Introduction

According to Thomas Princen, a principal reason for the unsustainability of modern societies is their reliance on efficiency as at least a central member of the set of principles according to which such societies organise their use of resources, both natural and human. (Princen 2005, p.49) By a principle of organisation for the use of resources Princen means a society-wide behavioural regularity or pattern in the way decisions are made to exploit resources, to deploy technologies, to produce items for sale, to employ people and to organise the processes of production and consumption.¹ In modern industrial societies, these decisions are made according to whether the action under consideration yields more output per unit of input than previously.

For example,

(i)f cutting trees in 1,000-acre blocks is efficient—that is, if more trees can be cut per day or per worker or per unit of invested capital—such clear-cutting supersedes selective logging. If lopping off the tops of mountains is the cheapest means of getting out coal, deep-shaft mining is abandoned. If mile-long drift nets catch more fish per vessel per voyage, smaller nets and long lines are jettisoned. If a construction company can buy brand-new wooden concrete forms cheaper than it can clean and reuse its old forms, it dumps the new forms after a single use in the nearest landfill. (Princen 2005, pp.49-50)

From these examples it is clear that what Princen means by a principle for the organisation of the use of resources is not necessarily something consciously appealed to by the relevant decision-makers.² In other words, the deliberation foresters engage in when deciding whether to replace selective logging by clear-cutting can still be subject to, or governed by, the principle of efficiency even when they do not explicitly appeal to efficiency in their reasoning. Foresters can be still operating according to a principle of efficiency even if they simply think, “Clear-cutting yields more trees cut per day and/or per worker and/or per unit of invested capital than selective logging. Therefore, we will replace selective logging by clear cutting.” They do not have to think, “Clear-cutting yields more trees cut per day and/or per worker and/or per unit of invested capital than selective logging. So clear-cutting is more efficient than, hence preferable to selective logging. Therefore, we will replace selective logging by clear cutting.”

This point is not quite as pedantic as it seems. Princen justifies his claim that modern societies are governed by efficiency as at least a core component of the set of principles according to which they use resources by appeal to the hegemony of the discourse of efficiency in specific phases of modern history. In particular, he appeals to the cult of efficiency which sprang up around the so-called American Progressivist movement from the 1890’s to the 1920’s.³ This creates a problem for him. Clearly, one cannot justify the claim that all modern industrial society, whether capitalist or socialist, whether late 18th, 19th or 20th century, is governed by a principle of efficiency simply by pointing to the fact that at a certain period in a certain country in these 250 years of history there has been, at the level of public discourse, a craze about efficiency. Why, after all, should we think that a fad in public discourse should ever translate itself, even during its heyday, into practical reality, as a principle which governs all or even most social decision-making about the use of resources?

In fact, there is something deeply right, both in Princen’s claim that all modern industrial society is governed by a principle of efficiency and in his claim that this is a principal cause of the unsustainability of such society. But it takes some reconstruction in order to make this clear. And only when this has been made clear can we see the truly deep reason why Princen is right in his intuition that in American Progressivism this principle comes, as Hegel might say, to consciousness of itself. We must therefore try to reconstruct and elaborate Princen’s position through an examination of the ideology of American Progressivism.

Taylor and Taylorism

Emblematic of American progressivism are the views of one man, Frederick Winslow Taylor (1856-1915).⁴ It is, by now, a familiar observation that the culture of work required for operation of modern industry was not something people slipped into naturally; they had to be disciplined into it. Thus, we are all familiar with the stories of the problems early industrial capitalists had in keeping their workforce on the job, stamping out the tendency of pre-modern peasants and farm labourers to do only as much as they wanted to do and then simply to down tools and go home. What is less well-known is that much more than acceptance of a fixed working day is required to create the culture of work we have today. This is shown by the problem Taylor saw himself confronted with at Midvale Steel Company in Philadelphia, where he worked as a foreman supervising the production of forge steel locomotive axles and wheels in the early 1880’s. These axles and wheels

… were cut to size with such tools as lathes and boring mills. Those great belt-driven machines could cut metal only so fast without burning themselves up or prematurely dulling their tool bits. But how fast was that? How should the machines be set?

Taylor [who wanted to get more out of those under him] would have liked to tell his men: “You’re running it was too slow. With this kind of steel and that depth of cut, and that feed, you should be able to go twenty feet per minute, easy. …” But he couldn’t say that, because he didn’t know for sure. The grizzled machinists under him would say, “No, that’s too fast. It’ll burn up the tool. It’ll wreck the machine.” But they didn’t know either, not really; much of their wisdom was guesswork, built up over the years into serviceable rules of thumb. Besides, what was the big hurry? For them informed guesses were good enough. (Kanigel 1997, p.18)

But guesswork and semi-intuitive craft knowledge—Aristotle would speak here of techné—were not good enough for Taylor. He began experimenting to find out how fast tools could cut and men could work. “(H)e offered workers double pay, warned them their windfall would be over at the first sign of slacking off, … and calculated the work they did per day in foot pounds and its rate in horsepower.” (Kanigel 1997, p.19) At first, Taylor got no data consistent enough to use in determining lawful wage/performance levels. But in 1883, he had a moment of epiphany. Until then, he had taken whole complex tasks as what he had timed in the effort to determine how much harder men would work as pay levels rose. But a stop watch had been used to time the tasks given to his men and a stopwatch could as easily time “what a man did in ten seconds as in ten hours.” (Kanigel 1997, p.19) And so, as he later wrote, it occurred to Taylor that ““it was simpler to time each of the elements” [Taylor’s word!] of a job and establish an overall time for it by adding up those times than it would be to search through records of past jobs and guess how they might apply to a future one.” (Kanigel 1997, p.19)

At first, when timing someone at work, their performance seemed to be “no more than a smear of featureless movements.” (Kanigel 1997, p.19) But Taylor found that in reality this was not so. One could analyse work into its component movements and it became “as if you were seeing it in slow motion, each element of it highlighted, dissected, splayed open for view.” (Kanigel 1997, p.19) And now one could determine, in a consistent, repeatable manner, what got in the worker’s way, how a machine could be modified to facilitate the operation of it, what elements of a process of work could be improved or eliminated, and so on. An early target for Taylor’s analysis was the production of steel locomotive wheels. Previously, one worker had produced each wheel, earning a fixed amount for each wheel they completed.

But once Taylor was through, their job was no longer to machine a [whole wheel] but was a succession of smaller tasks: Set [wheel] on machine ready to turn. Rough-face front edge. Finish-face front edge. Rough-bore front. And so on, each step minutely described and timed to the tenth of a minute. (Kanigel 1997, pp.19-20)

And now a crucial re-organisation of the whole structure of production became possible. Whereas previously one first class machinist had been employed to produce the one wheel, from start to finish, now ““men … trained up from a lower class of work,” typically laborers or machinists’ helpers could be employed to perform the individual smaller tasks, and indeed to do so in accordance with management’s instructions as to optimal speed, feed, tool shape, the kind of metal being cut, etc. Ultimately, almost all such variables “could be reduced to a mathematical problem solvable on a special slide rule, in twenty or thirty seconds.” And the slide rule was wielded by management. In other words, Taylor’s innovations in the organization of work meant that the thinking, problem-solving aspects migrated up the chain of command to management while the workers themselves were left to perform tasks in relatively unthinking obedience to management. This was, of course, Taylor’s famous, not to say infamous doctrine that in well-organised, efficient workplace there is a radical separation between mental and physical labour, with the former falling to management. And the successful implementation of this doctrine meant that

(t)raditional craft know-how [i.e., techné], reduced to scientific data, was passing from the workman to manager, from the shop floor to the front office. On the big boring mills at Midvale, Taylor was creating a new world of work. (Kanigel 1997, p.18)⁵

Of course, if through analysis of a complex task into its various component movements, one could now meaningfully determine such things as what got in the worker’s way, how a machine could be modified to facilitate the operation of it, what elements of a process of work could be improved or eliminated, etc., then it became possible for the first time to provide, in consistent, repeatable fashion, quantified accounts of the efficiency of the productive process. After all, an efficiency measure is simply the ratio of certain inputs to certain outputs. The higher the latter in relation to the former, the more efficient the process with regard to how it uses its inputs to generate its outputs. And so it now became possible to optimise the design of the task in the way an engineer might optimise the design of a machine, e.g., a car engine. This is nicely illustrated by one of Taylor’s many disciples, William H. Leffingwell, who

applied scientific management to the office, experimenting with typists and clerks the way Taylor had with machinists and shovelers. There was one best [that is, one optimal] way, he held, to insert paper into a typewriter, pin pages together or sit at one’s desk—“well back in the chairs, with the feet place squarely on the floor and head and shoulders erect.” Leffingwell saw malingerers behind every file cabinet, just as Taylor did behind every lathe. He and his followers tracked how many minutes a day typists types, fixed work standards of so many square inches of typed work per hour and awarded bonuses to those satisfying them. (Kanigel 1997, p.21)

Applications of Taylorism such as these can mislead one into a superficial objection. Surely, one will say, this example shows how inhuman Taylorism is: it reduces workers to mere cogs in a machine, demeaning them by getting them to perform mindless tasks over and over again. But this kind of objection⁶ takes a historically contingent feature of Taylorism to be its essence. In so doing, it makes one blind to the ongoing presence of Taylorism in the contemporary world. Such blindness is amply illustrated by trade union leaders such as Laurie Carmichael, who welcomed the introduction of new computer technology into machine shops. This technology certainly meant job losses, as unskilled labourers were replaced by computers and robots. But, he thought, it also restored the trade: now the turner and fitter was a skilled expert in wielding the computer-programmed lathe. Computer-programmed machines and robotics spelt, he thought, the end of Taylorism.

The mistake made by Carmichael and others was to think that according to Taylor the parts or elements into which he analysed more complex tasks and activities just had to be performed by human beings. In fact, Taylor quite rightly saw his system as implying the opposite. He saw it as a first step towards realising an enormous potential for automation—in his eyes, of course, not so much for the sake of liberating human beings from drudgery but for the sake of massively increasing productivity, although he certainly appreciated that it would release workers from drudgery. In other words, Taylor’s position was much closer to that of those contemporaries of Carmichael who, unlike Carmichael himself, sought the deployment of computer technology not because it restored the dignity of work, but because it made them more competitive. For the essence of Taylorism lies in the idea that work processes have parts or elements such that the whole they constitute can be reliably optimised for efficiency. This is the central idea, one exploited precisely in the design of computer-steered machine tools, car-making robots and the like, hence an idea which is still very topical today. And in order to see, not so much what is wrong in this idea but what is dangerous about it, one must proceed from a serious investigation of the ontological constitution and structure of production. To this I now turn. As we shall see, this will allow us to answer to the central question of this paper, namely, whether and in what way Princen is right in his claim that modern industrial society is dominated by a principle of efficiency to the detriment of the environment.

Towards an Ontology of Production

As a first step towards an ontology of production we need to distinguish between what I shall call a productive unit as such and a productive unit as deployed. Alternatively put, we need to distinguish between a productive unit and a particular deployment of this unit. The notion of a productive unit as such is quite intuitive: it is any set of human beings spatiotemporally organised around realising a certain kind of goal, together with the tools, machines and material inputs they need in order to engage in the activity of realising this goal. One example of a productive unit would be the press shop in a car-making factory. This is obviously a group or set of human beings, organised around stamping car panels and frames out of sheets of metal, and supplied with the tools, machines and sheets of metal they need in order to stamp car panels and frames out of sheets of metal. A second example would a team of loggers employed by a timber company, together with their chain-saws, bulldozers and fire pellets. This is obviously also a group of human beings organised around a certain kind of goal, in this case, supplying timber mills with lumber, and supplied with the tools, machines and trees it needs in order to supply timber mills with lumber. A final example would be a farm: this is also a group or set of human beings organised around a certain kind of goal, namely, producing food or, more specifically, say, growing wheat and it is supplied with the tools, machines and land it needs for producing food.

The notion of a productive unit as deployed is less intuitive but only marginally so: it is a productive unit as concretely or specifically arranged or set up spatiotemporally in the service of a goal around which it is organised. By the deployment of a productive unit I thus mean the specifc way in which the productive unit is spatially and temporally set up to accomplish a goal around which it is organised, for example, the location of the presses and their operators in the press shop, the sequence of activities and operations which together constitute the process of turning a sheet of metal into a car panel, and thus all the various actions and component actions of the operators implicated in this sequence. The productive unit as deployed is, one might say, a specific instantiation of the productive unit. In the case of that productive unit which is the team of loggers along with their saws, bulldozers, fire pellets and trees, the operators and their tools and machines come to the inputs rather than the inputs coming to them. So the deployment of this productive unit is the specific sequence of activities and operations within the forest which together constitute the specific process of turning a number of trees into lumber for a timber mill, a sequence which of course implicates a whole series of actions and purposive behaviours on the part of the loggers, e.g., filling the chain-saw up with petrol, starting up the bulldozer, chopping trees down, loading up the timber trucks., etc.

Consequently, the same productive unit can have different deployments. In other words, the same productive unit as such can be different productive units as deployed. The press shop in the car factory, the logging team in the forest and the farm could be organised spatiotemporally in all sorts of different ways: the various pieces of equipment they use could be arranged in different ways, indeed, each piece itself could have a different spatiotemporal form and other design qualities. The whole productive unit could even have, across different deployments, different defining goals. Indeed, from deployment to deployment an individual productive unit could be re-organised around a new defining goal. For example, the press shop could be, and in fact in real-life car factories often is, reorganised around producing panels and frames for a different model of car. And it could in principle be reorganised around an even more different overall goal, e.g., producing panels and frames for different kinds of vehicle, e.g., buses or railway carriages. Similarly, the logging team could be deployed, or indeed redeployed from clear-cutting, for selective logging or even building forest roads. Finally, the farm could be organised or indeed re-organised around producing barley rather than wheat. In each case, of course, since the overall goal around which the productive unit is organised is different, so, too, is the concrete spatiotemporal deployment of the unit. We thus have different productive units as deployed. For in general the difference in overall defining goal determines differences in tools and machines, differences in the way tools and machines are set up, differences in inputs, differences in the way in which tools and machines are applied to these inputs, and indeed sometimes differences in human beings: workers are put on or off, as need requires.

Now it is clear that the Taylorist analysis of work into its component elements applies to productive units as deployed. That is, it applies to particular deployments of productive units. What the Taylorist efficiency expert works on when conducting such an analysis is the press shop-as-spatiotemporally-deployed-in-the-service-of-stamping-metal-sheets-into-such-and-such-kinds-of-car-panel and frame, or again, the team of loggers with their equipment-as-organised-around-clear-cutting. The efficiency expert does not work on the press shop or team of loggers as such since the press shop or team of loggers can, as we have just seen, be deployed in all sorts of ways to do all sorts of different things and for each of these different things there will be a different set of component behavioural movements to be identified, measured with a stop watch and then made more efficient. Rather, he works upon a particular deployment of the press shop or of the team of loggers with their equipment, organised around some specific overall purpose or task. Notice now that a productive unit as deployed embodies or instantiates a function or mapping from inputs to outputs. It is a spatial and in particular temporally extended entity to whose very identity certain inputs of such and such kinds, a certain output of such and such a kind and finally a certain ratio between these two is inherent.⁷ This is precisely why only the productive unit as deployed, and only this, can be a target of analysis and investigation for the Taylorist efficiency expert. This latter seeks to identify, and ceteris paribus improve on, the actual efficiency of something from the outset understood as embodying a mapping from certain quantities of input to certain quantities of output.

The importance of the distinction between the productive unit as such and that which constitutes the true target for the efficiency expert, namely, the productive unit as deployed, can be brought by considering an example derived from one used by Princen.⁸ Imagine two wheat farmers A and B living alongside one another. Farmer A gets more tonnes of wheat per hectare than B while farmer B spends less on machines per hectare. Clearly, each farmer can claim that his farm is more efficient in some respect: A’s with regard to maximising output, B’s with regard to minimising capital cost. Which farmer is right to seek which kind of efficiency is evidently a function of which farmer has got the priorities right: should one in the circumstances seek to maximise output or should one seek to minimise capital costs? The fact, however, that there is this issue of priorities at all shows that neither farm is simply a spatiotemporal deployment of humans, tools, machines and inputs. Each farm is more than a mere deployment, a mere productive unit as deployed, precisely because, over and above its status as such a deployment, it embodies a decision as to what the right priority is. On A’s farm, the priority is that of maximising output and this priority determines a particular deployment of individuals, tools, machines and inputs. On B’s farm, the priority is different, namely, that of minimising capital costs, and so B’s farm constitutes a different deployment of individuals, tools, machines and inputs.

Now there is a good reason why a farm is always more than the particular deployment of individuals, tools, machines and inputs it currently implements, a deployment governed by a particular prioritisation and thus subject to a particular notion of efficiency. For in order for there to be any rational expectation of longevity, beyond a particular deployment of it, a productive unit like a farm must include, as part of its identity, the capacity to redeploy in the light of changed circumstances. Precisely for this reason, the character of each farm as having some prioritisation which is reflected in its deployment is essential to its very identity. Each farm is thus ontologically constituted in part by a capacity to respond more or less wisely to its environment in order to ensure its survival beyond the point at which its current deployment becomes dysfunctional. Perhaps farmer B used to seek maximisation of output just like his neighbour A. But then he realised, well ahead of A, that the price of agricultural machinery would go through the roof. And so B decided that in order to keep the farm going, he needed to re-deploy toward cost minimisation.

This brings out the important ontological point to be made about productive units such as a farm: it is underdescribed as something which exists for a purpose, say, producing food. It is more fully and accurately described as something which maintains or sustains itself in its social and biophysical environment as something which is for producing food. The farmer confronts questions of prioritisation of the kind illustrated precisely because the job of the farmer is not just to produce food, but to maintain or sustain the farm as something which is for producing food.

But what capacity must a farmer possess in order to maintain the farm as farm, in other words, in order not just to manage one particular deployment efficiently but to redeploy in the light of changed circumstances? It is not a matter of craft knowledge or skill, of techné, since this pertains to the running of the productive unit as deployed, that is, the particular deployment of human, technological and natural resources. What, then, is it? A clue is provided by the fact that this capacity must always involve the capacity to deliberate and decide ethically. Farms and indeed all productive units exist within a web not just of biophysical but also of social relations, which are subject to norms and values.⁹ Consequently, the kind of practical deliberation at issue here is, at least when it is most adequately realised,¹⁰ what Aristotle called phrónesis, i.e., practical wisdom. For by practical wisdom Aristotle meant precisely the kind of deliberation through which human beings maintain themselves and the things important to them in a social, hence ethically structured environment.

Taylorism, Efficiency and Sustainability

At this point, the real critique of Taylorism, the critique of its essence, can begin. Taylor was concerned to improve the efficiency—today we call it productivity—of those kinds of productive unit which exist only as proper parts of a larger productive unit responsible for deliberating and deciding how best to prioritise so as to maintain itself as the productive unit it is in its wider biophysical and social environment. That is, he was out to improve the efficiency of such things as the press shop in the car making factory. And the press shop does not determine, either for itself or for the larger factory of which it is part, what the appropriate deployment and in particular re-deployment of resources should be, given changed circumstances. This task falls to the whole factory, or rather, the management of the whole factory.

Now in his efforts to improve the efficiency of such units as the press shop, Taylor accomplished something remarkable: he and his theory of ‘scientific’ management created a deployment of the press shop which integrates the latter completely into the larger productive unit, ultimately, of course, the factory and firm. For his re-organisation of the workplace meant, as we have seen, that even the craft knowledge of the machinists, i.e., their techné, that which defined their workshop as pre-Taylorist, was taken from them and handed over to management. This crucial tendency to centralisation of ‘mental’ labour—a tendency massively displayed, of course, by contemporary managerialism—is a defining feature of the kind of organisational reform Taylor recommended. For this reason it is quite appropriate to describe contemporary managerialism as neo-Taylorism. And this centralisation represents a loss of power on the part of those on the shop floor to resist and thwart demands coming from higher up the chain of management. Those working in productive units of the kind illustrated by the press shop no longer had to have the specialist craft skills and knowledge which gave them and their productive unit the capacity to resist and thwart demands from above.

Now certainly Taylorist re-organisation of the workplace meant, as a matter of historical fact, the introduction of a style of work whose stultifying effects upon the worker had already been criticised by Adam Smith in his famous account of the pin factory. But such dehumanisation is, to repeat, a contingent, historical feature of Taylorism, not its essence. This essence consists in the idea that productive units as deployed can be analysed into their component actions and indeed even bodily movements, on which basis then their actually efficiency can be calculated and measures implemented in order to make the deployment more efficient. But the successful implementation of such measures has a distinctive effect. By making the efficiency of each productive unit, for example, the press shop, within a larger productive unit, in this case, the car-making factory,¹¹ consistently and reliably determinable, one makes the efficiency of the whole factory thus determinable. The whole factory can now be optimised for efficiency in its deployment of human, technological and natural resources. Thereby it gains, at least as long as current circumstances remain as they are, a competitive edge.

Innovations, however, which give a competitive edge tend to be adopted eventually by the competition. So Taylorist organisation of the work place will tend to become the norm across a given industry or sector. And it will inspire adaptation and appropriation of Taylorist reforms by productive units in other industries and sectors. Taylorist analysis and revision of work practices, insofar as it genuinely delivers on the claim to efficiency reliably determinable, hence reliably improvable, thus encourages a tendency to homogenisation and standardisation of work practices around a presumed optimum deployment within a given industry or sector. Precisely for this reason, contemporary managerialism, which is, as just intimated, merely a new phase or form of Taylorism driven by such things as intensity of competition coming from globalisation, ongoing cuts to government funding, or even government regulation to enhance competition, is riven with rhetoric about best practice, industry standards and the like.

Note now an implication of this elevation, across an industry or sector, of a certain kind of deployment to the status of a norm which participants in the industry or sector ignore at their peril: it creates, across an industry or sector, an inherent tendency or pressure to conform to a certain kind of deployment and the prioritisation underpinning it. Productive units such as factories and farms thus find themselves constrained in their capacity for maintaining themselves in the face of changed circumstances through re-deployment of their human, technological and natural resources. They find themselves having to behave more like the productive units under them, rather than as that level of productive unit which responds to the environment in which they and the productive units under them occur. The car factory as a whole behaves more like the press shop within it, the timber company more like the logging team, the cattle station more like the feed lot. For they are now all subject to an industry- or sector wide understanding of how resources are to be deployed to which it is rational to conform since this is what the others do; not to do so would put one at a disadvantage. Crucially, this tendency to conform can persist even in the face of changes to the wider environment, changes which make the standard way of deploying resources dysfunctional in the longer term. And so there is an inbuilt resistance to respond to these changed circumstances, even on the part of productive units which are aware of the fact that circumstances have changed and that re-deployment away from the industry- or sector standard is rationally required in the longer term.

One might put the matter this way: because it tends towards forcing productive units like the car factory, timber company and cattle ranch to act more like entities whose interactions with the environment are mediated by a larger system containing them, the Taylorisation of work practices creates a situation in which, at the level of interaction directly with an environment rather than a larger, encompassing system, the capacity to act in the light of what is rational longer term becomes circumscribed. Note that because at this level there is, by definition, no higher level to which the exercise of this capacity could be delegated, this situation is one of ontological ambiguity: the productive unit at this level exists in an ontological no man’s land, half sub-system contained within a larger one, half larger system itself, adapting to changed circumstances in an environment rather than responding to demands from above. And this means that the capacity to act on what phrónetic practical deliberation tells a productive unit at this level about how it, individually, should re-prioritise and re-deploy is circumscribed. It is as if the productive unit had lost a little of its practical wisdom, as opposed, of course, to its mere technical expertise. More simply put, the Taylorisation of work practices creates a situation in which one finds it difficult to act on changed circumstances, hence meaningfully to plan for how things must be given changed circumstances. One is forced constantly to respond to how things are now rather than to anticipate how they will be. This ontological ambiguity explains, of course, the sense we all have that the contemporary economic system is out of our control. And the Taylorisation of work practices is both a response to and a necessary component of the cause of this ontological ambiguity.

In short, this ontological ambiguity is what we are sensing when we feel that, for good or ill, we must work within “the system”. One may call this condition one of complexity if one likes but I now think the terms ‘complexity’ and ‘complex system’ are misleading. Talk of complexity obscures the fact that the complexity at issue is a human creation, one engendered in part precisely by the re-organisation of work practices which Taylorism represents. And Taylorism is something it is in principle possible to resist politically, both from the perspective of opposing something which contributes to systemic unsustainability and from the perspective of opposing something which contributes to a loss of autonomy, hence a worsening of our lives as practically deliberating, self-conscious selves.

At this point, I think, we can claim to have explained in what sense Princen is right to maintain that efficiency is a principle dominant in contemporary industrial society which accounts for much of its unsustainability. The principle of efficiency is a dominant, hegemonic one in the sense that we have developed an economic system in which acting in the short term, in response to merely to immediate threats, has come into conflict with, and displaced, acting from the perspective of the longer view. We have also explained why Princen is right to identify the historical phenomenon of Taylorism as attesting to the dominance of this principle in this sense. For Taylorist reform of work practices has played a major role in the development of such an economic system. But one thing remains unexplained: why is this principle one which, as Princen maintains, is dominant in all industrial society, even pre-Taylorist industrial society? Thus far, I have relied upon the idea that competitive pressure induces the spread of Taylorist reform, as competitors adopt these reforms in order to neutralise the competitive advantage which it gives. But this is a comparatively superficial explanation of why Taylorism could spread across whole industries and sectors. One can always ask where the pressure to compete comes from, indeed, to such a degree that one feels compelled to adopt such drastic reforms of traditional work practices.

Here the explanation must surely be that the kind of economic system which modern industrial society represents, which indeed made such society possible, has an inbuilt tendency to generate the competitive pressure to which Taylorism is both a response and an accelerant. Industrial capitalism is not just a system in which goods and services are produced for exchange, in which the use value of goods and services for their producer is their exchange value. This fails to capture the distinctive character which the profit derived from exchange possesses in a distinctively capitalist economy. The capitalist economy is, after all, distinctively capitalist, i.e., capital oriented: profit has the significance not, or not just, of being a means of acquiring either the necessities of life or indeed its luxuries. Nor does profit have simply the significance of being a means of acquiring power and influence. In addition, it has the significance of being a means of acquiring productive capacity, that is, of being capital. And this significance of profit, which is only truly or fully realisable possible given certain economic institutions, for example, stock markets, appropriate systems of finance and credit, such structures of governance as the limited liability corporation, etc., means that it becomes very easy to re-deploy, indeed to qualitatively transform, via the universal medium of money, production in response to opportunities and obstacles. This is an economy in which the very character of production—how things are produced and in particular what is produced—becomes massively flexible and adaptable as a result of extremely efficient, ubiquitous procedures for investment and disinvestment. Under this regime, then, a premium will be placed on keeping productive activities competitive in the distinctively capitalist sense, namely, generative of capital, i.e., of productive capacity itself. And so there will be an inbuilt, ubiquitous pressure to rationalise production in the manner epitomised by Taylorism. Taylorism thus emerges as a logical phase in the emergence of distinctively capitalist production. It is programmed into the DNA of the industrial society which precedes it.

Notes