Can We Apply the Theoretical Conceptualization to The New Unions in South Africa

In the late 1980s and early 1990s, an international debate took place among scholars active in and/or researching the new labor movements that had emerged in "developing" countries as divergent as Brazil, the Philippines, South Africa and South Korea (Scipes, forthcoming; see Lambert and Webster, 1988; Munck, 1988; Waterman, 1991a, b). These scholars were trying to theoretically understand the new phenomenon, and advanced the concept of "social movement unionism" to understand these new labor movements.

However, there were different understandings of what was meant by "social movement unionism," and the debate was an effort to refine the conceptualization for possible further generalization. The importance of this effort was to try to learn from the successes so as to inform other, newer labor movements, so they could use the knowledge developed from earlier efforts to inform the current ones, and hopefully advance them.

The final papers in the debate were published in 1992, when Kim Scipes published two articles that reviewed the development of the debate and refined the conceptualization of three different types of trade unionism-business, political, and social movement unionism (Scipes, 1992a)-and then differentiated social movement unionism from Marxist approaches (Scipes, 1992b). Scipes based his theoretical conceptualization on his early research on the Kilusang May Uno (KMU) Labor Center of the Philippines, and then followed these articles with a monograph on the KMU (Scipes, 1996) that was informed by his conceptualization of social movement unionism.

While Scipes believes the concept of social movement unionism could be applied to other emerging labor movements in addition to the KMU-he specifically mentions CUT (Central Unicado do Trabalhadores) in Brazil, COSATU (Congress of South African Trade Unions), and KTUC (Korean Trade Union Congress) in South Korea (Scipes, 1992a: 123, f.n. 6)-no one has tried to apply this theoretical conceptualization to any of these other labor movements.^¹In other words, despite his in-depth study of the KMU and its support for the conceptualization, it is not yet clear that the concept of social movement unionism can be applied to any of these other new labor movements, much less further generalized. However, the importance of the debate, although since muted, has not gone away.

To advance this debate, as well as to inform labor researchers and activists, I test Scipes' conceptualization in this paper by examining the new black trade unions that emerged in South Africa between 1973-1992, and specifically the unions that joined to create COSATU (Congress of South African Trade Unions) in 1985: is this a valid conceptualization, can it be generalized to include these black unions in South Africa and, if so, what can be learned to help other emerging labor movements?^²I ground this examination in both labor movement and social movement theory, and then place the emergence of these unions in their socio-historical context, before examining black workers' mobilization in that country.^³

Labor movement theory allows me to suggest how researchers should approach the study of labor movements, both regarding how to conceptualize labor movements as well as how to understand their development, and it allows me to theoretically differentiate the types of trade unionism practiced by unions at the heart of a labor movement. I then argue that labor movements are developed by their members making conscious choices, and that they are not merely reflections of structural situations or reflections of changes in structural situations.

But labor movement theory assumes that workers are already in motion-it does not tell us how they get active. I discuss general social movement theory to understand how social movements emerge.^⁴I specifically look at how individuals join with like-minded others to form groups, create a collective identity for each group, and then engage in collective activities. From this, I look at how they understand these actions and how they communicate their understandings to others in attempts to enlarge their forces and counter opposing social forces. This helps us to understand how labor movements emerge and then create the ground work for their further development as social order-challenging labor movements.

I follow this with an account of the particular context in which this labor movement has emerged. This mobilization must contextualized, because mobilization takes place within a particular socio-historical context. Without having some sense of the context in which they operate and then how workers choose to respond to that context-first individually, then collectively-we cannot understand the processes by which they made the choices they did regarding trade union conceptualization and the development of their unions and labor movement.^⁵

My specific methodological approach is thus: I examine key monographs of the extensive literature that has developed on COSATU, its component unions, and the broader labor movement in which it is located to try to discern their development patterns and to test these theoretical claims derived from the discussion. To contextualize this study, I place the mobilization that took place within South African efforts to industrialize, and then place these industrialization processes within the racial dictatorship of South Africa, "apartheid." I then examine the mobilization processes by workers that created this labor movement and its component organizations. By focusing on these developmental processes, I can then analyze workers' efforts in light of Scipes' conceptualization of social movement unionism, and then evaluate whether they confirm, challenge and/or refine his conceptualization. From that, I then suggest the findings that appear to be of most use to emerging labor movements.

COSATU and the labor movement in which it and its unions are part are excellent subjects to test Scipes' conceptualization of social movement unionism. First, and most importantly, international recognition among labor activists and writers is that COSATU is one of the most dynamic labor movements in the world. Secondly, because of the close linkages between COSATU and a number of academics/intellectuals in the country, there is an extensive literature available to use: COSATU and its component unions arguably have been studied by contemporary researchers to an extent unmatched by any labor movement in the world. Third, these unions have been successful in helping overthrow a racially-based dictatorship, as well as improving pay, benefits and working conditions for workers in workplaces. And fourth, because South Africa is at a higher level of economic development than is the Philippines (Broad, 1987), establishment of social movement unionism as a valid theoretical conceptualization would suggest the validity of generalizing this concept at least to labor movements in the newly industrializing countries, which would include CUT and KTUC.

In short, Scipes' conceptualization of social movement unionism would be considerably enhanced should it be confirmed by a study of the new unions of South Africa. This, in turn, would support efforts to generalize social movement unionism to other countries. It is to such a study that I now turn.

In this literature review, I examine relevant writing in labor movement and social movement theory.

Labor Movement Theory: Different Approaches to Labor Movements

There are different ways of looking at labor movements, and Simeon Larson and Bruce Nissen (1987) discuss how labor theorists have done this. First, they note that there is not just one theory of the labor movement, but there are a number of theories. They also point out that theorists have concerned themselves with different questions-some with the origin of the labor movement, some with the role of unions in an industrial society, some with ideological direction of labor organizations, and some with their ultimate goals, etc. Accordingly, Larson and Nissen write that "not only are there sharp differences over the function and content of labor movement theory, but it is virtually impossible to integrate the various theories into one overall theory of the labor movement." They suggest that it makes more sense to develop a typology of labor theories, and they briefly discuss Mark Perlman's fivefold classification. However, for their purposes, "we have found it more useful to group the theories according to the overall social role each [theorist] assigns the labor movement" (Larson and Nissen, 1987: 3-4). They see seven different social roles that theorists have given the labor movement:

(1) an agent of revolution, (2) a business institution for economic protection of its members, (3) an agent for extending industrial democracy, (4) an instrument for achieving the psychological aims of groupings of workers, (5) an agent for moral and spiritual reform; (6) an antisocial, destructive monopoly, or (7) a subordinate mechanism with 'special interest' functions in a pluralist industrial society (Larson and Nissen, 1987: 4).

But following Bonnell (1983) and others such as Marian Golden (1988) and Kim Scipes (1996), I suggest it makes more sense to try to understand how the workers-particularly activist workers-have looked at labor movements rather than how the theorists have looked at them. In other words, rather than placing a labor movement into a theoretical straight-jacket according to the ideological presuppositions and analysis of a particular theorist, I think we should look at them as do workers: I argue that workers see trade unions and labor movements as having multiple purposes,^⁶ and that theorists must recognize this multiplicity.

Modifying the above list somewhat, I suggest that workers can see labor movements as being (1) agents of social change, (2) means to improve their economic situation; (3) initiators of industrial democracy; (4) forces that improve the psychological well-being of workers; and (5) vehicles of moral and spiritual reform.^⁷Further, a particular worker can see a labor movement serving one or more purposes at any one time and, of these purposes, this worker might differently prioritize the importance of one or more at any particular time. Thus, recognizing this multiplicity of approaches that are common among workers, I'm going to argue the necessity of taking a worker-centered approach to the study of labor movements.

At the same time that we recognize the multiplicity of purposes of labor movements as seen by workers, we must also recognize that there is not just one common group or "class" interest that exists simply by workers participating in the social relations of production, distribution or exchange (cf., Marx and Engels, 1978 [1848]); i.e., there is not an automatic workers' group interest nor consciousness that exists. A collective consciousness must be developed and collective interests must be created-they cannot be assumed (Melucci, 1989).^⁸

Labor Movement: A Conceptualization

So far, however, I've assumed common agreement about what is a "labor movement"-it is now time to develop this conceptualization before going further. Accepting McAdam and Snow's definition of a social movement as "a collectivity acting with some degree of organization and continuity outside of institutional channels for the purpose of promoting or resisting change in the group, society or world order of which it is a part" (McAdam and Snow, 1997: xxiii), I argue that a labor movement is one type of social movement-at least during its initial period of emergence and challenge to the established social order-and what makes it particularly important is its social location: a labor movement's primary organizational manifestation is located within the production, distribution and exchange sphere of society. Thus, a labor movement has something no other social group has: the ability to stop operations from inside the production-distribution-exchange process (Aronowitz, 1973). It is this position and potential social power that makes labor movements important.

Organizationally, the heart of a labor movement are trade unions and agglomerations of trade unions that are joined by a labor center which, in turn, works to further unify and strengthen the member unions.^⁹ Around these labor organizations are located supportive intellectuals, individuals, and other organizations; and these organizations include both those that serve the labor movement directly-such as educational institutions designed to provide labor education for labor center members-and those that choose to ally politically with the labor movement, indirectly strengthening the social power of the labor movement. It is this combination of labor center and supporters, mobilized into a mutually-reinforcing social network, that comprise a labor movement. Accordingly, labor movements are not simply collections of worker: labor centers cannot survive without the support of a larger social community, however constructed.

At the same time, there is a range of involvement and interest in a labor center or in a labor movement as a whole among members: I certainly do not want to argue that each member has a high degree of interest and involvement, no matter how much I might wish it were so. Accordingly, I classify labor movement members into one of three categories: activists, participants, and bystanders. "Activists" are those with high involvement and interest, and who play a leadership role. These leadership efforts can be formal, such as serving as an officer of a union or of a supporting community-based organization; they can be informal, such as volunteers-and they can support or oppose the formal leadership; and/or they can act as "bridges" (see Robnett, 1996) between the leaders and members, or between labor centers and allied organizations. "Participants" are those members who can be mobilized to participate around particular issues or campaigns, but generally do not initiate self-involvement. And "bystanders" generally have low interest and low involvement-they are generally those who Mancur Olsen (1965) calls "free-riders"-although they can be mobilized at particular times and around particular interests should an effort be made to do so. And members can change their level of involvement at any time-these are not static or immutable categories.

Understanding the Development of Labor Movements

Accordingly, I am going to suggest-following Golden (1988)-that researchers can best understand the development of a labor movement by focusing primary attention on activists. I argue that activists play key roles because they are the ones that do the conceptualizing and thinking for the movement and, through the framing process (Snow, et. al., 1997/1986), are the ones that interpret the situation to members and to the outside public. Therefore, activists-both inside the labor center and throughout the supporting network-are central to the development of the labor movement as a whole, serving specifically to mobilize their members while attempting to neutralize and/or counter movement opponents.

At the same time, however, we cannot collapse our understanding of a labor movement to the activists. No matter how good or how innovative activists are, unless a substantial number of the members respond affirmatively to their efforts, there is not a labor movement but simply a collection of activists and/or organizations-and these are not the same! As Gay Seidman perceptively writes, "? while individual activists and clandestine groupings may help shape the discourse of an organization, they can hardly determine how that discourse is received or acted upon" (Seidman, 1994: 41). Thus, activists must present a program sufficient to motivate the activation and mobilization of rank and file (i.e., non-activist) workers.

Labor movements derive their power from their ability to mobilize large numbers of people as a unified force to disrupt production, distribution and/or exchange, and to withstand counterattacks from capital and/or the state and/or their respective allies. Accordingly, labor movements at all times are potentially at risk from state repression and/or countermovements that can challenge the interpretations of activists and weaken or destroy connections (and mobilization networks) between activists and members of the movement (see Meyer and Staggenborg, 1996).

Concurrently, despite a general tendency to use the term labor "movement" almost unconsciously to refer to any collectivity of labor organizations, I argue this is unsupportable: not all collectivities of labor organizations are labor movements. A labor movement is qualitatively different than a collection of labor organizations: a labor movement is the combination of a labor center and its affiliated organizations and their supporters mobilized into a mutually-reinforcing social network. And there can be more than one labor movement in any single country.

Types of Trade Unionism

Key to understanding labor movements is to recognize the type of trade unionism practiced by the labor organizations at the heart of each movement. Three radically different approaches to understanding this are offered by the works of Selig Perlman, Victoria Bonnell, and Kim Scipes. These three authors are chosen because their approaches to trade unionism exemplify the three types of trade unionism-economic, political, and social movement unionism-that Scipes delineated in his 1992 articles (Scipes, 1992a: 124-134; 1992b: 82-87).

Selig Perlman (1968/1928), in his comparative study of labor movements in Germany, Britain, the US and Russia during the nineteenth and early twentieth centuries, argued that activists in the US chose a conservative ideology in the face of particular structural conditions and in response against radical intellectual intervention in their movement. He conceptualized unionism in the US as an economic institution, based on "job consciousness" that limits itself to "wage and job control" (Perlman, 1968/1928: 169). Perlman saw labor's vision limited to survival in a terribly hostile social-political environment, and one in which maintaining organizational survival was extremely difficult (Perlman: 160-169). Key to its survival, in his opinion, was in rejecting radical intellectuals and their various projects, including a labor party. Success in this project meant the institutionalization of labor.

Perlman's analysis, although he might not agree with all of the conceptualization, is of a trade unionism that Scipes labels "economic unionism":

I define 'economic' unionism as being unionism that accommodates itself to, and is absorbed by, the industrial relations system of its particular country; that engages in political activities within the dominant political system for the well-being of its members and its institutional self but generally limits itself to immediate interests ... (Scipes, 1992a: 126; 1992b: 86).

Perlman's analysis is implicitly challenged by Victoria Bonnell's (1983) study of workers in St. Petersburg and Moscow between 1900-1914. Bonnell has argued that activists chose a radical ideology in the face of particular structural conditions and in support of radical intellectual intervention in their movement. When trying to identify how Russian workers could develop "revolutionary consciousness," she identified "endogenous" and "exogenous" understandings within social theory:

But how do workers arrive at this rejection of the prevailing arrangements, and how do they develop an alternative vision? These issues are often conflated, but from an analytic point of view they represent distinct if interrelated problems. It is conventional in the literature to draw a distinction between two basic approaches to these issues: theories that focus primarily on revolutionizing circumstances external to the workers themselves and their milieu, and those that locate the roots of rebellion in the workers' own experiences acquired at the workplace, in the community, or in society. For the sake of brevity, I will call them exogenous and endogenous, respectively. They are not mutually exclusive, and elements of both can be found in some studies (Bonnell, 1983: 7).

She points out that "Exogenous theories share a common assumption that workers cannot develop revolutionary consciousness on their own," but require outside intervention by forces such as a political party or radical intellectuals. Endogenous theories, on the other hand, argue that workers are revolutionized by their own experiences, without the intervention of an outside agency (Bonnell, 1983: 8).

Bonnell's study, in fact, was designed to resolve the dispute: she concluded that workers organized themselves through creating a new understanding of their collective identity and then, from that collective identity, evaluated interactions with capital and the state and interpretations of these interactions by radical intellectuals. And, then, seeking a radical change to their situation, they decided to subordinate themselves to the intellectuals' organization with the program that they felt best met their needs. Success of the intellectuals' project in Russia also resulted in the institutionalization of labor, albeit under a more friendly, but ultimately controlling, regime.

Bonnell's analysis, although she might not agree with all of Scipes' conceptualization, is of a trade unionism that Scipes labels "political unionism":

I define 'political' unionism as unionism that is dominated by or subordinated to a political party or state, to which the leaders give primary loyalty-and this includes both the Leninist and "radical nationalist" versions. This results in generally but not totally neglecting workplace issues for "larger" political issues (Scipes, 1992a: 127; 1992b: 86).

And in explicit distinction from either Perlman or Bonnell, Kim Scipes' study of the radical wing of the Philippine labor movement delineates still another type of trade unionism, social movement unionism. In his study of the Kilusang Mayo Uno (KMU) Labor Center, Scipes (1996) argues that activists chose to develop a different type of trade unionism, wherein they are ultimately masters of their own fate, rather than relying on intellectuals and/or intellectuals' organizations of either the right or the left. In other words, Filipino activists in the KMU have been willing to consider various approaches to their particular social conditions and then have decided, on the basis of their own understandings, their course of collective action. Thus, activists created a type of trade unionism that recognized their particular situation, but did not straight-jacket their efforts by limiting their conceptualization to either an economic or a political conceptualization nor did they act only in response to intellectuals. Thus, based on this, Scipes proposes a conception of unionism that differs from Perlman's economic conceptualization or Bonnell's political conceptualization, and that he-along with others-calls social movement unionism.

Social movement unionism is a type of trade unionism that differs from the traditional forms of both economic and political unionism. This type sees workers' struggles as merely one of many efforts to qualitatively change society, and not either the only site for political struggle and social change or even the primary site. Therefore, it seeks alliances with other social movements on an equal basis, and tries to join them in practice when possible, both within the country and internationally.

Social movement unionism is trade unionism democratically controlled by the membership and not by any external organization, and recognizes that the struggles for control over workers' daily work life, pay and conditions are intimately connected with and cannot be separated from the national socio-political-economic situation. This requires that struggles to improve the situation of workers confront the national situation-combining struggles against exploitation and oppression in the workplace with those confronting domination both external from and internal to the larger society-as well as any dominating relations within the unions themselves. Therefore, it is autonomous from capital, the state and political parties, setting its own agenda from its own particular perspective, yet willing to consider modifying its perspective on the basis of negotiations with the social movements that it is allied with and that it has equal relations (Scipes, 1992a: 133; 1992b: 86-87).

It must be noted that, however, in each case referred to above, workers are considered active subjects and are not merely determined by the situation that they confront.

Nonetheless, I suggest it is the ideological conceptualization of trade unionism, by activists who mobilize workers to unite and form labor collectivities, that plays a central role, if not the central role, in determining the type of trade unionism that will guide particular labor movements. In other words, while it may be structural conditions that cause workers to decide to act collectively to create and/or join labor collectivities, it is the ideological conceptualization of trade unionism by activists within the collectivities, and their abilities to convince co-workers and supporters of the superiority of their approach, that determine the direction of each collectivity. Thus, the type of trade unionism will shape the direction and possibilities of the labor movement in how it interacts with its larger environment.

But labor movement theory assumes that workers are already in motion-it does not tell us how they get active. I first discuss social movement theory to understand how social movements emerge in general, and then follow this with an account of the particular context in which this labor movement emerged.

Social Movement Theory

Within the literature on social movement emergence, there is material directly relevant to how labor movements emerge, and I want to incorporate this knowledge into our understandings of labor movements. Two general approaches are structural explanations and cultural explanations (see McAdam, 1994).

Emergence-Structural Explanations

The literature explaining why social movements emerge traditionally focuses on structural changes in the society, and arguably the pre-eminent work is by Frances Fox Piven and Richard Cloward who write, "The emergence of popular uprisings reflects profound changes in the larger society" (Piven and Cloward, 1979: 7). Focusing on "protest movements" among the poor stratum of the working class, Piven and Cloward divide theories into those that see pressures on people forcing eruptions on the one hand, and those that see breakdowns in regulatory control of society, allowing eruptions to occur, on the other. However, they see these two factors as being connected. Focusing on the impact of sharp economic change, they point out that

But even if sharp economic changes or the breaking down of daily life do explain why at least some social movements emerge, this does not tell us how movements emerge.

Emergence-Cultural Explanations

However, even Piven and Cloward do not see structural explanations as sufficient; they add cultural ones as well to their approach. Sharp economic changes generally do not result in the rise of protest movements; there must be changes in how people understand their respective situations: "For a protest movement to arise out of these traumas of daily life, people have to perceive the deprivation and disorganization they experience as both wrong, and subject to redress" (Piven and Cloward, 1979: 12). In any case though, Piven and Cloward see these changed meanings as products of external, structural factors: large scale social distress (muting "the sense of self-blame, predisposing men and women to view their plight as a collective one" and blaming their rulers "for the destitution and disorganization they experience"); failures of dominant institutional arrangements; and/or splits within elite groups (pp. 12-13). However, and unfortunately, they have no suggestions as to the processes by which individuals come to recognize these situations as "wrong," how they join to address these concerns, how they make decisions to act, how they withstand repression, and how they continue challenging the defined "wrong" over time.^¹⁰

Alberto Melucci (1989; 1995) argues that a major weakness of social movement research is that movements have been treated as empirical realities, as though they already exist. He believes researchers need to recognize the constitutive processes by which they are constructed; i.e., that if we want to understand the emergence and development of social movements, we should not treat a movement as a given, but rather focus on how it has been built.

Melucci uses cultural explanations to do this. He focuses on the importance of individual needs as to why social movements emerge, as people recognize the need to think about society and their social experiences differently than has been done in the past [see also Mueller (1994)].^¹¹ Further, he recognizes the process of individual activity in the creation of collective identity, and he argues that creation of a collective identity is a precursor for consciously collective action.

Melucci defines collective identity as "an interactive and shared definition produced by several individuals (or groups at a more complex level) and concerned with the orientations of action and the field of opportunities and constraints" in which the proposed action is to possibly take place (Melucci, 1995: 44). He sees this as a process that includes a "notion of unity, which establishes the limit of a subject and distinguishes it from all others?" but he also recognizes that "a certain degree of emotional investment" is required in creating collective identity, which means that participation in collective identity can never be totally negotiable. He continues: "The notion of identity always refers to these three features: the continuity of a subject over and beyond variations in time and its adaptations to the environment; the delimitation of this subject with respect to others; the ability to recognize and be recognized" (Melucci, 1995: 45).

The process of creating a collective identity, an on-going process, is important: success allows potential actors to act together in ways that enable them to be in control of their own actions, and that is why it is important to understand on a theoretical level:

In other words, Melucci recognizes the interactive process between individuals that leads to collective identity and then, through on-going negotiation and renegotiation, the process that can lead to collective action once the group decides to act.^¹²But he also recognizes the interactional process that occurs between conflicting groups, and then the negotiation and renegotiation within each group as a result of that conflict (Melucci, 1989).

Hank Johnston, Enrique Laraña and Joseph Gusfield take an understanding of collective identity and focus on its interactions between individual identity and public identity (Johnston, Laraña and Gusfield, 1994). But they don't limit themselves to describing this; they go beyond this to try to connect ideology, grievances and collective identity as precursors to collective action. They admit that grievances and collective identities are not the same, but they claim that their relationship is close because "how social movement adherents think about themselves is structured in important ways by how shared wrongs are experienced, interpreted, and reworked in the context of group interaction" (Johnston, Laraña and Gusfield, 1994: 22).

Anthony Marx' 1992 study of the internal opposition within South Africa develops a parallel argument, and then takes it farther. He focuses on different ideologies and how they developed over time, and shows how particular ideologies were surpassed as the organizations that had mobilized around them had failed to resolve particular problems facing the movement based on that particular ideology. In my view, ideology is important because "? it is a way of experiencing reality, a form of 'practical consciousness', based on a combination of culture and individual and historical processes" (Marx, 1992: 236).

Ideology helps one to understand, order and operate in her world. If one accepts that whites are racially superior to blacks, then that person's behavior, consciousness and general approach to life-especially in relation to the "other"-will qualitatively differ from one who accepts that all people are equal as human beings. Or, if one accepts "free enterprise" capitalism as being desirable, then that person will respond differently to strong unions and strikes than one who has some form of socialism as his goal. Likewise, the choice of a particular type of trade unionism-economic, political or social movement unionism-will effect how one acts as a trade unionist (and, when consciously chosen for a union, how a particular union will act) regarding issues, such as race relations in the union, workplace and the community (Scipes, In progress). Ideology can join with grievances to help create a collective identity, and the choice can effect individual and organizational behavior and development.

The connection between the creation of collective identity and grievances has also been mediated by the social psychological dynamics of collective attribution and social construction, what David Snow and colleagues have labeled "frame alignment processes" (Snow, et. al., 1997/1986). They argue that "grievances or discontents are subject to differential interpretation" and thus, it is not the particular problem that is so important, but rather how it is interpreted by potential actors. And adding to the complexity of the issue, they recognize that interpretations can vary "across individuals, social movement organizations, and time" (Snow, et. al., 1997/1986: 236). And although they refer to social movement organizations (SMOs) instead of broader social movements, as I do in this paper, their point is quite clear as to the importance of understanding and dealing with these processes: "The ways in which SMOs manage and control these frame vulnerabilities, as well as interpretive resources in general, thus seem as crucial to the temporal viability and success of an SMO as the acquisition and deployment of more tangible resources ..." (Snow, et. al., 1997/1986: 250).

Building on the different understandings of framing processes and the social construction of collective identities, Hunt, Benford and Snow try to address how they interact. They examine the way that "frame alignment processes can be conceived as rhetorical strategies to affect the alignment of collective and personal identities" (Hunt, Bradford and Snow, 1994: 191). They point out that there are three basic "identity fields" of actors-"protagonists" or advocates of movement causes; "antagonists" or opponents of these causes; and "audiences" who are the uncommitted for which the protagonists and antagonists compete in the battle to win support and generate political power in furtherance of their cause-and that they each interact so as to help develop individual and collective identities which, they claim is important "to understand the emergence of particular expressions of collective action" (Hunt, Benford and Snow, 1994: 192-204).^¹³

But it is Carol Mueller who ties these concepts of collective identity and framing together. In an analysis of Melucci's work through examining the origins of the women's liberation movement, she notes that Melucci's focus is on "submerged networks" of "small, separate groups engaging in cultural experimentation," and that

Further, she notes, "the status quo must be challenged at the cultural level in terms of its claims to legitimacy before mass collective action is feasible" (Mueller, 1994: 239).

Mueller, building off work by Bert Klandermans' conception of levels of social construction, suggests four levels of analysis: "public discourse, persuasive communication initiated by movement organizations, 'consciousness raising' from participation in episodes of collective action, and the creation of collective identities in submerged networks" (Mueller, 1994: 258).^¹⁴Obviously the processes of social construction will move back and forth, largely depending on the situation examined, but what her analysis suggests is that the creation of collective identities is crucial, and out of this, people engage in collective action from which they further learn, and these understandings are communicated to other organizations and then, ultimately, to the public.

Melucci describes the overall process by which groups produce collective action:

If we couple Johnston, et. al.; Anthony Marx; Snow, et. al; and Hunt, et. al.'s understandings with Melucci's and Mueller's, we see that social movements develop through cultural processes. These cultural processes begin with individuals joining together to form groups, and then creating a collective identity for each group. These groups, through a process of interaction, negotiation and conflict, then frame the situation so as to legitimate collective action. As they act, and learn from their taking such action, these groups further frame their efforts in a way so as to legitimate their actions to themselves (i.e., their members), other organizations and the public at large. They attempt to do so in a way that builds additional support for their actions among the larger public, while limiting or negating efforts by opponents that seek to counter or delegitimate these collective actions.

Application of these Social Movement Theories to the Labor Movement

While the importance of sharp economic changes recognized by Piven and Cloward obviously can be related to the rise of labor movements, the work on collective identities is more difficult to relate. The main reason is because the theorist who is key to understanding the importance of collective identities, Alberto Melucci, argues that "this process of constructing collective identities is a unique characteristic of highly complex societies" (Mueller, 1994: 238). A reading of Melucci leaves no doubt that he is talking about contemporary, post-industrial capitalist societies when he uses the term "complex" (Melucci, 1989).

However, Carol Mueller specifically ties these processes with labor movements. She argues that

I agree with Mueller: we cannot understand the process by which labor movements have risen without understanding how the roles of new collective identity and institution building have affected this process.

However, we also have to recognize, especially in labor struggles, that collective identity is not solely created through conscious, rational, cognitive processes-it can also be created through partaking in collective action, which can be emotional but is in many cases, non-rational when measured by the traditional standard of self-interest. Rick Fantasia (1988) uses a wildcat strike in a factory to suggest how "cultures of solidarity"-"more or less bounded groupings that may or may not develop a clear organizational identity and structure, but represent the active expression of workers' solidarity within an industrial system and a society hostile to it"-can develop among a group of workers.

This can be clearly seen in his analysis of one stage of the wildcat strike: "At this stage [after the men had taunted managers who had threatened to call the police-KS], the workers had defied company authority at its highest level and a strong sense of solidarity had emerged?. Solidarity among the workers was not an a priori 'fact', but grew out of this interactive process of negotiation between workers in their confrontation with authority" (Fantasia, 1988: 88).

The point here being that while collective identity is crucial to the rise of labor movements, it is not always created through a deliberative process. Workers are confined within certain contexts during their working time, and situations can develop, as Fantasia has shown, that workers are forced to respond simply out of their gut: will they stand in solidarity with co-workers against management, risking their jobs, or will they side with management? No one knows beforehand which way people will jump. But when workers respond solidaristically , this leads to creation of collective identity among those that stood together. They then can decide if they want to go farther or not-and develop a more conscious collective identity,^¹⁵ and possibly even engage in collective action-or they can choose not to do so but, for workers, deliberation is not always a choice in the initial mobilization process.

And now that we have examined both labor and social movement theory, it is now time to shift to our empirical evidence. To be able to evaluate whether social movement unionism exists in South Africa or not, we must place the development of the new unions and their coalescing into COSATU in November 1985 into a larger social context. This mobilization must contextualized, because mobilization takes place within a particular socio-historical context. Without having some sense of the context in which they operate and how workers choose to respond to that context-first individually, then collectively-we cannot understand why they made the choices they did regarding trade union conceptualization and the development of their unions and labor movement.

Mobilization in South Africa

The mobilization of workers in South Africa-particularly in manufacturing-must be placed in a historicized context which, in this case, includes industrialization under a racial dictatorship.

However, when considering the impact of industrialization, we must be careful. Although some analysts of industrialization in the "developing" countries argue that industrialization in general-especially rapid industrialization-causes mobilization (e.g., Seidman, 1994), others argue that the type of industrialization can either inhibit or promote mobilization-thus whether the industrialization is oriented for export or domestic usage can effect mobilization (e.g., Deyo, 1989; cf. Hutchison, 1992; see also Gereffi, 1990). Based on Scipes' work in the Philippines (Scipes, 1996), I would also add that levels of industrialization effect mobilization, so that mobilization can be effected by whether a country is a NIC (newly industrializing country) or at a different level of development. Therefore, I argue that both levels of industrialization and types of industrialization within levels can be seen as independent variables that have differential effects on mobilization. However, what structural models of mobilization cannot do is determine the type of mobilization that takes place when workers do decide to act.^¹⁶

Industrialization in South Africa

Industrialization in South Africa generally began in the 1920s, as the government began what would later be known as an ISI (Import Substitution Industrialization) program. This started to create an economic base of light manufacturing, although this shifted to include heavy industry after World War II. This was an economic program of industrialization, but also a political program as well: "... the importance of the state to South Africa's industrial development in the postwar era should not be underestimated," and under this program, "... controlling black workers and supporting Afrikaans-speaking whites were at least as important to the National Party as economic growth." But regardless of its motives, "the South African state had, by the early 1960s, laid a basis for rapid industrial growth" (Seidman, 1994: 75).

In the 1960s, this industrial growth took off. This was largely based on foreign capital. As Seidman notes, "... the industrial boom of the 1960s, depending as it did on foreign technology and capital to expand into new sectors, was dominated by large firms, closely tied to international investors" (Seidman, 1994: 81). This caused rapid growth of the manufacturing sector and large new factories were erected (which would bring large numbers of workers together in increasingly larger units of production) (Bonner, 1987: 55). Between 1960 and 1970, the economy grew by 56%, "expanding at a rate matched only by Japan"; foreign direct investment doubled, growing from R (Rand) 1819 million to R3943 million (MacShane, Plaut and Ward, 1984: 20).

However, Gay Seidman discusses how contemporary industrialization has differed from the earlier industrialization experiences:

In any case, we can see the extent of this growth from the following figures: in 1960, the manufacturing sector had a total output of R7.121 billion; by 1970, this reached R16.267 billion; by 1980, this reached R27.342 billion.^¹⁸In fact, by 1980, the output value of manufacturing surpassed that of mining, which has traditionally been South Africa's leading economic producer (COSATU, 1992: 43).^¹⁹Accordingly, employment in the manufacturing sector reached 28.8% of total employment by 1980 (COSATU, 1992: 45). By 1989, South Africa's Gross Domestic Product (GDP) had reached $80.370 billion, and its GDP per person was $2,296 (COSATU, 1992: 7).

Because of Apartheid, economic benefits were stratified by race. In 1972, average money wages of Africans working in manufacturing were only 17.12% of whites working in manufacturing (MacShane, Plaut and Ward, 1994: 51). However, change was in the air.

Over a slightly different period, we can see that there were 300,000 jobs for unskilled workers in manufacturing in 1965, but only 200,000 jobs in 1985. At the same time, in 1965, most African male workers were unskilled, but by 1985, most African males were semi-skilled (COSATU, 1992: 46). This demand for skilled labor in manufacturing (as well as the increasing strength of the new unions who organized these workers) led to an relative increase in black wages: where average money earnings of blacks had been 17.12% of white workers in 1972, it was 31.60% by 1982 (MacShane, Plaut and Ward, 1984: 51).

However, in the 1980s, manufacturing sector growth all but stopped: between 1980 and 1990, the yearly growth rate was 0.1%. Manufacturing output, which was R 27.342 billion in 1980, totaled only R 27.596 billion in 1990 (COSATU, 1992: 48 and 49, Figure 51). The annual average growth rates for major sections of the manufacturing sector declined between 1981-1989: -8.9% for transport equipment,-4.93% for motor vehicles, -4.36% for machinery and equipment, -3.12% for wearing apparel, -2.53% for textiles, and -1.07% for industrial chemicals, while tobacco production increased 2.77%, beverages, 3.38%, non-ferrous metals, 3.38%, and paper, 4.20% (COSATU, 1992: 51). Relative employment also decreased in manufacturing, from providing 28.8% of all jobs in 1980 to providing only 26.6% of all jobs in 1989 (COSATU, 1992: 45, Figure 46).

Apartheid

However, industrialization does not take place in a vacuum: it takes place within a specific social order, usually within a particular nation-state. This is especially important to recognize in considering developments among black workers in South Africa.

To a degree unmatched in the contemporary world, South Africa was a racial dictatorship:

As one contemporary report characterized it, "In South Africa, a 4½ million minority rules over a 21.5 million black majority. In order to perpetuate this system, the racist minority authorities have, through a series of enactments, established a system of untrammeled power which affects every aspect of human life" (ICFTU, 1984: 7).

While the social brutality of apartheid was unparalleled, blacks had fought against racial oppression since its introduction into the country. To understand black workers' mobilization, we must place it within the context of the struggle against racial oppression, which developed particularly since the late 1940s, when the system of apartheid was introduced by the Nationalist Party (ICFTU, 1984: 7).

The BC conceptualization of "race" recognized that race was a social construct, not a biological reality, and the movement's leaders chose to define it in a way that challenged the government's definition. For example, they included Asians and "coloureds" in the category of black, specifically undercutting the government's efforts to separate the different groups through racial categorization. And while BC excluded whites from their movement, it was a situational response rather than a categorical one, as they saw that all racial categories would dissolve after "liberation."

The BC movement was ultimately an idealist project of achieving social change by reshaping values and attitudes. However, when the material conditions of society changed, BC provided a way of interpreting the changes, and did so in a way that was grasped by the large number of blacks, and particularly poor, black, urban youth. "When a wider array of the oppressed were aroused by changing material conditions to join the opposition, the mood and language of Black Consciousness proved to be more pervasive than the breadth of formal BC affiliation had suggested. At least among urban youth, the days of bowing and scraping were long gone?" (Marx, 1992: 61).

By the mid-1970s, material changes were widely affecting the population. The economy was drastically slowing, with a concurrent downturn in employment: where 2,850 new manufacturing jobs for blacks had been created each month during 1974, "fewer than half as many new jobs were created in the subsequent eighteen months" (Marx, 1992: 61). The burden of apartheid on the state's resources-duplicated administration, additional military and policing expenditures, restrictions on growth of domestic economic markets and skilled labor by blacks, and investment too inefficient to overcome the oil and arms embargoes-limited the state's ability to respond to the economic slowdown (Marx, 1992: 61).

And, of course, it was the poor and those at the bottom of the social order that the crisis struck hardest. General living standards declined, as the black population doubled in the largest urban centers but housing only grew in these areas by 15% during the 1971-75 period. Black secondary school enrollment leaped by 160% between 1970-75; in the black township of Soweto, outside of Johannesburg, "there were over sixteen thousand families for every high school, as compared with thirteen hundred families per high school in white Johannesburg" (Marx, 1992: 62).

Black youth began building their own BC organization, the South African Student Movement (SASM) in late 1975-early 1976. SASM became involved in opposing the state's new effort in early 1976 to have mathematics taught in Afrikaans-the "language of the oppressor"-in the schools instead of English. The secretary of a SASM in a high school was to be arrested by police in early June, but students fought the police, forcing them to retreat off campus. On June 13th, students called for a demonstration on June 16^th to protest the state's attack. The police responded harshly to the students' efforts: 25 students were killed, and local workers were also attacked without provocation. "? the casualties of June 16, 1976, were only the start of a violent conflict, exacerbated by discontent over material conditions and assertiveness encouraged by BC rhetoric?," and "Six days after the initial uprising, one hundred and thirty people were officially listed as having been killed" (Marx, 1992: 68).

Ultimately, the mobilization inspired by the June 16^th demonstrations ended by the end of 1977, as the BC movement had not built the organization necessary to advance its politics:

But what the BC movement accomplished was to break the chains of mental subjugation that had kept large numbers of blacks politically immobilized, and this change was most obvious in the urban areas; areas in which most of industry was located.

The 1976 students' uprising directly effected the emerging labor movement. "Following this, a new spirit of militancy and a younger generation of factory floor leaders schooled in this environment began to emerge" (Bonner, 1987: 58). Gay Seidman supports this: "? years later, workers on the East Rand attributed increased worker militance after 1976 to anger at repression and shame that students had taken the lead in resisting apartheid" (Seidman, 1994: 181).

The increase in political consciousness stimulated by the BC movement also resulted in a revitalization of the "Charterist" movement, the political mobilization headed by the African National Congress and motivated by the Freedom Charter. The concept of "national unity as a strategic necessity" began being adopted by a growing majority of the opposition movement after the uprising of 1976. "An ideological transition from a primary concern with values to a greater focus on mobilizing physical resistance was taking shape within the opposition, driven by the external pressures of state repression as these were assessed and interpreted" (Marx, 1992: 80).

The primary form that this development took was the creation of the United Democratic Front (UDF) in 1983. Where the BC movement had been led by blacks and for blacks, the UDF was a non-racial project that was intended to mobilize the greatest number of South Africans in the struggle against apartheid, and thus accepted anyone of any "race" that opposed apartheid. Its project was based on mass mobilization, and worked through community mobilization, publicity, and massive demonstrations.^²⁰

While the BC movement had been stymied by its lack of attention to organization building, the UDF was limited by its type of organization. Perhaps most importantly, as a political coalition, its multi-class membership immobilized it when facing class-related questions and, accordingly, then on how to proceed in the struggle against apartheid-officially, it was reduced to a politics of "least common denominator" in efforts to maximize oppositional mobilization. However, in practice, local UDF-affiliates mobilized extensively in the townships and communities and acted quite militantly at the local level. The other limitation was that because it was based in the townships and communities, and dependent on publicity through the mass media, it was very susceptible to repression by the state. Yet the UDF was able to mobilize increasing numbers of people across the country in militant political opposition to the apartheid state (Marx, 1992: 106-146)

Beginning in late 1984, a massive revolt by blacks across the country shook the state. Stimulated by the development of the UDF and worsening economic conditions, mass organized revolt emerged and was met by heavy repression. Between August 1985 and the end of that year, between 650 and 879 people had been killed; 371 by police. In July 1985, a partial state of emergency was declared by the state. This failed to halt the upsurge, was withdrawn for a few months, and then surpassed as governmental leaders imposed a total state of emergency nationwide in June 1986. This hampered the opposition, and in February 1988, the UDF itself was banned (Marx, 1992: 147-188).

It was out of the unrest that COSATU was launched in November 1985 (Baskin, 1991: 87-90).

The range of oppositional politics-along with those of an "independent worker" bloc-were reflected within the labor movement. BC-inspired unions stayed in black (as differentiated from non-racial) labor centers such as CUSA (Council of Unions of South Africa) and AZACTU (Azanian Congress of Trade Unions), and these two labor centers merged in 1986 to form NACTU (National Council of Trade Unions). Although NACTU has co-operated with COSATU to a greater or lesser extent over the years, it has remained a separate labor center. Nonetheless, in November 1985, a number of regionally-based general unions-unions not limited to any particular industry-inspired by the ANC/UDF movements, joined with the independent industrial-based unions of FOSATU (Federation of South African Trade Unions) to create a new labor center, COSATU (Congress of South African Trade Unions). And interactions between these different political ideologies and traditions-which played out differently depending on the issue at hand (Baskin, 1991: 101-104; and, more generally, see Kraak, 1993: 180-205)-helped shape the internal politics of COSATU and its organizational development.

With some idea of the political-economic situation, it is now time to focus on worker mobilization during the period of industrialization. I focus here on the black, but non-racialist, trade unions in South Africa that eventually came together to form COSATU in 1985.

Black Worker Mobilization

Although there had been African unions since 1919, the black union movement had been basically wiped out by the mid-1960s. Although never banned as an organization, the non-racial unions of SACTU (South African Congress of Trade Unions) were devastated when their leadership was "decimated by arrests, detentions and bannings": "The capacity to organize workers inside South Africa had been destroyed" (MacShane, Plaut and Ward, 1984: 118; Maree, 1987: 2). A listing of strikes between 1962-72 shows that only in one year, 1972, that more than 8,000 workers went out on strike nationwide (MacShane, Plaut and Ward, 1984: 20).

However, in January 1973, in Durban, 2,000 black workers at the Coronation Brick and Tile Company went on strike. This touched the nerve of other workers: within a month, over 30,000 workers were on strike in and around Durban-29 firms had been struck during January, and then strikes spread to Durban municipal workers-and strikes had begun to spread throughout the province of Natal (Institute for Industrial Education, 1979).^²¹Within the first three months of 1973, over 61,000 workers had been on strike, which was more than the total for the previous eight years (Baskin, 1991: 17-18). By the end of the year, around 100,000 workers in the Durban area had struck (MacShane, Plaut and Ward, 1984: 21).

"The most significant immediate consequence of the [Durban] strike wave was that masses of militant African workers poured into newly founded working class organizations in Durban and Pietermaritzburg," writes Johann Maree. "This heralded the start of the resurgence of African trade unionism as a whole?" (Maree, 1987: 2).

In an article providing an overview of the emergence of the new trade unions between 1979-1984, Maree delineated three periods of development. Between 1973-1976, there was the "struggle for survival" where unions emerged and then, toward the last part of 1976-after the students' uprising-were almost destroyed. In fact, union operations were "completely ruptured" in the areas around Cape Town and Johannesburg. Between 1977-1979, the new unions went through a period of "reconstruction and consolidation," and in April 1979, FOSATU (Federation of South African Trade Unions)-further discussed below-was founded. Also, in 1979, the state decided to recognize African trade unions after the Wiehahn Commission's first report. And then, between 1980-1984, the new unions expanded rapidly (Maree, 1987: 1-7):

During the four year period from the end of 1979 to the end of 1983, signed-up membership of the emerging independent unions went up more than fourfold from about 70,000 to almost 300,000?. But the unions made their most impressive headway in gaining formal recognition from companies. This is indicated by the enormous increase in the number of signed recognition agreements: between 1979 and 1983, they increased from a mere five to no less than 406 with the FOSATU unions accounting for 285 or seventy percent of the agreements (Maree, 1987: 7).

As stated above, strikes in Durban were the first step in building the black trade unions nationwide: in 1972, there had been 13,381 work days lost to industrial action by black workers-in 1983, there were 390,314 work days lost (Marx, 1992: 194). But along with noting the great increase in striking, we need to keep in mind why workers were striking and some suggestive data is available from 1982. Out of 69 strikes in the metal industry that year, 36 (52.17%) were primarily over wages, and the number of strikes over wages were more than three times greater than the number of those over retrenchments, the next most significant issue. Of the 26 strikes in the motor (automobile) industry that year, 14 (53.85%) were primarily over wages, and there were more strikes over wages than over the next two issues-dismissals and retrenchments-combined (MacShane, Plaut and Ward, 1984: 60; percentages calculated by author).

But there were more to the strikes than just the identified "main demands." Gerald Kraak explains:

Phil Bonner supports this claim, noting that two things developed from the strike wave: "workers provisionally won a de facto right to strike" and "a large number of strikes centered on issues of management control and affronts to personal dignity" (Bonner, 1987: 60).

Central to the emergence and development of the new unions, and their increasing shopfloor and strike power, were early decisions by workers and their unions that the unions had to be internally democratic, and that rank-and-file members had to be actively involved at the workplace and in the union. This meant that the members had control over union affairs and officials. And that the union organization on the shop floor had to be in-depth, "with shop stewards being given a key role" (Maree, 1987: 3). "The emphasis on strong, self-reliant shop steward committees" was what, ultimately, allowed the new unions to survive (Bonner, 1987: 57).

In the middle of this upsurge of labor militancy, as well as to address the problem of increasing dependence on black labor, the South African state legalized the new unions "in the hope of creating a more disciplined and complacent 'labor aristocracy'" (Marx, 1992: 247-248).^²³This program failed to constrain the new unions; union organizers used the opportunity to expand and strengthen these unions.^²⁴

This was reflected in national-level strike activities (not including stoppages or lockouts). Between 1970 and 1979, the yearly number of strikes varied between a low of 69 (in 1971) and a high of 384 (1974), with the average over these 10 years being 179 strikes. The numbers of strikers for the same 10 years varied between a low of 4,146 (1970) and a high of 98,379 (1973); the average was 27,903 strikers per year. The number of working days lost varied from a low of 10,558 (1978) to a high of 229,281 (1973), with an average number of working days lost between 1972 (the year records first became available)-1979 being 51,373. However, between 1980 and 1987, comparable figures report a significant increase in worker militancy. Strikes fluctuated between 207 (1980) and 1,148 (1987), with an average of 510 over these eight years. The number of strikers ranged from a low of 61,785 (1980) to a high of 591,421 (1987), averaging 224,773 strikers. And the number of working days lost varied from 124,594 (1983) to 5,825,231 (1987), averaging 1,135,408 days lost due to strike activity (Data from Kraak, 1993: 129, Table 6.1, Computations by author).^²⁵

The April 1979 founding of FOSATU (Federation of South African Trade Unions), a non-racialist labor center with strong organization in the factories, was a key development in consolidating these new independent unions (MacShane, Plaut and Ward, 1984; Maree, ed., 1987; Marx, 1992: 194-198; Seidman, 1994: 183-193; Adler and Webster, 1995: 79-81). FOSATU started off with 12 unions that had 20,000 members (Baskin, 1991: 25). By November 1983 , FOSATU-affiliated unions had organized over 106,000 workers in almost 500 factories (MacShane, Plaut and Ward, 1984: 38).

FOSATU's development is crucial to understand since it was the largest and most militant of the new labor centers,^²⁷ and then it later played a key role in the development of COSATU, an even larger and more politically advanced labor center that was founded in November 1985. In 1985, FOSATU's paid-up membership reached 139,917, and its power came from its affiliates' position in the industrial economy. FOSATU unions "had become the majority unions in the motor, tyre and rubber industries, as well as in chemicals, metal and engineering, paper and printing, furniture, transport sectors and (outside the Western Cape) textiles" (Kraak, 1993: 185).

Glenn Adler and Eddie Webster also detailed FOSATU's approach to the strategic use of power: "... (1) democratic processes to win voluntary consent from members for mobilization and for restraint when necessary, and (2) tactical flexibility, which included a capacity to distinguish principles from tactics and to choose those tactics most likely to succeed, including negotiation and compromise" (Adler and Webster, 1995: 80).

Key to FOSATU's organizational development was the principle of members' controlling their unions. "FOSATU argued strongly that workers on the shopfloor should dominate all union structures and control union officials, and insisted on mandates, report-backs, and worker control" (Baskin, 1991: 31). According to Anthony Marx (1992: 195), "FOSATU was most clearly distinguished on building unions from a strong base of member participation and on pursuing its members' economic interests as workers."^²⁸

Politically, however, FOSATU limited itself. Because it prioritized the building of strong unions in workplaces, it generally refused to get involved in "community" issues that focused on issues outside of the workplace as it "concentrated on consolidating its industrial presence" (Kraak, 1993: 220).^²⁹This caused considerable tension, especially early in FOSATU's development, between community-based organizations and FOSATU unions. Part of this hesitation was from an understanding of SACTU's (South African Congress of Trade Unions) experiences in the 1950s-early '60s where engagement in larger political campaigns brought on the direct enmity and repression of the state, which SACTU unions had generally been unable to withstand. But part of the hesitation was due to recognizing the diverse membership of FOSATU-affiliated unions, a considerable number who had no experience of community organization or political struggle, or who belonged to outside organizations such as Inkatha, which would oppose mobilization around community issues (Hindson, 1987; see also Seidman, 1994: 186). Thus, there was a real need to consolidate its members, and this took time.

The perspectives of the FOSATU unions were gradually changed in the recognition battles and struggles against the Industrial Council system. Regional and cross-union co-operation brought a greater uniformity of perspective among its affiliates, and in February 1982, FOSATU led the stoppage to protest at the death in police detention of AFCWU organiser Neil Aggett. Until then, the stay-away had been the primary way in which workers had asserted political demands by withholding their labour power. The Aggett stoppage was a political gesture at the workplace and it widened the methods of struggle used by workers (Kraak, 1993: 220).

Despite the unwillingness to engage in community-based issues, FOSATU came to understand the importance of community issues, not only for its members for all members of black communities. Hindson (1987: 213) argues that the key force in this change was that union members "were becoming involved in community based action" and that there was a high degree of frustration "at the lack of union assistance in fighting township issues" (see also Seidman, 1994: 197-203; 227-254).^³⁰

FOSATU also came to recognize the validity of the community organizations' mobilizing abilities; the challenge was how to consolidate organization after successful mobilization. One development that had important ramifications for FOSTATU and the labor movement was the establishment in the early 1980s of local area-based shop stewards' councils. These brought together stewards from a township or an industrial area, and they began taking up more general issues, such as developing strike support for colleagues in neighboring factories, and stopping shack destruction in nearby communities (Baskin, 1991: 30-31). This ultimately led to a broader conceptualization of politics by the union center.

By FOSATU's second congress in April 1982, delegates took positions on a number of explicitly political issues:

The labor center also sought to join other unions with whom they could ally with, and indicated that it was willing to dissolve FOSATU to create a larger, more inclusive labor center.

At the same time that FOSATU unions were building strong organizations on the shop floor, what became known as "community unions" were emerging. These were unions that had been more directly inspired by community-bases struggles and then later, the UDF-they believed it was impossible to separate workers' demands in the factories from their township problems. These unions were more obviously political than the "workerist" unions in FOSATU, but not as well organized. These unions tended to organize widely across entire communities, but not very extensively within the factories: "their organizational structures were unequal to the task of maintaining a massive post-strike membership or the winning of effective economic gains" (Baskin, 1991: 29).

Members in the two wings of the labor movement-the "workerists" of FOSATU, and the "populists" of the various community unions-recognized the need to build a united labor movement. On the ground, these various unions formed an alliance with civic organizations and youth and student congresses, "actively participating in the series of general strikes and political organizations from late 1984" (Adler and Webster, 1995: 81).

The first mass mobilization was in November 1984, when "800,000 workers stayed away from work and 400,000 students boycotted classes." This was "the beginning of united mass action between organized labour, student and community organizations, with unions taking a central role" (Adler, Maller and Webster, 1992: 318).^³¹Hindson points out the importance of this 1984 stayaway to political developments within the union movement: "What distinguished the action taken by the Transvaal unions in support of the students and residents in the townships was that the union and community organizations came together and forged a plan of joint action with specific objectives to take place over an agreed period of time" (Hindson, 1987: 217). "For the trade unions," write Adler, Maller and Webster (1992: 318), "it marked a decisive break with its previous strategy of remaining aloof from township struggles."

At the same time as these developments were taking place "on the ground," these unions engaged in a series of "unity meetings" between 1981 and late 1985. Finally, on the last weekend in November 1985, unions from five different traditions joined together to create COSATU (see Baskin, 1991: 49-50).^³²At the founding meeting in Durban, there were "760 delegates from 33 unions, representing over 460,000 organized workers" (Baskin, 1991: 53).^³³

The organizational ramifications of COSATU were fairly obvious; if nothing else by the size and nation-wide geographic reach of the new labor center. However, the political ramifications of the emergence of COSATU were perhaps even more important: this merger between the industrial-based and community-based unions signaled "? a strategic compromise in which the integrity of the industrial unions was acknowledged while the new [center] committed itself to participation in the national democratic struggle under the leadership of the ANC" (Adler and Webster, 1995: 82). By February 1986, a resolution was put forward that discussed COSATU's position:

This resolution was adopted a mere two months after the launching of COSATU. Anthony Marx relates that both economic and political goals had to be integrated in the labor movement's experience: "The experience of confronting both employers and the state in constructing unions had led them to appreciate how economic exploitation and political domination were linked and had to be confronted together." Further, they recognized that "an exclusive focus on short-term economic interests could result in complacency once concessions were granted, unless the workers had broader goals" (Marx, 1992: 204-205).

The importance of this type of trade unionism and its moving to the center of the liberation struggle simply cannot be under-emphasized. After the apartheid state enacted the State of Emergency in the country in June 1986, and then banned the UDF in 1988, it was COSATU that kept the anti-apartheid movement inside the country together, allowing it to survive (Baskin, 1991); in fact, "the labor movement emerged at the height of the state of emergency as the de facto leader of the internal democratic opposition" (Adler and Webster, 1995: 92).^³⁴Had COSATU been destroyed, it is all but certain that Nelson Mandela would still be residing in his cell on Robben Island-assuming he had remained alive-instead of being the former President of South Africa!

But COSATU was not destroyed, despite swimming in a maelstrom. It's paid-up membership grew from approximately 460,000 when launched in late 1985 to almost 1.5 million members in late 1990 (Baskin, 1991: 448). It helped create the political-economic conditions that forced the state to un-ban oppositional organizations (including the African National Congress, the Pan African Congress, and the South African Communist Party); free a number of political prisoners-most notably Mandela; and began dismantling apartheid between 1990 and 1992, all which led to free elections in 1994.^³⁵As of the middle of the decade, COSATU remained in the center of the post-apartheid transitions (Marcuse, 1995).

Discussion

In this paper, I have presented the context and development of the new trade union organizations in South Africa between 1973-1992. Although this paper is primarily written to respond to labor-related issues, it provides material related to a number of social movement theory-related issues. I mention these only briefly before addressing the labor-related issues:

With these issues now mentioned, there are three issues on which I wish to comment: the first two-which support some of my earlier claims-I touch on only briefly; when I discuss the third issue, concerning whether the new unions of COSATU can be conceptualized as 'social movement" unions, I do so in greater detail.

To address the earlier claims that labor movements involve more than just workers and their organizations, and the central role of collective identities in labor mobilization, I refer to the work of Gay Seidman, who addresses both of these issues. She points out that the emerging labor movements, especially in Durban in 1973, included networks of students and intellectuals-who publicized low wages and legitimized workers' grievances to the public and other workers-as well as clandestine factory activists, although she does not know the extent of the latter (Seidman, 1994: 176). I would go farther and add that networks of women, unemployed workers, students and other groups-both on an individual basis and as mobilized members of community organizations-helped support the emerging unions. I think the point has been sufficiently established: labor movements include more than just workers, and many, if not most, unions could not survive without this support, at least initially.

Supporting my earlier claim of the centrality of collective identity to mobilization, Seidman focuses on the class identities of the workers: "Undoubtedly, the most striking features of the labor movements that emerged in Brazil and South Africa in the 1970s was the creation of working-class organizations and identities?" (Seidman, 1994: 149). However, she downplays the racial identities of the workers, ignoring that the overwhelming majority of workers who mobilized in the new unions in South Africa were black (including Asian and "coloured") workers-this was not an across-the-board class mobilization as the vast majority of white workers did not join the new unions. The point stands that collective identity is crucial to labor mobilization; I only disagree with Seidman on the importance of racial identity, which I believe she downplays.

Where she is correct is that it was only with the creation and development of black workers' identities as workers (i.e., their class identity) that these workers were able to build powerful labor organizations that (1) surpassed the efficacy of multi-class organizations such as the UDF in the struggle (see Marx, 1992), and (2) were able to withstand the repression of the apartheid state. This eventually forced the state to negotiate with the opposition movement, dismantle the apartheid state, and hold free elections in 1994 (Baskin, 1991; Kraak, 1993; Adler and Webster, 1995; Marcuse, 1995).

The key issue at hand, however, is whether the development of COSATU and its member organizations between 1972-92 fit Scipes' conceptualization of social movement unionism or not. Therefore, I return to Scipes' conceptualization, of which there are five basic components:

Although Scipes does not set any specific requirements for fitting his conceptualization of social movement unionism-are three out of five or four out of five conditions sufficient, or must a labor center or affiliated unions meet every condition?-I think it is safe to say that any labor organization that largely incorporates these components into the self-conceptualization on which it acts qualifies as implementing social movement unionism.

In this case, it seems unquestionable that the type of trade unionism created and carried out by COSATU and its affiliated unions qualifies as social movement unionism: they see trade unions as only one site of struggle, not necessarily the only one or even the pre-eminent site, although they would probably argue that the unions are the "most important" site, and they ally with other social movements when possible; they see the unions as being controlled by their members and not by any external organizations; they see conditions in the workplace as being intimately linked with the national political-economic situation; they fight exploitation and oppression in the workplace along with domination from within and without the larger social order; and they are autonomous from other political organizations.

This finding, however-that COSATU meets the criteria for social movement unionism-also testifies to the validity of the conceptualization. We now have findings that two of the four most dynamic labor movements in the world-KMU and COSATU-specifically meet the criteria of this conceptualization.^³⁷

This finding is even stronger when it is realized that these labor movements exist in two very different countries: South Africa is classified as a "newly industrializing country" (NIC), with extensive industrial development, while the Philippines is at a lower-level of economic development; the countries, both colonized at one time, have completely different colonial histories, which has affected subsequent "development"-South Africa was colonized by the Netherlands and then England, while the Philippines was colonized by Spain and the United States; South Africa's colonization was by settlers who established a racial dictatorship, while the Philippines was occupied by outsiders who intermarried and created a mestizo elite; and South Africa has incredible amounts of extractive natural resources, while the Philippines' has qualitatively less. While I'm sure that other useful comparisons could be made, it should be clear that a new type of trade unionism has been developed in these countries, countries that differ qualitatively: this new type of trade unionism is not a product of similar structural factors.

But out of these findings that are of immediate relevance to researchers, what has been learned that is of direct relevance to new emerging labor movements? I project four:

Conclusion

Starting with labor movement theory, I clarified a number of concepts that have been used in a myriad of different ways, seeking to encourage researchers/activists to use common understandings for similar concepts, and suggesting a fruitful way to examine labor movements. I then concluded that it was ideological conceptualizations of trade unionism-initially advanced by activists but necessarily ratified by rank and file workers-that was the most important factor in the development of labor movements, and not structural factors.

I then turned to social movement theory, applying it to labor movements in their emergent stage, to explain how labor movements emerged. I particularly focused on the development of collective identity as being central to this process.

Turning to empirical data, I shifted to the mobilization processes in South Africa. I placed these processes within the context of industrialization in a racially stratified social order. I looked at the mobilization of blacks, beginning in the late 1960s, and suggested the impact of their conscienization on other mobilization processes.

And then I looked at the emergence and development of the new unions in South Africa. Based on their location within the social order's production, distribution and exchange processes, the new unions in COSATU built strong, shopfloor-based organizations that were able to withstand repression and united to become a political force for liberation in the country: specifically, they played an absolutely central role in the struggle to overthrow apartheid.

The mobilization efforts by the South Africans were all the more impressive in face of an economic system that was largely controlled by foreign investors, and under the domination of a repressive state that was bound and determined to maintain the racial hierarchy. While working and living conditions for African workers under this late industrialization process have been terrible, rather than to accede to their situation in a country that has an oversupply of labor, they organized one of the most dynamic labor movements in the world.

These workers' choice of social movement unionism as the type of trade unionism they wanted to develop is inspiring because they identified their interests as workers as part of those of the majority of the population and not separate from nor superior to the majority. This certainly was not predetermined for them. It might be more understandable had they chosen to monopolize a few positions and engaged in some type of economic unionism, but that would not have solved the problem of the racial hierarchy in the country. And while it is not known whether this conceptualization will take them beyond a "defensive" form of social movement unionism, it at least indicates their willingness to confront the larger issues faced by all people of color in South Africa.

A turn to political unionism, say in deference to the African National Congress for example, also would not be totally out of the realm of possibility. Certainly by sending a number of senior COSATU leaders to Parliament after the 1994 elections-i.e., after the end of the emergence and challenging period-COSATU weakened itself, and has since been trying to get back on its feet (Adler and Webster, 1995). But COSATU has refused to just submit to ANC leaders, preferring to maintain an independent position and stance, and there has been some significant debate about its relationship with the ANC, as well as with its other formal alliance partner, the South African Communist Party, with some unions calling for COSATU to withdraw from their formal alliance that was established in 1990. Certainly in light of shared political ideologies, additional research must be done in the post-1992 period to see how these relationships have developed, and their effects upon workers.^³⁸

But how can we understand this selection of social movement unionism as the basis for COSATU's trade unionism, particularly in light of strong arguments that could have been made for engaging in either type of rival unionisms, economic or political unionism? This is where my discussion of social movement theory, with the detailed explanation of movement cultures, becomes so important. Basically, workers moved toward a social movement unionism type of labor movement as they maneuvered in a very difficult field of action. Nothing was predetermined. These workers were trying to figure things out, but from the standpoint of principally standing up for their rights both as economic actors and (potential) political citizens within the social order.

This article has found that COSATU and its members unions have been developing a type of trade unionism now known as social movement unionism. This type of trade unionism differs qualitatively from economic- and political-types of trade unionism. And by demonstrating its existence in South Africa, in addition to the Philippines, I have strengthened arguments for the validity of this conceptualization.

Based on all of these conclusions, I then suggested four findings of immediate interest to those who are involved in building new labor movements. The idea has been to learn from past experiences so as to help those who are following in the footsteps.

The effort to examine the emergence and developments of the new unions in South Africa has been a very fruitful one. Not only has it validated Scipes' version of social movement unionism in general, but it has extended this concept to incorporate the new unions in COSATU along with the previous included unions in KMU. Further, this examination has allowed me to extract a number of important findings from the experiences in South Africa, which can now be utilized by subsequently emerging labor movements around the world. It is hoped that this will lead to the further development of other unions and labor centers as social movement-type labor organizations.

References

Adler, Glenn, Judy Maller and Eddie Webster. 1992. "Unions, Direct Action and Transition in South Africa" in Norman Etherington, ed., Peace, Politics and Violence in the New South Africa. London: Hans Zell Publishers: 306-343.

Adler, Glenn and Eddie Webster. 1995. "Challenging Transition Theory: The Labor Movement, Radical Reform, and Transition to Democracy in South Africa." Politics and Society, Vol. 23, No. 1, March: 75-106.

Aronowitz, Stanley. 1973. False Promises: The Shaping of American Working Class Consciousness. New York: McGraw Hill Book Company.

Baskin, Jeremy. 1991. Striking Back: A History of COSATU. Johannesburg: Ravan Press, and New York and London: Verso.

Bonner, Phil. 1987. "Overview: Strikes and the Independent Trade Unions" in Maree, ed.: 55-61.

Bonnell, Victoria. 1983. Roots of Rebellion: Workers' Politics and Organization in St. Petersburg and Moscow, 1900-1914. Berkeley and Los Angeles: University of California Press.

Brenner, Johanna. 1998. "On Gender and Class in U.S. Labor History." Monthly Review, Vol. 50, No. 6, November, downloaded from www.monthlyreview.org/1198bren.htm.

Broad, Robin. 1988. Unequal Alliance: The World Bank, the International Monetary Fund, and the Philippines. Berkeley and Los Angeles: University of California Press.

Buhle, Paul. 1999. Taking Care of Business: Samuel Gompers, George Meany, Lane Kirkland, and the Tragedy of American Labor. New York: Monthly Review Press.

COSATU. 1992. Our Political Economy: Understanding the Problems. Johannesburg: COSATU Education Department.

Deyo, Frederic. 1989. Beneath the Miracle: Labor Subordination in the New Asian Industrialism. Berkeley and Los Angeles: University of California Press.

Dropkin, Greg. 1985. "The Making of the 'Superfed'." International Labour Reports, No. 12, November-December: 18-21.

Fantasia, Rick. 1988. Cultures of Solidarity: Consciousness, Action, and Contemporary American Workers. Berkeley and Los Angeles: University of California Press.

Freeman, Richard B. and James L. Medoff. 1984. What Do Unions Do? New York: Basic Books.

Friedman, Steven. 1987. Building Tomorrow Today: African Workers in Trade Unions, 1970-1985. Johannesburg: Ravan Press.

Gereffi, Gary. 1990. "Paths to Industrialization: An Overview" in Gary Gereffi and Donald L. Wyman, eds., Manufacturing Miracles: Paths of Industrialization in Latin America and East Asia. Princeton: Princeton University Press.: 3-31.

Golden, Miriam. 1988. Labor Divided: Austerity and Working-Class Politics in Contemporary Italy. Ithaca: Cornell University Press.

Hindson, Doug. 1987. "Overview: Trade Unions and Politics" in Maree, ed.: 208-217.

Hodson, Randy. 1991. "The Active Worker: Compliance and Autonomy at the Workplace." Journal of Contemporary Ethnography, Vol. 20, No. 1, April: 47-78.

Hunt, Scott A., Robert D. Benford and David A. Snow. 1994. "Identity Fields: Framing Processes and the Social Construction of Movement Identities" in Laraña, Johnston and Gusfield, eds.: 185-208.

Hutchison, Jane. 1992. "Women in the Philippines' Garment Export Industry." Journal of Contemporary Asia, Vol. 22, No. 4: 471-489.

ICFTU. 1984. African Worker Under Apartheid. Brussels: International Confederation of Free Trade Unions.

Institute for Industrial Education. 1979. "The Durban Strikes: South Africa 1973" in Robin Cohen, Peter C.W. Gutkind, and Phyllis Brazier, eds. Peasants and Proletarians: The Struggles of Third World Workers. New York: Monthly Review Press: 401-417.

Johnston, Hank, Enrique Laraña and Joseph Gusfield. 1994. "Identities, Grievances, and New Social Movements" in Laraña, Johnston and Gusfield, eds.: 3-35.

Johnston, Hank and Bert Klandermans. 1995. "The Cultural Analysis of Social Movements" in Johnston and Klandermans, eds.: 3-24.

Johnston, Hank and Bert Klandermans, eds. 1995. Social Movements and Culture. Minneapolis: University of Minnesota Press.

Kraak, Gerald. 1993. Breaking the Chains: Labour in South Africa in the 1970s and 1980s. London and Boulder, Colorado: Zed Press.

Lambert, Rob and Eddie Webster. 1988. "The Re-emergence of Political Unionism in Contemporary South Africa?" in William Cobbett and Robin Cohen, eds. Popular Struggles in South Africa. London: James Currey: 20-41.

Laraña, Enrique, Hank Johnston and Joseph Gusfield, eds., New Social Movements: From Ideology to Identity. Philadelphia: Temple University Press.

Larson, Simeon and Bruce Nissen, eds. 1987. Theories of the Labor Movement. Detroit: Wayne State University.

Lewis, Jon. 1987. "Overview: The Registration Debate and Industrial Councils" in Maree, ed.: 170-175.

MacShane, Denis, Martin Plaut and David Ward. 1984. Power! Black Workers, Their Unions, and the Struggle for Freedom in South Africa. Boston: South End Press.

Marcuse, Peter. 1995. "Transitions in South Africa: To What?" Monthly Review, November: 38-52.

Maree, Johann. 1987. "Overview: Emergence of the Independent Trade Union Movement" in Maree, ed.: 1-11.

Maree, Johann and Debbie Budlender. 1987. "Overview: State Policy and Labour Legislation" in Maree, ed.: 116-123,

Maree, Johann, ed. 1987. The Independent Trade Unions, 1974-1984: Ten Years of the South African Labour Bulletin. Johannesburg: Ravan Press.

--- 1992. Lessons of Struggle: South African Internal Opposition, 1960-1990. New York and Oxford: Oxford University Press.

--- 1998. Making Race and Nation: A Comparison of the United States, South Africa and Brazil. Cambridge: Cambridge University Press.

Marx, Karl and Friedrich Engels. 1978/1848. "Manifesto of the Communist Party" in Robert C. Tucker, ed., The Marx-Engels Reader, 2nd Edition. New York and London: W.W. Norton & Company: 469-500.

--- 1982. Political Process and the Development of Black Insurgency, 1930-1970. Chicago and London: University of Chicago Press.

--- 1994. "Culture and Social Movements" in Laraña, Johnston and Gusfield, eds.: 36-57.

McAdam, Doug and David A. Snow. 1997. "Introduction--Social Movements: Conceptual and Theoretical Issues" in McAdam and Snow, eds.: xviii-xxv

McAdam, Doug and David A. Snow, eds. 1997. Social Movements: Readings on Their Emergence, Mobilization, and Dynamics. Los Angeles: Roxbury Publishing Company.

--- 1989. Nomads of the Present: Social Movements and Individual Needs in Contemporary Society. London: Hutchinson Radius.

--- 1995. "The Process of Collective Identity" in Johnston and Klandermans, eds.: 41-63.

Meyer, David S. and Suzanne Staggenborg. 1996. "Movements, Countermovements, and the Structure of Political Opportunity." American Journal of Sociology, Vol. 101, No. 6, May: 1628-1660.

Moody, Kim. 1997. Workers in a Lean World: Unions in the International Economy. London and New York: Verso.

Mueller, Carol. 1994. "Conflict Networks and the Origins of Women's Liberation" in Laraña, Johnston and Gusfield, eds.: 234-263.

Munck, Ronaldo. 1988. The New International Labour Studies: An Introduction. London: Zed Press.

Nederveen Pieterse, Jan P. 1989. "Limits of World Systems Theory," Chapter 2 in Empire and Emancipation: Power and Liberation on a World Scale. New York: Praeger.

Perlman, Selig. 1968/1928. A Theory of the Labor Movement. New York: Augustus M. Kelley, Publishers.

Pillay, Devan. 1990. "COSATU: Driving Out Apartheid." International Labour Reports, No. 39, May-June: 7-10.

Piven, Frances Fox and Richard A. Cloward. 1979. Poor People's Movements: Why They Succeed, How They Fail. New York: Vintage Books.

Rachleff, Peter. 2000. "The Current Crisis of the South African Labor Movement." Presented to the North American Labor History Conference, Detroit, October.

Robnett, Belinda. 1996. "African-American Women in the Civil Rights Movement, 1954-1965: Gender, Leadership, and Micromobilization." American Journal of Sociology, Vol. 101, No. 6, May: 1661-1693.

SACCAWU. 1991. Sharing the Load: The Struggle for Gender Equality, Parental Rights and Childcare. Johannesburg: South African Commercial, Catering and Allied Workers Union.

--- 1992a. "Social Movement Unionism and the Kilusang Mayo Uno." Kasarinlan [Third World Studies Center, University of the Philippines], Vol. 7, Nos. 2 & 3 (4th Qtr. 1991-1st Qtr. 1992): 121-162.

--- 1992b. "Understanding the New Labor Movements in the 'Third World': The Emergence of Social Movement Unionism." Critical Sociology, Vol. 19, No. 2: 81-102.

--- 1996. KMU: Building Genuine Trade Unionism in the Philippines, 1980-1994. Quezon City, Metro Manila, Philippines: New Day Publishers.

--- 2000. "It's Time to Come Clean: Open the AFL-CIO Archives on International Labor Operations." Labor Studies Journal, Vol. 25, No. 2, Summer: 4-25.

--- Forthcoming. "Social Movement Unionism: A Call for Theoretical Clarification." Comparative Labour Movements Research Committee (RC44) Newsletter [Johannesburg], International Sociological Association.

--- In progress. "Business versus Political Forms of Economic Trade Unionism: Trade Union Conceptualizations in Chicago's Steel and Meatpacking Industries and Their Effects on Race Relations, 1933 to Deindustrialization." Ph.D. Dissertation, Department of Sociology, University of Illinois at Chicago.

Seidman, Gay W. 1994. Manufacturing Militance: Workers Movements in Brazil and South Africa, 1970-1985. Berkeley and Los Angeles: University of California Press.

Snow, David A., E. Burke Rochford, Jr., Steven K. Worden and Robert D. Benford. 1997/1986. "Frame Alignment Processes, Micromobilization, and Movement Participation" in McAdam and Snow, eds.: 235-251.

--- 1991a. " 'Social Movement Unionism': Beyond 'Economic' and 'Political' Unionism." Amsterdam: International Institute for Research and Education, Working Paper No. 19.

--- 1991b. "Social-Movement Unionism: A New Model for a New World." Working Paper No. 110. The Hague: Institute of Social Studies.

--- 1998. Globalization, Social Movements and the New Internationalisms. London and Washington: Mansell.

Footnotes

1Gay Seidman used the term "social movement unionism" in her comparison of the new labor movements that emerged in Brazil and South Africa (Seidman, 1994: 2). She appears to have been unaware of the international theoretical debate at the time, which is not surprising considering the diffused sites in which the articles of the debate were published (Scipes, forthcoming). Nonetheless, her term was not adequately developed on a theoretical level and, despite using the same term, the discussion in her book was not part of the international debate.

Although the term "social movement unionism" has been used to a growing extent in popular discourse, especially in the United States-e.g., Kim Moody (1997), as well as others-popular usage follows Seidman's conceptualization and not, for example, Scipes'. Therefore, popular usage differs qualitatively from the way it is used in this paper.

2I specifically limit my examination to the 1973-92 period. Political changes in South Africa since 1992 have been so extensive-effecting the union movement as well as the entire society-that a separate analysis is required for the period after 1992. At the same time, it is this earlier period of emergence that is of most interest to other emerging labor movements.

3The apartheid state classified the South African population into three "racial" categories: white, colored and black. The "colored" group was further divided into Asians-mainly Indians whose ancestors were brought into the country for their "cheap labor" by the British during the 19^th Century-and peoples from mixed-race relationships.

Following the general usage of workers in South Africa, I refer to all people of color in this paper as "black," unless a finer distinction is warranted, and then the specific distinction is used. See MacShane, Plaut and Ward, 1984: 7.

4While the term "labor movement" continues to be used by labor centers that have become institutionalized (and largely ossified), I do not use the term in this manner. In this paper, I specifically focus on labor movements in their pre-institutionalized periods, when they emerge as social movements and challenge the social order from which they emerge. Accordingly, in these pre-institutionalized periods, such as examined herein, labor movements are just a particular type of social movement and, thus, appropriate social movement theory can be applied to them.

5I want to be clear: my argument is not based on a structural analysis of society. All of the new labor movements that Scipes considers beyond KMU-again, CUT, COSATU and KTUC-emerged and developed in countries undergoing rapid industrialization and which were ruled by dictators. (The dictatorship in South Africa was a racial dictatorship, which directly affected workers, both black and white, albeit differentially.) However, workers in other countries facing similar conditions either did not try to build social movement unionism (e.g., Malaysia) or have been unable to do so to date (Indonesia). Accordingly, any explanation of how social movement unionism emerged must also allow for it not to emerge. Thus, I argue that we must not only look at the particular socio-historical situation in which the organizing takes place, but also must account for why the workers responded to it in the ways that they did.

6 Richard Freeman and James L. Medoff (1984) examine the effects of unions on US society in an empirical study to evaluate the social benefits of American trade unions. They see unions having two different "faces"-a monopoly face, which restricts competition and seeks gains for their members above any and all considerations, and a "voice/response" face, which allows workers through their unions to "dialog" with management and both resolve grievances and increase productive practices in their workplace. Thus, even this institutional analysis of what trade unions do-a much more limited and conservative approach than I take in this paper-recognizes at least dual purposes for labor organizations.

My approach looks at labor movements from the inside, i.e., from the perspective of members of labor movements, and thus could be looked at differently by workers outside of them. For example, while particular workers inside a labor movement could recognize that their power comes from their ability to monopolize skills, I doubt any of them would describe this monopoly as being anti-social or destructive; workers outside, however, could see this as such, especially if they were denied entrance to a monopolized craft, such as African-American and/or women workers denied entry into the building trades because of their race and/or gender (Buhle, 1999), or if they were members of a competing labor movement and had their more militant efforts undermined or destroyed (see Scipes, 2000).

8Randy Hodson (1991) looks at workers and how they deal with work, seeing these activities as multidimensional and not easily reducible to any single dimension-the prerogatives of autonomy are crucial to understanding worker behavior. He also points that "Worker behavior also includes a tremendous amount of activity directed against other workers" (Hodson, 1991: 72). This view also supports my statement that no automatic workers' group interest or collective consciousness exists and that collective interests must be constructed.

Johanna Brenner (1998), bringing women workers into our understanding, discusses how different factors-specifically including gender-affects the culture of workers and how they construct their "class" consciousness.

9The number of labor centers vary by country: some countries, such as Australia, Germany, Great Britain, and the United States, each have one labor center; others countries, such as Brazil, France, Italy, Japan, Mexico, the Philippines, South Africa, South Korea, and Sweden, each have at least two different labor centers, if not more. In the US context, but using international labor terminology, the AFL-CIO is a labor "center."

10Doug McAdam (1982) advances a "political process" model to explain social movement emergence and development that is a more elaborate and sophisticated effort than is Piven and Cloward's, but remains, at heart, a structural explanation, with its attendant weaknesses. It is not necessary for this paper to elaborate further.

McAdam (1994) has already begun moving away from a more structural explanation toward a cultural one.

11Melucci, who concentrates on movements in the so-called "developed" countries, argues that movements emerge particularly out of a felt need to challenge the cultural "codes" that shape knowledge and understanding in contemporary societies. His research focuses on the production of information and symbolic resources as the location of conflict in these countries (Melucci, 1989).

12Melucci sees movements not as entities with "unity of goals" but rather as "action systems": "They are systems of action, complex networks among the different levels or meaning of social action" (Melucci, 1995: 53). This means that any social movement should not be seen as a monolithic whole, but rather should be seen as having a multiplicity of politics, interpretations, possibly ideologies, etc., out of which a common politics is negotiated and from which action is constructed.

14Klandermans, according to Mueller, does not include the development of collective identities in his conceptualization (Mueller, 1994: 256-257). In a later article, Klandermans reiterates his three-level conceptualization of meaning construction (Johnston and Klandermans, 1995: 10).

16Without wanting to take much space on this issue, I argue that if structural factors determined the type of mobilization, then in a country that has multiple labor movements, the types of trade unionism would be the same across all of these movements. However, different types of labor movements exist in a number of countries, including Brazil, the Philippines and South Korea.

In South Africa, and particularly within the province of KwaZulu-Natal, there has been extensive conflict between the United Workers Union of South Africa (UWUSA), a federation based upon political unionism, and COSATU (Congress of South African Trade Unions), which I argue is a labor center based on a social movement type of unionism (Baskin, 1991). A structural analysis cannot explain the differences between the two labor groupings, nor the extensive conflict.

18South Africa's manufacturing sector has had fairly high rates of growth over a long period of time: from 1920-30, it averaged 4.5% per year; from 1930-40, 9.1%; from 1940-50, 6.8%; from 1950-60, 6.0%; from 1960-70, 8.6%; and from 1970-80, 5.3% (COSATU, 1992: 48).

19All valuations used in COSATU (1992) are in Rands (R) and listed in constant 1985 values, unless otherwise indicated.

While I do not have comparative economic data for 1985, I have found some indicators for April 1984, which are suggestive: the exchange rate was 1R=US $.80 (one Rand was equal to 80 cents), the monthly minimum wage in the metal industry was equivalent to $215.00; and "Prices in the shops for food, clothes, leisure were about the same as Western Europe or North America" (MacShane, Plaut and Ward, 1984: 8).

20This is an incredibly truncated account of the changes taking place among the opposition in South Africa during these years-I limit myself to highlights, as I understand them, that had a direct impact on the development of the new unions that ended up in COSATU in the late 1980s.

There is much I simply do not begin to include. I do not discuss the Pan African Congress (PAC), or any of the organizations that emerged in the late 1970s or '80s that continued in the "black consciousness" tradition, such as AZAPO (Azanian People's Organization), or the unions that developed from them. As important as they were, their importance to the opposition movement pales in comparison to organizations of the non-racial Charterist tradition, which includes the African National Congress (ANC), the United Democratic Front (UDF) and COSATU. For information on the black consciousness-inspired political organizations, see Marx, 1992; for information on black consciousness-inspired trade unions, see Kraak, 1993. I also do not discuss the important role of the South African Communist Party, which has long been allied with the Charterist organizations ( see Pillay, 1990). I also do not discuss external efforts, whether by forces in exiles such as the ANC leadership or SACTU, or the global solidarity movement that emerged around the world in support of the peoples of South Africa.

21This is an extract from a publication titled "The Durban Strikes," which was initially published in 1973. In the extract, not only is there a description of the spread of strikes, but a more in-depth examination and analysis of conditions, strikes, and developments in the textile industry is provided.

Seidman (1994: 174-175) presents evidence that there was some clandestine activity by activists linked to SACTU in the Durban strikes, although she is unable to determine to what extent or how extensively these activists were involved.

22Kraak (1993) emphasizes "class" issues, which are certainly important; unfortunately, he tends to downplay racial issues. [Seidman (1994) takes a similar approach.] Trade unions had long-been recognized as legitimate for white workers-it was the efforts by black workers to extend this legitimacy to black workers that made these battles so intense and violent. See Baskin, 1991; Friedman, 1987; MacShane, Plaut and Ward, 1984; Maree, ed., 1987; Marx, 1992: 189-234.

23For a discussion of the development of industrial law-which attempts to regulate labor relations-within the context of state regulation and control, see Maree and Budlender, 1987; for a discussion of the development of industrial law between 1979-1988, see Kraak, 1993: 113-125.

24For a description of union organization in a number of industries-automobile, building and construction, chemical, longshoring (docks), food, insurance and clerical, metal and engineering, mining, municipal workers, print and paper, retail, textile and clothing, and transport-see MacShane, Plaut and Ward, 1984: 91-110.

25Kraak (1993: 127-173) provides an wealth of information regarding workers' strikes between 1970-1987. He analyzes them chronologically as well as by the issues over which they were fought; e.g., registration under industrial law and recognition; struggles against official bargaining machinery (including industrial councils by selected industries); and struggles against employer and government controls. He notes strikes against racism in the workplace, and struggles to improve women workers' situations including maternity (such as for 12 months unpaid maternity leave with guaranteed return to job) and parental leave. He analyzes strikes against retrenchments (lay-offs), for pensions, against disinvestment. And he reports new tactics and new trends in workers' fights, including company wide strikes (beginning in 1983) and factory occupations (1985), as well as efforts to build unity through shopfloor activity, and efforts to build alliances between unions and community organizations.

27Although not all unions were affiliated to a labor center-some remained independent-there were three major labor centers of new unions by 1984: FOSATU (Federation of South African Trade Unions), CUSA (Council of Unions of South Africa) and AZACTU (Azanian Congress of Trade Unions). Of these, only FOSATU was organized on non-racial grounds; members, elected leaders and full-time staff could be of any racial background, including white. In 1984, FOSATU-affiliated unions had 122,302 paid-up members. CUSA-affiliated unions claimed 99,223 members in 1983/84, but these were only signed-up members and not the more accurate paid-up membership category. AZACTU-affiliated unions claimed 30,513 members in 1983/84, although only about half were paid-up (Kraak, 1993: 256-259).

These labor centers also differed by militancy. "In 1982, FOSATU-affiliated unions were involved in 145 strikes with 90,000 workers taking part. This compares with CUSA where 10,000 workers took part in 13 strikes, or SAAWU [South African Allied Workers Union, an unaffiliated black union-KS], which organized six strikes involving 2,600 workers" (MacShane, Plaut and Ward, 1984: 58).

28For details on FOSATU's organization and structure-specifically addressing the key role of shop stewards, branches and locals, full-time officials, white organizers, recruitment, how unions are financed, communications, training and education-see MacShane, Plaut and Ward, 1984: 64-74.

29MacShane, Plaut and Ward (1984: 126-129) discuss the political role of the black unions, comparing the position of the General Workers' Union (representing FOSATU's position), CUSA, and two UDF-affiliates, the Municipal and General Workers' Union and the General and Allied Workers' Union. They present the positions of these different organizations in their appendices.

30Seidman tends to see developments in community organization and struggle as a product of class consciousness; i.e., that political consciousness developed among workers in the factories, who then transmitted their analysis and understandings to community members who, in turn, adopted them as their own and acted accordingly. I do not think this is an accurate description of what happened, nor do I think it is an accurate analysis of what did happen.

It is clear that some workers (as well as probably most community members) got politicized in community struggles (such as those against influx controls, or in those for adequate housing, schools, services, etc.) and took their new understandings into the factories. (This is a schematized version of what I understand as happened, as workers are community members at the same time, as are their family members, friends, etc.) (Marx, 1992). In any case, I think a much more accurate analysis would see the interpenetration of race- and class-related understandings, with racial-consciousness being primary at some times and class-consciousness being primary at others. (I would also include gender-consciousness in this interpenetration as well-see, for example, SACCAWU, 19991.) Further discussion is beyond the scope of this paper.

31For details of labor's-and particularly COSATU's-involvement in mass mobilization, which was central to forcing the democratic transition that repudiated apartheid between 1985-1991, see Webster, Maller and Adler (1992). Adler and Webster (1995) place this labor mobilization in the context of "transition theory," arguing that "a mobilized civil society and powerful social movements-especially the labor movement-played a central and constructive role in creating the conditions for the transition, shaping its character, and indeed in legitimizing the transition process itself" in South Africa (Adler and Webster, 1995: 77), and thus need to be included in transition theory (pp. 99-100). Scipes (1996) makes a parallel argument in his study of the KMU in the Philippines, especially in the transition from Marcos to Aquino.

33Baskin's 1991 book is the recognized standard on the emergence and development of COSATU during its first five years. His is an incomparable account from inside, at the highest levels of the organization.

Kraak (1993: 180-205) provides additional information on the new unions, including more on the unions that joined to create NACTU. Kraak also places the struggle to build the new unions in the context of the struggle against apartheid, despite underplaying it. Kraak presents data on workers responses to calls for stay-aways and on-site stoppages between 1982-1987, by region and by specific call (Kraak, 1993: 232-233), as well as specific information on unions affiliated to COSATU and NACTU, and growth of union membership of the affiliated unions between 1980 and 1986 (Kraak, 1993: 252-259).

34Key to COSATU's emerging power was the "stayaway," the joining of unions, community organizations and students to fight for demands, both in the industrial relations realm as well against state policies not directly related to the workplace. In November 1991, COSATU led a stayaway against state efforts to institute a Value Added Tax. The stayaway lasted two days, with estimates that 3.8 million people participated on the first day, and 3.4 million on the second (Adler, Maller and Webster, 1992: 332). For a discussion of the development of stayaways, from 1950 to 1991, see Adler, Maller and Webster, 1992: 312-337.

35For a contemporary report of the situation and issues facing COSATU-and its allies, the African National Congress (ANC) and the South African Communist Party (SACP)-immediately after the unbanning of these organizations and the release of Nelson Mandela, see Pillay, 1990.

36The concept of political opportunity was even less relevant to Scipes' study of the KMU in the Philippines. Despite helping to replace the dictator Marcos with the "elite" democrat Corazon Aquino-a situation I think most researchers who accept the concept would label as a "political opportunity"-the KMU suffered more repression (deaths, torture, physical assaults) under Aquino than under Marcos (Scipes, 1996).

37Seidman (1994) makes a strong claim that the new unions in Brazil that joined in CUT developed similarly to those in South Africa. I think she is correct: evidence I've seen to date suggests that she is. However, she focuses on the collective identities developed out of the industrial processes and, particularly in South Africa, downplays collective identities developed in struggle against racial oppression. Anthony Marx (1998) certainly argues the case for more attention to racial issues in both Brazil and South Africa (and the United States).

Until someone more familiar with CUT, and Brazil, makes a stronger argument than I'm capable of making, I do not think that the conceptualization of social movement unionism should be definitively extended to the new unions of that country.

Seidman (1994: 264-272) also suggests similar development among the new unions in South Korea. Again, I think she is correct. But these unions, too, need a stronger argument.

38Peter Rachleff (2000) reports a case, at Volkswagen South Africa, that took place in early 2000, whereby insurgent workers refused to accept a deal negotiated by their union (NUMSA) and supported by leaders of both COSATU and the ANC, including Thabo Mbeki, the President of South Africa. Further research is needed to determine if this is an individual case or symptomatic of something much larger. But it raises the question for future investigation on the relationship between social movement unions and political movements they support. Further comment is beyond the scope of this paper.

letzte Änderung am 05. Juni 2003
LabourNet Germany ARCHIV! Aktuelle Meldungen im neuen LabourNet Germany
Home -> Diskussion -> Gewerkschaftsstrategien -> SMU -> Südafrika	\| Suchen

Years covered:	Average annual number of strikes:	Average annual number of strikers:	Average annual number of working days lost:
1970-1979^²⁶	179	27,903	51,373
1980-1987	510	224,773	1,135,408

LabourNet Germany ARCHIV! Aktuelle Meldungen im neuen LabourNet Germany

Can We Apply The Theoretical Conceptualization To The New Unions in South Africa - And Beyond?