A child as he grows older finds responsibilities thrust upon him. This is surely not because freedom of the will has suddenly been inserted into him, but because his assumption of them is a necessary factor in his future growth and movement.(1)

This is a mistake. Even fairly young children are capable of making some decisions. More importantly, treating them as if they were devoid of autonomy hampers their becoming fully autonomous adults. As toddlers become preteens and then adolescents, they become increasingly able to assume responsibility and to make decisions about their own lives. We must nourish these abilities if children are to become responsible, autonomous adults. That requires that we treat them as if they were already partially autonomous. Of course we cannot treat children, especially young children, as if they were fully autonomous. Rather, we must find ways to accommodate children's volitional and experiential deficiencies while respecting and cultivating their burgeoning autonomy. That is easier said than done. The difficulty of accommodating both is vividly illustrated in the current debate about whether children should have the authority to decide where and with whom they live.

Deciding Where and with Whom to Live
Should children ever be able to divorce their oppressive or uncaring parents? Or, when parents divorce, should children be able to decide with which parent they will live? Those who think children are not autonomous will find these questions wrongheaded, silly, or devilish. On the other hand, those who think children are sufficiently autonomous so that they have rights, roughly equivalent to the rights of normal adults, will likewise find these questions foolish.(2)

However, many of us find ourselves in the proverbial middle. We want to say what both sides are prone to see as the height of evasiveness: "It all depends." That is, we will want to say that children should undoubtedly be able to free themselves from abusive or uncaring parents however that can be achieved. We also think that although older child should not singularly determine who their primary care-giver should be, their preferences should have substantial weight. Thus, we will want to say children have some, but not complete, autonomy, and that older children should have considerable say in custody disputes. However, the middle ground is always difficult to defend: the position is vulnerable to attack from both sides. So we must explain why we should give children some say while denying them complete say. We must also explain just how much say we think they should have.

This current debate is important in its own right. It also provides an opportunity to think more generally about the nature and scope of children's autonomy. To answer these practical concerns as well as the underlying general issues, we should distinguish three questions: (1) Do children have the rational and experiential wherewithal to make informed decisions about their futures? (2) Should parents (or guardians) permit children to make decisions about their futures? (3) Should the state legally protect children's prerogative to make decisions about their futures? The first question concerns descriptive autonomy, empirical questions about the intellectual and volitional abilities of children. The second and third questions concern normative autonomy, questions about how parents and legal authorities should relate to children.

Historically we have conflated these questions in our discussions about how to treat children, and even in our discussions about how to treat adults. We assumed that adults, but not children, were descriptively autonomous. We then inferred that adults, but not children, should be normatively autonomous. In short, we construed both descriptive and normative autonomy as all or nothing. We assumed that an individual either is or is not descriptively autonomous, and that if she is, then she should have complete normative autonomy, complete say over her self-regarding choices. We correspondingly assumed that if someone is not descriptively autonomous, then we should not give them any normative autonomy. This view puts excessive moral weight on questionable empirical claims and blurry conceptual distinctions. Thus, we confine children to practical purgatory where they have no socially recognized autonomy until, upon reaching the magical age of 17, 18, or 21 (depending on where they live), they suddenly become infused with it. As Dewey attests, that is surely a mistake.

What we need to do is grant children, even young children, circumscribed (normative) autonomy. Although there are good reasons why we should not grant children (especially very young children) complete normative autonomy, we should not deny them a say in important matters. We must train children to become autonomous, and that requires, among other things, that we treat them in some respects as if they were already descriptively autonomous.

This is not merely the old view in a different guise. I am not simply saying that children should not have to live in oppressive environments in which their physical and psychological needs are ignored or thwarted. Virtually everyone even those emphatically opposed to granting children normative autonomy acknowledges that. On the standard view, children should be free of oppressive environments because that is in their best interests, not because that is their choice. That is, governmental officials, serving in loco parentis, decide for children based on their judgement of what is in the children's best interests. Children should have no direct say in these important matters, although, at least occasionally especially when they are older we might ask them for their views, perceptions, and preferences. Nonetheless, we would treat the children's perspective as one additional bit of information, not as an authoritative statement by someone with normative autonomy.

Why Children should not be (normatively) autonomous
Those who are uncomfortable with giving a child any substantial choice over her life, typically rely on one of two grounds. First, they may argue that the child is not descriptively autonomous, that she is not even capable of making a decision, let alone making a reasonable decision. That is, the child cannot comprehend the scope of the decision or appreciate its significance; the child may not even care about what is significant. Second, someone may claim that even if some child were descriptively autonomous, giving her the authority to decide where and with whom she lives will create an environment that will be antithetical to her or other children's best interests. Laura Purdy combines both objections to explain why children should not generally have such authority.

Children are not just miniature adults; they need time to mature and develop the traits that make it possible for adults to live good lives. These morally relevant differences undermine the attempt to show that it is unjust to deny children the same rights as adults; they also suggest that the consequences of equal rights would be harmful to children and society as a whole.(3)

Children's lack of descriptive autonomy
Clearly there is more than a mite of truth in Purdy's claim. Much, however, depends on what she means by "a child." Is Purdy referring to a two-year-old who is not yet potty trained? Or is she referring to a 15-year-old who is watching her parents' marriage fall apart or a 12-year-old who has been systematically ignored by her parents? If it is the former, then of course, the child cannot understand the options, let alone make an informed choice. Consequently, it would not merely be misguided to let the child have a choice, it wouldn't make sense. If, however, we are thinking of either of the latter children, then I want to say of course they probably have a considerable ability to make informed choices, and, indeed, ought to have a defeasible say in controlling their future. I would not suggest that the children's say can never be legitimately overridden by a court exercising its role of in loco parentis. However, that judicial role should be exercised infrequently and then only when there is compelling evidence that these children's decisions are not only unreasonable but demonstrably imprudent. For, without such compelling evidence, we have no reason to think they cannot make a decision that is reasonable enough, or that they should not be permitted to decide.

Those defending this standard position would respond that this ignores the profound intellectual and volitional deficiencies of children. These theorists claim we are not justified in making basic decisions for (most) adults since their cognitive and volitional abilities are sufficiently well-developed so that their decisions are rational, even when unreasonable. That is, since adults can make reasonable decisions (are descriptively autonomous), we are not justified in denying them normative autonomy, even when a particular decision is unquestioningly unreasonable. Children, however, do not just make bad decisions, they lack descriptive autonomy: they are incapable of making informed choices.

Deficiencies which undermine descriptive autonomy
To evaluate this argument, we should isolate five different types of deficiencies that can constrain descriptive autonomy: (1) an intellectual deficit: an inability to mentally process information; (2) ignorance of relevant factors; (3) insufficient understanding of the implications of one's actions for the future; (4) emotional instability; and (5) temporary incapacity.

For present purposes we can ignore the first since we are concerned about how to treat intellectually normal children. We can all agree that a severely retarded child or adult cannot be given the same autonomy as an intellectually normal adult. Therefore, we can limit discussion to the latter four deficiencies.

Ignorance straightforwardly undermines a person's ability to make a reasonable decision. If I do not know the relevant facts, then I cannot weigh those facts. If I do not weigh those facts, then my decision cannot be based on the evidence. Hence, my desires cannot be reasonable. I might stumble on the proper decision. However, that would be luck, not wisdom. Young children are often ignorant of relevant facts. So, they are less likely than adults to make wise decisions. For instance, a child who does not know that arsenic is dangerous will not be able to rationally decide whether to take arsenic. A child who does not understand what alcoholism is will be unable to rationally decide to live with her alcoholic father.

It is not enough that someone knows the relevant factors, she must also understand and appreciate the consequences of her choices. For example, many young children fail to appreciate the consequences of smoking, of unprotected sex, of a failure to study in school, or of living with their clinically depressed mother. Or, even if they abstractly understand these consequences they can recite what someone else told them the consequences would be they fail to appreciate their likelihood or seriousness. That is why parents are disinclined to let children decide whether to smoke or study. Instead, they require their children to study and forbid them from smoking. It also explains why the courts might ignore a child's wishes to live with her depressive mother.

Additionally, even if the child knows the facts and appreciates their significance, she may make an irrational choice if she lacks emotional strength, perseverance, or personal confidence. Even if these emotional deficiencies do not bar her from making a good choice, they may bar her from executing that choice. Someone who lacks self-confidence or is emotionally fragile may be swayed at the last minute by irrelevant factors, while someone who procrastinates or gives up quickly may fail to do what she thinks she ought to do. Finally, if a child is temporarily incapacitated if she is extremely angry, ill, exhausted, inebriated, or depressed then, as in the previous case, she will be both less likely to judge what is reasonable, and less likely to do what she judges to be reasonable.

No doubt these are obstacles to wholly rational choices. Moreover, these four deficiencies are more prevalent in most children than in most adults. That is why we have reason to generally think children are less descriptively autonomous than adults.

The unreasonable and the irrational
Nonetheless, even if children are less descriptively autonomous than adults, we should not conclude that adults should never or only rarely be treated paternalistically while children should standardly be so treated. We can justify treating children and adults this differently only if we can draw a clear empirical and conceptual line between unreasonable choices and irrational choices. Here's why. Everyone acknowledges that adults sometimes make stupid choices. Children also make stupid choices. Why should we force children but not adults to act in their self-interests, against their express (but admittedly stupid) wishes? Presumably because the adults' choices are rational even if unreasonable, while children's decisions are irrational, full stop. (4)

Those who draw this distinction claim unreasonable choices, although not in the agent's best interest, are nonetheless chosen by that agent and are therefore that agent's decisions to make even when others are convinced that the decision is imprudent. Adults are presumably capable of rationally weighing and processing relevant information. Thus, their choices, even when unreasonable, are nonetheless rational. Children, on the other hand, are incapable of weighing or appropriately evaluating relevant features of the available options. Hence, their choices are not fully voluntary. That is why, on this view, children should not be able to make important decisions about their lives, including decisions about where and with whom they will live. These decisions are best left to adults.

However, this distinction between irrational and unreasonable choices cannot carry this moral load. Although we can make a rough distinction between the two terms, we should not conclude that unreasonableness is qualitatively distinct from irrationality. Irrationality is nothing more than the disposition to act unreasonably. This is apparent once we try to distinguish (a) an irrational person from (b) one who always makes unreasonable choices, but is nonetheless rational. How could we distinguish between them? The latter person might display a knowledge of logic; she might cite presumed reasons for her behavior. However, if she standardly made unreasonable choices, we cannot conclude she is really rational.

This point can be made clearer by examining cases where children are supposedly more volitionally limited that adults. We generally assume children, especially young children, are more ignorant than adults. This is a safe assumption. Nonetheless, children are not invariably more ignorant than adults. Some children are highly knowledgeable, and some adults are astoundingly ignorant. Or consider Feinberg's example of the a person who cannot understand the implications of her actions. He claims that children's cognitive disabilities are "not only inabilities to make correct inferences, but also failures of attention and memory, failures to understand communications, and even failures to care about a belief's grounding and implications, leading in turn to a failure to grasp its full import, or adequately to appreciate its full significance."(5) Thus, he claims a 15-year-old is incapable of having a "full visceral appreciation of the significance of an irrevocable transaction for his future interests over the course of a lifetime." That is why we should not let her decide where and with whom she should live.(6)

This sounds convincing enough. Surely we do not want children to suffer long-term damage because of their inabilities to grasp the significance of their decisions and deciding where and with whom they live are very significant decisions. Yet any way we specify this third disability of children will either (a) fail to justify our decisions to systematically deny children a say in these matters, or (b) it will justify us in determining living arrangements for adults in far more instances than we would countenance.

For although it is doubtless true that the run-of-the-mill 15-year-old will fail to have a "full visceral application of the significance of an irrevocable transaction for his future interests over the course of a lifetime," those among us having such an appreciation should throw the first paternalistic stone. Adults like children often fail to appreciate the long term effects of our choices. To the extent that they can appreciate them, it is because they have learned by trial and error. It is not though a clear application of the principles of reason that these adults see clearly and children see only darkly. Consequently, even if a diluted version of the claim were true (as it most certainly is) e.g., that the 15-year-old fails to largely appreciate the significance of a difficult-to-revoke transaction for her future interests over many years, that is equally true of the 40-year-old chain smoker and the 29-year-old helmetless cyclist. On the other hand, if we weaken the criteria of reasonableness so that most adults are properly deemed reasonable, then we can no longer plausibly claim that many children certainly most older children have substantially impaired capacities.

Forcing Children to Become Adults: worries about normative autonomy
Of course most people recognize that children are not wholly incapable of making rational choices. Nonetheless, they may contend that parents and especially the state should not grant children normative autonomy. These objectors fear that in granting children some normative autonomy, we will grant them too much: that we will force them to make significant personal decisions they are ill-prepared to make.

This is a sensible concern if we are talking about a 5-year-old making a complex decision with important consequences for the remainder of her life. What, though, about older children? The previous arguments give us some reason to think that even late preteens are, in relevant cognitive and volitional respects, already mature enough and reasonable enough to decide. They are already more like adults than infants. Most children past infancy are more capable of judging and acting on their own than we typically grant. They are less capable than many adults, no doubt, but they are capable nonetheless.

However, let's consider a child who is admittedly less rational than most adults. Suppose Sue is an emotionally limited 12-year-old living in an abusive home. Should she be given the right to divorce her parents? Some people might claim we already have social mechanisms in place to rescue such children. This is highly dubious. Social service agencies are unaware of many cases of abuse and neglect. Moreover, many abused children will be reluctant to approach those agencies for help as long as they think the agency is likely to simply return them to their abusive parents. Often the child's only realistic hope is to divorce her parents the old-fashioned way: to run away. If we permitted children to legally divorce their parents, then Sue would have another option.

Should we deny her the authority to "divorce" her abusive parents simply because she is incapacitated? No! For the reason she is incapable of making an informed choice is almost certainly that her parents, who are, ex hypothesi, abusive, have not prepared her to make them. Consequently we cannot justifiably deny her the authority to divorce her parents on the grounds that giving her that choice will not prepare her for life as an adult. That assumes she would be so prepared were she left in the current environment. We know otherwise. The best chance to prepare her for life as an adult is for her to escape from her parents, by whatever means, before it is too late.

The real fear here is that establishing legal and social institutions giving Sue the right to divorce her abusive parents will necessarily give the same right to all children. Yet giving all children the authority to decide where and with whom to live will damage many children's (children other than Sue) chances of being well-prepared for life as adults. That is, children in non-abusive homes children who would, in the current conditions, be well-prepared for life as adults will be harmed by prematurely giving them normative autonomy.

Those defending this tack would note that a law which benefits some individuals may be inappropriate for others. However, we should be morally queasy about denying abused children the right to free themselves from detrimental environments so that luckier children do not have to face decisions they might be ill-prepared to make. Nonetheless, if the dangers to other children are substantial, perhaps we must sacrifice children like Sue to protect these other, already luckier, children.

Before we do so, however, we must ascertain the nature of the dangers children will face if we give them normative autonomy. Some fear many children will abuse the power given them: perhaps hordes of them will be scrambling to divorce their parents because their parents did not give them an adequate allowance or forced them to eat their spinach. I see no reason, however, for thinking that more than a minuscule number of children would even consider this option, and even less reason to think courts would permit many of these children to act on their preferences.

I think this objection is more plausible if interpreted as a worry about the indirect costs of giving children normative autonomy. To give children autonomy is to grant them certain rights. However, if children and their parents get mired in squabbles about which rights the children have, it will create an atmosphere detrimental to children. Some two decades ago, Frances Schrag argued that giving children rights, even limited rights, was detrimental to them, and to society. He claimed that the clamor for children's rights would mean that the relationships between parents and children would be

. . . defined increasingly by mutual rights and obligations and that the natural affection and sympathy most parents feel for their children would be undermined. Parents would focus more and more on meeting their obligations or if not on meeting them, at least appearing to meet them in the eyes of the law. Parents would begin to practice "defensive parenting", i.e., the art of meeting the letter of the law to forestall the threat of a future suit . . . Parents . . . would tend to see themselves more and more as sub-contractors to the state performing a definite service in exchange for pay . . . The present relationship of a competent nurse to a psychiatric patient would be the paradigm for the new relationship.(7)

Furthermore, although Schrag raises legitimate worries, these are wholly irrelevant when discussing the fate of a child especially an older child whose parents are divorcing. Here the child finds herself in circumstances that demand that she make a choice. Perhaps she is not as well-equipped to make as wise a decision as she will be able to make later in life (let us hope not). Perhaps it is unfortunate that she must make such a choice. In a different world, political, legal, and social institutions might make it easier for children to maintain regular contact with both parents. Nonetheless, in our world, circumstances often demand that a decisionbe made, and it is difficult to know why the child whose welfare is directly and substantially affected by this decision should not have a say.

Some people might claim we can accommodate these cases simply by requiring appropriate judicial and administrative authorities to seek the child's view. However, I fail to see why this is a better solution. If giving children the (defeasible) choice of where and with whom to live is unduly onerous, then merely asking them what they want to do is likely to be as onerous. It may be more problematic if the children know their opinion will simply be one opinion among many and far from determinative, then they might say what they think these officials want them to say, rather than what they really believe.

Finally, although Schrag's worries are not wholly unfounded, I fail to see that they justify completely denying children normative autonomy, especially if there is a better option.

Circumscribed Autonomy
The preceding analysis supports my earlier contention that our difficulties in thinking about children and autonomy arise because we construe autonomy as all or nothing. We assume an individual is either descriptively autonomous in which case she should have complete say over her self-regarding choices or else that she is not descriptively autonomous in which cases, she has no say at all. We can better understand what is wrong with this standard view once we identify a sixth factor that undermines autonomy: the lack of practice making decisions. John Stuart Mill notes what at some level everyone recognizes: that the ability to make informed choices does not descend on a person, as from heaven. It is acquired by effort and practice.

The human faculties of perception, judgment, discriminative feeling, mental activity, and even moral preference, are exercised only in making a choice. . . . The mental and moral, like the muscular powers, are improved only by being used. The faculties are called into no exercise by doing a thing merely because others do it, no more than by believing a thing only because others believe it.(9)

The lack of volitional exercise not only directly diminishes autonomy, it also intensifies the other deficiencies, especially the second and third. If someone has never had to live with the consequences of her decisions, then she is more likely to be ignorant of factors relevant to her decisions. After all, we want to know the relevant facts only if we realize that we need to know them. Likewise, someone who has not made significant decisions on her own cannot adequately appreciate the significance of her decisions. How could she? Simply hearing someone say that our decisions matter is different from discovering that they matter. If someone is always intervening to save us from our bad choices, then we will neither see nor appreciate the need to make wise choices. That is why as Dewey puts it, talking on responsibilities and making decisions are "necessary factor(s) in our future growth and movement."

In summary, the standard view of children's autonomy errs in two ways: One, it assumes children are less capable of making informed decisions than they are. Two, it ignores the crucial fact that the only way children can learn to become fully autonomous is by first being permitted to act on their growing descriptive autonomy.

To avoid these errors, we should stop thinking about autonomy as all or nothing. We should begin to think of children as having circumscribed normative autonomy, albeit in three different stages: (1) administered autonomy, (2) monitored autonomy, and (3) minimally constrained autonomy.

(i) administered autonomy. When children are very young, they have little ability to decide. However, they must learn, and they must start to learn at an early age. Yet, parents and the states must be careful to protect children from serious repercussions (for themselves or others) of children's unwise decisions. So, although parents should give children some prerogatives, it should be carefully administered, much in the same way as one might administer a medication. The parent tries to let the children acquire a sense of making choices and of taking responsibilities in small matters. Nonetheless, the choices here are a bit more apparent than real since the parent is always on the ready to step in.

One related option is to give the child certain responsibilities, and then expect the child to fulfill them: for instance, expecting the child to carry out the garbage, and then expecting that they will do just that. This will help the child to see that success or failure matters. However, although this may help the child learn to assume responsibility, unless the child also has some autonomy, even if limited, she is unlikely to learn how to make wise choices on her own, rather than simply fulfilling the demands of others.

(ii) monitored autonomy. When children are still quite young, but after they have coped with the minimal doses of administered autonomy, the parent should give them greater choice and greater responsibility. Here parents do not parcel out autonomy in a precise regimen that gives children little latitude. Instead, the parents let the children make choices and assume responsibilities and, to some degree, let the child cope with the consequences of her decisions. Nonetheless, the parents will loosely monitor their child's choices. That is, the parents let the child act in ways that hurt her a bit; but they are prepared to intervene if it appears the decision is dangerous or highly detrimental. Nonetheless, they will not intervene as quickly or often as they would in the previous stage.

(iii) minimally constrained autonomy. As a child becomes still older and has developed the increasing intellectual means and emotional stability to generally make wise choices, parents must give her increasing autonomy. The parent may still intervene from time to time, but only infrequently, and even then not with a heavy hand. The parent will knowingly let the child make more serious errors than they would have at earlier stages, although they might still intervene in ways they might not intervene in the life of another adult. Moreover, if the parents have taken the children through the first two stages, then typically they can intervene best by simply talking with the child.

The Parent and the State
The previous discussion explains why the parents rather than the state play the pivotal role in giving children normative autonomy. The child's metamorphosis from the irrational through the partly rational to the mostly rational must be overseen by a caring and proximate parent, not by a distant judge. Judges may be required to intervene, however, if relationships between children and parents fail. In the ideal world, though, astute parents will give imperceptibly increasing normative autonomy to their increasingly descriptively autonomous children. There are no clear rules that tell them how much autonomy to grant or when to grant increasing prerogatives to their children. Parents must be attentive to the ways in which their children have behaved in the past, and be ever alert to what they can do to enable their children to make better decisions. Often the best way is by letting them make decisions for themselves and then to expect them to cope with the consequences.

Still, the state does play an important role. Skeptics will not want the state to play more than a minimal role. They will defend the current view of the legal status of children by saying that although some children are descriptively autonomous much of the time, that many are not. The state, however, must "draw the line" somewhere: they must distinguish those people to whom we should grant normative autonomy from those to whom we should not. Of course, there is no way to draw the line that always leads to the correct result. Anywhere we draw the line we will give normative autonomy to some adults who are not descriptively autonomous, while denying normative autonomy to some children who are descriptively autonomous. What the state cannot do, they say, is decide on a case by case basis how much scope to give children. That, they could would have even worse consequences. Therefore, we should deny all legal rights to children until they reach the age of majority.

There is, of course, a morsel of truth to this view. As I argued earlier, there are dangers of granting children too much normative autonomy. However, there are still greater dangers to children, and ultimately to society, if we cannot find some model of autonomy that allows that children are descriptively and normatively autonomous in different degrees. More important, this model must recognize that the only way children will ever become descriptively autonomous is by granting them increasing normative autonomy. Even if the legal system does not get into the business of deciding how every parent should treat every child, legal pronouncements can set appropriate expectations that would make it more likely that children are given increasing normative autonomy.

We see a move in this direction within the both US and UK legal systems. US federal courts have ruled, in various cases, that we cannot categorically deprive all children of normative autonomy, since many older children are, in fact, descriptively autonomous at least as autonomous as many adults. For instance, the courts have acknowledged that high school students should have some freedom of speech, even in the public schools.(10) Likewise, the British Children Act of 1989 specifies that in cases where children must be placed, the courts must "give due consideration (having regard to his age and understanding) to such wishes of the child as they have been able to ascertain" (section 20). These decisions will not only benefit children who should be given more normative autonomy, they send a message to parents a legal sermon from the legislative and judicial pulpits encouraging parents to grant their children increasing normative autonomy.

Of course neither the courts, nor the legislatures, nor parents can find a consistent way to resolve all these questions: one law or decision appears to permit what another denies. I think we should see these confusions and difficulties, however, not as signs of sloppy judicial, legislative, or parental thinking, but as a faithful rendering of the confusing psychological facts and fuzzy moral boundary between the unreasonable and the irrational. If my earlier arguments are correct, then we should expect fuzzy rulings, laws, and decisions after all, when it comes to children, there are no clear, "objective" standards on which we can always rely. Instead, parents, the state, and judges must ascertain the salient facts and attend to relevant considerations. Deciding how much normative autonomy children should have will be difficult. Deciding whether we should honor the wishes of a seven-year-old child in the midst of a heated custody dispute will not always be evident. However, that in no way implies that we should condemn children to abusive homes from which they cannot escape, or to letting others, no matter how well-meaning, make all profound decisions about the future of their lives.(11)

(1) John Dewey, The Moral Writings of John Dewey, ed. J. Gouinlock. Buffalo: Prometheus, 1994, p. 195.

(2) Howard Cohen, Equal Rights for Children. Totowa, NJ: Littlefield, Adams, and Company, 1980.

(3) Laura Purdy, "Divorce, 90's Style," The World and I., 1994, p 369.

(4) Hugh LaFollette, Personal Relationships: Love, Identity, and Morality, Oxford: Blackwell, 1996.

(5) Joel Feinberg, Harm to Self, Oxford: Oxford University Press, 1986, pp. 317-8.

(6) Ibid. p. 325.

(7) Francis Schrag, "Children: Their Rights and Needs," in W. Aiken and H. LaFollette (eds), Whose Child: Children's Rights, Parental Authority, and State Power. Totowa, NJ: Rowman and Allenheld, 1980, pp. 246-7.

(8) Hugh LaFollette, "Two Forms of Paternalism." In A. Peczenik and M. Karlsson (eds), Law, Justice, Rights, and the State, Franz Steiner Verlag, 1995, 188-94.

(9) John Stuart Mill, On Liberty. Indianapolis, In: Hackett Publishing Company, 1978/1865.

(10) Tinker v. Des Moines Independent School District, 393 US 503,111.

(11) I wish to thank members of the Philosophy Department at East Tennessee State University, members of the Philosophy and Religion Department at the University of Northern Iowa, members of the audience at the 1997 meeting of the Political Studies Association (UK), the editors of this volume, and especially David Archard, John Hardwig, and Eva LaFollette for insightful comments on earlier drafts of this paper.