Zambezia (1991), XVIII (i).'WE ARE TAKEN AS SHOVELS, USED AND PUTASIDE ... ': ANTHROPOLOGICAL PERSPECTIVES ONTHE ORGANIZATION OF WORK AND WORKERSIN ZIMBABWEAN INDUSTRY IN THEFIRST DECADE OF INDEPENDENCEANGELA P.CHEATERDepartment of Sociology, University of ZimbabweIN 1980 MUCH was expected of Zimbabwe's newly independent government.It had promised, in the party's Election Manifesto, that 'Tomorrow, aZANU Government will ensure that the economic system is controlled andoperated in the interests of the People as a whole' (Zanu[PF], 1980, 8).who would be involved as full participants 'in both the decision-makingprocess, management and control of the industries concerned . . . and inthe sharing of benefits in accordance with their contributions of in-puts oflabour' (ibid., 4). Of course, read carefully with the benefit of ten years'hindsight, these statements promised only strong central state controlover the economy, which was indeed delivered, before new promiseswere made in 1989 to dismantle its more stifling aspects in order toliberalize trade and rejuvenate the failing economy. But workers and theowners of capital alike may be forgiven for having in 1980 interpretedthese electoral promises as implying a greater future involvement ofworkers in the management of industry.In fact, very distant from the concept of worker self-management,government's main concern was to redress colonial racism, which hadoverlaid the racial difference between Black and White on the divide(which some would call a 'class divide*) between workers and management.Existing hierarchies of top-down managerial control, inherited from thecolonial past, have not been questioned as Blacks have taken the reins ofmanagement in both public and private sectors. If anything, new Blackmanagers have tended to widen the already large gap between managementand workers in ways that the newly introduced workers' committees havebeen powerless to control, even given their participation in works councils.The declines in output and worker productivity experienced by manyindustrial firms after Independence, including the two considered in thispaper, perhaps reflect the workers' growing disenchantment with the newmanagerial class' as distinct from the old 'managerial race'.From the perspectives of different categories of employee, this articleexamines work and its organization in independent Zimbabwe, drawing on6970 'WE ARE TAKEN AS SHOVELSresearch carried out at two separate enterprises at two different times(1982-3 and 1989-90). I have given both firms pseudonyms: 'Zimtex' is atextile factory and 'Zimcor' a heavy industrial enterprise. Both invited meto investigate what their managements perceived to be organizationalproblems. Zimtex thought that its workers were taking excessive sickleave with the connivance of local health officials. Zimcor asked me toinvestigate 'why the workers won't work and the managers won't manage'.I agreed to both requests on the conditions that I would define bothproblems within the wider context of company organization, that I wouldbe free to use whatever techniques of investigation I deemed appropriate,and that I would have free access to both the shop-floor and companyrecords and meetings. In both companies I used standard anthropologicalmethods of investigation: direct observation and participation wherepossible (though not infrequently this meant that I had to disentanglemyself from attempts to involve me, as a neutral referee, in the resolutionof disputes, for example in disciplinary meetings, over what, exactly, hadbeen said). Unstructured, in-depth interviewing of individuals in theirworkplaces, plus group discussions among different categories of employee,supplemented my own observations of work and social interaction.The two companies investigated are not really comparable. Thoughboth are large employers, the primary industrial producer, which was stillexperiencing a decline in production at the end of the decade, was tentimes larger than the textile firm, which had by then overcome its crisis ofthe early 1980s (described in Cheater, 1986 and 1988). Their ownership,too, is very different: Zimcor has a majority shareholding by the state andminority shares held mainly by multinational corporations, while Zimtexis owned mainly by individual local and overseas investors through itslisting on the Zimbabwe Stock Exchange (Cheater, 1986,5-7). While Zimcoris still is constrained in its profitability by state-controlled prices and thenecessity to export through a parastatal, Zimtex has successfully exploitedits freedom to explore export markets independently and to negotiatesales in the domestic market without price constraints. Yet, despite theirdifferences, the views of workers in these two enterprises showed someremarkable structural similarities which I shall explore and attempt toexplain here.In order to do so, it is necessary to consider the two workplaces, eachwith its multiple productive units, within the context of the organizationalinterfaces among workers, management and the state or government (seeCheater, 1986, chs. 7 and 8). In examining these interfaces, it is alsonecessary to bear in mind that, although enterprises have formal rules bywhich they are organized, each also has an 'informal structure' of personalrelationships linking role incumbents (Roethlisberger and Dickson, 1939;Blau, 1964). Within such informal, personal relationships, workplace issuesANGELA P. CHEATER 71are actually negotiated among those differentially placed in the workhierarchy. Workplaces are, therefore, sites of micro-political strugglesconcerning production: they are not places in which rules are simplyenforced without resistance. 'Management' is one description of thisprocess of negotiation in which, if it is successful, labour, raw materialsand other resources are combined to yield a new, marketable product. Butsuch negotiations are also influenced, if not 'managed', by those whocontribute their labour to the production process. They can go slow,refuse to work overtime, sabotage equipment or use delaying tactics inmeetings. While they do not have the formal authority vested in manage-ment, workers do have informal power over the production process.Management tactics that treat workers as 'shovels', to be used and thenput aside, are therefore generally unsuccessful. Rules alone are insufficient:they have to be supplemented by effective personal interaction.BUREAUCRATIC RULE AND ITS EFFECTS IN THE WORKPLACEAs I have argued elsewhere (Cheater, in press), the new Zimbabweangovernment appears to have regarded bureaucratic theory as the solutionto its problem of how to de-racialize the economy in a fair and equitablemanner. The legal authority of an impersonal bureaucracy, as describedby Weber (1947, 329, 341), is based on rationally-agreed rules laid downin writing, which are applied consistently by officials within clearly-defined, specialized spheres of competence. These offices are ranked ina hierarchy of supervision, and their incumbents, who are responsible tothe system of rules and appointed on the basis of their technical qualifi-cations, are remunerated at standard, defined rates. Officials have nopersonal claim on the resources of the offices they hold. Offices are not'appropriated' by their incumbents. In theory, then, and leaving aside theproblem of 'red tape', a system of bureaucratic rules and offices,irrespective of the individuals who apply or hold them, should guaranteeequity of treatment among Black and White Zimbabweans. Rules of recruit-ment and promotion should treat all equally on the basis of their qualifi-cations. With the exception of subsidiaries of transnational corporations,however, industrial workplaces in colonial Rhodesia were generally notbureaucratized. Formal recording procedures that might have beenexpected to be part of a large work organization, as 1 have described forZimtex (Cheater, 1986, 105), tended to emerge as a result of quarrelsamong interdependent individual managers over the availability of re-sources necessary to ensure production in their own sections. Only afterIndependence did the ethos of bureaucracy become widespread asgovernment legalized new rules of procedure, insisting, for example, onthe formal recording of wages paid and received even among domestic72 'WE ARE TAKEN AS SHOVELS .workers, as well as instituting new procedures for dismissal which nowrequire government approval in each individual case.It should be noted that the effects of bureaucratization, which areperceived by the workers and managers themselves to have a negativeimpact on their work motivation, have resulted at least in part from thisexternal imposition of a new system of work rules. However, where thoseconcerned have instituted the rules themselves to protect themselvesagainst their colleagues, bureaucracy has grown 'from below' without thesame threatening impact.Prior to Independence, workers, as well as management, generallypossessed experience rather than formal qualifications. Those workerswho had proved themselves on the job and had been promoted into linemanagement as supervisors and junior managers were often very poorlyeducated and totally 'unqualified' in bureaucratic terms (see, for example,Cheater, 1986, 58-63). Their on-the-job skills, specific to particular enter-prises, were acquired over extended service with a single company andwere not easily transferable. This type of skill acquisition, then, lockedthem Š usually voluntarily Š into loyalty to their individual firms and (asKapferer (1972) has also noted of this type of employee) into a commitmentto resolving problems in that organization over the long term, rather thanfacilitating their move to another firm. In the new bureaucratized order,however, their informal, uncertificated skills have been severely under-mined by the emphasis on formal qualifications and the discounting ofexperience.There is, of course, a problem of quantifying and comparing the on-the-job experience possessed by most employees before Zimbabweanindustry was bureaucratized. At Independence managerial experience inparticular was held mainly by Whites and for that reason was especiallyproblematic. Therefore, those in favour of a bureaucratic solution toZimbabwe's racial imbalances can argue that 'twenty years' experience' isnot actually the experience of twenty years but, rather, 'one year's experi-ence repeated twenty times' where a person has done the same job for allof that time. Such arguments, if upheld, terminate the further promotionprospects of employees with experience only.The discounting of experience in the new bureaucratic order has notonly been especially detrimental to unqualified, unskilled workers, butalso to those with practical qualifications. For example, over the pastdecade many jobs have been regraded, particularly in large organizationsadopting a simplified grading system such as the Paterson Method. Suchregrading is done, in the language of management, to 'rationalize' minute'discrepancies' of relative skill and status which actually mean a verygreat deal to a formally unskilled workforce. In some large enterprises theloss of job status has led to considerable disaffection among workers andANGELA P. CHEATER 73has contributed to declining productivity. Workers have been demotivatedby a bureaucratization which they do not understand except in its negativeimpact on their own systems of ranking themselves. It was precisely in thecontext of re-grading within the Paterson Method that a worker who hadhad his job down-graded made the remark that I have used in the title ofthis paper: 'We are taken as shovels, used and put aside . . . workingwithout a job title. Even road sweepers have a job title.' And in the samecontext, another worker noted that 'deprived one can be very annoyedand go off disgruntling'! The workers' problem is that they do not have thenecessary education, or often even the skills of literacy and numeracy, tobe able to understand or manipulate a bureaucratized workplace. Bureau-cratization has increased their disadvantaged position at work, andworkers' representatives who have understood the source of this disad-vantage have pin-pointed it very precisely in their resolve 'to fight this evilwritten down to the workers' (my emphasis).The bureaucratization of recruitment is worth specific mention herein the context of unskilled as well as skilled and managerial jobs. Colonialsystems of hiring at the company gate men recommended by supervisorsor existing workers was anathema to the new government, which insistedon formal hiring, wherever possible through a labour exchange whichkeeps a waiting list of job applicants. As I have pointed out in the case ofZimtex (Cheater, 1986, 43-4), unskilled workers have a different view.Their view is a result of the past mode of recruitment, which attracted,from a limited number of chiefdoms and villages, numerous chains ofrelated migrants (shown in detail for Zimtex in Cheater, 1986, 43-58). AtZimtex, these migrants, over 40 per cent of whom came from beyondZimbabwe's borders, settled in the company village, married (often locals),and produced families which had very little contact with their fathers'rural homes. For these 'urbanized' offspring of proletarians, their sub-sistence depended not on access to rural land but on gaining employment.Zimtex's workers therefore regarded their offspring, kin and friends ashaving a moral entitlement to employment in their company which shouldtake priority over all claims to employment by outsiders. They saw Zimtexnot as an impersonal bureaucracy but as a 'moral community', whosemembers had obligations to, as well as rights over, one another. AlthoughI did not gather comparable data from Zimcor (which also had a companyhousing estate) concerning its workers' origins, they, too, regarded theirkin as having a moral entitlement to employment with their company. Yetthe preferential employment of kin is part of an existing 'corporate culture'which bureaucratization seeks to eliminate.Those with formal qualifications of a practical rather than a theoreticaltype have also been disadvantaged in recruitment and especially promotionby the rigid ranking of qualifications. A degree, for example, outranks a74 'WE ARE TAKEN AS SHOVELS . . .'Higher National Diploma in engineering, so even if the HND-holder is morecompetent as a result of his longer experience on the job than the degreedengineer, the latter is likely to leapfrog the former through promotion.And, while workers generally do not comprehend the new bureaucraticrequirements, managers, Black and White alike, who are experienced onthe job, are scathing about the real value to production and output ofhighly-ranked formal qualifications, especially when they are not accom-panied by experience, since their holders often prove, if not incompetent,at least very slow on the job for an extended period of time while they gainpractical experience. Perhaps the last weapon available to those whodefend personal experience against formal qualifications is ridicule: 'Theywill want degree people even in the mill here Š to pick up the scrap!'Executive managers who favour the bureaucratization strategy, rightlynote the threat to the continuity of the management of an enterprise thatis inherent in the experience of individuals: 'What happens when you areno longer here? We must write it down. We must create a company historyof procedures, so that the company can continue.' For immobile workerswho assume that they will be with a company for life and will be succeededby their children and grandchildren as company workers, there is, ofcourse, no such problem. Their personal expertise will be available onrequest, even after retirement in the locality. But the bureaucratizationstrategy assumes that people will move from one job to another.Indeed, the emphasis on formal qualifications and discounting of on-the-job experience are designed precisely to ease job mobility among theformally well-qualified rather than to promote loyalty to an individual firmover a long time. This point is absolutely apparent when one looks atfigures for length of service among different categories of employee: theaverage length of service was 8,9 and 13,5 years in the two companiesstudied but, in both cases, managerial and administrative staff had servedtheir respective companies less than half of these average times. Yet theysought to impose their bureaucratic view of ephemeral work commitmenton those whose cosmology of work included lifelong service.Lifelong service also represents a form of job security which seemed,at least at Zimtex, to be predicated on the distances from which workershad migrated to the company. Perhaps because they had come so far,these workers' visits home were few and irregular. They seemed disinclinedto change jobs when the security of home was so far away. This attitudepredisposed towards a stable workforce with little turnover but, togetherwith their acquisition of company-specific skills referred to earlier, it alsoexposed workers to grave threats when their companies went throughlean times and had to retrench the older, infirm, long-serving employeesno longer capable of doing heavy manual labour (see Cheater, 1986, 135).There is yet another way in which workers are disadvantaged byANGELA P. CHEATER 75bureaucratic work organization: in the distribution of rewards accruingfrom the value of their labour. The bureaucratization of managerial know-how requires, at least initially, a considerable expansion of administrativestaff to produce and apply the written rules required. Administrative andmanagerial jobs carry high salaries and many perks, yet these are notproductive offices in an industrial enterprise and must, therefore, be paidfor from the value of workers' labour realized in the sale of output. Ifworkers are sometimes hazy about the precise mechanism by which thistransfer of value is effected from themselves to others, they have nodifficulty in understanding the principle when their own transport servicesare cut to save money while managers simultaneously receive new cars,as had actually happened in the larger of the two companies I studied. Andthey are well capable of arguing (as at Zimtex: Cheater, 1986, 135) thatworkers' jobs and wages may be preserved by the withdrawal of expensiveperks from management.As administrative management has been expanded in the bureau-cratization of work, one result has been the shifting upward of respons-ibility. Trivia like leave and overtime forms, once the responsibility ofsupervisors, gravitate up the management hierarchy. Higher-levelmanagers, burdened by the new paperwork, abandon their shop-floors toattend meetings and sign papers, and things start going wrong morefrequently. The new rules, for example concerning dispute settlement,remove disciplinary authority from individuals and vest it in committees.Feeling themselves under threat, individuals gradually withdraw fromtheir remaining responsibilities while complaining that the newbureaucratic systems do not allow managers to manage or supervisors tosupervise. Finally, decisions which should properly be taken bymanagement are put before the Board of Directors, who complain abouthaving to perform managerial functions. Meanwhile, those lower-levelmanagers who have lost their responsibilities regard themselves as nolonger trusted to perform the jobs for which they were hired. Indeed, theirjobs may be formally redefined as the hierarchy is expanded to create newadministrative jobs.THE COMMUNICATION OF INFORMATION IN THE NEW HIERARCHYAs the new hierarchy expands, communications are, paradoxically,restricted to single channels. No longer do junior managers have directaccess through an 'open-door' policy or by accidental meetings in theworks to those more than one level above them in the hierarchy. They nolonger control the information fed upward from their own level but theyare in a position to remove or alter the content flowing upward from theirsubordinates and, sometimes, that coming down from superiors. Control76 'WE ARE TAKEN AS SHOVELS . .'over the content of information substitutes for control over its direction.Information is thus more easily politicized as a tool to create relations ofpatronage over subordinates Š which contradicts the theoretical intentof bureaucratization Š as this 'gate-keeping' function has been slippedinto formal offices to the advantage of individual incumbents.Where official communications have been 'streamlined' in this way,one would none the less expect that information would be exchangedinformally through the social networks of individual managers and workers.While this certainly happens among workers, since Independence it hastended to happen less and less among managers. Some firms, like Zimtex(Cheater, 1986), underwent their managerial revolutions around this time,as the founding age-cohort of managers retired simultaneously and werereplaced by strangers who tended not to socialize with one another afterwork. White managers resigned from larger enterprises, like the heavyindustry studied, and were replaced by Blacks who tended not to beabsorbed into the existing managerial social networks as friends nor toremain very long in any given job before moving on to better prospects.Nor, it would appear, did the job competition among upwardly-mobileBlacks permit the creation of even temporary new social networks amongBlack managers. Instead, competitive suspicion resulted in small cliquesof temporary allies exchanging gossip (usually detrimental to the interestsof 'the opposition') in local pubs.In addition to the formal communications flowing up and down thesupervisory hierarchy of the enterprise itself, other organizations also actas channels of communication. In the past, the trade unions, through theirshop stewards, have handled grievances over pay and other matters.Since Independence some union functions (especially those dealing withshop-floor grievances) have been taken over by the workers' committees.For this and other reasons the relationship between workers' committeesand unions in most workplaces, including both of those I studied, hasbeen one of considerable suspicion. Some enterprises, like the heavyindustry studied, refuse to recognize any other associations of workers(such as supervisors' or secretaries' associations) which could potentiallyact as channels of communication, and insist that all issues be handledthrough the workers' committees (as does the state's Department ofLabour Relations: see Cheater, 1986, 130). Such bureaucratization ofcommunication has been enormously frustrating to workers with interestsdifferent from those of the mainstream which they can no longer com-municate to management.Moreover, the state has required a single organization (the workers'committee) to validate communications not only between workers andmanagement but also between management and the state Š for example,in signing applications for retrenchment. This bureaucratization of corn-ANGELA P. CHEATER 77munication channels may simplify the state's tasks but, as I have indicatedelsewhere (Cheater, 1986, 130), it has imposed on these new and poorly-trained committees an additional impediment to their ability to functionas 'workers' representatives'. In addition to their new intercalary role uis-a-vis management, they are required to line up with management againstthe workers if the enterprise is get what it wants from the state Š notablythe right to fire workers. Members of workers' committees thus run the risk ofbeing categorized as vatengesi (sell-outs) as a consequence of performing thecommunications role that they are required by the state to fulfil.INTERCALARY AND OTHER UNCOMFORTABLE WORK ROLESIn all representational systems some roles are 'caught in the middle' oftwo conflicting sets of expectations which arise from the brokerage functionof these roles. Classically, in those parts of Africa colonized by the British,this intercalary 'squeeze' affected indigenous chiefs and headmen(Gluckman and Barnes, 1949) who simultaneously represented the interestsof their people to the colonizers and acted as the lowest-rankingadministrators of the new state system. In industry supervisors occupy asimilarly uncomfortable position, being caught between the workers theysupervise in particular work units spatially separated from others and themanagement whose lowest rank they are. 'Workers think that when it suitsus we join management and when it suits us we join workers.' Executivedirectors may be similarly squeezed between the interests of the employeesof the enterprise they manage and the Board of Directors representing theinterests of its owners (see Cheater, 1986, ch. 5). The workers' committees,because they are required to support management views in relationsbetween the enterprise and the state and because they are constituted asthe only legitimate channel of communication between workers andmanagement, have also been caught more and more in an intercalarysqueeze, which may not have been intended but certainly could Š andshould Š have been predicted from the structural principles involved.Towards the end of the colonial period in Zimbabwe, many Blackworkers were promoted to supervisory posts vacated by emigrating Whites.In the mid-1970s, as I have described elsewhere in detail (Cheater, 1986,ch. 5), Zimtex's executive management promoted Black supervisors withoutinforming the Board of Directors (presumably because it feared theiropposition to these promotions) and without clarifying the managementstatus of the new Black appointees. Black supervisors were an anomalywithin the Zimtex system because, in the past, Blacks had always beenworkers and supervisors had always been White. The responsibilities ofthe new Black supervisors' jobs were never spelt out and the conditionsunder which they worked (including their pay, which was much lower78 'WE ARE TAKEN AS SHOVELS .than that of the Whites whom they replaced) suggested that they wereseen by the company primarily as Blacks rather than as supervisors. Forfive or six years this anomaly continued until the Black supervisors wenton a four-day strike late in 1981. They were summarily dismissed and losttwo days' pay but were immediately reinstated in their jobs withoutinterruption to their continuity of employment and the company finallypublicized their position and status as part of management in a formalnotice to the entire works.Clearly, this example is precisely the reverse of the 'window-dressing'Black appointments so often complained of after Independence, when Blackswere hired to positions with good pay and high status but few productiveresponsibilities. Yet one of the post-Independence managerial complaints isthat Black supervisors will not exercise the responsibilities of their position,which complaint has also been used as an excuse for the relocation of theseresponsibilities to a higher level in the managerial hierarchy. White managersin the heavy industry I investigated believed Black supervisors to be corruptedby the extended family system and reluctant to exercise work authority overkinsmen. Black managers noted that the localized training for supervisorswas not of the same quality or standard as the overseas courses used beforeIndependence. The supervisors themselves said: 'Supervisors can't disciplineanyone. When the workers' committee get up in arms, management areafraid to make decisions and stand on your side', a view with which somemanagers themselves concurred.Supervisors (of all races) were also acutely aware of the importanceof personal, face-to-face relations on the shop-floor. When a supervisorrequires extra labour to finish an urgent job, he cannot order workers towork overtime, for that is not part of their contractual obligations to thecompany. He must persuade them to give up part of the leisure which isthe company's contractual obligation to its employees. For a 'good man',workers will oblige; but for a 'hard supervisor', who has refused their earlierrequests for leave at a specific time or other 'favours', workers may refuse anovertime request, and jeopardize specific contracts. The more disorganizedand inefficient a company is, the greater will be the demands on its supervisorsin this area of work negotiations. If increasing bureaucratic rigidity in theapplication of company rules removes their freedom to negotiate, supervisorscannot supervise work effectively: and 'if we can't supervise effectively,we're useless'. At both the companies studied, the Black supervisors usedtheir personal knowledge of workers' private lives in such work negotiationsand regarded living in the same company township as a useful source ofsuch personal information. Workers feared that their supervisors wouldtransmit such information further up the management hierarchy.One of the functions of brokers is, of course, to communicate, butsupervisors were particularly incensed (because it impaired theirANGELA P. CHEATER 79negotiating ability on other matters) at always being required to breakbad news to their workers (about job regrading, for example, or the loss ofbenefits and perks). At Zimtex, there was no personnel department tohandle such issues, but at Zimcor the supervisors felt that their personnelsection was evading its newly defined bureaucratic responsibilities in thisrespect to the detriment of supervision generally.WORKER-MANAGEMENT RELATIONSAt both workplaces studied executive management regarded their enter-prises' difficulties as resulting from the failure of the workers in their over-manned organizations to do a decent day's work. 1 witnessed, in workscouncil meetings in both workplaces, executive managers argue that smallperks recently withdrawn from the workforce could possibly be restoredonly if output rose. These small, non-taxable perks Š in the poorly-paidworkers' view, 'small things [which] matter most' Š were the colonial 'in-kind' insulation against low wages, and ranged from food rations throughcompany transport to sub-standard company products. They have becomeeven more important in the late 1980s as inflation has completely erodedthe workers' earlier nominal increases in wages, which have, of course,also exposed them to direct taxation from which they were formerlyshielded. While executive management removed perks from wage workers,they regarded managerial perks as contractually guaranteed, defensible inlaw and, therefore, not to be tampered with. In contrast, workers who hadinvested much of their working lives in these moral communities forwhich they produced, argued that, at least with respect to entitlements to(differentiated) perks, 'tose we are equal, management nevashandf ('weare all equal, management and workers').The executive view that idleness among the workers was responsiblefor falling output was disputed, in both companies studied, by middlemanagement, who were largely responsible for the manning levels. In bothcompanies, middle management acknowledged severe managerialdeficiencies, going right to the top of their organizations, and were much lessinclined to blame workers than management for the crises of their companies.Bureaucratized 'management by committee' and particularly collectivizeddisciplinary action, they felt, were the prime internal causes of decliningoutput Š shortages of foreign currency, spares and raw materials being themost important external constraints, over which firms continue to have littlecontrol. Those middle managers whose stinginess in the allocation of theproductive resources under their own control paralleled that of their Boards(see, for an example. Cheater, 1986, 105-7) tended also to be the mostcritical of their own workers.Middle managers were generally regarded by workers as more80 WE ARE TAKEN AS SHOVELSreasonable than 'them up there' Š the newly bureaucratized executivemanagement, rarely if ever seen on the shop-floor. In part, even whenindividual managers were personally disliked, this view stemmed fromtheir face-to-face relationships with workers, their experience and abilityto sort out technical problems on their shop-floors, and their willingnessto experiment with work organization in ways that were financially bene-ficial to their workers even though they violated bureaucratic requirements(see, for a Zimtex example, Cheater. 1986, 113-4). Middle managers alsodefended their workers against the rest of the enterprise, objecting, forexample, to changing production runs in ways that disadvantaged theirworkers' capacity to earn bonus wages, even if those run changes were inthe company's short-term interests. Like supervisors, middle managementalso had to negotiate with their workers if their departments or divisionswere to run smoothly.This negotiation often involved managers in paternalistic relations(reminiscent of the colonial period and anathema to the new bureaucraticorder) which were articulated by workers in ways designed to emphasizemanagerial obligations to themselves, often using the idiom of kinship.Whereas in the colonial past White managers had accepted this rolewithout much difficulty, post-Independence bureaucratization nowprovided a rationale for those who wished to evade such obligations: 'I amnot your "father", much less your mother, and it's not my business. Timeshave changed, you know: we live in Zimbabwe now!'Black managers and supervisors in independent Zimbabwe were morelikely than their White peers to respond to such pleas to exercisepaternalism. But Black managers also, paradoxically, tended to be morebureaucratic in the way« in which they exercised their managerial functions.The result of this paradox was the growth of patron-client relations withinbureaucratic structures, together with a heightening of workers' per-ceptions of their conflicts with management as a class of patrons whotoday treat workers as bureaucratized objects ('shovels', or "pens') ratherthan as human beings with individual needs and wants.Nowhere was this conflict more apparent in the companies I studiedthan in symbolic aspects of the worker-management relationship, whichsymbolism the new bureaucratic managers (and some of the older ones,too) regarded as 'petty". The provision of refreshments to those workingin hot and humid environments; on-duty celebrations when a shift brokeaproduction record; the knowledge of workers' personal names and theobservation of basic politesse; the failure to offer subordinates a chairwhen discussing production issues in managers' offices; insanitary toiletfacilities; the managerial insistence that workers queue in a straight line toreceive their pay; the workers' (lack of) access to telephone messages;their entitlement to and use of soap, mutton-cloth and protective clothingANGELA P. CHEATER 81Š all of these issues recurred again and again on the agendas of thecompanies' works councils' meetings, much to their managements'frustration. But, as the workers pointed out when this managerial frustrationwas verbalized, the reason why they remained on the agendas was manage-ment's failure to resolve these issues to the satisfaction of the workers.'You expect the demoralized worker to come begging on his knees' for therestoration of minor privileges removed to penalize falling production, ifyou want a big tree to grow, water the roots, those small things thatcollect nourishment' among workers concerned with their own statuswithin a productive organization.Status is recognized as being important to management, where it isreflected in differentials of pay, the type and quality of fringe benefits, thesize and furnishings of office accommodation, and access to certainexclusive company facilities (the Boardroom, sporting facilities, on-siteparking, etc.). Managers report themselves to be very status-conscious.Managerial committees, for example, prefer to keep company housingempty rather than allow occupancy by ranks not entitled to that grade ofhousing, which preference is often specified in the bureaucratic rulesgoverning housing allocation; and execut ive salaries are usually kept secreteven from the personnel and data pr<» essing departments responsible forpaying wages and salaries. At company social occasions (as 1 havedescribed elsewhere: Cheater, 1986, 103) distinctions of rank within thework organization are translated into spatially-discrete groupings ofsimilarly-ranked employees. Yet enterprise management appears singularlyignorant of the importance of status gradations among unskilled and semi-skilled workers.When differences of earned income are reduced, even to zero, otherindicators of relative status within a large organization become more, notless, significant to those workers who, because their 'concerns are sofrequently focused on simple survival and subsistence', fail to be 'suitablyworthy or valuable' in the eyes of management (Cheater, 1986, 138). InChina, for example, where unskilled manual road-workers used to earn thesame as health clinic attendants, the latter job was seen as infinitelypreferable for its working environment, relative ease and cleanliness.Clinic attendants consequently enjoyed a much higher job status thanroad-workers, and much corruption was involved in individuals obtainingtheir allocation by the state bureaucracy to such jobs for which 'back-door connections' (through kinship, friendship and the party) wereessential.Among impoverished Zimbabwean workers, status inheres inpermanent employment in a designated, stable job with a clear title, notmerely an impersonal pay scale shared with hundreds if not thousands ofothers in a large bureaucracy. Jobs which permit access via skills82 WE ARE TAKEN AS SHOVELS . . .'acquisition to a 'labour market' or ladder of promotion internal to theorganization, are particularly prized, and may rank higher in the workers'estimation than their formal value to the organization (Cheater, 1986, 58-63). 'Front-line' production workers (such as weavers in a textile factory),on whose labour (cloth) output is critically dependent, enjoy a specialstatus among their fellow workers.The corollary of workers' rankings of their own importance to anenterprise is that threats to their systems of ranking themselves have aparticularly demoralizing effect on the organization as a whole. Re-gradingexercises, especially if they are undertaken regularly, are the equivalent of'permanent revolution' and undermine the entire foundation on whichworkers construct their own hierarchical reality in the workplace. Likewise,demotion as a disciplinary measure is a total threat to a worker's self-esteem, since it forces him to re-rank himself relative to former subordinatesof lesser status. Some workers prefer dismissal to such a reversal of statusand loss of face. Indeed, in one of the firms investigated, a worker demotedfor a year and whose resignation was refused on technical bureaucraticgrounds, repeated his original offence in order to be fired: dismissal was,for him, the only release from an even more unacceptable loss of statuswithin the workplace. Some managers thought him mad, assuming that hisresponsibility to his family had to outweigh his own self-image. But whilefamily responsibilities and subsistence do undoubtedly loom large as amechanism by which the majority of workers may be controlled, theactions of those who are not prepared to accept this definition of theirown unworthiness as human beings or to be held to ransom by theirdependants' subsistence requirements show with great clarity the import-ance of status to ordinary workers.Workers' ranking of themselves is also affected by bonus systems,particularly where some workers and managers earn their bonuses on thebasis of productive work by others. These 'others' may be workers inother units, as when service staff, including administration, are paid ageneral bonus' based on the averages of all production units, or, in thecase of managers who are paid bonuses not on the quality and efficiencyof their own performance of managerial tasks but on the output from theirunit, these 'others' may be workers in their own units. Having othersriding on our backs' in a bonus system, may cause those workers affectedto diminish their own work input, to ensure that the others do not earn abonus which, via increased income, may enhance their status relative tothose on whom the bonus-earning capacity depends. As I have pointed outin my study of Zimtex (Cheater, 1986,66), resentment of what is perceivedas an unfair distribution of bonus pay may well 'drag output down to alevel that threaten[s] the Company's financial viability' as workers devisestrategies to defend their own status in the organization 'as a series ofANGELA P. CHEATER 83worker-production units defending their own interests against other, similarunits'. If, under such circumstances, a firm was to go bankrupt, it would beat least as much the fault of management for not understanding that thesource of the problem lies within the social relations of the 'informalenterprise' as the fault of workers for withdrawing their labour becausetheir position has not been understood.CONCLUSIONIn this article I have approached the organization of work and workers inZimbabwean enterprises 'from within' and 'from below', concentrating onthe relationships between social interaction in the workplace and theeffective organization of labour. Such an approach emphasizes theimportance of symbolic as well as social elements in the ways in whichproduction is effected. It is an approach uncongenial to organizational andbureaucratic theories which play down the significance of human andsocial considerations. Yet any investigation of work, in Zimbabwe orelsewhere, that fails to take account of the fact that 'workers, too, arehuman', will paripassu offer an incomplete analysis of the social processesthat enable workplaces to produce at all.Indeed, many Zimbabwean firms have, since Independence,experienced declines in output and worker productivity. Some elementsof capital and its management have attributed these declines to 'over-protective' legislation, which first defined minimum wages and thenimposed piecemeal controls on hiring, firing and worker representation indecision-making. Later these provisions were consolidated into the LabourRelations Act, which was criticized even in its draft form, by labour andcapital alike, for concentrating bureaucratic authority over the workplacein the hands of the state and government. This legislation has affectedproduction detrimentally, but perhaps not primarily as a result ofconcentrating authority in the hands of state bureaucrats who makedecisions very slowly. Instead, I would argue, the failure of this labourlegislation is rooted much deeper: in its organizational unacceptability toboth management and workers involved in social relations of productionon the shop-floor.