Schools of thought in and against total quality.

In the early 1980s, a new concept entered managerial discourse: Total Quality Management (TQM). Later called "Total Quality" (TQ), TQM was heralded by governments, major corporations and the business media as the most effective and elegant way out of the economic crisis and into the global market. It should be noted, however, that the preoccupation with quality is by no means new. In a 1956 article, Feigenbaum describes the new stakes confronting corporate managers in the early 1960s as follows: "Customers - both industrial and consumer - have been increasing their quality requirements very sharply in recent years. This tendency is likely to be greatly amplified by the intense competition that seems inevitable in the near future" (Feigenbaum, 1956: 93). In the 1980s, TQM became a product in itself, nearly a billion-dollar industry (Harari, 1993b). Like many fashionable concepts, TQM has spawned abundant literature aimed principally at managers confronted with problems of implementation (Reed et al., 1996). Only recently has there been a more indepth analysis of the total quality movement, with several academics publishing the results of theoretical and empirical research. For instance, an entire issue of The Academy of Management Review (19(3), July 1994) was devoted to studies of this nature. TQM also earned its share of detractors who accused it of being merely a fad. Several authors pointed out that while total quality approaches have met with considerable success, their failures, though less publicized, have been even greater (Ahire, 1996; Dreyfuss, 1988; Hammonds, 1991; Krishnan et al., 1993; Roberts and Corcoran-Nantes, 1995; Schaffer, 1993; Sherwood and Hoylman, 1993; Spitzer, 1993). Others questioned TQM's conceptual soundness, its applicability and its ideological basis (Schaffer and Thomson, 1992; Dean and Bowen, 1994; Hill, 1995; Webb, 1995; Steingard and Fitzgibbons, 1993; Tuckman, 1995; Wilkinson et al., 1991). Given the existence of diverse discourse on and against TQM, we believe that a literature review and a classification of this literature may be of value to both academics and practitioners who seek to better understand the value and limitations of TQM. The purpose of this article is to propose a classification of different schools of thought and critiques in quality management, to show the relationship between each of them and how they have engendered what is now called total quality. Table 1 shows excerpts of the schools of thought classification and Table 2 shows excerpts of the critique classification. We believe that some of the harshest critiques of TQM originate in the nature of the evolution of the field of quality management, which is marked by the presence of several shifts. Classification helped us develop a theoretical analysis of the historical and discursive dimensions that we believe are being neglected. We think that our analysis may cast the current debate among believers, nonbelievers, and skeptics in a different light. TWO SCHOOLS OF THOUGHT IN QUALITY MANAGEMENT Several authors have described the confusion created by the existence of distinct definitions of, and perspectives on, quality management (Chatterjee and Yilmaz, 1993; Dean and Bowen, 1994; Gehani, 1993; Krishnan et al., 1993; Sitkin et al., 1994; Wilkinson and Willmott, 1995). For example, Dean and Bowen concluded that "Despite thousands of articles in the business and trade press, total quality remains a hazy, ambiguous concept. The difference among frameworks proposed by writers such as Deming, Juran and Crosby have no doubt contributed to this confusion" (1994: 394). In our view, however, the confusion is due not only to differences in approach, but also to the lack of interest shown by academics. To our knowledge, the attempts at differentiating schools of thought in the literature are limited to a distinction between "hard" and "soft" approaches. The former are associated with the systematic control of work using statistical tools, while the latter emphasize qualitative and human aspects (Hill, 1995; McArdle et al., 1995; Warnotte, 1992). It is always difficult to group and label approaches, particularly when they do not have well established foundations in management theory. Unfortunately, this is the case for most TQM approaches (Dean and Bowen, 1994). We nevertheless believe that it is important to try. Thus, we have borrowed the categories proposed by Barley and Kunda (1992) to chart the evolution of American managerial discourse. According to these authors, the managerial discourse consists of alternating waves of normative and rational rhetoric and forms of control. Rational control is characteristic of theories such as scientific management and systems rationalism, according to which productivity stems from expertise, careful work design, and coherent support systems. It presumes that employees are "calculative actors with instrumental orientation to work" and "understand the advantages of an efficient system or . . . [are] powerless to resist a well-designed structure" (1992: 384). Under this assumption, control is achieved by manipulating systems, and if the systems are adequate, compliance is assured. On the other hand, normative control (associated, for instance, with welfare capitalism in the 1930s and 1940s, and the more recent trend of organizational culture) assumes that the organization is "a locus of shared values and moral involvement" where "cohesion and loyalty . . . [are] the ultimate source of productivity. . . . Control therefore rested on shaping workers' identities, emotions, attitudes, and beliefs" (1992: 384). Barley and Kunda associate TQM with the more recent normative wave of organizational culture. However, when one considers the evolution of the quality management movement, it appears that both forms of control are present in the quality management literature. In fact, we argue that these forms of control constitute a prominent difference between the schools of thought. The schools of thought we refer to as rational and normative both share a common goal: customer satisfaction. However, each school proposes a different means to achieve this goal. We must point out that the classification we propose is not the result of a systematic content analysis of the literature on TQM. Rather, it is inferred from an extensive (if not exhaustive) literature review. Quotes from the main authors associated with the two schools of thought are presented in Table 1. The Rational School The guiding principle of the rational school is the importance of acting at the level of systems and processes and, contrary to the normative school, not blaming poor quality on individual attitudes. This is the broadest, best documented, and oldest school of thought in quality management. Shewhart (1931) was one of the first to develop statistical tools allowing managers to make more enlightened decisions regarding not only product quality, but also the need to intervene in a given process during manufacturing. Statistical process control represented an important break with the traditional view of quality control, which until then had been limited to product inspection (Garvin, 1988). Even today, the rational school views statistical tools as a fundamental ingredient of quality management in a context of large-scale production using standardized components. In the 1950s, however, Deming - another statistician - argued that statistical tools are inadequate if managers use them improperly or impractically (for instance, by looking for a culprit), or when they foster managerial practices that inhibit workers from applying information acquired from statistics to improving processes. Indeed, Deming recognized that data on product quality and process stability are useless if management does not subsequently act to rectify the situation or if management's actions are limited to laying blame and exhorting improvement (Deming, 1982, 1986). Deming, thus, developed a set of "lines of conduct" for managers, which have come to be known as "Deming's 14 points."(1) Deming's message, introduced in Japan, was subsequently extended by Juran (1951), who also advocated changes in management practices. This was followed by the publication of Feigenbaum's Total Quality Control, which proposed to broaden quality management to include the functions of product design and maintenance (Feingenbaum, 1961). The introduction of the concept of "Total Quality Control" (TQC) represented a turning point in quality management because it attributed the responsibility for finished product quality to a larger number of key individuals. Ultimately, responsibility was spread over the entire company - "Company-Wide Quality Control" - and involved not only functions directly related to product creation but also all company services involved in supplying production support and resources. Within this framework, it is no longer enough to manage the quality of finished and unfinished products, but attention must also be paid to the quality of raw materials, worker training, equipment, etc. (Ishikawa, 1985). To summarize, the rational school presupposes that individuals want to do a good job. If they do not, the reasons are twofold: they may be hindered by the obligation to conform to methods prevailing in their workplace, or they may not possess the necessary information to manufacture good products. Within this framework, any attempt at improving quality can only succeed by means of a systematic, critical examination of existing management systems and by developing tools to gather and analyze data about the quality produced. The Normative School The normative school stresses the individual responsibility of employees (whether line workers or managers) with regard to quality and the need to correct behavioral flaws. This school emerged in the 1960s with Crosby's "zero-defect" theory, which argues that there is only one acceptable level of quality (zero-defect), that the only way to ensure quality is "to do it right the first the time," and that the only true measure of quality is the price of nonconformance resulting from a lack of prevention. This is what led him to claim that Quality Is Free (the title of his best known book; Crosby, 1979), as well as to suggest that the solution to nonconformance consisted of involving individuals and refusing to accept that "to err is human" (Crosby, 1964). In 1974, in the third edition of the Quality Control Handbook, Juran (a rationalist) argued that Crosby's perspective was based on spectacular examples in which human error was the cause of a catastrophe and that these examples were not representative of the real problem of poor quality encountered by managers. In his view, this "inattention" school was no more than a flash in the pan, even though he recognized that many company executives could find it quite attractive (Juran, 1974). The unfolding of events, however, would prove him wrong. The 1980s saw the normative school grow in importance with the push toward excellence (Peters and Waterman, 1982), associated with quality and "zero-defect," as well as with the rise of economic liberalism with its emphasis on individual responsibility. With the rise of the concept of excellence and its pairing with quality management, the 1980s witnessed a distortion of the original TQC concept (in which it is "control," not quality, which is total(2)), which was transformed into TQM (Total Quality Management), then often simply reduced to total quality. It appears to have been by an implicit and gradual consensus that total quality took root, even though precise definitions of the term are rare and divergent (for examples of definitions of total quality, see Dean and Bowen (1994), Ebel (1991), and Kelada (1996)). Many who claim to be talking about total quality trace their authority to Deming and Juran, two of the best known authors in the area of quality management. To our knowledge, however, these authors have never used the expression total quality, even in their recent works. For example, when asked about the future of TQM, Deming replied "there is no such thing. It is a buzzword. I have never used the term, as it carries no meaning" (Romano, 1994: 22).(3) In summary, the normative school's basic hypothesis is that poor quality can largely be attributed to worker negligence and to companywide carelessness. In this context, the objective is to develop and disseminate a flawless argument to illustrate the role played by the individual in attaining quality (with management setting an example) and to stress economic rationality and the performance obligation facing companies. By means of repeated messages and appropriate incentives, the goal is to allow for the integration and actualization of this normative vision in daily behavior. THREE CRITICAL STREAMS TQM is quite recent, and critiques have begun to appear only in the last five years or so. We grouped the critiques into three main streams according to the subject under discussion. Our categorization was based on an extensive but not exhaustive review of the literature. Some believe that the approaches undertaken have not yielded striking improvements and question the ability of TQM to generate the promised results: the pragmatic critique Certain authors have examined the approaches themselves from a conceptual perspective and offer a critique of their content: the theoretical critique There are those who contest the social and political premises underlying TQM and raise questions about its relevance: the ideological critique. The main arguments of each critical stream are shown in Table 2. The Pragmatic Critique: Does TQM work? The pragmatic critique (e.g., Schaffer and Thomson, 1992; Harari, 1993a, 1993b) originated from two main groups: quality "specialists" who critique the approaches put forward by their colleagues (and sometimes competitors), and executives and workers who failed at implementing these approaches. Juran's (1974) critique of Crosby's vision was presented above, yet the opposite argument also exists. For example, rationalists are often accused of proposing an idealist vision which over-emphasizes processes to the detriment of results. As such, some have argued that companies should adopt timely approaches intended to produce short-term results which have a visible impact on company profitability (Schaffer and Thomson, 1992). These kinds of approaches clearly identify causal links between actions and results, thereby contributing to an ongoing learning process based [TABULAR DATA FOR TABLE 2 OMITTED] on past experience. This would be impossible with long-term approaches aimed at improving systems since too many variables influence the results. Rationalist approaches are also criticized for confusing ends and means, allowing managers to hide behind an anticipated long-term performance improvement so as to avoid having to explain difficulties encountered and the absence of tangible results (Schaffer and Thomson, 1992). Moreover, data concerning alleged TQM failure rates and the description of particular cases have given rise to a new debate: did the companies which experienced failure really adopt a TQM approach? According to some observers, businesses which met with success were those which had adopted a "true" TQM approach, while those who experienced failure had either not adopted TQM or had poorly implemented it (Ahire, 1996; Becker, 1993). In response to this claim, others maintain that the "true" TQM is that which is described as such by managers and not what it is supposed to be in theory (Harari, 1993b). Pragmatic critiques can also be found in the informal comments made by managers and workers who have unsuccessfully attempted to implement TQM. Criticism sometimes surfaces in the distortion of TQM vocabulary and associated concepts: TQM becomes the "total quality myth" (Bemowski, 1995); QWL - quality of working life - is rendered as "quietly whipping labour" (Landen and Landen, 1990). Other expressions from the total quality discourse have been discredited and forcibly eliminated (Landen and Landen, 1990; Tuckman, 1995). For example, the term quality circles has often been replaced by continuous improvement work-teams or other similar expressions. As consultants and teachers, we often hear disillusioned and cynical remarks concerning the official quality discourse. Although this is only anecdotal evidence, such remarks are so frequently and spontaneously expressed that we feel compelled to include them as informal manifestations of pragmatic critiques. The following examples are drawn from our experience or provided by research and practitioner colleagues(4): "invited suggestions are heeded only when they reflect the desires of management," "management itself is not able to implement its own principles," "quality is only important to the extent that production quotas and delivery deadlines have been met," and "quality is something that is done only between ten and eleven o'clock, Thursday morning." During a training session on statistical process control, we even heard one of the participants say "don't worry, here we make sure that the control charts always say what they are supposed to say." But the type of remark we most often heard is "there was a lot of talk about total quality for a few months, then it just disappeared into thin air...." The failures of TQM (and the disillusionment of the workforce) are often attributed to a lack of top management commitment, a lack of rigor during implementation (which leads to a failure to develop a quality-oriented culture), or a lack of understanding of the "real" nature of TQM (Ahire, 1996; Becker, 1993). However, contrary to the pragmatic critique, the two following critical streams question the very nature of the proposed change as well as its underlying hypotheses. The Theoretical Critique: Is TQM scientifically sound? This critical stream (e.g., Dean and Bowen, 1994; Hill, 1995; Kerfoot and Knights, 1995; Roberts and Corcoran-Nantes, 1995) encompasses a wide variety of works on TQM as an approach to change, the contents of which are examined from a conceptual perspective. Generally speaking, they criticize TQM advocates because of the purely prescriptive nature of the approaches they propose, their lack of nuance with respect to difficulties in the implementation of change, and the absence of any reference to management theory. Some researchers attempt to compensate for the latter failing by establishing links between TQM and current knowledge about organizational management. For example, Dean and Bowen (1994) claim that total quality prescriptions regarding worker leadership, training, and participation could be corroborated by human resources management theory. Furthermore, Spencer (1994) proposes that TQM, generally associated with an organic management model, could in some way be brought closer to the mechanistic model and the cultural model. While the TQM literature can find a degree of support in management theory, several authors argue that it has important limitation as well. For example, the TQM literature has been criticized for its failure to take into account the presence of different interest groups within the organization, social issues, and the problem of industrial relations (Hill, 1995; Roberts and Corcoran-Nantes, 1995; Wilkinson et al., 1991). According to Hill, "Resistance to improvement is properly highlighted, but its nature and extent are not understood and the proposed solutions are too restricted.... the mechanisms required to persuade people to "buy into" quality management are limited to leadership from the top, systematic education and training, learning the benefits by doing and recognition for achievements. The thrust of these prescriptions is that top managers should win hearts and minds without compulsion" (1995: 40). Moreover, Dean and Bowen (1994) consider that total quality places too much emphasis on the formal analysis of information, which is inconsistent with research demonstrating the importance of maintaining a certain degree of informality and ambiguity in organizational contexts characterized by uncertainty and political issues. Dean and Bowen also argue that total quality imposes a strategic formulation that centers exclusively on client expectations, without taking into account considerable literature on strategy that recommends that consideration of an organization's strengths and weaknesses and its competitiveness is just as important as that of the market needs. The critiques above raise the issue of the universality of TQM's concepts. The total quality discourse typically suggests, or explicitly claims, that the implementation of TQM is both desirable and feasible in any and all public or private services or industrial organizations. Organizational contingencies are either completely forgotten or viewed as having little importance (Dean and Bowen, 1994; Reed et al., 1996). Sitkin et al. (1994) go even further to suggest that this "overselling" of TQM might be the cause of many of the reported fail- There has also been some interest in the psychological dimension of the total quality discourse. An American study conducted at AT&T (Lader, 1988) shows that for Americans "quality" is an emotionally charged concept related to personal feelings of success and failure, serf-esteem, and the fulfillment of parental expectations, which are subsequently transferred to authority figures. This study shows that expressions such as "do it right the first time" and "zero-defect" (associated with the normative school) are perceived negatively, and the terms "control" and "total" are spontaneously associated with inflexibility and totalitarianism, leading to the conclusion that the expression "total quality control," notwithstanding its underlying good intentions, may be a very poor choice of words. The study also suggests that the archetypes of quality are represented in films like Rocky and Star Wars, in which the final result of attaining an ideal is less important than perseverance and the struggle against adversity, without which there can be no real success. The presence of these archetypes would explain why the concept of continuous improvement is much better received by workers, because it focuses on the struggle more that on an ideal result. With certain nuances, this conclusion has been supported by a second study (Bemowski, 1993). While the studies above suggest the existence of a transfer between management-defined quality improvement objectives and individual aspirations, they do not directly question the relevance of such a displacement. Other researchers, however, have looked into this matter. Their arguments constitute what we call the ideological critique. The Ideological Critique: Does TQM lie? Contrary to the pragmatic and theoretical critiques, the ideological critique (Aubert and de Gaulejac, 1991; Delbridge et al., 1992; du Gay and Salaman, 1992; Kerfoot and Knights, 1995; Steingard and Fitzgibbons, 1993; Webb, 1995) does not focus on either the theoretical content or the applicability of TQM approaches. Rather, it examines the social consequences and political significance of the movement. The critique is based on the observation of a considerable gap between TQM rhetoric and the reality experienced by organizations implementing TQM (Webb, 1995; Kerfoot and Knights, 1995). First, while TQM rhetoric advocates autonomy and worker empowerment, the fact is that its implementation is generally accompanied by an increase in control (Delbridge et al., 1992; Webb, 1995). Second, TQM is supposed to be a collective effort in pursuit of a common cause. Yet the concept of an internal customer on which it is often based inherently conveys the other side of any market relations: exploitation based on power (Webb, 1995). Finally, TQM promotes worker participation, their personal involvement in the organization's success, and the possibility of improving their working conditions and their chances for advancement by initiating a true "meritocracy" (Webb, 1995). However, critics believe that TQM implementation is often undertaken within the framework of massive streamlining, accompanied by a flattening of hierarchical structures, which reduce chances for advancement (Kerfoot and Knights, 1995; Roberts and Corcoran-Nantes, 1995). These divergences might be attributed to the fact that the total quality discourse was rashly adopted by many organizations seeking to keep up with trends and improve their public image. However, a much harsher criticism argues that the total quality rhetoric is just a way of gaining acceptance for changes which serve ends other than those it purports to pursue, thus constituting a vocabulary of motives (Webb, 1995). Some researchers have focused on the impacts of TQM approaches on employee life. For example, Aubert and de Gaulejac (1901) conducted in-depth studies of the harmful effects of approaches advocating excellence, which fit in well with the culture of anxiety currently prevailing in industrialized societies, but which frequently lead to burnout. It is also argued (du Gay and Salaman, 1992; du Gay, 1906) that in creating the cult of the customer and making everyone a supplier responsible for customer satisfaction (which is a key principle of TQM), the needs of the organization are superimposed on individual projects. The financial difficulties experienced by companies are transferred onto the shoulders of individuals. Employees are made to believe that, as responsible suppliers and alert customers, they will become more virtuous and more empowered individuals. In fostering an image of omnipotence, perfection, and expansion, it is argued that corporate growth is fueled by the self-actualizing capacities of employees, who see progress as giving their lives a meaning which they cannot find elsewhere. When the company fails, even partly, in its quest for excellence, the result is either a collapse of the ideal image of self (burnout) or a psychological withdrawal (reduction of involvement). In internalizing the corporate culture and in perceiving themselves as corporations whose success they must secure, employees are said to endorse their own exploitation and willingly, even at times enthusiastically, participate in what McArdle et al. (1995) have termed "management by stress." Empirical data show that TQM introduction increases the control over workers, though it is supposed to reduce it (Delbridge et al., 1992; Webb, 1995). It is argued that the obligation to be accountable to internal customers introduces a neighborhood watch, which some express as "the chain metaphor represents manacles as well as links" (Tuckman, 1995: 58). Moreover, TQM techniques themselves, with their insistence on the establishment of indicators of quality and control charts that must be made visible (graphs are often displayed on the walls of the workplace), always give managers access to "scientific" data, easily considered irrefutable. According to McArdle et al. (1995), the increase in control might be all the more facilitated, given that it is justified by a rhetoric of the market and competition, which play on the fear of job loss in order to eliminate resistance. Some effort has been made to emphasize the political nature of the total quality movement. Munro (1995) and Webb (1995) argue that within organizations, TQM constitutes a new discursive space used by managers in their own political struggles. For example, Webb concludes from a case study that "Instead of the prescribed equal dialogue.... Marketing had emerged at the top as a result of the rhetoric of customer responsiveness. It was able to use its direct contact with customers as a legitimating device" (1995: 117). Other authors (Steingard and Fitzgibbons, 1993; Tuckman, 1995) more interested in the sociopolitical nature of organizations, view TQM as an element within a larger hegemonic project which presents market relations as an ideal to be overlaid onto social relations. Steingard and Fitzgibbons maintain that the pursuit of this ideal leads employees to be "totally managed" as parts of the "TQMachine." They write: "this [TQM] organizational panacea for the twenty-first century conceals a capitalist schema of alienation, dehumanization, and totalitarianism" (1993: 31). This political critique is not without its detractors, however, who question the characterization of TQM as part of a Machiavellian plan for worker exploitation (Kerfoot and Knights, 1995; Webb, 1995). They argue, for example, that increases in the control of workers might be a perverse effect of TQM rather than a hidden objective. Hill (1995) also argues against the theory of indoctrination and maintains that "people are not cultural dopes." DISCUSSION It is quite normal and very healthy that TQM should have proponents, critics, and "impartial" observers, like any valid management approach. It is only through the discussion of ideas that advances are made in the social sciences. We believe, however, that the study of the evolution within the field of quality management exhibits certain characteristics which may provide a new understanding of the current situation. The presence within quality management of schools of thought which differ fundamentally, but which are not recognized as such, leads to confusion and contradictions within the discourse on quality. Several authors and consultants have attempted to integrate the views of the rational school with those of the normative school in order to present what they see as a "more complete" approach. It is for this reason that certain approaches simultaneously advocate both continuous improvement and "do it right the first time," and both "zero-defect" and the right to make mistakes. These approaches borrow the most seductive concepts and striking slogans from various authors without concern for the coherence of the resulting message. To be sure, some have attempted to reconceptualize these expressions with a view toward eliminating contradiction. However, the often complicated definitions that ensue do not necessarily clarify matters for the average worker. It is sometimes forgotten that the meaning given to words is as much a function of those using them as those defining them. For example, in everyday language it is quite natural to associate the word "total" with the idea of absolute or perfection, and the nuances developed by theorists change nothing. It is possible that the existence of several definitions could be advantageous if the concept can be adapted to the particular situations of those using it (Garvin, 1984; Reeves and Bednar, 1994). However, when talking about quality, it is necessary to take the time to define one's terms precisely before beginning an in-depth discussion. Moreover, divergences must be acknowledged. Finally, even though some may contend that this change in the vocabulary is the result of natural evolution (for instance, replacing "control" with "management" because of the former's negative connotation and subsequently eliminating "management" to avoid the impression that TQM was a program limited to management), we believe that the new wording is indicative of a more fundamental conceptual shift. Another paradox emerges from the juxtaposition of rational and normative schools of thought. Rhetoric from the rational school purports that TQM is founded on the desire for scientific rigor (management by facts) and relies heavily on formal systems and technical tools. Rhetoric from the normative school purports that its very justification and the context in which it is implemented is the struggle for survival waged by organizations in a global market, but in this case TQM is articulated by a moral, religious vocabulary(5) (Steingard and Fitzgibbons, 1995; Webb, 1995). The "quality crusader" must "have faith," but its actions must be based on objective facts. More importantly, the discourse surrounding quality has shifted considerably over the years (for examples, see Table 3). The concept of quality management has been broadened to such an extent that it is now a synonym for "sound management" (Table 3, Column 1). The economic argument evolved from a rhetoric of reducing costs as a means of justifying improvements in product quality to a rhetoric of improving quality in order to justify reducing costs (Column 2). The discourse shifted from the need for everyone w be involved in the improvement of product quality to everyone becoming a "quality person" (Column 3). To better understand these shifts and their interrelations, we present an historical perspective and contextualize the prescriptions of those currently referred to as quality "gurus." Originally, quality management had nothing to do with a search for "total quality." Rather, it was a more prosaic response to the need to offer products and services, the characteristics and attributes of which corresponded to what customers had been promised. This was the underlying principle for Shewhart, Deming, Juran, and Feigenbaum, who considered deficiencies in management systems as merely a hindrance to product quality, and company-wide quality as only a means to improving product quality. In Feingenbaum's TQC, "control" is total, not quality.(6) TQC's objective is to secure the participation of all company functions in improving the quality of products and services offered to customers. Indeed, the Japanese ideograms for the TQC concept literally signify "quality together" (Aubert and de Gaulejac, 1991). With Deming, the tendency is to retain only his 14 points, which [TABULAR DATA FOR TABLE 3 OMITTED] when taken out of context, appear as prescriptions for sound management. However, the idea behind the development of these management concepts was the improvement of product quality. If "drive out fear" (point 8) or "eliminate numerical quotas" (point 11) were deemed necessary, it was because fear led employees to camouflage defects and because production quotas emphasized production quantity to the detriment of quality. In the early work on quality management, cost reduction was the argument used by specialists in this field to convince managers to adopt the tools and methods they were proposing. In effect, managers at the time thought improving product quality would entail high costs, and it was necessary to convince them that the costs of poor quality were greater than the required investments. The objective of Deming, Juran, and other pioneers of quality management was the improvement of product quality. The possibility of reducing production costs by reducing rejects and reworking due to poor quality was an incentive, a happy consequence, or almost a bonus that occurred in concert with improvement of quality. In the 1980s, however, TQM became a "product" to be sold, borne by the popularity of the concepts of competitiveness and excellence. In order for TQM to be even more attractive to the company heads for whom it was intended, market share and profitability were emphasized. The concept of quality was thus broadened to "to do a quality job," which was associated not only with the reduction of error, but also with the efficiency of work methods and the reduction of waste due to bureaucracy. All company services had to improve the quality of their work not only for production personnel to turn out better products, but (first and foremost) to improve the entire company's productivity and competitiveness. As such, TQM became the symbol of sound management, as illustrated by the following: "total quality is simply, and somewhat astonishingly, not really about quality at all! . . . [It's about improved organizational Performance]" (Schaffer, 1993: 19). Far from being superficial, this redefinition of the concept of quality constitutes a complete reversal of the situation. Initially, advocates of quality management employed a rhetoric of cost reduction to persuade business leaders to invest financial resources in quality management. Now, business leaders (often led by consultants) employ a rhetoric of quality to enlist employees and executives in a program of change. The primary objective of the program is to improve the company's competitive position, not only with respect to its customers, but to its shareholders as well. One can observe a transition from the management of quality to the quality of management (Brocka and Brocka, 1992, Feigenbaum, 1997), and, finally, to management by quality. These shifts in the discourse surrounding quality have important consequences. First, they permit a better understanding of the ideological critique. This critique accuses TQM of using a virtuous discourse of improving quality to exploit workers for the benefit of capital (even though it is understood that this is more a question of following passing fads and the manifestation of management's desire for miraculous solutions than the result of a Machiavellian maneuver). Second, if TQM is the equivalent of sound management, then it reduces good management to a set of prescriptions. Hence, the theoretical critique is even more justified in reproaching the TQM discourse for not grounding itself in management research and neglecting the social dimension. This also makes TQM an incredibly ambitious project - all the more so, given that the emphasis is on the word "total." This might explain the low success rate of the TQM projects surveyed by the pragmatic critique. Above all, we believe that these shifts are a hindrance to quality management. By disguising the elimination of waste as efforts to improve quality, quality management is tainted by deceptive maneuvers for which it is not responsible. TQM is thus vulnerable to being discredited. One might wonder whether in trying to achieve "total" quality, "plain" quality is neglected. "Plain" quality refers to the simple quality of the product or service which corresponds to the legitimate expectations of the purchaser and should fulfill the producer's promises. In effect, the TQM concept is now charged with so many different meanings that the place of quality in itself is being increasingly reduced. For example, in a recent trade show on the implementation of quality at which representatives of various companies gave the results of their quality improvement programs, a good proportion of exhibitors showed what were in fact programs for the elimination of waste (reduction of set-up time, optimization of a production line, reduction of material loss in pattern cut-outs, etc.). Our intention is not to minimize these efforts, but to point out that, while quality management and efficiency management are both important and can often overlap and help one another, they nonetheless remain two distinct elements of organizational management. We believe that talking in terms of simple quality has the advantage of taking us back to the very concrete and, ultimately, relatively humble origin of quality management. Such an approach might not appear very ambitious to advocates of excellence, but we think that the simple quality of products and services already is in itself a very important objective and that the means to attaining it are very complex. To speak in terms of simple quality is also to reframe quality management within organizational constraints and environmental contingencies and to remove it from the hands of an evangelical and universalist discourse which leaves little room for discussion and real participation. We should add, however, that we are not suggesting that simple quality is "purer" than total quality, nor that it is free of all value or ideology. Its ideological charge is no greater than that of any other dimension of management in a market economy and capitalist society. To speak in terms of simple quality would not in our view constitute a step back or a rejection of efficiency, but rather a re,emphasis of the hard reality of the arbitration sometimes needed between opposed interests and objectives. IMPLICATIONS FOR MANAGERS AND FURTHER RESEARCH In this article, we have demonstrated that the field of quality management is not always as homogenous and unified as some are led to believe. On the contrary, it is comprised of distinct schools of thought. Our classification is a preliminary proposal which could (and should) be supplemented by further scientific analysis. We believe that although classifications are always imperfect and subjective, they provide managers with useful criteria to characterize and comprehend the many quality management programs available. We also believe that the labels "rational" and "normative," as borrowed from Barley and Kunda (1992), may be useful in several respects. They call attention to the possibility of internal contradictions in cases where the program in question proposes an orientation that seems to be both rational and normative. In addition, they evoke the fact that different quality management approaches may be based on quite different assumptions about the nature of human actors and their attitude toward work. Therefore, before selecting a particular program, it might be worthwhile to consider whether or not the assumptions on which the program is based correspond to one's own beliefs and to the organizational reality. Finally, these concepts remind us that quality management is also a form of control, anchored in rhetoric, and is not free of ideological content. As such, quality management cannot be idealized as being above other organizational contingencies, conflicting interests, and political struggles. We believe that these aspects of our classification may help managers. recognize the pitfalls they may encounter when they implement quality management approaches. The framework can be a valuable tool and foster further research in the area. For example, Barley and Kunda (1992) identify several orientations (tenors) that prevailed within the successive normative and rational waves. Similarly, it might be possible to identify subgroups within the two schools that would display some distinct characteristics. Moreover, the proposed framework could help researchers and managers examine quality management or TQM implementations' successes and failures in terms of the chosen approach, whether rational or normative, thus avoiding sterile debate about "true" TQM approaches. Does one approach guarantee better chances of success or is it preferable to refer to best-fit approaches? That is, depending on some organizational characteristics (type of structure, leadership style, size, etc.), one approach could potentially increase the chances of success. We are convinced that the discourse surrounding quality has shifted considerably over the years. Moreover, we have provided evidence that substantiates the existence of such "shifts." However, much work remains to be done. For instance, argument analysis or other forms of content analysis could strengthen the evidence in favor of our proposition. With the theoretical analysis presented here, we sought to show how the evolution of the discourse surrounding quality allows for a better understanding of some of the criticism levelled at it. We hope that this fosters a climate of exchange for proponents and opponents of TQM. We firmly believe that it is not by rejecting criticism or by responding with truisms and equivocation that the cause of quality will be advanced. We believe that criticism should be encouraged, analyzed and classified to better organize ongoing discourse. Our intention was to highlight the risks inherent in diluting the concept of quality within a discourse that is both too encompassing in its grasp and overly reductive in its applications. We wanted to reduce the frustration that managers and employees have when confronted by TQM so as to reduce the chances that they will dismiss it as yet another management fad. On the one hand, we believe that what happened to the discourse surrounding quality management may reflect a more general process: the discursive creation of a "miracle cure." If this is the case, managers should watch for the warning signs of this phenomenon when they try to select a new management approach. On the other hand, a better understanding of the historical reasons and of the discursive processes which have led to the misappropriation of the concept of quality may prevent managers from becoming discouraged by the failures of so many TQM programs.