convictions are such that it is simply psychologically impossible for them to have the kind of relationship their children need without exceeding the constraints on the exercise of parental authority that I have argued for? Consider someone who is convinced that her child will suffer eternal damnation, or merely that he will live a worthless life, unless he comes to endorse her own doctrine. Her spontaneous sharing and revealing of herself is bound to exert a level of emotional pressure that is inimical to the child’s enjoying a genuinely “open future”. And there may be no way for her to refrain from deliberately guiding him towards a doctrine the endorsement of which will have (she believes) such a huge impact on his wellbeing: there may be no gap, for her, between acting in ways that the child needs her to act if he is to feel loved, on the one hand, and at least nudging him towards a flourishing life, on the other. Here, it seems, there is no morally costless resolution to the conflict. Such a parent is unable to provide her child with the kind of loving relationship that the child needs without simultaneously threatening her autonomy and exceeding the proper limits on her parental authority in the process.

To be sure, in such cases we might excuse the parent’s wrongful treatment of her children. To find oneself unable to give one’s child the love he needs without failing fully to acknowledge his independent moral status is very different from simply treating him as a means to a greater good or as a vehicle for the pursuit of one’s own preferences. To see this, and to differentiate between different kinds of failure, imagine the children of four kinds of religious parent. First, one whose parents, entirely unconcerned for his wellbeing or agency, see him only as a means for increasing the number of true believers in the world; second, one whose parents love him and care for his wellbeing, but, confident in their belief that they know what is best for him, show no concern at all for his autonomy; third, one whose parents, though mindful of their duty not to hamper his capacity for autonomy, have strong religious convictions of such a kind that, despite their best efforts, spontaneous family life makes it hard for him to develop and exercise it; fourth, one whose parents, while careful to ensure that he does develop the capacity for autonomy, find that their love for him sometimes takes the form of deliberately guiding him towards their own religious views.

Although, if my account is right, all four exceed the proper limits of the parental role, it seems clear that they do so to different, and decreasing, degrees. All these children have a complaint against their parents, but those complaints become progressively less severe. In the first case, there is a complete absence of concern for the child; in the second, a failure to recognize the importance of the child’s being able to form and act on her own religious views. The third and fourth cases have a more subtle relation to my proposed account. In both, the parents are well motivated and are providing their child with the right kind of relationship—the problem is that, given their religious views, the only way they can do that either impedes the child’s autonomy (third case) or involves a regrettable compromise between different aspects of the parental role (fourth case). In these two cases we can readily imagine the child forgiving or excusing his parents—after all, they were doing their best for him, given their convictions—and their misjudgments about their proper role were less serious. (See Cormier (2018) for the related thought that, where parents’ deliberate shaping of children’s values can be justified as necessary for the kind of relationship that children need to develop autonomy, children might retrospectively consent to it.)

I want to end by considering two objections to the proposed view of the rights that parents have to influence their children’s religious views. Both worry that I go too far in prioritizing children’s interests and do not give enough weight to parents’, and both will be relevant to the question of religious schooling. First, imagine a critic who says: “I’m willing to grant not only your suggestion that parents’ rights are those that are needed to fulfil the parental role but also your account of what the parental role is. I’ll even accept your view about what that means for parents’ rights with respect to their children’s religious upbringing. But parents are not only parents. As you said at the beginning, they are individuals with their own lives to lead. Surely there must be some scope for them to lead those lives in ways that suit them, and for those ways to affect their interactions with their children, even if they are not thereby serving their children’s interests”. The suggestion here is that a full account of how parents may treat their children must take account of non-parental rights, rather than relying entirely on an account of rights defined in terms of the parental role and its fiduciary responsibilities. Alternatively, suppose someone says: “I agree with you that parents’ dealings with their children must be constrained by the duty to facilitate their children’s autonomy and in other ways to provide children with the kind of relationship that serves their interests. But, as long as they meet those demanding conditions, I don’t see why their rights with respect to their treatment of their children shouldn’t also reflect their own views about what matters in life. Rather than merely “hoping” that her child will come to endorse her own views, surely a parent should be permitted deliberately to guide him towards them.” The issue here is whether we should adopt a “strict” or “lax” interpretation of parents’ rights. On the strict view, parents are limited to those exercises of authority that are demanded by children’s interest in familial relationship goods; on the lax view, they have more discretion.

Although analytically distinct—one appeals to non-parental rights, the other adopts a lax interpretation of parents’ rights themselves—these two objections make a similar point. Both deny that parents should be limited in their dealings with their children to those interactions that they need to engage in to fulfil the parental role and provide the requisite relationship goods; both hold that there is some space for them to act on their own preferences where doing so is not justified by their children’s interests—as long they do not thereby fail to deliver what their children need from the relationship. Although more sympathetic to the former way of formulating the point—which leaves my proposed analysis of parents’ rights intact—I will treat them together.

In my view, the general objection gets something right: parents’ interactions with their children may permissibly sometimes reflect their own preferences. Not only in the spontaneous, unmonitored way discussed above, but also as a result of deliberate considered decision. Parents are people too and it is reasonable for the shared life of the family to reflect their own enthusiasms and interests to some extent. This is partly because it is in children’s interests that their parents are experienced as people with lives of their own. The familial relationship goods account itself emphasizes the value, to children, of parents’ sharing themselves—who they are and what matters to them—with their children. But, even where these distinctively familial relationships are not at stake, i.e. where a parent cannot claim that a particular choice about how to exercise authority is serving her children’s interests, she should be able to continue, at least to some extent, her own, independent, life, by, for example, taking her children with her on holiday to places she wants to see, or to visit friends of hers. This is so even though doing that kind of thing tends to influence, as it surely will, the values and beliefs her children will come to hold. It is in children’s interests that their parents take them on holiday; at that level of description the exercise of authority does indeed comply with Burtt’s (2002:17) claim that “authority over other human beings should extend only so far as making up the deficits that legitimate their subordination”. But that principle is too strict if it is taken also to apply to the particular choice of destination, which—precisely because parents are adults with their own lives to lead—may properly reflect the parent’s own interests and enthusiasms.

None of this, however, justifies the use of parental authority deliberately to guide or direct children towards particular religious views. They are, obviously, free to explain the merits of their preferred doctrines to other adults; in this and other ways they can live their religious lives as they see fit when it comes to those over whom they do not exercise authority and who are not captive audiences. But their authority over their children derives from children’s interests in a relationship of a certain sort, not from their own interest in pursuing and promoting their own religious convictions. If they want to visit a holy place and their children’s interests are served by turning the trip into a family outing, then they make take their children with them. In doing so, they are not exercising a distinctively parental right; they are pursuing their non-parental interests in a way that permissibly affects their children. And since those children had no say in the matter of who their parents are, and have no escape from subjection to parental authority, those non-parental interests may permissibly be pursued only where doing so is compatible with discharging their fiduciary duties.

Religious Schools

Religious schools come in many shapes and sizes. They cater to children of different ages. Some are funded by the public purse, some rely on private resources. Some accept only children raised in a particular faith, others offer them preferential access, still others take no account of the religious background of their would-be students when choosing whom to educate. They vary in their purposes: some seek to direct their students towards a particular faith; others offer a non-directive and autonomy-promoting education that is nonetheless intended for those raised in a particular religious tradition and suffused by its ethos; still others aim to educate all children, without regard to their religious origins or destinations, albeit in a way that is somehow informed by a particular religious standpoint. These cross-cutting differences generate many distinct types of religious school. Contrast, for example, a publically funded school for children aged between 5 and 11 that, while mildly religious in its ethos, is equally available to all and makes no attempt to guide its students towards any particular faith with a privately funded school for children aged 13-18 that admits only those who identify as members of a particular religion and teaches that doctrine as truth. My account of parent’s rights might mean that we are morally required to respect their freedom to choose of some of these types but not others.

In general terms, the implications of that account for issues concerning religious schooling are rather straightforward. Most obviously, the duty to ensure that their children develop the capacity for autonomy precludes forms of education—whether at school or in the home—that deny children the knowledge, skills, attitudes and dispositions needed for them to make and act on their own judgments about the variety of ways in which they might choose to live their lives. Children are wronged if they do not learn about a range of different views on religious matters. And they are wronged if, though informed about that range, alternative views are presented in ways that preclude their coming to see them as real, rather than merely hypothetical, options—whether because the alternatives are presented as unworthy in themselves or because of the excessive emotional and psychological cost of choosing them. This implication—which applies to private or independent schools just as much as to those funded by citizens collectively, and to schools propounding atheism just as much as to those promoting other religious views—is already enough to impugn a great deal of the schooling that is tolerated throughout the world.

Though widely rejected in practice, the state’s duty to protect children’s autonomy is fairly uncontroversial among liberal political philosophers. True, by focusing on the fiduciary nature of the parental role—which includes the duty to help their children to become autonomous—and on the interest in playing that role as the basis of the right to be a parent in the first place, my view offers a distinctive frame. Where most theorists would see the state as restricting parents’ rights for the sake of children’s interests, I suggest a more integrated, less conflictual, way of conceptualizing the issues. But the implications of the familial relationship goods approach for religious schooling go beyond providing a new frame for a familiar conclusion. If I am right, parents do not have the right to send their children to schools that will direct them towards a particular religion, or away from all religions, even where those schools also succeed in providing the kind of autonomy-facilitating education that is demanded by the more conventional liberal position. The fact that one is a child’s parent does not give one the authority deliberately to guide her towards one’s own religious views even in the conduct of family life at home. It certainly doesn’t justify sending her to a school for that purpose.

Might the familial relationship goods account of parents’ rights offer alternative justifications for parents’ decisions to educate their children in religious schools? In the previous section, I noted the difference between the spontaneous and deliberate mechanisms by which parents might influence their children’s religious views. Clearly sending one’s child to a particular kind of school cannot be regarded as the kind of unplanned, un-self-monitored, interaction that parents need to be free to engage in for family life to go well. Just as choosing elite private schooling differs from spontaneous helping with homework (Brighouse and Swift, p.142), so the proper concern to protect valuable familial interactions from counterproductive regulation or self-monitoring yields no support for religious schooling. I also suggested that, since parents are not only parents, it is permissible for them to pursue their own independent interests, at least to some extent, in ways that could be expected to exert that kind of effect. While it’s true that participation in a school’s activities, and other forms of association with co-believers facilitated by a shared school, might serve parent’s own interest in pursuing their religious life, directing their children to attend such a school is different from, for example, taking them on holiday to visit a site of religious significance. In the latter, the parent is deliberately pursuing her own interests through her choice of the particular activity by which she is playing the fiduciary role. The former has no equivalent justification.

But I also considered various ways in which the proposed account might regard deliberate decisions likely to influence their children’s religious views as legitimate exercises of parental authority, so there remains a question about whether any of them apply to religious schooling. Before getting on to those, let me bring out one more general point about the relation between family and school. Where many see the school as an extension of the home, with parental authority naturally extending from home to school, the account I have proposed instead regards the school as a corrective to the home, a crucial safeguard against the risks that inevitably confront those engaging in the challenging task of parenting. The former view has no problem with, and may even prize, continuity between family and school; the latter sees discontinuity as valuable in assisting parents to discharge their fiduciary duties to their children.

Raising children to become autonomous adults is challenging, especially for parents with strong religious convictions. For their healthy emotional, moral and cognitive development, young children need to feel securely attached to, and to identify with, their parents. Yet attachment and identification themselves can easily hamper the development of children’s capacity to make and act on their own judgments about who they are, what matters to them, and how they want to live their lives. Although the familial relationship goods account insists that parents should be mindful of the ways in which the conduct of family life may pose dangers to that development, it leaves room for spontaneous, un-self-monitored, familial interactions that have the potential to obstruct it even though they fall within the scope of parents’ discretionary authority. And quite apart from these psychological developmental factors, even well-intentioned parents may simply be badly placed to provide their children with the requisite knowledge, skills attitudes and dispositions. Schools exist partly because parents cannot be expected to educate their children properly on their own—think about science, mathematics, humanities. The same applies to autonomy. It is through schooling that the state is most easily able to supply the raw materials needed for autonomy: through the curriculum children can experience intellectual and emotional encounters with ideas, values, and traditions that are different from, and sometimes conflict with, those they are raised with in the home. Perhaps more importantly, in a socially and culturally diverse school they can become acquainted with different ideas, values and traditions through the friendships they make and through intimate interactions with their friends’ families. A culturally diverse teaching force can provide children with a range of adult role models who are unlike them and whom they can come to admire. A robust and well-designed extra-curriculum can lead them to discover enthusiasms and interests that would never have been stimulated by their home culture. Discontinuity is educationally valuable (Brighouse 2005).

Somewhat simplistically, then, the more a school takes on the task of promoting children’s autonomy, the less parents need to worry about the autonomy-inhibiting effects of the ways in which, on my account, they may permissibly interact with their children. We might think of this as a division of labour between family/home and school. The former meets children’s affective, emotional and psychological developmental needs through intimate-but-authoritative relationships with particular adults. The latter supplements and complements that familial contribution by widening children’s horizons, by introducing them to perspectives different from those they are exposed to at home, and by teaching them to reflect critically on the choices available to them concerning how they are to live their lives.

The division of labour cannot be complete. However well they play their part, schools have only limited potential to counteract the threat that parents pose to their children’s autonomy, so we should not think that an appropriately constituted school system could leave parents free to conduct family life along religious lines (Weinstock 2018). Parents can adopt strategies to immunize their children from the autonomy-facilitating lessons and experiences that the school provides. Those strategies may sometimes fail, but they will succeed often enough for parents’ choices about how to respond to schools’ attempts to promote critical reflection on the values they themselves hold to remain important. Even where parents make no deliberate attempts at immunization, the emotional pull of their parents’ enthusiasms will be enough to prevent many children from responding to the schools’ messages. This is especially likely if parents take the awareness that another institution is taking care of autonomy as giving them the freedom to be completely uninhibited in their promotion of—or even the non-directive revealing and sharing of—their own values. And because it is important that schools not damage healthy familial relationships, the autonomy-promoting role of schools itself places limits on what parents may teach children at home. Consider parents who, thinking themselves freed from the responsibility for promoting their child’s autonomy by the presence of an autonomy-facilitating school, teach her that homosexuality, or apostasy, are sins punishable by eternal damnation. One can easily imagine that what the school would have to do to facilitate that child’s autonomy would interfere with the familial relationship. The same applies to parents who are deeply hostile to anything other than atheism. We cannot relieve parents of the duty to participate conscientiously in developing their child’s autonomy.

On the question of whether deliberate choices for religious schools might be justified by appeal to familial relationship goods, there are a number of different scenarios to consider. I argued, first, that a parent might be justified in sharing himself with, or revealing himself to, his child in ways that go beyond spontaneous informal interactions. He might permissibly take his child to his place of worship, for example, and he might even require her to experience—and not merely observe—some aspects of his religious practice. As far as schooling is concerned, this suggests a right at most to a school that will teach the child about her parents’ religion. Whether that implies a school with a distinct religious ethos is far from clear: one could imagine a secular school system teaching children enough about their parents’ various religious views that no supplementation was required. In some contexts, perhaps, this rationale might extend to sending one’s child to Sunday school, or its equivalent in other faiths, but only for a limited period and again, and crucially, only where the faith is presented in non-directive terms. The point is to inform and educate the child about the parents’ views—ex hypothesi in ways that the parent cannot do at home—not to guide her towards their endorsement. Where parents do indeed need to draw on this kind of ancillary provision, they should bear the cost privately, just as non-religious parents are expected to use their own resources when sharing their sporting or cultural enthusiasms with their children.

What are the schooling implications of cases where parents’ religious views are such that they can only give their children the loving relationship they need if they are deliberately directing those children towards (their view of) the truth on religious matters? These are those parents for whom sharing, revealing and hoping are not enough. I suggested that they should be regarded as wronging their children—even if excusably so, and even if their children might reasonably forgive them—with the degree of wrong varying across different specifications of the case. The state cannot police familial interactions within the home without denying parents the discretion they need to discharge their relationship duties—so it has to permit parents the space to misuse their authority in that context—but, at least in principle, it can identify schools that are complicit in those wrongs. To ban religious schooling that guides children towards a particular religious view is to protect children from the illegitimate exercise of parental authority. This is so even where that schooling also provides them with the capacity for autonomy.

Readers may wonder whether my claim about the value of discontinuity applies to children at all ages. Given my focus on the developmental significance of parent-child relationships and the obvious point that healthy development involves children’s relationships with their parents changing over time, it may seem strange that I have felt able to ignore the difference between pre-school children and those at primary school, or between primary and secondary education. Even those who agree that a concern for autonomy rules out religious schooling at the secondary stage may think that it is permissible that children attend primary schools that immerse them in a particular religious tradition and teach them to reason in accordance with the ethical framework that it provides. One reason offered for that view is precisely that it is important, for the development of autonomy itself, that a young child’s school experience be consonant with the primary culture that she receives at home. If they are to develop the kind of secure and stable identity that is a precondition of autonomy, young children need their schooling to reinforce the messages they get from their parents, not to undermine them (MacMullen 2007, ch.8).

Suppose this claim is right. It leaves the issue of primary schooling hostage to the way that parents choose to conduct life within the home. Parents can indeed instill in their young children such firm religious beliefs, and beliefs about issues of such high stakes (such as a fear of eternal damnation), that those children may suffer if they are informed about, never mind encouraged to consider the merits of, alternative views. The same can apply in the case of more moderate doctrines, if family life is so suffused with religious practice and observance that children experience the world entirely in its terms. Some of the harm may be specifically to the stability of their identity, and affect the development of autonomy that way, some may be to the familial relationship more generally. What this shows is the extent of the power that parents have over their children—which is why it matters so much that they not exceed its proper exercise. The familial relationship goods account of parents’ rights permits parents to share and reveal their religious views but not the deliberate, let alone systematic, direction of children towards those views. We cannot—should not—police what goes on within the home, so public policy with respect to religious schooling will doubtless have a remedial aspect, adjusting itself to parents’ illegitimate choices and doing the best for children in the circumstances. But if parents observe the proposed constraints, children will not need to be protected from different perspectives, even at the primary stage of education.

This last point takes us into questions about non-ideal circumstances. My argument has been pitched at a purely philosophical level, bracketing real-world considerations and enquiring into the nature of parents’ rights over their children’s religious upbringing in a rather abstract and idealised manner. The focus has been almost entirely on what freedoms the state must, in principle, grant in that domain and on how parents may permissibly exercise those freedoms if granted. This question about how policy should respond to parents’ exceeding their legitimate authority in the home is just one of many issues that demand attention once we factor in the various different ways that parents, or policies, may fall short of the prescribed ideals. For example, most of my arguments apply both to private schools and to those that are publicly funded; going back to basics means denying the public/private distinction its conventional status as an organizing principle to guide policy. But it remains an interesting and important question how the state sector should respond if, as is in fact the case—and as many believe it should be—the regulation of independent religious schools, or of homeschooling, is less strict than that of schools funded at taxpayers’ expense. Just as parents’ acting beyond the proper limits on their power over their children may warrant remedial special measures when it comes to primary schooling, so policies for the public sector may justifiably be affected by the other options that are available. It could be appropriate to adjust the regulation of state schools so as to accommodate parental preferences and thereby reduce parental exit into even less regulated private alternatives. Another example concerns the situation, where one particular religion—such as the Church of England—is granted privileged status within the state system. It may be wrong that Anglican parents can send their children to schools that endorse their preferred religion, but it may also be unfair that they can do so while members of other faiths cannot. Is it right, all things considered, to extend equally to all parents the option of misusing their parental authority, or is that a kind of “levelling down” that unjustifiably imposes a wrong on all children, rather than just on some of them, in the name of fairness?

I don’t have space here to discuss such matters. (See Clayton et al (2018) for a philosophically informed but realistic proposal for a regulatory framework for religious schooling in England.) But one general point is too important to pass by. When I claim, for example, that parents have no right to send their children to a directive religious school, I mean specifically that the state would not be wronging parents if it prevented them from doing so: a policy banning all such schools would not violate parents’ rights. I do not mean that, where such schools are in fact permitted, a parent would never be justified in sending her child to such a school: the options she faces may be such that she not only has the right to choose such a school but even that she has the duty to do so (Swift 2003). Schools vary in many different ways, and conscientious parents will take into account the full range of their obligations to their children. Sometimes parents might choose a religious school not for its religion but for its other properties.

Suppose, for example, that, of the options available to her, only a religious school is “good enough”—all the available non-religious schools are inadequate, in the sense that, in one way or another, parents who chose such a school would be failing to discharge their fiduciary duties to their children. Perhaps the other schools are dangerous, or rife with religious harassment, or perhaps the educational standards are so low that, given her other circumstances, the child does not have a realistic prospect of achieving self-respect or avoiding a life of poverty. (See Merry (2018) for the claim that Islamic schools’ ability to protect their students from stigmatic harms justifies parents’ choosing such schools despite the risk of indoctrinatory harms.) Perhaps, indeed, the alternative schools are less likely to facilitate children’s autonomy than the religious option. After all, nothing in my argument has ruled out the possibility that even a directive religious school may be more conducive to the development of children’s autonomy than the available non-religious alternatives. That will depend, in part, on the content of the religious views to which the child is being directed and the continuity, if any, between those views and the parents’ own.

Many parents believe that they have a right to choose the best available school for their children; some think that they have a duty to do so. If an option is legally available, they are justified in taking it. That is not my claim. Nor, in my view, is their choice justified simply because they believe it is the only one that is “good enough”. Many parents have implausible moral views about what counts as “good enough” and many have false beliefs about what the schools available to them are actually like. My point is simply that we should distinguish the question of whether the state would wrong parents if religious schools were banned, which is what I have been discussing here, from that of whether and when parents may be justified in using such a school where it is available. That is by no means to condone all those parents—whether religious or otherwise—who choose a religious school because they think it better than the alternatives.


By getting clear on what parents are for, and why exactly it is so valuable for many people that they get to be one, we put ourselves in a position to think coherently about the proper scope of parents’ rights. If the familial relationship goods account presented here is correct, then there are many ways in which our current practice is too deferential to parents. Allowing them to raise their children as adherents of their their own religion is one such way. Where that is done at the expense of children’s developing the various capacities needed to make and live by their own judgments on religious matters, the wrong we thereby permit to inflict on their children is grave indeed. But even where autonomy is not impeded, parents nonetheless exceed the proper limits of their authority if they use their power over their children deliberately to guide their children in their preferred direction. Children’s interest in a particular kind of relationship with their parents means that we must leave plenty of room for interactions by which parents will, in fact, tend to influence their children’s views about religious matters. The relationship itself, and the discretion it affords to parents, pose a standing threat to children’s developing the requisite independence. By following the misguided view that policy must respect parents’ preferences for their children’s schooling, and so allowing schools to reinforce the religious messages they get from home, we are depriving those children of their key protection against that threat.

But it is not only their children who suffer from excessive deference to parents. As I suggested at the beginning, that deference also obstructs the legitimate pursuit of civic goals. My argument has focused entirely on parents’ rights over, and duties to, their own children, but many similar policy conclusions would follow simply from giving proper weight to the interests of their fellow citizens. We all have a legitimate interest in how other people’s children are raised; that interest extends beyond the concern that they be trustworthy, capable of trusting others, and able to limit their pursuit of self-interest for the sake of mutually beneficial cooperation. It matters also that they are equipped to play their role as democratic citizens in a liberal state, which requires a range of deliberative and moral capacities that are best developed through contact with, and understanding of, others raised in different religious traditions, and none. Schools are the obvious place for that contact and understanding to be accomplished.

One might reject my restrictive view of parents’ rights, and grant parents more extensive authority over their children’s religious education, while recognising that parents, qua citizens, also have civic duties that properly inform policy with respect to religious schooling. As far as that issue is concerned, my main aim has been to propose a different way of thinking about the relation between parenthood and citizenship. Just as my approach offers an unusually integrated account of parents’ and children’s rights, so too it reduces the conflict between people’s roles as parents and citizens. Rather than balancing parents’ right to send their children to a religious school against the legitimate pursuit of civic goals, we should deny that they have that right in the first place.

Acknowledgements: This article develops—and in part reproduces—work done with Harry Brighouse (Brighouse and Swift 2014). I am grateful for improving suggestions from Matthew Clayton, Tom Douglas, Andrew Mason, two anonymous referees and members of UCL’s Political Theory Workshop. Work on this article was supported by a Spencer Foundation grant for the project Faith Schools: Principles and Policies.


Brighouse, H. (2005). “Channel One, the anti-commercial principle, and the discontinuous ethos.” Education Policy 19:528-549.

Brighouse, H., Ladd, H.F., Loeb, S. and Swift, A. (2018). Educational Goods: Values, Evidence and Decision-Making, (Chicago: Chicago University Press).

Brighouse, H. and Swift, A. (2014). Family Values: The Ethics of Parent-Child Relationships, (Princeton NJ: Princeton University Press).

Burtt, S. (2002). “What Children Really Need: Toward a Critical Theory of Family Structure.” The Moral and Political Status of Children. Eds. D. Archard and C. MacLeod., (Oxford: Oxford University Press) 231–52.

Clayton, M. (2006). Justice and Legitimacy in Upbringing, (New York: Oxford University Press).

——— (forthcoming). The Independence of Children, (Oxford: Oxford University Press).

Clayton, M., Mason, A., Swift, A., and Wareham, R. (2018). How To Regulate Faith Schools, (London: Wiley).

——— (2019). “The Political Morality of School Composition: The Case of Religious Selection.” British Journal of Political Science First View: https://doi.org/10.1017/S0007123418000649 [accessed 25 November 2020].

Cormier, A. (2018). “On the permissibility of shaping children’s values.” Critical Review of Social and Political Philosophy 21:333-350.

Dailey, A. C. (2006). “Developing Citizens.” Iowa Law Review 91:432-503.

Ferracioli, L. (2016). “Why the Family?” Law, Ethics and Philosophy 3:205-219.

Fried, C. (1976). Right and Wrong, (Cambridge, MA: Harvard University Press).

Galston, W. (2002). Liberal Pluralism: The Implications of Value Pluralism for Political Theory and Practice, (Oxford: Oxford University Press).

Gerhardt, S. (2004). Why Love Matters: How Affection Shapes a Baby’s Brain, (London: Routledge).

Locke, J. (1689/1988). Two Treatises of Government. Ed. P. Laslett, (Cambridge: Cambridge University Press).

Macleod, C. (2002). “Liberal Equality and the Affective Family.” The Moral and Political Status of Children (Oxford: Oxford University Press), Eds. D. Archard and C. MacLeod, (Oxford: Oxford University Press), 212-230.

MacMullen, I. (2007). Faith in Schools? Autonomy, Citizenship and Religious Education in the Liberal State, (Princeton NJ: Princeton UP).

Merry, M. (2018) “Indoctrination, Islamic Schools, and the Broader Scope of Harm.” Theory and Research in Education 16:162-178.

Overall, C. (2012). Why Have Children? The Ethical Debate, (Cambridge, MA: MIT Press).

Page, E. (1984). “Parental Rights.” Journal of Applied Philosophy 1:187-203.

Rawls, J. (1971). A Theory of Justice. (Cambridge MA: Harvard University Press).

——— (1963). “ The Sense of Justice.” Philosophical Review 72:281-305; also in his Collected Papers (Cambridge MA, Harvard University Press): 96-116, edited by S. Freeman.

——— (2001). Justice as Fairness: A Restatement, (Cambridge, MA: Harvard University Press).

Reshef, Y. (2013). “Rethinking the value of families.” Critical Review of Social and Political Philosophy 16:130-150.

Richards, N. (2016). “Raising a child with respect.” Journal of Applied Philosophy 35:90-104.

Schoemann, F. (1980). “Rights of Children, Rights of Parents, and the Moral Basis of the Family.” Ethics 91:6-19.

Swift, A. (2003). How Not To Be A Hypocrite: School Choice for the Morally Perplexed Parent, (London: Routledge Falmer).

Weinstock, D. (2018). “For a Political Philosophy of Parent-Child Relationships.” Critical Review of Social and Political Philosophy 21:351-365.

Wölfer, R., Hewstone, M., and Jaspers, E. (2018). “Social Contact and Inter-Ethnic Attitudes: The Importance of Contact Experiences in Schools.” in Growing up in Diverse Societies: The Integration of the Children of Immigrants in England, Germany, the Netherlands, and Sweden, Eds. Kalter, F., Jonnson, J.O., van Tubergen, F. and Heath, A., (Oxford: Oxford University Press).