Jean-Jacques Rousseau founded his political philosophy on the myth that people existed originally in a state of nature without any network of social relationships.  Isolated individuals were perfectly free, but also desperately insecure: vulnerable to attacks from bears, gryffons, monsters, hailstorms, and other human beings; hunting and gathering food with the use of their own strength and wits without the help of family or neighbours; struggling through good and bad times without unemployment or sickness benefit, with­out social workers, or district nurses; facing old age without residential care; and dying in a ditch without medical help or even herbal remedies, except for any they had found for themselves.  Societies were formed when these dis­connected individuals came together to make a contract in which they gave up part of their freedom in return for the protection and help of others.  Social order arose from the freely given consent of equals who might have decided to remain in the original state of nature.  Values, norms, rules, the law were the outcome of a rational decision by free men and women to allow a portion of their capacity for choice to pass permanently to a central authority.

One reason why this has been such a fruitful myth is that, as Rousseau in­tended, it provided a powerful argument against the arbitrary exercise of power.  But it was much more than a convenient justification for a political programme.  It represented a shift of the paradigm used to think about the nature of social order.  Before Rousseau it was usually assumed that societies must have a basically unchanging hierarchical form, and that order spreads down from the top of the pyramid.  In material terms, order was maintained by the sanction of force, with higher ranks having the monopoly of weapons, armies, and the technology of warfare.  In terms of beliefs and attitudes, the hierarchical order was supported by a pow­erful ideology which made it seem unthinkable that society could be other­wise.  The power of the ideology that legitimised absolutist government is conveyed in the famous speech of Ulysses from Shakespeare’s Troilus and Cressida.

The heavens themselves, the planets, and this centre

Observe degree, priority, and place,

Insisture, course, proportion, season, form,

Office, and custom, in all line of order…

                              but when the planets

In evil mixture to disorder wander,

What plagues, and what portents, what mutiny,

What raging of the sea, shaking of earth,

Commotion in the winds, frights, changes, horrors,

Divert and crack, rend and deracinate

The unity and married calm of states

Quite from their fixure!  O! when degree is shak’d,

Which is the ladder to all high designs,

The enterprise is sick.  How could communities,

Degrees in schools, and brotherhoods in cities,

Peaceful commerce from dividable shores,

The primogenitive and due of birth,

Prerogative of age, crowns, sceptres, laurels,

But by degree, stand in authentic place?

Take but degree away, untune that string,

And, hark! what discord follows…

Not illustrated in that play, because it is set in ancient Troy, is the additional backing to the social hierarchy given by the state religion of Christianity which represented the ranks and stations of people on earth as reflecting a heavenly or spiritual hierarchy, with the monarch at the apex of the political order taking his authority directly from God.  From the seventeenth century that model of society came under increasing strain.  Rousseau’s essay on the social contract is a particularly clear indication of the resulting paradigm shift.  After Rousseau, people were no longer restricted to imagining social or­der as a natural arrangement that had to be maintained by authority im­posed from above with the backing of physical force and divine endorsement.  Instead, it became possible to conceive of social order as the outcome of a ne­gotiation between equals.

The thought that I want to develop today is that order is negotiated, not simply imposed.  Historical changes over the past 200 years have made this in­sight more and more important.  I want to illustrate ways in which the broad principle of negotiated order can help us to understand parenting, youth crime, and juvenile justice.  First I need to say more about the principle itself.

THE PRINCIPLE OF NEGOTIATED ORDER

In recounting the myth of the social contract, Rousseau was trying to grapple with a number of paradoxes.  First among these is the consensual nature of authority.  The theme was later taken up by the German sociologist, Max Weber.  Order seems to be imposed, authority commands, with force as the ultimate threat, and the less powerful usually obey.  Yet even in a slave soci­ety, the exercise of power is not primarily through direct physical force, but through shaping objectives, priorities, perceptions, moral precepts, and ideology.  Power is ultimately backed by the threat of force, but the threat cannot be implemented without control of minds, otherwise, for example, the police or the army may rebel.  Naked force may be used to intimidate one group, but then the ruler needs first to win the confidence of another group that will do the intimidating.  The ruler must therefore start, not with intimidation, but by winning the support of a powerful group of protectors.

Even among the intimidated, the exercise of power depends on persuasion.  As in present-day Afghanistan, vastly superior power may be capable in the end of doing no more than devastating the country, and either slaughtering or starving the inhabitants.  People can always choose death rather than submission, and some do.  Others may be frightened into obedience for the time being, but that must be a precarious means of control.  In the long term, power will only be effective if it persuades by changing percep­tions, assumptions, and beliefs.

Hence power without legitimacy is always insecure, and even in authoritarian systems leaders have to negotiate in order to achieve legitimacy.  In that sense, order always depends on negotiation to some degree.  Rousseau’s myth expresses the idea that governments do exercise power, but only with the consent of the people, so that the ruler’s power is borrowed from the people.  This is a comment on society as it is, and in that respect it draws attention to the fragility of power: the consent of the people can be withdrawn, and gov­ernments do lose control because they lose legitimacy, as they did in the Soviet Union, East Germany, and Romania, for example.  The myth is also a comment on the moral basis of authority: the government has no absolute moral right to rule, but rather acts as the steward of the people’s will.

Although it clearly expresses the conditional nature of power, the social con­tract myth is also open to fundamental criticisms.  Two of these are worth discussing, even though they are very well known, because they go to the heart of the problem of understanding child development and juvenile justice.  The myth pre-supposes that individual people, much as we see and under­stand them today, existed in an original state of nature without ever having had a network of social relationships.  This is not intelligible.  It does not make any kind of sense.  From the story of Romulus and Remus onwards, there have been tales of infants raised by wild beasts in the forest who later became normally functioning human beings, but these stories are fabulous and completely incredible.  All documented cases of infants raised without human contact show that they are damaged beyond repair: cru­cially, they never learn to speak.  These facts show that a person cannot develop apart from society, because what we think of as a person is the product of a long period of development, involving a myriad of interactions with other people.  Language can only be acquired through these interactions.  Furthermore, an organism without social relationships could not form a concept of its own self, because we see ourselves through the eyes of other people.  Once through the critical process of socialisation, someone may live in complete isolation for many years and still remain recognisably a person, which is why the story of Robinson Crusoe is perfectly credible.  But a collection of humanoids who had always lived in isolation could not come together to make a social contract, any more than a collection of cats could.

This is not just a pedantic objection, not just a refusal to look past the details to see the essential truth of the myth.  On the contrary, it highlights a point of principle: that individuals do not exist prior to society.  Instead, individu­als are constituted by social relationships.  An individual person is a history of social interactions and a network of social relationships.  Different psy­chological theories give different accounts of how the history of the individ­ual’s social interactions is deposited within the person, and becomes who the person is and what he or she is like.  On a Freudian account, for example, the commands of the father and mother are introjected and become an aspect of the self, a super-ego that stands over the ego and constantly tells it what to do, and more important, what not to do.  On a behaviourist account, a person repeats actions that other people notice and reward, and avoids actions that other people ignore or punish.  All theories imply that the structure of social relationships is built into the developing individual, so that each individual person contains a great deal of evidence about how society as a whole is or­ganised.

It follows that the normative order of society could not be created by individuals coming together and freely contracting to give up some of their free­dom and adopt agreed rules.  This is because the normative order of society and the choosing individual with consciousness of self and moral judgement come into being together.  They develop jointly out of a myriad of social inter­actions.  Neither the normative order of society nor the choosing self comes first.  Both grow out of social relationships.  To use a biological metaphor, individual psychology and social relationships are symbiotic, or co-dependent.  Each one is constituted by the other.

The second criticism of the social contract myth is closely related to the first.  According to the myth, the social contract was negotiated just once at a dis­tant epoch during some shadowy pre-history.  This does not fit with the facts of secular change, or with comparisons between existing societies.  The level of order has changed substantially in European countries over documented historical periods.  For example, Manuel Eisner is currently collating evi­dence to show that the rate of homicide in several European countries has de­clined by a factor of 100 or more since the middle ages.  Again, making cross-sectional comparisons, it is clear that the level of order now is substantially higher in Switzerland and Japan than in Albania or Russia.  Equally impor­tant, compliance with societal norms is certainly not constant over the life cy­cle.  On the contrary, infants are amoral creatures who are impossible to con­trol, but they usually grow into biddable children then self-regulating adults.  In the light of the unmistakable facts of secular and life-cycle change, it is fruitless to maintain that the foundations of social order were built only once in a dim pre-history that nobody can remember.  Instead, social order is con­tinuously made and re-made in every country and in each generation.

The story of the development of each individual person from the egg is the story of the growth of three intertwined systems: a choos­ing self, social bonds arising from reciprocal relationships, and a normative order which must be constantly re-created.  Events in these three domains interact repeatedly and rapidly.  For example, developing moral ideas are immediately influenced by exchanges with parents and siblings, but these ex­changes are influenced in turn by the growing person’s moral state, for example their level of self-control.  People become people through belonging to networks of social relationships.  It is through social exchanges that they come to understand the consequences and moral meaning of actions.  Only through being located in social networks that exert constraints can people be­come moral agents who evaluate alternatives and are therefore free to choose.   Only through relationships with other people can they form an idea of themselves as actors in a social world.  Development can only be a double process involving both the internal state of the individual and external social relationships.  A normative social order arises both from the relationships themselves and from the moral perceptions that mirror the relationships, yet also influence them.

The problem, then, with the social contract myth is that it suggests that al­ready existing people get together to form an ordered society, whereas if there were no ordered society, these individuals would not be people.  The solution to the problem is that people and society constitute each other, and necessar­ily develop in tandem.  Social order therefore emerges organically out of the way that children develop and the way that people relate to each other in small groups.  Rousseau’s basic insight remains valid, however.  At every step in the sequence of interactions that build social relationships and frame child development, order is negotiated, not imposed. 

APPLYING THE PRINCIPLE  

Crime can be understood as a breakdown of negotiated order, and the insight can be applied in many fields: to fraud and corruption as well as to domestic violence, to international trafficking in women as well as to housebreaking and vandalism on run down council estates.  However, I want to concentrate in these remarks on the wide spectrum of ordinary, locally-based crimes that are mostly committed at a particular stage of the life cycle: in adolescence and early adulthood.  Typical examples are theft, taking cars, spraying graf­fiti, vandalism, dropping objects onto trains and railway lines, starting fires, housebreaking, robbery, fighting, and assaults, most of them relatively mi­nor.  It is possible to argue that major frauds are more important, for exam­ple in terms of the total values involved, or that violence in the home is more important, for example in terms of the number of incidents if they could be properly counted.  However, what I have called ordinary crime still domi­nates the criminal justice system, and it constitutes at present levels a major disruption to social order, especially in certain neighbourhoods, where most of us don’t live.  It peaks at the age of 18, and becomes much less common from the mid 20s onwards.  Ordinary delinquency, therefore, is a disruption of social order that is specifically associated with the transition from child­hood to adulthood; it is evidence of turbulence arising in the process of re-creating the normative and social order in the next generation.

Child development and parenting can best be understood as the stage by stage construction of negotiated order against the background of the order existing in the wider social world.  Ordinary crime, and especially rising lev­els of ordinary crime, are a sign that this work of construction and re-construction is failing, or at least encountering increasing difficulties.  The creation of separate juvenile justice systems in every jurisdiction is a recognition that a long process of development is needed to create the freely choosing moral agent who fully belongs to the social and moral order.  Crucially, the peak age of offending comes at the moment of maximum ambiguity in the process of development from childhood to adulthood, when young people are not quite ready for independence, but cannot be given summary instructions like children.  Juvenile justice systems recognise the crucial  importance of re-negotiating order with young people where it has broken down at that critical, ambiguous stage.

Over the past 20 years, there has been a growth of initiatives and programmes of restorative justice in many countries.  A recent international re­view[1] shows just how diverse these programmes are in terms of aims, objec­tives, underlying principles, the stage of criminal process at which they oper­ate, and substantively what the interventions consist of.  For all its complexi­ties and confusions, this movement signals an increasing emphasis on nego­tiating order rather than imposing it.  In fact, the principle of negotiated or­der could well be used to clarify the aims of restorative justice.  Advocates of these programmes give many conflicting answers to the question ‘What is being restored, and to whom?’  A possible answer would be ‘Order is being restored through negotiation with all parties’.

HISTORICAL CHANGE

The principle of negotiated order, then, has a particularly clear application to child development and parenting, and to the way the official systems respond to a breakdown of order in the development of young people.  A number of historical changes come together to make the principle especially important now.  Rousseau’s writings were part of the shift from the seventeenth century onwards away from absolutism towards democracy, which has now become the official religion of much of the world.  In democratic societies, there is ex­plicit recognition that the legitimacy of the political order depends, and should depend, on transparency, information, discussion, negotiation, flexi­bility, openness to change, the possibility of redress.  Also, there is increasing emphasis on the value of participation in decision making in many domains, even if there is generally little active participation in the political process itself.  What is meant by democracy is a set of values or aims, such as transparency, negotiation, and redress, much more than a set of processes, such as voting at elections.  The increasing emphasis on those values is bound to in­fluence the relationships between parents and children.  Some historians have probably exaggerated the brutal and authoritarian treatment of chil­dren in the middle ages and up to the sixteenth century and beyond.  Others[2] may have overstated the idea that children in the middle ages were treated as minia­ture adults, and that childhood was not regarded as a form of life with its own internal coherence and satisfactions.  Still, there is no doubt that the authoritarian hierar­chy of earlier societies with absolutist governments was reflected to some extent in attitudes towards children and parent/child relationships.  The change in political culture is paralleled, particularly over the past 50 years, by a change in everyday relationships, for example in the workplace, and by a huge effort to make service providers in both public and private sectors more accountable.  These are pervasive changes that influence all relationships in society, so they must also have a profound effect on the way that parents re­late to their children.  An authoritarian style of parenting, admitting little or no negotiation, used to be widely accepted as normal.  It is no longer acceptable now, because it is out of step with the values of democracy, and because it conflicts with practices in many other spheres, such as the workplace or even the school.

An equally important change over the past hundred years has been the lengthening of the period of transition from childhood to adulthood. At the beginning of the twentieth century, most children left school at the age of 12 or earlier.  As soon as they joined the labour market, they were likely to have contact with many adults as well as others of their own age.  Of course, in many spheres of life they would not make the transition to adulthood until much later: for example, it would be unusual to get married, leave home, and set up a separate establishment before the early twenties.  Still, the short period of education meant that most young people were segregated by age group for only the first twelve years of their lives.  Now, young people typically remain in full-time education until the age of 18, and around half continue until the age of 22 or more.  For a longer period of their lives than before—six to twelve years longer—young people now spend most of their time in the company of other people in their own age group.  Effectively, the period of youth has been considerably extended. 

Age segregation has been accompanied by the creation of a distinctive youth culture, ex­pressed for example in music, dress, and language.  Young people now spend a much longer period than before preparing for full adulthood, which takes longer and longer to arrive.  An increasing emphasis on individualism gives them more scope to take their own decisions in some domains, such as sexual relationships, while at the same time many of them remain economically de­pendent on their parents for a longer period than before.  These complex and in some ways contradictory trends have clearly identified adoles­cence as a separate phase between childhood and adulthood.  They have marked it out as an ambiguous phase, where a person is more than a child but less than an adult; expected to assert their individuality, perhaps in an aggressive or rebellious way, yet often still dependent on their parents; largely segregated among people of their own age group, but charged with the task of preparing for life as an independent adult.  A further set of ambiguities is created when the values, preoccupations, and social practices of youth culture start to have a wider influence, so that adults imitate or nostalgically re-create the habits, pursuits, and manners of adolescents.

An obvious consequence of these changes is that old-fashioned parental au­thority is bound to be called into question during a lengthy, ambiguous phase of youth.  The need for parents to negotiate order during this period is par­ticularly acute.  Negotiating successfully is important especially because the long, ambiguous period of adolescence includes the peak age of offending.

PARENTING AND DELINQUENCY

Fifteen years ago, Rolf Loeber and Magda Stouthamer-Loeber completed a review of the extensive evidence on links between parenting and later delin­quency among children.[3]  They concluded that there was extensive evidence to show that several aspects of parenting have an influence.  Where parents supervise and monitor the children closely, and where they are involved with children and do things with them, delinquency is less likely.  Where parents are themselves criminal, where discipline is inconsistent and harsh, and where there is family conflict, delinquency is more likely.  These conclusions would be much the same if the review were to be updated today.

However, we need to take a closer look at some of the research findings in or­der to understand what they imply about relationships between parents and children and how they influence child development.  Unlike most of the psy­chologists who carry out this research, we need to be candid about inquiring into the power relationship between parent and child, and to ask whether and how this allows scope for negotiation.

Some of the most interesting and important research in the field has been carried out by Gerald Patterson and his colleagues at the Oregon Social Learning Centre.[4]  Like much of the relevant research, this programme has the severe limitation that it was restricted to boys.  The great difficulty in finding out about family functioning and child/parent relationships is that neither parents nor children give unbiased or accurate accounts of what goes on.  Individual interactions are quick and fleeting, and may run into hun­dreds or thousands a day.  Most researchers therefore have to rely on broad summaries or generalisations about complex interactional sequences: for example, they ask mothers  a question like ‘Once a punishment has been decided, how often does Johnny go on arguing about it?’ or ‘How often do you give up when you ask Johnny to do something and he doesn’t do it?’.  The great strength of Patterson’s work is that he uses a range of methods to de­scribe interactions between parents and children in some detail.  Although he does use questionnaires to parents and children that incorporate the usual generalisations and over-simplifications, he combines that with many hours of observation in the home, video-taped problem-solving tasks, and stan­dardised tests.  This still does not provide a blow-by-blow account of specific interactions, but it does give a much more textured description of kinds of behaviour and how frequently they occur.  Also, his team has given close attention to the development of robust measures from several convergent sources.  The penalty to be paid for adopting this detailed approach is that the number of families covered has to be fairly small: 200 were covered in the research programme I am describing now.  This means that only large effects will be statistically significant.

Patterson focuses on unpleasant behaviour which has the function of getting a result: for example, mother refuses to give the child a sweet, the child begins to whine and goes on whining until mother gives in and lets the child have a sweet after all.  Patterson does not assume that the child intends to get the sweet by whining, nor that the child self-consciously starts to whine as a means of getting the sweet.  His approach is behaviourist, and he tends to disown explanations in terms of conscious thought processes.  Instead, he argues that the child adopts the whining behaviour because it has been reinforced in the past by getting a sweet.  From the observer’s perspective, therefore, the function of the whining is to get a sweet, although the child may or may not be conscious of that, and whether or not the child is conscious of it probably is not important.

Exchanges like this one, in which one person uses unpleasant behaviour to get what they want, are described as coercive exchanges, and some families are found to be far more coercive than others.  How do they come to be coer­cive?  Patterson argues that this starts from a breakdown in the effectiveness of parents in their disciplinary confrontations with children.  This leads to an increase in coercive exchanges between the child and other members of the family.  By repeating these exchanges, the child finds that aversive behaviours are effective.  They work for him both by turning off the unpleasant behaviours of others (for example, hitting sister stops her teasing) and by pro­ducing positive benefits (for example, he gets to play with her toys).  This leads to an escalation of the coercive exchanges: they become more frequent, and the child’s behaviour also becomes more severely unpleasant.  That leads in turn to a decline in parental supervision.  Effectively the child coerces the parent into allowing more unsupervised time.

From the description so far, it sounds as if the child’s behaviour, although very annoying to the family, is essentially trivial.  However, there is good evidence that it will have serious long-term consequences.  Through many repetitions of these coercive exchanges, Patterson argues that the child will start to possess a stable antisocial disposition, which will tend to be expressed in all situations.  At the next stage, the social environment reacts to the child’s antisocial behaviour.  Other children, if they are reasonably well-behaved, are likely to reject the antisocial child.  Parents and teachers are likely to avoid or reject him.  The child therefore starts to fail at school, and may stop attending.  Rejected by most well-behaved children, he seeks out similar antisocial peers.  Peer influence and school failure are likely to lead to a serious criminal career in adolescence and early adulthood.  In contrast to those criminologists who have been intoxicated by the romantic notion of the young offender as innovator and cultural pathfinder, Patterson insists on the drab and dislocated failure of most criminal lives.  ‘Theirs is not the selective resistance to authority that characterises the effective rebel.  Instead, they defy all authority figures and rules that restrict them.  They are not even very good at fighting or stealing.  These boys are losers.’[5]

It is argued that all of these long-term consequences flow from the way that fractious behaviour, trivial at the beginning of the sequence, is handled in the home.  In the most common sequence, described as ‘escape conditioning’, the child’s aversive behaviour is a way of escaping from something unpleasant.  For example, sister teases her brother, he hits her, she stops teasing.  Because the teasing is stopped, he will be still more likely to hit her again on a future occasion.  Those who are socially skilled learn to neutralise or deflect unpleasant behaviour, for example by ignoring it, or by using humour.  In coercive families, nobody responds in that kind of way.  From their detailed observations, Patterson’s team find that in coercive families the daily routine consists of a steady torrent of low-intensity aversive exchanges punctuated by occasional powerful explosions.  The annoying behaviour is going on all the time at a low level, and is mostly ignored, but every now and then someone reacts, and each time that happens the conditioning sequence is set off.  These sequences act as a training programme in coercion.  Through that programme children learn that coercion works for them in the immediate situa­tion, although in the long run they find themselves trapped by the long-run consequences of their own behaviour, which are extremely damaging.

Coercion training depends on having frequent interactions with people who start behaving unpleasantly but reliably withdraw once they are counter-at­tacked.  Why does this happen much more in some families than in others?  One reading of the work of the Oregon team—although it is not entirely ex­plicit in their writings—is that the explanation for coercive family function­ing lies fundamentally in the balance of power within the family.  In coercive families, parents and children have roughly equal power, and alternate in as­suming the roles of victim and aggressor.  Parents allow children to win about as often as they do.  In normal families, power is unequal, and the par­ent usually wins.  There is a corresponding difference in the clarity of rules and boundaries.  Coercive exchanges are more likely in families where the rules for child behaviour or the roles of family members are not well defined.

How then do parents in normal families gain control and establish an unequal balance of power, and why do parents in coercive families lose control?  Patterson’s team argue that there are two aspects of parenting practice that are important: parental monitoring, and effective parental discipline.  Monitoring is frequently collecting information about what the child is doing, where he is, who he is with, and so on.  Effective discipline has three compo­nents: accurately tracking and classifying problem behaviours; ignoring trivial coercive behaviours; and ensuring that unpleasant consequences consistently follow whenever punishment is necessary.  They argue that parents with problem children fall down on all three tests of effective control.  They are too inclusive and undiscriminating in what they classify as deviant, so that they waste their energy on trying to stop behaviour that is inconsequential.  They are less likely to ignore coercive behaviours, when ignoring them is often the best response.  They are more likely to ‘natter’ in response to them, that is to respond in a mildly annoying but ineffectual way; and they are also more likely to respond coercively themselves.  Their coercive responses often don’t function as punishments, but instead stimulate immediate further bad behaviour.  Finally, parents with antisocial children typically do not back up their requests, so there are no reliable bad consequences for the child if he does not comply.  Also they fail to reward compliance consistently.

So far these results do not demonstrate the importance of negotiation.  In­stead, they emphasise the primordial importance of establishing parental control, and identify the techniques for achieving it.  However, compliance is not secured by authoritarian tactics: not by aggression, violence, instilling fear (the tactics of absolutist government).  In fact, it is the parents who are not in control who tend to use these tactics.  Control is secured by unruffled consistency, whereas unconditional aggression (unconditional in the sense that it is not consistently related to the child’s behaviour) does not secure control and stimulates intensified sequences of bad behaviour in the child.  That kind of parental aggression is analogous to the unpredictable severe punishments of a weak ruler.  This whole analysis by the Oregon team is also analogous to the argument by Beccaria and Bentham that moderate punish­ments combined with a high probability of being caught are the most effective deterrent against crime.  An important implication of these arguments is that parents who are in control are strong enough not to be aggressive.  This suggests that in the parent/child relationship, power is productive, and weakness is destructive.

Although they argued in favour of the primordial importance of parental monitoring and effective discipline, Patterson and his colleagues also proposed that three aspects of positive parenting would have an important influence: joint problem solving, tracking and encouraging co-operative behaviour by the child, and parental involvement, which means spending time doing things with the child.  The first of these, which they call problem solving, is in fact negotiation.  They explain that in the paradigm case of problem solving, there are four stages of negotiation between two people, in this case parent and child: one clearly states the problem; the other paraphrases to check there is a common understanding; they both brainstorm solutions; finally they jointly choose a solution through negotiation involving compromise, and write a contract (the very word used by Patterson and colleagues) that describes the agreement, the positive con­sequences flowing from following it, and the negative consequences from breaking it.  It cannot be a coincidence that this description owes something to Rousseau.  But here the contract is not a once-and-for-all agreement that flips isolated individuals out of a state of nature into society.  Instead it is one of a series of negotiations that take place at the many different steps along the road of the progressive socialisation of the child.

Patterson and colleagues found the effect on antisocial behaviour of the three measures of positive parenting, including negotiation, to be much weaker than the effects of parental monitoring and discipline.  That may sound de­pressing, but it may also be misleading.  The analysis presumes that negotiation, if it is to work, must be between parties with unequal power.  This is in sharp contrast to Rousseau’s model of free and equal individuals coming to­gether to make a contract.  The problem with that model is that the contract cannot be enforced, as there is no established power structure to enforce it.  Order first needs to be guaranteed by a powerful ruler before negotiation takes place.  Although this may run counter to intuition, negotiation between unequal parties is the typical case: for example, negotiation of a peace treaty after war.  If it is asked ‘Why negotiate once the battle is won?’ the answer is that the loser is seldom utterly defeated, and the winner does not want to start fighting again, so there is usually room for compromise.

The model, then, is that the parent first establishes control by quiet, unruffled consistency, and this needs to be overwhelmingly secure control in the toddler and infant years.  Then from a position of strength, the parent begins to encourage negotiation, which allows the child to have an influence on successive revisions of the ground rules.  So the parent stays in control and acts as guarantor for the enforcement of the contract, but also participates in repeated negotiations leading to changes in the terms of the contract.

In these circumstances, the effect of parental control may seem overwhelming, but this may be misleading.  Parental control creates the conditions in which negotiation can occur.  Without parental control, negotiation will have negative effects; it will be merely a weakness to be exploited by the coercive child.  Given parental control, negotiation will have positive effects, but one of the most important of these effects will be to strengthen parental control itself!  The effect of negotiation may be difficult to observe because it is so closely bound up with parental control, and because strong parental control is a pre-condition of effective negotiation.

FINDINGS FROM THE EDINBURGH STUDY

We can take this analysis one step further by considering findings from our own study of young people, the Edinburgh Study of Youth Transitions and Crime.  In this case, the measures of parent/child relationships are based on the child’s reports alone, although detailed responses from parents will also be available soon.  The statistical base is much more robust for the Edin­burgh than for the Oregon results, however, since the study includes 4,300 young people who were aged 13 at the time when the results that I will quote were collected, in the autumn of 1999, the second year of the study.  Unlike the Oregon programme, the Edinburgh Study covers roughly equal numbers of girls and boys.

I will consider just three measures of parent/child relationships: parental supervision (similar to Patterson’s parental monitoring); child autonomy; and parent/child conflict.  To measure parental supervision, young people were asked: When you go out, how often do your parents know (1) where you are going, (2) who you are going out with, and (3) what time you will be home?  To measure the trust or autonomy afforded them, young people were asked how often their parents (1) let me do things I like doing, (2) trust me, (3) treat me like a baby, (4) try to control everything I do, (5) let me make my own decisions.  To assess the level of conflict between parent and child, young people were asked how often they disagree or argue with their parents about (1) homework, (2) friends, (3) how tidy your room is, (4) what time you get in, (5) what you do when you go out.  The responses to the three groups of questions were used to compute three scales measuring supervision, trust/autonomy, and conflict.  Self-reported delinquency was measured by a series of questions about 15 kinds of delinquent act: fare dodging, shoplifting, public rowdiness, riding in stolen vehicles, theft at school, carrying a weapon, vandalism, housebreaking, graffiti, robbery, theft at home, fire setting, injuring or fighting, breaking into cars, truancy.  A variety of delinquency score was com­puted for the number of kinds of delinquency that the young person had taken part in during the past 12 months.

Table 1: Correlation between delinquency and parent/child relationships: Edinburgh Study, sweep 2 (age 13) 

Note: all correlation coefficients are non-parametric (Spearman’s rho).  All are significant at better than the 99.9% level of confidence.

As can be seen from Table 1, there was a high correlation between parental supervision and self-reported delinquency (the negative sign indicates that high parental supervision is associated with low delinquency).  This parallels Patterson’s findings on the primordial importance of parental monitoring.  However, a point not covered by the Oregon study, there was also a high cor­relation between child/parent conflict and delinquency (this time the positive sign indicates that high conflict is associated with high delinquency).  There was also a rather weak correlation between trust/autonomy and delinquency (the negative sign means that high trust/autonomy is associated with low delinquency).

But what of the links between these different aspects of the parent/child relationship?  As expected, there is a strong correlation between high conflict and low trust.  At first sight, the other two relationships may be somewhat unex­pected.  Firm supervision must restrict the freedom of the child, which could be a cause of conflict.  There is in fact a fairly strong correlation between pa­rental supervision and parent/child conflict, but the negative sign indicates that high supervision is associated with low conflict.  Similarly, children who are closely supervised could feel that they are not trusted and not allowed to make their own decisions.  There is in fact a correlation between supervision and trust/autonomy, although it is not particularly strong; however, the positive sign indicates that strong supervision is associated with a sense of autonomy and being trusted, and not the reverse. 

Although they may seem unexpected, these findings from the Edinburgh Study strongly confirm the theoretical analysis that I have put forward.  They suggest that where there is strong supervision, there is also negotiation and accommodation.  Parents who monitor their children’s movements tend not to be in conflict with them, and tend to make them feel that they are trusted and allowed to make some of their own decisions.  That is because pa­rental control creates the conditions in which negotiation can occur.  It is the parents who are not in control of their children who are in conflict with them; and where parents are not in control, children do not feel that they are trusted and allowed to make their own decisions: quite the contrary.  On Pat­terson’s analysis, that is because it is aggressive and inconsistent parents who lose control, and arbitrary, authoritarian behaviour does not inspire trust or confidence.

Although high supervision is generally associated with low conflict, we can distinguish between young people having a high versus low level of conflict with their parents, and consider the effects of supervision on delinquency among these high and low conflict groups.  The results show that high par­ent/child conflict reduces the effect of supervision on delinquency.  Supervi­sion is most strongly related to delinquency among those who are not in con­flict with their parents.  This suggests that consensual, negotiated supervision is most effective.  Our findings show that this need for supervision to be negotiated is more clear-cut at the age of 13 among girls than among boys.

IMPLICATIONS FOR JUVENILE JUSTICE

I suggest that the idea of negotiated order could be developed into a guiding principle for the juvenile justice system.  This might be a better foundation than the nebulous concept of ‘the welfare of the child’.  One problem with the idea of welfare is that it leaves open the question of who should decide where the child’s interests lie.  This problem is carried over into the actual operation of the children’s hearing system in Scotland.  Although there is some recognition that the child’s views should be heard, the welfare principle implies that well-meaning volunteers will probably know best, an essentially patronising assumption.  Explicit recognition that solutions are to be negotiated would be far preferable.

A related problem with the welfare principle is that it sweeps the issue of control under the carpet.  For the majority of children who come before the hearings, the clearest need is for their lives to be set within a clear frame­work so that authoritative adults, subject to stringent accountability, regain control over them, and help them to develop self control.  As I have argued earlier, negotiation can only take place within an ordered framework of that kind.  Unfortunately, juvenile justice systems tend to share most of the characteristics of the ineffective parent: they do not notice bad behaviour until much too late; they pay too much attention to trivial matters; they respond inconsistently, and fail to back up their requests; and they are subject to sud­den bouts of arbitrary severity.  I suggest that the principle of negotiated or­der would form a good foundation for a campaign to tackle these weaknesses.