Industrial conflict as a source of technical innovation: three cases

Tine Bruland

Abstract

This article is concerned with the history and theory of technical change. The central problems in this area involve explaining not simply the general propensity of the capitalist economy to technical development, but also the impulses to specific lines of technical advance: why, for example, are some production processes subject to innovation while others are not? Why does the urgency with which possible innovations are examined differ between processes? There are a wide range of possible explanations here. This article explores one of them: lt argues that industrial conflict can generate or focus technical change in production processes which are prone to such conflict. This relationship is demonstrated concretely by reference to three textile process innovations in the nineteenth-century British economy.


The economic and social history of Britain in the late eighteenth and early nineteenth centuries is written in large part around the impact of new techniques of production, which — it is usually held—changed methods, products and places of production, and which formed, under their impact, the industrial working class. The central concept in the historiography of this period is of course that of 'industrial revolution'. Quite recently, the validity of this concept has been questioned by various writers (Bruland and Smith, 1981a; Tribe, 1981; Fores, 1981) on a number of grounds among which is that the term 'industrial revolution', while being constituted around processes of technical change, offers no satisfactory explanation of such change. This is a serious problem since although technical change can and should be rejected as an autonomous determining: factor in the social and economic transformations of this period, it did happen end requires explanation: perusal of the patent series, for example, reveals very sharp and sustained increases in inventions patented after 1760, on such a scale as to be significant even allowing for the fact that patents may be a weak indicator of the actual adoption of new processes (Deane, 1969, p. 128).

There are, in general, two kinds of approach to technical change in the literature on the late eighteenth and early nineteenth century British economy, each with attendant limitations. In the first approach, no serious account of the impulses to technical change within the British economy of that period is offered at all, and technical change is seen—at least implicitly—as self-generating; sometimes, quite startling forms of technological determinism result from this.' On the other hand, an explanatory account may be offered of a very general kind: Landes, for example, in his Unbound Prometheus, writes of technical change as an effect of a conjunction of Western 'rationality' (by which is meant means-end calculation) and a 'Faustian spirit of mastery' (Lances, 1978, p. 21). Rather more sensibly, Samuel Lilley emphasizes the causal effectivity of the control, decision-making capacity and incentives to innovate of the capitalist entrepreneur:

The capitalist entrepreneur is aware—to a degree that no previous exploiter is aware—of how much he stands to gain from this or that technical change. He probably also has enough technological knowledge to judge the practicability of an invention, perhaps even to invent for himself. And the cold steel of competition reinforces this awareness and eliminates those who do not possess it. Hence derives the extreme sensitivity of response to technological opportunity that eighteenth century entrepreneurs repeatedly exhibited [1978, pp. 219 - 20] .

This approach, which has a Marxian-Schumpeterian ancestry, seems to me considerably superior to Landes's. Yet both share the weaknesses of their generality. They seem at least potentially to explain too much, for if taken seriously they would imply an impetus to technological change across the whole range of technical processes of the capitalist economy. However, the history of technological change is in fact one of advance in quite specific directions, often concentrated not just on particular sectors of the economy, but on particular processes within sectors subject to change. In a word, there appear to be priorities. The theoretical problem here has been most succinctly outlined by Nathan Rosenberg:

In the realm of pure theory, a decision maker bent on maximizing profits under competitive conditions will pursue any possibility for reducing costs...What forces, then, determine the directions in which a firm actually goes in exploring for new techniques? Since it cannot explore in all directions, what are the factors which induce it to strike out in a particular direction? Better yet, are there any factors at work which compel it to look in some directions rasher then others? [ 977, pp. 110 - 11]

For any economic history of technical change, or for an economic theory of technical innovation, there are clearly a range of theoretical problems, at different levels. There might be, for example, a quite abstract level at which the general propensity of a system to such change is explored: this is, as I take it, the level at which Landes and Lilley, referred to above, are writing. Then there are questions about why particular sectors of the economy exhibit a propensity to technical change: here one would have to consider questions of industrial structure, of development of markets and patterns and levels of demand, of the structure and capacity of producer goods industries, of state economic policy.

Landes's account of the determinants of change at this level is couched in terms of the effects on marginal costs of output expansion in growing sectors: steeply rising marginal costs a general incentive to technological advance (1978, p. 42). But is still quite a general level of explanation, for, even given all this, questions arise concerning why it is that particular processes within an industry are the object of an innovative effort (at some particular time) while others are not. Why, in the 1820s, for example, were there major efforts to automate cotton spinning— even though this was already the most technically advanced of cotton manufacture—and not preparation of the raw material, a process which was relatively backward, technically speaking? These kinds of questions are clearly crucial in understanding particular technical structure of production emerges.

The problem here might be put in a different way, in terms of the accounts, noted above, of Landes and Lilley. Like many economic historians, they emphasize the role of the as a rational calculator. This emphasis probably derives from two sources: on the one hand, from neo-classical economics, which is essentially a theory of rational decision making in the face of a field of given costs and prices, and, on the other, from a position often ascribed to Weber, Sombart and Schumpeter, in which the differentia specifica of capitalism is seen as the calculating entrepreneur, a kind of rational risk-taker. The difficulty with such approaches is that they remain at what I referred to above as the most abstract level of analysis. This is because they isolate the process of calculation from the objects of calculation; that is to say, capitalists do not just calculate, they calculate about something, and it is this 'something' on which we should concentrate if we wish to understand the impetus to specific lines of innovation. This involves focusing on the problems which innovation might solve for enterprises, rather than the process of calculation through which potential solutions to such problems are evaluated. Instead of seeing calculation as a relatively neutral activity directed to a given field of costs, attention should be paid to costs and their real basis more directly, since it is here that specific problem areas, which in some sense require innovative solutions, can be identified.

Nathan Rosenberg has proposed three such 'problem areas'. They are: technological complementarities, in which imbalances between technical processes induce correcting innovations; supply disruptions of various kinds, leading to innovations to provide substitute products and processes; and labour conflict, in which strikes or plant-level struggles generate 'a search for labour-saving machines' (1977, pp. 110, 117, 121). It is with the last of these problems that this article is concerned: I shall attempt to show that within the British textile industry in the early nineteenth century, direct causal relationships can be identified between particular arenas of industrial conflict and particular kinds of technical innovation. We can thus understand certain specific avenues of technical change, and hence certain aspects of the technical structure of production, in terms of the struggles over power, control and distribution which are endemic within production systems in capitalist economies.

There are reasonable theoretical grounds for assuming that some such relation exists; this can be considered by asking how labour discipline affects costs, and how choice of technique appears, within this context, as a problem for entrepreneurs.

An enterprise usually faces a given wage rate, or more accurately wage rates; does this suffice to specify labour costs per unit of output? The answer, of course, is no, because such costs depend not only on the wage rate but also on the always variable amount of production which can be generated out of the labour which the enterprise buys (the assumption that there is a given, technically determined level of output associated with given amounts of capital and labour rather than a variable output is one of the major weaknesses of the production function approach to choice of technique). Thus unit costs for the capitalist entrepreneur are not pre-given, and the enterprise must organize a system of pro-auction which will not only co-ordinate the labour and fixed capital which has been assembled, but also extract maximum output from it; because workers are not mere robots, this involves the creation of a managerial organization of production in which the entrepreneur strives for the power to attain control of the output rate. In a competitive market the entrepreneur is compelled to do this, for failure to achieve average sector levels of productivity will eventually lead to products being priced out of the market; thus the construction of an apparatus of labour discipline, with the aim of achieving regularity and continuity of output under the direction of management, is an integral part of capitalist economic organization. As Pollard has shown in The Genesis of Modern Management, much of the early history of industrial capitalism shows a deep concern with such problems on the part of (embryo) managers.

The real unit cost of labour to the entrepreneur is thus composed of two elements, the wage rate and a rate of physical output. In consequence, in organizing the technical side of a production process, the capitalist entrepreneur might be expected to have an eye for aspects of a technique which extend well beyond its price tag. in achieving a certain rate of output, what kinds of labour power would be required—skilled adult males? children? Is the potential labour force organized in unions, what are its strengths and weaknesses in wage-rate bargaining? To what extent is the technique depend on skilled labour (which may be able to exert some control over the pace and quality of production) or on unskilled labour-power (in which the replaceability of labour can be used as a bargaining counter)? Such questions are crucial decisions concerning innovation, both in assessing available technical possibilities and, as I show, in deciding where innovative opportunities lie; thus entrepreneurs may seek a solution to problems of supervision, control and conflict by innovating around the problem area, for example. Within economic history as presently constituted, these questions are rarely addressed; when labour problems are considered in relation to technical change, they tend to be related either to the substitution possibilities of new techniques (Deane, 1978, p. 179), or to labour difficulties in manufacture, which the mechanized factory is assumed to have solved (that is, labour obstacles have some effects in the transition from manufacture to the industrial factory, but are not thereafter decisive—see, e.g., Landes, 1978, p. 43).

Perhaps the only body of work in which relationships between technique and labour conflict are systematically addressed is the Marxist literature on the labour process. In the main these see techniques not as 'neutral' with respect to labour, but on the contrary chosen in order to facilitate —as far as possible—a general system of control and coercion by capitalists and manage meets; techniques are weapons in struggles over the 'contested terrain' of production, in R. E. Edwards's phrase. There are clearly links between those conceptions and the ideas and examples developed here. But they should not be over-emphasized, principally because this article is about the development of new techniques, not their use. It is not concerned, therefore, with the labour process as such (if it was I would want to develop some criticisms of the Marxist literature on this topic), but rather with the way in which labour problems should form a part—but only a part—of the wider theory and history of technical change.

The idea that there is some fundamental connection between industrial conflict and technical change is not a new one. This article was in fact stimulated by a passage in Andrew Ure's Philosophy of Manufactures (Ure, 1835, Book Three, ch. 1, pp. 364 - 71) in which Ure describes what he claims are the circumstances surrounding the development of Roberts's automatic mule and a number of other devices; all were developed, in his account, in order to circumvent problems of wage demands or labour discipline. Ure's remarks are discussed and derided for their internal contradictions by Marx in Capital, vol. I (Marx, 1976, part 5, ch. 15, section 5, pp. 562 - 4), though Marx unhesitatingly accepts the general relationship posited by Ure (they can safely be said to have agreed on little else). And brief discussions or passing references can be found elsewhere in the nineteenth-century literature on mechanized production, by historians and biographers (e.g. Baines, 1835, p. 208; Smiles, 1863, pp. 267 - 8, 271, 294 - 5), by engineers (e.g. Fairbairn, 1877, p. 163; Nasmyth, 1883, p. 307), and by those concerned to am Iyse the impact on production of the new mechanized techniques (e.g. Babbage, 1835, pp. 298 - 300;Gaskell, 1836, pp. 145 - 7, 349). Unfortunately, Ure's account is essentially anecdotal, being part of a virulently anti-union polemic; and Marx's treatment, despite his recognition of the importance of the matter, consists of little more than quotations from Ure and Fairbairn.

In fact, with a few exceptions—a brief section of a paper by Nathan Rosenberg, noted above, which relies heavily on Marx's already derivative account, a paper by Stephen Marglin on the origins of hierarchy in capitalist production and the work of William Lazonick on the impact of technical change on the technical division of labour2 —the place of industrial and social conflict in technical change has not even been accorded a narrative history, let alone a full discussion and analysis. The objective of this paper is to contribute to such a history, by showing that Marx's assertion that 'it would be possible to write a whole history of the inventions made since 1830 for the sole purpose of providing capital with weapons against working class revolt' can be substantiated.

The following sections ask this question: how, historically, have concrete technical choices been conditioned or determined by antagonisms and struggles between workers and textile entrepreneurs? This is considered through an examination of three examples of technical change in the late eighteenth and early nineteenth centuries Richard Roberts's automatic spinning mule; cylindrical calico printing machinery; and devices for wool-combing in the worsted manufacture.


1 The development of the automatic spinning mule

Cotton spinning was the first part of the textile industry to be mechanized, with the development in the late eighteenth century of first the water-frame, based on a system of rollers, and then the spinning jenny, 'which reproduced the actions of the hand spinner' utilizing a system of spindles with a moveable carriage (Singer, 1958, p. 278). But the major technical change was the development and diffusion of the 'mule', so-called because it was a cross between the frame and the jenny, employing elements of both. The first mule, invented by Samuel Crompton, was working by 1779, and over the next fifty years was subjected to a great number of improvements which considerably increased its productive capacity (Singer, 1958, pp. 279 - 99); variants of this machine formed the staple device around which the development of the textile industry occurred. It permitted large increases in productivity: so much so that the technical development of the cotton sector as a whole is often written in terms of the imbalance between spinning and the other processes of cotton manufacture. Stedman Jones, for example, writes of 'the structural crisis which afflicted the cotton industry.... Its cause lay in the disbalance between the mechanized spinning and the unmechanized weaving sectors' (1975, p. 38)3 and Singer points out that the invention of spinning-machinery stimulated efforts to speed up the preparatory processes (1958, p. 280). These lags and imbalances are unquestionably important in the technical development of the industry, but they should not deflect too much attention away from the developmental effort which continued to be made in spinning itself and which cannot, of course, be explained by reference to the advanced character of other parts of the sector. The reason for continued emphasis on new techniques of spinning was simple: 'Despite the developments which had taken place, the final stage of spinning remained by far the most expensive single operation and it was here that efforts of cost education were concentrated' (Calling, 1970, p. 48).

These efforts had one central objective: quite apart from modifications to the basic device produced by Crompton, which included the application of power, increases in the number and speed of rotation of spindles, and so on, there was a major effort to render the power-driven mule 'self-acting', or automatic: 'viz. a mule to act itself, without the aid of attendants, further than merely to piece broken threads of yarn, and to clean, oil, and keep the machine in order' (Montgomery, 1886, p. 196). Montgomery lists some of the many attempts in this direction:

Many have been invented and secured by patents, but few have been found practically efficient. And it is believed that the inventions of the following parties only have been put into operation beyond the purposes of mere experiment, viz. the Messrs. Eatons of Wiln, in Derbyshire and in France; Mr. De Jongh of Warrington; Mr. Roberts and Mr. Knowles of Manchester; Mr. Buchanan of the Catrine Mills; Mr. Smith of Deanston, Scotland: and Dr. Brewster of America [1886, p. 197] . What did the spinner do during putting-up which was of such complexity as to confound the efforts of those early nineteenth century inventors? . . . During the actual operation of putting up, the spinner does three things simultaneously and in a closely co-ordinated manner:

1. He pushes the carriage in towards the roller beam.

2. By means of the faller shaft and wire he controls the winding of the yarn onto the copchase in such a way as to form a correctly shaped cop.

3. He turns the spindles, by means of the fly, at such a rate as to just wind up the yarn being released by the inward movement of the spindles. This must be done sufficiently precisely as to avoid breaking (or even unduly stressing) the yarn yet without ever permitting the yarn to go slack. . . .

Inability to provide sufficiently precise co-ordination of 3 and 1 was the Achilles heel of the Eaton and all other preRoberts self-actors. Much of the difficulty experienced in automating putting-up arose from the fact that the cop was very much the preferred form of yarn passage [1978, p. 431.

The particular character of the cop package was dictated by the needs of weaving; winding the cop was subject to a number of performance parameters which are not strictly relevant here, but as Catling says, 'even to a skilled operator, it was a task of the utmost delicacy, very dependent on continuous visual monitoring of the changing situation as the carriage approached the roller beams' (1978, p. 45).

The skilled operative spinner was thus a technically important figure in the production process, which was reliant on his dexterity, judgment, etc.; without his technical skills the process could not be carried on. He was aided by a number of ancillary workers, however. The general organization of the process is described by Ure:

In working the common, or as it is, for the sake of distinction, now termed, the 'hand-mule', various persons are employed to perform different portions of the work; viz. the 'spinner', who directs the general operation of the machine, gives to the yarn a suitable degree of twist during the spinning, and, when spun, winds the yarn in a certain form around the spindle to make what is termed a 'cop'; one or more 'pieces' to join the threads which break during the spinning, and to remove the cops, when formed, from the spindles; a 'creel-filler' to place the 'rovings from which the yarn is to be spun, in a part of the machine termed a 'creel'; and a 'cleaner' or 'scavenger', to remove the waste cotton, termed 'fly', which accumulates during the spinning, and to clean the machine generally. The 'spinner' being the principal person of the set thus employed, and, in most instances, an adult; the others being subordinate to him, and always young persons, or children; the set, thus arranged, working one pair of mules [1861, p. 152].

Ure emphasizes the general directive, almost managerial, role of the spinner, as do other writers (Calling, 1970, p. 154). What he does not mention is the system of sub-contracting through which the spinning process was set up: the piecer and other assistants of the spinner were in fact his employees (indeed often also his relations), and the supervisory role of the spinner was thus a consequence not purely of his technical place in the production process but of his place in the economy of the process in an attempt, presumably, to weld his supervisory function to a concern for output the sub-contracting system was linked to a payment system based on rates for the piece. The whole organization of spinning was thus oriented around the central role of the spinner. This centrality of the spinner is crucial to the innovation effort which will be described below: but it is important to recognize that, although the spinner was indeed a skilled worker, it was not just his skills which made him so important. These skills, and the bargaining position which they strengthened, were often acknowledged; but they were probably not an insuperable barrier faced by the cotton entrepreneurs. It was certainly possible for piecers and other ancillary workers to become mule rninders; but, as the Webbs pointed out, 'the cotton operatives display a remarkable aptitude for trade unionism.' One of the most important objectives of the spinners' union was the control and restriction of entry into spinning; to read some of the spinners' union literature is almost to step into the world of a craft guild, with its severe regulation of recruitment, training and standards. In the delegate meeting of operative spinners at the Isle of Man in 1829, the subject was almost an obsessive one:

Mr. Doherty thought that now was the proper time to come to the very important question of piecers spinning. So much had already been said on the subject, and there seemed such a unanimity of feeling as to the existence of the evil, and the absolute necessity of a remedy being applied, that he would not take up their time with offering any arguments in support of his proposition, but would merely move a resolution, that no spinner should allow a piecer to put up, except a son, brother, or orphan nephew [Operative Spinners, 1829, p. 36].7

These sentiments were 'heartily seconded', 'cordially supported', 'concurred entirely with', with only the Glasgow delegate disagreeing because 'it scarcely went far enough' (Operative Spinners, 1829, p. 37). The wages received by the spinners were thus not simply a reflection of the skills of the operative, but were also based on his quasi-managerial role in production, and the ability of spinners to restrict entry. This 'quasi-managerial role' could well be, as we shall see, most important in understanding the pace of diffusion of the self-actor.

Thus, though replaceable in the production process, the operative spinners were clearly difficult to replace in numbers, for the technical and organizational reasons outlined above, and their relative wages were indeed high. Wages fluctuated, of course, according to particular phases of the trade cycle, but they were consistently higher than those of other employees. The scale of relative wages which Gaskell estimated in the early 1830s was probably typical:

Spinner 20s to 3 Ss

Piecers 4s to 8s

Card-room Hands 6s to 14s

Reelers 7s to 1 3s per week

[1836, p. 394]

Such wages for spinners were the result of the translation of the central productive and quasi-managerial role of the spinners into concrete bargaining positions; high wages were both a result of success in wage conflicts, and (at least according to Ure) a factor in shaping the ability of spinners to undertake major strikes: 'the spinners . . . from their ample pecuniary resources were able to continue them (i.e. strikes) for long periods' (Ure, 1861, p. 153).8 This judgment seems confirmed by the care and detail with which the operative spinners conference of 1829 considered problems of strike funding. To sum up, in the words of Cooke-Taylor:

It is obvious that the spinner is a very important workman when such mules . . . are employed: on him depend not merely the machinery and its work, but the employment of the young piecers and the 'scavengers' [ 1843, p. 151 ] .

And the spinners were consistently ready, if circumstances warranted, to use the strike weapon in support of their wage objectives: 'The spinners knew their strength, and, though they received very large remuneration, frequently turned out for higher wages' (CookeTaylor, 1843, pp. 151 - 2).

Their general strategic approach is described by Turner:

Thus, the spinners' early strategy is essentially self-reliant and opportunistic. When trade is slack, resistance to the erosion of their high wage standards is confined to mill actions, stubborn though these might be. The spinners' major strength is reserved for the times—1810, 1818, 1823 - 4, 1830 and again in 1834 - 6—when its exercise will be most embarrassing to their employers [1962, p. 76].

The strikes exhibited a range of tactics and objectives, and although nearly all failed, 'the indirect effect of the spinners' vigor must not be discounted' (Turner, 1962, p. 77): by and large the spinners succeeded in preventing wage reductions and in maintaining the wage differentials noted above. But whatever the direct economic effects of the strikes for the spinners, the wider economic effects were very marked, for a spinners' strike almost invariably mean a full-scale stoppage:

The spinners do not form a greater proportion than one-tenth of those employed in a cotton mill, but their labour is absolutely necessary to the working of the establishment; consequently by refusing to work themselves, they force all their fellow labourers to be thrown out of employment at the same time. Thus, however unwilling nine-tenths of the work people in a factory may be to strike, they have no power of refusal in the matter, but are subject to the uncontrolled and despotic sway of the remaining tenth, who can order them to cease working, whenever it suits their will and pleasure [Tufnell, 1834, pp. 12-13].

That these stoppages, though often defeated in their primers aims, were of considerable effect in establishing a challenge to the power of the cotton-masters is indicated, in a sense, by the tone o Tufnell's passage: his reference to 'the uncontrolled and despot)' sway' of the spinners is hardly what one would expect to fin said of a perennially defeated group of workers. This tone o resentment, and perhaps fear, can in fact be found elsewhere in' the literature of the time. Ure, for example, in The Philosophy' of Manufactures (1835) remarks that 'the mule spinners have also abused their power beyond endurance, domineering in the most arrogant manner . . . over their masters' (p. 366): and Baine indicates that the activities of the spinners were not simply an irritation to outside commentators with axes to grind: '[the spinners'] combinations and stoppages of work have often been extremely annoying to the masters' (Baines, 1835, p. 208).

Although the spinners' high wages were clearly partly to do with their supervisory functions, their ability to carry out tines' strikes and stoppages was widely ascribed to their technical place in production; and the question thus arose as to whether a change in the technique might subdue the unruly spinners. It appears the this was enough of an issue to unite a number of manufacturer in recognition of their common interest vis-à-vis the spinners with a major strike in 1824 (following the repeal of the Combination Acts) providing the crucial impulse: although, as Wheele remarked in 1836, 'the object has occupied the attention o ingenious men, more or less earnestly, for the last four and tweet, years', and goes on to note that 'it has been most eagerly pursued when the manufacturers were labouring under the inconvenience and loss of an actual or threatened turn-out of their hands' (183t p. 538). There are a number of nineteenth-century accounts of the events of 1824 (e.g. Ure, 1835, 1861; Axon, 1825; Wheeler,1836; Cooke-Taylor, 1843; etc.); perhaps the most comprehensive is that of Samuel Smiles (1863), which is, incidentally, much plagiarized: see Lang (1928), who uses it almost word for word. All of these accounts concern an approach made to Richard Roberts, a well-known mechanical engineer and toolmaker, and all seem to follow the narrative of Roberts himself, delivered to a House of Lords select committee on 19 May 1851:

1334. The self-acting mule was made in consequence of a turnout of the spinners at Hyde, which had lasted three months, when a deputation of masters waited upon me, and requested me to turn my attention to spinning, with the view of making the mule self-acting. l said that I knew nothing of spinning and therefore declined it; they called a second time, that was on the following Tuesday; I declined again; but before seeing me on the third Tuesday, they saw my partner, the last Mr. Thomas Sharp, and requested that he would do what he could to induce me to turn my attention to it; on the third visit which they made I promised to make the self-acting mule [PP, Inventions general, 1: 1829 - 51] .

Smiles puts matters thus:

Mr. Roberts, notwithstanding his reticence, had been occupied in carefully pondering the subject . . . the very difficulty of the problem to be solved had tempted him boldly to grapple with it, though he would not hold out the slightest expectation to the cotton spinners of his being able to help them in their emergency until he saw the way perfectly clear. It was arranged that he would proceed with it at once, and after a close study of four months he brought out the machine now so extensively known as the self-acting mule. The invention was patented in 1825, and was perfected by subsequent additions, which were also patented [1863, pp. 267-9].

The device of 1825 was improved over a period of five years, with a further patent being taken out in 1830; the development effort was a major one, for the firm spent £12,000 developing the second patent (Ure, 1835, p. 368; Singer et al., 1958, p. 288; Catling, 1970, p. 64). It apparently achieved its major aim: Ure quotes a statement by Sharp, Roberts and Co. on the advantages of the self-actor:

First, the advantages connected with spinning. The saving of a spinner's wages to each pair of mules, piecers only being required, one over-looker being sufficient to manage six or eight pairs of mules upwards [1861, 2, p. 156] .

Or as Baines puts it, more bluntly: 'One of the chief advantages of this machine to the spinners is, that it renders them independent of the working spinners' (1835, p. 266). It may well be, to engage m some speculation, that some of the more recalcitrant spinners recognized the seriousness of the threat from the very beginning: a fire at the Sharp Roberts works at that time 'was believed to be the work of an incendiary' (Axon, 1825, p. 170).

The invention was described in laudatory terms in the Manchester Advertiser of 8 November 1825, the article claiming that the new device was not only automatic, but simple in construction, fast in operation, and little more expensive than the hand mule:

'we have to announce its complete success—a success, indeed, so decisive, as to astonish even those who were acquainted with the extraordinary talents of the inventor' [cit. Rose, 1825, p. 2] .

In view of these virtues, one might have expected the self-actor to diffuse rapidly throughout the industry, and to succeed in its aim of subjugating the skilled spinner. The actual results are ambiguous: diffusion was relatively slow, principally because — as noted above—replacement of the spinner was not simply a change in a technical aspect of the production process; it also involved reconstructing the managerial functions performed by the spinner. As William Lazonick has shown in a detailed study (1979) this was for various reasons out of the question, and the production management system of sub-contracted spinner exercising supervisory powers was retained. This limited the possibility of cost reductions in spinners' wages, which in turn was a major component, as I have argued elsewhere (Bruland and Smith, 1981b), in inhibiting the diffusion of the self-actor. Nevertheless, spinners' wages were relatively depressed in the period following the innovation of the self-actor, strikes and industrial conflicts declined markedly, and the self-actor was used as a 'deterrent weapon' against a resurgence of spinners' militancy (Bruland and Smith, 1981b, pp. 14 - 33).


2 Calico printing machinery

Ure, in The Philosophy of Manufactures, gives as an example of the labour-taming effects of 'science' (when suitably deployed by capital), the machines which 'render calico printing an unerring and expeditious process':

This elegant art, which embodies in its operations the most exquisite problems of chemistry as well as mechanics, had been for a long period the sport of foolish journeymen who turned the liberal means of comfort it furnished them into weapons of warfare against their employers and the trade itself [1835, p. 369].

The operative printers did this, according to Ure's rather bizarre metaphor, in 'the spirit of Egyptian task-masters' until the resources of 'science' ensured that the capitalists 'were speedily re-instated in their legitimate rule, that of the head over the inferior members' (ibid., p. 369). Despite the Manichean imagery and the medieval conception of the social order, Ure's account, though compressed, is not inaccurate. The expansion of cotton output in the eighteenth century had placed considerable pressure on some subsidiary processes which were, for technical, or technological, or organizational reasons very difficult either to expand or speed up; in bleaching, for example, the constraints imposed by a process based on large quantities of sour milk are obvious, and led to successful experimentation with sulphuric acid. This was also the base with calico printing, which was an important part of the general cotton production process, for the very elastic demand for cotton cloth was not simply a function of its cheapness, lightness, Nashability, etc., but also of its colourfulness (the way it contracted, for example, a world of fashion accessible to those other hen the very rich). But in calico printing as traditionally practiced - as described in detail, for example, in O'Brien's Calico Printers assistant of 1792—there lay formidable obstacles to any rapid expansion of output: obstacles which were partly technical in :character, but also managerial. The technical process is described n Baines (1835); his account is worth quoting at length, for it implies some of the social or organizational difficulties:

The old method of printing . . . was by blocks of sycamore, about 10 inches long by 5 broad, on the surface of which the pattern was cut in relief, in the common method of wood engraving. On the back of the block was a handle by which the workman held it; the surface was applied to a woolen cloth stretched over a vessel containing the colour, so as to be saturated by it, and was then laid upon the piece of cloth (there being wire points at the corners of the block, to enable the workman to apply it with exactness) and struck with an iron mallet. Thus the figure was impressed upon the cloth, one colour only being used at once; and if other colours were required to complete the pattern, it was necessary to repeat the operation with different blocks. In order to produce more delicate patterns than could be engraved on wood, copper plates were introduced in the neighborhood of London, and the cloth was thus printed from flat plates, with the kind of press used in copperplate printing. Each of these modes was tedious, as no more of the cloth could be printed at once than was covered with the wooden block or copper plate; and a single piece of calico, 28 yards in length, required the application of the block 448 times [1835, pp. 264 - 5].

Such a process was not only very slow, but dependent on the skills of the particular workmen for the alignment of blocks and the general quality of the work, for within the constraints imposed by the hand nature of the process, control of quality and rate of output was in the hands of the operative printer. Now this was not necessarily an obstacle to increased output or expansion of the sector; it can be seen even from the account above that the skills involved in printing were far from esoteric, and certainly far less than would have been involved in actually carving the blocks. Although the process was utterly dependent on the human operative, it was not dependent on particularly complex training or judgment or experience of that operative. This means that it must have been at least theoretically possible for output expansion to have occurred through 'capital widening' without running up against a skilled labour constraint; that this occurred, and also exercised a downward pressure on wages, is clear from the following passage in the text by O'Brien mentioned above. He is speaking of the firm Livesey, Hargreaves and Co.:

the price of labour was latterly reduced as much as possible; by converting (as done at other places) herds of Lancashire boors into drawers, cutters, printers, machine workers, etc., and the work was latterly proportionally execrable [Treatise on Calico Printing, cit. Turnbull, 1951, pp. 72 - 3].

Whether or not the quality deficiencies were due to the employment of 'Lancashire boors', it is pertinent to ask whether in fact there were impulses to expansion and cost reduction at the level of the firm in this sector, and, if so, what precisely were the obstacles to such expansion.

Calico printing, like other aspects of the general production process of cotton goods, expanded rapidly in the late eighteenth and early nineteenth centuries; Potter, in a lecture on calico printing given to the Society of Arts in 1852, estimated from excise returns that production expanded from about 50,000 pieces (each of about 30 yards) in 1750, to 1,000,000 pieces in 1796, to 7,000,000 in 1821. By 1851 total production reached 20,000,000 pieces (Turnbull, 1951, p. 81). At the same time competition was intense: Graham's extraordinary manuscript account of the industry (Graham, and) gives a detailed history of almost every calico printing firm in the Manchester area over a period of about a century to 1846, lists many hundreds if not thousands of enterprises, and details the bankruptcy of most (with many works taking part in more than one bankruptcy). Turnbull quotes a famous remark of Thompson, who has been described as 'the Duke of Wellington of calico printing', to the effect that 'once you become a calico printer there are but two courses before you—the Gazette or the grave.' In large part these competitive pressures derived from the low capital costs in, and hence ease of entry to, the industry. In fact, the competitive pressures within calico printing were even more extreme than those in spinning, weaving or bleaching, because the finished product was subject, when it finally reached the market, not just to the vicissitudes of the trade cycle and price competition but also the changing dictates of fashion:

While at times a very lucrative occupation, the fluctuation from one extreme to the other was very rapid indeed, depending largely on the caprice of the reigning fashion and the temper of the merchants. During the busy times, when orders were accumulating, the hours of employment were excessive, owing to the slow rate of production, especially in blockprinting [Turnbull, 1951, p. 201] .10

The 'seasonality' would pose major problems for a 'capital-widening' approach to output expansion, for it would require a permanent pool of adequate and available labour; but this need not necessarily have been an insuperable problem, for some of the evidence considered above indicates a fairly elastic labour supply. However, there was another, rather more intractable, obstacle to such expansion, namely the operative printers and their union organization.

Calico printing developed as an industry not in Lancashire but in London during the seventeenth century; from the outset the operative printers were well organized, but this organization owed a great deal to what Turner has called 'the availability of a tradition of regulating working conditions derived from a previous age'. That tradition, in London, was the form of organization of the medieval craft guild, the association of artisans and journeymen who regulated quality of output and quantity of entry into the trade, through a system of apprentices. As the 1807 pamphlet 'History of the Combination of the Journeymen Calico Printers' remarks, the Londoners 'had no corporation laws, yet they made use of all possible means to keep the business to themselves' (Anon., 1807, p. 18). Enterprises which, in the early days of the Lancashire industry, set up in the north used these workmen:

the works were first supplied with men from London, they appeared in rather curious costumes having Cocked Hats: Scarlet Coats and Top Boots and in other respects being very aristocratic in their ways holding the Lancashire men in great contempt who were learning the business [Graham, and, p. 345].

And it is quite clear that the Lancashire men learned not only the trade but also the union aims and forms of struggle from the Londoners, for from the time of the first Union of Block Printers (formed, according to Turnbull, in 'the last few years of the seventeenth century') the aims of the printers—regulation of entry and earnings, etc.—were strongly guild-like in character. As such, they were heavily criticized:

By their combination they prevent the master employing any journeymen they do not approve of, who as they say is not a fair man; and all journeymen must ask the constable of the shop, for the time being (an officer appointed by the combination) for work, before they ask the master. They can discharge a journeyman from service without his master's consent; they can advance their wages and, in many instances, prevent the master from making reasonable abatements for spoiled work. While they are all at work they prevent the master taking more apprentices than they approve of, and such a number, as the nature of the work requires. In any case, when the master does not show a readiness to comply with their commands, they order both journeymen and apprentices, to strike and turn out (as they term it) [Anon., 'Facts and Observations . . .', 1807, p. 7] .

The same author goes on to discuss in some detail the use of fines and premiums by the union both to discipline its members and maintain its strike funds. (e.g., p. 17). The union was a very active one: in 1790 it mounted a 'general turnout' (against an attempt by the masters to prevent journeymen moving from one employer to another) which lasted up to six months in some places; although it ended in a partial defeat for the union it did not particularly affect the wage position of the printers (Turnbull, 1951, p. 185). As well as such spectacular strikes, there appears to have been a continuous undercurrent of struggle which occasionally ruined even major firms: the collapse of Livesey, Hargreaves and Co., a major bankruptcy, was blamed on the 'Mosney strike' of 1786, and in Graham's text it is not uncommon to find passages such as this:

Sunnyside, near Burnley, John Brooks and Co.... Has made a very large concern and done a very extensive trade, have had great contests with the block Printers union . . . had the military billeted in the neighbourhood for the safety of life and property, much ill feeling was engendered on both sides for several years [nd, pp. 431 - 2].

This is the context in which the mechanization of the industry should, ~ think, be considered. An inherently slow production process was exacerbated by its dependence on a stratum of workers whose possibly replaceable skills were offset by a formidable union organization, which in turn created problems in achieving quality and regularity of output especially in periods of expansion; a sine qua non for the breaking of this organization and its restrictive effects on enterprise expansion was a change in the technical process itself. And this occurred chiefly through the invention of a printing machine based on a cylindrical copper roller. The major change has been described as follows by Baines:

The grand improvement in the art was the invention of cylinder printing, which bears nearly the same relation in point of despatch to block printing by hand, as the throstle or spinning mule bears to spinning by the one-thread machine. This new mode of printing may be thus described: A polished copper cylinder, several feet in length (according to the width of the piece to be printed), and three or four inches in diameter, is engraved with a pattern round its whole circumference, and from end to end.... It is then placed horizontally in a press . . . the colour being left only in the engraved pattern, the piece of calico and muslin is drawn tightly over the cylinder, which revolves in the same direction, and prints the cloth. [1835, pp. 265-6].

The cylinder printing machine was in fact the most spectacular breakthrough in a general innovative effort, which included other devices such as improved and mechanized blocks, and so on. The introduction of the cylindrical machine was, as we shall see below, strongly opposed by the operative printers. But before examining that, I want to ask the question, what was the relation between the struggles of the printworkers and the print masters, and the innovation and diffusion of the above device, and devices like it? Ure's answer was, as we have seen, that there was a direct causal relation, that 'capitalists sought deliverance from this intolerable bondage in the resources of science' (1835, p. 369). Although it is not clear, in fact, whether he is talking of the Bell machine, his contention is supported elsewhere in the literature of the period. In one document it is suggested that entry restrictions imposed by the journeyman printers had induced product modification and mechanization:

It is a certain fact, that the restraining of masters from taking apprentices, has induced many to print great quantities of single colours, on machines, than they would otherwise have done [Anon., 'Facts and Observations', 1807, p. 9].

The 'History of the Combination . . .' offers an account of the struggles in a single firm, Livesey, Hargreaves and Co. (who subsequently bankrupted spectacularly) between London journeymen and the Lancashire management, an account which traces a whole cycle of struggle, mechanization, and further struggle against mechanization:

these journeymen had so long complained of the country people doing their work low, their great object was to advance wages, and to keep them as high in the Country as in London . . . but this was not the worst, so ill disposed were these journeymen to the Country masters, that they executed their work in a very imperfect manner, and would often tell the overlooker, when he complained of bad work, that it was good enough for the country.... This was the principal cause of (the house's) ruin; for though the execution of their work cost such enormous sums of ready money, yet it was so imperfectly done, they were constantly pushing the markets with quantities of spoiled work, which they often sold for less than the price of the cloth and the amount of the duty. This soon brought the house into an embarrassed situation. The acting partners now attempted to lessen the expense of the work by means of machinery and apprentices, but the journeymen, who to prevent the use of machinery and the taking of apprentices, entered into a combination, 'turned out', and succeeded in taking from their work the apprentices also, for the purpose of enforcing their demands [Anon, 'History of the Combination . . .', 1807, p. 14].11

This passage describes, in fact, the Mosney strike, one of the first important priming strikes in Lancashire (Turnbull, 1951, pp. 72-3). Put simply, the problem for the employers was one of control: control of labour costs, and control of quality and quantity of output; the quality aspect is particularly important, for 'lessening the expense of the work' would have been not simply a matter of replacing the journeymen printers with cheaper labour, but of doing so in a way which did not detract from the quality of the product. Given the choice of machinery or apprentices, one might expect that machinery would have certain quality advantages (see, for example, the diatribe quoted above on the quality of the work produced by the 'herds of Lancashire boors': it was 'execrable').

But perhaps the most explicit account of the relation between problems of discipline and struggle, and the introduction of machinery, from the employers' point of view at least, can be found in the 1815 pamphlet 'Considerations Addressed to the Journeymen Calico Printers by One of their Masters'. The anonymous master begins with a recitation of the sins of the ('you set all subordination and good order at defiance', etc.), and then goes on to ask and answer, with a certain rhetorical flourish, the main question addressed in this study:

What has principally contributed to the increase of cylinder and surface machines, so that there are twenty now, where, fifteen years ago, there was only one?—Your turbulence. What has given birth to the union machinery within these last eight years;—Your turbulence. What has produced, and is daily producing, all the varieties of two, and three, and four coloured cylinder and surface machines?—What but your turbulence? The machine makers say you are their best friends; and well they may, for they have been supported and encouraged by your proceedings [Anon., 'Considerations', 1815, p. 6].

The struggle here was clearly not one-sided, for according to the master quoted here the printers did have some success in the implementation of the various machines (Anon., Considerations', 1815, pp. 4 - 5); and it should be remembered that the Livesey, Hall and Co., who according to Baines first introduced the cylinder machine, and who, according to the 'History' quoted above, were engaged in such a major struggle with their journey men, were the same firm whose collapse 'convulsed the whole country for some time' (Graham, nd, p. 420). There were a number of strikes against the new machinery, but by 1844 we find Rashleigh writing of 'the block printers . . . 4,000 of whom were thrown out of work by new machinery, and the remaining 3,000 of their body not half employed (Rashleigh, 1844, p. 82 ). Of his period Turnbull remarks 'by now the block printers were fighting a stern battle against other forces (i.e. other than anti-union legislation) - forces against which they were powerless to compete — the introduction of the printing machine (195, p. 190). The struggle against machinery seems to have been a major component of the strong support for Chartism among block printers (Turner, 1962, p. 103); and like Chartism the printers' union declined dramatically after the 1840s, with Lancashire membership declining from 5,337 in 1845 to a mere 429 in 1888.


3 Devices for wool-combing

In so far as the history of early British industrialization is written around the textile sector, it is usually cotton which is accorded pride of place: Hobsbawm remarks that 'Whoever says Industrial Revolution says cotton' (1972, p. 56), and Landes is typical in deploying cotton statistics as a graphic indicator of the pace of industrial advance (1978, pp. 41 - 2). But the quantitative importance of cotton should not lead us to think that similar qualitative changes (in the organization of enterprises, or of direct production or technologies) were not taking place elsewhere. In fact, as Landes himself points out, the first of the major inventions which transformed textile production—Kay's flying shuttle, and Wyatt and Paul's spinning frame—were designed for the woolen manufacture (1978, p. 84); and we can trace, within the woolen sector, many of the forces and relationships which have been noted, in the two sections above, as characterizing cotton.

During the eighteenth century the industry had major centers in London, Norwich and Yorkshire; in the last half of the century, the West Riding rapidly became the major producing area. The competitive pressures were intense Sigsworth, in a preface to a catalogue of woolen enterprise records, mentions almost a thousand firms located in the West Riding (Hudson, 1975, p. ix). 'Here also,' says James, 'before the termination of the century, the inventions of spinning machines were first applied to the production of worsted yarn, and were destined ere long to produce such a remarkable transformation and growth in all the departments of the trade' (1857, p. 258). The major organizational change was a transition from an outwork-based system to factory production; from the beginning there was a strong workers' organization, founded during the period of domestic outwork:

The very early development of workers' combinations in the wool, worsted and finishing trades, arose, of course, from the growth in them (before a significant cotton industry existed) of a capitalist domestic and workshop economy . . . the woolcombers" national confederation was one of the most formidable of the later eighteenth-century labour organizations—in 1801 it was alleged to have 60,000 members [Turner, 1962, p. 174].

It is with these wool-combers, and the machinery that replaced them, that this section is concerned.

Wool-combing is one of the preparatory processes of worsted manufacture. It is succinctly described by the Hammonds:

Before worsted can be spun the fibres of the wool must be laid in a parallel direction, and this work used to be performed by the wool-combers, whose stock in trade was two hand combs with two or three rows of teeth apiece, a stove at which to heat the combs, and a post on which to fix one of them. The work was hard and skilled, the atmosphere in which it was done was generally vitiated by the fumes from the stove [1967, p. 195].

A complete description of the process can be found in James (1857, pp. 249 - 50): the chief technical difficulties resided in adjusting the combing process to the characteristics of different wools, with different degrees of tangle, etc.; in the simultaneous manipulation of two combs, with a delicate and constantly changing relationship between the teeth of the two combs; in the adjustment of appropriate temperature levels of the combs out the multi-stage process. The skill of the wool-comber lay in the ability to carry out this process such that the spinnable wool was undamaged and prepared for spinning, and the rest was in suitable state for baize and coarse cloth manufacture, or for ordinary spinning on the woollen system.

The combers' union was a formidable and ancient one, with aims and methods based (like the calico printers' combination) on aspects of guild organization. Throughout the eighteenth century they had a reputation for uncompromising militancy. The following 1741 pamphlet, by an anonymous 'Lover of his Country', is worth quoting, for one finds similar descriptions and charges until well into the nineteenth century:

Our combers have for a number of years past, erected them selves into a sort of corporation (tho' without a charter) . . . and when they had become a little formidable, they gave laws to their masters, as also to themselves, viz.:—that no man should comb wool under two shillings per dozen; that no master should employ any comber that was not of their club, if he did they agreed one and all not to work for him; and if he employed twenty, they all of them turned out, and often times were not satisfied with that, but would abuse the honest man that would labour, and m a riotous manner beat him, break his combpots, and destroy his working tools; they further support one another, insomuch that they are become one society throughout the kingdom [cit. James, 1857, p. 232].

The litany of combers' activities covers almost the whole industrial conflict. The complaints begin with entry and high wage rates:

No woolcomber was permitted to take n apprentice except his eldest son, and they not only dictated their own rate of wages but sought to prescribe the prices which the masters should ask for the products of their labour [Burnley, 1889, p. 164].

And they go on, more generally, to depict the wool-combers as both industrially and politically dangerous:

Since the commencement of this century, the wool-combers had become a powerful, organised body, who frequently by their commotions, strikes and insubordinate conduct, occasioned much difficulty with the masters . . . a well informed class, (they) have been memorable for strikes and general improvidence, and strongly impregnated with doctrines of the democratic school [James, 1857, pp. 250, 559] ,14

So powerful were the combers that specific statutes were passed against them, such as 12 Geo. I, ch. 34, 'An Act to Prevent Unlawful Combinations of Workmen Employed in Woollen Manufactures. But such legislation was largely unsuccessful.

Wool-combing was, because of its technical structure and the degree of control exerted by workers, a major bottleneck in the worsted process. As with the automatic mule, I think one can see the scale of the inventive effort directed at overcoming this bottleneck as an indicator of its importance. Mann (1860) lists seven major devices invented between 1792 and 1852; James remarks that

a volume would be required to describe in detail the numerous complicated contrivances for combing wool which, during the last sixty years, have been patented, some of which never into practical use; and others again are only remembered as failures after short trials [1857, p. 564].

The first significant machine was invented by Edward Cartwright of Doncaster, and patented in 1790; it was subsequently improved by a Nottingham mill-owner called Hawksley, but in neither version was it particularly successful. The first really practical device was the so-called Platt and Collier machine, patented in 1827; further advances were made by Donisthorpe in the 1840s, and then by a Frenchman, Heilman, who was able 'to effect a total and vitally important change in the principle and working of combing machinery' (James, 1857, p. 567). Donisthorpe and his partner, Lister, continued to develop in the 1850s, applying for patents which bore a suspicious similarity to Heilman's principle; litigation ensued, which ended with Lister, who had become sole owner of the Donisthorpe/Lister patents, buying—though never bringing into production—the Heilman patent. This period marks the emergence of the first technically adequate combing machines; subsequently others were produced, most of them attracting a writ from Lister.

Both the invention and diffusion of the various wool-combing devices were conditioned by the militant struggles of the combers' union, and the consequent inability of the mill-owners to control the production process. Tufnell writes of a union demand (apparently for the dismissal of non-union members) that 'Compliance with this request would have deprived [the company] of the control necessary to carry on their business', and says of the accompanying strike

This strike was the cause of the invention of the wool-combing machine, which wholly superseded the labour of that class of men, who were the chief ringleaders in this affair, and which has struck a blow at their combination, that it can never recover. The highest credit is due to Messrs. Hindes and Derham for the spirit in which they resisted the tyranny it was attempted to impose on them [1834, pp. 61 - 2; my emphasis] .

James also presents a connection, less direct than in the case above, between the activities of the union and the diffusion of the Cartwright machine:

Like the wool-combers of Norwich fifty years before, those of the West Riding were a turbulent ill-ordered class, and occasioned much trouble to their employers . . . thus when Cartwright's combing machine became public, the manufacturers eagerly availed themselves of it, in order to dispense with the services of the hand-comber [1857, p. 322].

In fact, this machine had its technical limitations, and we subsequently find James remarking, of a major strike in 1825, that 'as one of the results of this strike, attention was again directed to the combing machine (1857, p. 406). Lipson repeats James's verdict on this strike: 'The only result of this ruinous contest was to stimulate the introduction of machinery. The year 1825 marks the turning point in the history of the hand-combers, whose condition now underwent rapid deterioration' (1921, p. 204) However, about ten years after this supposedly decisive strike, in 1834, we find Tufnell saying:

The Wool-Combers' Union has been celebrated above a century, and several Acts of Parliament have been passed with the object of suppressing the power which it had acquired, and exercised with the usual bad consequences. Hence many endeavors have been made to comb wool entirely by machinery, but with very partial success, till last year, when the whole of the combers in a large factory struck, upon which the proprietors turned their attention to this machine, applied their skill and capital to its improvement, and in a short time brought it to such perfection, as to completely supersede the employment of wool-combers [1834, p. 112].

In James's text, thirty pages on from the remarks quoted above, we find him agreeing with Tufnell's chronology, thus associating the impetus to mechanization with the 1832 strike:

the year 1832 witnessed another wool-combers' strike . . . but it was the last strike of the wool-combers of any importance, for it compelled the masters to seek in the combing machines, which had recently been improved by Messrs. Platt and Collier, relief from the intolerable conduct of their workmen [1857, p. 435].

As it happens, all of these texts represent a wish rather than a fact, for the transition to mechanized combing was nothing like as abrupt or decisive as they imply; as late as 1852 we find Forbes writing that wool-combing machinery 'is just at present coming into use' (1852, p. 315). But even if the objective of replacing the combers took many years to achieve decisively, at least that objective is reasonably clear as a component of the innovation decision. In fact, on this matter, we have an explicit statement by one of the aspiring inventors, who was himself a manufacturing employer. The Hammonds, in The Skilled Labourer, quote at great length a petition to Parliament by one William Toplis, a Nottinghamshire worsted manufacturer, published in the House of Commons Journal, 31 March 1794. Part of the petition reads as follows:

for several years past, the Petitioner has employed from 100 to 150 woolcombers in combing Wool, which he has spun into worsted yarn by Machinery, and that, experiencing great Inconveniences from the Inadequate number of Wool-combers for the purposes of trade, and from their irregular and improper conduct in forming themselves into Societies and Combinations, the Petitioner, after great Labour and Study, invented a machine for combing wool [Hammond and Hammond, 1967, p. 198].

The combers resisted mechanization with all the means at their disposal. These ranged from Parliamentary activity ('upwards of forty petitions, from various parts of the kingdom, were presented to Parliament during the course of the present session for [the machines'] suppression; and for this purpose a Bill was brought into the House of Commons by the friends of the Petitioners. It was, however, rejected by a large majority' James, 1857, p. 566), to machine wrecking (Anon., 1812), to the publication of elaborate though not particularly cogent economic arguments against the use of machinery in wool-textile industries as a whole (e.g. Anon., 1803).

The combers' resistance was essentially a rearguard action, doomed to ultimate failure. As a weapon against them, the machine adequately performed its function even in its imperfect versions. Partly this was a matter of threat: Burnley, for example, quotes uncompromising notices issued by the masters during the 1825 Leeds strike: 'No advance—take combers at old prices... The combers had better turn in and do the work than let it be done by Big Ben (the nickname for the Cartwright machine)' (Burnley, 1889, p. 174). This threat had a serious effect on the combers even where the machine was not widely in use (Sigsworth, 1958, p. 9); nevertheless, mechanization proceeded, utterly transforming the division of labour, and by the early twentieth century Clapham could write of wool-combing that it was light and simple work, dominated by the automatic machine, and suited to a workforce of unskilled women (1907, pp. 178, 187, 195).


Conclusion

The historical evidence collected in the previous sections is intended to offer support for the argument that the economic effects (on costs of production, on rates, prices and regularity of output) of industrial conflict are of great importance in understanding the direction of innovative efforts. As I noted in the introduction, this relation was widely remarked upon in the nineteenth century and even regarded as rather obvious:

Notwithstanding the losses and suffering occasioned by strikes, Mr Nasmyth [a major nineteenth-century engineer] holds the opinion that they have on the whole produced much more good than evil. They have served to stimulate invention in an extraordinary degree. Some of the most important labour-saving processes now in common use are directly traceable to them [Smiles, 1863, pp. 294 - 5; my emphasis].

The general point here, as I see it, is not simply that innovation may occur in some cases to destroy production bottlenecks created by the recalcitrance of workers. It is rather that problems of technical innovation — and problems of the development, application and structure of technologies more generally - in capitalist economies, should be located within a clear conception of the social and managerial field of forces underlying capitalist production, and which structure many decisions relating to technique. The calculations of entrepreneurs should be conceived as including this kind of component. Again, there is nothing necessarily obscure about this, for such matters are often referred to by economic historians. Landes, for example, has written that 'the essence of the factory is discipline—the opportunity which it affords for the direction of and co-ordination of labour' (1966, p. 14); and Saul has remarked that 'in engineering, technical change was not just a matter of buying machines, but possibly even more of workshop organization and control' (1970, p. 19). But as Marglin (1976, p. 29) and Hartwell (see below) have pointed out, the supervisory facilities afforded by the factory may have been acknowledged by economic historians, but the implications of this have not been developed, and indeed remain quite in the background in the mainstream of economic history:

The usual analysis is in terms of the economic and social consequences of technological change rather than the technological consequences of economic change [Hartwell, 1971, p. 287].

The machinery question in the early nineteenth century, for example, as Keith Tribe has recently pointed out, is invariably treated in terms of total employment (that is, as a question of whether or not new techniques displace labour) rather than in terms of the serious problems of supervision and control which were really at stake, and which exercised contemporary commentators such as Ure, Gaskell and Babbage (see Tribe, 1981, ch. 3). Locating the process of technical change within this field of discipline and conflict offers just one line of approach to the solution of the real problems in the history of technical change. The limits of such an approach should, perhaps, be emphasized in conclusion: innovation is a complex matter; technical change has a wide variety of determinants and purposes—ranging from broad structural changes in the economy to small-scale problems in specific firms—and a wide variety of forms. These range from decisive breakthroughs' in basic technology to tiny modifications of existing machines. Giving some emphasis to labour problems should not therefore mean giving priority to them.


Notes

I would like to thank K. G. Ponting, Keith Smith, Grahame Thompson, James Thomson and Nicholas von Tunzelmann for advice, comments and criticism in the preparation of this study and on previous drafts; none should, of course, be implicited in the results.

1. With, for example, the development of accounting systems being held to derive from the impact of new technologies; for a brief account of this and similar ideas, see Bruland and Smith, 1981a, p. 92 and fn 5.

2 Rosenberg, 1977; Marglin, 1976; Lazonick, 1979.

3. See also Hobsbawm, 1972 p. 58.

4. See also, e.g., Dickinson, 1945-7, p. 127;Singer, 1958, p. 288.

5. Montgomery, 1886, pp. 199 - 200; Dickinson, 1945 - 7, p. 127.

6. 'the workman is paid by the piece, and not by the day, and in this case the workman hires the children who assist him in attending the machinery of his employer' (T. Jones Howell, PP, Industrial Revolution Factories, 6: 1835 - 41, p 17)

7. See also pp. 17-21.

8. See also Tufnel, 1834, pp. 14-15.

9. See 'Operative Spinners', 1829.

10. In the 1807 pamphlet, 'Facts and Observations', it is remarked that calico printing is 'entirely a fancy trade, and subject to continual changes of fashion' (Anon., 1807).

11. See also p. 20.

12. The 'union' machine was another name for the so called 'surface' machine. See Baines, 1835, p. 271.

13. See, e.g., Turnbull, 1951, p. 85.

14. See also Burnley, 1889, p. 160.

15. This point is emphasized by William Lazonick in his detailed study of the technical division of labour associated with the hand and automatic spinning mules:

attempts by managers to change the technical structure of production on the basis of rational cost-benefit decision-making may lead to conflicts and compromises between capitalists and workers, the nature of which will depend on the existing structure of industrial relations. These conflicts (e.g strikes, sabotage, slowdowns) and compromises (e.g. wage settlements, co-operation over intensification of labour) will change the relative costs and benefits of alternative techniques, perhaps causing managers to make quantitatively or qualitatively different, but equally rational, decisions concerning the technical organization of production than they would have made before the conflicts occurred and/or the compromises were worked out. That is, the ultimate decisions on technical change, made at a 'point in time', cannot be understood in abstraction from the dynamics of capital-labour relations internal to an industry (1979, p. 256n).