PART FOUR
THE WORKERS’ COMPANY

Seven
An Industrial Labor Force

Large companies like the PGC were primarily responsible for the transition from an artisanal to an industrial work force.^[1] The results of the transformation require more investigation than they have yet received, since most labor studies have been concerned with craftsmen and their responses to changes at the workshop. Though historians no longer portray factory workers simply as the victims of the "dark, satanic mills"— utterly atomized, regimented, and undifferentiated—the present state of knowledge about the group is largely a matter of negative statements: industrial workers did not serve as the activists in the labor movement; they comprised a minority within the French working class; they were usually quiescent; when they did act, their protest was generally ineffective and poorly focused.^[2] The industrial army that emerged at the gas plants was complex and underwent dramatic changes even in the absence of technological upheaval. Its evolution provides an opportunity to see factory workers in a more positive light.

[1] For an effort to conceptualize these types of work, see Michael Hanagan, "Artisans and Skilled Workers: The Problem of Definition," International Labor and Working-Class History, no. 12 (November 1977): 28-31.

[2] William Sewell, "Artisans, Factory Workers, and the Formation of the French Working Class, 1789-1848," in Working Class Formation: Nineteenth-Century Patterns in Western Europe and the United States , ed. Ira Katznelson and Aristide Zolberg (Princeton, 1986), pp. 45-70; Géard Noiriel, Les Ouvriers dans fa société française: XIX -XX   siècle (Paris, 1986); William Reddy, Money and Liberty in Modern Europe (Cambridge, 1987), p. 208. Among the recent studies that will lead beyond the negative statements are William Reddy, The Rise of Market Culture: The Textile Trade and French Society, 1750-1900 (Cambridge, 1984); Donald Reid, The Miners of Decazeville (Cambridge, Mass., 1986); Michael Hanagan, Nascent Proletarians: Class Formation in Post-Revolutionary France (Oxford, 1989).

The Stokers

One reason for the failure to come to grips with the shift from artisanal to industrial labor is the complexity of the process. Occasionally employers were able to install machines and obliterate in a single blow craft-based production methods. This seems to have happened at the bottle plant in Carmaux, brilliantly described by Joan Scott. Supine machine tenders suddenly took the place of skilled glassblowers, who had trained for fifteen years to do their jobs.^[3] Such a clean break was exceptional. Indeed, em-ployers even today have rarely succeeded in whittling down labor to a ho-mogenized, easily replaceable body of operatives doing simple, repetitive work.^[4] Nineteenth-century factory work usually involved skilled workers doing crucial operations. Sometimes these were "industrial craftsmen," who had nearly complete control over the work process. More commonly, employers were able to subdivide craft work into simpler, but still skilled, activities. They also took control of hiring and training, usually leaving supervisors with these responsibilities.^[5] Some new industries had never known an artisanal phase but revolved around skilled work of the latter sort. The replacement of artisans and industrial craftsmen by skilled workers was a long and gradual process. It was surely the central development in French labor history between the Commune and World War I.^[6] The stokers who produced gas for the PGC provide an example of the type of factory laborer then coming to dominate production and protest.

The PGC was born rather advanced in terms of the social evolution of the labor force. A process-centered industry, gas production did not re-quire the fine shaping and fitting that assembling industries did. Though stokers performed the manual operations that were crucial to gas production, there was nothing intricate about their performance. Nor were there preindustrial traditions to stoking; the job arose with the nineteenth-century gas industry. Stokers did not have to adapt basic operations to particular circumstances, innovate at the job, or complete a multiplicity of

[3] Joan Scott, The Glassworkers of Carmaux (Cambridge, Mass., 1974).

[4] Patrick Fridenson, "Automobile Workers in France and Their Work, 1914-83,' in Work in France: Representations, Meaning, Organization, and Practice, ed. Steven Kaplan and Cynthia Koepp (Ithaca, N.Y., 1986), 514-547; Roger Penn, "Skilled Manual Workers in the Labor Process, 1856-1964," in The Degradation of Work? Skill, Deskilling, and the Labor Process , ed. Stephen Wood (London, 1982), pp. 90-108.

[5] Michael Hanagan, "Urbanization, Worker Settlement Patterns, and Social Protest in Nineteenth-Century France," in French Cities in the Nineteenth Century , ed. John Merriman (New York, 1981), p. 218; Lenard R. Berlanstein, The Working People of Paris, 1871-1914 (Baltimore, 1984), pp. 92-100.

[6] Reid, Miners of Decazeville , p. 72.

complex operations. They did the same task repeatedly and had only a small degree of autonomy too little for the concept of craft control to be relevant. Engineers and foremen set the parameters of the tasks and modified them at will.^[7] The distinguishing feature of stokers, as skilled workers, was their market power. Not many humans could perform the strenuous work of stoking in the torrid distillation moms. And this was not the sort of work that an employer could allow to stop for very long. If pressure fell and air entered the gas mains, then the danger of catastrophic explosion was great,^[8] Stokers found strategies for exploiting their market power in ways that differentiated them from common laborers and gave shape to the emerging industrial labor force.^[9]

The banner of the union newspaper portrayed the stokers as barrel-chested, muscular giants, stripped to the waist and covered with coal dust and sweat. Outsiders who occasionally visited a distillation room never failed to be impressed by stokers at work. The oppressive heat, the strength required to fill a retort, the dexterity necessary for a neat job, and the physical dangers stokers abided filled visitors with admiration. The journalist Gustave Babin waxed eloquent at the sight of the PGC's stokers in action: "Athletes, for the most part, as they need to be to do such labor, these workers are superb to watch in action, in the play of reflexes they reveal. They have their special skills, their professional dexterities, which are marvelous. With great precision the stokers . . . know how to throw the shovelful of coal into a burning furnace, withdrawing their unprotected arm just as flames are about to lick it. With equally great poise they empty coke from the retort, projecting their heavy iron hooks into the incandescent tube."^[10]

[7] AP, V 8 O , no. 764, "Main-d'oeuvre de distillation (26 avril 1870)"; "Etude des modifications à introduire dans la réorganisation dans le régime du personnel ouvrier de la distillation (29 avril 1870)." These two reports by the head of the factory division are the starting points for understanding the work situation of the stokers.

[8] Ibid., no. 161, report of Euchène, March 18, 1907.

[9] The industry-specific skills of the stokers made their positions analogous to those of the elite of mining teams, the hewers and timbermen, as well as to puddlers in the steel industry. See Rolande Trempé, Les Mineurs de Carmaux , 1848-1914, 2 vols. (Paris, 1971), 1:112; Reid, Miners of Decazeville , pp. 72-78; lean-Paul Courtheoux, "Privileges et misères d'un métier sidérurgique au XIX siècle," Revue d'histoire économique et sociale 37 (1959): 159-189. For the argument that nineteenth-century industry created many new types of skilled workers, see Raphael Samuel, "Workshop of the World: Steam Power and Hand Technology in Mid-Victorian Britain," History Workshop 3 (1977): 6-72.

[10] " Le Gaz d'éclairage," L'Illustration, no. 3084 (April 5, 1902): 251-253. An-other admiring observer was Léon Duchesne, Hygiène professionnelle: Des Ouvriers employes d la fabrication du gaz de l'éclairage (Paris, n.d.).

A stoker firing the furnace. From L'Illustration, no. 3084 (April 5, 1902).

The entire operation of stoking consisted of several distinct activities, none of them easy and each with its particular physical demands and threats. First, the workmen brought several hundred kilograms of coal from the coal yard to the distillation room and pulverized it. The preparation of a hot retort for charging could be hazardous or at least noxious. Gas that had not escaped into the evacuation pipes might back up into the stokers' faces and burn or suffocate them. To prevent that, workers would open the mouth of a retort carefully and ply a glowing piece of coke so as to burn off the remaining gas. Then shoveling could commence. This op-

Stokers at work: shoveling, loading coal with a scoop, and
emptying the retorts. From L'Illustration, no. 3084 (April 5, 1902).

eration involved propelling heavy shovelloads of coal into a retort in such a way as to fill it evenly. Obviously, practice and dexterity were needed. During the four to six hours of roasting, a special stoker, the fireman, kept the furnace filled with fuel (usually coal or coke). In the meantime, other stokers had to pay attention to the evacuation pipes. When they became clogged with condensed residues (especially naphthalenes), a stoker had to clear them—quickly—with steam. Blockage produced frequent emergencies and not a few incidents of scorched skin. When distillation ended, the job of the stokers was to clear the retorts of the coke that carbonization had produced. This operation, one Babin especially admired, was per-formed with a heavy iron pole that had a ring on the end. It weighed eighteen kilograms, and workers had to manipulate it in the long, hot tube in such a way as to remove the coke without unnecessarily scratching the surface of the retort. They used the pole to push the burning coke into a wagon, which they then wheeled to the yard, where it could be extinguished,^[11] In the opinion of some observers, transporting the burning coke was the most hazardous activity stokers did since occasions for being burned were numerous. Reports on work accidents confirm the assessment of danger. At the PGC's largest factory, La Villette, stokers were ob-

[11] AE V 8 O , no. 764, "Etude des modifications."

ligated to push wagons of coke seventy meters back and forth eighteen times during each charge.^[12]

Stokers also had maintenance chores that complicated the job. Firemen were responsible for the care of their furnaces and had to remove the cinders after each charge. Such upkeep had to be done faithfully for the state of the furnace determined how hard the work was and how much stokers earned.^[13] Stokers checked the retorts for fissures in the walls between charges and repaired them by spreading molten earthenware with an iron pipe. The inspection of evacuation pipes was also an important part of the maintenance activities.

Such labor was all the more difficult in that it was performed under debilitating circumstances. Temperatures in the distillation rooms rarely fell below 100 degrees Fahrenheit, and stokers were continually exposed to heat radiated from glowing surfaces. Director Camus cited the heat as the principle reason the job was beyond the capacity of most wage earners.^[14] Work in the distillation rooms during the summer was reputed to be especially painful, but it did have one advantage over winter work: when stokers wheeled coke into the yard during the cold weather, the sudden change of temperature contributed to illness.

For obvious reasons the PGC needed men of rare physical endurance. Ideally, the physical giants who did stoking should also have had the dexterity for packing coke tightly.^[15] In all probability the company could not insist on perfection in packing the retorts and had to settle for men who could do the work with minimal competence. One factory manager used the term "apprenticeship" to describe the training of a potential stoker. The exercise entailed setting up a retort in the courtyard and having the fledgling practice filling it with the shovel. The superintendent went on to

[12] Ibid., no. 717, 'Voles ferrées"; no. 151, report of Aline to director, May 11, 1897; no. 718, report no. 120, June 21, 1873. There were easier and less dangerous ways to handle the burning coke, but the company chose the method that would preserve the most value for its sale to domestic customers. For the work accident records of the La Villette plant, see AP, V bis 19 Q⁶ .

[13] AE V 8 O , no. 764, "Etude des modifications"; no. 151, report of Euchène, November 11, 1898.

[14] Compte rendu des travaux de la commission nominee par Monsieur le Ministre de l'Intérieur le 31 janvier 1890 en exécution de l'article 48 du traité entre la ville de Paris et la Compagnie parisienne de l'éclairage et du chauffage par le gaz (Paris, 1890), p. 139.

[15] One superintendent wrote, "The proper or improper execution of charging determines the yield of gas from a given quantity of coal and at a given temperature." Inexperienced stokers raised the cost of production 20 percent during the strike of 1890. AE V 8 O , no. 148, report of Cury to director, June 25, 1865; report of Gigot to director, October 8, 1892.

underscore the overwhelming importance of endurance by pointing out that such an apprenticeship was "as much physical as manual."^[16] He no doubt meant that building stamina was crucial. The superintendents did not expect more than one in ten common laborers to be able to perform the infernal work.^[17]

The harsh conditions meant that a stoker could not grow old in this line of work. The PGC's engineers professed to know hardly any who worked at the job for more than fifteen years or beyond the age of fifty.^[18] Stokers inevitably spent the last third of their working lives as common laborers. If they tried to extend their careers at the furnace a year or two beyond their capacity, they paid dearly. One stoker wrote poignantly and from experience that "the last years on the job are the hardest times in a stoker's life."^[19]

The evolution of stoking within the PGC illustrates how a job characterized by stable production methods could undergo profound transformation. A stoker of 1850 would have been disoriented by the work his comrade performed thirty years later. The very meaning of the stokers' work in relation to the rest of their lives had shifted over that generation. We can understand the transformation as a change from relative independence from the gas industry as a source of employment to dependence on the industry, and on the PGC in particular, for a livelihood.^[20]

By any standards and under the best of circumstances the stoker's work was debilitating, yet it became far more so during the PGC's first decade of operation. The rationalization of production during the 1860s was responsible for the deterioration of conditions (see chapter 4). At that time the company became aware of the benefits of hot roasting. Raising the temperature of distillation allowed the firm to increase the number of charges per day, the size of the retorts that had to be filled, and the amount of coal in each retort while still producing gas of sufficient quality for lighting. The implementation of hot roasting caused unit labor costs to plummet and output per stoker to soar, all achieved without any mechanization whatsoever. When the PGC began operations in 1855, its factories completed only four charges in a twenty-four-hour period. Each furnace had no more than four retorts, and sometimes only one, to heat.

[16] Ibid., no. 148, report of Jones, April 26, 1865.

[17] Ibid., "Chantiers à coke"; "Union des agents principaux des chantiers h coke (19 novembre 1892)," p. 6.

[18] Ibid., no. 151, "Etat nominatif des chauffeurs ayant été conservés tout I'été."

[19] Ibid., no. 150, "Affaire Mercier."

[20] I have borrowed the concept of dependency from David Brody, Steelworkers in America: The Nonunion Era (New York, 1960).

Those vessels were short (two meters in length) by later standards.^[21] Stokers did not have an easy job, but the demands of work were far less than they would soon become. An indication of the endurable pace of labor before 1860 was that stokers did twenty-four-hour shifts, working every other day. They were able to do so because they had time to rest and sleep (on piles of coal) between the six-hour roasting periods.^[22]

The rationalization of production initiated by factory chief Arson soon brought the number of charges per day up to six, one every four hours. Stokers packed 100-120 kilograms instead of 75 kilograms of coal into each retort for each charge. Thus, with hot roasting the amount of coal stokers handled per day nearly doubled between 1860 and the late 1870s and continued to increase through 1905:^[23]

Maurice Lévy-Leboyer has argued that gains in productivity during the Second Empire usually resulted from intensified deployment of labor rather than from an increase in capital, and such was the case for stoking at the PGC.^[24] The higher quantities of coal roasted depended entirely on the stokers' exertions.

Not only were stokers working much harder; they were doing so under increasingly adverse conditions. Hot roasting raised the heat under which stokers had to exert themselves, all the more so when Siemens furnaces were introduced. Superintendents reported temperatures of 120 degrees Fahrenheit in the distillation rooms when it was only 56 degrees outside.^[25]

[21] AP, V 8 O , no. 709, report of Servier to director, May 7, 1860.

[22] Ibid., no. 148, report of Letreust, April 29, 1865.

[23] Ibid., no. 763, "Tirage des cheminées (28 avril 1869)'; nos. 148, 151, 159, 163, 709, 711, 716, 768, 1520.

[24] Maurice Lévy-Leboyer, "Capital Investment and Economic Growth in France, 1820-1930," in The Cambridge Economic History of Europe (Cambridge, 1978), 7 (1): 267.

[25] AP, V 8 O , no. 675, deliberations of May 8, 1872. Another report put the temperature at 115 degrees Fahrenheit when it was 75 degrees outside. Ibid., no. 768, "Usine de Saint-Mandé" report of April 18, 1869.

Moreover, the greater number of charges per day necessitated quicker emptying of the retorts and unloading of the coke. It also made the job of clearing the retorts more difficult because the burning coke dung to the sides.^[26] At the same time, maintenance chores became more troublesome and time-consuming. Hot distillation caused a much more frequent clogging of the evacuation pipes. Furnaces and retorts were also harder to keep in good repair. Rest time between the more closely spaced charges was thus severely squeezed. Indeed, some observers asserted that there was no longer any rest time.^[27]

The company contrived to rearrange work patterns so as to derive all the benefits it could from hot roasting. It broke down some of the stoker's work into specialized operations. Up to the 1850s stokers, working in teams, would take turns charging, clearing, moving the coke, and firing the furnace. Each of these steps had its distinctive demands, and stokers tried to even the burdens by exchanging tasks. The company forced the stokers to specialize in one of these operations.^[28] The introduction of the Siemens furnace was another way the company gained more control over production. With simple furnaces the temperature of distillation de-pended entirely on the firemen: they could sabotage hot roasting at will by withholding fuel from the furnaces. The Siemens furnaces, however, were not really under their control.^[29] Another change in the stoker's work pattern was the abandonment of the twenty-four-hour shift. Workers could not exert themselves for so long over so many consecutive hours, and the twelve-hour shift became the rule (except in one or two plants). Workmen resented this change because the twenty-four-hour shift had allowed them to rest or take other jobs on their off days.^[30]

Thus, by 1870 the work of stokers had deteriorated drastically. Indeed,

[26] Compte rendu des travaux de la commission nominée . . . le 31 janvier 1890 , p. 39. The engineers who authored this report affirmed that stokers much preferred to work with ordinary furnaces.

[27] AP, V 8 O , no. 670, deliberations of May 25, 1864. Babin, the journalist who observed the stokers at work, remarked on the absence of a rest period. "Gaz de l'éclairage," p. 14. Even a superintendent was skeptical that stokers had much of a rest between charges. See AP, V 8 Oé, no. 160, report of superintendent of Ivry plant to Euchéne, July 16, 1899.

[28] AP, V 8 O , no. 764, "Main-d'oeuvre de distillation (26 avril 1870)"; no. 709, report of Servier to director, May 7, 1860.

[29] Ibid., no. 711, report no. 43, "Règlement du salaire de la distillation"; Compte rendu des travaux de la commission nominee le 4 février 1885 en execution de I'article 48 du traité intervenu le 7 février 1870 entre la ville de Paris et la Com-pagnie parisienne de I'éclairage et du chauffage par le gaz. Procès-verbaux (Paris, 1886), p. 189.

[30] AP, V 8 O , no. 764, report of April 29, 1870 (fols. 18-21); no. 768, "Usine Passy,' report of December 2, 1860.

it is hard to find a similar tale of deterioration in the annales of French labor history Skilled workers were not usually passive as their work be-came degraded. They fought and usually drew employers to a standoff long before their work load had increased to the extent the stokers' had. Were stokers an anomalous case, or did they exemplify the helplessness of industrial workers—even skilled ones—in the face of a determined employer?

To the extent that historians have examined that industrial work force, they have concluded that the group was helpless either to resist employers' demands or to find an organizational and ideological basis for the resistance.^[31] The sociologist Craig Calhoun provides a theoretical argument that purports to explain the malleability of the factory operative in contrast to the craftsman. Calhoun's formulation shifts the emphasis from resistance to adaptation.^[32] He posits that skilled industrial workers, unlike artisans, were not inalterably opposed to industrial development by their very way of life. Whereas artisans had to defend the craft basis of production, skilled factory workers could use their ability to create bottlenecks at crucial steps in production to bargain with their employers. Calhoun believes that industrial workers naturally took advantage of capitalistic development to obtain improvements in their earnings and living conditions. Calhoun's model of bargaining is applicable to the stokers, who were not at all helpless before their bosses. They did use their market power to obtain immediate material gains in exchange for greater output. Such comportment, however, characterized stokers in a particular phase of their development. The model of adaptation was not relevant under all circumstances, and stokers altered their strategies as the role of work for the PCG in their lives changed. Calhoun did not see that skilled industrial workers, French ones at least, had the capacity to become more like craftsmen in their confrontations with managerial authority.

The shop floor in and of itself provides too narrow a perspective for understanding how stokers evaluated their options on the job. Their re-action to the rationalization of gas production depended, at least in part, on their way of life.^[33] Stokers were one example of the sort of factory

[31] Scott, Glassworkers of Carmaux, chap. 7; Yves Lequin, Les Ouvriers de la région lyonnaise (1848-1914), 2 vols. (Lyon, 1977), 2:242-348; Michael Hanagan, The Logic of Solidarity: Artisans and Industrial Workers in Three French Towns, 1871-1914 (Urbana, Ill., 1980).

[32] Craig Calhoun, The Question of Class Struggle (Chicago, 1982).

[33] Michael Hanagan demonstrates the importance of integrating a consideration of the workers' private lives into their work histories in "Proletarian Families and Social Protest: Production and Reproduction as Issues of Social Conflict in Nineteenth-Century France," in Work in France , ed. Kaplan and Koepp, pp. 418-456.

workers who did not rely on one and only one kind of labor for permanent, year-round employment. The gas industry was highly seasonal be-cause of the limited contours of consumption: the number of stokers needed for production in July was only about 55 percent of that required in January. Stokers were very much part of a floating population that alternated between factory work and other activities, often rural employment. In 1865 the assistant chief of production, Servier, stated that "nothing attaches the stokers to the company. Our jobs are for them a pis aller, and they often prefer to earn less while working in the open air."^[34] Three decades later the journalist Babin was still able to describe the stokers as "Bretons and Auvergnats, for the most part, who come to spend winter in the hellfires of Paris, when the earth enters its annual sleep, and who, at the first springtime sun, return to a much healthier life in the country."^[35] This description, a rough approximation of reality at best, captures an important truth: stoking was only a part of the working lives of the stokers.

As members of a floating population, the stokers were by no means a special case of industrial workers. The convention of dividing social phenomena into urban and rural is often highly misleading. Yves Lequin has shown that temporary transfers of population between the countryside and the city provided a major part of industrial labor through most of the nineteenth century. "Industry founded its takeoff" Lequin writes, "on the periodic use of a circulating labor force more than on the assembling of a fixed force."^[36] The phenomenon was a consequence of a pattern of rural proletarianization that was quite pronounced in France. The countryside had long known a proliferation of very small holdings, and cultivators had to supplement their incomes, often through temporary emigration and seasonal industrial activity Rural population growth also produced underemployed village artisans, who sought seasonal work in cities. On the demand side, small and irregular markets for industrial goods made a large, permanent industrial labor force unnecessary. In the end, gas production was unexceptional in its seasonality and its use of a floating labor force. The stokers were not at all peculiar as peripatetic proletarians.

The stokers' acquiescent response to the rationalization of production until the 1880s must be understood in the context of their relative independence from the PGC. The good pay that the gas firm offered for part of

[34] AP, V 8 O , no. 148, report of Servier, June 20, 1865.

[35] "Le Gaz de l'éclairage," p. 252. On the seasonality of gas production, see Frank Popplewell, "The Gas Industry," in Seasonal Trades , ed. Sidney Webb and Arnold Freeman (London, 1912), pp. 148-209.

[36] Lequin, Ouvriers de la région lyonnaise , 1:vii; Noiriel, Ouvriers dans la société française , pp. 44-55.

the year was an attractive option for the stokers, but it was not their only option. Their commitment to stoking as an occupation was conditional on being able to earn more than through other lines of work. Such was the instrumental logic of their employment with the PGC.

The stokers were essentially rural people in background and in habits. The earliest electoral lists for the neighborhoods they inhabited show that the vast majority came from the provinces, usually the deep provinces (table 18). Almost all were born in villages. They could not follow the pattern of some industrial workers, who settled near factories and begat a second generation of laborers sur place. The special "skills" stokers had, unusual strength and endurance, could not become a family legacy; and stokers might have wanted less demanding jobs for their children, anyway. Thus, each generation of stokers was constantly renewed from rural France, not from established factory communities.^[37]

In the early years of the gas industry stokers seem to have organized

[37] For the more common case of immigrants who eventually established a community around the factory, see Lenard R. Berlanstein, "The Formation of a Factory Labor Force: Rubber and Cable Workers in Bezons, France (1860-1914),' Journal of Social History 15 (1981): 163-186.

their lives in a manner similar to that of the better-known "Limousins,' or building workers from that province. These masons left home each year to spend the winter in Paris and rarely renounced being cultivators. Their ultimate goal was to put aside money to survive as farmers or to save up enough to round out their plots.^[38] The geographical origins of the stokers tied them to the same world of proletarianized peasants. As we have seen, Babin assumed that most stokers were still peasant-workers who, even at the dawn of the twentieth century, returned to their farms each spring. That might have been true at one time, but growth and change in Paris during the Second Empire interrupted the pattern. The PGC's engineers noted agricultural labor as one alternative for stokers, but construction in Paris provided another option, probably a more important one by the 1860s. Various building activities were viable for stokers because of the seasonal patterns, the relatively high pay, and the physical requirements. Some stokers also went into the brickyards for work during the warm season. Most often the PGC expected to lose stokers to masonry, ground clearing, and foundation work.^[39] Opportunities to take jobs in construction were plentiful in the Paris of Baron Haussmann.

Because stoking was not a permanent occupation but rather a part of an economy of makeshifts, often with an agrarian component, stokers acted with a great deal of independence regarding their employer. Thus, the Auvergnat Pierre Garrigoux had to be rehired eight times between July 1876 and March 1880. He had departed voluntarily seven times and was laid off only once. His leadership in the 1880 strike effort finally cost him his job with the PGC.^[40] The most visible manifestation of this independence vis-à-vis the gas industry was the paradox that the PGC had a shortage of stokers during the season it needed the fewest. The difficulties in recruiting a sufficient number of stokers were greatest in July and August, when gas consumption was at its lowest point of the year. At the root of the paradox was the stokers' practice of leaving gas plants to take jobs in construction or returning to the farm, both of which were then at their peak of activity. During the moments of especially powerful demand for labor in construction, gas engineers complained that stokers left even before winter ended so as to get the best-paying jobs at building sites. There they

[38] Alain Corbin, Archaïsme et modernité en Limousin au XIX   siécle (Paris, 1975); David Pinkney, "Migrations to Paris during the Second Empire," Journal of Modern History 25 (1953): 1-12; Gabriel Désert, "Aperçue sur l'industrie fran-çaise du bâtiment du XIX siècle," in Le Bâtiment: Enquête d'histoire économique, XIV -XIX siècles, ed. Jean-Pierre Bardet et al., (Paris, 1971), 35-119.

[39] AP, V 8 O , no. 148, report of Jones, April 26, 1865; report of Gigot, April 30, 1865; report of Letreust, April 29, 1865; report of Servier, June 20, 1865.

[40] Ibid., no. 150, "Affaire Garrigoux."

could earn as much, or almost as much, as in the distillation rooms while escaping the frightful physical toll of a summer near the furnaces.^[41] Stokers further asserted their independence from the gas company by resisting the shift from a twenty-four-hour to a twelve-hour day. As late as 1870 a particularly acute shortage of stokers forced the PGC to offer a return to the old schedule to attract more workers.^[42]

The independence signals what stokers cared about most in working for the PGC—earnings. Stokers did not constitute an occupational community rooted in a craft. They did not define their status in terms of mastery at the workplace. Distilling coal was a way of scraping together income. Alain Corbin has identified the "ferocious will to save" as the guiding rule of the Limousins, and the same point applied to the stokers.^[43] The quest to maximize earnings was the focus of their working lives, and it created both areas of compromise and areas of conflict with the company. On the one hand, managers were eager to raise productivity and the stokers were willing to accommodate, for a price. On the other hand, the engineers' moral view of wages did not honor the principle that harder work necessarily deserved greater compensation.^[44] Thus, there was a large potential for conflict between management and the workers even though the latter accepted the tradeoffs inherent in the logic of capitalism.

<><><><><><><><><><><><>

The reports of factory superintendents to their superiors during the 1860s place the role of labor into perspective. The managers apparently regarded neither the supply of workers nor their malleability as major concerns. Superintendents rarely noted that workers refused to increase their output or alter their routines. Even complaints about shortages of stokers were sporadic, not chronic. The manager of the Ivry plant pointed out in 1862 that the booming construction industry for the most part helped the PGC find stokers, for it attracted many hands to Paris, and they needed work in the winter. Other problems preoccupied the superintendents far more than labor: the quality of the coal, the supply of well-

[41] Ibid., no. 674, deliberations of April 30, 1870; no. 148, report of Jones, June 26, 1865; report of Philippot, April 25, 1865. The stokers could have remained with the PGC, working in the courtyard, if they had been content to earn only 3.25 francs.

[42] Ibid., no. 674, deliberations of April 30 and May 4, 1870. The Belleville plant ran on twenty-four-hour shifts until 1874. See no. 677, deliberations of December 12, 1874.

[43] Corbin, Archaïsme et modernité, p. 205.

[44] See chapter 4. For a theoretical perspective on consensual relations in industry, see Michael Buraway, Manufacturing Consent: Changes in the Labor Process under Monopoly Capitalism (Chicago, 1979), which underscores the contingent nature of workers' consent.

constructed retorts, and the state of the equipment. Labor agitation occurred from time to time, but managers would hardly have been inclined to define industrial relations in terms of a crisis, as they would in 1899.^[45]

The distilling halls of the PGC were disturbed by strikes or by serious threats of strikes roughly every other year between 1859 and 1869. Peasant-workers though they may have been, the stokers had no trouble cooperating with one another and making the company aware of their discontents. In each biennial action they forced the company to give them a raise. Between the founding of the PGC, in 1855, and the Franco-Prussian War stokers had obtained a 70 percent increase in their pay. Several forces were driving their agitation. The cost of living had risen sharply in Paris, and stokers were in part safeguarding their purchasing power.^[46] At the same time, stokers were demanding compensation for their greater exertion on the job. Inflation and harder work may have formed the basis for the common understanding that allowed stokers to take collective action, but they did not determine the timing or the form of the strikes. These aspects of collective protest derived from the stokers' independent status. Stokers struck when they could take advantage of their attachments to the construction industry.

The labor agitation in the PGC up to the 1880s coincided with peaks in building activity.^[47] Through the nature of their strikes the stokers re-minded the company of their independence from it. In "classic" strikes, workers withhold their labor and inform their employer what it would take to bring them back to the factory. The stokers instead staged a collective act of quitting and left en masse for jobs on construction sites. That is why all the strikes of the 1860s took place in the spring or summer, at the time of hiring for construction but in the dead season for gas production. Apparently, the stokers did not always specify the conditions that would keep them at the plant.^[48] Such practices might appear "archaic," the action of workers who were not fully inured to the industrial process or who possessed a "preindustrial mentality"; but they were nothing of the

[45] See the collection of reports from the plant superintendents, AP, V 8 O , no. 768; no. 148, report of Gigot, April 30, 1865.

[46] On price levels in Paris, see Alain Plessis, De la fête impériale au tour des fédérés (Paris, 1979), p. 93.

[47] On building activity in Paris, see Françoise Marnata, Les Loyers des bourgeois de Paris, 1860-1958 (Paris, 1961), and Jacques Rougerie, "Remarques sur l'histoire des salaires à Paris au XIX siècle," Le Mouvement social, no. 63 (1971): 71-108.

[48] For examples of early strikes, see AP, V 8 O , no. 148, "Grève et réclamations"; no. 1286, report of superintendent of Ternes plant to Arson, April 26, 1867.

kind.^[49] The stokers' form of striking was a rational strategy arising from their position vis-à-vis the gas industry. The action might also have been designed to avoid the harsh treatment that the company reserved for strikers. Instead of firing them for making trouble, the superintendent had to convince them to remain by offering concessions, so

The object of collective action was not yet a complex matter of control but rather a matter of pay. The stokers were willing, perhaps eager, to labor harder provided they received higher compensation (table 19). Whether they were paid by the day or by the amount of coal distilled did not seem to make a difference to the stokers; they preferred the mode that would bring them the most income. The rationalization of gas production met with little opposition from the skilled workers because they aspired to maximize gains from lucrative work that was available for only part of the year The raise did not even have to be proportional to the increase in exertion. When the PGC took over a gas plant near the suburb of Saint-Mandé, stokers demanded the pay level that prevailed in the company's other plants, which meant a raise of twenty-five centimes a day (6.7 per-cent). The new superintendent yielded to the demand only after he reorganized teams so that five men would have to do the work that six did formerly. Despite the imbalance between the raise and the increased labor extracted, no further trouble occurred. Indeed, the superintendent re-ported that the stokers were quite happy with the arrangement.^[51]

Superintendents were often surprised by the amount of work they could extract from stokers by according a small raise. The manager of the Passy plant had not been happy about giving the stokers twenty-five centimes more per day in 1859, for it meant they were escaping the internal labor market he had aspired to construct. Nonetheless, he appraised the results of the raise with satisfaction. Not only was the superintendent able to replace the popular twenty-four-hour shift, but he was also able to introduce charges every four hours. Thus, the 7 percent raise yielded a 55 percent increase in coal distilled.^[52] At the same factory nine years later, another acute labor shortage compelled the superintendent to offer the stokers a choice of work routines. They could continue at the same pay and load short (2.5 meter) retorts, requiring seventy-five kilograms of coal

[49] For intriguing efforts to link the form of collective protest to workers' culture, see Reddy, Rise of Market Culture, chaps. 7, 10; Michelle Perrot, Workers on Strike: France 1871-1890 , trans. Chris Turner (New Haven, 1987).

[50] Organizers of collective protest were inevitably fired, if not at once, then at a convenient moment.

[51] AP, V 8 O , no. 768, "Usine Saint-Mandé."

[52] Ibid., "Usine Passy." The manager reported producing 249,000 cubic meters of gas a day in 1861 and only 160,000 in 1859.

per charge, or they could load long (2.75 meter) retorts, taking a hundred kilograms of coal, and receive a raise of twenty-five centimes. The stokers chose the higher pay even though they were not going to be fully compensated for their exertion.^[53] On two occasions at least, in 1865 and 1870, top production managers believed that stokers had reached the limits of human endurance and that productivity could climb no higher without basic changes in the work process.^[54] Yet output per worker continued to rise without resort to mechanization, though more slowly than in the past. The incentive bonus introduced in 1869, twenty centimes per hundred kilograms of coal over twenty-five hundred kilograms distilled, apparently motivated stokers to boost output.

Though wages were the stokers' principal focus at the time, there were other issues that could unite them. Cury, the ancient superintendent of the La Villette plant, once sought to improve the lighting power of coal by treating it with heavy oils derived from coal tar. The stokers refused to abide the experience because the odor of the oil made them ill and the treated coal stuck to their shovels. Curd had to abandon the initiative.^[55] Likewise, firemen occasionally resisted the endless experiments with dif-

[53] Ibid., no. 749, report of superintendent of Passy plant to Arson, April 6, 1869.

[54] Ibid., no. 764, "Main-d'oeuvre de distillation (26 avri11870)"; no. 148, report of Gigot, April 30, 1865.

[55] Ibid., no. 768, "Usine de La Viilette,' report of December 15, 1861.

ferent sorts of fuels.^[56] The work routine on Sunday was another source of friction. The company attempted to make the stokers on the day shift work a full night shift too. The stokers successfully united to compel the company to accept four charges on Sunday instead of six.^[57] These incidents suggest that stokers had the potential to stop or at least curtail the dramatic changes that management made in the work routine during the 1860s, but the workers did not apply their collective power to that goal.

In addition to seeking more pay, even at a high cost to themselves, the stokers held the company to a rudimentary sense of justice entailing equal pay for equal work. Rumors that stokers at another plant were earning more sparked some agitation. But the stokers were never completely successful in standardizing wages among the different plants because work routines and output differed from one to another. Furthermore, the company often succeeded in keeping them ignorant about pay levels outside their small spheres.^[58] On the other hand, the stokers quickly disposed of corporate plans to impose on them the career-based pay under which unskilled workers suffered. The company tried in 1858 to create three "classes" of stokers, with twenty-five-centime differences for each step, but agitation forced management to back away from the attempt within a few months.^[59]

Lacking from the stokers' repertoire of collective protests were con-frontations with supervisors over unwelcome interference, insults, tests of will, and the like. The records on labor from the 1860s and 1870s, ad-mittedly incomplete, contain no reference to such conflicts.^[60] The absence is all the more noteworthy in that the stokers' increasing productivity was achieved through closer supervision, badgering, and pressuring from fore-men. The number of immediate supervisors increased significantly. In 1858 there was one foreman for every thirty-seven distillation workers; in 1884 the ratio was one to twenty-three. Stokers may have resented the increased supervision but they had to deal with it as individuals, quitting or standing up for themselves. In view of the trouble that foremen would cause by the end of the century, the absence of such conflicts is telling.^[61]

[56] Ibid., "Usine Ternes,' reports of September 16 and 23, 1860.

[57] Ibid., "Usine Beileville,' report of April 14, 1861; no. 717, report no. 95, Rigaud to Arson.

[58] Ibid., "Usine Passy,' report of April 14, 1861; no. 151, report of Arson to director, May 1, 1890.

[59] Ibid., no. 666, deliberations of August 3, 1858, and March 9, 1859.

[60] Ibid., no. 768; no. 148, "Rdponses aux demandes concernant le personnel." Engineer Audouin reported no more than two or three suspensions a year for rebellious behavior.

[61] Ibid., no. 665 (fols. 500-514); no. 153, "Personnel." On agitation against the authority of the foreman at the end of the century, see chapter 8.

The stokers clearly created consternation among management by defeating its plans for an internal labor market. Nonetheless, these physical giants proved pliant in serving the larger interests of the firm. The managers' complaints about "outrageous" demands were a sign of having been spoiled by fortunate circumstances.^[62] Ultimately, engineers implicitly recognized their good fortune and showed it by tolerating wage demands without seeking to weaken labor's power through technological displacement. Over the long run, however, the firm's advantageous position was destined to disappear. It was not that stokers suddenly became "class-conscious." Rather, the conditions that kept them independent, interested primarily in higher wages, and responsive to the carrots and sticks that managers chose to wield ended during that prime transition point in Parisian social history, the economic crisis of the mid-1880s.

<><><><><><><><><><><><>

Historians have long recognized the quickening pace of structural change within the French labor force in the last two decades of the nineteenth century. Machinery challenged handicraft production; international competition compelled managers to reorganize and intensify work routines.^[63] One further force creating a "proletariat" in France was the curtailment of temporary migrations to urban industry from the country-side. Industrial workers settled in cities.^[64] The stokers participated in the last pattern of change, which entailed far more than taking up new residences. It called into question all the ways the stokers had heretofore related to their employer.

Ceasing to be rural laborers who sought winter work with the PGC was one aspect of a broader alteration in the stokers' independent status. They became dependent on the firm for year-round employment. No longer were they workers who put together a livelihood through varied tasks, among them stoking; they became gas workers, pure and simple. By the 1890s, 45 percent stayed in the distillation room during the summer. Most of the rest struggled for the now coveted "privilege" of having year-round work with the PGC and took jobs as common laborers in the courtyards to be on hand when stoking jobs became plentiful in the fall. Superinten-

[62] For the internal labor market, see chapter 4. For a fine example of a factory manager spoiled by fortunate circumstances, see AE V 8 O , no. 749, report of superintendent of Passy to Arson, April 6, 1869.

[63] Jean-Marie Mayeur, Les Débuts de la III République (Paris, 1973), pp. 66-72; Michelle Perrot, "The Three Ages of Industrial Discipline in Nineteenth-Century France," in Consciousness and Class Experience in Nineteenth-Century Europe , ed. John Merriman (New York, 1980), pp. 149-168.

[64] Lequin, Ouvriers de la région lyonnaise , 1:123-156; Noiriel, Ouvriers dans la société française, p. 83.

dents were now pleased with the number of job seekers available. Indeed, the assistant head of the factory division maintained that some of those applicants took the trouble to procure letters of recommendation from influential people.^[65] The sources of this transformation into dependent workers were manifold. Some causes reached back into the deep provinces where the stokers had originated. The heavy rural exodus from Brittany, Auvergne, and Savoy, after centuries of overpopulation and temporary migration, affected the fate of the stokers. The decline of rural industry, the reconstitution of larger farms in response to more commercialized agriculture, and the broadening of peasants' cultural horizons sent thousands of rural hands to Paris and the other metropolises, where they remained permanently.^[66] Like the stonemasons who began to settle in Paris some decades earlier, stokers no longer considered returning home. The demo-graphic shifts assured a plentiful work force for the PGC, all the more so because of two crucial changes in Paris.

The local changes undermined the stokers' independence by calling into question their ties to the construction industry as an alternative to the PGC. First, the wages of stokers pulled decisively ahead of building workers by 1880 at the latest. Pay for skilled masons stagnated at around eight francs a day through the rest of the nineteenth century.^[67] In the meantime, stokers had taken advantage of the PGC's critical labor short-age in 1880 to strike and win a pay rate of three francs per thousand kilo-grams of coal distilled. That rate assured them a daily wage of almost nine francs and eventually more. Even if stokers were willing to settle for the lower pay of construction workers, jobs would not have been easy to find. The golden era of building had passed by that time. The Parisian housing market was in recession after 1882: Paris had too many apartments, and rents stabilized until the century ended.^[68] Baron Haussmann's projects had once made it possible for stokers to escape their painful work for part of the year; alternatives to stoking now became much more problematical. Stokers were likely to remain in gas production as long as the company needed them or as long as their stamina allowed. It is not surprising that when two stokers were fired in 1895 for having ruined a retort through

[65] AP, V 8 O , no. 162, report of Euchène to director, April 10, 1906; no. 163, report of superintendent of Ivry plant to Euchène, March 23, 1902; no. 148, "Re-vendications. Service des usines," responses of Gigot.

[66] Françoise Raison-Jourde, La Colonie auoergnate ti Paris au XIX siècle (Paris, 1976), pp. 179-190; Maurice Agulhon et al., Histoire de la France rurale, vol. 3, Apogee et crise de la civilisation paysanne , 1789-1914 (Paris, 1976), pp. 469-487.

[67] Rougerie, "Remarques sur l'histoire des salaires," p. 99-100.

[68] Michel Lescure, Les Sociétés immobiliéres en France au XIX siècle (Paris, 1980), p. 78.

negligence, their immediate concern was whether the superintendent would take them back next year.^[69]

The consequences of the stokers' dependency were soon visible in pat-terns of labor turnover (table 20). The changes after just a few years of the new conditions were quite dramatic. The portion of stokers who had remained with the PGC three years or less dropped by more than half. More stokers were staying on the job longer. The figures also attest to a reduction in purely seasonal employment. That is why the portion of stokers who lasted fifteen years was actually a good deal higher in 1878 than in 1888; men who stoked year round could not have lasted so long. Most would have been incapacitated after a decade of year-round work.

Another consequence of dependence on the PGC was an end to the paradoxical shortage of stokers during the slow season. Plant superintendents confidently reported having a surfeit of applicants during the 1890s.^[70] Though stokers had once hastened to return to the countryside or snatch a good construction job when the warm weather approached, those options were no longer attractive or available. After 1880 stokers maneuvered to avoid seasonal layoffs. Indeed, the PGC's policies regarding layoffs now became a major source of grievance and frustration.

Not surprisingly, the issue surfaced first in the mid-1880s, when the stokers' dependency became manifest. The PGC had granted a large wage increase to stokers on the occasion of a strike in 1880 and then attempted

[69] AP, V 8 O , no. 150, report of Hadamar, July 10, 1895.

[70] Ibid., no. 148, report of Gigot to director, October 8, 1892. The assistant head of the factory division asserted that each factory received about forty applications a day.

to reduce its wage bill by hiring Belgians and Italians at lower pay. By 1887 more than 40 percent of the stokers at the La Villette plant, the company's largest, were foreigners.^[71] French stokers complained that the foreigners were retained at the plant during the slow season whereas they were sent away and then not always rehired in the fall. Pressured by the city and by the national administration to favor French workers, the company announced a formal policy of layoffs based on seniority. The concession settled very little, however, for the PGC altered its interpretation in a mercurial manner and never clarified it in detail. Was seniority determined on a plant-by-plant basis, or was it companywide? Did a stoker who worked two full years have more or less seniority than a stoker who had returned for five consecutive winter seasons? The director allowed plant superintendents to apply the rules as they wished and to improvise. He undoubtedly did not mind that arbitrary layoff procedures sowed dissention among stokers even as they darkened relations with the company.^[72]

The shift to dependency brought prominence to the distinction between stokers with the newly acquired right to work all year and those subject to layoffs. Superintendents reported that at least three years of employment with the company were necessary to establish that right, and the slow expansion of gas consumption in the 1890s must have lengthened that term. In the meantime, misunderstandings over rules were numerous, and fights among stokers, each defending his rights, broke out.^[73] The solidarities promoted by common frustration with the company were partially eroded by jealousies among stokers over the issue. That these tensions could cut deep is demonstrated by the aftermath of the wildcat strike of March 1899 (see chapter 4). The company it will be recalled, agreed to take back all strikers, but it was necessary to put them to work in plants other than the ones in which they had been employed. The stokers already at those plants protested the strikers' presence as a threat to their own layoff status. Thus, comrades-in-arms one day became squabbling job seekers the next.^[74] The irony was that the March strike had been triggered by a foreman's arbitrary application of rules on layoffs. The quest for se-cure, year-round work had come to color relations among workers and relations between workers and the PGC in ways it could not have during the 1860s.

[71] Ibid., no. 90, letter of Camus to consul general, October 1, 1887.

[72] Ibid., no. 151, "Règlement à l'ancienneté pour toutes les catégories"; letter of Lajarrigue to Hadamar, November 8, 1901; Le Journal du gaz. Organe officiel de la chambre syndicale des travailleurs du gaz, no. 108 (June 5, 1897): 2.

[73] AP, V 8 O , no. 163, report of superintendent of Ivry plant to Euchène, March 23, 1902; no. 151, letter of Lajarrigue to Hadamar, November 8, 1901.

[74] Ibid., no. 159, report of Gigot, March 25, 1899.

The dependent status of stokers had not yet solidified by 1880, but al-ready the strike of that year was different from earlier ones because of the change. The defiant stokers sent a delegation to management, not to announce a collective departure for construction sites, but rather to make specific demands regarding their situation at the PGC. Moreover, stokers were now determined to profit from their higher productivity to a greater extent than earlier. They pushed managers to accept piece-rate pay, which the engineers had been postponing for a decade.^[75] Still another novel element of the stokers' strike was the nonwage grievance. To restore labor peace, the company took its first step to lighten stokers' burdens. It agreed to hire workers to carry coal from the yards to the distillation hall and to break the coal into small pieces suitable for charging. Eventually management would turn this concession to its own advantage by converting the new crew into a cadre of potential replacements for stokers, but initially it was a benefit stokers won through collective action.^[76] The strike of 1880 was a relatively minor outburst—quickly settled because the company had no choice except to yield—but it did portend a reshaping of the stokers' protests.

In the end, the stokers' new status owed very little to the PGC's conscious labor policies. Management had long resisted the demand for piece rates in 1880 that eventually drove stokers' wages decisively above those in competing lines of work. The company had never offered benefits, like housing or pensions, that might have made workers cut their ties with rural life or with alternate urban jobs. Dependency came about, as we have seen, through the operation of impersonal forces. Nonetheless, the new status of labor did offer the PGC important strategic tools for dealing with its most obstreperous workers, and the company was not slow to exploit them.^[77]

It would be a serious mistake to confound dependency with submissive-ness. To the contrary, dependency eventually forced on stokers a new and potentially explosive set of issues that would create confrontation with the firm. The previous independent situation of stokers had reconciled their interests with the business calculations—though not the moral sensibili-

[75] Ibid., no. 716, "Rapport sur les primes," no. 66, March 1880; Préfecture, B/a 176, reports of March 15-20, 1880.

[76] AP, V 8 O , no. 151, "Communications faites à la presse parisienne"; no. 695, deliberations of June 8, 1899. It should not come as a surprise that the PGC took the step of creating a pool of potential replacements for stokers when labor tensions heightened in 1899.

[77] By contrast, David Brody, Steelworkers in America , pp. 74-102, depicts the classic example of the United States Steel Corporation explicitly encouraging de-pendent status among its skilled workers.

ties—of management. Under these conditions Calhoun's model for skilled industrial workers approximated reality. But as we shall see in chapter 8, dependency brought an end to these circumstances. A brief comparison of strikes of the 1860s with those of 1890 and 1899 illustrates the changes in protest that were about to ensue. In 1890 the stokers' work stoppage won them a raise, but they still defined the outcome as a failure. In 1899 their strike demands allowed for a sharp reduction in pay. In effect, dependency gave stokers a set of demands that were more, not less, difficult for the company to satisfy. The stokers started to behave more like craftsmen, conscious of control over the workplace, than anyone would have hereto-fore imagined possible.^[78]

The Common Laborers

The multifarious business activities of the PGC created a demand for general laborers to perform a wide variety of simple tasks. Unlike stoking, these jobs could be mastered quickly and most any worker could do them. Taken together, the many sorts of common hands comprised a corps more than three times as large as the stokers. Wage earners of this kind have not had much of a presence in labor history owing to their silence. By and large they accepted the conditions offered them, either because it was the best they could hope for or because they were powerless to do otherwise. One of the noteworthy features of industrial relations within the PGC was that the common laborers acquired their own voice, for a time a clearer one than that of their comrades with market power.

With the company recognizing sixty-seven categories of unskilled workers, the diversity defies easy categorization.^[79] Whatever grounds these laborers had for collective action did not derive from common work conditions. A large group of unskilled laborers was composed of coke handlers—nearly 850 of them in 1890 (table 21). These workmen took the extinguished coke from the stokers and processed it for residential use. This task involved piling the coke in large heaps, filling burlap bags with it, carrying the bags on their backs to delivery wagons, and loading the wagons. Obviously, there was nothing intricate about such work, but it was exceptionally taxing. The PGC did not begin to mechanize these op-

[78] For confirmation of this tendency among skilled workers of modern industry, see Reid, Miners of Decazeville , pp. 130-139; Margot Stein, "Working to Rule: The Social Origins of a Labor Elite. Engine Drivers, 1837-1917" (Ph.D. diss., Harvard University, 1977), pp. 246, 278-292.

[79] AP, V 8 O , no. 159, report of Bodin to Euchène, March 3, 1899.

erations until the last years of the century. By the company's own estimates a coke handler carried each day 186 sacks weighing (altogether) eight thousand kilograms for a total of thirteen kilometers.^[80] This toil brought the handlers a meager wage of 3.75 francs a day in 1890. The ten-hour day was often extended by the irregular arrivals and departures of the delivery carts.

Nearly three hundred coke deliverymen completed the labor force attached to this department. They formed a highly problematical service. Every year the company had a dozen or so prosecuted for theft or fraud. At one point the PGC hired undercover agents to spy on the delivery carts as they made their rounds but finally decided that the expense of such surveillance was too great and that it was easier to rely on customers' complaints to detect fraud. The record of criminality gives substance to the assertion made by the chief of the coke department that the company

[80] Ibid., no. 148, "Examen des réclamations concernant le service des cokes (4 avril 1892)."

paid too poorly to attract carters of higher integrity.^[81] For their part, carters had their own grievances against the firm. They bore civil and criminal responsibility for any accidents they may have caused on the crowded streets. A majority of carters affirmed their willingness to contribute two francs a month from their meager pay for a mutual insurance fund, but the company refused to help them organize one. Moreover, management forced high standards of punctuality on these workers, who were accustomed to the more casual schedules characteristic of outdoor workers. If one was five minutes late for the 5:00 A.M. call, the company assigned his wagon to another worker, and he lost a day's pay.^[82]

Almost seven hundred common laborers worked in the courtyards of the distillation plants. Their jobs included unloading shipments of coal, packing by-products, and performing general maintenance chores. The company made an important distinction between the few courtyard men who could replace absent stokers and those who could not. It attempted to retain as many of the former as possible by paying them an extra fifty centimes a day.^[83]

In close proximity to the courtyard workers were the two hundred or so who operated the by-products plant. Their jobs were particularly unpleasant because of the stench or fumes from the chemicals they worked with. Some had to stand all day knee deep in pools of noxious liquids. Others operated the machinery or fueled the furnaces. Eye ailments and skin rashes were common occupational hazards for these laborers. Despite these hardships, the by-products workers were among the poorest paid in the company.^[84]

Another large group worked not in production but in the distribution service, as lamplighters, greasemen, waxmen, and repairmen. Because their labor was fixed, regular, and predictable, the company commissioned them after a probationary period. The thousand or so lamplighters came closest to thinking of themselves as quasi-municipal workmen: part of their pay came from a municipal subsidy (three centimes per streetlight per day), and their labor was directly related to a public service. The municipal council could alter their work routine by changing its regulations

[81] Ibid., no. 90, report of Hallopeau to director, January 14, 1867. The deliberations of the comité d'exéution contain a large number of decisions to prosecute carters for theft or fraud.

[82] Ibid., no. 148, "Revendications. Service des cokes (17 septembre 1892)"; re-port of Montserrat, December 23, 1891; report of Montserrat, April 27, 1892.

[83] Ibid., no. 151, "Revendications générales—personnei ouvrier."

[84] Ibid., no. 157, "Conditions du travail et d'hygiène des ouvriers des anthracènes"; "Procès-verbaux de la commission administrative de la Caisse de prévoy-ance," deliberations of March 11, 1864.

concerning street lighting. The PGC regarded its lamplighters as part-time laborers. It contended that the rounds of lighting, cleaning, and extinguishing the sixty-five or so assigned streedamps required at most five hours a day. It noted that the majority of the lighters had secondary occupations, as concierges, street peddlers, shoemakers. The lighters them-selves asserted that their work demanded an eight-hour day. At one point the company had a spy follow a union leader on his rounds and found that the tasks did indeed require eight hours—but only because of frequent stops in pubs or other breaks. The lamplighters, maintaining a preindustrial conception of their work routine, did not wish to define these breaks as external to the job.^[85] In any case, the lighters' secondary occupations gave them some independence from the company and they used it to be among the most turbulent of unskilled laborers.

The greasemen, waxers, and repairmen had an advantage that most wage earners would have welcomed. As "fixed workers," they were paid by the month, and their employment was permanent. Their pay was very low, however, twelve hundred to fourteen hundred francs a year The manual workers put in thirteen-hour days, outside, painfully exposed to the elements.^[86]

There were another two hundred workmen who practiced a trade— plumbers, wheelwrights, fitters, turners, and others. These laborers were not easily categorized. They were not affiliated with the Parisian crafts. Perhaps because their jobs were ancillary to the PGC's central production concerns, the tradesmen were never able to take a directing role in any protest. There was nothing about their comportment that exemplified the proud, contentious craftsmen. For the most part their fate was more closely linked to the common hands than to the stokers.^[87]

In important ways the differences between stokers and common laborers was artificial, or at least superficial. Bargaining power on the job was one of the few social distinctions between stokers and common hands. The social background of the two groups could not have been more similar. Both participated in the temporary rural immigration to the capital as long as it lasted. The rural exodus from the deep provinces supplied an

[85] Ibid., no. 151, report of Saum to director, June 3, 1893; reports of Saum to director, April 2, 14, and 17, 1892; "Revendications des allumeurs"; "Renseignements sur les professions des allumeurs (24 mars 1900)." In 1900, 846 of 1,028 lamplighters were reported to have second occupations.

[86] Ibid., no. 148, "Syndicat. Service extérieur."

[87] Mechanics who worked for the company making motors, stoves, and valves did so at piece rates under sweated conditions. They were the sort of degraded hand-workers that fully trained mechanics called "small hands." See ibid., no. 151, re-port of Maubert, August 25, 1897.

increasing portion of both sorts of gas workers. Most stokers had roots in both camps. The company recruited them among the common hands; the stokers spent summers in the courtyards, working as general laborers. Stokers eventually returned to unskilled labor when they could no longer endure the work.^[88] In a sense, stoking was not much more than a stage in the life cycle for the minority of physically qualified common laborers.

Nonetheless, the common laborers had a distinctive set of problems on the job—above all, being caught in the internal labor market that managers created for them.^[89] That situation meant, as we have seen, low and uneven pay. Table 21 must be read in this context. As a result of the step system of pay which the company imposed on them, a third to a half of each category earned twenty-five centimes less than the average par and a like portion earned twenty-five centimes more. Managers hoped to impart moral lessons as well as divide workers through segmented pay, and they probably achieved the second goal. In 1892 a union delegate from the coke department admitted that "young and older workers do the same job at different pay and there results an antagonism between the two ."^[90] Unlike stokers, the common hands could not mount a successful offensive against unequal pay for equal work.

When, after 1890, convincing the public that its pay scales were generous became important to the PGC, it could do so by publicizing the stokers' rates alone, even then neglecting to mention the seasonal quality of the pay. Average wages of common hands were less than half of the stokers' in 1889. The comparative data that engineers gathered from neigh-boring firms showed that the PGC's compensations were in line with the lowest-paying industries—fertilizer, animal-rendering, and glue companies. Engineer Audouin, who led the fact-gathering mission, was surprised to learn that the Lebaudy and Sommier sugar refineries, notorious exploiters of Italian and Belgian labor, paid the unskilled as well as the PGC. Refinery workers were hired at 3.75 francs a day and could earn 4.50 to 5.50 francs with experience.^[91] A clearer understanding of market wages came as something of a revelation to the PGC's managers. Up to this point they had hoped to be able to ignore the market. Whether the new aware-ness undermined their self-righteousness about labor policies is doubtful. The low pay that the gas company offered inevitably generated a rapid

[88] Ibid., no. 163, "Assimilation des ouvriers (4 février 1903)."

[89] On the efforts to create an internal labor market, see chapter 4.

[90] AP, V 8 O , no. 148, "Procès-verbaux de la quatrième audience donnée le 3 mars 1892 aux délégués" (remarks of Gabanon).

[91] Ibid., "Renseignements sur les salaires journaliers des ouvriers pris chez différ-ents fabricants."

turnover of the labor force. After all, common hands could earn more elsewhere, suffer no less from seasonal layoffs, and often have tasks that were less painful. Even the greasemen, who were commissioned, had a resignation rate of 7 percent in a year of recession, like 1889, and 19 per-cent in a year of dramatic labor shortages, like 1881.^[92] Wage earners, how-ever, were by no means committed to a peripatetic existence by preference or by life-style. Michelle Perrot has correctly identified an "aspiration for rootedness and for security of employment and retirement" among common hands, and it is dear that an appreciable portion of those at the PGC sought those arrangements.^[93] They became possible when political pressures and bourgeois reform altered the wage policies of the firm. Average pay rose 16 percent, to 5.63 francs, by 1895, not counting the profit-sharing bonus introduced in 1893 (30-38 centimes a day).^[94] Furthermore, the small pensions of 360-600 francs promised by the company in 1893 made employment more attractive.^[95] Such remuneration still did not attain the level of largesse the PGC publically proclaimed, but the unskilled nonetheless seized on the improvements to construct an element of stability in their own lives. Twenty-nine percent of the ditchdiggers in the distribution department, 36 percent of the workers in the mechanical shops, and 54 percent of the chemical-products laborers accumulated at least ten years of seniority by the end of the century. Despite the grueling routine in the coke department, 52 percent of the laborers stayed on the job at least ten years.^[96] One greaseman aptly expressed the new outlook on the PGC as employer. Admitting to having committed a serious error on the job, he begged the director to inflict no more than a long suspension. "Being fired—that would be death" he lamented.^[97] General laborers seemed eager to define themselves as the personnel of one firm and make long-term commitments to it as soon as competitive pay made that situation viable. Like stokers, common hands came to depend on the PGC for steady employment.

The intervention of the city of Paris in the labor affairs of the gas com-

[92] Ibid., no. 148, report of Perrin to director, December 24, 1881; no. 90, report of Arson to director, January 11, 1889. Up to 1890 the PGC seemed to hire roughly fifteen hundred new workers a year.

[93] Michelle Perrot, "Une Naissance difficile: la formation de la classe ouvrière lyonnaise," Annales : Economies, sociétés, civilisations 33 (1978): 834-835.

[94] AP, V 8 O , no. 148, "Etat comparatif des salaires."

[95] So eager were wage earners for the benefits of a pension that they offered to contribute to the pension fund through a withholding of pay. See ibid., no. 155, "Règlement relatif aux versements à effectuer par le personnel ouvrier (1892)."

[96] Ibid., no. 151, "Personnel des ouvriers'; "Ouvriers des travaux chimiques"; no. 156, "Augmentation des salaires depuis 1 mai 1890."

[97] Ibid., no. 162, letter of Le Blais to director, n.d.

pany ended a prolonged period of helplessness. The unskilled may have admired the ability of the stokers to force wage concessions on management, but they had rarely been able to duplicate the success. Hence the earnings gap between the two sorts of laborers widened continuously. When the company was founded, general laborers received two-thirds of the stokers' daily pay; by 1890 the former received less than half.^[98] Only in the very few years of exceptional labor shortages in the capital did the unskilled acquire the bargaining power that the company could not easily resist. Common hands found such power only in 1867, 1876, and 1881. Otherwise, the few job actions that occurred ended in repression.^[99]

The lamplighters were the most aggressive of the general laborers. The second jobs that they held gave them some independence; furthermore, they knew that the disruption of public lighting even for a brief time would create public ire against their parsimonious employer. These advantages, real though they were, profited lamplighters not at all in 1865. Inspired by a successful strike among coach drivers, lighters demanded a raise from 65 francs to 90 francs a month and threatened to desert their rounds if it was not conceded. This threat made managers organize one of their acts of "salutary intimidation." They created a substitute corps of lighters, composed of inspectors, greasers, and auxiliary lighters (those still in the midst of their probationary periods). On the day before the threatened strike the director sent the substitutes to meet the regular workers with letters of dismissal. Eventually the company took back most of them, excluding only the perceived agitators. The rehired lighters had to accept a loss of their seniority in step pay and of their annual bonuses. If the potential strikers of the 1860s expected the municipal council to plead for leniency on their behalf, they received no support whatsoever from that quarter.^[100] The exquisite stagecraft choreographed by the com-pany in response to the strike threat told unskilled laborers that they could expect neither success nor clemency.

Even the modest wage gains that laborers were able to win during the boom of 1876-1882 came at a price. The PGC soon compensated for the raise by hiring foreign workers in large numbers. The result was new ten-

[98] Ibid., no. 149, "Augmentations des dépenses annuelles résuitant des améliora-tions depuis 1889."

[99] For examples of wage demands by common laborers, see ibid., no. 771, report of April 12, 1865 (fol. 46); no. 768, "Usine de la Villette," report of March 20, 1861; no. 752, report of May 6, 1864; Prefecture, B/a 176, report of March 6, 1880.

[100] AP, V 8 O , no. 148, "Etude relative à une augmentation du salaire des allumeurs (15 juillet 1865)"; "Rapport de la commission nominée à l'occasion de la grève des allumeurs (28 juin 1865)."

sions and divisions among the common hands.^[101] The PGC clearly knew how to dominate this underprivileged personnel—and intended to do so. Given the sour state of the gas industry during the 1890s, it seems unlikely that common hands would have received meaningful improvements had the city not transformed the balance of power by intervening in their behalf.

<><><><><><><><><><><><><>

Until the last decade of the nineteenth century the PGC employed thousands of workmen but had not truly formed a manual personnel of its own. Almost in spite of managerial policies, the labor force became tied to the firm, and an identifiable proletariat of gas workers came into existence. By no means was this process of labor-force formation unique. Analogous developments occurred at mine pits, railroad lines, steel mills, and machine-building plants all over France at roughly the same time.^[102] What was distinctive about the gas workers was that they came by the means to make management listen to their grievances. Only part of the labor force responded pragmatically and flexibly to the opportunities opened by the crisis of the liberal order. Defining what the workers wanted, not to mention achieving the goals, proved to be a difficult task. Still, there is no reason to suppose that other sorts of industrial workers would have used the opportunity differently.

[101] Ibid., no. 157, report of Gravier to Audouin, November 29, 1897; Prefecture, B/a 176, report of March 8, 1880.

[102] Joan Scott, "The Glassworkers of Carmaux, 1850-1900," in Nineteenth-Century Cities , ed. Stephen Thernstrom and Richard Sennett (New Haven, 1969), pp. 3-48; Lequin, Ouvriers de la région lyonnaise , pp. 123-156; Trempé, Mineurs de Carmaux , 1:189-253; Jean-Paul Brunet, Saint-Denis, la ville rouge (Paris, 1980); Noiriel, Ouvriers dans la société française , chap. 3; Hanagan, Nascent Proletarians , chap. 3.

Eight
Labor Protest in an Era of Social Reform

The outstanding concerns of fin-de-siècle labor history—a general crisis of industrial discipline, achievements in building working-class institutions, and harsh repression of protest—were all visible at the Parisian gas plants. Studying the gas workers permits us to explore the dynamics of the larger trends in microcosm. The gaziers even traversed the long path of the French labor movement from direct, informally organized action to indirect participation within national, bureaucratic unions.^[1] How and why they did so clarifies many important points about the construction and collapse of peaceable industrial relations in France.

Reformist Syndicalism in Action

The logic of labor protest might seem to dictate that artisans would use their skills to defend their interests through direct action at the workplace but that industrial workers, especially recent rural immigrants, would have to rely on a centralized political party to advance their cause.^[2] Unrest at the gas plants belies such a tidy generalization. The PGC's manual per-

[1] Charles Tilly, The Contentious French: Four Centuries of Popular Struggle (Cambridge, Mass., 1986), chap. 10; Michael Hanagan, The Logic of Solidarity: Artisans and Industrial Workers in Three French Towns , 1871-1914 (Urbana, Ill., 1980), pp. 214-217; Val Lorwin, The French Labor Movement (Cambridge, Mass., 1966), chaps. 1-5.

[2] Bernard Moss, The Origins of the French Labor Movement, 1830-1914 (Berkeley and Los Angeles, 1976), pp. 25-26. The textile operatives of the Nord often form the basis for such a generalization. See Robert Baker, "Socialism in the Nord, 1880-1914," International Review of Social History 12 (1967): 357-389, and Claude Willard, Les Guesdistes: Le Mouvement socialiste en France (1893-1905) (Paris, 1965).

sonnel engaged in collective protest centered at the workplace. The choice was not between syndicalism and socialism but rather between a syndicalism centered on strike action and one based on solid, carefully constructed organization. The gaziers tolerated patient organization building and an incremental approach to resolving their grievances in inverse proportion to their bargaining power with their employer.

The strategy of forming trade unions to pursue collective interests does not appear to have originated with the workers themselves. The law of 1884, which legalized unions, produced no spontaneous movement among the gas personnel. Formal organization began with the opening of negotiations between the PGC and the prolabor municipal council of the 1890s. The initiative probably came from outsiders and politicians. Their motives were to improve the condition of the workers and to develop a clientele. The first union of PGC workers, a fleeting one, was the project of one Jules Roques. A shadowy figure today, Roques was a journalist who edited the Socialist L'Egalité .^[3] Roques chose to organize stokers exclusively. He presided over their union and penned their demands. On May 1, 1890, he led a strike. Even though the stokers won a raise, they were disillusioned by the outcome of the action (for reasons we shall explore later). They apparently quarreled with Roques, and the union fell apart.^[4]

A second union, having nothing to do with the first and even rejecting its legacy, was formed June 1891. Its membership was entirely different. Stokers were not involved in its formation, nor did they seek to join it. To the second—and more permanent—union belonged common laborers from the courtyards, the lighting service, and the coke department. Initiators of the new movement were Socialist municipal councilmen, who assured gas workers of their support and protection. Jean Allemane, a Socialist who subordinated politics to trade union activity, was quite active among the organizers.^[5] This time the fruits of outside intervention took root. An indigenous leadership appeared to exploit the favorable political circumstances that make unionization possible. The common laborers employed by the PGC joined the union in large numbers. By the mid-1890s the Chambre syndicale des travailleurs du gaz was one of the largest, if

[3] Claude Bellanger et al., Histoire g é nérale de la presse française, 5 vols. (Paris, 1969-1976), 3:372. The police appeared to have no file on Roques even though he was prosecuted for publishing pornographic illustrations. See AP, V 8 O , no. 1063, "Grèves de 1890 et de 1899," dipping from L'Egalité, May 4, 1890.

[4] AP, V 8 O , no. 151, "Grève des chauffeurs (1 mai 1890)."

[5] Ibid., no. 148, "Rapport sur ia réunion du mercredi le 11 novembre 1891 à la Bourse du Travail"; "Rapport sur la réunion du 19 juillet 1891"; Préfecture, B/a, 1424, reports of June 27, November 12, 1891, January 16, and November 22, 1892. The editor of Le Rappel, Charles Bos, was also active in encouraging the union.

not the largest, union in Paris. It was also wealthy and had influence with local politicians and the press. As such, it was rather an exception in French labor history, but one whose evolution is nonetheless instructive.

The origins of this union were inseparable from the foundering of the liberal order at the end of the nineteenth century and the development of a political will to attempt the integration of wage earners into the republic. The municipal council announced its intention of helping workers win their rights, defined largely in terms of better pay. The situation has important implications for the question of relations between French workers and the nineteenth-century state. Scholars have debated whether the state functioned as an arm of capital at moments of labor unrest or whether it pursued a conciliatory role in the interest of public order.^[6] Whatever the intentions of the authorities, the comportment of gas workers points to a vast chasm of distrust between wage earners and the state. Workers did not act spontaneously to gain the state's support. They had taken no steps to importune officials during the numerous confrontations prior to 1890. Even when the company had behaved in a heavily repressive manner, neither stokers nor common hands (with the exception of lamplighters, who were paid in part by the government) had tried to draw the authorities to their side. In truth, the gas workers did not reach out to the state until the state—in the form of the municipal council—reached out to them. That happened as the left majority committed itself to a new round of republicanization. Such an extended hand was necessary to make workers active contenders for power within the existing political order. Even with the hand outstretched from above, a large number of gas laborers, as we shall see, turned their backs on the state. This segment of the labor force was not able to explain its behavior in terms of theoretical principles, but it was acting on the syndicalist assumption that the liberation of workers was the task of workers alone.^[7]

The indigenous leaders who directed the second union never fully renounced their distrust of the state, but they soon developed policies firmly

[6] Arguing for the repressive nature of the state, at least until the end of the century, is Michelle Perrot, Workers on Strike: France , 1871-1890, trans. Chris Turner (New Haven, 1987), pp. 280-303. Describing a more flexible policy are Charles Tilly and Edward Shorter, Strikes in France , 1830-1968 (Cambridge, 1974). The economist Gerald Friedman finds that the American government was more repressive than the French and that French strikes were more likely to be successful as a result. See "Strike Success and Union Ideology: The U.S. and France, 1880-1914," Journal of Economic History 48 (1988): 1-25.

[7] See Peter Stearns, Revolutionary Syndicalism and French Labor: A Cause without Rebels (New Brunswick, N.J., 1971). His assertions that militant syndicalism had the shallowest of roots among the mass of workers and that the vast majority sought only moderate reforms are not fully borne out by this study.

aimed at grasping the extended hand. They were also eager to listen to the paternalistic rhetoric of management set on bourgeois reform. They staked the success of their leadership and of the union on the viability of piecemeal reforms in the seemingly favorable climate of the 1890s. The same nine or ten men guided the union from its origins through the strike of 1899. Their backgrounds were humble and rooted in the milieu of the rank and file. Four had been team heads in the coke yards. The other members of the union's governing board had been commissioned workers in the lighting service, lamplighters, greasemen, and waxers.^[8] These leaders quickly escaped from the influence of Socialist politicians and stamped their own strategies and theoretical positions on the union. Manifesting varying degrees of ideological awareness, they all became reformist syndicalists. That stance by no means isolated them in prewar Paris. Indeed, revolutionaries like Victor Griffuelhes frequently lamented that the capital had no major unions of the ideological coloration of which they approved. The programs of the reformists have largely been forgotten by historians but deserve to be far better known.^[9]

The strongest influence over the Chambre syndicale des travailleurs du gaz in its early years came from Alfred Brard, a former team head from the coke yards. He was the general secretary of the union until 1893, when he won a seat on the municipal council from the district that contained the La Villette plant. Brard's views on labor movements had many affinities with those of Maurice Claverie, his counterpart at the employees' union. Brard was just as much a reformist syndicalist and just as flexible. His electoral platform proclaimed that he "had always thought that theories do not suffice; that fine words do not provide bread; that actions are better than words; that realities are better than hopes, which often cannot be realized anyway."^[10] Though Brard ran successfully as an "independent Socialist" in two municipal elections, he was a stranger to the Socialist group at City Hall. He was too committed to the role of councilman in service of the unions to support the policies of the Socialists on the gas question. Eventually, he changed his political label to a more appropriate

[8] AP, V 8 O , no. 148, "Renseignements sur les principaux ouvriers dits délégués des Travailleurs du gaz"; report of September 17, 1892.

[9] It is interesting to compare the French leaders to the head of the British gas workers' union, Will Thorne. See his autobiography, My Life's Battles (London, n.d.), and his biography, Giles Radice and Lisanne Radice, Will Thorne: Constructive Militant (London, 1974). Thorne was clearly more energetic and imaginative than the people who led the French gas unions.

[10] Prefecture, B/a 195, "Election de 1900, XIX Arrondissement"; AP, V 8 O , no. 148, "Procès-verbaux des audiences données aux délégués du syndicat"; report of July 25, 1891.

"syndicalist." Several Socialist candidates challenged Brard in his last campaign in 1900, and by vying for votes on their right, he not only retained his seat but also won a reputation for being a Nationalist in some circles. The visibility of Brard's rightward drift was reinforced by his qualified support for the Chamon project of 1901, which he viewed as giving workers as much as they could expect. In reality, Brard had not joined the anti-Dreyfusard right; he was simply a reformist syndicalist who avoided large ideological issues and lacked an abiding animosity for capitalism. This position did not make him an anomaly among the union's activists.

After Brard departed from the board to serve the gas workers at city hall, the leadership passed to Jean Darène, who was still more naive about the use of power than Brard and more opportunistic in his quest for immediate gains. His understanding of the workers' position in French society transcended classic divisions between the left and the right. The union newspaper, which Darène edited, expressed his anti-Semitic views. Some-how Jewish stockholders of the PGC were especially guilty of exploitation, and Darène saw the possibility of cooperating with the gentile ones. He was also xenophobic and blamed foreign workers for intensifying the problems that French laborers faced. The antipathy for foreign wage earners that Darène expressed was probably widespread in this union, which may at times have excluded them. n

Darène's hopefulness about reform in this era of liberal crisis was coupled with gullibility and perhaps even an overweening respect for hierarchy. Darène was completely duped by Director Godot's ploy of agreeing to enforce labor rules only to have lower managers violate them. He retained his faith in the director's word long after it was reasonable to do so. The union leader explained the many abuses of authority that his own newspaper reported as a matter of ignorance on the part of top management, which, he insisted, sought to be just in the face of intransigent supervisors.^[12] Darène's guiding vision was some sort of permanent cooperation between well-intentioned management and the moderate union. He perceived in the politics surrounding the gas concession the occasion for a corporatist accord with the PGC.^[13] Ultimately, Darène's lack of critical in-sight led to inept strategies at moments of crisis.

[11] Le Journal du gaz. Organe officiel de la chambre syndicale des travailleurs du gaz , no. 81 (April 20, 1896): 3; no. 90 (September 5, 1896): 3; no. 21 (October 30, 1893): 3.

[12] Ibid., no. 83 (May 20, 1896): 1.

[13] AP, V 8 O , no. 150, letter of the executive commission of the Syndicat des travailleurs to director, September 1892. The union leaders reminded Godot of "the work the union has done to realize a renewal of the monopoly."

The corporatist implications of Darène's approach to union affairs mer-its further investigation, for they were not at all unique. We have already observed the strain of thought in Claverie and Brard. Scholars usually associate corporatism with social thought inspired by Catholicism, preindustrial tradition, and hostility to class analysis. Yet reformist syndicalism and corporatism shared several key assumptions and could readily dove-tail. Both outlooks focused on the interests shared by capital and labor. Both assumed that the employer had the capacity to make major improvements in the lives of bis workers. Corporatists and nonrevolutionary syndicalists agreed that the enterprise, not the political arena, was where reform had to occur. Corporatists were intent on combatting "state socialism," and the union leaders thought of political influence mainly as a source of leverage to negotiate with the employer.^[14] The boundaries between the two outlooks were likely to be especially blurred in this era of nascent corporate capitalism. The separation of ownership and control confused reform-minded syndicalists, it allowed them to represent managers as simply a part of the personnel.^[15] The bureaucratic structure of management at the PGC led Claverie, Brard, and Darène to envision engineers and workers in an easy cooperation—one as the brains, the other as the musde. Furthermore, the freedom that large enterprises enjoyed from market constraints opened the possibility of the company's sharing its monopolistic profits with the personnel. Reformist syndicalists inevitably found attractions in corporatist solutions to labor unrest, particularly at a moment when managers advocated measures of social peace.

Other union leaders were somewhat less optimistic or naive than Darène. They were sometimes openly skeptical about his roseate assessments of management's intentions. Nonetheless, they were just as committed to reformist syndicalism. The policy of delivering ultimatums to the PGC or of striking was not one they advocated. They differed from Darène in having more faith in the good intentions of the municipal council than in the company. They believed that the city could wring further concessions from an employer who was unresponsive, either out of ill will or lack of thought.

Though reformists too, the Socialists on the municipal council found their relations with the gas union leaders to be highly problematical. The Socialist councilmen strove to improve the lives of the gas personnel, but

[14] At the CGT's celebrated Congress of Amiens in 1906, the reformist syndicalists were just as insistent as the revolutionaries on independence from political parties.

[15] At a union meeting in July 1891, Brard affirmed that director Camus was not "an enemy of workers" hut simply an "employee of the stockholders." See AP, V 8 O , no. 148, "Rapport sur la quatrième réunion du syndicat (25 juillet 1891)."

their solution to the gas question, municipalization, required waiting until after 1905 for dramatic ameliorations. Union leaders inevitably found Radicals, and especially the Nationalists, more open to discussing immediate gains for the rank and file. Thus, dose ties between Socialists and syndicalists failed to develop despite a shared concern for the plight of the gas workers. Ironically the leaders of the gas workers might have found more common ground with reformist Socialists had they been less reformist as syndicalists.

The progress achieved by the union was ultimately dependent on the employer's will to grant concessions, and the breakdown of negotiations with the city at the end of 1892 dimmed those prospects. From 1893 to 1898 the union leadership faced frustrations, for the company was not responsive, and the municipal council was not able to impose reforms. There were moments when even Darène lost his optimism. In May 1896 the secretary expressed his disappointment with Director Godot: not that the director had abandoned his concern for the personnel, but he placed too much trust in his subordinates. Darène reluctantly concluded that Go-dot cared more for the principles of hierarchical authority than for social justice. Reaching the acme of disillusionment in late 1896, Darène broached the possibility of some sort of "demonstration" to shake management out of its routine.^[16] The impasse forced the union to reformulate its demands and its strategies. It dropped the long-standing negotiating points of a minimum wage of five francs and equalization of pay and put in their place far-reaching demands, the eight-hour day and tenure for all workers who had been with the firm five years.^[17] The boldness of the new projects was more a sign of resignation than one of radicalization. The union leaders realized that little would be accomplished until the larger gas question was resolved one way or another. They were establishing an agenda for the PGC's successor.

In a speech to the assembled members the union president stated that the goal of collective action was assuring "a certain material well-being [une petite aisance ] for workers."^[18] Though there is no conclusive way to survey the views of the rank and file, there is equally no reason to suppose they were fundamentally out of step with the goals and tactics of the indigenous leaders. Gas workers did flock to the union in large numbers;

[16] Journal du gaz , no. 92 (October 5, 1896): 3.

[17] The minimum wage was realized as a result of the 600,000-franc grant in 1896. Municipal politicians may have been behind the new demands. The city tenured its own workers in 1895. See AP, V 8 O no. 154, "Mesures prises par le Conseil municipal en vue de l'amélioration de la situation du personnel."

[18] Ibid., no. 148, "Rapport sur la réunion du mercredi 11 novembre 1891."

the company, the police, and the municipal council concurred that the majority were members. The figure of 6,783 members that Darène advanced in May 1899 was creditable and implied that nearly every common laborer belonged.^[19] The superintendent of the La Villette plant understood the strength of the union's attraction enough to be troubled and insulted by it. Cury noted that until 1892 workmen had come to him to air grievances or ask favors. Once the union era opened, the personnel relied on the dele-gates.^[20] It is clear that the leaders did try to discover the will of the rank and file and act on it. We have already seen the popularity of the five-franc minimum wage, and union officers took up that demand for a long time, often against their better judgment. When the PGC first granted the profit-sharing plan, some leaders were inclined to celebrate it as a major victory and give low priority to the minimum wage. They soon learned, along with management, that workers were attached to the latter goal and would not count their year-end bonuses as part of the wage. Godot insisted that officers convince workers of their good fortune, but the leaders expressed helplessness to do so. The lamplighter Chapelle affirmed that he personally "was in agreement with the company in terms of principles" but that his members "were not mathematicians and did not understand these calculations."^[21] Union leaders prudently abandoned the effort to make the rank and file understand. The union ended its meetings in 1893 with the cry "Long live the five-franc minimum wage!" but not necessarily because leaders wanted that to be the rallying point.^[22]

The leaders' respect for managerial prerogative and their optimism about the good intentions of upper management also had resonances among the mass of wage earners. Common hands did have trouble with immediate supervisors, and their vocabulary expressed it. They referred to the foreman as le patron, with all the resentment and fear that that title implied.^[23] Indeed, the alleged injustices perpetrated by the foreman had long been sources of complaint and would soon become major themes of protest. Nonetheless, the rank and file shared their leaders' expectation that the engineers who headed departments would be men of wider vision and of equity. That is to say, Emile Cheysson's view of engineers as disinterested authorities ready to replace the arbitrary foreman had credence on the shop floor. Similarly, Darène's faith in the director's justness had its counterpart below, in the internalization of a paternalistic ethos among

[19] Journal du gaz , no. 154 (May 5, 1899): 1.

[20] AP, V 8 O no. 148, report of Cury, March 24, 1892.

[21] Ibid., "Procès-verbaux des audiences données aux délégués du syndicat," p. 3.

[22] Ibid., no. 148, clipping from Le Journal des débats , May 28, 1893.

[23] Ibid., no. 159, Louis Saget to director, n.d.

general laborers. Workers initiated contact with the chief executive to ask for his aid and for his intervention in their personal problems. One coal hauler, who was barely literate, wrote a rambling letter to Godot, full of complaints about his ungrateful and lazy sons. The worker had nothing to request from the director, but he apparently needed to explain his situation to his employer.^[24]

One historian of British labor has commented on workers' fractured and indistinct notion of employers as a class, and still more as "the enemy."^[25] This vision is clearly relevant for the mass of workers at the PGC as well as for union leaders. The identification of friends and enemies was all the more difficult in the new sort of enterprise that was arising, with its bureaucratic layers of authority and its encumbered face-to-face relations. Thus, the failure of union officials to confront management aggressively and the respect they generally accorded engineers met with comprehension from the rank and file.

The mass membership achieved by the gas union provides insights into the French labor movement because it was so uncharacteristic. The ability to mobilize so many laborers as permanent, dues-paying members casts doubt on the common arguments that French wage earners were too fiercely independent to submit to formal organization or too entranced by a romantic, revolutionary faith in apocalyptic action.^[26] French laborers were willing to unionize when there was little danger of revenge from the employer and when that was the best option for obtaining results. The gas workers' union also stands out for having attracted the unskilled rather than the most qualified laborers. The aberrant situation points not only to the exceptional political circumstances but also to the influence of the workplace situation in determining the shape of collective action. At the mercy of their employers, the common hands readily seized the opportunity to gain some control over their fate by grasping the extended hand of the state. Stokers, by contrast, could rely on informal cooperation to re-dress grievances because of their relative power to disrupt operations at the workplace. Such power encouraged them to pursue goals that Darène's

[24] Ibid., no. 162, J. Depoix to director, n.d.; no. 150, "Demandes de secours, 1875-1899." Elinor Accampo has argued that paternalism had a strong appeal to the workers she studied. See Industrialization, Family Life, and Class Relations: Saint Chamond, 1815-1914 (Berkeley and Los Angeles, 1989), chap. 5.

[25] Patrick Joyce, "Labor, Capital, and Compromise: A Response to Richard Price," Social History 9 (1984): 67-76.

[26] Reaffirming these long-standing stereotypes are Gordon Wright, France in Modern Times, 3d ed. (New York, 1981), p. 291, and Gérard Noiriei, Les Ouvriers dans la société française: XIX -XX siècle (Paris, 1986), pp. 263-267.

union dared not countenance.^[27] The organization of skilled industrial workers did not proceed along the same lines as that of the unskilled, and reformist syndicalists were not prepared to deal with the skilled workers.

The union officials were ideologically and temperamentally out of their element with the upsurge of labor militancy created by the stokers in 1898. Darène in particular developed tactics that were inept and ultimately unreasonable. The union had always hoped that the stokers would join, but when they finally did, the leaders were alarmed by the spirit of vindictiveness. The "winds of revolution" blowing through the union, against which Darène warned, appeared quite threatening—all the more so in that further concessions from the PGC suddenly appeared possible once again. Another round of talks between the company and the city opened in 1899. Hope of movement on one front and fear that militants could undermine progress pushed Darène to undertake a bold, but futile, experiment. The union secretary was ready to abandon a neutral stance regarding the gas question and use the moral authority of the union to obtain a prolonged charter for the PGC. In return Darène had a list of benefits that management would grant to the workers—larger pensions, the eight-hour dar and tenured positions. In July 1899 he began to establish his stance by writing in the union newspaper that "the most resolute enemy of the worker is not the company but the city. The company would grant our demands, at least in part, if it could be certain of a renewed charter." A few days later (July 20) the union journal openly announced the corporatist policy of working with the PGC to obtain benefits for both it and the personnel but added that the union was "ready to combat our ally if it does not give us complete satisfaction."^[28] The tactic was of course infused with Darène's naive optimism regarding the good intentions of management. It was particularly ill timed in that the war of nerves between management and stokers was near an explosion. Indeed, Darène made his plan public just before Godot gave permission to crush the union should any strike occur. Clearly, the union leaders were badly out of touch not only with the engineers' intentions but with the stokers' militancy as well. The stokers were forcing the union to accept a new agenda that reformist syndicalists wanted desperately to eschew.

[27] See Alain Cottereau, "The Distinctiveness of Working-Class Cultures in France, 1848-1900," in Working Class Formation: Nineteenth-Century Patterns in Western Europe and the United States, ed. Ira Katznelson and Aristide Zolberg (Princeton, 1986), pp. 111-154. Cottereau would find the stokers' tactics and strategies characteristic of a wide segment of the French labor force.

[28] Journal du gaz , no. 158 (July 5, 1899): 1; no. 159 (July 20, 1899): 1. On the behind-the-scenes cooperation, see AP, V 8 O no. 160, "Août 1899."

The Crisis of Industrial Discipline

That there were major differences in collective behavior and outlook be-tween stokers and general laborers is a revealing finding. After all, the two groups were one in most ways. They had identical social backgrounds. Stokers were recruited among the common laborers and did unskilled work for part of the year.^[29] As stokers aged, they reentered the ranks of the general hands. Both types of wage earners became dependent on the PGC at roughly the same time. Despite these shared experiences, there was an undeniable gulf between stokers and general laborers. The stokers failed to provide leadership for their more oppressed comrades, and union formation divided the two still further. The stokers' indifference to permanent organization made them far more representative of the nineteenth-century French labor movement as a whole. Moreover, their periodic, direct challenges to managerial authority after 1890 also reflected wider currents of unrest among skilled laborers in France.

Though the stokers did not join Brard's organization in significant numbers, they were not adverse to unionization in principle. Indeed, they had formed the first trade union in 1890. The police believed that all or nearly all of the stokers had joined it. Its organizer, Jules Roques, had used his newspaper to defend the stokers' cause when agitation surfaced in the spring of 1890. Under Roques's presidency the union strove to be exclusive. Its high monthly fee of five francs (more than a day's wage for the unskilled) opened it only to stokers.^[30] From the beginning of the union era stokers were destined to act separately from general laborers.

Stokers did not need a union to win concessions from their employer, as their history before 1880 demonstrated. Yet their concerns shifted, and new interests encouraged new strategies. Stokers eagerly joined a union in 1890 because they were angry, and the threat of a strike emerged at once. The issues mobilizing them now were quite different from the wage concerns that had generated agitation before 1880. As dependent workers, stokers of 1890 were upset by the arbitrary pattern of layoffs the PGC practiced. The stokers understood seniority as the valid basis for according the right to continue to serve in the distillation rooms during the slow season. But factory managers refused to honor the principle, which the director had accepted, in a systematic manner. This grievance first brought the stokers and the Socialist journalist Roques together. He pleaded their cause in L'Egalité, the only journalist to do so other than Henri Rochefort.

[29] AE V 8 O , no. 163, "Personnel ouvrier—jours de congé."

[30] Ibid., no. 148, report of July 25, 1891.

The stokers were also discontent with the company's administration of their sick-pay fund. Though the stokers contributed 1 percent of their nine-to-ten-franc wages to the fund, the company awarded them only two or three francs as "half pay" when they missed a day of work because of illness. Moreover, there were changes of details in the work routine that stokers found irksome. Managers had recently changed the mix of coals, increasing the portion of a hard, heavy variety (cannel coal). The change made the stokers' task more difficult. The stokers now wanted some checks on the arbitrary power of the company and limits on their exploitation.^[31]

The chargers relied on Roques to translate their grievances faithfully into formal demands, but he did not do that well. The statement he penned on April 29 mentioned the augmentation of the "already super-human work load" but he stressed the need for higher pay. The journalist made a raise from three to five francs per metric ton the central demand.^[32] A pay increase, however, was no longer the primary matter to stokers, and gaining it did not produce satisfaction. Caught by surprise, the PGC immediately promised a raise, but the stokers struck anyway on May 1, 1890, France's first Labor Day Some 350 stokers remained off the job for a week. The timing of the action, coinciding with the new festival of labor power, expressed the complexity of the stokers' challenge better than did Roques's negotiating position.^[33] He and the company quickly came to an agreement on a 10 percent raise (3.30 francs per metric ton of coal distilled) and justice on the matter of sick pay. In the meantime, the PGC had come close to reestablishing full production by recruiting seasonal stokers from the provinces. The strikers drifted back to the gas plants to the extent their employer would have them.^[34]

In a narrow sense, the strike had been a success: the company had granted a raise, and management was licking its wounds over the defeat. Nonetheless, the stokers did not define the settlement as successful, though they surely would have prior to 1880. A police report noted that workers "would rather have had [concessions] on lighter work loads than on wages."^[35] Moreover, the raise by no means ended the grumbling

[31] Ibid., no. 1063, "Grèves de 1890 et 1899."

[32] Ibid., no. 151, Roques to director, April 29, 1890.

[33] On this festival of labor, see Michelle Perrot, "The First of May 1890 in France: The Birth of a Working-Class Ritual," in The Power of the Past: Essays in Honor of Eric Hobsbawm , ed. Geoffrey Crossick et al. (London, 1984).

[34] AP. V 8 O , no. 151, "Grève des chauffeurs (1 mai 1890)."

[35] Prefecture, B/a 176, "Grève des chauffeurs gaziers, 1890." See, especially, the report of May 13. There was a similar pattern to the large strike wave that hit the textile industry of the Nord on May Day, 1890. See William Reddy, The Rise of Market Culture (Cambridge, 1984), pp. 305-309.

among the stokers, and managers reported on rumors of new strike actions through June.^[36] Far from being bolstered by a recent success, the stokers' union soon fell apart as a result of recriminations and disappointments among the rank and file. Eight years later Darène and his comrades could still cite the 1890 strike as a disillusioning experience that kept stokers from joining the gas workers' union.^[37]

The sudden effervescence of 1890 was indeed followed by eight years of somnolence among stokers. Immediately after that strike the company was able to undertake its most aggressive attack to date on the source of the stokers' bargaining power—without meeting any visible resistance whatsoever. Factory managers finally substituted the scoop for the shovel (see chapter 4), thereby reducing the dexterity and, to a lesser extent, the endurance necessary to charge retorts. The threshold of recruitment was lowered without quite bringing stoking to the level of common labor.^[38] The stokers' acquiescence made managers crow about their recovered sense of control over the labor force. Cury, the ancient superintendent of the La Villette plant, was so confident of his ascendancy that he even sup-pressed the fifty-centime bonus that chargers had long received. That move earned him a slight rebuke from his superior, who was not eager to press his luck, on the grounds that the change had been done "not to save money but rather to reassert control over the workers."^[39]

Disillusionment with the recent strike may have been one reason for the stokers' passivity in view of the fundamental threat to their bargaining power. It was also the case that the chargers conspired in their own deskilling. The company had imposed the scoop during the height of the slow season. Not only were the workers aware of their good fortune to have jobs, but also their opposition to the tool was weakened by the thought that it could prolong their days in the distillation room. By rendering charging somewhat less taxing, the scoop favored a longer career at high pay. Apparently disillusionment, coupled with the caution imposed by dependency, undermined protest over a serious attack on the stokers' status as skilled workers.

[36] AP, V 8 O , no. 151, report of Cury to Euchène, June 13, 1890.

[37] Journal du gaz , no. 134 (July 5, 1898): 3.

[38] AP, V 8 O , no. 148, "Résumé des avis donnés par les chefs de service sur les demandes formulées par la chambre syndicale (décembre 1891)."

[39] Ibid., report of Gigot, October 21, 1890. The bonus had been given for many years to the stokers who worked the shovels; Gigot insisted that superintendents give the bonus to the stokers who took the handle of the scoop.

The birth of a new gas workers' union in mid-1891 left the vast majority of stokers indifferent. Managers continually congratulated themselves that stokers were staying away from the Chambre syndicale des travailleurs du gaz.^[40] Even the union newspaper revealed (in 1898) that although eight-ninths of the common hands had joined, only a quarter of the stokers, at most, were members.^[41] The aloofness of the stokers, reflecting the very different perspective on protest separating them from workers with whom they otherwise had much in common, entailed divergent strategies and goals. The skilled workers now disdained the wage-centered concerns of the reformist union. Moreover, they distrusted the patient organization building and negotiations with public authorities. The stokers' power on the shop floor, weakened but not eliminated by the scoop, allowed them to remain faithful to a direct-action strategy even though the public authorities were willing to help them procure material gains. The stokers in-tended to use their bargaining power to extract concessions directly from their employer Their passivity regarding the union was not at all a matter of contentment. In fact, the stokers were seething with resentment, but the union was unable to tap it. That anger finally exploded in the second half of 1898.

During the summer and autumn of that year—a moment of general labor unrest in Paris—the stokers began to flock to the union. Eventually as many as a thousand may have joined within the space of just a few months.^[42] The union leaders, at once pleased and apprehensive about the new recruits, organized plant-by-plant meetings for the purpose of hearing complaints and channeling the resentment into acceptable directions.^[43] Yet stokers were not interested in listening to the reformists; they had their own reasons for entering the union en masse. They were intent on forcing the union to confront their central grievance—the arbitrary power of superintendents and foremen to insult, punish, lay off, fire, and reward. The sudden unionization of the stokers announced that they were willing to abide such authority no longer

The source of the stokers' vexations in 1898 was not a recent problem, nor was the grievance felt by chargers alone. Darène's organ, the Journal du gaz, had from its initial issues reported on supervisors' unjust or vindictive treatment. Long before the stokers provoked collective protest over the matter, the union sheet had lamented the freedom of foremen to ter-

[40] Ibid., report of Gigot, April 15, 1892.

[41] "Usine d'Ivry," Journal du gaz , no. 134 (July 5, 1898).

[42] Journal du gaz , no. 137 (August 20, 1898): 1; no. 142 (November 5, 1898): 1.

[43] Ibid., no. 135 (July 20, 1898): 1. Every issue thereafter (up to the strike of August 1899) had lengthy reports on these rallies.

rorize workers by threatening to fire them, deprive them of a bonus, or demote them to poorly paid work. The supervisors' arrogance, bribe taking, utter lack of civility and pleasure in seeing workers cower were common themes in the union press.^[44] Louis Sagot, a barely literate courtyard worker, complained to Director Godot that his patron, the team head, had laid him off while retaining a man who had far less than Sagot's seven years of seniority. Sagot felt the sting of the injustice all the more in that he had followed the team head's orders to frequent a particular café "Besides," Sagot wondered, "weren't layoffs supposed to be by seniority [par numéros] ?"^[45] The Journal du gaz may well have expressed a common sentiment among the native workers in linking this sort of complaint to immigrants. The charge was that foreign workers were especially supine, and their subservience brought them favors from supervisors while proud French workers suffered. The union press found young French stokers, newly arrived from the provinces, almost as culpable of forgetting their dignity and buying favors from foremen. The Journal du gaz fumed against spineless Auvergnats, "whose approach to the foreman is distinguished by heavy servile steps, and breath redolent of sauerkraut and sausage.^[46] Quite possibly, the union paper was projecting on a clearly identifiable minority the self-hatred gaziers felt as a result of their own subservience. In any case, the authority of supervisors was clearly a painful dilemma for all the PGC's workers. Unlike common laborers, stokers committed themselves to resisting it as the century drew to a close.

In truth, resentment of the power of immediate superiors was an international problem that inflamed industrial relations in all advanced countries of the late nineteenth century.^[47] Only in France, though, did the grievance precipitate a widespread industrial crisis, and the stokers were full participants in it. The stokers' resentment arose from a corporate power structure that retained strong face-to-face elements despite the firm's large size and bureaucratic division of labor. Within the PGC, at least, the distribution of authority was the result of management's conscious designs. We have seen that the director sought to distance himself

[44] See, for example, ibid., no. 11 (May 20, 1893): 3; no. 20 (October 5, 1893): 2; no. 29 (February 20, 1894): 2. The police were well aware of the grievance too. See Préfecture, B/a 1424, report of June 13, 1893.

[45] AP, V 8 O , no. 159, Louis Sagot to director, n.d.

[46] Journal du gaz , no. 68 (October 5, 1895): 3.

[47] See, for example, Walter Licht, Working for the Railroad: The Organization of Work in the Nineteenth Century (Princeton, 1983), pp. 171, 254; Tamara Hareven, Family Time and Industrial Time: The Relationship between the Family and Work in a New England Industrial Community (Cambridge, 1982), p. 304; Will Thorne, My Life's Battles (London, n.d.), p. 74.

from the management of individual factories. Responsibility for daily operations resided with the division head and plant superintendents. They, however, were preoccupied with a multitude of matters and in practice left personnel matters to immediate supervisors, foremen, team heads, and paymasters (pointeurs ). The rules of the corporation formally denied foremen the right to hire, lay off, or fire, but these powers came to be part of their duties. So much was this the case that both the engineers and the workers casually spoke of foremen doing the hiring and firing.^[48] Even when an engineer may have regretted his foreman's decision in a particular instance, he dared not reverse it for fear of calling authority into question. Before the union era the informal power of immediate supervisors had been a matter of convenience for busy engineers, but once workers began to organize, it became part of a managerial strategy to combat the union. The "good intentions" Darène attributed to Director Godot quickly disappeared down the hierarchical chain of command. Commitments made at the top were never enforced, and the director could defend himself by pro-claiming his powerlessness. Engineers eagerly harnessed the foremen's venom for the unions in their struggles to preserve what they could of authoritarian paternalism. Delegating authority in that way often meant losing control over supervisors and permitting them to use their command for personal gain.^[49] Yet the engineers accepted the abuses as a lesser evil.

Whether the corruption and oppression that the stokers denounced were as pervasive as they contended is impossible to say. There is no doubt, though, that their grievances had a basis in reality. The PGC disciplined a foreman and a paymaster for running cafes in 1888 and tried to prevent supervisors from operating shops near the gas factories.^[50] Yet as long as the engineers relied on foremen to intimidate workers, they could not easily control the supervisors' lives off the job. An incident that occurred in 1899 involving a union delegate substantiates the charges of venality The delegate was about to be laid off as a delayed punishment for his participation in the March strike, and he offered his foreman ten francs to retain his post. The factory engineers learned about the bribe and sought to use it to cast opprobrium on the union. Management ostentatiously challenged workers, "all too ready to accuse foremen and team

[48] AP, V 8 O , no. 763, "Rappel du règlement, 20 avril 1869"; no. 150, Darène to Lecoeur, February 16, 1893; no. 148, report of Cury, October 22, 1891; no. 151, "Revendications générales—personnei ouvrier."

[49] Workers often complained about foremen who exploited them for their personal interest, but the company never followed up on the charges. See Journal du gaz , no. 37 (June 20, 1894): 3; no. 148 (February 5, 1899): 2.

[50] AP, V 8 O , no. 685, deliberations of January 25, 1888.

heads of venality" to stop tempting supervisors.^[51] The company may have been successful at embarrassing the union, but it also admitted, in an oblique manner, that payoffs were a problem. There is no reason to sup-pose that the practice was of recent origin, but the stokers' unwillingness to abide the abuses any longer was new. Their militancy in 1898 was in effect a revolt against both the abuses and the very nature of face-to-face authority.

When the union began its round of mass meetings in response to the influx of stokers, the leaders' hope was to convince the new members that an eight-hour day was what they needed. The proposal entailed handling two four-hour charges a day instead of three. A serious drawback for the proposal was that stokers would have to accept a large pay cut for doing a third less work. Darène never tired of explaining to skeptical chargers that they were better off earning 7.50 francs for eight hours than 11.50 for twelve. He insisted that they would have to purchase only one restaurant meal a day. More important, cutting the workday would raise the number of posts the company needed to fill by a third. Thus, the stokers' jobs would be secure, and the chargers could easily get year-round work if they wished. Moreover, Darène reasoned that with the cruel physical toll of three charges a day reduced, stokers might expect to prolong their years in the distillation room.^[52] Such arguments might have been influential in 1890, when stokers had refrained from protesting the scoop, but the logic no longer convinced them. The urge to put supervisors in their place over-whelmed practical arguments this time.

The stokers barely disguised their impatience with the eight-hour proposal. Instead, they took over the rallies and focused the discussions on their own issue—the discriminatory authority of supervisors. Stokers complained about their humiliation at the hands of vindictive foremen. They lamented the unjust fines and arbitrary job assignments and bitterly recalled the many times supervisors had abused rules on layoffs. The stokers used the meetings to represent themselves as victims of venality, vanity, and vengence. Foremen, they asserted, rewarded the most subservient of workers and persecuted those who stood up for their rights. They argued forcefully that bribery permeated power relations on the shop floor. Supervisors required them to patronize the stores their wives ran, drink in cafes with which they had a financial connection, and lodge in boarding houses that made payments to the foreman. Avoiding a seasonal layoff, the stokers affirmed, required a bribe. The chargers made it clear in the

[51] Ibid., no. 159, report of Hadamar to Euchène, March 9, 1899.

[52] Journal du gaz , no. 133 (June 20, 1898): 2; no. 135 (July 20, 1898): 2.

fall of 1898 that the eight-hour day, even with its practical promise of steady work, was not what they cared about most deeply Instead, they manifested a powerful drive to gain control over the work experience— not in the sense of enforcing a "craft control" over production, which they had never enjoyed, but rather by compelling superiors to govern by fixed, uniform, and equitable rules to which workers had given their consent.^[53] Their surge of militancy was a gesture of rebellion against managerial authority.

So forceful was the stokers' expression of victimization that they compelled union leaders to abandon their own agenda. The leaders had to admit that the eight-hour day was not winning a commitment from the chargers. Already in June 1898 Darène announced his intention of using the union newspaper to expose the acts of arbitrariness committed by supervisors.^[54] Having been impressed by the urgency of the issue, Darène took it upon himself to raise it with Godot. By the end of the summer the union was claiming a victory: the director had promised to order superintendents to follow general rules and administer punishments uniformly,^[55] Yet Darène was not able to stem the rising tide of militancy because the new order brought no better results than had previous ones. Throughout the fall and winter the Journal du gaz lamented that vicious supervisors were destroying trust between the company and the union and forcing the latter to become confrontational. Finally Darène capitulated entirely to the stokers and adopted as the union's primary goal the implementation of fixed and uniform rules (règlement unique ) governing all aspects of work.^[56] The militant stokers had come to dominate the union after having ignored it for so long.

As we have seen, the stokers had always been practitioners of direct-action syndicalism avant la lettre, and true to their habits, they did not wait for reformist union leaders to guide, channel, or propose peaceful means of negotiating over authority. Instead, they began an incipient insurrection within the gas plants during the fall of 1898. It was this offensive that made Darène warn of "a revolutionary wind, the reign of dangerous agitation." He was not the only union official to lament the members' sudden will to "crush everything . . . to forment a revolution."^[57] Another

[53] Ibid., nos. 137-160. Licht, Working for the Railroad, pp. 260-271, makes the important distinction between control over the work process and control over the work experience .

[54] Journal du gaz , no. 133 (June 20, 1898): 2.

[55] Ibid., no. 137 (August 20, 1898): 2.

[56] Ibid., no. 150 (March 5, 1899): 1.

[57] Ibid., no. 148 (February 5, 1899): 1; no. 150 (March 5, 1899): 2.

reformist retrospectively described the situation prior to the strike of 1899 in precisely the terms that management might have used: "Workers burned with a desire to start a struggle. . .. In the factories, managers were no longer masters. . .. The acceptance of discipline had disappeared."^[58] The Socialist councilman Blondeau might have agreed with that account of a crisis of authority too. When he attended a mass meeting of stokers in March 1899, he was frightened by the soaring spirit of vengence and the cries of "Vive la grève générale!' Blondeau's calls for caution brought expressions of disdain from the assembled stokers.^[59]

Production engineers, for their part, were painfully aware that stokers had suddenly turned the tables on them and were administering some "salutary intimidation" of their own. The so-called Laurent affair of October 1898 was the opening round of the showdown that the engineers had long predicted. On the advice of a hated foreman, Laurent, the superintendent of the La Villette plant fired a stoker. His comrades viewed the firing as an act of injustice and put down their tools. The superintendent, now alarmed by the gravity and solidarity of the stokers, backed down and admitted that a mistake had been made.^[60] This incident, occurring at a moment of intense labor unrest in Paris, set a precedent for future confrontations in gas factories.

A repetition of the general features of the Laurent affair next March pushed tensions to the breaking point as far as management was concerned. This outburst occurred at the Clichy plant, which was run by the most severe superintendent, Hadamar, and by the most vindictive fore-men. Correspondingly, the spirit of revolt was strongest among the stokers of that factory, according to Darène.^[61] Under these circumstances the aggressive behavior of Hadamar was exceedingly provocative—probably intentionally so. The superintendent had planned for some time to fire two general laborers for poor work and insolence. Instead of taking the step at the moment of their offense, Hadamar waited for the slow season to commence and laid them off in spite of six and twelve years of seniority. Stokers, highly sensitive to the principle under attack and in no mood to compromise, immediately issued an ultimatum to rehire the workers or face a strike. The next day 200 of the 230 stokers left their posts. At least half of their comrades at the La Villette and Landy (Saint-Denis) plants joined them. Hadamar was more than eager to pursue the fight, but the director could not allow the conflict to continue because of negotiations

[58] Ibid., no. 164 (October 5, 1899): 2.

[59] AP, V 8 O , no. 159, "Rapport sur la réunion du 21 mars 1899."

[60] Journal du gaz , no. 141 (October 20, 1898): 2-3.

[61] Ibid., no. 148 (February 5, 1899): 2.

with the city. He ordered the rehiring of the two workers and allowed all the strikers to return to their jobs. The superintendent and foremen had to overcome their sense of humiliation in the face of the reported "insolence" as the workers triumphantly reentered the plant.^[62] At this point the engineers concluded bitterly that the stokers had become their masters. Though the stokers had previously called foremen their patrons, they now seemed resolved to escape such subordination.

These victories gave stokers not only a heightened sense of power—the "insolence" to which the engineers referred—but also a sense of elation and even generosity. The latter sentiments emerged when they tested authority one more time before the disastrous strike of August. In the wake of the strike at Clichy, superintendents prepared for more agitation by substituting a smaller and lighter scoop, which could be handled by workers of average strength. Stokers were quite aware of the strategic implications of the new tool and, in addition, found that it prolonged their work routine.^[63] Thus, delegations at each plant asked superintendents to return to the large scoop. Significantly, the delegations bypassed normal union channels and went directly to the plant managers. Once again the superintendents backed down (though perhaps only because they now had a more sinister trap in mind).^[64] The manager of the Passy plant could not help noticing how congenial his stokers became the moment their request was granted: the stokers treated the large scoops with great care and worked with renewed energy. He reported that the workers lost the "scowling air" they had had for the past weeks. Now, the superintendent noted, the stokers had become "good-natured; they banter with the fore-men—something they had not done for a long time."^[65] Evidently; the stokers were basking in their newly discovered power. It did not take much to make these physical giants satisfied—even submissive. Unfortunately for all concerned, managers were not the least bit inclined to compromise even when they recognized that the stokers were easily appeased by token gestures and quick to obey those who accorded them a modicum of respect.

That the eight-hour day was the subject of the last and fateful test of authority shows how little power the stokers had, in the end, to shape issues. Yet that was hardly the only peculiar feature of the strike of August 1899. "What a bizarre strike it was," asserted Darène when it was

[62] AP, V 8 O , no. 159, "Grève, mars 1899."

[63] It took two fillings of the large scoop to load a retort and three fillings of the small scoop.

[64] AP, V 8 O , no. 160, "Renseignements sur le service."

[65] Ibid., report of Bodin to Gigot, July 1, 1899.

over. "Its cause was a demand for lower pay!"^[66] In a superficial sense the strike broke out over a trivial pay demand; but that demand ultimately stood for the issue of authority at the workplace. It will be recalled that the engineers, probably following a plan to manipulate the stokers into striking, initiated a companywide trial of the eight-hour day with two charges.^[67] Management did this in mid-July even though the union was no longer pressing the issue. The company decreed a pay formula for the trial that was bound to be confusing and inflammatory. Workers would receive four francs per metric ton of coal distilled for the first 1,750 kilo-grams and then only 3.30 francs for each additional ton. It was inevitably troubling that the base pay was higher than the incentive rate. Stokers understandably believed rumors that the company was prepared to pay four francs for each ton.^[68] To the extent that the seventy-centime difference directly contributed to the strike, it was because the higher pay cushioned a bit the one-third reduction in earnings that would result from the eight-hour day.

The pay issue became confused with larger matters when the union hastily sanctioned a strike over the demand for four francs. That action in itself was a sign of events spinning tragically out of control, for the union leaders were not at all favorable to aggressive action. At a poorly attended general meeting at the Bourse du Travail on the night of August 7, the demand for four francs was made, and Darène solemnly warned against it. The eight-hour day was part of the accord he wished to make with the company, in return for which he would put the union behind a renewed charter for the PGC. Moreover, he had been urging management for some time to bolster the prestige of the moderates and calm the "winds of revolution" in the union. Darène apparently believed that the eight-hour trial was the answer to his pleas. When that very concession produced further agitation, he became paralyzed. Late in the night he allowed the handful of workers present to take a strike vote, and it passed, 40 to 15. Then Darène fatefully committed the union to the action he had just op-posed, a move that would later be the basis for charges that he was cooperating with the company to rid the PGC of its militants. About 750 out of 1,100 stokers answered the union's strike call.^[69]

[66] Journal du gaz , no. 161 (August 20, 1899): 2.

[67] On the likelihood that managers plotted to foment the strike, see chapter 4.

[68] There are some suggestions that the union leaders themselves spread the notion that the PGC would pay workers four francs, perhaps to promote enthusiasm for the eight-hour trial. See AP, V 8 O , no. 151, report of Euchène to Hadamar, July 24, 1899.

[69] Ibid., no. 159, "Pièces adressées à Monsieur le juge de paix du XV Arrondissement"; "Réunion extraordinaire du 7 août," Journal du gaz , no. 161 (August 20, 1899): 1.

There can be no doubt that authority, not pay, was the real issue in the strike. The action was entirely out of proportion to the demand. With four francs stokers would have earned fifteen to thirty centimes more a day, a pittance considering they were already taking a huge pay cut. The strike was a gesture of insubordination. Stokers disliked the way the company had suddenly disrupted their work routines. They resented the peculiar pay scale, about which the company had been absolutely inflexible, and they were indignant about having to face such a deep pay reduction. With the union having committed itself to the strike, the stokers felt they had its authority to uphold. Above all, the stokers were attempting to maintain the momentum of the Laurent affair and the March strike. They were unwilling to allow a return to the company's authoritarian paternalism. Management understood the strike in precisely that manner. The engineers explained it as being rooted in a "spirit of indiscipline," which was raging out of control. Granting four francs would have trivial financial consequences, the chief of production admitted, but "the loss of authority would be incalculable." In his view, the strikers were truly "rebels" who physically threatened foremen and were 'a danger for the company, for Paris, for society."^[70]

The engineers' inveterate pessimism about workers' character at last yielded genuine insight. Though the strike was not about wages, neither was it simply about regular, uniform work rules. There is a case to be made that stokers looked past the matters of security and equity to engage in a gesture of rebellion for its own sake. For several years they had disregarded the blandishments of politicians intent on being generous to la-borers. They had given no support to the campaign for an eight-hour day, despite its practical promises. And their alarming militancy, which terrorized normally self-confident managers in 1898-1899, signaled an anger that transcended reasoned calculation of interests. The stokers surely wanted an application of the règlement unique, but they were also groping to declare their utter opposition to managerial authority.

The strike of August 8 had little chance of success. Stokers were divided and perplexed by the eight-hour issue with such a steep reduction in pay. Those at the Passy plant chose not even to participate in the trial.^[71] The demand for four francs addressed no basic issue whatsoever. The company,

[70] AP, V 8 O , no. 159, "Grèves de mars et août 1899," report of Euchène, September 14, 1899.

[71] Ibid., "Instructions è Messieurs les régisseurs (7 août 1899)."

for its part, had made excellent preparations. It quickly rounded up seasonal stokers from the provinces and promised them year-round jobs. During the especially steamy weeks of August 1899, demand for gas was minimal anyway. Moreover, the union treasurer absconding with the strike funds devastated morale. It is little wonder that superintendents found the atmosphere in the distillation rooms entirely transformed after the strike was over Workers were "respectful" of the foremen and happy to have jobs.^[72]

Management was adamant about excluding many strikers from the gas plants, not only in the immediate aftermath of the strike but also permanently.^[73] The municipal council passed one resolution after another over the next decade in favor of the unemployed strikers and even appropriated funds to aid their families, but the engineers could not be moved. They held to their view of the strikers as dangerous rebels who attacked legitimate authority. Indeed, the managers were ultimately ready to pay the price of career security to defend the principle of authoritarian paternal-ism. The PGC came to an end in 1907 without having rehired all the strikers of 1899.

<><><><><><><><><><><><>

The militancy of 1898-1899 was at once complex, puzzling, and revealing. It is time to proceed from narrative to analysis in order to explore the implications of the gas workers' protests. Examining the strike from the perspective of the gas firm alone is inadequate. By no means were the stokers unique in challenging managerial authority as the nineteenth century drew to a close. We have seen that gas clerks and other white-collar employees had already been protesting discretionary power when the stokers' militancy surfaced. Manual workers were also turbulent. The Laurent affair occurred just as Paris was in the grips of an upsurge in labor unrest. Navvies had declared a strike; railroad and construction workers had followed them in the fall of 1898. This was the first nationwide at-tempt at a general strike. Stokers first forced gas management to back down in a city filled with more than forty thousand strikers.^[74] Challenges to the authority of foremen had already been a major theme in Parisian strikes for several years, and workers' aggressiveness on the issue climaxed as stokers were rallying to Darène's union. The machine-building

[72] Ibid., "Grèves de 1899—articles de journaux."

[73] Ibid., report of Gigot to director, November 10, 1899. The company considered 147 stokers too "dangerous" to rehire.

[74] France, Office du travail, Statistique des grèves et des recours à la conciliation et à l'arbitrage survenus pendant l'année 1898 (Paris, 1899), pp. 108-109, 112, 252-270.

industry of Paris and the suburbs erupted just before the PGC did. At the beginning of 1899 automobile workers at one firm walked out because a foreman fired a worker for being five minutes late. Only a few weeks be-fore the stokers initiated their ill-fated strike, mechanics at the imposing Dion-Bouton Automobile Company just outside Paris insisted on the firing of a foreman whom they despised. The haughty marquis de Dion could not crush their revolt and in the end yielded. The militancy against industrial authority persisted for several more years in the automobile plants.^[75] Gas managers were quite correct to view the Laurent affair and the March strike in Clichy as part of a generalized movement against fore-men.^[76] The engineers clearly saw that their right to dominate skilled in-dustrial workers would have to be defended with force.

The challenge to managerial authority was especially strong in Paris, but it was not confined to the capital. The stokers' offensive coincided with a decisive takeoff in union membership throughout France as a whole. The unions they entered were increasingly radical; revolutionaries were pushing aside the proponents of cooperation with a reform-minded govern-ment.^[77] What that radicalization meant on a local level is well illustrated by the revolutionary tide in Limoges. John Merriman has found that between 1895 and 1905 "economic, social, and political conflict seemed to merge with everyday life" in that provincial city.^[78] The tumultuous decade contained not only angry confrontations between workers and the bourgeois state but also a series of violent strikes led by skilled porcelain workers. "The most explosive strike issue," Merriman notes, "was that of industrial discipline, particularly the role of the foreman, who emerged as the most visible and vexing representation of the boss."^[79] Part of the same general confrontation were the angry and highly publicized strikes of Le Creusot and Monceau-les-Mines in 1899. As at the PGC, workers in the company towns were resisting the heavy hand of managerial paternalism. The steel workers of Le Creusot insisted on union recognition, respectful treatment from their superiors, and suppression of the employer's private

[75] Lenard R. Berlanstein, The Working People of Paris , 1871-1914 (Baltimore, 1984), pp. 173-188.

[76] AP, V 8 O , no. 160, report of Masse to Gigot, September 5, 1899; report of Maubert to Lévy, July 22, 1899.

[77] Yves Lequin, Les Ouvriers de la région lyonnaise , 2 vols. (Lyon, 1977), 2:304-307. The entry of the Socialist Alexandre Millerand into the government in 1899 set off a struggle between syndicalists eager to cooperate and those who feared entanglement.

[78] John Merriman, The Red City: Limoges and the French Nineteenth Century (New York, 1985), p. 183.

[79] Ibid., p. 219.

police force during their vigorous September strike. The confrontation ended with the creation of a workers' council to advise management on disciplinary measures. Premier René Waldeck-Rousseau, who arbitrated that dispute, was well aware that the job actions at this moment had more than immediate gains at their roots. He sensed that they engaged the wage-earning population emotionally. Daily demonstrations by workers monopolized the streets of Saint-Etienne, Le Havre, Carmaux, Vienne, and Marseille for times in late 1899. Strikes in the Nord, the Loire, the Saône-et-Loire, and the Doubs provoked repeated rallies and fund-raising activities of exceptional magnitude. The manifestations of class solidarity persisted until 1901.^[80] Clearly, the last years of the nineteenth century— known mainly for the dramas of the Dreyfus Affair—also contained a crisis of skilled factory workers, who rebelled against their subordination at the workplace.

To the limited extent historians have discussed the crisis of industrial discipline, they have portrayed it essentially as resistance to proletarianization.^[81] Skilled laborers were presumably protesting their employers' efforts to rationalize manufacturing, tighten discipline, and curtail their mastery of the production process. The crisis, in this view, was a stage in the inevitable evolution of capitalism, with industry progressing to the point of intensive mechanization, international competition, and excessive capacity. Foremen were caught between employers, who tried to assert greater control over production, and workers, who were unaccustomed to tight discipline and unwilling to accept it. As the most accessible symbols of managerial authority, they bore the brunt of workers' resistance.

The labor disturbances at the turn-of-the-century gas plants do not sustain this interpretation and instead advance a very different one. The intensification of work and the interference of supervisors were not closely connected to the stokers' attack on managerial authority. Both developments had long histories prior to 1890. Proletarianization did not impinge on stokers in 1898; they had experienced a deterioration in their work situation long before. Indeed, the most thoroughgoing act of deskilling undertaken by the company, the introduction of the scoop, passed without

[80] Léon de Seilhac, Les Grèves (Paris, 1903), pp. 152-229; Pierre Sorlin, Waldeck-Rousseau (Paris, 1966), pp. 467-469; Reddy, Market Culture , pp. 309-323.

[81] Michelle Perrot, "The Three Ages of Industrial Discipline in Nineteenth-Century France," in Consciousness and Class Experience in Nineteenth-Century Europe , ed. John Merriman (New York, 1980), pp. 149-168; Merriman, Red City , chap. 7; Berlanstein, Working People, chap. 5.

incident. Far from being the result of the rationalization of production, the stokers' militancy was the direct cause of a new round of it, with the precipitous introduction of the stoking machines. In short, the crisis of industrial discipline at the PGC was not the outcome of a particular moment in capitalistic development.^[82]

The crisis, rather, was an emotional manifestation of rising aspirations on the part of the stokers. The arbitrary power of supervisors had always been a fundamental feature of their work experience. They had to put up with foremen who insulted them, discriminated against them, and made crucial decisions about their lives. The stokers, along with many other sorts of wage earners and employees across France, were now willing to take steps to end or reduce their subordination. Vengeance against the supervisors was on their minds. Whether spontaneously or not, they came to accept the goal of fixed bureaucratic rules to which they had consented as the proper solution to their oppression. In 1899 the stokers believed—and their superiors fearedwthat a new form of managerial authority, promising predictability and equity, was within their grasp.

The problem of industrial discipline flared into a general conflagration in turn-of-the-century France because of exceptional political circumstances. It was a moment of hope, restlessness, and tension on the part of republicans. For several years the Republic had sympathetically considered ways to make the industrial order more acceptable to the masses, and the rise of Nationalist agitation before and during the Dreyfus Affair tilted the regime still more to the left. Wage-earners, along with other working people, had long hoped for a further round of republicanization of the social order, extending even to the workplace. Sensitive to the emerging possibilities as the liberal order foundered, wage earners and employees were quick to force their agenda to the fore.^[83] It was certainly not a coincidence that stokers exploded over their subordination to foremen just as the Chamber of Deputies was considering a bill to regulate finings and firings. Indeed, Maurice Claverie saw in the legislation the solution to the antiauthoritarian grievances that were stirring the gas workers,^[84] The

[82] See chapter 4 for a discussion of technological change within the PGC. The crisis of discipline of 1899 was not simply a matter of a sudden revival of the business cycle. If it had been, the stokers' militancy would have surfaced in 1896. See appendix, fig. A3.

[83] Michelle Perrot, Workers on Strike , p. 314, points out that strike waves often coincided with left-leaning governments. She condudes that the strikes "though rarely political in their objective . . . frequently had political origins." The finding is relevant to the gas workers at the turn of the century.

[84] AP, V 8 O , no. 149, Claverie to president of Conseil al'administration, July 20, 1899.

turn of the century was a moment when le peuple came to expect much from their Republic—more than republican politicians were ultimately prepared to deliver.

Hatred of the kind of power that foremen wielded was not simply learned on the shop floor from experience. There is the visceral nature of the workers' emotions to consider. The stokers' outburst expressed a political culture that was profoundly antiauthoritarian and antihierarchical. The skilled workmen contested not just abuses of authority but also the principle of discretionary authority itself. Their categorical rejection of the legitimacy of face-to-face authority differentiated them from labor aristocrats in Britain or from the members of the Freie Gewerkschaften. At the same time, stokers shared their aversions with French white-collar employees, with whom they did not otherwise identify For all their differences, the two groups accepted a binary vision of authority: there could be either equality or arbitrariness. This political culture found expression even in the highest circles of power as the twentieth century dawned. Only in France did prominent members of the governing party, now the Radicals, denounce personal authority as an unacceptable privilege. In this sense the crisis of 1899 was the outcome not only of shop-floor strife but also of a distinctive political culture that had deep roots in popular politics and flourished temporarily under the Radical Republic.

There is an irony in the stokers sharing an ideology of legitimate authority with office clerks and with petit bourgeois politicians while failing to join with their unskilled comrades at the gas plant in united action. An account of protest at the PGC testifies to the palpable weakness of factory labor as an experience that could forge a cohesive working class. However much the stokers and common hands shared on and off the job, they could not find a basis for agreeing on interests, strategies, and goals. The cleavages were largely the result of differences in their situation at the work-place. The stokers' bargaining power allowed them to contest grievances when the laborers were helpless and to formulate goals that the unskilled could not hope to redress even under the best of circumstances. Factory labor produced compartmentalization, not a melding of differences among the exploited.

Political realities after 1890 reinforced the cleavages created by the work situation. General laborers surmounted their suspicion of the state because they had no other options. Powerless before their employer, the only form of collective action open to them was the sort the authorities sanctioned. Hence they consented to patient organization building and sent their reformist leaders to negotiate for them. The skilled workers, by contrast, had the option of using their own proficiencies to extract conces-

sions from the company. Hopes for small improvements in pay or even for pensions did not entice them to abandon direct action. The stokers set their own agenda, and at its core was control over the work experience. Different political options and uneven workplace power operated in tandem to keep the gas workers divided.

Not only did the gaziers fail to unite as a class; they also failed to grow progressively in class consciousness. The convention of taking the linear development of the working class as normative, so common in labor studies, does not receive support from the case of the stokers.^[85] These workers simply did not evolve from a preindustrial stage to a class-conscious one. The stokers of the 1860s, even as peasant-workers, were fully capable of frustrating the designs of their stern superiors. Their status as independent workers made them interested in wages, and they pursued the agenda with sophistication and success. The stokers of the 1890s had nothing to teach the earlier generation in terms of tactics and consciousness. The later generation had not become notably more interested in organization nor more likely to identify their concerns with those of other sorts of laborers. The stokers of the 1890s had a different agenda because they were now dependent workers, and they pursued their interests as best they could. Common hands, for their part, were no less responsive to immediate circumstances when they rallied to their union in the 1890s. Alain Cottereau proposes that historians put aside the preoccupation with de-grees of consciousness and focus on "how the practices of workers proceed from the logics and strategies adapted to the situation they confront."^[86] His situational approach to class formation accurately characterizes social conflict at the Parisian gas plants.

A final irony of 1899 is that it completely misrepresented the stokers' future. The tensions and antipathies expressed during the strike linked stokers to many sorts of working people throughout France. The stokers' action reflected the principal contours of the French labor movement as a whole—its local autonomy its ephemeral commitment to trade unions,

[85] Older and current works in labor history reflect the expectation of linear development, hut the questioning of that convention has recently become intense. In-deed, class analysis itself has been subjected to serious scrutiny, with the assertion that class is an externally imposed category, not a lived experienced. See William Sewell, "Uneven Development, Autonomy of Politics, and the Dockworkers of Nineteenth-Century Marseille," American Historical Review 93 (1988): 604-637; William Reddy, Money and Liberty in Modern Europe (Cambridge, 1987); Gareth Stedman Jones, Languages of Class (Cambridge, 1983); Joan Wallach Scott, "On Language, Gender, and Working-Class History," in Gender and the Politics of History (New York, 1988), pp. 53-67.

[86] Cottereau, "Working-Class Cultures," p. 113.

and its direct action. Yet the strike of 1899 was their last outburst with such features. As the twentieth century began, the stokers would no longer exemplify the labor movement as it currently stood. Instead, they would point to its future by rallying to a bureaucratized union that operated within the parameters set by the bourgeois state. This change, too, was a legacy of the pursuit of social peace by a Radical Republic at its height of social experimentation.

Advancements in Political Integration

The engineers should have realized that their victory over the union would inevitably be Pyrrhic. The rank and file wanted a union, and since political realities deprived the company of the weapon of intimidation, the workers would have one again. The principal question was what kind it would be. Darène's organization had fallen apart after the August strike, and two options existed after that disaster and after the Nationalists' municipal victory in May 1900. Some of Darène's associates formed a union that sought close ties to the corporation. Whether this was a bona fide com-pany (or "yellow") union remains unclear since its official relations with the PGC were tenuous at best. Nonetheless, the leaders did not hesitate to proclaim, "We are 'yellow' and we don't get 'red' about it."^[87] They argued, as management did, that the fatal strike had been the fault of dangerously rebellious laborers. The company unionists believed that the municipal revolution of 1900 presented a golden opportunity for workers to support the PGC in its negotiations with the city and be rewarded for doing so. They imagined that the company would accept all the workers' legitimate demands and that the engineers would learn to cooperate with the personnel. Only a few general laborers rallied to the company union. The firm's failure to rehire the strikers of 1899 may have discouraged membership. Moreover, the option lacked warm support from management and from the Nationalist councilmen. The disintegration of the Nationalist move-ment in Paris after 1902 undermined its raison d'être entirely.^[88]

Far and away the more successful option was the revival of reformist syndicalism under the leadership of Louis Lajarrigue. This former mechanic for the PGC had joined the strike of 1899 without enthusiasm but out of a sense of duty. The yellows subsequently attacked him as a radical, but he was nothing of the sort. In fact, he became a worthy successor to

[87] Journal du gaz , no. 186 (November 5, 1901): 2. On the yellow unions, see Zeev Sternheil, La Droite révolutionnaire: Les Origines françaises du Fascisme (Paris, 1978), chap. 4.

[88] Préffecture, B/a 1424, report of April 12, 1903.

Brard and a supportive colleague to Claverie, of the clerks' union. Lajarrigue's proposed solution to the gas question was dose to Claverie's call to give "the gas works to the gas workers." He accepted much of the class collaboration Claverie envisioned without duplicating Darène's naive trust in the engineers' goodwill.^[89] Lajarrigue would never attempt to initiate deals with the company that could alienate the councilmen. The former mechanic took over Brard's seat on the municipal council when the latter died (in 1903) and continued his pragmatic defense of organized workers' interests. Lajarrigue was no closer to the Socialist aldermen than Brard had been, but his syndicalism scrupulously respected political realities.^[90]

The new union secretary's success derived from his judicious assess-ment of the political situation. He foresaw that Parisian councilmen would bestow on gas workers benefits that were well beyond realistic expectations for wage earners in other industries. With the strike of 1899 having served to reinforce public sympathy for the PGC's personnel, gaziers would receive decent pay, civil-servant status, job tenure, bureaucratic rules regulating work conditions, and retirement benefits. These advantages would be guaranteed in the charter of whatever firm carried on gas production, and the engineers would have to accept them as final. Lajarrigue correctly envisioned that the gas company would become a quasi-public service that used some profits to support a privileged personnel. By the time the PGC's old charter had expired, Lajarrigue's expectations had been entirely realized.^[91]

Not surprisingly, common laborers rallied to a union that operated within that framework, as they had before 1900. This time, however, so did most of the stokers. Perhaps the PGC's intransigence on rehiring had eliminated or intimidated the most militant of them, but the positive reasons for accepting the path of moderation were surely dominant. The striking stokers of 1899 had belatedly achieved their goals by other means. The ones who labored in the distillation rooms in 1906 no longer had to face supervisors who held arbitrary power over them. Wage earners at the PGC had fixed careers, clear and uniform rules governing the work experience, and a review board to resolve disputes when irregularities arose. The gaziers had gained far more security, better pay, and more control over their work conditions than impoverished immigrants to the capital had

[89] AP, V 8 O , no. 1632, "Personnel, 1906-1920"; no. 150, "Rapport sur la manifestation en faveur de la paix, accompagnè des renseignements sur l'industrie du gaz à Londres."

[90] See Conseil municipal , July 3, 1903, for votes on which Lajarrigue and the Socialist group of aldermen parted company

[91] AP, V 8 O , nos. 162-163.

any right to expect. Even the stokers seemed to recognize their good for-tune and now protected it within the context of a bureaucratic and reformist union.

The impulse for direct action and the anarchic rage characteristic of 1899 had not disappeared entirely among the stokers. A small group re-mained outside Lajarrigue's organization. Some established contacts with the revolutionary wing of the Confédération générale du travail and tried to win their comrades to class conformation. Their militancy was not simply a legacy of the recent past; it was also the response to a new problem that younger stokers confronted—technological displacement. The PGC and its successor firm speeded the introduction of stoking machines in the decade prior to World War I. Even in the face of the new threat, most stokers no longer supported taking matters into their own hands and spiting a regime that offered valuable benefits. The militants proved to be no more than a minor annoyance to the reformists.^[92]

Lajarrigue held to an ambitious and idealistic vision for his union that was not altogether compatible with the consummate pragmatism of his policies. He wanted Parisian gas workers to help gaziers all over France obtain similar benefits. To this end he participated in the formation of, and presided over, a national federation.^[93] Furthermore, he strove to make his privileged corporation into a model and an inspiration for the syndicalist world. He set about using the growing wealth of the gas union to build workers' institutions. Under his direction the union established a vacation home by the sea and a retirement colony for aged gaziers. During Aristide Briand's ministry Lajarrigue read the premier's Socialist tracts and claimed inspiration from them when he proposed that workers serve on the board of directors of the new gas company.^[94] Of course, Lajarrigue was never able to explain how other sorts of workers could obtain the advantages his followers enjoyed while operating within legal and pacific channels. There were also clear limits to his idealistic endeavors. He never asked members to make sacrifices to the larger working-class movement,

[92] Les Travailleurs du gaz: Organe officiel de la chambre syndicale des travailleurs de la Compagnie du gaz, no. 147 (April-May 1906): 1; no. 153, (September 5, 1906): 1.

[93] Actually, the impetus for a national federation appears to have come from the provinces following a protracted strike among gas workers in Bordeaux, but Lajarrigue was happy to lead the movement. See Journal du gaz, no. 125 (February 20, 1898): 1.

[94] AP. V 8 O , no. 1257, "Project de maisons ouvrières"; Travailleurs du gaz , no. 253 (April 20, 1910): 1; Préfecture, B/a 1425, report of February 24, 1911. On Briand's project, see M. Dezès, "Participation et démocratie sociale: L'Expérience Briand de 1909," Le Mouvement social, no. 87 (1974): 109-136.

nor did he take steps to come to the aid of oppressed wage earners outside the gas company. Most stokers willingly followed his lead.

Though Lajarrigue hesitated to admit it, his aspirations and strategies virtually precluded the use of the strike as an instrument of class action. He intended to rely on reasonable demands, impressive displays of solidarity, and, above all, on political pressure. Of course, he had to be careful to distance his union from any public demonstration of "revolutionary" in-tent. Thus, he kept his followers out of the massive May Day strike move-ment of 1906, the denouement of the crisis of industrial discipline.^[95] When the electrical workers struck the next year, Emile Pataud, the head of their union, pleaded with gaziers to support the action, but without success. Relaying orders from his vacation house in Brittany Lajarrigue made sure that gas workers remained aloof.^[96] He even conspired in the exploitation of other wage earners to maintain the advantages of his privileged following. The tenure system under which the gas personnel operated soon became bloated with laborers who had the requisite one year of seniority. Lajarrigue protected his membership by arranging with the Parisian Gas Society (the successor to the PGC) to penalize newly hired la-borers. They would be fired on their 364th day of work so that the pool of workers with "permanent" status would shrink. Some laborers were to be fired at once to hasten the process.^[97] There seemed to be little discontent within the tenured rank and file over these pragmatic—even egocentric— decisions.

As a national figure who claimed to represent the future of syndicalism, the leader of the gas union came to have prominent enemies. Revolutionary syndicalists like Victor Griffuelhes saw in Lajarrigue all that was wrong with reformism. Lajarrigue became the particular bête noire of Pataud, the "prince of darkness" who had led the electricity strike. This revolutionary believed that the gaziers had become "incapable of any energetic action. . .. All their activity consisted of appeals to the authorities." He concluded that Lajarrigue's respect for legality and gradualism had "mummified [the workers] in attitudes of submission." The only hope Pataud saw for the gaziers resided in a handful of militant stokers.^[98]

[95] Préfecture, B/a 1424, report of April 23, 1906. The police reports (as well as the union newspaper) make it clear that Lajarrigue had to overcome some internal opposition to his decision not to participate in the strike. For a good description of May Day 1906 and its context, see Tilly, Contentious French , pp. 315-321.

[96] Prefecture, B/a 1424, report of October 5, 1907.

[97] AP, V 8 O , no. 1632, Rouland to prefect, June 14, 1910; report of Keck to prefect, lanuary 17, 1910; report of Rouland to prefect, June 30, 1910.

[98] Emile Pataud and Emile Pouget, Comment nous ferons la révolution (Paris, 1909), pp. 37, 42.

Thus, the gas workers' union survived the PGC. It became large, influential, and rich. Admirers of British unionism regularly cited it as an ex-ample of what French workers could achieve. Radical critics saw in it a model of how easily wage earners could succumb to bourgeois egoism if leaders did not pursue the tactics of class confrontation. The stokers had eventually followed the lead of the common hands. So recently the practitioners of unmediated conflict in disregard of the political consequences, the stokers now settled for sending leaders of a formal organization with national affiliations to negotiate for them. They became indirect contenders for power within a capitalistic and parliamentary regime. The stokers may have extracted an exceptional price for such behavior, but their choice was in step with many other skilled industrial workers in France.^[99]

Conclusion

The case of the gaziers demonstrates that French factory proletarians did not readily entrust their fate to a centralized political party Nor were the more qualified industrial workers quick to renounce localized, direct action as their principal means of resolving grievances. Even the common hands looked to indigenous leaders who were resolutely syndicalist, not Socialist. So little respect did the leaders have for Socialists that they cooperated more easily with Radicals and with Nationalists on the municipal council. The widespread commitment to workplace-centered action (whether direct or not) probably owed much to workers' long-standing sense of internal exile within the republic and also to the revolutionary inheritance within French political culture.^[100]

Two contradictory features marked the gas workers and many other French laborers around the turn of the century. On the one hand, some wage earners tried to attack managerial authority, a cause that brought stiff resistance from employers and public charges of outright "rebellion." On the other hand, many workers relaxed their fundamental distrust of the state and committed themselves to large, organized interest groups working within channels established by the state. The synthesis of syndicalism and socialism toward which the leftist parliamentary leader, Jean Jaurès, had been building became increasingly relevant under the circumstances.^[101] The second trend did not simply reflect underlying structural

[99] Merriman, Red City , pp. 245-250.

[100] Tony ludt develops this explanation for the weakness of socialism in "The Labor Movement in the Nineteenth Century," in Marxism and the French Left (Oxford, 1986), pp. 24-114.

[101] Michelle Perrot, "On the Formation of the French Working Class," in Working Class Formation: Nineteenth-Century Patterns in Western Europe and the United States, ed. Ira Katznelson and Aristide Zolberg (Princeton, 1986), pp. 108-110.

shifts from a nineteenth-century artisanal economy to a twentieth-century industrial economy. The gas workers, quintessential nineteenth-century factory proletarians, passed through each phase of the transition from direct action to a nationally organized labor movement.

Both of these turn-of-the-century features had their source in the intensified involvement of the state in pacifying industrial society. Even the rather modest reforms of the Radical republicans had important repercussions. A serious problem for the reformers was the failure of factory work to generate a unified working class and, still less, one that was uniformly attracted by practical gains. Depending on whether laborers had much or little bargaining power at the workplace, they took the favor of the state as a signal either to challenge their employers' mastery or to build bridges to politicians of whatever stripe who promised benefits. To the extent that the public authorities appeared to be encouraging "rebellion" against industrial discipline, they had to draw back from efforts to encourage moderation. This paradox plagued the Radical Republic.^[102]

By no means was the insurrectionary impulse put aside once and for all, even though political integration achieved decisive gains around the turn of the century The events of the 1890s at the Parisian gas plants illustrated the waves of action and disengagement that characterized the protest of industrial workers.^[103] Later, the shocks of World War I and of the world depression would spark new bouts of unrest. The gas workers' experience showed that only exceptional benefits from the state could permanently reconcile industrial workers to their subordination at the work-place. Rarely was the liberal order in France so thoroughly shaken that it would generalize such concessions.

[102] For Ciemenceau's experience, see David Watson, Georges Clemenceau: A Political Biography (London, 1974), pp. 169-206.

[103] See Noiriel, Ouvriers dans la société franqaise , pp. 263-267, on the cycles of militancy and quiescence among workers.