Abstract
Despite significant achievements in empirical research, considerable unease exists about the lack of conceptual and theoretical debate within the sociology of work. One potentially significant problem is the uncritical use of concepts that have their origins in Marxism and purport to explain the essential features of the employment relationship. Using evidence from a systematic review of four highly ranked British journals I chart the growing influence of the concept of contradiction, notably within the labour process perspective where it has become a key concept, especially in relation to the problem of labour control.
In spite of its popularity, I argue that the concept contains two sets of flaws. The first set, which relate to its utility as a concept, include problems of logic, differentiation and operationalization. The second set relate to the substantive use of the concept, especially its dependence on supporting assumptions, and its expectation of social change. The article concludes by calling for a moratorium on further usage.
Keywords
Introduction
More than half a century ago, Robert Merton observed that a significant strand of sociological theorizing was ‘… taken up with the clarification of concepts – and rightly so’ because clearly defined concepts were indispensable for empirical inquiry (Merton, 1958: 114). This was primarily because concepts provided a definition of what was to be observed and therefore played a crucial role in linking sociological theory to empirical research. Unlike the natural sciences, however, concepts in the social sciences do not always have the kind of secure and shared meaning that makes the perpetual quest for generalization somewhat easier. Instead, ambiguity, confusion and debates about mis-specification are common even in relation to such celebrated concepts as social capital (Portes, 1998). The danger is that any concept (or set of concepts) remains so vague that it cannot be used to identify appropriate observations, with the result that the related theory cannot be refuted. It is therefore something of a truism to state that sociology, as a social science, cannot hope to generate an accumulating body of knowledge if it fails to keep the meaning and use of its concepts under scrutiny. After all, the formation and interpretation of concepts raises one of the most basic questions in social science: what exactly are we talking about?
In this article I shall argue that a striking example of this failing can be found in the sociology of work, especially as it is practised in Britain. One of the more significant problems, I suggest, is the persistent and uncritical use of theoretical concepts that have their origins in Marxism and purport to explain much of what is sociologically significant within the employment relationship. A striking example is the concept of contradiction, which was originally developed within Marxist perspectives on industrial relations and the capitalist labour process before subsequently enjoying a remarkable surge in popularity, partly because of the emergence of the Labour Process perspective. Yet the relative success of Labour Process Theory can hardly account for the extraordinary surge in usage over the past three decades, especially as this has been a period in which Marxism has been in retreat as both an intellectual and a political force. Whatever the reasons behind its meteoric rise, it is remarkable that such a prominent concept has never been subjected to any kind of scholarly critique.
In the next section I briefly set out the Marxist origins of the term before presenting a systematic review of its emergence and usage within four leading British journals. Having established the significance of the concept and considered some of the main ways in which it is being used, I shall focus on two sets of flaws. The first set, which relate to its utility as a concept, are problems of internal logic, differentiation and operationalization. The second set relate to the substantive use of the concept, especially its dependence on supporting assumptions, and its expectation of social change.
Origins and Meaning
The concept of contradiction has both an analytical and a political meaning within classical Marxism. Analytically, contradiction was a central component in Marx’s dialectical analysis of capitalism. One of the defining characteristics of dialectical reasoning is the assumption that conflict and struggle are not temporary or superficial features of capitalism but are an intrinsic underlying reality. Capitalism is a ‘contradictory’ system characterized by the interpenetration of opposites, or dualisms, with the result that any social structure is imbued with an essential negativity. For instance, Marxist theory insists that the prevailing form of economic organization acts as a constraint on the development of the productive forces within society. This well-known contradiction between the relations of production (ownership, control, etc.) and the forces of production (productive potential) is, in turn, compounded by the deep-seated and enduring struggle between the two major social classes: the capitalist and the proletariat. The class struggle is itself exacerbated by the growing contradiction between the increasing socialization of production and the persistence of private appropriation. As capitalism develops into a universal social power, the contradiction between it and the private power of the capitalist to own and control the social conditions of production will inevitably lead to the dissolution of private property relations.
These contradictions lead to chronic instability and regular crises. Within the ‘crisis of overproduction’, for example, each individual capitalist finds that when they strive to maximize profits by securing more labour this leads to lower profits for the capitalist class as a whole since they are forced to bid up wages. What must also be emphasized is not just that capitalism is prone to contradictions of this kind but that the concepts of contradiction and class conflict provide the mechanisms by which social change occurs across history. In this context, it must be remembered that the classic Marxist notion of contradiction is essentially that of a fatal flaw or fundamental weakness. Capitalism is a contradictory system that generates the seeds of its own destruction from within. In Marx’s memorable words: ‘what the bourgeoisie therefore produces, above all, are its own gravediggers’ (Marx and Engels, 1992[1848]: 18). Contradiction therefore implies change rather than merely the presence of strain, tension or conflict. 1
Symbolically, contradiction is but one of a number of critically oriented concepts, such as alienation and exploitation, which aim to raise consciousness about, if not actually encourage, the belief that capitalism contains moments or crises that may (be used to) lead to its destruction. In other words, they have an obvious political purpose. Although an attempt is made to explain social phenomena, this is ultimately subordinate to the aim of justifying a normative position that is directed towards the goal of human emancipation.
In Search of Contradictions
In contrast to Marx, who used the notion of contradiction at a highly abstract level to refer to the conflict between the forces and relations of production, scholars in industrial relations and the sociology of work and organizations deploy the term at a more concrete level to capture what they perceive to be the underlying antagonisms that shape the management of the employment relationship. Such applications have proven to be so influential that they have been introduced to generations of students through textbooks in industrial relations and the sociology of work (e.g. Edwards and Wajcman, 2005; Hyman, 1975; Noon and Blyton, 2007; Watson, 2008). Textbook appearances do not, however, indicate the prevalence of the term, the ways in which it is used, or the kind of research in which it appears.
Research Methods
Searching for the term across the voluminous literature on the sociology of work and organizations, as well as the related literature on industrial relations and human resource management, would be an enormously time-consuming task given the unknown number of books that would need to be included. Also, a search of this kind would run the risk of including books and papers that may have relatively little influence or might be viewed as being of low quality. For these and other reasons, I decided to restrict my search to articles published in four leading academic journals, namely, the British Journal of Industrial Relations (BJIR), Organization Studies (OS), Sociology and Work, Employment and Society (WES). The selection of journals was based on four criteria. The first is obviously that each of the journals has a reputation for publishing papers on the sociology of work (or at least papers with a sociological orientation relating to the world of work). Second, all are considered to be leading journals and, as such, might reasonably be expected to exert some influence over their respective sub-fields. Aside from Sociology, each journal obtained the highest rank in the most recent set of journal rankings published by the Association of Business Schools (Kelly et al., 2009). Sociology, which is the flagship journal of the British Sociological Association and a leading general journal, was included because it has the strongest reputation among the general journals for publishing papers relating to the sociology of work. Third, the journals were selected to represent the subject areas of industrial relations, the sociology of work and organizational studies, in order to demonstrate the influence of the term across cognate areas. Finally, these journals were chosen because they all cover the period since the specialist journal Work, Employment & Society was launched in 1987 and the opportunities for publication in the sociology of work increased.
The search for contradictions within these journals was also based on a conservative strategy that was deliberately designed to strengthen claims of validity. For instance, everyday or conventional language usage was excluded, as well as claims of ‘apparent’, ‘potential’ or ‘seeming’ contradictions. Similarly, contradictions in discourse or identity, which appeared in writing influenced by post-structuralism, were also omitted. Instead, the emphasis was on searching for social contradictions that referred to contradictions that were built into the social order and manifest in capitalism, capitalist firms, and managerial strategies.
Furthermore, papers were only selected if the concept was used repeatedly and in a meaningful fashion. Papers that merely had a passing reference to contradictions of any kind were not considered. Instead, papers had to include at least three references to such contradictions (or contradictory phenomena) and, in many cases, use of the term in either the abstract or concluding section to denote its significance within the paper. The extensive search activity was greatly helped by the use of electronic search tools that accompany most Portable Document Format (PDF) files that journals have been using since the late 1990s.
Excerpts from those papers describing contradictions were coded and analysed using the NVIVO software package. The initial list of codes (‘Nodes’ in NVIVO) was revised and extended to capture the diverse ways in which the concept was used. Other details, including definitions and source references, common phrases relating to contradictions, as well as the year of publication, title of journal and type of article (e.g. empirical, theoretical or commentary) were also included. The excerpts from the articles were first copied and then coded onto text documents before being checked and entered into NVIVO.
Contradictions at Work
In the 25-year period since 1987, I identified a total of 63 papers that made substantial use of the conception of contradiction using the search criteria outlined above (WES 27, OS 14, Sociology 12 and BJIR 10). 2 Significantly, papers using the term on at least three occasions became more prevalent over time. In the 1990s for instance, there were a total of 17 but by the 2000s this had doubled to 34 papers. When divided into five-year periods starting in 1987 there was a decline from 10 to seven papers between the first and second periods, before rising to 22 papers in the fourth period and falling back to 16 papers between 2007 and 2011. In other words, use of the term peaked during the mid-2000s but continued strongly through to the end of the period. Given the impressive rate of diffusion, it is difficult to understand why the concept is not deemed worthy of any discussion in expositions on ‘core labour process theory’ (Jaros, 2010; Thompson, 1990), even though it is probably that perspective’s single most influential concept.
Four major types of contradiction emerge from the search. These relate to contradictions of labour control, the employment relationship, capitalist firms and capitalism generally. Those familiar with classical Marxism will know of the contradictions within capitalism, notably between the forces and relations of production. However, what is striking about the literature surveyed here is the emergence of new notions such as those of contradictions within capitalist firms, within the employment relationship and, in particular, in relation to the control of labour. Contradictions of labour control were by some margin the most widely discussed, having been identified on 102 occasions across 38 articles (Table 1). The most well-known paper here is, of course, Richard Hyman’s (1987) elaboration on the contractions facing the management of capitalist enterprises which was published in the first issue of Work, Employment & Society and would subsequently go on to become one the most cited papers ever published in that journal. 3 Other influential papers include those by Knights and Willmott (1989) and Armstrong (1989).
Forms of contradictions listed in elite journal articles, 1987–2011.
Notes: Articles (N = 62) may contain more than one form of contradiction
The publication of Hyman’s essay coincided with the emergence of a generation of sociologically oriented critical management writers who have applied the concept of contradiction to a range of new management rubrics and to work in the growing service economy. Consequently, two further sub-types can be identified within the contradictions of labour control: the contradiction between control and consent and those relating to new management rubrics. Examples of the latter include studies of Total Quality Management (TQM) (e.g. Knights and McCabe, 1999, 2000), team-based forms of work (Danford, 1998) and the introduction of New Public Management in the public sector (e.g. Cooke, 2006; Foster and Hoggett, 1999). 4
The contradictions within capitalist firms include two further types. The first is where initiatives that seek to provide employees with more discretion and autonomy, possibly through new ‘empowering’ forms of work, fail to meet expectations because they are contradicted by a managerial preoccupation with cost reduction, labour intensification and short-term profitability. For instance, Knights and McCabe claim that their ‘case study illustrates a number of tensions and contradictions surrounding the implementation of a TQM programme, not least of which is the preoccupation with cost constraints and short term profitability’ (1998: 436). Also, an influential paper by Bolton and Boyd on emotional labour reports that: … efforts to recruit, train and socialize workers to deliver ‘sincere performances’ will be undoubtedly constrained by the range of contradictions created by those cost-cutting strategies which undermine the quality of cabin crews’ working conditions and, ultimately, their physical health and well-being. (2003: 301)
The second contradiction, which reflects the rise of the ‘service economy’, is that between ‘quality and quantity’. This arises in front-line service work where employees have to respond directly to customer needs. In this case, the emphasis that management place on employees being customer oriented is contradicted by the demand that they should deal with large numbers of customers regardless of the quality of customer care (e.g. Darr, 2011; Korczynski et al., 2000; Taylor and Bain, 2005).
Although a relatively small proportion of the papers refer to these contradictions (which is why they are not listed in Table 1), the most substantial recent discussion of the contradictions of capitalism and their manifestation in capitalist firms can be found in the debate surrounding Adler’s Marxist critique of Labour Process Theory in Organization Studies (Adler, 2007; Delbridge, 2007; Knights and Willmott, 2007; Vallas, 2007). I return to Adler’s conception of contradiction later. Meanwhile, I propose to concentrate on the contradictions of labour control, as these were by far the most prevalent across the four journals.
Contradictions of Labour Control
The journal search identified two basic ideas relating to the problem of labour control. The first, which was a central theme in Friedman’s celebrated historical study of managerial strategies of labour control, claims that any strategy of labour control is inherently contradictory, or unstable, when undertaken within a class-divided society (Friedman, 1977; see also Littler and Salaman, 1982: 264; Storey, 1985: 197; Vallas, 2007: 1383). For Friedman, the contradiction exists because workers are not machines and have independent and often hostile wills that cannot be destroyed. However, this does not mean that employers find it impossible to implement any labour control strategy. Rather, Friedman argues that a contradiction of this kind may ‘be suppressed, or disguised or bypassed’ by new managerial strategies, but its continued existence will inevitably lead to the re-emergence of this basic contradiction and further attempts at suppression or disguise (1977: 106). Friedman elaborated on this point by claiming that neither of the two labour control strategies he identified as being the most prominent historically, ‘Direct Control’ and ‘Responsible Autonomy’, were able to resolve this problem. In fact, he argued dialectically to the effect that the more these two strategies were implemented, the more they were likely to expose the contradictions that lay underneath.
The second and most well-known contradiction is the supposed perpetual tension between the need to control labour while simultaneously eliciting its cooperation. In an influential neo-Marxist paper on industrial democracy and the control of labour, Cressey and MacInnes (1980) argued that the inherent dualism of conflict and cooperation in capitalist forms of work creates a two-fold relationship between employers and employees. Employers have to control their employees yet they must also, to some degree, seek a cooperative relationship in order to tap their creativity. At the same time, employees have an interest in resisting their own subordination while also needing to cooperate with employers for their livelihoods. Contradictions arise because the employer knows that relinquishing control to the workforce can enhance its bargaining power, especially when the workers view cooperation as an opportunity to enhance wages and conditions.
Hyman’s celebrated elaboration insisted that: … the function of labour control involves both the direction, surveillance and discipline of subordinates whose enthusiastic commitment to corporate objectives cannot be taken for granted; and the mobilization of the discretion, initiative and diligence which coercive supervision, far from guaranteeing, is likely to destroy. (1987: 41)
In addition, he emphasized the dynamic nature of this inherent contradiction. New fashions in management which sought to address problems of discipline inevitably aggravated the problem of consent and vice versa. So, what do these and other contradictions mean? The answer for Hyman is the much quoted assertion that the likelihood of any strategy being successful is undermined by the inherent contradictions within capitalist enterprises that create a fruitless search for management panaceas: ‘… there is no “one best way” of managing these contradictions, only different routes to partial failure’ (Hyman, 1987: 30).
Contradiction as a Concept
It should be acknowledged that there are schools of thought in sociology which consider debates about the specification of individual concepts to be unimportant. Concepts should not be reduced to either observational statements or a set of operations that yield a measurable set of indicators. The implication, as Outhwaite notes, is that: ‘what counts are the theoretical structures in which these concepts are combined’ (1982: 3). To put it another way, any attempt to evaluate a concept or set of concepts has to evaluate the theory to which they contribute. Or, more crudely, concepts are only as good as the theories they serve.
In the case of contradiction there is, of course, such an extensive literature on the failings of Marxism as social theory that it need not be rehearsed here (e.g. Kolakowski, 1978; Popper, 1966[1945]; Schumpeter, 1943). To be fair, some of the leading proponents of Labour Process Theory have long since abandoned many of the central tenets of Marxism, such as those emphasizing the role of class struggle in the transformation of capitalist society (e.g. Edwards, 1986; Thompson, 1990). In any case, one of the striking aspects of the articles selected from across the four leading journals is that they frequently lack any reference to a conceptual framework or theoretical perspective of any kind (only half of the papers identified with a theoretical perspective and only one-third claimed to offer a Marxist or Labour Process perspective). In other words, it would be exceedingly difficult to evaluate the concept as part of a general theoretical framework when the latter is often absent.
Even so, I argue that it is possible to have a fruitful examination of a concept without having to engage in a detailed appraisal of the related theory. My basic orientation is therefore one of conceptual pragmatism in which the focus is on understanding how concepts are used by their originators and their followers for the purposes of empirical research (see also Mouzelis, 1995: 8–9).
Logic and Conceptual Precision
According to the Oxford English Dictionary, contradiction usually means ‘a combination of statements, ideas, or features which are opposed to one another’. It would not therefore be surprising if those who are not familiar with Marxism or the literature on the labour process should assume that the term has a basis in logic. Indeed, in classical logic, a contradiction consists of a logical incompatibility between two or more propositions. Aristotle’s law of non-contradiction, for instance, states that ‘the most certain of all basic principles is that contradictory propositions are not true simultaneously’ (Aristotle, 1976: 1011b, 13–14). Of course, this seems obvious. Either the sun is or is not shining but it cannot do both at the same time and in the same place, as Popper noted in his long-forgotten critique of dialectics (1940). If, as Popper argued, contradiction really means contradiction then a deductive science is not possible because contradictory propositions do not convey any sort of useful information. Rather, they are essentially meaningless and, if allowed, would mean that anything goes (1940: 410).
Some Marxist scholars have sought to side-step this issue by arguing that the requirements of formal logic do not apply in the social world where there are real oppositions, tensions and conflicts (e.g. Colletti, 1975). Accordingly, Hyman defines contradictions as ‘incompatible social forces’ (1975: 4), while Edwards states that ‘a contradiction is not a logical impossibility but is a state of tension between two or more features of the social structure’ (1986: 68; see also Watson, 2008: 282–3). More recently, Adler, in his critique of Labour Process Theory, states that the term ‘… is here used in a Hegelian sense, to designate a complex type of relation between real forces rather than merely between a logical incompatibility between propositions’ (2007: 1319). 5 Contradiction, for Adler, exists ‘at a deeper layer of causality’ such as between the forces and relations of production or between the socialization of production and the persistence of private property-based relations of production (1319–20). 6
But once the term ignores simple everyday logic it suffers an obvious loss of meaning that limits its social scientific value. Elster made this argument some years ago in an attempt to rehabilitate the concept within what later became known as Analytical Marxism. Analytical Marxism sought to salvage some viable ideas from Marxism by adopting the methods of analytical philosophy and conventional ‘bourgeois’ social science. The emphasis was no longer on a philosophically oriented interpretation of capitalist processes, as exemplified through dialectically pleasing formulations of contradiction, but on the kinds of causal chains or mechanisms that might be intelligible to mainstream social scientists (Wright et al., 1992: 6–7). Accordingly, Elster insisted that: If by ‘contradiction’ we mean only opposition, conflict or struggle, then we should say opposition, conflict or struggle. We should firmly resist the temptation to play upon the logical connotation in order to make our opinions seem interesting, and then fall back upon the non-logical connotations in order to make them look plausible. (1978: 3; see also Elster, 1985: 43–4)
One striking example that Elster used to demonstrate where the notion of contradiction was erroneously applied happens to be one that reoccurs in articles highlighting the contradictions of capitalism. This is the idea that a contradiction exists between the principles of socialized production and private appropriation (Adler, 2007: 1320; Knights and Willmott, 2007: 1372; Willmott, 1987). Such contradictions are important to Marxist and critical realist scholars because they believe that they reveal something about the deep structures of capitalism that non-Marxist theories are incapable of comprehending. But, for Elster, the contradiction is a merely verbal one, such as that between a tall wife and a short husband. Although such contrasts are striking, Elster insists that they are irrelevant precisely because they provide no reason why the apparent contradiction should not persist. His summary of the point is both pithy and devastating: ‘the notion of a social contradiction has the theoretical function of identifying causes of instability and change, not of locating symmetry violations’ (Elster, 1985: 48).
External Differentiation and Concept Redundancy
Gerring, in his insightful examination of ‘what makes a good concept’, states it should be possible to distinguish the boundaries of one concept from another though there may be some degree of overlap. Apples and pears, for instance, may share many similarities but they are still different types of fruit. So, even if there may be some overlap between concepts, problems arise when the boundaries between them are poorly defined (Gerring, 1999: 375–6). In the extreme case, where concepts are indistinguishable from each other, concept redundancy ensues. It was precisely because of this problem that the venerable Italian political scientist Giovanni Sartori insisted that ‘no word should be used as a synonym for another word’ when forming concepts (1984: 63). This is more than a mere matter of semantics. If contradiction is to be defined or used in such a way that it encroaches upon the notion of conflict then the theory from which this concept is derived is compromised. In other words, the utility of a concept is influenced by the extent to which it can be differentiated from neighbouring concepts (Gerring, 1999: 364; 375–9). More generally, if we are unable to distinguish between our concepts then we can no longer subscribe to one of the more widely shared views across the social sciences, which is that clearly defined concepts are the essential building-blocks of any theory.
The example of the conflation between contradiction and conflict is deliberate because it is among the two stock phrases that keep reoccurring within these leading journals. For instance, one of the aims of Heyes’ case study of training in a chemical plant is to show how ‘… actors seek to make sense of and manage the conflicts and contradictions that arise’ within the workplace (1996: 355), while Bolton and Houlihan insist that their Labour Process approach to the study of emotional labour ‘highlights the conflict and contradiction involved in customer-service work’ (2005: 691). Finally, Rubery and colleagues’ study of employment relations at a UK airport finds that: When workers are employed in a multi-agency setting, they attempt, through the same single act of labour, to satisfy simultaneously their obligations to two employers or agencies – the temporary work agency and the client. This simultaneity raises questions about organizational commitment and loyalty, and presents conflicts and contradictions with respect to lines of authority and workload. (2003: 270)
However, the other phrase, which places the emphasis on ‘tensions and contradictions’ has become much more common, though it is no less flawed. This phrase appears on some 24 occasions across 18 papers, while the ‘conflict and contradiction’ phrase appears 10 times across seven papers. Armstrong, for instance, claims that ‘… the real question of management, from the labour process perspective, concerns the tensions and contradictions within the agency relationship’ that characterizes capitalist management hierarchies (1989: 312). McCabe insists ‘… that as TQM progresses it will increasingly face tensions and contradictions’ (1999: 682), while Knights and McCabe believe that while ‘… the tensions and contradictions are inherent in the design of TQM, they may only be made manifest when seeking to implement them in practice’ (1998: 436). However, what is evident on reading through these papers is that the authors invariably fail to distinguish between the two terms. As much of the ensuing discussion concerns tensions of some kind, the ‘contradictions’ in ‘tensions and contradictions’ becomes little more than an empty cliché. In any case, the tendency to stretch the concept so that it includes the apparent contradiction between ‘quality and quantity’ or that between cost-cutting strategies and new management rubrics means that any gains made from extending the term have been matched by losses in connotative precision (Sartori, 1970).
Problems of Operationalization
One possible defence of contradiction is that it is a high-order theoretical concept that should not be treated as a directly observable concept such as births, deaths and votes. Abstractions of this kind should be assessed only as part of the wider theory and its usefulness in making sense of the underlying causal forces that drive capitalism. To analyse them at workplace level rather than at the level of the labour process generally is to mistakenly conflate two distinct levels of analysis.
The problem with this argument is that it fails to appreciate that all abstract concepts are not directly observable and a great deal of effort is put into attempts to locate their referents empirically. Given that almost all of the empirical research that uses contradiction is of a qualitative nature, it is important to stress that the requirement that abstract concepts are identifiable through empirical characteristics applies to both quantitative and qualitative research (Gerring, 2012: 155–8; Hammersley and Atkinson, 2007: 174–5). If anything, those working within the qualitative tradition like to claim that their standards for conceptual validity are at least as high, if not higher, than those working on survey research (e.g. Becker, 1996). But the crucial point is that we have to be able to recognize a concept when we see it so that we can, for instance, distinguish the experience of emotional labour from manual labour or that of conflict from one of contradiction.
In any case, the fact that contradiction might be an abstraction of a higher-order has not prevented researchers from using the term to interpret the evidence produced from the 27 qualitative studies identified in the journal search. Although such studies became more common over the 25-year period, there is no evidence to indicate that this was because the workplaces being studied had somehow become more contradictory. 7 There is certainly a growing number of studies that examine the ‘inherently contradictory’ new management rubrics, such as Total Quality Management and New Public Management, but the contradictions that they uncover tend to be of a rather conventional kind (i.e. between management policies or between policies and practice). Even then, a striking feature of this work and, indeed, of the case studies generally, is that the concept of contradiction is generally read into the evidence rather than emerging directly from the views expressed by employees. This includes the surprisingly few examples where the interpretation follows directly on excerpts from interview that might presumably be viewed as direct evidence (Foster and Hoggett, 1999: 29–30; Korczynski et al., 2000: 679).
A common refrain within the qualitative tradition is that the purpose of qualitative research is to understand the actors’ interpretations of their experience rather than those of any researchers, especially when they might have an ideological axe to grind (e.g. Becker, 1996; Hammersley and Atkinson, 2007: 3). My concern is that the imposition of contradiction creates two rather different but equally intractable difficulties. The first is that the emphasis on contradiction not only excludes other possible interpretations but that it may also prevent the researcher from following the unanticipated leads and emergent concepts that are among the strengths of qualitative research (e.g. Becker, 1996). This means that the preoccupation with contradiction may push the actor’s own interpretations into the background, especially if they do not agree with the ideological implications of the concept. Also it limits the chances of capturing new insights either through other sensitizing concepts or from the discovery of local folk concepts such as the ‘making out’ game in Burawoy’s celebrated study of Allied Corporation (1979).
The second difficulty is that the conclusions of these studies frequently contain the well-worn claim that labour control contains an inherent dualism of cooperation and conflict and, furthermore, that these co-exist in contradiction given the capitalist imperative to generate profit (e.g. Danford, 1998: 426–8; Dobbins and Gunnigle, 2009: 565; Knights and McCabe, 1999: 217–8; Korczynski et al., 2000: 684–5; O’Connell Davidson, 1994: 41–2; Webb, 1992: 490–1). This dualism, which is never voiced by those interviewed, is derived ultimately from the Hegelian belief that there are contradictions everywhere in reality. In other words, all phenomena in nature, society, or thought contain struggles or oppositions. The problem here, as Elster remarked, is that: … such statements say more about the theory than they do about social reality, especially when the meaning of the term becomes so vague that it is always possible to offer a post hoc claim to the effect that at least some feature of a given process is ‘contradictory’ in some sense. (1978: 68–9)
The Dependence on Additional Assumptions
As I indicated earlier, Marxist, Labour Process and other writers have used the concept of contradiction to explain developments in the control of labour under capitalism. The much cited contradiction between control and consent states that employers have to control their employees while simultaneously requiring their cooperation and goodwill. At the same time, employees depend on their employers for an income but also have an interest in resisting their subordination (e.g. Cressey and MacInnes, 1980; Edwards, 2003; Hyman, 1987). 8 Although this formulation of the problem has a certain dialectical elegance I would argue that it suffers from three major weaknesses.
The first is that it only makes sense logically if buttressed by a number of implicit assumptions. For instance, control and cooperation must be assumed to be incompatible or at least prone to undermining each other. 9 Logically, it could also be assumed that increased control might bring more rather than less cooperation. Workers might, for example, respond better to firm management because they like to know who is in charge. Indeed, some Marxist studies of workplace relations have actually recorded workers expressing a need for strong managerial authority, though this has been interpreted as a ‘contradictory desire’ (e.g. Collinson, 1992: 176). Either way, the claim only makes sense if it is based on certain assumptions about how employees respond to their experience of employment.
Second, and relatedly, the argument must assume that workers are alienated and exploited (e.g. Adler, 2007: 1324; Brook, 2009: 544; Cressey and MacInnes, 1980: 14; Danford, 1998: 426–7). Accordingly, extracting work is not only problematic for employers but also any attempt to concede control will be used to the employer’s disadvantage. Friedman even goes so far as to claim that ‘because labour power is alienated under the capitalist mode of production, each strategy, in its ultimate vision, is based on a contradiction’ (1977: 7). The problem here is that the debate over alienation, which once marked a long-running divide between Marxist and non-Marxist sociologists, has more or less ended. Along with the related concept of exploitation, it is still not clear how it might be subjected to any kind of serious empirical examination. On this point, it is worth recalling the position taken by Goldthorpe and colleagues when they tried to apply it within the Affluent Worker project. They concluded that the concept of alienation ‘is not a specifically sociological concept: it is rather a notion expressive of a certain human and social philosophy which often figures crucially in a rhetoric of revolution. It is not intended to be tested against fact’ (Goldthorpe et al., 1969: 179).
A third problem with dialectical reasoning of this kind is that it is unable to account for social action at the individual level since the changes that occur flow from the exigencies of systemic or structural conflict. Even if some notion of an alienated worker were to be accepted it is quite possible that workers may experience collective action problems in responding to either coercive or cooperative forms of labour control. If some decide to leave and others wish to remain loyal there is no guarantee that the remainder will be capable of organizing themselves to the point where they actually bring about a change in the form of labour control. In fact, an influential review of working-class consciousness reported that those ‘who are most alienated and most desperate are those who are least confident of their ability to change their situation’ (Mann, 1973: 70). In short, it seems that this particular contradiction and the changes in the labour process that are associated with it reflect a wider problem with Marxist theories of social change which is that they are not so much examples of critical thinking as wishful thinking. 10
Contradiction and Change
Finally, it should be remembered that Marxists view contradiction as a central component in processes of social change. Within the literature on the sociology of work and industrial relations, two not unrelated claims can be identified. The first is that the dialectical relationship between control and resistance creates a dynamic that generates new forms of control. Such writers insist that control over the labour force is never complete and that the inherently contradictory nature of such control explains its evolution from one form to another (Burris, 1989: 2–3; Friedman, 1977: 106–7; Storey, 1983: 185–7). While this argument has attracted many supporters within the labour process tradition, it is open to the obvious challenge that nobody has been able to demonstrate empirically that processes of labour control have developed dialectically and, in particular, that worker resistance has proved to be either a productive or consistent source of managerial innovation (see, for instance, Brody, 1984; Edwards, 1986: 39–41; Lazonick, 1983: 131–2).
The second claim in relation to social change is, of course, the classic Marxist argument that contradictions represent self-destructive forces within capitalism that generate so much instability and conflict that they eventually lead to its demise. Leaving aside the possibility of societal transformation, which featured in many of the Marxist papers (e.g. Adler, 2007: 1328; Barrett and Rainnie, 2002: 424; Burris, 1989: 18; Hyman, 1987: 52) the idea that contradiction would generate instability and the prospect of change was also evident from the journal search. Knights and McCabe, for example, noted ‘… how some of the contradictions [of TQM] leave space for staff not only to resist but also to extend aspects of the quality programme beyond the parameters of management intentions’ (1998: 436). Other studies also saw similar opportunities for resistance and or collective organization (e.g. Danford, 1998: 427; McCabe, 1999: 688; Russell and Thite, 2008: 630–1).
Focusing any further on individual studies runs the risk of missing the general point which is that a substantial strand of research in leading academic journals insists that capitalism, capitalist firms, management strategies, and the employment relationship are riddled with contradictions. Assuming that this is indeed the case and assuming that such contradictions do provide opportunities for collective organization and resistance, then the obvious question to ask is why union membership, collective bargaining, strikes and other forms of industrial action have collapsed over the past few decades (Godard, 2011; Kersley et al., 2006; Lyddon, 2007). Nor is there any evidence to indicate that the firms that introduced any of the ‘inherently contradictory’ new management models subsequently struggled or went out of business. Perhaps nothing is so damaging to those who insist on the significance of contradictions than the massively awkward fact that they have not generated the kind of workplace instability and change that the concept originally implied.
Conclusions
If the sociology of work is to advance on theoretical and empirical fronts then it has to encourage a shared commitment to clarity and rigour in scholarship, terminology and meaning. I have argued that the concept of contradiction and the way it is used by a substantial number of scholars has failed to meet these requirements because of weaknesses in logic, inadequate differentiation, problems of operationalization, and its dependence on flawed assumptions. Even if the associated problems of concept stretching and concept redundancy are set aside this still leaves what is probably the most damning weakness – the expectation of widespread instability and upheaval within capitalist workplaces.
The danger is that if the current tendency for piling up case studies of contradiction continues then it is unlikely to generate the kind of conceptual and empirical innovation that were once the hallmarks of British industrial sociology. In short, I believe the time has come to call for a moratorium on the concept and possibly for abandoning it altogether. The experience of some of the leading Analytical Marxism scholars is informative on this point. Elster substituted the related sociological concept of unintended consequences (Merton, 1936) for the ‘real contradictions’ that can be found in Marx’s work. Indeed, he suggests that one of Marx’s central contributions to the methodology of the social sciences was his analysis of unintended consequences, especially those of a negative or counter-productive nature (Elster, 1978: 106–24, 1985: 22–7, 44–8). Significantly, Elster himself no longer makes much use of contradiction, ‘real’, ‘social’ or otherwise, while the idea of unintended consequences or externalities is presented as one of a range of social mechanisms (Elster, 2007: 300–11). In any case, what matters is that whatever concepts we use are not only clearly and consistently defined but that they also help explain social reality. Otherwise, we leave ourselves open to the old allegation that sociology is little more than political prejudice dressed up in academic jargon.
Footnotes
Acknowledgements
I would like to thank Paul Edwards, Stephen Hill, Richard Hyman, Sandy Jacoby, John Kelly, Bill Roche, Tim Strangleman and his colleagues at the Sociology seminar at the University of Kent for helping to sharpen my arguments even when they are not necessarily ones that they would agree with. The usual disclaimer applies.
Funding
This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.