Jump to ContentJump to Main Navigation
Dominating Knowledge$

Frédérique Apffel Marglin and Stephen A. Marglin

Print publication date: 1990

Print ISBN-13: 9780198286943

Published to Oxford Scholarship Online: October 2011

DOI: 10.1093/acprof:oso/9780198286943.001.0001

Losing Touch: The Cultural Conditions of Worker Accommodation and Resistance

(p.217) 7 Losing Touch: The Cultural Conditions of Worker Accommodation and Resistance
Dominating Knowledge

Stephen A. Marglin

Oxford University Press

Abstract and Keywords

This chapter examines the influence of the Western intellectual model on the cultural conditions of worker accommodation and resistance. It argues that the transformation of work into its modern Western form was the result of deliberate attempts to take control of process and product from the worker. It suggests that the factory does not impress contemporary Westerners as a curtailment of freedom since it is because people have grown used to this particular form of repression.

Keywords:   worker accommodation, worker resistance, Western model, intellectual model, factory, repression

1 Introduction

This chapter continues a series of enquiries (Marglin 1974, 1979, 1984) which have focused on the role of profitability and capitalist class interest, as distinct from efficiency,1 in shaping the organization of work. At issue here is the workers' side of the story, specifically the cultural underpinnings of resistance and accommodation to the capitalists' project of domination.

The argument is built up from several separate propositions. The first is that cultural variables are central to the outcome of conflict over the organization of work. What people value, how they know—both systems of values and systems of knowledge—affect whether, how, and with what degree of commitment people will defend themselves and their work against the capitalist (or the commissar, for that matter). Indeed, this chapter, although it assumes the importance of class struggle in determining the organization of work, asserts equally strongly that the parameters of class struggle are determined by cultural values. Even class (p.218) itself is not born of economics alone, but of the union of economic interest and cultural justification. Classes act in history only when they are armed and legitimized by cultural values which are generally and widely held throughout the society. And culture not only empowers, it sets limits to class conflict. People struggle only to the extent that their common cultural heritage permits of different interpretations.

The first concrete application of this idea is that the tenacity with which workers defend their work arrangements depends on the meaning that they attach to their work. This meaning can take one of two forms, holistic or individualistic. Holistic meaning attaches significance to one's work because it is an integral part of a whole which commands the allegiance and assent of the community; individualistic meaning, by contrast, is a significance that one creates oneself through work.

The preconditions of holistic and individualistic meaning differ. Holistic meaning requires that work be embedded in the cultural fabric, that it be an expression of one's relationship to the cosmos rather than simply a matter of earning one's daily bread. Individualistic meaning requires that one be in control of process and product, without which the very project of creating meaning becomes unthinkable.

The problem for workers has been twofold, for Western culture fosters neither holistic nor individualistic meaning for most forms of work. First, the Judaeo-Christian and the Greek traditions disembed most work from a context that might make it meaningful. Second, starting with Plato and Aristotle, one system of knowledge has become hegemonic. The problem is that workers' knowledge is generally organized in terms of other, ‘inferior’, systems, with the result that it becomes the inferior knowledge of inferior people. The culture thus undermines workers' attempts to defend their control of work.

These arguments are developed in the sections that follow. After a sketch of the relevant background (Sections 2–4), we shall turn to the main ideas, the disembeddedness of work (Section 5) and the devaluation of workers' knowledge systems (Sections 6–10) in the West. Sections 11–13 highlight these ideas by contrasting the conception of work in Hindu culture as revealed by a case-study of a weaving community.

2 Meaning—holistic and Individual

Work is, and undoubtedly will remain, a core element in establishing a sense of purpose and accomplishment, a sense of significance for one's entire life. Even to speak of the meaning of work is therefore necessarily to speak of the meaning of life, and to oppose the nihilism of Macbeth (‘[Life] … is a tale/Told by an idiot, full of sound and fury,/Signifying nothing’) or the utilitarianism of Freud (‘What decides the purpose of life is the … pleasure principle’ (1961: 23)).

(p.219) Notwithstanding the scepticism, and indeed downright hostility, of the modern, secular, West to any attempt to invest life with meaning—‘human presumptuousness’ Freud called it (1961: 22)—human beings expend considerable energy to disprove Macbeth. Our loves, our friendships, our relations with parents and children, our work, are at least partially a function of a deep-seated drive to make our lives signify something. Certainly Freud's pleasure principle falls well short of explaining attitudes towards work and love, which Freud himself is supposed to have made the measure of successful psychological adjustment.

Social organization may facilitate or hinder the search for meaning. In non-individualistic, ‘holistic’, societies, the meaning of work is a cultural given, but this is the case only exceptionally in individualistic societies. When one is part of a ‘cause’—the last one which commanded broad allegiance in the United States was the cause of defeating Germany and Japan during World War II—it is much like being part of the effort to build a cathedral in the non-individualistic Middle Ages: the particular job can be subordinated to the whole and derive its meaning from the meaning of the whole.

But the general situation of modern individualistic society is that there is no such transcendent cause. Instead, one must be both playwright and leading actor; we each must create our meaning both through the roles we write for ourselves and the way we act them. Doubtless it was much easier when parts in the theatre of life were socially given, and all one had to do was to interpret one's part.

There are thus two kinds of meaning that attach to work, or perhaps it is more accurate to speak of two separate paths through which work feeds into the meaning people attach to human existence. The first is a holistic meaning, with significance and purpose assigned to work by the common consent of the community. The glory of God or the defence of democracy, or for that matter the construction of socialism, all exemplify transcendent causes that have invested work with meaning at some places and times.

The second form of meaning is individualistic, a significance and purpose that the worker herself creates and asserts through her work. ‘Symbolic immortality’, to borrow a term from Robert Jay Lifton (1983: 21), ‘something to point to’ in the words of a steelworker whose story is the preface to Studs Terkel's Working (1972).2 The achievements of work, more than the result of any other activity we undertake, give us grounds (p.220) believing that we might transcend our physical mortality, a belief already implicit in holistic meaning.

Control over process and product becomes central to any discussion of the meaning of work once we recognize the absence of a holistic meaning in contemporary Western society. For control is a necessary, if hardly sufficient, condition for investing one's work with individualistic meaning. But individual or even collective control by workers clashes with the capitalist project of domination. The central proposition of my first paper on work organization (Marglin 1974) was that two crucial steps in the history of work—the extension of the division of labour at the subproduct level and the concentration of production in factories—were instituted by capitalists in order to enhance their control. Subsequent papers (Marglin 1979, 1984) have been variations on the theme of control.3

3 The Division of Labour and the Rise of the Factory

Let me now briefly recapitulate the role of the division of labour in the perspective of control. My ‘enemy’ is evidently Adam Smith, for whom the efficiency of a highly developed division of labour was the starting-point of economic analysis. Indeed, Smith is commonly given credit for ‘discovering’ the division of labour. There is some justice in this, for certainly nobody before Smith (and, with the exception of Durkheim, nobody since) has made the division of labour so central to his argument. But the credit should be for emphasis, not discovery. The Greeks anticipated Smith in all important respects, down to the size of the market as the main factor limiting the division of labour (Xenophon). Plato went one better than Smith. In the Republic, the putative efficiency of the economic division of labour is the basis of the argument for a political division of labour. The chief drawback of democracy is precisely that everybody does everything, a defect which Plato's own blueprint of a specialized political caste was supposed to remedy.

(p.221) Neither Smith nor the Greeks distinguished between the division of labour at the product level and the division of labour at the subproduct level, the division of labour between workshops and the division of labour within the workshop. Both were seen as reflecting only efficiency considerations, the second being simply an extension of the first as the size of the market becomes large enough to warrant further specialization.

Karl Marx perceived an important difference once the division of labour ruled within the workshop as well as without: the division of labour at the subproduct level, the technical or detailed division of labour in Marx's terminology, makes the worker dependent on the capitalist. By contrast, the division of labour at the product level, the social division of labour, in no way presupposes the capitalist as integrator of the production process.

In the main, however, Marx followed Smith in invoking efficiency to explain the extension of the division of labour from the social to the technical level. The increased dependence of the workman on the boss was simply the unfortunate by-product. In this respect I differ from Marx; I argue that what Marx understood to be a by-product was in fact intentional.

The problem of the capitalist, as seen in my perspective, is to maintain a position for himself in the production process. Sometimes an adequate position is made possible by high capital requirements and indivisibilities, which put the activity beyond the reach of the individual workman or even of an association of workmen. Coal-mines and shipyards and even Arkwright's spinning mill come to mind as examples. But these are, I believe, the exceptions rather than the rule.

In the general case indivisibilities and associated capital requirements are not sufficient to maintain a position for the capitalist in the production process. Here, I have argued (Marglin 1974), the capitalist maintains his role by specializing workers to particular tasks, reserving to himself a critical task, frequently the task of integrating the separate operations into a marketable product. This allows him to stand between the workers and the product market, obliging the workers to sell their labour power rather than the product of their labour. Robert Cookson, a British woollen manufacturer in the period of the industrial revolution, succinctly put the essential point to a parliamentary committee: ‘Suppose a man goes into a room and is confined in a room where there are twelve, thirteen, or fourteen looms, how is that man to be proficient in any part of the business than that?’ (Report from the Select Committee to Consider the State of Woollen Manufacture in England, British Parliamentary Papers (1806), 1, p. 74. Quoted in Morris 1972.) Lest the committee harbour any doubts about his meaning, Cookson answered his own question: ‘People trained up in a manufactory are never likely to set up for themselves.’

(p.222) Thus, division of labour at the subproduct level, or more accurately, specialization at the subproduct level, became an instrument of capitalist control. Even if less efficient than alternative forms of work organization in which the worker exercised more control, a highly developed division of labour would enhance the capitalist's control over the production process and increase his profits.

Specialization ensured the dependence of the worker on the capitalist, but it did not get him to work. The problem of discipline was of course not a new one. Xenophon in the fourth century BC, Columella in the first century AD, Walter of Henley and other medieval writers, addressed the problem of worker discipline, anticipating such thoroughly modern ideas as incentive systems (Columella) and even time and motion studies (Walter). But at the time of the industrial revolution the legal freedom of the workers compounded the problem of discipline. The worker laboured primarily under economic rather than political compulsion, and the market was not much help.

Capitalists consistently complained about the perversity of the labour market. Higher wages, instead of coaxing out more effort, served only to reduce the worker's dependence, to the point that one eighteenth-century observer noted ‘the sentiment universal’ among cotton manufacturers ‘that their best friend is high provisions’ (Arthur Young 1770). A little later, the same observer put the point in somewhat stronger terms: ‘every one but an idiot knows that the lower classes must be kept poor, or they will never be industrious’ (quoted in Thompson 1963: 358).

The putting-out system made matters worse: in his own cottage the worker had control of raw materials and set the pace of his work. His control over raw materials led to endless squabbles over product quality as well as over embezzlement and fraud. (The worker was in a position to substitute inferior goods; give short weight; put ‘wastes’ to his own, rather than the boss's profit.) Control over the pace of work increased the possibilities for perverse—from the boss's point of view—responses to the incentive of higher wages.

No wonder that capitalists sought ways of limiting workers' control: organizational forms in which the boss, not the worker, fixed the hours and intensity of work, and where the worker would labour under the eye of a boss so that the fringe benefits (the worker's view of fraud and embezzlement) that accrued from control of raw materials would be harder to come by. The end result of this search was the factory.

This is not to suggest that efficiency considerations (in the sense of note 1) played no role in the rise of the factory. Indeed, in some industries, such as cotton textiles, efficiency considerations may have been very important. The transition from cottage to factory in cotton appears to have been predicated on the greater efficiency of the water-mill and steam-engine and the associated technical compulsion to centralization. (p.223) Boilers could not be practically adapted to the dispersed production of the putting-out system.

But two aspects of the transition from cottage to factory make it difficult to accept the proposition that efficiency considerations were generally decisive. On the one hand, the transition from cottage to factory also took place without the appreciable change in technology that accompanied the shift in cotton spinning and weaving. Woollen textiles are a case in point: the first woollen mills did not make use of water-or steam-based technologies, but utilized the same hand-powered technologies that workers had in their own homes. On the other hand, where workers were sufficiently powerful and motivated to resist factory organization, as in the case of the Coventry silk ribbon weavers, they were able to adapt steam-powered technologies to their own purposes. In Coventry, weavers ‘rented’ capital from capitalists: the capitalist's steam-engine powered a shaft that ran through the lofts of a row of attached houses, the lofts being the workshops in which the weavers' looms were located (Prest 1960).

The existence of industries like woollens in which factories arose without any technological innovation argues against the necessity of a technological basis for the factory. The existence of non-factory forms of organization around a new technology like steam, as deployed by the Coventry ribbon weavers, argues against the sufficiency of a technological explanation.

4 The Evolution of Production Relations

The subproduct division of labour and the factory were of course neither the beginning nor the end of capitalist innovation in the organization of work. Capitalists established their dominance over production only gradually; even in the late nineteenth century, workers—at least some of them—exercised substantial control over important aspects of work which would today, in the United States at least, be regarded as ‘management prerogatives’: for example, hiring and firing of assistants, setting work rules, and distributing the workers' share of the pie (Montgomery 1979, Buttrick 1952).

The basis of this control was twofold. First, the working class maintained a cohesion reflected in strikes and other forms of mutual support. Second, knowledge and skills were carefully cultivated and protected from outsiders—particularly from the boss and his agents. ‘The boss's brains’, according to a popular aphorism of a hundred years ago, ‘are under the worker's cap.’ In the words of Frederick Winslow Taylor, the father of scientific management,

foremen and superintendents know better than anyone else, that their own knowledge and personal skill falls far short of the combined knowledge and (p.224) dexterity of all the workmen under them. The most experienced managers therefore frankly place before their workmen the problem of doing the work in the best and most economical way (1967: 32).

To consolidate and extend their dominance over production, capitalists had a double task, first to break the solidarity of the workers and, second, to restructure production so as to reduce to a minimum the role of workers' knowledge and skills. The destruction of solidarity focused on transforming the orientation of the trade-union movement to ‘business unionism’, with a narrow focus on the dollar and an almost total abdication from issues of control. The role of confrontations, of which the Pullman and Homestead strikes are representative US examples, was central to this process. These were setbacks from which the labour movement took decades to recover.

In the restructuring of work key steps were the development of what Richard Edwards (1979) has labelled technical and bureaucratic control, symbolized by the assembly line and the rule book. Both were attempts to depersonalize authority relations, to legitimize capitalist control by appealing to the shared cultural value of the superiority of impersonal to personal relations (Banuri, Chapter 3 above). The intention of both technical and bureaucratic control was to create and foster the impression of a transcendent authority—the assembly line or the rule book—to which all, boss and worker alike, are subject. What the capitalist forgets to mention is that somebody makes the rules, just as somebody sets the speed of the line.

The capitalist project of reorganizing work was epitomized by scientific management, or ‘Taylorism’ as it is sometimes called in honour of its principal founding father, whose description of the traditional, ‘pre-scientific’, state of affairs was earlier quoted at length. Contrasting his proposal with traditional systems of organizing production, Taylor (1967) had this to say:

Under scientific management … the managers assume … the burden of gathering together all of the traditional knowledge which in the past has been possessed by the workmen and then of classifying, tabulating, and reducing this knowledge to rules, laws, and formulae (p. 36). These replace the judgment of the individual workman (p. 37). Thus all of the planning which under the old system was done by the workman, as a result of his personal experience, must of necessity under the new system be done by management in accordance with the law of the science (p. 38).

Scientific management has been the subject of considerable controversy. Even the degree to which it has succeeded in eliminating the worker as a thinking participant in the production process is controversial. My own view is that friend and foe alike have misread Taylor's project as (p.225) accomplished fact.4 There is however no gainsaying the considerable success achieved by capital in carrying through the larger project of which scientific management was a part, that of both reorganizing the workplace and channelling the expression of worker opposition into ‘bread and butter’ issues of pay and employment security.

It has been emphasized that these changes could not be brought about by unilateral fiat on the part of capital. They were rather the outcome of struggle, in which both resistance and accommodation played important parts. But notwithstanding protracted and sometimes heroic resistance, the fact is that accommodation predominated.

There are many reasons for this. Capitalists wielded disproportionate power, politically as well as economically. Capitalism also delivered the goods, to use Herbert Marcuse's (1966) apt phrase; the restructuring both of work and of the field of resistance took place during a period of rapid growth of real wages. This is well known and needs no elaboration.

What is less well understood is the cultural dimension of class struggle. Indeed, a premiss of this chapter is that classes do not act in history until they are armed by culture, as culture normally finds expression in setting the terms of class conflict. Resistance to oppression, like oppression itself, is invariably the union of class interest with cultural justification. The accommodation of the working class to capitalist domination can only be comprehended when we understand the limits of the cultural basis for working-class resistance.

My thesis is that the capitalist class held the high ground not only economically and politically, but culturally as well. For, on the one hand, work was never embedded in the life of the Western community and the Western cosmic order, as it is in holistic societies. (Lutheran Protestantism and the communities coming under its sway present an exception that tests this rule.) Neither was the worker empowered by culture to defend the union of conceptualization and execution which underpins control by the individual worker. So the worker was unable to find the holistic meaning immanent in embedded work and he was discouraged by cultural values which he shared with the capitalist from defending the conditions of individualistic meaning.

5 The Cultural Roots of Disembeddedness

In the West work stands outside, if not actually opposed to, life. As part of the economic sphere, work is ordinarily reduced to an instrumental (p.226) status within a pleasure/pain calculus, questions of meaning reduced to quality of work-life or on-the-job satisfaction. The disembeddedness of work is doubtless related to a larger separation of the economic from its social, political, and moral context, the central focus of Karl Polanyi's (1944) analysis of the development of capitalism. One can fault Polanyi for telescoping into the eighteenth and nineteenth centuries a process which had gained considerable momentum by the thirteenth. But if we recognize the origins of the ‘economy’ in medieval society, particularly the transformation of land and labour into ‘commodities’, we must reckon with the even greater remoteness of the disembeddedness of work.

This disembeddedness has roots in both Greek and Judaeo-Christian conceptions of work, particularly manual work, and its relationship to the purposes of human existence. Plato and Aristotle held the workman in equal contempt, though each had a different theory of how and why manual labour disqualified one from the higher pursuits of philosophy and politics. Plato, as has been observed, modelled the ideal polity of the Republic on the putative efficiency of the division of labour and specialization, buttressing this argument with an elaborate theory of knowledge (to which we shall return in another context). Aristotle, disputing both Plato's views on the polity and his theory of knowledge, nevertheless shared his views on the division of labour.5 Aristotle held even the labour of directing other people's work in low regard. ‘This science [of direction] is of no particular importance or dignity … Therefore all people rich enough to be able to avoid personal trouble have a steward who takes this office while they themselves engage in politics or philosophy’ (Politics, 1255b 32–7).

Xenophon was even more condemning:

The mechanical arts utterly ruin the bodies of those who work at them, compelling them to sit still and remain indoors, or in some cases even to spend the whole day by a fire. And when the bodies are made effeminate, the souls too become more diseased. Lack of leisure to join in the concerns of friends and of the city is another condition of those that are called mechanical; those who practice them are reported to be bad friends as well as bad defenders of their fatherlands. Indeed, in some of the cities, especially those reputed to be good at war, no (p.227) citizen is allowed to work at the mechanical arts (Oeconomicus, iv. 2–3, Strauss 1970: 17).

Xenophon makes an exception for agriculture. Agriculture was argued to be physically superior, complementary to rather than subversive of military preparedness (Oeconomicus, v) and thus at the service of and embedded in the highest aims of the political community. But for Xenophon agricultural labour was exceptional in its moral superiority too, for it was embedded in the cosmos in ways in which the craftsman's work was not and could not be. ‘[T]he earth, being a goddess [emphasis added], teaches justice to those who are able to learn, for she gives the most goods to those who serve her best’ (ibid. v. 12).

In contrast with the Judaeo-Christian origin myth, labour on the land is not punishment for sin but the fall-out from internecine struggle among the gods; according to Hesiod, it all starts with Prometheus, or rather with Zeus' retaliation for Prometheus' transgressions:

[T]he gods keep hidden from men the means of life. Else you would easily do work enough in a day to supply you for a full year even without working … But Zeus in the anger of his heart hid it, because Prometheus the crafty deceived him; therefore he planned sorrow and mischief against men (Hesiod, Works and Days, 1. 42–50).

Hesiod goes on to sing the praises of agriculture and the agriculturalist, praises that are echoed in Western conceptions of the good life and the good society from Virgil to Thomas Jefferson. The French classicist Jean-Pierre Vernant summarizes the distinctive embeddedness of agriculture in Greek thought this way:

Agriculture remains … integrated into a system of religious representation … Working the land … is participation in an order superior to man, at once natural and divine. It is in this religious context that the aspect of effort, in agricultural labour, takes on a particular significance: the confrontation with the task imposed, the difficult and strained occupation, acquire value and prestige in the measure in which a relationship with the divine is established, a kind of reciprocal tie. Labour can then appear meritorious as the counterpart of exigencies and of divine justice, in the most general sense, arete [virtue]. There is in this a theme which comes to equilibrate in the moral reflexion of Greece the affirmation of the superiority of pure thought over action (Vernant 1982 ii. 23–4).

The Judaeo-Christian tradition does not afford even this exception. At least until the Reformation, work appears to have largely negative attributes. Indeed, it lies outside the fundamental concerns of the Judaeo-Christian world-view. True, agriculture starts off promisingly enough: ‘The Lord God took the man and put him in the garden of Eden to till it and keep it’ (Genesis 2: 15). But after Adam and Eve eat of the forbidden fruit, God condemns the one to labour in the fields and the (p.228) other to labour in childbirth.6 As punishment for original sin, work clearly has religious connotations, but these are largely negative.

The specifically Christian innovations hardly help. The Sermon on the Mount directs us to imitate the fauna and flora: ‘… the birds of the air … neither sow nor reap nor gather into barns, and yet your heavenly Father feeds them … the lilies of the field … they neither toil nor spin; yet I tell you, even Solomon in all his glory was not arrayed like one of these’ (Matthew 6: 26–9).

St Paul is generally represented as having a more positive attitude towards work. To be sure, Paul said ‘If any one will not work, let him not eat’ (2 Thessalonians 3: 10). But Paul found no positive value in labour. Rather, work was preferable to idleness, breeding ground for ‘busybodies’. No more is said for honest toil than that it avoids the temptations of idleness and sloth.

By the Middle Ages, the intransigence of the Sermon on the Mount had been further mellowed by the exigencies of living in and governing this world. But Christianity remains ambivalent towards work. The ideal remains trust in God to fulfil all of one's needs, but the reality is a social, political, and economic order which requires a diversity of labours for its maintenance and reproduction.

Piers Plowman, the dream allegory written by William Langland, an otherwise obscure fourteenth-century English cleric, illustrates the ambivalence.7 The first version extant has ‘Truth’ pardoning the sins not only of Piers, but along with him the sins of ‘every labourer on earth who lives by his hands, who earns his own wages and gets them honestly, living in charity and obeying the Law’ (Book vii). This version of the text ends with Piers recognizing the ‘pardon’ as nothing but an excerpt from the Athanasian Creed. In a rage, he destroys the bogus pardon, invoking the Sermon on the Mount to justify his decision to ‘give up my sowing, and cease from all this hard labour’.

A later version of Piers carries the ambivalence further. Additional text transforms Piers into St Peter, who, as Christ's vicar, presents the division of labour as ‘diversities of gifts, but the same Spirit’. For the sake of social harmony ‘each craft must love the other’ (Book xix).8 A key difference from the first portion of the text is the injunction that ‘you must clothe and feed yourselves’, a departure from the birds-of-the-air, lilies-of-the-field view with which the earlier text concludes.

(p.229) Three points are worth making here. First we should take note of a problem with which medieval Christians—from more or less orthodox philosophers and theologians to heretics and rebels—wrestled: the problem of coming to terms with the transformation of Christianity from a religion of oppressed and marginalized minorities in the Roman world to Christianity as the official religion of the oppressors as well as the oppressed. In this respect, the transformation of Piers into St Peter is significant, reflecting the change of Christ as the Humble and Humiliated Redeemer into the Powerful and Glorious Ruler, and his Church from a federation of local bands of the faithful into the most powerful institution of Christendom (!), an institution which could hardly afford to take seriously the birds of the air or the lilies of the field as models for human attitudes towards work.

But even then, and this is the second of the three observations suggested by the later Piers, work is not embedded in the Christian scheme of salvation. It reflects rather the Pauline disposition towards the lesser of evils. Each may be endowed by the Holy Ghost (Grace in the allegory) with a ‘gift’ to permit him to gain an honest living, but this is to fulfil no cosmic plan, no divine order. No one comes into touch with the Godhead through his gift. Rather, the point is, as for Paul, ‘that Idleness, Envy and Pride should never get the better of him’ (Book xix).

At the end of Piers's story, however, and this is the final point to be made here, the institutional needs of the Church are again subordinated to the Sermon on the Mount. ‘Nature’ invokes the power of love:

‘Learn to Love, said Nature, and give up everything else.’

‘But how shall I clothe and feed myself and make a living?’

‘If you love sincerely’ he said, ‘you will never lack food and clothing as long as you live’ (Book xx).

How do we account for Langland's pervasive ambivalence? It undoubtedly owes something to the ambivalence of his own position as a lowly cleric. Closer to the mass of peasants than to any seat of temporal power (he has harsh words for priests who prefer the high life of London to their country parishes), Langland could pay closer attention to the spiritual message of Christianity than to the administrative exigencies of the Church.

In short, in so far as Piers Plowman speaks for medieval Christians, work is part of the social order, but this social order remains, even after a millennium, a makeshift and transitory order. It is a holding pattern of social existence still awaiting the Antichrist and the Last Judgement. Like marriage, work continues to be represented in the Pauline manner, as a lesser of evils. Do-Well, Langland's Everyman, may require work to avoid (p.230) idleness, but the seeker after salvation who breaks out of the holding pattern can trust to God for all his needs—like birds and lilies.

Thus, at the end of the Middle Ages, as at the dawn of Christianity, work remains outside the central purpose of human existence. But the Reformation brings a dramatic change in attitudes towards work. Specifically, in the Lutheran dispensation, the ‘calling’ becomes central to the religious life; work is firmly embedded in the Lutheran conception of divine and human purpose.

It is, however, important to separate the embeddedness of work in Lutheran communities from the more general question of the link between Protestant religion and capitalist economy. Max Weber (1930) and R. H. Tawney (1938), though arguing opposite sides of the case, both affirm a strong link between the Puritan doctrine derived from John Calvin and the rise of English capitalism. From our perspective, even if we accept Weber's position, the effects of Calvinism are ambiguous. For one thing, Calvinism fundamentally alters the notion of calling so as once again to disembed work. Instead of a calling to a particular line of work, Calvinism emphasizes election or predestination. One's worldly activity is now important only as a sign of God's favour or disfavour: economic success is a superficial characteristic of (if, to be sure, hardly a sufficient condition for) salvation. As a sign of divine grace, the work one does becomes less important than the rewards one receives for the work. Calvinism embeds money, not work. Richard Baxter's memoirs, written in the waning years of the seventeenth century, make this clear (quoted in Tawney 1938: 241):

If God show you a way in which you can lawfully get more than in another way (without wrong to your soul or to any other), if you refuse this, and choose the less gainful way, you cross one of the ends of your Calling, and you refuse to be God's steward.

Within Germany and the Nordic countries, Lutheran ideas about work may still be a motive force behind a continuing (craft) dedication to quality. But elsewhere, modern Protestantism would appear to owe more to Calvin than to Luther. In the United States, for example, virtually all that survives of the Lutheran idea of a ‘calling’ is that the non-academic, blue-collar wing of the American secondary school often goes under the name of ‘vocational education’.

The Marxian vision of a post-capitalist, even a post-socialist, Utopia no more embeds work than does the Greek or Judaeo-Christian world-view. Marx had no conception of an organic or holistic society, with a unified world-view in which work would be embedded. Louis Dumont (1977) has rightly emphasized the extent to which Marx shared the individualism of his culture: of course he differed fundamentally from the dominant ideology in his insistence that capitalism stifled the (p.231) individual, but this is not the same as taking issue with individualism in and of itself. His paean to work in the German Ideology

In communist society, where nobody has one exclusive sphere of activity but each can become accomplished in any branch he works, society regulates the general production and thus makes it possible for me to do one thing today and another tomorrow, to hunt in the morning, fish in the afternoon, rear cattle in the evening, criticize after dinner, just as I have a mind, without ever becoming hunter, fisherman, shepherd, or critic (Marx and Engels 1846: 22).

—has to be understood in terms of the possibilities of individualistic rather than holistic meaning.

Indeed, it is questionable whether Marx would consider hunting, fishing, cattle-rearing, and criticism as work at all. Heir to both Greek and Judaeo-Christian attitudes, he would, arguably, assign all this to the realm of leisure. In the realm of work, however society might be organized, the best Marx hoped for (as we have observed in note 4) is ‘rationally regulating the interchange with Nature … and achieving this with the least expenditure of energy’. To quote the relevant passage at length:

In fact, the realm of freedom actually begins only where labour which is determined by necessity and mundane considerations ceases; thus in the very nature of things it lies beyond the sphere of actual material production. Just as the savage must wrestle with Nature to satisfy his wants, to maintain and reproduce life, so must civilized man, and he must do so in all social formations and under all possible modes of production. With his development this realm of physical necessity expands as a result of his wants; but at the same time, the forces of production which satisfy these wants also increase. Freedom in this field can only consist in socialized man, the associated producers, rationally regulating their interchange with Nature, bringing it under their common control, instead of being ruled by it as by the blind forces of Nature; and achieving this with the least expenditure of energy and under conditions most favourable to, and worthy of, their human nature. But it nonetheless still remains a realm of necessity. Beyond it begins that development of human energy which is an end in itself, the true realm of freedom, which, however, can blossom forth only with this realm of necessity as its basis. The shortening of the working-day is its basic prerequisite (Marx 1894: 799–800).

We have considerably shortened the working day since Marx's time, but we are no closer to a society in which either work or leisure is meaningful. Probably we cannot approach either goal so long as the two are separate and distinct, with work—like chicken-pox—conceived of as something to be got over as quickly as possible.

6 Knowledge Systems: Techne and Episteme

The disembeddedness of work means that a defence against the capitalist project of assuming control of process and product could not be based (p.232) on the cultural significance of the worker's activity, on its holistic meaning.

But what if cloth-making or gun-making or wheel-making had the personal significance for the weaver or the gunsmith or the wheelwright that teaching and research have for the university professor? What if the same possibilities for investing work with individualistic meaning were perceived in the first set of occupations as in the second? Is it not plausible that the prerequisites of meaningful work would have been defended more passionately? The premiss of the argument that will be made in this section is that workers' accommodation to the capitalist project of controlling work was facilitated by the shared cultural assumptions of workers and capitalists which devalued the worker's efforts. Specifically, it will be argued, first, that a basis of workers' control of the production process was a system of knowledge that intimately linked conception and execution; and, second, that this system of knowledge was implicitly regarded—by workers and capitalists alike—as inferior to the knowledge system which capitalists used to restructure production so as to separate conception from execution, the better to bring execution under their control.

Scientific management provides a useful and important text for examining this argument, and the cultural values at issue are once again deeply rooted in Western culture. But before we attempt to read the text or explore its roots, we require much more clarity concerning what is meant by the notion of a system of knowledge. ‘Knowledge system’ has by now come to have a fairly wide currency, frequently being associated with French structuralists and post-structuralists of the ilk of Claude Levi-Strauss and Michel Foucault. But the concept is used too idiosyncratically to have a standard meaning to which one can refer.

Knowledge system here is used to characterize different ways of knowing in terms of four characteristics: epistemology, transmission, innovation, and power. Epistemology is the first issue: how do we know what we know? Every system of knowledge has its own theory of knowledge, that is, its own theory of what counts as knowledge. Transmission is closely related to epistemology. How do we go about distributing and receiving knowledge? Innovation refers to the process of change: how does the content of what we (collectively) know get modified over time? Finally, power: what are the political relationships between members of a community who make use, in greater or lesser measure, of the same system of knowledge? And how does a particular knowledge community relate to other knowledge communities?

The point of the term system is twofold. Its first purpose is to suggest that epistemology, transmission, innovation, and power are not attributes of knowledge in general but characteristics of particular ways of knowing. There is no single epistemology, but specific epistemologies which belong (p.233) to distinct ways of knowing. Equally there are distinctive ways of transmitting and modifying knowledge over time. And different ways of knowing imply different power relationships among the people who share knowledge and between ‘insiders’ and ‘outsiders’.

The links among these several characteristics are a second systematic aspect of knowledge. How we know and how we learn and teach, how we innovate and how we relate to power—these characteristics of knowledge mutually interact, as well as interacting with the basic constructions that underlie each particular way of knowing.

All this may become clearer if we concretize the discussion in terms of distinct knowledge systems which I shall call techne and episteme. It should be explained that the Greek terms are intended to evoke rather than to define; techne and episteme will be defined, not by (approximate) English equivalents of the Greek, but by a series of oppositions.

On the one hand, episteme is knowledge based on logical deduction from self-evident first principles. The best model is perhaps Euclidean geometry, though Euclid's axioms have turned out with the passage of time to be less self-evident than had once been supposed—we now have a variety of geometries each with its own axiomatic basis. ‘Logical deduction’ implies proceeding by small steps with nothing left out, nothing left to chance or to the imagination. Besides the mathematical theorem, the computer program comes to mind as a model of epistemic knowledge.

Epistemic knowledge is analytic. It decomposes, breaks down, a body of knowledge into its components. It is thus directly and immediately reproducible. It is fully articulate, and within episteme it may be said that what cannot be articulated does not even count as knowledge.

Episteme lays claims to universality, to being applicable at all times and places to all questions. Indeed adherents of episteme do not in general see it as one system of knowledge among many, but as knowledge pure and simple.

Epistemic knowledge is purely cerebral. Mind is separate from body, and episteme pertains to the mind alone. The statement ‘I feel there is something wrong with what you are saying’, which is to say ‘I sense something is wrong, but I cannot articulate what or why’, has no place within episteme.

Even when pressed into action, episteme is theoretical. Once the tentative and provisional nature of any axiomatic scheme is recognized, epistemic statements are necessarily hypotheses. Indeed, without entering into the nuances of the debate between Karl Popper (1968) and his critics (Kuhn 1970; Lakatos 1970; Putnam 1974) it can be said that episteme is geared one way or another to verification. Its very procedure, the insistence on small steps that follow immediately and directly upon one another, precludes discovery and creativity. To discover or to create through episteme would be like the proverbial monkey typing Shakespeare: he (p.234)

Table 3. Corresponding Attributes of Episteme and Techne



Logical deduction/Self-evident axioms
















Egalitarian internally/ Hierarchical externally

Hierarchical internally/ Pluralistic externally

might some day do it, but we would be hard pressed to find the wheat among the chaff.

Finally, episteme is impersonal knowledge. Like the Christian God (Romans 2: 11), episteme is impartial; it is in principle accessible to all on equal terms. It is thus not only theoretical knowledge, it is theoretical knowledge of theoretical equals. By contrast, like the Christian faith, episteme not only distinguishes those within from those outside the knowledge community; as Christianity denies the possibility of salvation to unbelievers, episteme disenfranchises those outside. From the universalistic claim of episteme it is an easy and direct step to the view that those lacking in episteme are lacking in knowledge itself. Table 3 opposes these attributes of episteme to corresponding attributes of techne.

In contrast with the basis of episteme in logical deduction from self-evident axioms, the bases of techne are as varied as the authority of recognized masters to one's own intuition. Opposed to the small steps of episteme are both received doctrine and the imaginative leaps which all at once enable one to fit the jigsaw puzzle together. It is in either case a knowledge of the whole, difficult to break down into parts. In contrast with the analytic nature of episteme, techne is indecomposable.

Possessors of techne often find it impossible to articulate their knowledge. They are generally aware that they possess special knowledge, but their knowledge is implicit rather than explicit. It is revealed in production of cloth or creation of a painting or performance of a ritual, not in manuals for weavers, artists, or priests.

Technic knowledge makes no claims to universality. It is specialized in nature and closely allied to time and place. It always exists for a particular purpose at hand; techne is contextual.

Techne belies the mind/body dualism which is basic to episteme. Under techne one knows with and through one's hands and eyes and heart as well as with one's head. Techne is knowledge which gives due weight to what Martha Nussbaum and Amartya Sen (1989: 316) have called the (p.235) ‘cognitive role of the emotions’ as well as to the knowledge of touch. Feeling, in both senses of the term, is central to techne; techne is at once both tactile and emotional.

Techne is intensely practical, to the point that, as has been suggested, it reveals itself only through practice. This is not to deny the existence of an underlying theory, but the theory is implicit rather than explicit, not necessarily available, perhaps not even usually available, to practitioners.

Technic knowledge is geared to creation and discovery rather than to verification. Even a mathematical theorem is largely the product of techne, although the proof must, by the very requirements of the knowledge system on which mathematics is based, be cast in terms of episteme.

Finally, where episteme is impersonal, techne is not and cannot be. It normally exists in networks of relationships and cannot be transmitted or even maintained apart from these relationships. The normal avenues of transmission—parent-child, master-apprentice, guru-shisha—are intensely personal (cf. Chapter 3).

These are not normally relationships among equals. There is a hierarchy blending age, power, and knowledge. But it should be observed that the hierarchy is typically linear rather than pyramidal, as wide at the top as at the base. Thus those at the bottom have a reasonable expectation (though no guarantee) of moving up to the top with the passage of time. It is the hierarchy of the guild, where every apprentice can expect to be a master, not that of the factory, where few workers can become foremen, let alone executives.9

If techne is internally hierarchical, it is more open externally. Laying no claim to universality, recognizing limits of time, place, and purpose, techne does not inherently subordinate those outside a particular community of knowledge to those inside the community. Techne may not be inherently egalitarian in terms of its external relations, but it is at least pluralistic.

In terms of the four characteristics that have been proposed to distinguish knowledge systems, the differences between episteme and techne are striking. Episteme recognizes as knowledge only that which is derived by the rules of logic from axioms acceptable as self-evident first principles. Techne by contrast recognizes a variety of avenues to knowledge, from authority to immediate experience: the test of knowledge is practical efficacy.

The transmission mechanisms are as different as the epistemologies. (p.236) Epistemic knowledge in principle is accessible through pure ratiocination, but in practice, episteme is generally acquired through formal schooling. Indeed knowledge in the West has more and more come to be equated with what is taught in the schools, and the schools in general are dedicated to episteme,10 so much so that a young friend suggested opposing book knowledge and street knowledge in place of the opposition between episteme and techne. The canonical way of transmitting techne is, as has been indicated, through a personal nexus epitomized by the master-apprentice relationship. The master's example more than any precept instructs the apprentice, who absorbs almost unconsciously what he is taught. Almost anybody can acquire the rudiments of a craft in this way, but quality is a matter of intuition, of a heightened sense of touch and feel developed through years of practice.

Epistemic innovation leads a double life. The formal model allows one only to replace an erroneous logical derivation with a correct one or to change the assumptions. One can supplement existing axioms, or, more rarely, replace existing axioms by new ones, as Newton did for his predecessors and Einstein did for Newton. With new axioms one can proceed to new theorems by old methods: the new theorems are simply logical entailments of the new assumptions. In practice, as has been noted, a considerable admixture of techne is involved even in epistemic innovation: the innovator has to know where he is going and the map is provided by his intuition rather than his logic. Technic innovation is largely a matter of trial and error. This is not to say it is haphazard, but the underlying structure of technic innovation, like the techne it modifies, is often hidden from the innovator himself.

If knowledge is a text, the canonical form of epistemic innovation is criticism. Innovation takes the form of a direct assault, a challenge to logic or to first principles themselves. The canonical form of technic innovation, by contrast, is commentary, emendation and explanation of the text. The authority of the fathers is not challenged but reinterpreted. For this reason, epistemic innovation can flourish only in a community of equals, where respect for personal authority is relatively attenuated. The attacks of Peter Abelard on the doctrinal authority of popes and saints presupposed that in the West even religious knowledge had, by the year AD 1100, come to be regarded as epistemic in nature.

Episteme and techne reverse internal and external power relations. As has been noted, episteme assumes a community of equals, whose superior (p.237) knowledge makes them collectively and individually superior to those outside. Techne, by contrast, presupposes a hierarchy of knowledge and a corresponding hierarchy of power within the knowledge community. But the community as a whole can relate in different ways to other communities. According to context, it can be more knowledgeable, and hence wield greater power, or less knowledgeable and correspondingly weaker.

It will be recognized that episteme and techne are ideal types. Episteme comes close to what many mean by science, and ‘science’ is indeed one of the words used to translate the Greek. (‘Knowledge’ is another translation, which suggests that the claims of episteme to universality are long-standing.) Techne is more difficult to pin down. As the root indicates, it contains elements of ‘technique’; ‘art’, one translation of techne, conveys some of its flavour. But contemporary scholarship glosses both episteme and techne as ‘science’, a translation perhaps more congenial to my purposes in its emphasis on the common field of these terms (see Nussbaum 1986: 444). In any case an opposition between ‘science’ and ‘art’ would unduly and prematurely constrain the meaning of these terms, and the very openness of the Greek terms (to those of us who would claim to be largely innocent of classical Greek) has much to commend it at this stage of our proceedings.

Oliver Sacks has written of a condition called Tourette's syndrome, a nervous disorder ‘characterized by an excess of nervous energy, and a great production and extravagance of strange motions and notions’ (1985: 87). An element of Sacks's story is that Tourette's syndrome simply dropped out of sight soon after its identification in the late nineteenth century. ‘Dropped out of sight’ may appear a peculiar locution, but as Sacks tells the story it is quite accurate: there is no indication that Tourette's syndrome occurred with any less frequency, but it went unnoticed by the medical profession for the better part of a century, until it was restored to respectability, in part through Sacks's own efforts, in the 1970s. It is now a flourishing disease—in the sense of having not only medical legitimacy but also an association of victims and their sympathizers. The interesting question for our purposes is how an obvious and dramatic disorder like Tourette's syndrome could go from visibility to invisibility and back.

In our terminology, Sacks's answer is that as Western medicine became increasingly epistemic in the twentieth century, there was less and less place for disorders for which no organic basis had been found. The medical knowledge system defined medical knowledge, and what didn't fit simply dropped from view. As far as the medical profession was concerned, Tourette's syndrome ceased to exist. The rediscovery of Tourette's syndrome is a tribute to the efforts of Sacks and those affected (p.238) by Tourette's to resist the monopoly of medical episteme. Although friends and disciples of medical episteme might take comfort in the subsequent identification of an organic basis for Tourette's syndrome, Sacks emphasizes that neither the disorder nor its management can be conceptualized in purely epistemic terms.

Sacks remarks in passing that he was alerted to the importance of Tourette's syndrome by an experience that occurred the day after he clinically observed a patient with Tourette's: within the space of a single hour, he noticed three more examples of a disorder that the more charitable of his colleagues had labelled extremely rare and the less charitable had termed mythical.

I have much the same feeling about techne and episteme. Having once been sensitized, I now see them everywhere. This is of course simply a way of recognizing that the distinction between episteme and techne has considerable overlap with a variety of dichotomies that others have proposed to distinguish ways of knowing. The best known, perhaps, is Robert Pirsig's distinction between classical and romantic knowledge in Zen and the Art of Motorcycle Maintenance (1976). ‘A classical understanding’, according to Pirsig, ‘sees the world primarily as underlying from itself. A romantic understanding sees it primarily in terms of immediate experience’ (p. 66). The emphasis is different but ‘classical’ clearly resonates with epistemic, and ‘romantic’ with technic. Even closer, perhaps, is Michael Polanyi's (1958) characterization of ‘tacit’ knowledge as a distinctive way of knowing, in which touch and feel play an essential role—-as they do in my characterization of techne. In a different context, Ian Hacking (1975) has argued that the persistence until the Renaissance of a conception of knowledge as logical deduction from self-evident first principles was the major obstacle to the development of a theory of probability. Probability, for Hacking, can emerge only when the warp of a stochastic conception of events becomes enmeshed with the weft of less than certain knowledge, knowledge which varies with empirical evidence, a conception completely at odds with the certainty of logical deduction from a sound axiomatic base.11 Finally, in yet another context, Jerome Bruner (1962) has distinguished between ‘right-handed’ and ‘left-handed’ knowledge, the first stressing logic and the second intuition, a distinction easily assimilable to the opposition between episteme and techne.

Free association (a techne in itself) leads from Jerome Bruner back to Oliver Sacks. His best-seller of 1985, The Man Who Mistook His Wife for a Hat, has already been mined for Tourette's syndrome. But there is much more to this book. The Clinical Tales of its subtitle are the vehicle (p.239) for a plea for a new system of medical knowledge (my terminology, not Sacks's), one that integrates techne and episteme and is thus worthy of the complexity of human order and disorder, in which mind, body, and soul are inextricably intertwined.

The disorders Sacks recounts can be seen in my terms as imbalances of techne and episteme. The patient from whom the book takes its name, for instance, suffering from visual agnosia, had totally lost the techne required to distinguish a hat from a head. He lacked judgement, quintessentially a matter of techne rather than episteme. Looking at pictures, ‘He failed to see the whole, seeing only details, which he spotted like blips on a radar screen’ (p. 9). A rose which Sacks offers the patient for examination becomes ‘“a convoluted red form with a green linear attachment”’ (p. 12).

With much assistance from his wife, Sacks's patient managed to maintain a semblance of carrying out the daily functions of life by reorganizing them into an epistemic system. As long as he could break down an activity like dressing or eating into a series of detailed operations, he could cope, but, just as he was totally at a loss in seeing a whole picture, he was totally unable to carry through an operation which could not be decomposed into details. Episteme is clearly insufficient to organize experience.12

Now to the Bruner connection. Imbalances of techne and episteme can frequently be related to right- and left-hemisphere lesions, and one is therefore tempted to reduce techne and episteme to an organic phenomenon, locating techne in the right side of the brain and episteme in the left. In fact, when I met Dr Sacks in the autumn of 1986, I put this possibility to him, but he was quite hostile to the idea. It would be a great mistake, he warned, to attempt to assign normal mental functions to one hemisphere or the other on the basis of pathologies. The mind, one may infer, is indecomposable, like techne itself.

It has been remarked that techne and episteme are ideal types. It is not surprising therefore that in pure form, by itself, techne or episteme will function with comic, bizarre, or grotesque results, as in Sacks's clinical histories. But the practical requirement that episteme and techne co-operate (p.240) does not prevent a universalizing episteme from pushing its claim to a monopoly of knowledge.

Having illustrated this tendency by the ups and downs of Tourette's syndrome, let me turn to an example closer to home, one drawn from economics. Milton Friedman is well known outside the economics profession for his extreme opposition to government intervention in the economy; Victorian England looks positively dirigiste in comparison with Friedman's ideal. Friedman is equally well known inside the profession for his contributions to economic theory, most of which are regarded as quite within the mainstream, if somewhat towards one end of the spectrum. Among the less controversial contributions is an article on methodology which has become a classic in the field, ‘The Methodology of Positive Economics’ (1953). One strand of Friedman's argument is particularly important for our present purposes.

The point of the article is that we should judge a theory by its conclusions, by how well it squares with observation, not by its premisses—a position on its face perhaps not uncongenial to those who operate primarily in terms of techne. But Friedman's purpose is not to defend techne; rather it is to defend an episteme shorn even of the minimal grounding in experience implied by the requirement that its premisses be self-evident.

One of Friedman's arguments is particularly salient for our purposes. His text is the postulate that individual behaviour is governed by utility maximization, economists' jargon for Jeremy Bentham's pleasure/pain calculus. It really doesn't matter whether or not individuals actually perform such calculations, Friedman assures us. What matters is whether they act as if they maximize utility. Friedman offers a striking analogy to illustrate and defend the ‘as if’ argument. He invites us to consider a superb billiard player lining up a delicate shot. This billiard player does not write down, much less attempt to explicitly solve, the complicated differential equations of the interactions of the billiard balls. It is enough, according to Friedman, that the successful billiard player acts as if she were an expert in differential equations.

Economists have generally found the billiard analogy compelling, and indeed it is compelling—if we are prepared to grant that all knowledge is epistemic. For in that case the billiard player must willy-nilly be a closet mathematician (and a superior one at that, since no closed-form solution exists for the three-body problem which is the epistemic essence of billiards). But this identification of knowledge with episteme is precisely the point at issue. Friedman has no warrant—apart from the weight of Western intellectual tradition—to reduce the billiard player's techne to the mathematician's episteme. In the present perspective, her knowledge is of an entirely different kind, belonging to a different system, and it is no more instructive to suggest that the billiard player lines up her shots (p.241) ‘as if’ she were a mathematician than to suggest that the mathematician solves her equations ‘as if’ she were trying to make a score at the billiard table. Techne cannot be reduced to episteme any more than episteme can be reduced to techne.

Let me take another example from economics. In 1921 another great economist of the so-called Chicago School, Frank Knight, proposed to distinguish two kinds of chance, ‘risk’ and ‘uncertainty’. In Knight's language, risk applies to the class of chancy events for which an underlying probability distribution is known. Customary examples are the chances of an American woman aged fifty-three surviving for ten years, or the temperature in Helsinki exceeding 30°C on a July day between now and the year 2000. Uncertainty applies to the class of events which are not only chancy, but for which not even an underlying probability distribution is known. Examples are the extent of the market for videophones by the end of the century, or Israeli acceptance of a Palestinian state before this book is published. In Knight's view, there is an essential difference between the two classes of events: a fifty-three-year-old woman can buy ten-year term insurance, the cost of which will reflect the actuary's mortality tables, but, when all is said and done, the videophone manufacturer must rely on his hunches about the future of communications technologies: he cannot insure an investment in videophone-making machinery against the failure of a market to materialize.

For Knight, the point of the distinction is that whereas risk can be made routine (‘epistemized’, I should say) through such devices as insurance, uncertainty requires a special class of economic agents. Profit is the reward of those whose superior techne (once again my word, not Knight's) guides them to business strategies which are eventually validated by the market. Loss is the penalty for inferior techne.

This is not the place to examine or criticize Knight's theory. Indeed, my interest lies elsewhere, with the subsequent history of the economics of chance. In a word, like Tourette's syndrome, and for much the same reason, uncertainty has more or less disappeared from sight. Mainstream economic theory assimilates uncertainty to risk by the device of subjective probabilities. The first step in the argument is to blur the Knightian distinction by observing that, at the risk end of the spectrum, we can never ‘really’ know a probability distribution; we only have more or less relevant information on particular samples. By the same token, at the uncertainty end, we are never totally without information about the likelihood of alternative outcomes. Moreover, the institutional lines are blurred too: markets exist in a wide range of situations that are closer to uncertainty than to risk, for instance, futures markets in commodities and the ad hoc insurance contracts that have long been the speciality of Lloyd's of London.

(p.242) The blurring of the distinction between risk and uncertainty in practice ignores the obvious in Knight's theory—that these categories are ideal types. (Knight was after all a student of Max Weber.) The existence of mixed cases and fuzzy lines becomes the pretext for abolishing the distinction altogether. If one is disposed in that direction, it is an easy intellectual step from the fuzziness at the edges of the distinction to the idea that all probabilities are personal or subjective in nature. And this indeed is the dominant view in mainstream economic theory today. As with utility maximization, it does not matter for the theory whether individuals consciously calculate the subjective probability distributions required by the theory. ‘As if’ behaviour will do just fine.

What accounts for the success of subjective probability theory? Certainly not its predictive power. Ever since the theory was elaborated, critics have observed that untutored individuals violate the precept of uniform treatment of risk and uncertainty (see, for example, Ellsberg 1961). Indeed many persist in going their own way even in the face of instruction. But neither the criticism nor the supporting evidence has had perceptible influence on the status of the theory.13

It is not predictive power but theoretical unity which commends subjective probability to the mainstream. The hidden agenda is to eliminate techne and any action based on techne from economic theory so as to maintain the epistemic purity of the economic conception of knowledge and behaviour. The techne of coping with uncertainty, like all technai, is subsumed as an inferior form of episteme, rather than being regarded as a distinct, complementary system of knowledge and basis for action.14

These examples are meant to illustrate two basic points about episteme and techne, first, the complementarity of the two systems of knowledge (p.243) in everyday life, and, second, the attempt of episteme to crowd out techne in fields as diverse as economics and medicine. But all this is by way of introduction to the main theme: the interaction between techne and episteme in production and in the struggle over control of the production process.

7 Techne and Episteme In Production: the Wheelwright's Shop

Though practical knowledge invariably combines the two systems, workers' knowledge has traditionally been organized much more in terms of techne than in terms of episteme. The extent to which techne predominates is well illustrated by George Sturt's account of wheelwrights and their work in a small town in Victorian England. Sturt's account was first published in 1923 and has been reprinted ten times since, its enduring popularity a tribute to Sturt's ability to evoke the quality of work and life in a place and time both nearby and far away.

Sturt was a disciple of Ruskin, under whose influence ‘I felt that man's only decent occupation was in handicraft’ (Sturt 1923: 12). Unfortunately, although the son and grandson of master wheelwrights, Sturt had not learned the wheelwright's trade as a young man, a lack he was to regret after his father's terminal illness obliged him in 1884 to abandon the village schoolhouse, where he had been a teacher, for the wheelwright's shop, where for the next third of a century he was to be the proprietor.

Even allowing for Sturt's Ruskinian predilections, his description of the shop is arresting;

Reasoned science for us did not exist … eye and hand were left to their own cleverness … A good wheelwright knew by art but not by reasoning the proportion to keep between spokes and felloes; and so too a good smith knew how tight a two-and-a-half inch tyre should be made for a five foot wheel and how tight for a four-foot, and so on. He felt it in his bones. It was a perception with him. But there was no science in it; no reasoning. Every detail stood by itself, and had to be learnt by trial and error or by tradition.

… it was years before I understood why a cart wheel needed a certain (p.244) convexity … none of [the men], any more than myself could have explained why it had to be so (pp. 19–20).

Equally arresting is the power that knowledge gave the worker:

… I have known old-fashioned workmen refuse to use likely-looking timber because they held it to be unfit for the job.

And they knew. The skilled workman was the final judge. Under the plane (it is little used now) or under the axe (it is all but obsolete) timber disclosed qualities hardly to be found otherwise. My own eyes know because my own hands have felt, but I cannot teach an outsider, the difference between ash that is ‘tough as whipcord’ and ash that is ‘frow as a carrot,’ or ‘doaty,’ or ‘biscuity.’ In oak, in beech, these differences are equally plain, yet only to those who have been initiated by practical work (p. 24).

All the elements of techne are present. The emphasis on touch and feel, trial and error, and tradition: ‘My own eyes know because my own hands have felt …’. The knowledge that cannot be articulated: ‘but I cannot teach an outsider …’.

Tradition and rule of thumb come before understanding. Sturt indeed spends an entire chapter (ch. 18) explaining ‘dish’, the convexity necessary for a proper wheel to which the extensive quotation above alludes. Sturt emphasizes that wheelwrights—that he himself—was putting dish into wheels long before he understood why. For the men, that their fathers had built wheels this way was reason enough. Only ‘intellectuals’ like himself felt it necessary to epistemize the lore of wheel and wagon construction. As Sturt wrote in another context, ‘No rule of or scale of foreway was known in my shop, but wheelwrights of great experience could get the right effect by exercising their judgment on it’ (p. 137).

Sturt's summary deserves to be quoted at length:

The nature of this knowledge should be noted. It was set out in no book. It was not scientific. I never met a man who professed other than an empirical acquaintance with the waggon builder's lore …

In a farm yard, in tap room, at market, the details were discussed over and over again; they were gathered together for remembrance in village workshop … the whole body of knowledge was a mystery, a piece of folk knowledge, residing in the folk collectively, but never wholly in any individual (pp. 73–4).

Sturt must have been aware of the irony. Having lived far enough into the twentieth century to see a world which had no further use for the wheelwright's techne, Sturt sought to preserve this techne by recasting it as episteme: one almost feels capable of crafting an entire wagon after reading The Wheelwright's Shop, but undoubtedly this is the spell of Sturt's techne—as author rather than wheelwright.

The automobile must bear final responsibility for the demise of the wheelwright's craft, but even before the automobile, innovations in the (p.245) technology of wheel manufacture had combined with the extension of the market wrought by the railway to change the business environment in which Sturt and others like him operated. Moreover, the production relations of Sturt's shop had a dynamic of their own, which nicely illustrates the power relations that inhere in techne.

Long before the advent of the automobile, Sturt began a saga of innovation, one that started from disillusionment. Within a few years of taking over from his father, Sturt abandoned Ruskin:

I realized how impossible it would be to carry out any of the Ruskinian notions, any of the fantastic dreams of profit sharing, with which I started. The men in the shop, eaten up with petty jealousies, would not have made any ideals work at all (p. 200).

Sturt did not spare himself. He, as well as the men, was to blame: ‘under my ignorant management, the men had grown not so much lazy as leisurely’ (p. 200).

Sturt evidently felt himself caught between the proverbial rock and the equally proverbial hard place:

To discharge the men was not to be thought of. How could I ever find fault with those who had taught me what little I knew of the trade and who could but be only too well aware of how little that was? Moreover they were my friends. Business was troublesome even on the best of terms, but I could not have found the heart to go on with it at all at the cost of the friction which must have come if I had begun trying to ‘speed up’ my friends and instructors. Meanwhile, none the less, the trade these friends of mine depended on for a living was slipping away, partly by their own fault (p. 200).

‘What’, Sturt asks rhetorically, ‘was to be done?’ His solution was less original than the anguish it caused him:

Eventually—probably in 1889—I set up machinery: a gas engine, with saws, lathe, drill, and grindstone. And this device, if it saved the situation, was (as was long afterwards plain) the beginning of the end of the old style of business ….

… there in my old-fashioned shop the new machinery had almost forced its way in—the thin end of the wedge of scientific engineering … ‘The men’, though still my friends, as I fancied, became machine ‘hands’. Unintentionally, I had made them servants waiting on gas combustion … they were under the power of molecular forces. But to this day the few survivors of them do not know it. They think ‘Unrest’ most wicked (pp. 200–1).

Some will read the conclusion to Sturt's tale as simple sentimentality. One cannot have gas-driven machinery without the new production relations it entails. But a close reading suggests a different interpretation.

Consider the nature of authority in a shop like Sturt's. Formally the boss's power rests in ownership. But substantively ownership is an insufficient basis. To command effectively requires legitimacy, and to (p.246) achieve legitimacy requires that the boss be a superior wheelwright. A legitimate ruler commands by example as well as by fiat. He sets the pace because he knows by experience what constitutes a fair day's work and he can, if need be, do that much and more.

George Sturt's father commanded by example. So did his grandfather. As superiors in techne, both held the respect as well as the affection of the men. But lacking the apprenticeship necessary for the development of superior skill, Sturt could only, as he says, lay claim to their friendship—not to their respect. Unable to exercise power in the traditional ‘mechanical’ mode based on a superior command of techne, Sturt, in this reading, turns to a new ‘organic’ mode; a new technology makes the men dependent on him in a new way in which their traditional knowledge plays a much diminished role. With gas-driven machinery comes a new system of knowledge, ‘scientific engineering’, in which Sturt is on a more than even footing.

It is not machinery which gave rise to a new system of production relations, but the need for a new system of production relations which led to the introduction of particular kinds of machinery. Had Sturt been able to command in the way of his forefathers, gas-driven machinery need not have transformed production relations any more than the steam-powered looms transformed the production relations of Coventry ribbon weavers (Prest 1960, summarized above in Section 3).

8 Scientific Management

Sturt's problem was fundamentally the same as Frederick Taylor's. In the words of the father of scientific management, ‘The shop was really run by the workmen, and not by the bosses. The workmen together had carefully planned just how fast each job should be done, and they had set a pace for each machine throughout the shop’ (Taylor 1967: 48–9).

Appointed a gang boss in the Midvale Steel Company, Taylor set about challenging worker control. Lacking Sturt's fastidiousness—his father and grandfather had not run the shop before him—Taylor was initially determined not to let any bonds of affection (which in Taylor's case were unlikely in any case to have been reciprocal—see Kakar 1970) stand in his way.

The story of scientific management begins with his essentially pyrrhic victory in those early Midvale efforts. Despite considerable success, he regarded these efforts as a failure—for the same reason that made George Sturt hesitate to intervene at all. As Taylor later told the story (Taylor 1967: 52–3), it was anguish at the ‘bitter relations’ caused by these first attempts to take control of production which spurred him on to more fundamental efforts at reorganizing work. It is these later efforts that he (p.247) labelled ‘scientific management’ and which others have called simply ‘Taylorism’.

Evidently there was not only a residue of bitter relations, but continuing resistance to Taylor's efforts to speed up work. The immediate problem was ‘the ignorance of the management as to what really constitutes a proper day's work for a workman’ (p. 53).15 But underlying the problem of a fair day's work was a more basic problem of knowledge. The narrative continues: ‘He [Taylor] fully realized that, although he was foreman of the shop, the combined knowledge and skill of the workmen who were under him was certainly ten times as great as his own’ (p. 53).

In our terminology, the underlying problem, as perceived by Taylor, was a knowledge system based on techne. Within it, he could never hope for a decisive victory; as long as his project took the form of a simple appropriation of the worker's techne, success would be at best partial. His vision of total dominance required a thoroughgoing reorganization of the knowledge of production, as the basis for a thoroughgoing reorganization of production itself. Only a recapitulation of workers' knowledge in the form of an episteme to which management alone had access would provide a firm basis for managerial control. Only such a recapitulation would allow management to escape the constraints of subordinating production to the worker's techne, within which the best the manager can do is

frankly to place before the workmen the problem of doing the work in the best and most economical way … inducing each workman to use his best endeavour, his hardest work, all his traditional knowledge, his skill, his ingenuity, and his good will—in a word, his ‘initiative’, so as to yield the largest possible return to his employers (p. 32).

Taylor knew techne all right—he has articulated it in this passage as nearly as one can—and he knew it for the obstacle it was to managerial control. To remove this obstacle, not only must managers get hold of the knowledge of workers, they must change its form. Thus scientific management is not simply appropriation, it is also transformation. Reread the passage from The Principles of Scientific Management which was used in Section 4 to describe the basic idea of Taylorism:

Under scientific management, the managers assume … the burden of gathering (p.248) together all of the traditional knowledge which in the past has been possessed by the workmen and then of classifying, tabulating, and reducing this knowledge to rules, laws, and formulae (p. 36). These replace the judgment of the individual workman (p. 37). Thus all of the planning which under the old system was done by the workman, as a result of his personal experience, must of necessity under the new system be done by management in accordance with the laws of the science (p. 38).

Taylor was sanguine about the consequences of separating execution from control. As was indicated in note 4, he was never the lackey of capital which critics on the left have consistently portrayed him: his suspicions and resentments of capitalists ran as deep as his reservations about workers (Kakar 1970). Rather, he saw scientific management as a third way, a way of making ‘the interests of the workmen and the management … the same, instead of antagonistic’ (pp. 52–3). According to Taylor, and there is every reason to believe he was being sincere,

Scientific management … has for its very foundation the firm conviction that the true interests of employés and employers are one and the same; that prosperity for the employer cannot exist through a long term of years unless it is accompanied by prosperity for the employé, and vice versa; and that it is possible to give the worker what he wants—high wages—and the employer what he wants—a low labour cost—for his manufactures (p. 10).

Taylor paid less attention to the non-material interests of the worker and dealt with these interests in a contradictory fashion. He first suggested that scientific management would actually increase workers' on-the-job satisfaction; but he immediately fell back on the argument that the omelette of progress required a few broken eggs. On the one hand (pp. 120–1),

All of us are grown-up children and it is equally true that the average workman will work with the greatest satisfaction, both to himself and to his employer, when he is given each day a definite task which he is to perform in a given time, and which constitutes a proper day's work for a good workman. This furnishes the workman with a clear-cut standard, by which he can throughout the day measure his own progress, and the accomplishment of which affords him the greatest satisfaction.

On the other hand (p. 125),

Now when through all of this teaching and this minute instruction the work is apparently made so smooth and easy for the workman, the first impression is that this all makes him a mere automaton, a wooden man. As the workmen frequently say when they first come under the system, ‘Why, I am not allowed to think or move without someone interfering or doing it for me!’ The same criticism and objection, however, can be raised against all other modern subdivision of labour.

Taylor's guinea-pigs were less enthusiastic about separating conception and execution. Charles Shartle, a subordinate, later reminisced (p.249)

… he [Taylor] would always say that he had others to think, and we are supposed to do the work. I remember he said to me, many times, I have you for your strength and mechanical ability, and we have other men paid for thinking, and I think he used to try to carry this out pretty well. But I would never admit to him that I was not allowed to think. We used to have some pretty hot arguments just over this point. (Charles Shartle, Recollections, Taylor Collection, Stevens Institute of Technology, Hoboken, New Jersey. Quoted in Kakar 1970: 98–9.)

In the end, Taylor never achieved more than the partial success he enjoyed when working within the framework of techne. Running a factory on episteme alone was a fantasy. Like one of Oliver Sacks's more extremely afflicted patients, a factory run on pure episteme might be grotesque or comical, but it could never function with any semblance of normality.16

9 Numerical Control

But such is the power of ideas that the project of epistemizing production has never been abandoned. David Noble's book, Forces of Production (1984), is a superbly instructive chronicle of scientific management reincarnated in the automation of machine-tool manufacture. Noble's story describes how a particular strategy of automation, numerical control (or N/C) came to dominate in the United States. The other side of Noble's tale is why the alternative, record playback (or R/P) was put aside.

Recast in our terminology, the choice was between a technology based on episteme and one based on techne. Whereas N/C, the epistemic technology, attempted to bypass the skilled machinist altogether, R/P built directly on the skilled machinist's techne. N/C envisioned a direct transition from the blueprint of a part into a series of instructions expressed in mathematical form with the help of a computer, which, once encoded on a magnetic tape or punched paper, were supposed to activate the machine tool to perform an appropriate series of operations. By contrast, R/P prepared the tape by recording the motions of a skilled machinist. Noble describes the difference between the two technologies thus:

Whereas with the motional approach, the skills and tacit knowledge of the machinist were automatically recorded as he interpreted the blueprint and put the machine through its paces manually, without ever having to be formally or explicitly articulated, with N/C, all interpretation was performed by a ‘part programmer,’ at his office desk, who was required to spell out precisely in (p.250) mathematical and algorithmic terms what had heretofore been largely sight, sound, and feel (p. 84).

The merits of N/C in terms of efficiency were in Noble's view dubious, to say the least, though their economic defects were hidden for a long time by the generosity of the United States Air Force in supporting the research and development of N/C technologies. Rather, the appeal of N/C was the fantasy of a production process so thoroughly epistemized that workers, at least skilled workers, would no longer be necessary. His imagination stirred by a demonstration at the Massachusetts Institute of Technology (a prime contractor in the early development of N/C), one early enthusiast wrote to the MIT project leader in early 1952 that N/C ‘signals “our emancipation from human workers”’ (Noble 1984: 235).

Shortly thereafter, in 1954, the trade magazine American Machinist observed more soberly: ‘Numerical control is not a strictly metalworking technique, it is a philosophy of control’ (quoted in Noble 1984: 237–8). In 1976, with two decades of hindsight, Iron Age reached a similar conclusion: ‘The fundamental advantages of numerical control [is that] it brings production control to the Engineering Department’ (p. 238).

This was still part fantasy. The reality, in Noble's view, was that N/C was a costly and cumbersome technology which had little built-in flexibility to cope with the variety of conditions under which production actually takes place. ‘Different materials, temperatures, irregularities in the workpiece, tool-wear, machine malfunction—all of them would affect reproduction accuracy and their final quality’ (Noble 1984: 151). And require, willy-nilly, the intervention of the machinist's techne. Indeed, although, in the words of a production supervisor, ‘The whole purpose of N/C is to remove the operator from the process’ (p. 242), operators had to learn to read the tapes in order to make adjustments for changes in operating conditions. ‘N/Cs are supposed to be like magic,’ observed one operator, ‘but all you can do automatically is produce scrap.’17

Of course R/P technologies were not immune to these problems. The difference between R/P and N/C lay in how the control tapes were prepared rather than in how these tapes were operated.18 For the most part the architects of R/P as well as of N/C technologies had imagined that the system would function with minimal operator intervention, and what little was required could be handled by unskilled workers (Noble (p.251) 1984: 151). In short, the designers of R/P also saw little need to incorporate provision for the operator to ‘override’ the program when conditions warranted.

There was however an essential philosophical difference between these two approaches: the N/C system, based on pure episteme, presupposed the redundancy of the machinist and his techne; the R/P system by contrast built on that techne. Practically, under an R/P technology the skilled machinist would still be available on the shop-floor when his techne was required. The goal of N/C technology was to eliminate the skilled machinist altogether.

It is unfortunate that Nobel could not include more comparative material in his study; the experience of other countries, to judge from the tantalizing hints Noble drops along the way, has been different in ways of direct relevance to our primary concern. In particular, Japan provides a revealing contrast with the American predilection towards complex N/C systems, and the corresponding rejection of simpler R/P systems (see Noble 1984: 169 n. and 182 n.). In a comparative study of the Japanese and American firm, Masahiko Aoki (1988) remarks specifically on this difference; for Aoki it reflects a Japanese attitude towards workers' knowledge fundamentally different from the attitudes that characterize the West.

This brings us to one of the basic questions of this essay: how does the nature and role of knowledge in the production process figure in the odd mixture of resistance and accommodation with which workers have received technical changes that have undermined their autonomy? Resistance is perhaps the easier part of the story. Workers are able to resist successfully because their techne is essential to the production process. Noble's operator spoke volumes when he observed that ‘all you can do automatically [readepistemically”] is produce scrap’. The point is that episteme can never be a self-sufficient system for organizing thought, much less action.19 As Pirsig remarked of ‘classical’ knowledge in Zen and the Art of Motorcycle Maintenance, ‘the motorcycle, so described, is almost impossible to understand unless you already know how one works. The immediate surface expressions that are essential for primary understanding are gone. Only the underlying form is left’ (1976: 71). Practical managers, however disposed they might be to episteme and the control it promises, come eventually to learn its limits. One executive had this to say in 1957 to a meeting of the Electronics Industry Association:

(p.252) In my opinion there is too much talk about giant brains, computer-controlled factories, and the abolition of the factory worker … N/C will never be able to do away with factory workers … experience leads me to believe that the engineer is incapable of doing an efficient job without the know-how of the factory worker (quoted in Noble 1984: 236).

An article in Fortune put it this way: ‘Factory operations may seem orderly enough until you try to describe them in computer programs, then they begin to look quite irregular’ (quoted in Noble 1984: 344).

The limitations of episteme are a running theme of Noble's account of the introduction of N/C technologies. On the shop-floor, the focal point was control over machine speeds. On the one hand, machine speed was, as it were, the cutting edge of managerial control. On the other hand, the variations of conditions made it necessary for managers to permit workers to use their judgement in deciding when to override the computer program. But the capacity to override was at the same time the capacity to ‘pace’ and maintain ‘stints’—the very heart of the traditional system that management had been struggling against since well before Taylor's time. As Noble put it, ‘Management attempts to control the freedom of the work force invariably run up against the contradiction that the freedom is necessary for quality production’ (p. 277).

Accommodation is more difficult to explain. Since time out of mind workers have been concerned with the effects of technical change, beginning with the most obvious question of the effects on employment. But in the end, resistance on the shop-floor is limited to the kind of guerilla struggle that uses overrides to pace production or in the kind of rearguard action that defends job classifications against ‘deskilling’.

It is easy to understand why individual workers have not played an active role in influencing the course of innovation. Isolated from the design process by the very organization of work under capitalism and by their own limitations in the episteme of production, workers are necessarily on the receiving end.

But the same arguments cannot apply to the trade unions. All the same, trade union intervention in the process of technical change has, in the United States at least, long been marked by reticence, hesitation, and timidity—when it occurs at all.

The trade union imagination is limited to protecting the jobs and incomes of the existing labour force, a strategy successfully pursued for many years by John L. Lewis of the United Mine Workers. Unions no more than individual workers have sought a role in shaping the content of technical change.

Noble's account of the unions' role in the automation of machine-tool production fits this general picture. Even when their assistance was actively sought by the developers of ‘Specialmatic’, a rarity as a technology (p.253) more conducive to worker control, the labour unions showed little interest and less enthusiasm. According to Noble, ‘labour unions in the metalworking industries never championed the Specialmatic, or any other potentially labour-oriented technological advance for that matter, leaving such decisions to management alone’ (1984: 95–6). In short, the trade union attitude was that of Harry Bridges, the militant leader of the International Longshoremen, who ‘concluded that the fate of technology was irresistible and that [labour] would have to “adjust” to survive’ (quoted in Noble 1984: 259).

This fatalism, of course, has many roots. It has already been observed that the emergence of ‘business unionism’ in America was itself the outcome of a protracted struggle. But the abdication of trade unions with respect to questions of technology, more precisely the content of technology, cannot in my judgement be understood without reference to the cultural values attached to episteme. In the West episteme is held at the very least to be a superior form of knowledge, if not the only form of knowledge. Moreover this value is widely held. It may serve the interests of capital in legitimizing scientific management (or indeed any enterprise that can lay claim to being ‘scientific’), but it derives its legitimacy from the circumstance that workers as well as capitalists, craftsmen as well as schoolmen, believe in the superiority of episteme over techne. This shared belief, in my view, is a key ingredient in the accommodation of workers and their leaders to the project of capitalist domination of production. If you believe that episteme subsumes techne, it is hard to make a determined defence of your techne.

10 The Marginalization of Techne

The glorification of episteme in Western culture has a long history. The term, along with techne, is of course Greek, but there is much dispute among students of classical Greek civilization over how these terms were used and understood by different Greeks at different times. A formal distinction between the two terms somewhat along present lines is made by Aristotle in the Nichomachean Ethics (1139b14–40a24), but the salient issue is less the precise nature of the distinction between the two than the subordination of the one to the other.20 There is first of course class politics: in so far as techne referred to the craftsman's knowledge of production, it is to be expected that it would be subordinated along with (p.254) the craftsman himself. But if Plato is any guide, the upper-class Greek conception of the craftsman must have been ambivalent: the craftsman figures prominently in the Platonic origin myth, creation itself being the work of a demiurgos, a craftsman; and the craftsman's techne appears and reappears in the Platonic dialogues as the model of purposive knowledge (Klosko 1986: 28, 41; Vidal-Naquet 1983: 293). The ambivalence may stem from a very real tension between the essential role of the craftsman and his knowledge to the well-being of the polis on the one hand and the inferior position of the craftsman on the other (Vidal-Naquet 1983: 289–316).

But more than power politics is at issue: Greek theories of knowledge, Jean-Pierre Vernant has suggested, led to the devaluation of technical knowledge, the artisan's techne, because production involved contamination of pure knowledge, which deals with the unchanging and the certain, by the unpredictable. Unlike episteme, technical knowledge deals with approximation, ‘to which neither exact measure nor precise calculation applies’ (Vernant 1982, ii: 51). Thus

Artisanal techne is not real knowledge. The artisan's … techne rests upon fidelity to a tradition which is not of a scientific order but outside of which would hand him over, disarmed, to chance. Experience can teach him nothing because in the situation in which he finds himself placed—between rational knowledge on the one hand and tuche, chance, on the other—there is for him neither theory nor facts capable of verifying theory; there is no experience in the proper sense. By the strict rules which his art necessitates, he imitates blindly the rigour and sureness of rational procedure; but he has also to adapt himself, thanks to a sort of flair acquired in the practice of his profession, to the unpredictable and the chancy, which the material on which he acts always has in greater or lesser degree (Vernant 1982, ii: 59).

Citing the poet Agathon, Aristotle summarizes the problem succinctly (Nichomachean Ethics, 1140a20): ‘Art [techne] loves chance, and chance art’, and this, if we follow Vernant's interpretation, must lower the status of techne.

The association of techne with chance recalls Pierre Vidal-Naquet's eloquent evocation of the opposition between order and disorder in the evolution of Greek thought and social institutions. Lacking a theory of probability, the Greeks identified chance with disorder, and knowledge of random variability was not knowledge at all (Hacking 1975). One might suggest that for the Greeks episteme was not only the knowledge system of science but the knowledge system as well of social order, and its attractions the attractions of stability. The craftsman and his techne (p.255) represent—to borrow a phrase Vidal-Naquet employs in a different context—‘disorder and the individual exploit’ (1983: 174).21

Worse, techne, certainly the artisan's techne, was bound up with empeiria, experience, and therefore further contaminated by its contact with the concrete and the practical. ‘Empeiria, experience, … is neither experimentation nor experimental thought but practical knowledge obtained by groping (tâtonnements). To the extent it comes more closely into contact with the physically concrete, theory [i.e. episteme] loses its rigour and ceases to be itself. It is not applied to, but degraded in, facts (Vernant 1982: 52). Indeed, Plato appears to use the term empeiria to describe characteristics of craft production that I have described in terms of techne.22 But whatever the names, the distinction between types of knowledge is central to Plato's philosophy. The Republic is categorical about the inferiority of the craftsman's knowledge. Socrates expresses the Platonic view in terms of the relation between the knowledge of the horseman who uses the bit and bridle and the craftsman who makes them. ‘Is it not true,’ Socrates asks, ‘that not even the craftsmen who make them know [how they should be made] but only the horseman who understands their use?’ (601 C). At issue is a difference not between episteme and techne (see footnote 20) but one between episteme and techne on the other hand and ortha doxa (right opinion) on the other. Socrates continues:

It follows, then, that the user must know most about the performance of the thing he uses and must report on its good or bad points to the maker. The flute-player, for example, will tell the instrument-maker how well his flutes serve the player's purpose, and the other will submit to be instructed about how they should be made. So the man who uses any implement will speak of its merits and defects with knowledge, whereas the maker will take his work and possess no more than a correct belief, which he is obliged to obtain by listening to the man who knows (601 D–E).

The terminology may be different, but there is no question that the knowledge of the craftsman is of a different, and inferior, sort.

(p.256) For Aristotle too the craftsman left to his own devices could lay claim only to an inferior grade of knowledge. Indeed Aristotle even takes over the parable of the flute-maker and the flute-player and with Plato stigmatizes the craftsman's knowledge as simply ‘right opinion’ (Politics, 1277b27–30). (But it is for ‘experts in the science of mensuration to elect a land surveyor and for experts in navigation to choose a pilot’ (Politics, 1282a9–10). Foolish consistency is the hobgoblin of little minds.) Aristotle believed that there could be an episteme—albeit an inferior one—of even the slave's work, an episteme for instance of cooking (ibid. 1255b26–32). In this respect, Aristotle is the true precursor of Taylor.

It is evidently too much to assert that the conception of knowledge, and particularly of craft knowledge, held by certain Greek philosophers determined the Western conception for all time to come. In the first place, alternative readings of the Greeks are possible, as modern scholarship has amply demonstrated. For instance, in contrast to the dominant reading of ‘the Greek’ (that is, Plato's and Aristotle's) conception of knowledge as limited to that which is logically derivable from self-evident first principles, which is my notion of episteme, Martha Nussbaum (1986: 290 ff.) has suggested that Aristotle in particular had a much more elastic view. In Nussbaum's interpretation, Aristotle's conception of practical wisdom, the knowledge of life, differs from the episteme of mathematics and natural science precisely in its reliance upon the emotions, experience, and other aspects of what I have assigned to the realm of techne. In Nussbaum's reading, Aristotle assigns practical wisdom to a distinct place from episteme, but it is not an inferior one.

It is significant however that such an interpretation has a relatively recent pedigree, whereas the dominant reading goes back at least to Thomas Aquinas. And the dominant reading, while a matter of interpretation, is not an invention out of whole cloth. The power of episteme in the modern West is hardly conceivable without deep roots in the past.

11 The Handloom Weavers of Nuapatna

Much of the foregoing is necessarily speculative. It deals with a side of history that mainstream historians and social theorists have largely ignored. Those of a Marxist bent, seeing culture as derivative of economic and technological conditions, have, for their part, preferred to see the accommodation of workers to the capitalist project of controlling production in terms of the political and economic power of a dominant class, rather than in terms of a shared cultural heritage that devalues work along with the knowledge of the worker.

Even if there were more solid scholarly foundations to build on, it would be difficult, as long as we stay within the history of work in the (p.257) West, to gain any perspective on what might have been or, for that matter, what still might be. It is for this reason that we now shift gears abruptly, to examine the political economy and cultural basis of work in a traditional, non-Western, setting.

I first visited Nuapatna, in the Indian state of Orissa, during the winter of 1985–6, in search of cross-cultural counterpoint to my historical and theoretical studies of work organization in the West. I was attracted by the possibility of observing on the ground the social organization of an industry—handloom weaving—that in Europe had been consigned to the history books a century and a half ago.

I had little idea of what to expect. My information was limited to the bare fact that Nuapatna had for hundreds of years been a centre of a particular kind of handloom weaving, ikat.

Ikat is a form of tie and dye in which dye is applied to the yarn prior to the actual weaving process, rather than to the finished cloth.23 The origins of this art are lost in the mists of time, and centres of ikat production have been found from South-East Asia24 and Japan to West Africa and South America. Ikat was once produced in many places in India, but apart from a remnant of a few families who still produce ikat fabrics in Gujarat in Western India, only three centres of ikat survive in the subcontinent, two of which are in Orissa, the third in neighbouring Andhra Pradesh.

This is not the place to tell the story of Nuapatna ikat. I have here the more limited aim of examining the basis of its survival and growth, with particular attention to the issues of embeddedness and knowledge systems that have loomed so large in my analysis of the cultural constraints that operate in the West. Accordingly my description of the political economy of Nuapatna will be kept to a minimum consistent with providing the necessary context.

Nuapatna, located at 86° E, 20° N, had a population of 3,500 according to the census of 1971, the latest from which detailed data are available. Of this number, 650, two-thirds of the paid labour force, are listed in the census as engaged in household industry, and most of these people would be weavers. A round number of 500 would probably not be far off for the population of active weavers, but that would exclude a large number of ancillary workers, women and children, and is better interpreted as the number of looms rather than the number of those involved in weaving. Moreover, these numbers refer to Nuapatna proper. The adjoining villages (p.258) also have their share of weavers, and all in all 1,000 looms would be at least the right order of magnitude.

Nuapatna's population, though probably not the number of looms, has increased substantially since 1971, but that gets us ahead of the story. It is the largest village in Tigria tahsil (the tahsil is the smallest unit in the administrative hierarchy) but it is not the tahsil headquarters. That honour belongs to the village of Tigria, seat of the tiny princely state which bore this name in the pre-Independence period. Nuapatna is situated two miles from the Mahanadi, the great river that divides the state of Orissa in two, still not spanned by a bridge for hundreds of miles upstream of Cuttack, the large commercial centre near its mouth in the Bay of Bengal.25

The first reference to weaving in Nuapatna dates from the reign of Ranchandradev II, ruler of the kingdom of Puri in the early eighteenth century and suzerain of the Raja of Tigria and many other such minor rulers. A document in the temple archives—Puri is a pilgrimage centre and site of the great temple of Jagannātha—records an order of the Puri king concerning the production of Gītagovindakhanduā. By order of the king, the production of Gītagovindakhanduā is to be transferred from the village of Kenduli, the birthplace of the author of the Gīta Govinda, to Nuapatna, specifically to ‘eight brother weavers … who were to be remunerated for their labour … by way of Bhoga-praśād [food offered to the temple deities and then distributed to the people] and other gifts’ (Mohanty and Krishna 1974: 20).26

A mid-eighteenth-century British visitor, one Thomas Motte, also makes reference to weaving in Nuapatna. Motte recounts that the Raja of Tigria invited weavers who were victims of the depredations of a neighbouring raja to settle in Nuapatna (Mohanty and Krishna 1974: 20).

Interestingly, a gazetteer of the princely states of Orissa, written at the turn of the present century by a member of the Indian Civil Service, L. E. B. Cobden-Ramsay, makes no reference whatsoever to Nuapatna or to ikat. This absence is all the more striking since under the heading ‘occupations, manufactures, and trade’ other places, including a nearby village in the adjoining princely state of Baramba, are mentioned as centres of weaving of high-quality cloth and design (Cobden-Ramsay 1982: 81, 322–3).

(p.259) This may well indicate the low state into which Nuapatna weaving had fallen in the latter days of the Raj.27 The general view among the older weavers of Nuapatna, certainly, is that their craft had fallen on hard times during their youth. To be sure, the tradition of weaving for the Jagannātha temple never died. Moreover, elaborate ikat saris survive from the early part of the century, and there was apparently a limited market for silamajoḍa, a garment of raw silk traditionally worn by the groom at Hindu weddings. But most of the weavers were reduced to producing for the wage-goods market, in which demand was limited to coarse (20-count28 and lower) fabrics with little ornamentation.

Local people credit the late Arjun Subuddhi for the revival of Nuapatna ikat. Arjun Subuddhi, whose surname is an honorific granted by the Raja of Tigria for his service to the state as a tax-collector and for his proficiency as a musician, certainly lived up to his name (su = good and buddhi = intellect). Born in 1905 to a weaving caste family who had established themselves as putters-out, Subuddhi was trained in the traditional art of his caste, and according to his son (who provided most of the details of his father's life to my collaborator, Puma Chandra Mishra), he was himself a proficient weaver. But his life's work lay elsewhere. In addition to his duties as tax-collector, he took part in the family putting-out business, and in travelling to Cuttack to sell Nuapatna cloth—today an hour and a half by car on a hard-surfaced road, but before Independence a day and a half away—he acquired a knowledge of the market that he later put to use not only for his own benefit but for the benefit of the entire village.

One anecdote recounted by his son will convey something of Arjun Subuddhi's business talent. Seeing that middlemen and brokers, referred to by Subuddhi's son as ‘rich businessmen of Cuttack’, were taking an inordinate share of the profit, he contrived to learn the names and addresses of the people in Calcutta and Bombay with whom they dealt. He was thereby able to establish direct connections with Bombay and Calcutta businessmen and began to send cloth directly to these distant merchants by post. He also had sample books made up and distributed.

Through his business travels, Arjun Subuddhi became aware of the (p.260) popularity of Sambalpuri fabrics, Sambalpur being the other (and more renowned) centre of ikat in Orissa. Around the time of Independence he sojourned in the Sambalpur area and, a weaver himself, was able to assess the problems of learning the Sambalpuri technique and design.

He took three weavers with him on one of these trips, that they might learn at least the rudiments of the Sambalpuri art. Returning to Nuapatna, Subuddhi set the three (one of whom was still alive during my time in Nuapatna but too feeble to be interviewed) to work on Sambalpuri designs, advancing materials and guaranteeing to purchase the products even if there were mistakes in them that rendered them unfit for market. He would bear the risks.

The three gradually mastered the Sambalpuri art and the process of diffusion began. Once again Arjun Subuddhi provided encouragement, both economic and psychological, but according to many accounts the weavers were reluctant to experiment. Reportedly, the weavers feared that they would not become proficient enough to turn out quality goods and that even if they did, the goods would not find markets.

The difficulty with this explanation of the weavers' reluctance to adopt the new designs and techniques is that it was not their own resources but Arjun Subuddhi's that were on the table. For the weavers, the problem might have been some combination of identification with Arjun Subuddhi, concern for their own reputations and sense of self, and, finally, a fear of the unknown. These are not variables with which economists are comfortable, but having myself attempted a relatively minor marketing innovation in Nuapatna, I am disposed to believe the accounts of an unfocused reluctance.

In any event, the obstacles were overcome and the economy of Nuapatna was transformed. Ikat weaving became quite general, to the point that Nuapatna has come to rival Sambalpur as a centre of this art.

How much of the credit is due Arjun Subuddhi? A fair answer must allow for two factors which might diminish Arjun Subuddhi's role. The first is a cultural tendency towards cults of personality, the tendency to personify what might in the West be described in terms of an impersonal process of technological innovation and diffusion (cf. Chapter 3). The second is that at about the same time, other actors, chiefly governmental, were getting into the picture, and it is hard to sort out the contributions of all these agents of change. None the less, when all is said and done, I am persuaded of the general accuracy of the account I have rendered and its stress on Arjun Subuddhi's role.

A more difficult problem is to identify with any precision the nature of Arjun Subuddhi's innovation. The problem is not to identify Subuddhi's legacy, but to discover the situation before he transformed it. There is general agreement that Nuapatna weavers had fallen on hard times and that the majority of weavers produced coarse cotton cloth for a mass (p.261) market. What is in doubt is the extent to which the minority practised the craft of ikat. It is probable that fine weaving included a great deal of embroidery, especially on silamajoḍa. It still does, particularly on the borders of saris, but the effect is now achieved mechanically by means of dobbies on the loom whereas earlier it was a much more time-consuming, hand process.

But many people, including Arjun Subuddhi's son, assert much more, to wit, that Arjun Subuddhi introduced ikat to Nuapatna. On the face of it this assertion is implausible. Whatever their origin, ikat Gītagovindakhanduā survive from the late nineteenth century, as do ikat saris (Mohanty and Krishna 1974). If not a total novelty, however, it is likely that before Subuddhi, ikat was limited to a small number of weavers who worked in silk and produced either for the Puri temple or for the luxury market; and that apart from the expense of the fabric, the nature of the production process made it too labour-intensive and therefore too costly for ordinary consumers (even with weavers receiving the low wages which prevailed, and still prevail, in the handloom industry).

The technical difficulty was that the tie and dye process was applied to the weft, and therefore each pic had to be adjusted individually, a time-consuming and laborious process. The secret of the success of Sambalpur was that the tie and dye was concentrated on the warp, particularly on the border of the sari. According to the elaborateness of the design, some weft tie and dye might still be incorporated into the design, particularly at one end of the sari, called the pallav or anchal; the Sambalpuri method allowed considerable latitude in this matter. The ikat border provided a pleasing effect—and at a sufficiently low cost that the technique could be employed on cotton cloth destined for the mass market.

Thus the frequently heard claim that Arjun Subuddhi brought ikat to Nuapatna is probably correct up to a point. It is likely that he introduced warp ikat and made it possible for cotton weavers to utilize tie and dye yarns in their work. This was an enormous achievement, and it is no wonder that Arjun Subuddhi is revered throughout the weaving community.

Arjun Subuddhi died in 1961, but even before his death a second innovation was introduced that was to have a profound impact on the political economy of weaving. In 1955, the first co-operative society began operation, with Arjun Subuddhi as its president. Co-operatives in name only, the societies (the English word has been taken over into Oriya) are in fact modern-day putters-out, advancing materials to the weavers and frequently stipulating the design as well. Even the terminology is more capitalist than co-operative: the weaver gets majuri (labour charge) and the society gets the profit. Four societies now dominate the scene in Nuapatna, but they have not completely displaced the private-sector (p.262) putters-out. Traditional putters-out, locally called mahājans29 still flourish, and there are even a few small, private handloom factories in operation.30

The Nuapatna co-operatives compete with the mahājans, but it is a symbiotic rather than a rivalrous competition. Arjun Subuddhi carried on his private business while he was president of the No. 1 Handloom Weavers' Co-operative Society, and the tradition of mahājans holding office in the societies continues.

Although the overwhelming majority are independent in the sense that they own their own looms and weave in their own houses, I have never encountered nor even heard of a completely self-sufficient, independent weaver who provides all his own working capital. However, all other conceivable combinations seem to exist. It is not unusual for a father and son living in the same household to be members of different co-operative societies, or for a weaver of a batch of saris for the co-operative society to sell a few of the batch on his own account, refunding the cost of materials to the society. He might also piggy-back a few additional saris, for which he supplies the working capital, on to a batch produced for the co-operative society. One weaver of my acquaintance sold a sari to a local mahājan when the co-operative society, strapped for funds, could not pay him the labour charge.

The scales would appear to be loaded against the mahājan. The co-operative societies have access to subsidized credit and are the beneficiaries of other subsidies as well. But the mahājan has one great advantage. The co-operative societies are generally barred from providing credit to their members except for productive purposes and this takes the form of advances of materials only. The mahājan operates under no such restriction, and loans extended for anything from marriages and funeral expenses to house construction are an important aspect of the complex connections between mahājans and weavers.

It is not easy to explain the dependence of the weavers on the co-operative societies and the mahājans. A purely economic explanation has some power: the putter-out not only supplies credit, but keeps abreast of the market, a function that is particularly important in the market for saris, where, as in other clothing markets, fashion predominates. As early as 1613, Peter Floris, the principal merchant on the Globe, which undertook the seventh voyage of the English East India Company, learned the hard way the importance of judging the market correctly.31 This excerpt from Floris's journal describes a costly error of judgement: (p.263)

Butt a greate oversight hath bene committed in the bespeaking of the foresayd Maleys cloth, to witte, the pattas, dragans, salalus and theyr sortes, for they have all of theym, to witte, the Petapaolishe cl[o]ath, a little narrowe white edge, and the upright [proper] Maleys cl[o]ath muste bee withoute it, as the cl[o]ath of Paleacatte was; wherein those of Maleys are so curious [i.e. particular] that they will not once putte foorth theyr handes to looke uppon theym; and, yf I had not nowe founde it by experience, I had never believed it, that so small a faulte should cause so greate an abatement in the pryce … (Floris 1934: 71).

However, the role of the putter-out in interpreting the state of fashion does not fall from heaven but from the organization of marketing. There is virtually no local wholesale market for Nuapatna fabrics despite the possibilities that exist. In the first place there is a twice-weekly market to which local traders come to sell their wares. This market is heavily patronized by local people; it is the main source of fresh vegetables and a source as well of staple foods, rice and pulses. There is even a small section of the market devoted to cloth. There are also two or three permanent shops on the main road, as well as the co-operative societies, which deal exclusively in cloth goods.

But all this operates on the retail level. In the absence of a wholesale market regularly visited by outside traders, the individual weaver has no way, even if he were so disposed, of informing himself about the course of fashion and dealing with a thick (as opposed to thin) market, of selling on a regular basis the fruit of his labour rather than his labour power.

Not that the weaver is typically so disposed. There are, or can be, significant personal ties that bind the mahājan and the weaver to each other, as I have seen not only in Nuapatna, but in other weaving centres as well. It would be a mistake to reduce this relationship to an economic one.32

Even when the personal tie is much diminished, as is the case when the mahājan becomes the co-operative society, the putter-out plays a role that goes beyond the narrowly economic. Over several months in early 1987 I attempted to establish a direct connection between several weaving families and Central Cottage Industries, a Government of India (p.264) undertaking that has had considerable success in marketing handicrafts to affluent tourists and Indians alike in the chief cities. Once again, this is not the place for that story. Suffice it to say that for the most part my efforts failed. Nominally the inability of CCI representatives to provide an advance for materials was the sticking point; the weavers frequently complained of their poverty and attributed their reluctance to execute orders from the CCI to their shortage of working capital. But in my view the obstacle was not so much poverty, illiquidity, or the extra risks involved in working without an advance. It was the unfamiliarity of an unconventional mode, the same reluctance that made their fathers hesitate to adopt Arjun Subuddhi's innovations. The advance provided by the mahājan or the co-operative society plays an economic role of course; but it is also a symbolic guarantee of continuity and reciprocity.

Whatever its sources and reasons, the dominance of putters-out is significant. Whether it be the East India Company or Arjun Subuddhi or the co-operative societies providing the designs, for the producer the union of conception and execution is ruptured. A senior official of the state government with many years experience of the handloom industry, whom I interviewed in January 1986, was totally pessimistic: ‘In Nuapatna [he said] the weavers used to weave for pleasure as well as for bread, but now it is just for bread. The aesthetic touch is lost … Now his creative talent is lost, and his work has become like a carbon paper in a typewriter.’

This is very likely an overstatement. In the first place the halcyon days were never so halcyon. Those who produced 20-count plain saris were hardly in a position to exercise their creative talents. More important, a measure of artistic independence and creativity is present even today. The relationship of the putter-out to the weaver is certainly hierarchial, but it is at the same time a personal rather than an impersonal one (see Chapter 3 above). There is considerable room for give and take between the weaver and the putter-out, especially for experienced, accomplished, and senior weavers. There is also room for improvisation. The most skilled among the Nuapatna weavers are clearly artists—and regard themselves as such. Listen to one of them, Narasingha Paramanik, in a conversation with Purna Chandra Mishra in the spring of 1986:

Let my Suresh [Paramanik's son, then a young man of seventeen and already an accomplished weaver] learn everything that I know and then at least my soul will be happy to take rest. This is what I want—let him learn everything and then let him do what he feels like doing. [Suresh had, earlier on in the interview, been reported by his father as being more interested in money than in art.] I mean to say let this art live, let it not be sold.

The co-operative society says that I should give them these beautiful designs. ‘We will decorate the society.’ But I do not; they are my creations and I do not like to sell these designs and to earn money in this way. I do not like it.

(p.265) If you weave a design and that goes to market [Mishra asks], other weavers will be able to learn it, isn't that so?

Let them learn; this way I will give more exercise and additional tasks to the weavers. Only the weavers who have a good brain can copy the design, but when I see the designs being copied, as the primary man I will change and modify them. This is a khela [play or game]—this is what good weavers feel and need. I would like to stay in this art—playing, playing, thus finishing life, coming to the end to die. But before my death I want to see how my son plays.

Besides a measure of artistic independence, the Nuapatna weaver enjoys the considerable flexibility that attaches to cottage industry. He is free to begin and end work as he pleases, to determine the intensity and pace of work, in short, to set his own rhythm. I have seen at first hand varied uses to which this flexibility is put: on one visit to Nuapatna I was frustrated by my inability to interview Narasingha Paramanik because he was attending a week-long reading of a religious text in a local Hindu temple. On another visit, Dharmananda Sahu was away from the village altogether at his son-in-law's house in another part of the state. Another time, Arjuna Patra was unavailable, attending to preparations for the marriage of one of his sons.

The scope for creativity and the scope for flexibility play a role in the vitality of Nuapatna weaving. This is certainly part of the ‘compensating differential’ that economists invoke, dens ex machina, when individuals fail to respond to pecuniary incentives. For it is clear that the attractions of weaving are not financial. Nobody can say with certainty how much the weaver earns per hour or day of work—not even the weaver himself. For one thing the irregularity of his work—the flip side of the flexibility that makes weaving attractive—makes it difficult to measure productivity, and productivity measures are essential to estimate earnings since earnings are by the piece. For another, the preparatory work (talikāma) is carried out a bit at a time rather than at a single stretch, frequently by women, who do this work when they have free time between other household chores; or else it is done by children and old folks, whose rhythms make comparison with the labour time of adult males exceedingly difficult. But when all is said and done, I judge the range of earnings for ‘full-time’ adult males in 1986–7 to have been from Rs 10 to Rs 20 per day of eight hours, with most of the weavers crowded down towards the bottom end of the scale. At Rs 1.00 = $0.08, Rs 10 is not a high daily wage, not even by Nuapatna standards. Even if differences in the cost of living are taken into account—rice, the staple food, at that time cost from Rs 3 to Rs 3.50 ($0.25 to $0.30) per kilogram—this remains a low wage. By comparison, agricultural labourers reportedly earn Rs 10 for a five-hour day and Rs 15 for eight hours, although the relevance of this comparison is diminished by the seasonal nature of agriculture in Nuapatna.

More relevant is the possibility of factory work. On 30 October 1984, (p.266) Indira Gandhi inaugurated the J-Spin mill at Nuapatna, her last public act before being assassinated by her own security guards. J-Spin, short for Shri Jagannath Weavers' Co-operative Spinning Mill, is once again a story unto itself, and here I shall skip over most of what does not impinge directly on present concerns. J-Spin, like many public sector enterprises, was created for a number of purposes: rural employment and an assured supply of high-count cotton yarns to the co-operative sector at ‘reasonable’ prices are the official reasons, but it is hardly coincidental that the mill is situated in the constituency of the Chief Minister of Orissa, J. B. Patnaik.

The main point for present purposes is that J-Spin, carrying 750 people on its rolls in order to assure the presence of 550 in its three-shift operation, paid a minimum of Rs 12.50 per day (as of spring 1987—a rise to Rs 13.25 retroactive to the first of the year was in the works as I was leaving in April), plus fringe benefits (bonus, vacation pay, pension fund, medical insurance) that on my calculations are worth approximately 25 per cent of one's base pay. Many operatives earn substantially more. Piecers, for example, earned Rs 17.50 per day, and reelers (mostly women who are paid piece rates) reportedly averaged Rs 15–16 per day, with some earning as much as Rs 25. None of this work requires more than a few months' on-the-job training.

It is not particularly desirable work—income and security apart. The noise and the dust are oppressive, as was the heat on the day of my first visit in January 1986. (The heat becomes less of a problem, at least relatively, in March and April, when outside temperatures climb to 40°C.) The work is itself boring and the routine stultifying.

Blake's ‘satanic mills’ spring easily to mind; apart from neon lights, J-Spin hardly seems an improvement over nineteenth-century English spinning mills. And even here, the gain is not without its corresponding cost: artificial lighting obviates any necessity of windows, and one cannot in fact see the open sky from inside the mill. One mill executive, a particularly perceptive as well as open individual, told me, ‘people feel it to be like a big prison’.33

But there is nevertheless no problem recruiting workers: mill wages and employment security are not easily come by in India, particularly in rural Orissa. Recruitment is through a cumbersome process in which the local employment exchange supplies names to the mill; new engagements are then vetted by a committee of local notables whose function, it was explained to me, is to ensure an equitable distribution of the available jobs within the Nuapatna area, a special concern to the Chief Minister, in whose parliamentary constituency the mill lies.

(p.267) Local people are more sceptical of the committee's purpose. The going price of a job at the mill, I heard on more than one occasion, is upwards of Rs 1,000. In any case, the intricacies of hiring procedures were more important when the mill first went into operation than now. Precious few openings come up at the mill, and, shortly before my departure from India in the spring of 1987, I was told by a member of the committee of notables that no new workers had been hired in the last year, despite the ‘voluntary’ departure of 41 leaders of an unsuccessful strike over the position of temporary workers in the summer of 1986.

But one fact stands out. Practically no weavers work at the mill, and the few that do appear to be there for very special reasons, as in the case of one young woman in her twenties I interviewed who still lived at home in a society where for women of her class marriage was the well-nigh universal rule. Had her husband sent her home? Was she unable to bear children and therefore unmarriageable? The questions were too delicate to put.

As has been observed, economists will have no trouble digesting the reluctance of weavers to enter the mill: the standard explanation is that the income differential simply does not compensate for the added onerousness of factory work. I have no desire to pick a quarrel, especially since this explanation undoubtedly does capture some of what is at issue; the independence and flexibility on the one hand and the regimentation and physical discomfort on the other undoubtedly contribute to the weavers' willingness to forego the economic gains available in the factory.

But I think there is much more to the story. The weaver's work is embedded in the very fabric of his being in ways that are scarcely comprehensible to Westerners of this day and age. And his techne is on a more secure footing culturally than is the Western craftsman's. These considerations at the very least reinforce the pecuniary and non-pecuniary utility the weaver derives from his work. Arguably, these considerations dominate and make a utility calculus superfluous.

12 The Embeddedness of the Weaver's Craft

Let me take up the issue of embeddedness first. It has been noted that the earliest reference to Nuapatna weaving is from a temple document going back to the early eighteenth century. The connection continues to this day, and the weaving of Gītagovindakhanduā is as much a religious as an economic act, one moreover that is ritually loaded.

The weaver of these temple garments must take special care that the work is perfectly executed, and even more care not to pollute cloth destined for Lord Jagannātha or his brother or sister. His diet is restricted (p.268) to pure foods, to one meal of rice, boiled vegetables, and ghee (clarified butter); meat, fish, and vegetable oil are specifically forbidden. He may not chew betel nut or talk while weaving for fear of polluting the cloth with his saliva. The weaver cannot have his hair cut or his face shaven, again for fear of pollution. He must abstain from sexual intercourse. He cannot weave the Gītagovindakhanduā if a birth or death occurs in the house—even a birth, although auspicious, involves the household in ritual impurities (Apffel Marglin 1985). Special clothes are worn during the weaving, clothes that are kept clean and therefore free from pollution.

Nobody else touches the cloth or the loom, and even the weaver himself avoids touching the frame of the loom with his feet (the pedal excepted). Small children may be ritually contaminated by bodily excretions as well as contaminating their mothers and other caretakers, so both must be kept away from the loom. Menstruating women are particularly polluted, and to be on the safe side all women are kept away from the loom and the tantaghara (the room in which the loom sits, generally the room facing the street). Even the shadow of a child or woman should not fall on the loom.34

I cannot judge how far these traditional procedures and restrictions are actually followed. One informant, an old man who earlier had been a traditional supplier, suggested that the decline of the Puri temple had affected the weaving of Gītagovindakhanduā.35 According to this man, the weaving of these sacred garments has become a business like any other: ‘some big businessmen get [an] order from the temple administrator and come here and place the order with the co-operative society. The society gives this order not to us, [but] to some people of their own and they do not take care.’

On the other hand, my own attempts to place an order for Gītagovindakhanduā, as well as similar attempts of Central Cottage Industries, were repeatedly frustrated. There was great reluctance to take on the responsibility for providing Gītagovindakhanduā for any other purposes than the traditional temple ones. The tradition may be at risk, but it is hardly a thing of the past.

Everyday weaving is less ritually loaded, but it is still deeply embedded in the weaver's definition of himself and the meaning of his life. A religious attitude is pervasive. Most weavers stressed the importance of beginning the work day with a purificatory bath. A brief ceremony of (p.269) worship precedes the actual commencement of work. The post (or khuṇṭa) on the right side of the loom, which bears the weight of the loom, is worshipped as the seat of the god Visvakarma (who is worshipped in many crafts as the god of the tools of the trade). The Visvakarmakhuṇṭa is invariably made of wood of the neem tree, the same wood that is used for the images of Jagannātha and his brother and sister in the Puri temple. By contrast, the rest of the loom can be made of any wood, and the choice is dictated by price and quality. The Visvakarmakhuṇṭa is normally decorated with a bānā (a narrow strip of cloth) which represents the first offering of a new year or a new project. As new offerings are put in place, old ones are taken by an older member of the household to a pond, river, or tank (artificial pond) and put into the water.

Rules of purity and pollution are in force, but in an attenuated form relative to the weaving of Gītagovindakhanduā. For instance, menstruating women are forbidden to enter the tantaghara or to touch the loom and its ancillary equipment for a week. And though even untouchables can enter the tantaghara, visitors must take care not to touch the loom, especially with their feet.

Weaving is nominally a seven-day per week operation, but we have seen that work is frequently interrupted for other pursuits. It is also punctuated by holidays on which work stops altogether. Of special interest is the festival of aṇukūla (good beginning). At the time of the aṇukūla, unlike other holidays, it is imperative that some cloth be woven, for this is essential to the ritual of ‘good beginning’ The timing of the ritual differs according to caste (there are four weaving castes in Nuapatna), but the ritual itself appears to follow the same general form. The first step is a thorough cleaning of the tantaghara and of the loom itself. The Visvakarmakhuṇṭa is disengaged, so that the loom is at rest. During the day no work is done, and in the evening there is a ceremony of worship of the loom (tantapūjā) conducted by the head of the family (rather than by a Brahmin priest). On the next day, some cloth is woven. The newly woven cloth is immediately sold, traditionally to a mahājan, but the mahājan has in recent times been displaced by the co-operative society.

In December 1987 Purna Chandra Mishra attended the pūjā in the house of a leading weaver of Nuapatna, a collateral descendent of Arjun Subuddhi. His description of the pūjā includes offerings of food, water, flowers, and incense to the loom, particularly to the Visvakarmakhuṇṭa. The pūjā also included a reading from the caste purāṇ a,36 which recounts the origin of weaving and weavers and, seemingly unrelated, the difficulties of the original weaver with the original mahājan. A part of the ritual symbolically re-enacts these difficulties: a new bānā is reduced to ashes (p.270) over the flame of an oil lamp. A little oil is added to the ashes to make a paste, and the paste is applied to the face of a straw mahājan, sitting on a straw horse. The straw mahājan, his face blackened by the paste, is a silent witness to the pūjā. At the end of the pūjā, there is general rejoicing. The flame from the bānā burns big and bright, a portent of prosperity for the house of Viswanath and Bansidhar Das. Praśād, the food first offered to the gods, is distributed to family and friends.

It is inauspicious to see the blackened face of the bad mahājan after the ceremony. Accordingly, the next day, the head of the family will dispose of the figure of the mahājan on horseback in the village tank (the artificial pond that serves as a place for people and animals to bathe, and for the washing of clothes).

A weaver whom Mishra had interviewed the previous year interpreted the ritual blackening of the mahājan's face as a representation of the result of a bargain in which the weaver is unsatisfied with the price. ‘The weaver … becomes angry and … sets fire to a piece of cloth, and by its ashes the face of the mahājan is disfigured.’ This view is consistent with the sequel. The cloth woven the next day as part of the aṇukūla is sold to the mahājan or co-operative society for whom the weaver customarily works. The putter-out buys the cloth and adds gifts of food and clothing to the purchase price (a dhoti, the aṇukūla dhoti, was specifically mentioned by one weaver, but I do not know how common it is). All accounts stressed the importance of the money being paid over ‘quickly’, that is, without hesitation or quibble.

The purāṇa of one of the weaver castes, the Asini Pataras, tells the following story. The first weaver, Siva Das, is in debt to a mahājan and cannot repay the loan. Significantly, he has borrowed money to perform a proper pūjā at the festival of pauṣa māsa sukla pakṣa nabamī, the ninth day of the waxing moon in the lunar month of Pausa (December/January), on which the aṇukūla falls. The mahājan is repeatedly put off by Siva Das when he comes for his money. A year passes, and another nabamī festival claims the money Siva Das might have used to pay off the loan. Finally the mahājan proposes to foreclose on the bundle of cloth which has been left as security for the loan. Opening the bundle, the mahājan exclaims, ‘Oh, how beautiful—what is the price?’ Siva Das, encouraged by Visvakarma, replies, ‘I will not sell it for less than Rs 500. I would sooner throw it into the fire.’ The mahājan responds, ‘Then keep your cloth and pay back my money.’ Siva Das repays the loan in kind, saying, ‘Take as much of this cidr bastra [old and torn cloth] as is required to pay off the loan.’ This done, he throws the first bundle into the fire, and does pūjā to the family god (Iṣṭadeva).

It requires neither deep analysis nor arcane methodology to see how the religious and the economic meet in ritual and myth. Relations with the mahājans (old or new) are a constant concern of the weaver. Ragha (p.271) Rana, the weaver who had diverted saris from the co-operative society to a mahājan because the society was illiquid and could not pay the labour charge, remarked pointedly that the mahājan had paid at once. The conflictual part of the ritual is more ambiguous. Ostensibly an assertion of the weaver's power υis-à-υis the mahājan, it is more likely to be the assertion of an ideal little realized in practice than a statement of fact, present or past.

The myth is curious. The purāṇa depicts the weaver as improvident, the mahājan as patient and even accommodating. No Shylock he! The virtue of the weaver lies in his devotion to the gods and to his craft. Shame before the mahājan does not deter him from devoting his limited funds to the nabamī festival, funds which he might have used to repay his debt. The claims of men cannot be put ahead of the claims of the gods. Nor will he bargain about the worth of his art. Sooner let his creation be consumed by the fire than let it be the object of haggling.

The first part of this purāṇa reinforces an image of the weaver which in Western terms could at best be regarded as ambivalent, but which for the weavers themselves, whom I questioned extensively on this subject, is clearly positive. This part of the purāṇa recounts the invention of cloth. The gods, finding their nakedness ugly (not shameful, as did Adam and Eve but simply ugly) first contrive to steal the original cotton seed from a demon who has it firmly implanted in his forehead. But Kapisura, the demon, has been given the boon that ‘he will not be killed in the house or outside, nor in daylight or at night, nor by any weapon. Nor will he burn or drown.’ He must be tricked, and tricked he is.37 Krishna takes the form of lice and infests the demon's hair. Chakradhara (Vishnu with the discus) appears in the form of a kāminī (lustful woman) and offers to remove the lice that infest Kapisura's hair. On the pretext that the light is too faint in the house, it being twilight (and thus neither daylight nor night), the kāminī coaxes the demon out under the eaves of the house (neither outside nor inside) and Krishna, taking his own form, breaks the demon's head in two with his bare hands (no weapon is involved, and the demon neither burns nor drowns).38

After the cotton seed has been liberated from the demon, it is planted. The gods, attracted by the beauty of the cotton plants in bloom, pluck all the flowers. Once again a trick is employed to make them desist so that the flower might develop into fruit. Brahma tells the cultivators to smear a little blood on the flowers: the gods will not pluck flowers so (p.272) polluted. This done, the plants bear fruit. When the cotton is harvested and men are set to work spinning yarn, the weaver enters the picture.

Brahma, who has masterminded the operation from seed to yarn, turns over further responsibility to Vishnu. From a lump of earth Vishnu creates an image and gives life to the image; thus are created Siva Das, the first weaver, and Swati Devi, his wife. Siva Das is willing but ignorant. ‘What is our work?’ he asks. ‘What are we to do?’ Vishnu tells him and then leads him step by step through the preparatory processes of weaving. Siva Das is particularly troubled by the step at which the yarn is starched, since the customary starch is water in which rice has been cooked. ‘How will we avoid the pollution of the rice water?’ Siva Das asks Vishnu. ‘Wind the starched yarn on the spindle in the sun,’ Vishnu counsels the weaver, ‘and the pollution will go away.’

Then the gods, in heavenly convocation, ask Vishnu how the cloth will actually be woven. Vishnu names all the parts of the loom, and the gods volunteer to become these parts. ‘I will become the khila [the peg which holds the loom together],’ says Angira. ‘And I’, says Narada, ‘the dungi [needle].’ Brahma offers himself as the bharā [treadle]. And so on down the list.

But, Vishnu observes, as matters stand the loom is useless. Someone must bear the load and support the loom. ‘You know best,’ responds Brahma. But, says Vishnu, no one will bear the load and support the entire loom unless he is given the first cloth. The other gods reply, so be it. Then Vishnu says, ‘I will come as Visvakarma Vasudev and will bear the load. I will give the necessary support.’ Hearing this, the other gods bow down to Vishnu, and Vishnu orders Siva Das to begin weaving.

Siva Das inquires as to the exact day he is to make the loom ready and do the pūjā. Vishnu tells him that he should do the pūjā on the ninth day (nabamī) of the bright fortnight (sukla pakṣa) of the month of Pausa and on the tenth day he should begin to weave (tantabuna aṇukūla).

So it was done; the first fruits of the loom were duly offered to Visvakarma Vasudev, and subsequent products to the other gods. The gods were overjoyed and, wishing to reward the weaver, offered him a boon of his choice. ‘Let me be given the boon that I earn enough in one day's weaving to eat throughout the year,’ Siva Das said. And this was done too.

When Siva Das returns home, his wife asks him what had happened. He tells Swati Devi about the boon. Far from being pleased, she says, ‘What have you done? It is not good. Our children will be undisciplined (ayogya). Go back and tell the gods you have made a mistake. Ask for another boon. Say

  • Niti araji, niti khāibi
  • Jalāguḍi bāṭe, bela cāhinbi
  • (I will earn everyday, I will eat everyday
  • I will see the passage of time through the window).’

(p.273) Siva Das does as his wife bids him, and the gods grant him the new boon. ‘Both Siva and his wife are glad. The couple spend their time pleasantly.’

This ends the story of the creation. But at the same time it sets the stage for the tale of the weaver and the mahājan. The month of Pausa is about to come and Siva Das realizes that he has no money to prepare a proper pūjā for nabamī. He remembers the bundle of cloth and sets off for the house of the mahājan. The second part of the purāṇa, which has already been summarized, unfolds.

What must be remembered in interpreting this story is that the Nuapatna weaver, like most Indians, is much closer to his or her origin myth than Westerners are to theirs. Even ‘believing’ Christians and Jews, fundamentalists apart, are taught from an early age to take their biblical myths with a grain of salt. I can remember one of my children explaining to me at the age of six or seven that the parting of the Red Sea was ‘just a story’ and that it was ‘really’ tidal action rather than the hand of the Lord at work. Traditional Indians make no such distinction. The mythic and the real are not separate domains (cf. Chapter 4). The parts of the loom are not metaphorical gods or representations of gods, they are gods. And Visvakarma supports the loom today as he did for Siva Das. The concerns of Siva Das—why am I here? how do I avoid pollution? who takes precedence, the mahājan or the Iṣṭadeυa?—are equally the concerns of today's weaver.

And the answers are to be found in the purāṇa. The weaver works for his daily bread, to be sure, but he was put on this earth to add to the glory of the gods, not simply to make a living. And pollution? Pollution is to be surmounted, but it cannot be avoided entirely. Why, without pollution, the cotton bloom would never mature into fruit and the yarn could not be starched.

In this light, the ‘boon’ of working every day to eat that day is of particular interest. In one sense it is highly negative, very much like the Pauline injunction ‘if anyone will not work, let him not eat’: its chief virtue is to avoid the sloth and dissolution to which idleness must lead, the ayogya or indiscipline that Swati Devi fears for her children if they need work only one day per year. But in another sense, the contrast with both the Greek and Judaeo-Christian origin myths is striking: the by-product of conflict among the gods in the one case, a just punishment for man's transgressions against the divine in the second. For the Hindu weaver labour is a boon. The contrast with the Sermon on the Mount is also striking: the Hindu weaver is not to expect to be fed like the birds or clothed like the lilies, for God helps him who helps himself. However, the attitude towards time, calculation, and accumulation of worldly goods does recall the Sermon on the Mount: one is not to lay up stores of treasure, but to take each day as it comes with little heed for the morrow.

(p.274) With this attitude his dependence on the mahājan is less cause for wonder. It is as much a psychological as an economic dependence, and it serves no useful purpose to try to reduce the one to the other, whether the reduction be a Marxian one in which psychology is a reflection of economics (‘the weaver is psychologically dependent because he is poor’) or a neo-conservative one in which economics grows out of psychology (‘the weaver is poor because he lacks psychological independence’).

The bottom line is that the Nuapatna weavers' relationship to their craft cannot be reduced to the bottom line. Their work is deeply embedded in their life—who they are, why they are here. In the West, only the original Lutheran notion of a ‘calling’ would seem to approach the religious quality that work has for the weaving castes of Nuapatna. Narasingha Paramanik was no doubt posing a bit when he said to me, ‘More money is not good. I am satisfied with Rs 5; I don't want to run for Rs 10. If one gets too much, it is not good. Let Suresh [his son] earn name and fame as a weaver.’ But there is more than a grain of truth in this pose. If Suresh's future is in doubt—he talks of opening a shop to sell cloth—it is not for lack of parental guidance, but because the modern world is fast closing in on Nuapatna.

13 Ikat As Techne

Ikat is the embodiment of techne. No manuals, no texts, everything passed down from father to son (and increasingly, but that again is another story, to daughters and wives as well).39 On my second visit to Nuapatna, I enquired of a weaver how his teenaged son learned the craft. Easy, he replied. He simply watched how we adults did it. Much later, I asked a little girl of seven or eight how she had learned to draw raw silk from the cocoon and twist it into thread. She made no reply, but a little boy of five who was sitting at her side, presumably her brother, chimed in with a smile, ‘Spinning, spinning, spun’ (kāṭu, kāṭu, kāṭilā), which pithily sums up the process for learning any techne.

It is not to be doubted that ikat is a techne, closely adhering to the theories of knowledge, communication, innovation, and power outlined in Section 6 of this chapter. What remains at issue is the significance of this fact. Some will find it irrelevant to the survival of ikat, and the likelihood of epistemization, mechanization, and technological unemployment. It certainly can be argued that ikat remains a techne only because it is in nobody's interest to epistemize it. The economic motivation (p.275) to epistemize would be to lay the basis for mechanization, but with the best weavers commanding a wage of less than $2.00 per day and the majority hardly $1.00, there would seem to be little room for profit in mechanizing the weavers' art. Ikat, in this view, survives as a techne only because of the poverty of the craftsman.

At the other extreme, the technostructure of officialdom, both public and private, generally views the obstacles to mechanization as insurmountable. The tie and dye process is seen as inherently imperfect, and nothing but the judgement, skill, and dexterity of the weaver can compensate for these imperfections.

The weavers, for their part, live in continual fear of technological displacement. My collaborator and I were for a long time suspected of ulterior motives: that we would use the information we were gathering to take the very rice from their mouths. However benign our intentions, they had good reason for their suspicions. A Japanese delegation had reportedly visited Nuapatna a few years back and cleared out the stocks of the co-operative societies in an unsuccessful attempt to adapt Nuapatna ikat design and method to mechanization.

Though this report was never confirmed, it was a fact that cheap imitation tie and dye, so-called bapta print saris (which bear the same relation to real tie and dye that the classics comics of my youth bore to the classics themselves) had flooded the market between my first visit in the winter of 1985–6 and my return a year later. The one appeal of these mill-made fabrics from Bombay was price, but this was sufficient—for the time being at least—for Nuapatna products to accumulate on the shelves of the co-operative societies and the mahājans' shops.

One might be optimistic for the longer run. A similar thing had happened in Sambalpur a couple of years earlier, and the pendulum of fashion had turned back to the local product when the mill-made saris demonstrated over the course of a few months the inferior durability of both fabric and colour. But for the moment, the invasion was playing havoc. Large numbers of weavers were shifting from cotton to silk; as a luxury-market product, silk was, for the moment anyway, immune to competition from low-quality imitations. And everybody was nervous about the future of the trade.

My own judgement is that although the Nuapatna weavers are reasonably secure for the moment, there is in the long run no technical obstacle to a partial epistemization and mechanization of ikat. On the other hand, a completely automatic process seems no more possible than a completely automatic process of machine-tool manufacture.

The real issue, unless the continued poverty of the weavers decides the issue in favour of the status quo, is what the mix of episteme and techne will be: whether automation will be used to make the weaver increasingly an appendage of the machine and the weaver's techne subservient to the (p.276) episteme of the mahājans of the twenty-first century (whoever they might be), or whether automation will be used to enhance the power of the weavers by making episteme serve their techne.

The die is hardly cast. The predilections of India's political, business, and academic leadership are clear. Sharing the largely unconscious presuppositions of their Western counterparts with respect to the superiority of episteme over techne, they will see little to be gained from the second scenario. Indeed, in these quarters it will probably be dismissed as Utopian idealism, however much lip-service may be paid to that arch-idealist, Mohandas Gandhi.

But the possibilities on the other side should not be underestimated. The resilience of Orissan ikat weavers, I have suggested, is not finally reducible to an economic analysis, because the resilience is fortified by the embeddedness of the weavers' work in their life. I would add, somewhat more tentatively, the hypothesis that traditional Indian attitudes toward episteme and techne leave the weaver more cultural room to defend his techne.

Tradition of course grants the Brahmin superiority over other castes and accordingly grants his knowledge a superiority over the technai of other castes. But this hierarchy differs in a fundamental way from the hierarchical dominance of episteme over techne in the West; no attempt is made in India to reduce techne to an inferior form of episteme or to deny its very existence. On the contrary: just as each caste is accorded its distinct and necessary role in a well-ordered cosmos (see Chapter 4), so must the techne of each caste be recognized as distinct and necessary.

Indeed, the Bhagaυad Gītā can be read as (at the very least) denying superiority to the Indian equivalent of episteme over techne.40 Such a reading, to be sure, is hardly necessary or even obvious, for it requires us to interpret passages (such as chs. iii and v) where disciplined action is compared with contemplative reasoning as implicating alternative knowledge systems. The terms sāṁkhya and (karma) yoga, which are used in the Gītā to distinguish ‘reason-method’ from ‘discipline’ as paths to Truth, certainly do not map directly on to episteme and techne. But the indirect links, via jñāna and υijñāna, are at least suggestive. Franklin Edgerton, eminent Sanskritist of the inter-war period and translator/ commentator of the Gītā, identifies sāṁkhya with jñāna (1944: 166) and he elsewhere glosses jñāna as ‘theoretical knowledge’ and ‘abstract, unapplied knowledge’ (1933: 218, 220). Although karma yoga is never (p.277) equated with υijñāna, the emphasis in yoga on the ‘practical’ as against the ‘intellectual’ (Edgerton 1944: 165–6) arguably implicates υijñāna, which Edgerton glosses as ‘practical or applied knowledge’ (1933: 218–20). If we are on track in identifying yoga with υijñāna, it becomes easy to see a distinction like that between episteme and techne implicit in the distinction between sāṁkhya and yoga. But any insistence on a hierarchical ordering—on superiority and inferiority—perhaps misses the point. The main emphasis of the Gītā is on the unity of sāṁkhya and yoga (v. 4–5), and on the idea that context—in particular the individual's birth—determines the suitability of one path or the other, an idea which fits very well with the notion of complementarity between techne and episteme.

This is of course highly speculative. What is not speculative is that if the ikat weavers of Nuapatna are to defend their art, the best defence will be found in a critical assessment of their own cultural traditions, not in an uncritical acceptance of the West.

For all their conservative resistance to change, Nuapatna weavers have proved themselves remarkably adaptable. The innovation wrought by Arjun Subuddhi has already been recounted. But alongside these changes there have been numerous others. The substitutions of chemical dyes for vegetable dyes, modification of the length and breadth of the fabrics, technical improvements to the loom itself—individually these are much less dramatic than the new life given to ikat by adopting the Sambalpuri style of tie and dye; but together they indicate the possibilities for evolution and adaptation within the traditional system. Only time will tell if these possibilities extend to the computer age.

14 Conclusion

This essay has put forward the argument that in the West workers have largely accommodated themselves to the capitalist project of dominating the workplace because Western culture provides neither compelling reasons nor compelling means for workers to resist this project. In contrast to Hindu culture, workers who are heirs to the Greek and Judaeo-Christian traditions find both that their work stands outside the cosmic order and that the system of knowledge in which their work is organized is held to be an inferior system, if indeed it counts as knowledge at all. Lacking both cultural reason and cultural means to defend himself, the worker has been, to paraphrase Thomas Jefferson, easy prey for the designs of capitalist ambition.

It seems to me this argument has profound implications for the project of building a decent society. Socialists in particular have consistently stressed the importance of work organization in their project, and it is, in my opinion, right to do so. With less consistency, socialists have (p.278) neglected the role of culture in shaping both the possibilities for, and the constraints on, work organization.

We in the West probably do not have the option of embedding work in our cosmology. We perhaps have more control over our theory of knowledge. We must at least hope so, for, if not, the socialist project is probably doomed. As Robert Pirsig put it (1974: 94),

[T]o tear down a factory or to revolt against a government or to avoid repair of a motorcycle because it is a system is to attack effects rather than causes; and as long as the attack is upon effects only, no change is possible. The true system, the real system, is our present construction of systematic thought itself, rationality itself, and if a factory is torn down but the rationality which produced it is left standing, then that rationality will simply produce another factory. If a revolution destroys a systematic government, but the systematic patterns of thought that produced that government are left intact, then those patterns will repeat themselves in the succeeding government.

The same socialist project which has exalted workplace democracy as the foundation of a decent society denies this foundation by making participation conditional on accepting the dominant knowledge system: workers are free to participate on condition that they accept the knowledge system of the party cadres and the engineers. But that is an option open only to a minority. As there is room for only so many cadres and engineers, most must remain workers, consigned to inferior status by virtue of the inferiority of their knowledge.

The point of my argument is not to glorify but to rediscover and to legitimize techne; to argue for a balance and even for a tension between techne and episteme. I have neither the intention nor the need to denigrate episteme in order to attack the hierarchical ordering in which episteme is placed above techne; my purpose is to argue against an imbalance that puts the very project of erecting a decent society at risk.

Outside the West, conditions are at once more problematic and more promising. To take India as a concrete example, the problems of poverty and internal conflict are intertwined with the problems posed by the encounter, cultural as well as economic, with the West. The economic consequences of that encounter are ambiguous. Whether the British deindustrialized India or laid the basis for its industrialization (or perhaps did some of each) is still debated by scholars and the issue will perhaps never be settled.

But the cultural consequences seem to me to be much more clear-cut. If India is to build a society on the foundations of democratic and participatory work organization, it will do better to follow the impulse of Mohandas Gandhi and his kind to look to India's own tradition than to follow the impulse of the ilk of Rajiv Gandhi to look to the West.

(p.279) References

Bibliography references:

Aoki, Masahiko (1988) Information, Incentives and Bargaining in the Japanese Economy, Cambridge: Cambridge University Press.

Apffel Marglin, Frédérique (1985) [catalogued under Marglin, Frédérique Apffel] Wives of the God King: The Rituals of the Devadasis of Puri, New Delhi and Oxford: Oxford University Press.

Aristotle (1934) Nichomachean Ethics, trans H. Rackham, Cambridge, Mass.: Harvard University Press.

——— (1944) Politics, trans. H. Rackham, Cambridge, Mass.: Harvard University Press.

Braverman, Harry (1974) Labor and Monopoly Capital: The Degradation of Work in the Twentieth Century, New York: Monthly Review Press.

Bruner, Jerome (1962) On Knowing: Essays for the Left Hand, Cambridge, Mass.: Harvard University Press, 1979.

Buehler, Alfred, Fischer, Eberhard, and Nabholz, Marie-Louise (1980) Indian Tie-Dyed Fabrics, Ahmedabad: Calico Museum of Textiles.

Burawoy, Michael (1979) Manufacturing Consent, Chicago: the University of Chicago Press.

Buttrick, John (1952) ‘The Inside Contract System’, Journal of Economic History, 12: 205–21.

Cobden-Ramsay, L. E. B. (1982) Feudatory States of Orissa, Calcutta: Firma KLM (first published 1910).

Copley, Frank B. (1923) Frederick W. Taylor, Father of Scientific Management, vol. i, New York: Harper & Row.

Detienne, Marcel, and Vernant, Jean-Pierre (1974) Les Ruses de lʼIntelligence: la Mètis des Grecs, Paris: Flammarion.

Dumont, Louis (1977) From Mandeville to Marx: The Genesis and Triumph of Economic Ideology, Chicago: Chicago University Press.

Edgerton, Franklin (1933) ‘Jñāna and Vijñāna’, Otto Stein and Wilhelm Gampert (eds.), Festschrift Moriz Winternitz, Leipzig: Harrassowitz.

——— (1944) The Bhagavad Gita, Cambridge, Mass.: Harvard University Press.

Edwards, Richard (1979) Contested Terrain: The Transformation of the Workplace in the Twentieth Century, New York: Basic Books.

Ellsberg, Daniel (1961) ‘Risk, Ambiguity, and the Savage Axioms’, Quarterly Journal of Economics, 75: 643–69.

Floris, Peter (1934) Voyage of Floris to the East Indies, ed. W. H. Moreland, London: The Hakluyt Society.

Freud, Sigmund (1961) Civilization and its Discontents, ed. J. Strachey, New York: W. W. Norton (first published 1930).

Friedman, Milton (1953) ‘The Methodology of Positive Economies’, in Essays in Positive Economics, Chicago: University of Chicago Press.

Gilligan, Carol (1982) In a Different Voice: Psychological Theory and Women's Development, Cambridge, Mass.: Harvard University Press.

Hacking, Ian (1975) The Emergence of Probability: A Philosophical Study of Early Ideas About Probability, Induction and Statistical Inference, Cambridge: Cambridge University Press.

(p.280) Hesiod (1914) ‘Works and Days’, in The Homeric Hymns and Homerica, trans. Hugh Evelyn-White, Cambridge, Mass.: Harvard University Press.

Hirschman, Albert (1984) ‘Against Parsimony’, Economics and Philosophy, 7: 7–21.

Kakar, Sudhir (1970) Frederick Taylor: A Study in Personality and Innovation, Cambridge, Mass.: MIT Press.

Kline, Morris (1980) Mathematics: The Loss of Certainty, New York: Oxford University Press.

Klosko, George (1986) The Development of Plato's Political Theory, New York and London: Methuen.

Knight, Frank (1921) Risk, Uncertainty and Profit, Boston and New York: Houghton Mifflin Company.

Kuhn, Thomas (1970) The Structure of Scientific Revolutions, International Encyclopedia of Unified Science, vol. ii, no. 2, Chicago: University of Chicago Press, 2nd edn. (first published 1962).

Lakatos, Imre (1970) ‘Falsification and the Methodology of Scientific Research Programmes’, in Imre Lakatos and Alan Musgrave (eds.), Criticism and the Growth of Knowledge, Cambridge: Cambridge University Press.

Langland, William (1959) Piers the Ploughman, ed. J. F. Goodridge, Hammondsworth: Penguin.

Lifton, Robert Jay (1983) The Broken Connection: On Death and the Continuity of Life, New York: Basic Books.

Lyons, John (1969) Structural Semantics: An Analysis of Part of the Vocabulary of Plato, Oxford: Basil Blackwell.

Marcuse, Herbert (1966) One Dimensional Man: Studies in the Ideology of Advanced Industrial Society, Boston: Beacon Press.

Marglin, Stephen (1974) ‘What Do Bosses Do? The Origins and Functions of Hierarchy in Capitalist Production’, Part I, Review of Radical Political Economics, 60: 60–112, repr. in A. Gorz (ed.) The Division of Labour, Sussex: Harvester, 1976.

——— (1979) ‘Catching Flies with Honey: An Inquiry into Management Initiatives to Humanize Work’, Economic Analysis and Workers' Management, 12: 473–87.

——— (1984) ‘Knowledge and Power’, in Frank Stephen (ed.), Firms, Organization and Labour, London: Macmillan.

——— (1987) ‘Investment and Accumulation’, in John Eatwell, Murray Milgate, and Peter Newman (eds.), The New Palgrave, London: Macmillan.

——— and Bhaduri, Amit (1990) ‘Profit Squeeze and Keynesian Theory’, in Stephen Marglin and Juliet Schor (eds.), The Golden Age of Capitalism, Oxford: Clarendon Press.

Marx, Karl (1959) Capital, vol. iii, Moscow: Foreign Languages Publishing House (first published 1894).

——— and Engels, F. (1947) German Ideology, ed. R. Pascal, New York: International (first published 1846).

Mohanty, Bijoy Chandra, and Krishna, Kalyan (1974) Ikat Fabrics of Orissa and Andhra Pradesh, Ahmedabad: Calico Museum of Textiles.

Montgomery, David (1979) Workers' Control in America, Cambridge: Cambridge University Press.

(p.281) Morris, Frederic (1972) ‘From Cottage to Factory’, (unpublished senior honours thesis) Harvard College.

Noble, David F. (1984) Forces of Production: A Social History of Automation, New York: Oxford University Press.

Nussbaum, Martha (1986) The Fragility of Goodness: Luck and Ethics in Greek Tragedy and Philosophy, Cambridge: Cambridge University Press.

——— and Sen, Amartya (1989) ‘Internal Criticism and Indian Rationalist Traditions’, in Michael Krausz (ed.), Relativism: Interpretation and Confrontation, Notre Dame: Notre Dame University Press.

Pirsig, Robert (1976) Zen and the Art of Motorcycle Maintenance: An Inquiry into Values, London: Transworld.

Plato (1925) ‘Gorgias’ in Lysis, Symposium, and Gorgias, trans. W. R. M. Lamb, Cambridge, Mass.: Harvard University Press.

——— (1925) ‘Philebus’ in Statesman, Philebus, and Ion, trans. Harold N. Fowler and W. R. M. Lamb, Cambridge, Mass.: Harvard University Press.

——— (1941) The Republic of Plato, trans. F. Cornford, New York and London: Oxford University Press.

Polanyi, Karl (1944) The Great Transformation, New York: Rinehart.

Polanyi, Michael (1958) Personal Knowledge: Towards a Post-Critical Philosophy, Chicago: University of Chicago Press.

Popper, Karl (1968) The Logic of Scientific Discovery, New York: Harper & Row.

Prest, John (1960)The Industrial Revolution in Coventry, Oxford: Oxford University Press.

Putnam, Hilary (1974) ‘The Corroboration of Scientific Theories’, The Philosophy of Karl Popper, ed. K. Schilpp, La Salle, Ill.: Open Court.

Ricardo, David (1951) On the Principles of Political Economy and Taxation, ed. P. Sraffa with the collaboration of M. H. Dobb, Cambridge: Cambridge University Press (first published 1817, 3rd edn. 1821).

Roy, Donald (1952 a) ‘Quota Restriction and Goldbricking in a Machine Shop’, American Journal of Sociology, 57: 427–42.

——— (1952 b) ‘Restrictions of Output in a Piecework Machine Shop’, (Ph.D. dissertation) University of Chicago.

Sacks, Oliver (1985) The Man Who Mistook His Wife for a Hat: And Other Clinical Tales, New York: Summit Books.

Smith, Adam (1937) The Wealth of Nations, ed. E. Canaan, New York: Random House (first published 1776).

Staal, J. F. (1965) ‘Euclid and Panini’, Philosophy East and West, 15: 99–115.

Strauss, Leo (1970) Xenophon's Socratic Discourse: An Interpretation of the Oeconomicus (trans, of the Oeconomicus by Carnes Lord), Ithaca: Cornell University Press.

Sturt, George (1923) The Wheelwright's Shop, Cambridge: Cambridge University Press.

Tawney, R. H. (1938) Religion and the Rise of Capitalism: A Historical Study, Harmondsworth: Penguin (first published 1926).

Taylor, Frederick (1967) The Principles of Scientific Management, New York: Norton (first published 1911).

Terkel, Studs (1972) Working, New York: Random House.

(p.282) Thompson, E. P. (1963) The Making of the English Working Class, New York: Vintage Books.

Vernant, Jean-Pierre (1982) Mythe et Pensée chez les Grecs: Études de Psychologie Historique, vol. ii, Paris: François Maspero (first published 1965).

Vidal-Naquet, Pierre (1983) Le Chasseur Noir: Formes de Pensée et Formes de Societé dans le Monde Grec, Paris: La Découverte/Maspero.

Weber, Max (1930) The Protestant Ethic and the Spirit of Capitalism, trans. Talcott Parsons, New York: Scribner's and London: George Allen and Unwin (first published 1904–5).

Whyte, William F. (1955) Money and Motivation: An Analysis of Incentives in Industry, New York: Harper & Row.

Young, Arthur (1770) A Six Months' Tour Through the North of England, London: W. Strahan.

Yule, Henry and A. C. Burnell (1903) Hobson-Jobson: A Glossary of Colloquial Anglo-Indian Words and Phrases, and of Kindred Terms, Etymological, Historical, Geographical and Discursive, 2nd edn., ed. William Crooke, London: John Murray (first published 1886).


(1.) ‘Efficiency’ is the economists' preferred term for describing the virtue of the market. An efficient set of economic arrangements is one that precludes the provision of more of one good except at the sacrifice of some other good. ‘Goods’, it should be noted, has an elastic meaning: a good can be defined as narrowly as a specific commodity like bread or ice-cream or it can be stretched to mean the well-being of individuals. In its second meaning, an efficient set of arrangements is one which precludes increasing the well-being of one individual except at the expense of another.

The advantage of the more technical terminology of ‘efficiency’ over a more informal notion like ‘the size of the economic pie’ is that it permits comparison of situations where tastes differ: ‘size of the pie’ becomes ambiguous when some people like blackberry pie while others like apple pie. But taste differences are not matters of great moment for present purposes, and we lose correspondingly little if we take the informal route of identifying efficiency with pie size.

Observe that the economists' definition of efficiency is somewhat at odds with conventional usage. A businessman thinks he is increasing efficiency if he is successful in imposing a speed-up. But the economist will object that the higher output is at the expense of greater effort. It is not necessarily more efficient in the sense of providing more of one good without sacrificing something elsewhere.

(2.) ‘Pyramids, Empire State Building—these things just don't happen. There's hard work behind it. I would like to see a building, say the Empire State, I would like to see on one side of it a foot-wide strip from top to bottom with the name of every bricklayer, the name of every electrician, with all the names. So when a guy walked by, he could take his son and say, “See, that's me on the forty-fifth floor. I put the steel beam in.” Picasso can point to a painting. What can I point to? A writer can point to a book. Everybody should have something to point to’ (p. xxxii).

(3.) Observe that control is not being opposed to profit. Contrary to ‘invisible hand’ arguments, in which profitability is a function of efficiency alone, profitability is here held to depend jointly on control and efficiency. Efficiency at best creates a potential profit; without control the capitalist cannot realize that profit. Thus organizational forms which enhance capitalist control may increase profits and find favour with capitalists even if they affect productivity and efficiency adversely. Conversely, more efficient ways of organizing production which reduce capitalist control may end up reducing profits and being rejected by capitalists.

This is not to say that businessmen actually do maximize profits. This is a controversial point on which it is unnecessary for present purposes to take a view one way or another. To assume that capitalists are motivated solely by profits may be an overly narrow, economistic perspective, but there is no need for present purposes to assume that control is an end in itself which is sought independently of its contribution to profits. In the present circumstances Occam's Razor suggests that control be regarded simply as a means to profit. Entia non sunt multiplicanda praeter necessitatem. (But see Albert Hirschman 1984.)

(4.) Taylor himself was far from the lackey of capital he is portrayed by critics on the left (for instance, Braverman 1974). Over time he became as suspicious of capital as he was of labour. Of course, capitalists made use of Taylor. But then so did Lenin, whose encomiums for scientific management were already prefigured in Marx's characterization of ‘socialized man, the associated producers, rationally regulating their interchange with Nature, bringing it under their collective control … and achieving this with the least expenditure of energy …’ (Marx 1894: 800). Sudhir Kakar 1970 has written a rewarding psychobiography of Taylor, including an account of the experiences which soured Taylor on captains of industry.

(5.) Plato, Republic, 368–70; Aristotle, Politics, 1252–4. The agreement holds even as to the basis of specialization. We can distinguish between Smithian and Ricardian theories of the division of labour, the first arguing the efficiency of specialization in terms of environment (that is, learned behaviour) and the second in terms of heredity (endowment). Smith's arguments for such specialization are couched in terms of learned behaviour and in no way presuppose a genetic disposition or proclivity to one kind of work or another (1937). Ricardo's theory of specialization presupposes an innate ‘comparative advantage’ based on differences in endowments (1951). In this typology both Plato and Aristotle are at the Ricardian end of the spectrum.

(6.) Is it to Genesis we owe the double meaning of ‘labour’?

(7.) Why Langland rather than Aquinas or Dante as the representative medieval Christian? A thorough study of the issues raised here would doubtless include all three, but, as a low-ranking cleric, Langland is probably closer to popular belief than others.

(8.) Tolerance and mutual respect are enjoined in a remarkable passage which prefigures the Rawlsian emphasis on suppressing individual circumstances in arriving at a just social contract: ‘Some occupations are cleaner than others’ said Grace, ‘but you can be sure that the man who works at the pleasantest job could have been put to the foulest. So you must all remember that your talents are gifts from me. Let no profession despise another, but love one another as brothers’ (Book xix).

(9.) It can also be a hierarchy of sex, in which case the linearity notion obviously does not apply. But traditional societies typically allocate separate technai to men and women, so that in so far as power relationships between sexes are related to knowledge they reflect rules of power between rather than within knowledge systems. Carol Gilligan's well-known work, In A Different Voice (1982), can be interpreted as a defence of (feminine) techne against (masculine) episteme.

(10.) There are exceptions; the law and business faculties of my own university have since their inception given pride of place to the ‘case method’ of instruction, which attempts to condense the techne of the law office and the executive suite into a form accessible to students, and to allow learning to take place in a simulated environment in which mistakes are not too costly. But even these exceptions are coming increasingly under fire from the epistemic academic establishment.

(11.) In Hacking's formulation, ‘knowledge’ (episteme in present terminology) is opposed not to techne, but to ‘opinion’, roughly speaking, received doctrine. Of course, the role of received doctrine is sufficiently important in the present conception of techne as to suggest at least a family resemblance between the two concepts.

(12.) Adam Smith attached great importance to the division of labour as a stimulus to invention. ‘Men’, he observed in the Wealth of Nations, ‘are much more likely to discover easier and readier methods of attaining any object, when the whole attention of their minds is directed towards that singular object, than when it is dissipated among a great variety of things’ (1937: 9). An extraordinary claim, when we consider that invention requires an appreciation of the connections between various parts of the process, such as is achieved by integrating rather than dividing labour. Smith's claim makes more sense when it is recognized that he had in mind as potential inventors ‘philosophers or men of speculation, whose trade it is not to do any thing but to observe every thing; and who, upon that account, are often capable of combining together the powers of the most distant and dissimilar objects’ (p. 10). For Smith's ‘philosophers’, knowledge of production is in our terms necessarily epistemic. They stand outside the production process and comprehend it only if it is broken down into its components.

(13.) Subjective probability theory does have one great virtue. It emphasizes that reality is constructed rather than objectively given, an idea immanent in neo-Keynesian views of the dynamics of the capitalist economy (see Chapter 1 above, Marglin 1987, Marglin and Bhaduri 1989). But this is hardly the basis for its appeal to mainstream economists, for whom this aspect of Keynes is totally alien.

(14.) Economists will recognize that subjective probability theory does not necessarily do away with the need for a techne of coping with an uncertain future. As the standard theory is customarily stated, episteme suffices because subjective probability is teamed up with the somewhat fanciful notion of complete contingent commodity markets. In this theory, subjective probability is limited to expressing the likelihood of alternative ‘states of the world’, a notion that includes whatever information is relevant to the contracting parties, whether it be a temperature in Helsinki in excess of 30°C, which might be relevant to manufacturers of soft drinks, ice-cream, and sun-tan lotion, or the end of hostilities in the Middle East, which might be relevant to arms manufacturers. In the standard theory, the prices that would obtain in each state are assumed to be known with certainty, as the outcome of trading in contingent commodity markets. Contingent markets are markets in which delivery is contingent on a particular state of the world actually occurring. Conjured out of a warp of future markets (such as the market for no. 2 corn to be delivered in Kansas City next December) and a weft of ad-hoc insurance (à la Lloyd's of London), contingent commodity markets are not quite as fanciful as they might appear to the non-economist. The difficulty comes not in positing the existence of the odd contingent market, but in positing ubiquity. To reduce individual behaviour to epistemic calculations, subjective probability is coupled to the assumption of complete markets, so that there is provision for every possible commodity and every possible contingency, at least every contingency that is economically relevant.

Without complete markets, a subjective probability distribution over states of the world is not sufficient to allow us to take action on the basis of utility maximization. We would need a subjective probability distribution over contingent market prices as well. Indeed, while the point of the theory is to show the logical possibility of an epistemic treatment of behaviour in an uncertain world, the peculiarity of the assumptions required to reach this result points in the opposite direction, namely the impossibility of a purely epistemic theory.

(15.) Taylor had amplified this observation in testifying before a governmental commission, ‘When I got to be a foreman of the shop and had finally won out and we had an agreement among the men that there would be so much work done—not a full day's work, but a pretty good day's work—we all came to an understanding, and had no further fighting. Then I tried to analyze it, and said: The main trouble with this thing is that you have been quarrelling because there have been no proper standards for a day's work. You do not know what a proper day's work is. Those fellows know the times more than you do, but personally, we do not know anything about what a day's work is. We make a bluff at it and the other side makes a guess at it and then we fight. The great thing is that we do not know what is a proper day's work.’ (Testimony before the Industrial Relations Commission, cited in Copley 1923.)

(16.) Whyte 1961, ch. 10, and Burawoy 1979 provide excellent accounts of scientific management in operation in the same plant at an interval of 30 years. (The account in Whyte was contributed by Donald Roy. See Roy 1952a, 1952b for a more detailed discussion of the dynamics of boss-worker relations in the plant during the 1940s.)

(17.) And worse. One machine operator reported on the potentially fatal consequences of a programming error: ‘He tabbed it wrong one time and we were in a hole; I had a two-inch drill and I was seven inches in the hole. The machine took off and it threw that drill fifty feet down the shop. It weighed two pounds and it just missed me; it went over the helper's head and took his hat off. So after that I said, “I'm going to know this thing, because this might kill me”’ (p. 246).

(18.) But see Noble (1984: 164) for the view that R/P systems inherently gave the operator control over production speeds.

(19.) Even at the level of pure thought. Gödel, at least as he is translated for the non-specialist (see, for example, Kline 1980: 260–4), is supposed to have demonstrated the impossibility of a consistent and complete theory on any set of axioms: either there will be inconsistencies (theorems which will be at once ‘true’ and ‘false’) or there will be incompleteness (theorems whose truth status cannot be determined). With this unpromising beginning, it is hard to see how one would even be tempted to rely on pure episteme.

(20.) Aristotle, it should be observed, is inconsistent in his usage of techne and episteme (see Nussbaum 1986: 444), and earlier writers, including Plato, appear to have used the terms almost interchangeably, at least in those areas that are of concern to the present enquiry (Lyons 1969, Nussbaum 1986: 444). In particular, Nussbaum ascribes to techne (pp. 94–6) many of the characteristics that I have not only ascribed to episteme, but have made pivotal in distinguishing between the two.

(21.) Vidal-Naquet, along with Marcel Detienne and Vernant 1974 have emphasized mètis (after Metis, first wife of Zeus and mother of Athena) as a separate system of knowledge within classical Greek culture. Mètis connotes cunning, trickiness, deception—all relevant to the struggle over control of production as well as to the mythic struggles of gods and heroes. A separate study would be required to analyse the relationship between mètis, episteme, and techne in the production process. Suffice it to say that from my perspective it is not an accident that ‘crafty’ captures an important aspect of mètis.

(22.) In Gorgias (465 A) Plato has Socrates say with respect to the art of cooking: ‘I say it is not an art, but a habitude [empeiria], since it has no account to give of the real nature of the things it applies, and so cannot tell the cause of any of them. I refuse to give the name of art [techne] to anything that is irrational.’ In Philebus (55 D–E) Socrates asks his interlocutor to ‘consider whether in the manual arts [cheirotechnikai] one part is more allied to knowledge [episteme], and the other less, and the one should be regarded as purest, the other as less pure.’ He goes on to assert, ‘[I]f arithmetic and the sciences of measurement and weighing were taken away from all arts [technai], what was left of any of them would be, so to speak, pretty worthless.’

(23.) Dyeing is by the so-called ‘resist’ method, in which bundles of yarn are divided into several segments according to the design, and segments to which dye is not to be applied are wrapped in cylinders, originally of leaf, now generally of plastic, before immersion into the dye. Colours are applied sequentially with different portions of the yarn sealed off from the dyes in turn.

(24.) The name ikat is Malayan in origin (Mohanty and Krishna 1974). The Oriya word is bandha. Both ikat and bandha come from words that mean to tie or to bind.

(25.) In pre-Independence days the Mahanadi was a principal avenue of transport and trade, but not apparently for the Nuapatna cloth trade. Before the bus and truck arrived on the scene, however, at about the time of Independence, forty years ago, the chief means of transporting textiles to the Cuttack market was to carry a load of cloth atop one's head, a journey that took the better part of two days.

(26.) A khanduā is a shawl used to cover the upper body, and a Gītagovindakhanduā is the shawl used to cover the images of the gods in the Puri temple. The special khanduā in question gets its name from the 12th-century devotional poem, the Gīta Govinda, the song of Govinda (Lord Jagannātha), a verse of which is woven into the fabric by the ikat method.

(27.) An alternative explanation, favoured by Alfred Buehler and his collaborators, is that elaborate ikat motifs are a late 19th-century or early 20th-century innovation in Orissa, and that Gītagovindakhanduā were in earlier times embroidered rather than woven with tie-dyed yarns (Buehler et al. 1980: 70).

(28.) Count is a measure of the fineness of cotton yarn, denned as the ratio of the number of yards of a single thread required for a pound of yarn to 840. Thus if 8,400 yards of a single thread weighed one pound, the count would be 10. Coarse yarns are those below 40 count. Medium yarns are 40–60 count. Fine counts are 80 and above. Nuapatna yarns rarely go above 80 count, although I have encountered 100 count yarn and 2 × 120 count mercerized yarn. (Mercerizing is a finishing process for adding resilience to the yarn, and 2 × 120 means that 2 strands of 120 count yarn are twisted together for additional strength.)

(29.) According to Hobson-Jobson (Yule and Burnell 1903: 536), the word comes from the Sanskrit for great (mahā) people (jan). Its closest English equivalent is businessmen, but mahājan carries a greater connotation of power and importance.

(30.) The largest and most successful weavers' co-operative society in Orissa, the Sambalpuri Bastralya, itself maintains three handloom factories. But that is another story.

(31.) This voyage was innovative and portentous, in that its mission was not simply trade between Europe and the Indies. Most of the voyage, lasting from 1611 to 1615, was spent producing goods in India for trade in South-East Asia. Textiles were the chief product.

(32.) The most striking example was provided by a brief encounter with a prosperous and prominent family of weavers in Barpali, a centre of ikat in Sambalpur district. On a previous trip several years back, Purna Chandra Mishra had taken photographs of the family and particularly the head of the family, a distinguished weaver. As the photographs had been returned with the notation ‘addressee unknown’, Mishra was taking advantage of our visit to deliver the photographs in person. It turned out that the patriarch had died some months before, and our visit was a great consolation and even an occasion of joy: the family had no photographs of the old man. Amid the rejoicing, the three brothers, themselves proficient and even prize-winning weavers, sent immediately for the mahājan so that he could join in the joy of the occasion. The young men's warm feeling for the mahājan could not have been fabricated; clearly much more than an economic nexus was at work here.

(33.) My four-year-old was to have accompanied her mother on a tour of the plant during one of my visits. Jessica, who had with great equanimity dealt with all the novelties of India, began to scream upon entering the mill and could not be soothed within its walls. The tour was aborted.

(34.) Interestingly none of these restrictions apply to the talikāma (the preparatory arrangements), only to the actual weaving itself. I am unable to provide a reason for this distinction.

(35.) Prior to 1955, Puri temple was administered by the Raja of Puri, the only public role remaining to him after the British deprived him of his temporal authority in the 19th century. Since 1955, when the lands that supported the temple were confiscated, the temple has been subsidized and administered by the Government of Orissa (Apffel Marglin 1985).

(36.) Parāṇa means old, but refers in context to the document that mythically describes caste origins.

(37.) The mètis (craftiness) of the gods recalls the original Mètis. See n. 20 above, and Detienne and Vernant 1974.

(38.) The demon's boon is related explicitly, but the circumvention of the boon is not. The story simply recounts that the kāminī lures the demon outside under the eaves, on the pretext that there is insufficient light in the house to see the lice. There is no need for more explanation. The boon and its remedy, a standard device in Hindu mythology, would be familiar to the listener.

(39.) I err. Bijoy Chandra Mohanty and Kalyan Krishna 1974 have described the ikat process. But I defy anybody who is not already initiated into the art to make head or tail of their description, and those who are initiated have neither need nor facility for epistemic explanation.

(40.) It would take a much better understanding of Indian knowledge systems than I could ever claim even to outline the contours of the field. But a priori it is highly unlikely that Western episteme in particular will map isomorphically on to an Indian equivalent: J. F. Staal 1965 argues that Euclidean geometry is not the canonical form of rationality (read episteme) in India. In Staal's view, grammar rather than axiomatic mathematics is the canonical Indian episteme, and Panini rather than Euclid is the model.