Introduction

This file summarises a number of pieces of work, and connects to some Reading Guides on some background texts. Labelling theory was thought to indicate a radical new departure in theories of crime and deviancy, and it became very popular in the late 1960s and early 1970s in sociology generally -- there were even some spin-offs in the sociology of education (see file). Advocates made two basic claims for the superiority of this approach: 

The approach enabled us to generalise about deviancy in general, and to move away from the rather unsatisfactory category of  'crime'. Any kind of deviancy could be studied, including physical 'handicaps' or various stigmas. Deviancy became a sociological phenomenon in its own right, enabling us to break with the old views and definitions expressed in penology and social policy: we could finally 'make' rather than 'take' problems as a famous remark put it. This might have a critical impact on opening discussions in policy, which had been rather limited, it was hoped. The approach brought a much greater range of interest in social activities including semi- and non-criminal ones. 

The approach was based on the emergence of a new paradigm or perspective in sociology -- interactionism, either symbolic interactionism or particular versions of it. This perspective emphasises the collective rather than the individual nature of social action, advocated a study of interaction rather than causal studies, and focused on social processes including reactions and counteractions. For various reasons, the issues came to be focused on how 'labels' were developed, negotiated, exchanged and applied to people. Such a perspective was often critical, taking the side of the underdog, and criticising official agencies and their labelling work. 

Societal reaction theories. These are well described in pieces like those of Wilkins or Stan Cohen, both to be found in Carsons and Wiles  (1971). The basic argument is that in all social behaviours, the reactions of other people become important variables in the ongoing process of action. As an example, crashes on the Stock Exchange produce panic as a reaction, and this panic leads to more selling, which amplifies the original panic. It is the same with juvenile delinquency, although this is ignored in classic approaches: thus in the work of Albert Cohen, on criminal subcultures  (see file) the negative and malicious characteristics of juvenile delinquent acts, and the activities of juvenile gangs, are effective in producing a strong societal reaction, which classifies members as abnormal, helps the authorities denounce them as intolerable, which then leads to the denial of legitimate opportunities to juveniles, which then encourages them to turn back to deviant ones -- and so on around the cycle. 

Societal reactions are therefore very important, and we need more research on the strength and type of reactions, for example, on the degree of tolerance in our society. For Stan Cohen, this led to an interest in the activities of prominent agents of social reaction, especially in the media -- and he went on to coin the term  'moral panic' that was to be much developed later  (see file for a marxist version). His famous piece on how the media amplified the mod and rocker  'riots' actually can seem rather odd, with its undertones of crowd psychology -- but the main points are clear: the media stereotyped and polarised the contending groups  (not only mods and rockers, but young people and the police) and increased the sensitisation of the participants, encouraging further delinquent behaviour. This is a famous analysis, but still rather a speculative one which ends with a an important note of caution -- we still do not know very much about the effects of labels on actors, although there are reasons to believe these are complex ( see below). 

Social reactions have been studied on the micro scale, by examining the interactions between concrete participants, and this leads us to the classic work by Becker. The famous opening section in Outsiders ( Becker 1963), argues that since social groups create rules, they also create deviants or outsiders: no rules, no deviancy. Deviancy is thus not the quality of the act itself, but of the social process of recognition and rule enforcement. More famous implications follow -- it directs our attention to rule-makers and enforcers as much as to deviants, hence the classic piece on  'moral entrepreneurs' [ see file for a reading guide on this chapter] in Outsiders; it shows morality is an enterprise, not just a simple natural social process; it directs our attention to the values of labellers, and this nearly always involves a critique of them, since these values can no longer be seen as natural or neutral: in this sense, the approach takes the side of the underdog, Becker was to insist. On a more theoretical level, the approach can be defended using classic interactionist arguments -- we are demystifying strange concepts such as  'drives', denying the simple objectivity of acts, and drawing attention to collective processes. 

Becker is the usual source of radical variants of labelling. His work implies there is no need to explain deviance in the first place, that it is in fact a very common social activity, a normal one, which only becomes abnormal when it is to so labelled. Labelling itself then  becomes confirming, a self-fulfilling prophecy,  launching people on a deviant career. 

Becker's work is the usual target for criticisms too. These have been summarized very well in a number of pieces, including a sympathetic chapter by Plummer in Downes and Rock (1979): 

1. Labelling theory neglects the process of becoming deviant in the first place, the act of  'primary deviance'. It is not very good at describing important part of deviancy, such as murder  [often a classic example of primary deviance] It has particular problems in explaining certain kinds of sexual deviancy  (see Gove), including paedophilia, which does indeed seem  'psychological'-- many of the authors in Gove, not to mention a lot of public opinion, thinks that quite a bit of deviancy is driven by abnormal psychology. However, these criticisms still ignore the process of labelling and reaction, which are important and cannot be neglected even in these cases. And labelling clearly is very important in certain grey areas, such as deviancy in schools, where the rules to be broken are very variable and often unclear.  [This has led to some very interesting work on how schools can 'insulate' themselves against high rates of deviancy by sticking to a few clear-cut rules, and offering a  'truce' over a range of other behaviours such as wearing plastic jewelry, smoking in the cycle sheds, and using slang or patois in the corridor -- I have personally seen teachers reduced to apoplectic fury over these matters]. 
2. Labelling theory overdoes the idea of a self-fulfilling prophecy or a career -- there is no real empirical evidence for this, says Gove, and we know that most official agencies, including courts, are pretty ineffective. Later work draws attention to aspects of labelling neglected in Becker, especially what makes a label effective and permanent, how it becomes a master identity. A study of classroom labelling  (Hargreaves, Hester and Mellor 1975) shows lots of variation -- whether a label sticks, for example, depends very much on whether it agrees with the other labels being stuck on by a person's 'significant others'. 
3. Labelling theory neglects the power relations or structures in society. This is the usual criticism launched by marxists, whether in Critical Criminology, or in Policing The Crisis. In fact, it is not clear that Becker does simply ignore power differentials -- indeed, he sees labelling as aspect of symbolic power [ see his own reply to his critics]. Perhaps it is that the context of the symbolic power is unexplained, that it is not traced to a general theory of power in society -- but this gets close to saying that the problem with interaction is that it is not marxism. Not everything should be marxism, perhaps -- interactionists don't ignore relations of power, but just disagree about how the should be defined, and where they operate. 
4. Labelling theory is untestable by empirical standards -- and Gove again has some examples. Its main mechanisms are persuasive, but attempts to test them have been controversial, especially in education, in fact  (see Rist in Hammersley 1986 for a review -- and see file). Of course, empirical studies are also controversial, especially in the definitions they use. Plummer argues that the more complex variants of labelling theory are almost impossible to test or falsify, especially if labels can now be rejected or reversed. Sophistication may have been gained, but explanatory simplicity and power lost. 
5. Finally, there have been criticisms from a phenomenological or ethnomethodological direction, as in the piece by Philipson and Roche  (in Carsons and Wiles 1971 volume 2), or in the classic by Cicourel (1964). The terms in labelling theory lack precision, and there is no real accounts of central social processes, such as how every day behaviour actually needs to a societal reaction. Methodology generally is seen as lacking clarity, and incorporating too many commonsense definitions and assumptions. Instead, we need more detailed studies, says Cicourel (1968), of matters like police procedures, or the categories deployed and applied by social workers and lawyers -- in the case of police behaviour, for example, it is clear that much depends on the appearance, demeanor, or attitudes of potential arrestees, and that very different treatment can be advanced to suspects, depending on the collective and immediately-formed social judgements of the policemen. Similar work has been done, of course in classrooms  -- see file for some examples.

Carsons W and Wiles P (eds) (1971) The Sociology of Crime and Deviancy in Britain, vols 1 and 2, Oxford: Martin Robertson 
Cicourel A ( 1964) Method and Measurement in Sociology, New York
Cicourel A ( 1969) The Social Organisation of Juvenile Justice, New York
Downes D and Rock P (eds) (1979) Deviant Interpretations: problems in criminological theory, Oxford: Oxford University Press
Gove W ( ed) The Labeling of Deviance
Hargreaves D, Hestor S and Mellor F (1975) Deviance in Classrooms, London and Boston: Routledge and Kegan Paul
Hammersley M (ed) (1986) Case-Studies in Classroom Research, Milton Keynes: Open University Press