What Deviance and Control Reveal About Order

This Part has been about rule-breaking. It has really been about rule-following — because you cannot study deviance without discovering how much conformity there is, how it is produced, and how little of it is produced by the machinery everyone assumes holds society together.

The deepest finding of the whole field is a reversal. We ask why people break rules, and the study of that question keeps returning the harder and more important one: given how easy and often rewarding rule-breaking is, why do people conform as much as they do? This lesson draws the Part together around that question — which is the problem of social order from Part 1 (see 1.2.1), now answerable with evidence.

The night the lights went out, twice.

Take two power blackouts in the same city, a decade apart.

In the first, the city went dark for many hours on a hot summer night. And a great deal of it stayed orderly — people directed traffic, checked on neighbours, shared what they had. But in some areas there was widespread looting, arson and disorder, thousands of arrests, and lasting damage. The blackout is remembered partly for the breakdown.

In the second, a larger blackout hit the same city, and much of the surrounding region, also for many hours. And this time the overwhelming story was order — people again directing traffic and helping strangers, bars serving warm drinks by candlelight, and very little of the disorder of the first.

Same city. Same removal of electric light, of traffic signals, of much of the visible apparatus of control. Two very different nights.

Now the question the contrast forces. When the lights go out, the police cannot see, the cameras stop, the alarms fail, and the ordinary deterrents — certainty of detection (see 10.3.1) — collapse. On the theory that order is produced by external control, both nights should have been chaos.

Most of both nights was not chaos. Most people, in the dark, unwatched, with the machinery of control switched off, continued to behave much as they had with the lights on. They did not loot the shops they could have looted. They helped rather than robbed. The external control was gone and the order largely remained — which tells you that the order was never mainly produced by the external control.

And the difference between the two nights — why one broke down more than the other — is itself the sociological finding: the disorder tracked not the darkness but the underlying conditions of the areas where it happened — concentrated disadvantage, weak ties to conventional institutions, a history of hostility between residents and police (see 8.6.2, 10.2.3, and the legitimacy finding of 10.3.1). Where the bond to the social order was strong, the dark changed little; where it was already frayed, the dark let it tear.

Which is the whole Part in one image. Order is held, overwhelmingly, by something other than the police — and where that something is intact, the police are barely needed; where it has broken down, the police cannot supply it.

Why do people conform, when they could so easily not?

This is the problem of order (see 1.2.1), and the study of deviance answers it better than the study of order does — because you learn what holds a rule in place by studying what happens when it breaks, who breaks it, and what the breaking reveals.

Four answers have appeared across this Part, and they are not rivals but layers.

External control — the police, courts and prison, deterring by the certainty of detection. Real, and the smallest of the four. The blackout shows how little it holds; the crime evidence shows it works through certainty not severity (see 10.3.1) and cannot manufacture order where the other layers have failed.

The bond to conventional society — attachment, commitment, involvement, belief (see 10.1.2, control theory). People conform because they have something to lose — relationships they value, futures they are invested in, activities that fill their time, and a belief that the rules are legitimate. This is a larger part of the answer, and it is why the weakening of the bond, not the presence of a temptation, predicts deviance.

Internalisation — the rules taken inside, so that conformity is not a calculation but a disposition (see 2.3, 10.3.1, 5.8.1). Most people do not refrain from theft because they weigh the risk; they refrain because it does not occur to them, or because it would make them a person they do not want to be. This is the largest part, and the most invisible — which is why order looks so fragile in theory and is so robust in practice.

And the legitimacy of the order itself — people obey rules they experience as fairly made and fairly enforced (see 9.7.1, 10.3.1), and withdraw obedience from rules and authorities they experience as unjust. This is why the two blackouts differed: legitimacy had been withdrawn in some areas and retained in others, and the dark revealed which.

The Part's answer to the problem of order, then: society is held together far more by internalisation and the social bond and legitimacy than by external control — which is why it survives the lights going out, and why no amount of policing can hold an order whose legitimacy has collapsed.

What the Part established

Six findings, drawn together.

One — deviance is a property of the response (see 10.1.1). The same act is deviant or not depending on who did it, to whom, in what role, and how others reacted. This reoriented the field from the offender to the audience — to the rule-makers and enforcers who produce deviance — and it is why official rates are rates of processing, not of behaviour (see 7.4.4, 10.2.1).

Two — deviance is normal and structurally produced (see 10.1.2). It exists in every society, is generated by ordinary arrangements (the gap between goals and means), and is best predicted by the weakness of the bond to conventional life. It is not a symptom of individual pathology or social breakdown; it is a permanent feature, patterned by structure.

Three — the response shapes the deviant (see 10.1.3). Labelling can produce secondary deviance through closed doors and reorganised identity; the machinery that processes deviance is partly a cause of it. The push toward the first act and the pull into the career are both patterned by the inequalities of Part 8.

Four — the crime statistics measure the machine, not the behaviour (see 10.2.1). Five gates, each operated with discretion and incentive, filter by seriousness, relationship, trust and social position. A change in the figures is always ambiguous between the world and the machine.

Five — the field cannot predict and should be humble about explaining (see 10.2.2). The crime drop was confidently mis-forecast and is still incompletely explained; the confident explanations are mostly national accounts of an international event. This is the discipline's own limits, honestly stated.

And six — control and fear track the response, not the threat (see 10.2.3, 10.3.1, 10.3.2). Who is punished tracks inequality more than offending; discipline works by internalised visibility; and fear rose while crime fell. The whole apparatus of control is, in large part, a response to something other than the danger it names.

Two errors the Part exists to prevent, one on each side.

The control error: believing that order is produced by external control, and that more police, harsher sentences and more surveillance produce more order. The evidence is against it — the blackout, the certainty-not-severity finding, mass incarceration failing to track crime, and the fact that order survives the removal of control where the bond is intact. Order is mostly not made by the machinery that claims to make it, and treating deviance as a problem to be solved by more control mistakes the smallest layer for the whole.

The relativism error: concluding, from "deviance is a property of the response", that nothing is really wrong and all rules are equally arbitrary. The Part has been careful to reject this (see 10.1.1). The murder is real; the harm is real; the victimisation is real. What is socially produced is the category and its processing, not the underlying acts — and a society does have to decide what to prohibit and how to respond, which the analysis informs rather than dissolves.

The competent position holds both against the errors: order is real and mostly not made by control; deviance is real and mostly a property of the response; and the two are consistent, because the response is one of the things that produces both the order and the deviance.

And the reflexive point, which this Part must make about itself.

A sociology of deviance is itself a form of control, or can be — because to study, classify and explain deviance is to participate in the machinery that produces it (see 9.5.2 on the same reflexive problem for medicine, 6.8.1 on the course's own complicity).

The field has been most useful when it studied the labellers as well as the labelled, the powerful deviance as well as the powerless, the response as well as the act. It has been least useful — and most complicit — when it took the state's categories as given, studied only the deviance of the poor, and supplied the "socially accredited experts" who pronounce diagnoses in a moral panic (see 10.3.2).

Which is the discipline of this whole course, applied to its own subject. The sociologist of deviance is not outside the system of deviance and control; they are a potential part of it, and the honest response is to study the system including one's own place in it — not to imagine oneself a neutral observer of a machinery one is helping to run.

Because the way a society thinks about deviance and control determines what it does to the people at the bottom of it — and this Part has argued that the common-sense way of thinking is wrong in specific, consequential directions.

The common sense: deviance is bad acts done by bad people; crime is rising; order is produced by control; more control produces more order; and the fear of crime is a reasonable response to danger.

What this Part established against each: deviance is a property of the response as much as the act; crime has been falling; order is produced overwhelmingly by internalisation, bond and legitimacy rather than control; more control does not reliably produce more order and sometimes produces more deviance; and fear is decoupled from danger.

Every one of those corrections points the same way — toward a smaller role for the punitive apparatus and a larger role for the conditions that produce the bond and the legitimacy: the inclusion, the stake in conventional life, the fair treatment, the concentrated disadvantage undone (see 8.8.1, 9.7.1). This is not softness; it is what the evidence supports. The society that wants less deviance is, on this Part's evidence, better served by building the bond than by expanding the machinery of control — because the bond is what holds the order, and the machinery mostly cannot.

Three questions to carry out of the Part.

Is this a property of the act or of the response? For any deviance, ask how much is the behaviour and how much the reaction — and remember that the reaction is patterned by power and position.

Is this order made by control or by the bond? Where order holds, ask what is actually holding it — usually not the police. Where it fails, ask what bond has broken — usually before the control arrived.

And does the fear track the danger? For any alarm, check the fear against the risk. The gap is where control expands beyond what the threat warrants.

One closing observation, which returns the Part to where the course began. The problem of order (see 1.2.1) asked how society is possible at all, given that it requires people to restrain themselves. This Part's answer is that they mostly do restrain themselves, not because they are watched but because the order is inside them — and that the visible apparatus of control is a small, late, and partly counterproductive addition to a conformity that was already there. The lights went out and most people stayed honest in the dark. That is the most hopeful finding in the Part and the most important: order is not a thin crust over chaos held down by force. It is deep, distributed, and mostly self-maintaining — which is why a society is far stronger than its police, and far more dependent than it knows on the bonds and the legitimacy that the police cannot supply.

The study of rule-breaking is really the study of rule-following, and the field's deepest question is the reversal: given how easy and rewarding deviance often is, why do people conform so much? — the problem of order from Part 1, now answerable.

Two blackouts in one city: one with widespread disorder, one with overwhelming order, both with the apparatus of control switched off. Most of both nights stayed orderly in the dark, showing order was never mainly produced by external control — and the difference between them tracked the underlying conditions and the legitimacy of authority in the affected areas, not the darkness.

Four layers of conformity, not rivals: external control (real, smallest, works by certainty not severity, cannot manufacture order); the social bond (attachment, commitment, involvement, belief — something to lose); internalisation (the largest and most invisible — rules taken inside so conformity is a disposition, not a calculation); and legitimacy (obedience to rules experienced as fair). Society is held far more by the last three than by the first.

Six findings drawn together: deviance is a property of the response; deviance is normal and structurally produced; the response shapes the deviant; crime statistics measure the machine not the behaviour; the field cannot predict and should be humble about explaining; and control and fear track the response, not the threat.

Two errors: the control error (order is made by control — refuted by the blackout, certainty-not-severity, and incarceration failing to track crime) and the relativism error (nothing is really wrong — refuted by the reality of harm; what is produced is the category and its processing, not the acts). And the reflexive point: a sociology of deviance is itself potentially a form of control, most useful when it studies the labellers and the powerful, most complicit when it takes the state's categories as given.

Every correction points the same way: toward a smaller role for the punitive apparatus and a larger one for the conditions that build the bond and the legitimacy. The lights went out and most people stayed honest in the dark — order is deep, distributed and mostly self-maintaining, which is why a society is far stronger than its police and far more dependent than it knows on what the police cannot supply.

The problem of order — how self-restraint is produced at all; the question deviance answers better than order does.

External control — deterrence by detection; the smallest layer of conformity.

The social bond — attachment, commitment, involvement and belief; conformity from having something to lose.

Internalisation — rules taken inside, making conformity a disposition rather than a calculation; the largest layer.

Legitimacy — compliance with rules experienced as fairly made and enforced; withdrawn from those experienced as unjust.

Rate of processing — an official rate as an output of the control machinery, not a measure of behaviour.

The control error — the false belief that order is produced by, and increased by, external control.

The relativism error — the false conclusion that the reality of harm dissolves because the category is socially produced.

Reflexivity of the field — the sociologist of deviance as a potential participant in the machinery of control.

Self-maintaining order — the finding that conformity is deep and distributed and survives the removal of external control.

One — run the blackout. Ask what would actually happen where you live if the machinery of control switched off for a night. Then ask what would hold the order — and whether it is the police.

Two — locate your own conformity. Take one rule you keep and ask which layer holds it — deterrence, bond, internalisation, or legitimacy. For most rules it is the last three.

Three — separate act from response. Take one deviance and estimate how much is the behaviour and how much the reaction. The reaction is patterned by power.

Four — check the direction. Take one proposed response to a deviance and ask whether it builds the bond or expands the control. The evidence favours the first.

Five — apply the reflexive test. Take one study of deviance and ask whether it studied the labellers as well as the labelled. If only the labelled, it may be part of the machinery it describes.