Automation and Labor: Is Marx Equal to Adam Smith?

RETHINKING MARXISM VOLUME 16 NUMBER 4 (OCTOBER 2004) Automation and Labor: Is Marx Equal to Adam Smith? Benedito Moraes-Neto Translated by Paula Matvienko-Sikar The qualifications for employment within the modern microelectronic-based, automated systems can be understood as a negation of the Marxist claim that work would come to demand less skill as technology developed. This paper attempts to criticize this interpretation by seeking the work-deskilling concept in the writings of Marx himself. The result is the proposition that that which is observed in the modern factory*/that is, the radical dispensability of living work*/really mirrors work deskilling according to Marx. The more usual idea of work deskilling, attributed erroneously to Marx, is in reality Smithian in nature. Based on this analysis, a critical analysis is made of Labor and Monopoly Capital by Braverman, which has become accepted as the definitive interpretation of the ideas of Marx on the subject. The sole cause for confusion arising from equating the Marxist and Smithian analyses concerning technology and work should be attributed to an incorrect understanding of Taylorism and Fordism. Here we propose that recent technological developments in reality signify an end to the mistake of equating Marx with Smith, and also indicate the great relevance of Marx today. Key Words: Labor Process, Technological Automation, Deskilling, Marxist Analysis Development, Taylorism-Fordism, The effects of automation on labor content, particularly in microelectronic-based automated industry, can be used as an example of the theoretical failure of Marx in a very important area: the labor process under capitalism. In other words, the wellknown Marxist assertion that technological change would lead to growing deskilling of workers is not occurring because of essential training requirements in modern microelectronic-based, automated systems. This study will concentrate on analyzing that proposition. For this purpose, it is necessary to highlight the particularity of Marx’s theory concerning the implications of automation for labor directly applied to production. In addition, we will try to adapt Marx’s concept in the case of modern microelectronic-based automation. The effort to use Marxist theory to explain this important aspect of the present stage of capitalism will be followed by an attempt to distinguish the difference between the ideas of Marx and Adam Smith on the development of productive forces and the division of labor. This difference is of special theoretical relevance because recent literature in general wrongly attributes to Marx concepts typical of Adam ISSN 0893-5696 print/1475-8059 online/04/040407-16 – 2004 Association for Economic and Social Analysis DOI: 10.1080/0893569042000270898 408 MORAES-NETO Smith regarding labor and technological progress. The distinctions between Smith and Marx have consequences in the critical analysis made here of the well-known Labor and Monopoly Capital , by Braverman, which has generally become the ‘‘official’’ interpretation of Marx’s theoretical propositions about the nature of the labor process under capitalism. The whole basis for the discussion outlined above depends on whether Marx’s proposition about the tendency toward deskilling of the labor force as technology progresses is compatible with the nature of skilled labor now used in microelectronicbased automated plants. This paper will try to show that such a basis exists. In expressing Marx’s analysis in the terms that follow, the intention is to fashion an updated and clear formulation of the famous contradiction between productive forces and relations of production. A further aspect of this study is the emphasis given to the metal-mechanical industrial sector, which has a peculiarity largely responsible for present theoretical mistakes concerning the nature of the labor process. Microelectronic-Based Automation: Skilling versus Deskilling Automation and Deskilling: The Textile Machine as a Paradigm Before anything else, automation must be defined. Usually, it is understood to occur only when the workers in control of the machines are replaced. Alfonso Fleury expresses the difference between the concepts of mechanization and automation as follows: ‘‘mechanization is the result of using equipment . . . which substitutes manpower in the production process . . . [while] automation is based on information and replaces the worker in control of the equipment’’ (1988, 18). Using control as the basis for defining automation excludes rigid or dedicated automation, as can be deduced from a statement by Angelo Dina: ‘‘Consider a machine designed and constructed to produce only a certain part and thus reduced to a highly simplified operational unit. All necessary geometrical and technological information is condensed within its physical structure’’ (1987, 14). In this case, there is no control function to substitute and, therefore, not even the possibility of automation by Fleury’s definition. However, if it is understood as occurring with every replacement of manpower in the production process through the technological application of science, then automation, by its dedicated nature, has been well known since the beginning of the textile industry revolution. In the previous case, it is relevant that the spinning machine (patented in 1738 and advertised as able to ‘‘spin without fingers’’ [Marx 1952, 181]) and the most modern ‘‘open end’’ spinning machine are identical theoretically, in terms of the functions carried out by both the subjective and objective elements of the production process. This can also be said of the mechanized steam loom of the midnineteenth century and the modern shuttleless loom. In considering the extremely advanced nature of nineteenth-century textile machinery, Marx singled out the following defining characteristic: AUTOMATION AND LABOR 409 As soon as a machine executes, without man’s help, all the movements requisite to elaborate the raw material, needing only attendance from him, we have an automatic system of machinery. (1952, 186) The production process has ceased to be a labour process in the sense of a process dominated by labour as its governing unity. (1973, 693) [T]he entire production process appears as not subsumed under the direct skilfulness of the worker, but rather as the technological application of science. (1973, 699) Using the textile machine as a fundamental theoretical (and historical) reference, the metal-mechanical industrial sector will now be turned to. Its study may yield unexpected results. The Machine Tool of the Metal-Mechanic Industry Marx’s considerations about industrial machine tools indicate that in the nineteenth century this extremely vital innovation for furthering industrial development had already been adjusted by industry to the ‘‘machinery principle,’’ just as had happened in the case of textile machinery. Marx’s enthusiasm over the innovations brought about by machine-tool use was evident in the famous quotation from Capital regarding ‘‘machine production by machines:’’ The most essential condition to the production of machines by machines was a prime mover capable of exerting any amount of force and yet under perfect control. Such a condition was already supplied by the steam engine. But at the same time it was necessary to produce the geometrically accurate straight lines, planes, circles, cylinders, cones, and spheres, required in the detail parts of the machines. This problem Henry Maudsley solved in the first decade of this century by the invention of the slide rest, a tool that was soon made automatic, and in a modified form was applied to other constructive machines besides the lathe, for which it was originally intended. This mechanical appliance replaces, not some particular tool, but the hand itself, which produces a given form by holding and guiding the cutting tool along the iron or other material operated upon. Thus it became possible to produce the forms of the individual parts of machinery ‘‘with a degree of ease, accuracy, and speed, that no accumulated experience of the hand of the most skilled workman could give.’’ (Marx 1952, 187/8) We do not intend to deny the economic effects of the switch from manual machine production to the slide-rest lathe, which evidently increased labor productivity. However, it appears that Marx greatly exaggerated in concluding that such a machine tool was compatible with his theory on machinery. In fact, the slide-rest lathe is the prime example of multipurpose machine tools, so-called because of their flexibility*/that is, capacity to produce different part types, due to which multipurpose machine tools became essential for metalworking and are widely used even nowadays. The well-known operational characteristics of these machines are clearly explained by José Ricardo Tauile: 410 MORAES-NETO Due to frequent changes in the product (small series, batches, and specially ordered parts) workers must be experienced and highly skilled manually . . . The worker analyzes the project and, based on his own experience and judgement concerning the adequacy of the equipment, determines if changes are necessary . . . Having done this, the job can begin: the worker inserts the parts and tools into the machine, operates levers, cranks, and other commands related to parts and tools, sets the speed, turns the cooling fluid on, etc. The manufacturing process itself requires years of experience so as to ensure problem prevention and immediate action when necessary. A small change in the color of a chip might mean damage to the whole part, or a slightly different noise in the machine tool might result in a part being wasted. (1983a, 23/4) Based on the previous consideration, we are far from the worker conceived as an appendage, which is an inevitable consequence of the machine according to Marx. On the contrary, instead of understanding the worker as an appendage of the machine, doing work without meaning, superfluous and reduced to an abstraction, he can be conceived as having a function similar to that of a craftsman. And the machine that made this possible continues in importance precisely because of its productive flexibility, a direct consequence of the immense versatility of man as a production instrument. In fact, by the terms of the definition of automation previously presented, the multipurpose machine tool is by no means an example of automation but rather of its negation. This, in turn, raises the following question: How did automation manifest itself within the mechanical production process? As stated clearly by Angelo Dina, this was in the form of rigid or single-purpose automation, which reached its maximum development with the Cyclopean transfer machine, whose locus par excellence lays in the American automobile industry. The totality of mechanical production processes had been marked, until the incorporation of microelectronics, by the duality ‘‘rigid automation/flexible nonautomation.’’ It is in light of this duality that we must reflect on the impact caused by the coming of microelectronics on labor content and worker qualification requirements. But at this moment, it is important to stress that ‘‘only through the development of microelectronics was it possible to break the ‘rigid automation/ flexible non-automation’ pattern. This clearly occurred because microelectronics allowed the development of equipment that to some extent had capacities previously monopolized by man . . . Automation, i.e., material production without direct human intervention is no longer synonymous with rigid or single-purpose automation. So that automation now refers to a flexible process as well’’ (Moraes-Neto 1986, 36). Microelectronics incorporation at the machine-tool level in the metal-mechanical industry reflects a double movement: a. In substituting multipurpose machines by numerically controlled machine tools (NC), a totally automated machine of great complexity suddenly replaces a machine backward in conceptual terms because heavily dependent on skilled work. b. Theoretically similar, both the new machine and the dedicated (single-purpose) machine fit within the automation concept adopted here. In case of substituting AUTOMATION AND LABOR 411 the NC for the dedicated automated machine, what happens is not the emergence of automation as in the preceding example, but a change in its form. With respect to the first of the double movements, a question arises as to whether this change has led to a work deskilling process similar to that which happened in the transition from craftsmanship to classic automation. In the case of the second, the deskilling process is not even involved. The question of deskilling itself gave rise to a debate in which an essential reference is Braverman, who works with the notion that labor deskilling is emblematic of a broader process of increased deskilling as the automation level grows. On this matter, Braverman points out that NC are definitely replacing the multipurpose machines which have been widely used and have remained ‘‘the province of the skilled machinist’’ (Braverman 1974, 196). However, Braverman’s work overlooked the fact that the replacement of a machine tool ‘‘for any kind of purpose’’ by an automated machine at the end of the twentieth century is an astonishing fact in the social history of technology. Let us now concentrate on how Braverman treats the introduction of NC as it affected manpower. Initially he imagined it as an illustration of Babbage’s principle ‘‘now in a setting of technical revolution’’ (Braverman 1974, 200): the subdivision into three different operations of what had previously been the function of a skilled machinist.1 It is typical of Braverman to take a principle formulated within specific manufacturing conditions and extrapolate its use both indefinitely and indiscriminately. This can easily be shown, as will be done later, but in the present case Braverman exaggerated since a principle originating in a specific manufacturing system has little relevance in the case of such intense technological change. Let us now look at Braverman’s search for the effects of introducing NC into the work operation: ‘‘So far as the machine operator is concerned, it is now possible to remove from his area of competence whatever skills still remain after three-quarters of a century of ‘rationalization.’ He is now definitively relieved of all the decisions, judgment, and knowledge which Taylor attempted to abstract from him by organizational means’’ (Braverman 1974, 202). For this author, replacement of the multipurpose machine by NC doubtless results in a total loss of labor operation content. On this point, but more sharply expressed, is the comment of José Ricardo Tauile: ‘‘’The work of an NC operator from now on is reduced to loading the equipment with a raw piece and a suitable tool, setting the machine to begin, pressing ‘start,’ and supervising the process so as to stop it in case of partial or complete equipment breakdown’’ (1983b, 25). 1. Considering the importance in this study of Babbage’s principle, he himself should be heard from first: ‘‘[A]ny explanation of the cheapness of manufactured articles, as consequent upon the division of labour, would be incomplete if the following principle were omitted to be stated. That the master manufacturer, by dividing the work to be executed into different processes, each requiring different degrees of skill or of force, can purchase exactly that precise quantity of both which is necessary for each process; whereas, if the whole work were executed by one workman, that person must posses sufficient skill to perform the most difficult, and sufficient strength to execute the most laborious, of the operations into which the art is divided’’ (1971, 175/6). 412 MORAES-NETO As time went on, the link between numerically controlled technology and labor deskilling became a significant theoretical reference. However, this concept has gradually been revised in light of later empirical observations and ‘‘in the very recent period, the debate on deskilling has once again reopened’’ (Kaplinsky 1985a, 102). The reason for this revaluation, according to Kaplinsky, is the transition from NC to computerized numerical control (CNC) which ‘‘offers the ability of the worker to interrupt the automatic working of the machine and to reprogram it on the shop-floor’’ (1985a, 102). It seems to us, though, that the transition from NC to CNC has actually reinforced a tendency already existent in NC machinery. In interviews of engineers with experience in installation and management of the new technology, it was evident that the operation of machines based on NC technology requires some level of ‘‘knowledge.’’2 This assertion may constitute a basis for criticizing Marx’s theory as it concerns technology and work. To determine if it does requires examining this question more deeply, and perhaps the best starting point is furnished by Elenice Leite: With respect to machine operators, there are two types of change to be considered: as we know, NC practically takes over part production, thus simplifying or limiting the operator’s task. On the other hand, the job of preparing the machine grows in complexity, placing increased demands in knowledge and ability on the worker. This double tendency with respect to NC operator requirements, involving increased simplicity in one part of the process and complexity in the other, mainly results in companies’ clear choice of more highly qualified workers able to operate new (and expensive) equipment more safely, independently, and responsibly. There are only two companies which prefer operators with extremely limited tasks to carry out [supplying raw parts and removing finished ones]. In the other companies, the NC operator planner is, or will shortly turn into, the dominant category, with the possibility of becoming an operator/adjuster/ planner as the CNC generation spreads. (1986, 35) The multipurpose machine confers on the worker the responsibility for making a broad range of decisions necessary in transforming castings into parts, a function requiring considerable precision, skill, technical ability, and experience as the process proceeds from the drawing phase to actual production. In the NC case, on the other hand, direct human interference in production is minimum.3 This is because 2. In a pertinent assertion, Kaplinsky distinguishes skill from knowledge as follows: ‘‘[I]t is necessary to discuss briefly the relationship between skills and knowledge, which are related but not identical concepts. Knowledge comprises an understanding of a process or information at an abstract level, such that it can be transmitted to another individual in a similarly abstract manner. As such, knowledge must be explicitly rationalised in abstract terms that can be readily understood*/a process that we have come to know as science and technology. On the other hand, skill comprises a set of practiced experience, which may involve not only the acquisition of knowledge, but also a greater or lesser degree of natural aptitude and implicit rules of operation. Skills are individually acquired and involve a combination of abstract learning, aptitude and experience, but the same is not true of knowledge, which is essentially abstract and less individual-specific’’ (1985b, 435). 3. It concerns, in the Davis and Taylor (1973) project, the connection man-product. AUTOMATION AND LABOR 413 the NC is a machine in the unrestricted sense of the term: it completely transforms a given material, based on the instructions already programmed into the machine. Of the worker/product relationship, therefore, the only responsibility left to the worker is supplying the machine and removing the finished product. This is a radical work transformation process, rendering human labor completely empty and superfluous, exactly as happens with the automation process in general. In the NC case, the deskilling process is so extreme that a specific operator function might disappear completely from the labor division within the productive process. (To Braverman’s disappointment, we might have a ‘‘Babbage’s principle in reverse,’’ which could intensify in the case of CNC machinery.) Let us now analyze what differentiates a machine with flexible automation from a dedicated automated machine in operational terms (we have already seen that operational work in the strict sense is devoid of content in either case). As shown in the words of Angelo Dina, the rigidly automated or dedicated machine is quite simple in its functioning; consequently, the supervision operation is also. For this very reason, Marx considers it extremely simple to operate machines. A very different process occurs with the NC, a machine whose ‘‘geometrical and technological information’’ is no longer ‘‘condensed in its physical structure’’ but in a computer program ‘‘translated’’ by the machine and implying great complexity of its supervisory function. The pièce de résistance of the operational work of a NC is exactly the supervisory function, which has no parallel in multipurpose machine operation. (‘‘Knowledge’’ and ‘‘skill’’ are not comparable.) Let us now examine further the inversion of Babbage’s principle with the introduction of CNC, which represents an application of a principle of great importance in microelectronics-based automation*/integration, in the measure in which it integrates in a single machine the production of a program (computer) and the making of a mechanical product (machine tool), in such a way that the tasks of operation, preparation, and programming are carried out by one person only. This integrated activity can be characterized as technical system management which is heavily dependent on a high level of knowledge and represents an inversion of ‘‘Babbage’s principle.’’ Considering the technical nature of the new machinery and its economic repercussions (the high cost of equipment and irrelevance of salaries in the cost structure), it makes no sense to maintain together with a CNC an operator (in the strict sense, placing unfinished and removing finished parts) earning a low wage, a better paid worker to prepare and supervise the machine, and a programming worker, the most highly paid of the three. Although no technical-economic rationale demands this peculiar task fragmentation (unless supplied by some die-hard proponent of Marglin’s reference to ‘‘divide and conquer’’ [1974]), it can happen. Thus, the choice between unification or fragmentation of the tasks, in the view of some, becomes an organizational one. This manner of understanding the impact of new technology on labor*/as being undefined because dependent on organizational choice, which is a critical view related to technological determinism, has recently become more widespread. For some authors, it has even eliminated the question completely: [C]are must be taken with technological determinism which claims that what is required of labor is imposed by technology, whereas this is really defined 414 MORAES-NETO by work and production organization, together with human resources administration policy and labor demands. (Salerno 1988, 65) This abandonment of the attempt to explain tendencies involving work and technology comes, strangely enough, shortly after the author has stated: ‘‘In the Ferrari plant in Italy the introduction of a flexible manufacturing system (FMS) has reduced the number of workers from 100 to 9 (8 engineers and 1 worker) without diminishing production’’ (64).4 To be coherent, Salerno has now to convince us that this radical change had little to do with the FMS introduction and much more with organizational choice, not an easy task perhaps. Is he saying that it would be possible to keep the same 100 workers in spite of the drastic technological transformation? On the contrary, we believe that the actual tendency manifests ‘‘Babbage’s principle inverted,’’ the case of FMS constituting an excellent example. Equipment integration in an FMS leads to operational activity in a strict sense becoming superfluous, since all steps are supervised in the broadest sense by a versatile programmer. In an interview, an engineer told us that ‘‘integration did not eliminate the particularities of the various parts of this equipment. But it does not help to have a programmer familiar only with machine A, and another one who knows about machine B because in this way, understanding of the whole process is lost.’’ Therefore, the clear tendency is to unify all functions, which would characaterize management of a highly complex technical system. The next step in the advance of microelectronics-based automation in the metalmechanical industry will take us to computer-integrated manufacturing (CIM), a process leading to the ‘‘unmanned factory.’’ But first we must focus on another facet of metal-mechanical production: the ‘‘astonishingly innovative’’ assembly line. Automation and Flexibility in the Assembly Line Surely the metal-mechanical industry would have proved an appreciable surprise to Marx. Let us just imagine his reaction at reading a text published in England in 1985: ‘‘Consider, for example, the skill of metal-working: in the mould industry, it takes 10/12 years of learning before a craftperson matures’’ (Kaplinsky 1985b, 436). Not less surprising would be the sight of a still very contemporary ‘‘wonder of twentieth century industry,’’ the Ford assembly line, which illustrates the maximum development of the Taylor proposition. The reasons for this imagined reaction are found in our work Marx, Taylor, Ford: Discussing Productive Forces, where we present the idea that ‘‘Fordism is based on the most extreme development of work in the manufacturing version’’ (Moraes-Neto 1989, 59). But in the present case, the 4. A useful definition of FMS is found in Simmon: ‘‘FMS involves a control system based on a central computer made up of two or more CNC machines equipped with an automated system for automatic part transference from one machine to another, and capable of operating over long periods without any human interference. The central computer controls the production from supplying material through process finalization, according to a production program stored in its memory. In another words, FMS produces ininterruptedly a large variety of parts in small batches without any human interference in the operation, and in a programmed production sequence’’ (1986, 43). AUTOMATION AND LABOR 415 astonishment of Marx would be motivated by something else. In the multipurpose machine case, as has already been pointed out, the major difference is the great importance of worker skill. The Ford assembly line, however, is based on a work division requiring massive utilization of nonskilled manpower. However, unlike what happens in the case of labor as an appendage of the machine, extremely simplified manual tasks are not manual-labor leftovers capable of being mechanized; rather, they constitute the essence of the labor process itself. Even so, if they are extremely simple tasks, what makes them unfit for mechanization? The fact is that assembly line work is too simple for a ‘‘man-machine’’ (the Taylorized man), but too complex for the machine in its classic sense. ‘‘Consequently, to adapt an assembly line to the generic characteristic of production based on machinery (to transform the assembly line into a machine system) a new type of machine is required which is a product of microelectronic development: the robot’’ (Moraes-Neto 1986, 38). If the assembly line, by the nature of the operations involved, can only be automated by a microelectronics-based machine which, as we have seen, is able to incorporate tasks previously carried out by a worker, and if all microelectronics-based automation is flexible, then the automated assembly line is characterized by productive flexibility. It is not difficult to understand that automation of an assembly line can be conceived of only within an integrated automation system, through its backward links with mechanical production processes based on microelectronics and, therefore, equally flexible. Obviously, it does not make sense to integrate a flexible automated assembly line into a setup of mechanical processes rigidly automated. Now let us consider the effects of assembly line automation on manpower. We will start by quoting Henry Ford on the skills of the workers on the assembly line: ‘‘Regarding the time necessary for technical apprenticeship, the proportion is as follows: 43% require no more than a day; 36% need from one to eight days; 6% from one to two weeks; 14% from one month to one year; 1% from one to six years. This last category of workers must be highly skilled to deal with instrument production and calibration’’ (Ford 1926, 125). Let us imagine the impact represented by the comparison of ‘‘time used in training’’ with automated assembly line characteristics found in an article on ‘‘the wonders of the 54th warehouse:’’ [T]wo small mechanical fingers choose one specific alternator chain, among many of several types. Slowly and steadily the ‘‘hands’’ approach the engine and install the rubber chain, in an almost human way. The images of the 54th warehouse (belonging to Volkswagen in Germany) are already part of the twenty-first century . . . Some machines are frightening. One of them, which assembles the upper part of engines, has a photoelectric cell allowing it to identify the engine type and then choose the appropriate parts among 150 alternatives. (Folha de São Paulo , 29 December 1985, 49) It is not difficult to understand that the ‘‘assembly line which resulted from the most extreme development of the manufacturing process through the introduction of microelectronics (automation), adjusts abruptly to the machinery principle established by Marx’’(Moraes-Neto 1986, 39). What we are now facing is the genuine ‘‘technological application of science.’’ Observing the effect of this radical change on present work patterns, it is easy enough to see the application of all the conclusions 416 MORAES-NETO we have reached to the case of mechanical production processes automated by microelectronics in its most advanced form, the flexible manufacturing system. The management of a quite complex technical system demands professionals equipped with knowledge sufficient both for supervision and maintenance. At the same time and in the same way as in FMS, assembly line automation implies radical elimination of human work directly involved in production. Finally we arrive at the ‘‘unmanned factory,’’ the result of a connection between an automated assembly line and mechanical production processes based on FMS. The effects of this new technical basis on labor directly involved in metal-mechanic production are pointed out clearly by Adam Schaff: ‘‘We are not dealing here with science-fiction, but with reality which can often be more disturbing than fiction. It is a fact, for example, that total automation (found in Japan not as an experiment, but industrially implanted within what are called ‘unmanned factories,’ in which the production process goes on almost without interference) will entirely eliminate human work’’ (1990, 43). Marx versus Smith-Babbage-Braverman on the Division of Labor The ‘‘Theoretical Imprisonment’’ of Braverman It seems obvious that the ‘‘unmanned factory’’ and its workers, equipped with a considerable degree of knowledge, can constitute a refutation of Marx’s theory of labor deskilling accompanying technological development within the capitalist production system. Our aim here is to show why this premise to us appears entirely mistaken. For that, we have to analyze carefully the well-known book by Harry Braverman, Labor and Monopoly Capital , which became an ‘‘official’’ interpretation of Marx’s works. Besides, it has the concept of massive labor deskilling as a pièce de résistance. Let us start with the contradiction that has impelled Braverman: The more I read in the formal and informal literature of occupations, the more I became aware of a contradiction that marks much of the current writing in this area. On the one hand, it is emphasised that modern work, as a result of the scientific-technical revolution and ‘‘automation,’’ requires ever higher levels of education, training, the greater exercise of intelligence and mental effort in general. At the same time, a mounting dissatisfaction with the conditions of industrial and office labor appears to contradict this view. For it is also said*/sometimes even by the same people who at other times support the first view*/that work has become increasingly subdivided into petty operations that fail to sustain the interest or engage the capacities of humans with current levels of education; that these petty operations demand ever less skill and training; and that the modern trend of work by its ‘‘mindlessness’’ and ‘‘bureaucratization’’ is ‘‘alienating’’ ever larger sections of the working population. (1974, 3/4) After this first analysis of the labor division inside the productive process, Braverman develops his argument by initially defining the division of labor as the AUTOMATION AND LABOR 417 theoretical basis and starting point for understanding the capitalist labor process from its very beginning: ‘‘The oldest innovative principle of the capitalist mode of production was the manufacturing labor division, and in one way or another the labor division remains the essential principle of the industrial organisation’’ (68). But the intensification of labor division has damaging effects on the working class: In the progress of the division of labour, the employment of the far greater part of those who live by labour, that is, of the great body of the people, comes to be confined to a few very simple operations, frequently to one or two. But the understandings of the greater part of men are necessarily formed by their ordinary employments. The man whose whole life is spent in performing a few simple operations, of which the effects are perhaps always the same, or very nearly the same, has no occasion to exert his understanding or to exercise his invention in finding out expedients for removing difficulties which never occur. He naturally loses, therefore, the habit of such exertion, and generally becomes as stupid and ignorant as it is possible for a human creature to become. (Smith 1952, 340) We have much studied and perfected, of late, the great civilised invention of the division of labour; only we give it a false name. It is not, truly speaking, the labour that is divided; but the men: divided into mere segments of man*/broken into small fragments and crumbs of life; so that all the little piece of intelligence that is left in a man is not enough to make a pin, or a nail, but exhausts itself in making the point of a pin, or the head of a nail. (J. Ruskin, apud Braverman 1974, 78/9) One reporter cites the belief that ‘‘American industry in some instances may have pushed technology too far by taking the last few bits of skill out of the jobs, and that a point of human resistance has been reached.’’ He quotes a job design consultant at Case Western Reserve University, who said with disarming candor: ‘‘We may have created too many dumb jobs for the number of dumb people to fill them.’’ (Braverman 1974, 35) Something not at all trivial has happened: the first quotation is from Adam Smith, next to whom Braverman fits nicely, which is peculiar for one who produced his work ‘‘under the intellectual influence of Marx’’ (Braverman 1974, 8). It would not seem that Marx could be reduced to Smith. Besides reminding us of Adam Smith’s incurable anguish over the ill effects of labor fragmentation, Braverman, as we have pointed out already, confers permanency on ‘‘Babbage’s principle.’’ ‘‘Babbage’s principle is fundamental to the evolution of the division of labor in capitalist society,’’ argues Braverman (81). ‘‘Applied first to the handicrafts and then to the mechanical crafts, Babbage’s principle eventually becomes the underlying force governing all forms of work in capitalist society, no matter in what setting or at what hierarchical level’’ (81/2). However, ‘‘Babbage’s principle’’ was derived from the specific technical nature of work in the context of the manufactoring system, as it is explained by its author (see note 1). How are we to understand the intriguing contemporaneous aspect (for Braverman) of work in this context? Was the ‘‘Machinery and Modern Industry’’ chapter from Capital a fictional work? How comes it that an author who writes ‘‘under the intellectual influence of Marx’’ should simply ignore that, according to 418 MORAES-NETO Marx, machinery should radically surpass the manufacturing system? For us, the answer is as follows: Braverman is a theoretical captive of Taylorism and Fordism. By considering the Taylorist/Fordist way of organizing production a manufacturing development, the pertinence of one theoretical principle that has emerged from the manufacturing context becomes clear. In order to define more precisely this ‘‘theoretical imprisonment,’’ let us first of all concentrate on the importance given by Braverman to Taylorism/Fordism in his characterization of modern industry, and, therefore, its problems. He writes: ‘‘It is impossible to overestimate the importance of the scientific management movement in the shaping of the modern corporation and indeed all institutions of capitalist society which carry on labor processes’’ (86). Besides considering Taylorism, and its ultimate form, Fordism, a paradigm of modern industry, irrespective of its branch or nature, there is another premise theoretically relevant in Braverman*/that is, Taylorism is typical of capitalism: ‘‘A comprehensive and detailed outline of the principles of Taylorism is essential to our narrative, not because of the things for which it is popularly known*/stopwatch, speed-up, etc.*/but because behind these commonplaces there lies a theory which is nothing less than the explicit verbalization of the capitalist mode of production’’ (86). We would like to emphasize the theoretical significance of this premise because at this very point, we necessarily arrive at the conclusion that, besides characterizing modern industry, Taylorism will never be surpassed inside the capitalist production mode. And this will have serious consequences, as we will see further. It is curious to note that the ‘‘theoretical imprisonment’’ by Taylorism/Fordism leads Braverman to eliminate systematically Marx’s textile machine, and establish a ‘‘bridge’’ straight from Smith/Babbage to modern capitalism, which means toward Taylor: Charles Babbage, who not only wrote penetrating discussions of the organization of the labor process in his day, but applied the same concept to the division of mental labor, and who devised an early calculating ‘‘engine,’’ was probably the most direct forerunner of Taylor, who must have been familiar with Babbage’s work even though he never referred to it. (Braverman 1974, 89) As a demonstration of its power, his ‘‘theoretical imprisonment’’ induces Braverman to illustrate his conclusions with the presumably modern form, par excellence, of industrial production: the Fordist assembly line: Even pick and shovel work takes more learning before it can be done to required standards than many assembly or machine-feeding jobs. ‘‘Studies of final assembly line work in a major automobile company by the Technological Project of Yale University found the average time cycle for jobs to be 3 minutes. As to learning time, a few hours to a week sufficed. Learning time for 65 percent of the work force was less than a month.’’ And yet assembly jobs are the most representative type of operative jobs into which there has been so great an influx in the past three-quarters of a century. (432) AUTOMATION AND LABOR 419 The conclusions reached on the labor operation character of the NC, already discussed here, together with labor deskilling of the assembly line*/unsurpassed under capitalism*/allow Braverman to finish his book with the famous description of massive deskilling, entirely based on Taylorism: The perfect expression of the concept of skill in capitalist society is to be found in the bald and forthright dictums of the early Taylorians, who had discovered the great truth of capitalism that the worker must become the instrument of labor in the hands of the capitalist, but had not yet learned the wisdom of adorning, obfuscating, and confusing this straight-forward necessity in the manner of modern management and sociology. ‘‘What happens to unskilled labor under Scientific Management?’’ ask the Gilbreths in their Primer on this subject. ‘‘Under Scientific Management there is no unskilled labor; or, at least, labor does not remain unskilled. Unskilled labor is taught the best method obtainable . . . No labor is unskilled after it is taught.’’ The instruction of the worker in the simple requirements of capital: here, in the minds of managers, is the secret of the upgrading of skills so celebrated in the annals of modern industrial sociology. The worker may remain a creature without knowledge or capacity, a mere ‘‘hand’’ by which capital does its work, but so long as he or she is adequate to the needs of capital the worker is no longer to be considered or called unskilled. It is this conception that lies behind the shabby nominal sociology in which the sociologists find ‘‘upgrading’’ in the new names given to classifications by the statisticians. ‘‘Training a worker,’’ wrote Frank Gilbreth, ‘‘means merely enabling him to carry out the directions of his work schedule. Once he can do this, his training is over, whatever his age.’’ Is this not a perfect description of the mass of jobs in modern industry, trade, and offices? (446/7) Braverman?s final conclusion is entirely Smithian : task fragmentation is intensifying, and the complex Taylorist control of timing and movement of living work leads to the creation of an immense contingent of unskilled workers*/a ‘‘nation of Helots,’’ to recall the well-known expression of Ferguson (apud Marx 1952, 173). We need to know if the conclusion presented by Braverman is indeed a Marxist one, as has been assumed without any further debate, to the point where Marx himself has been replaced by Braverman, regarding the Marxist view on the subject. The question whether it is possible to reduce Marx to Smith might seem unusual in the context of current literature. Although Braverman in no way states that he considers Marx equal to Adam Smith*/the Smithian aspect of his analysis is the result of our interpretation*/such a comparison is already found in the more recent literature. Let us analyse how Marx was transformed into a minor ‘‘neo-Smithian,’’ to paraphrase Samuelson (7): The industrial revolution over the past three centuries was associated with the evolution of a radically different division of labour, a process charted by a large number of historians, including Adam Smith, Ricardo, Marx, Babbage and, more recently Fröbel, Heinrichs and Kreye. What these authors observed was a steady transition to increasing specialisation. (Kaplinsky 1985a, 13) 420 MORAES-NETO Except Ricardo, whose mention here seems improper, we will see that all Marx’s theoretical work on the development of the labor process under capitalism seems reduced to an observation on ‘‘increasing specialization’’ of the Smith/Babbage style. If the trend continues to grow, we will shortly be able to suggest to publishers, in the name of a presumed theoretical identity between Marx, Smith, and Babbage, that they delete the useless chapter from Capital called ‘‘Machinery and Modern Industry’’! This identity promptly assumed between Smith and Marx has a clear reason: the immense importance of the Fordist assembly line in characterizing modern industry, as we have already said. The question we pose now is as follows: if Taylorism/Fordism were insuperable in terms of capitalism, then would Braverman’s conclusion as to a ‘‘nation of Helots’’ make sense empirically, overshadowing his theoretical inconsistency with Marx (which we intend to show next). From what we have shown so far in item 1, an unmanned factory of the metal-mechanical industry (considering that the metal-mechanical is responsible for the theoretical turmoil) demolishes Braverman?s conclusion. If we take Braverman as Marx’s interpreter, par excellence, on the work content, then once more Marx has made a mistake. To say that Braverman is Smithian is not enough, because for many people it has already been assumed that Smith is equal to Marx. This will require defining the meaning of deskilling in Marx. Automation and Work in Marx: The Meaning of Deskilling and the Challenge of Our Times The quotations presented above (under ‘‘Automation and Deskilling’’) allow us to understand clearly Marx’s ideas on the effects of automation on work involved directly in production and, consequently, the meaning given by Marx to the process of deskilling as it concerns such work. The analysis starts with the transformation of the labor process into a ‘‘technological application of science.’’ This resulted in a profound movement in the direction of the deskilling process. Work deskilling is defined not only as devoid of content (which had already occurred at several stages of manufacture), but results in human labor becoming an appendage of the machine. In other words, the machine now ‘‘decides’’ on quality and frequency, leaving to the worker the tasks of supervision and, sometimes, intervention. As a consequence of the objectification of the labor process, Marx observes that ‘‘the worker appears as superfluous’’ (1973, 695). The idea of superfluousness of living work defines better than anything else the real meaning given to the deskilling process by Marx. It is not a self-contained process, which would perpetuate the deskilling of work*/that is Adam Smith’s idea*/but it is a radical move toward the negation of living work inside the production process itself, so that the ‘‘unmanned factory’’ represents the most advanced stage of the work deskilling process from Marx’s point of view. That an ‘‘unmanned factory’’ (of any kind) functions with a small number of highly qualified workers does not signify negation of Marx’s theoretical proposition, but rather, its consequence. AUTOMATION AND LABOR 421 How can we corroborate the interpretation given to the deskilling process in Marx? This is done by considering capital as a contradiction (which is basic to understanding his thought as a whole), as it is explained in a well-known passage in the Grundrisse : Capital itself is the moving contradiction, [in] that it presses to reduce labour time to a minimum, while it posits labour time, on the other side, as sole measure and source of wealth. Hence it diminishes labour time in the necessary form so as to increase it in the superfluous form; hence posits the superfluous in growing measure as a condition*/question of life or death */for the necessary. On the one side, then, it calls to life all the powers of science and of nature, as of social combination and of social intercourse, in order to make the creation of wealth independent [relatively] of the labour time employed on it. On the other side, it wants to use labour time as the measuring rod for the giant social forces thereby created, and to confine them within the limits required to maintain the already created value as value. Forces of production and social relations*/two different sides of the development of the social individual*/appear to capital as mere means, and are merely means for it to produce on its limited foundation. In fact, however, they are the material conditions to blow this foundation sky-high. (1973, 706) The contradictory aspect of capital is presented by Marx beginning from the negation of living work as a basic unit of the production process and its replacement by dead work. Considering that, how should this contradiction be explained when the technical basis is of a Fordist/Taylorist nature? The answer is that, in fact, no contradiction exists. The Taylorist/Fordist way of organizing the labor process is in accordance with capital as a social relation. Taylorism/Fordism, a technical form based on human work which leads to the employment of thousands of unskilled workers, is perfectly adjusted to function as a support for the capitalist social form. The capitalist dream of self-perpetuation has found a suitable technical basis. As we have said already, the application of microelectronics to the metalmechanical industry ‘‘will bring, as a consequence, this industry along the path of automation, as happened a long time ago in some technologically more advanced sectors of industry’’ (Moraes-Neto 1986, 39). Capitalist competition on an international scale and the possibilities opened by scientific achievements are changing the significant ‘‘Smithian/Bravermanian’’ fraction of the productive structure of capitalism toward a common path of automation, which is adjusted to Marx’s theory. By a curious irony of history in a critical moment in the socialist experience, nothing seems more contemporary to modern capitalist societies than Marx’s words: As soon as labour in the direct form has ceased to be the great well-spring of wealth, labour time ceases and must cease to be its measure, and hence exchange value [must cease to be the measure] of use value. The surplus labour of the mass has ceased to be the condition for the development of general wealth, just as the non-labour of the few, for the development of the general powers of the human head. (1973, p. 705) It is difficult to find a bigger challenge than this: how to manage politically, within the very shadow of the ruins of socialism, the increasing contradiction in 422 MORAES-NETO contemporary capitalism between productive forces and relations of production? From a Marxist perspective, this is the true question of our times. Acknowledgements Translated with the support of the FUNDUNESP*/Foundation for the Development of São Paulo State University. References Babbage, C. 1971. On the economy of machinery and manufactures. New York: Augustus M. Kelley. Braverman, H. 1974. Labor and monopoly capital . New York: Monthly Review Press. Davis, L., and J. C. Taylor. 1973. Design of jobs . Middlesex: Penguin Books. Dina, A. 1987. A fábrica automática e a organização do trabalho . Petrópolis: Vozes. Fleury, A. C. 1988. Automação, organização do trabalho e produtividade. Engenharia de Produção, no.1 (January): 17/20. Ford, H. 1926. Minha vida e minha obra . Rio de Janeiro-São Paulo: Companhia Editora Nacional. Kaplinsky, R. 1985a. Microelectronics and technical change revisited . IDS, University of Sussex. ///. 1985b. Electronic-based automation technologies and the onset of systemofacture: Implications for Third-World industrialisation. World Development 13 (3): 423/39. Leite, E. 1986. Novas tecnologias, emprego e qualificação na indústria mecânica. Revista de Administração 21 (2): 31/9. Marglin, S. 1974. Origens e funções do parcelamento de tarefas. In A. Gorz et al., Divisão do trabalho, tecnologia e modo de produção capitalista . Porto: Publicações Escorpião. Marx, K. 1952. Capital. Great Books of the Western World, vol. 50. Chicago: Encyclopaedia Brittanica. ///. 1973. Grundrisse: Foundations of the critique of political economy (rough draft) . London: Penguin Books. Moraes-Neto, B. 1986. Automação de base microeletrônica e organização do trabalho na indústria metal-mecânica. Revista de Administração de Empresas 26 (4): 35/40. ///. 1989. Marx, Taylor, Ford: as forças produtivas em discussão . São Paulo: Editora Brasiliense. Salerno, M. S. 1988. Automação e lutas dos trabalhadores. São Paulo em perspectiva 2 (3): 62/74. Schaff, A. 1990. A sociedade informática. São Paulo: UNESP/ Brasiliense. Simmon, M. 1986. Princı́pios de sistemas flexı́veis. Máquinas e metais, January/ February, 42/8. Smith, A. 1952. An inquiry into the nature and causes of the wealth of nations, Great Books of the Western World, vol. 39. Chicago: Encyclopaedia Britannica. Tauile, J. R. 1983a. Máquinas-ferramenta com controle numérico e seus efeitos sobre a organização da produção: o caso brasileiro . Universidade Federal do Rio de Janeiro, Instituto de Economia Industrial. ///. 1983b. Microelectronics, automation and economic development. Ph.D. diss., New School for Social Research, New York.