Capital flows,itinerant laborers,and time: A revision of Thompson’s thesis of time and work discipline

Pre-industrial, agrarian time

Thompson concluded his article suggesting that the emergence of the nexus of work discipline and time consciousness “is a problem which the peoples of the developing world must live through and grow through” (1967, 95). This comment essentializes the developing world and pre-industrial relations of time and work. It also suggests an inevitable transformation that societies must “grow through.” Since Thompson’s argument has been applied without sufficient reflection to conditions in many societies outside of Europe, his approach to be examined closely to understand the limits of his argument as well as to grasp the specific conditions that gave rise to the transition he documented for England.

Thompson’s portrayal of pre-industrial relations of time and work relied on an alchemy of ethnographic and historical sources: Evans-Pritchard’s The Nuer (1940), Hallowell’s discussion of the Salteaux (1937), and Bourdieu’s discussion of the Kabyle (1963). Thompson also referred to second-hand reports, such as Nilsson’s Primitive Time-Reckoning (1920) and Sorokin and Merton’s essay “Social Time” (1937). Thompson suggested that the time consciousness vastly different pre-industrial societies can be treated as equivalent to pre-industrial England. He assumes that all pre-industrial time reckoning shares the same features, namely, that “among primitive peoples the measurement of time is commonly related to familiar processes in the cycle of work or of domestic chores” (Thompson 1967, 58). These ethnographic accounts actually indicate far greater diversity of temporal ideas than Thompson admitted in his argument.

In his study of the Nuer, a society located primarily in southern Sudan, Evans-Pritchard’s discussion of time consciousness began with an emphasis on seasonal rhythms, but in his “Time and Space” chapter of The Nuer (1940) he moved to an argument that the rhythms and cycles of Nuer life are tied to social structure rather than merely a reflection of meeting needs posed by the natural environment—“In a sense, all time is structural” (1940: 104). Moreover, the close relationship between timekeeping and what Evans-Pritchard called the “cattle-clock” (1940, 101) is tied to the close identification of Nuer men to the cattle in their care rather than just the functional aspects of caring for cows. In fact, then, Evans-Pritchard offers a multi-layered portrayal that weaves together temporalities of seasonal production, social rhythms, and the reproduction of social structural arrangements. To reduce this portrayal to the mere cycle of work does it disservice since the point of Evans-Pritchard’s shift from the analytic frame of “oecological time” to “social time” is that the cycle of work is not sufficient for understanding Nuer time reckoning.

Hallowell discussed the Salteaux, a First Nations society within what is now Canada. In his study, Hallowell noted that Salteaux were familiar with clocks, but that clocks were not relevant to daily activity because daily activity was tied to the meaningful moments of the day, not abstract duration (1937, 656). Hallowell’s emphasis on meaningful moments of the day was documenting temporal logic not the nexus of time consciousness and work. In fact, he did not mention tasks, chores, or labor in his discussion. Finally, Hallowell’s conclusion that “temporal frames of reference vary profoundly from society to society” (1937, 647) does not support Thompson’s equating of Salteaux with Nuer and pre-industrial English time consciousness.

Bourdieu’s study of Kabyle seems to conform to Thompson’s view of essentializing agrarian time consciousness, but Bourdieu’s data was likely distorted because it was gathered under conditions of a refugee camp during the Algerian revolution (Bourdieu 1963, 55 n.1; Goodman 2003). Even under these conditions, Bourdieu wrote “the Kabyle peasant is part of nature, immersed within it” (1963, 57), adding that Kabyle were “Free from the concern for schedules,” and that they ignored “the tyranny of the clock.” (1963, 58). Yet, he contradicted his own conclusions. Buried in his text, Bourdieu provided an example of Kabyle factory workers being adept at conceptualizing a relationship between their wages and time worked. He documented that when one factory raised wages by 20%, the workers reduced the time they worked by 20%, but that when wages were doubled, work discipline changed considerably: “As if a threshold had been broken, the workers showed a desire to work, to earn even more, to work overtime, to anticipate the future by thrift” (1963, 71).

Bourdieu attributes this change to the increase in wages, but the workers had a very clear understanding of wages in relationship to time—that is why they reduced their working time by 20% when given a raise of 20%, and then increased their working time when given a raise that made each hour of work worth considerably more than before. This undermines viewing their time consciousness as dictated by “nature.”

Nilsson’s Primitive Time-reckoning (1920) is a dense survey of examples from many different societies, and it, too, contains examples that do not fit Thompson’s representation of pre-industrial timekeeping. Nilsson groups time reckoning techniques into those based on celestial cycles and those based on environmental cues from animals and plants. In doing so, he documents a dizzying number of techniques, only some of which are directly related to work. Some techniques, like the shadow-reckoning associated with Islamic prayers that diffused to Java and Sumatra, are not tied to cycles of work at all, but to daily prayer cycles (Nilsson 1920, 27). Nilsson’s book defies a facile equating of all non-industrial societies’ ideas of time.

Sorokin and Merton argue that “(t)ime systems are numerous and varied” (1937, 627). Instead of work, they emphasize social coordination. They see the scale of a society, not the mode of production or forms of labor, as an important factor in determining time reckoning practices. They argue that as societies grow, “local time systems of time reckoning are no longer adequate,” and they specifically not that natural cycles cannot serve as a “temporal reference as the area of interaction is enlarged, since these phenomena do not occur simultaneously in different areas” (Sorokin and Merton 1937, 627–628).

The details of the descriptions of pre-industrial time consciousness on which Thompson draws do not portray a simple, pan-cultural, primitive time consciousness. On the contrary, the ethnographic evidence demonstrates no single global agrarian, pre-industrial time consciousness, and in the case of the Kabyle, those who have not fully given in to industrial work discipline demonstrate an ability to directly relate wages and clock time. Despite the diversity of time concepts, one point recurs: whenever time and labor are discussed ethnographically, the labor is closely tied to season and place. In contrast to using material from a vast array of 20^th-century ethnographic sources to understand pre-industrial time consciousness in England, I turn to sources on medieval and early modern England and focus on the issues of seasonality and place.

Pre-industrial English timekeeping

Thompson complemented the materials from anthropology with historical evidence about medieval time consciousness mostly drawn from LeFebvre’s Critique of Everyday Life (2014) and Febvre’s The Problem of Unbelief in the Sixteenth Century (1982). Using these sources, Thompson suggested that there was a fundamental distinction between time that is measured and time than is lived.

Febvre argued that medieval time reckoning was not like the clock. He used phrases like “(p]eople were often still content to approximate it [time] like peasants” (Febvre 1982, 393). Febvre’s key point that time was not measured is only partially true, however—the measurement of duration was used in medieval Europe but found mostly in relationship to predicting tides (Frake 1985; Obrist 2000; Stevens 1995) and astronomy (McCluskey 1990 and 1998).

Thompson’s use of LeFebvre focused on the distinction between cyclic and linear time. LeFebvre suggests that cyclic time was typical of the Middle Ages and was close to nature (2014, 341–342), but Thompson does not explore LeFebvre’s idea of nature as a feature of living from which the worker became alienated (2014, 84, 92–93). LeFebvre’s book is more concerned with how concepts of space and time shifted in order to alienate workers from lived space and natural time than a simple narrative of a shift from cyclical/natural time to linear time. In fact, LeFebvre’s concern resonates with that of Sorokin and Merton’s idea that as the scale of society across space increases, the less useful local indications of time are. If that increased scale spans many latitudes, such local indications include the amount of daylight.

A common distortion in understanding medieval work is the conflation of time measurement and time reckoning. Measurement emphasizes documenting uniform durations; reckoning can include marking important transitions. For instance, dawn is an important time of day in time reckoning, but dawn is not typically measured. The use of transitional moments during the day, such as dawn, solar noon, or the bells rung at the ending of a Mass, were what structured medieval work. These were not units of measure, but they could be used to represent the day and adjust wages. To equate day work and task work, as Thompson and his sources on the Middle Ages do, neglects the possibility of representing time and wages in ways that relied on kairotic time reckoning rather than measurement.

For the medieval period, Thompson described the relationship of work, wage, and time as “irregular” (1967, 71), and suggested that this is the reason why there are no surviving “time budgets” (1967, 71). By emphasizing irregularity, it is difficult to conceptualize a relationship between work, wage, and time without clocks. Yet, if one shifts the focus away from uniform units of measurement to meaningful points of time during the day, the medieval period manifests a rich set of ideas about time that were employed to adjust wages according to seasonal variations in daylight. Contrary to Thompson, then, there was regularity, but it consisted of the same patterns being employed from year to year, rather than from day to day. With knowledge of these seasonal patterns, work could be reckoned with reference to fractions of a day without reference to clocks. Dohrn-van Rossum (1996, 291) observes that in the Middle Ages, labor was sold in relationship to time with the units being half days, weeks, or a year. It is estimated that at least one third of England’s population in the late medieval period engaged in wage labor (Penn and Dyer 1990, 356). While Thompson and later David Landes (1983, 229), viewed the use of wages in relationship to time as a means of punishing tardiness and imposing work discipline the emerged in the Industrial Revolution, evidence of time reckoning in relationship to work and the payment of wages pre-dates the 18^th century (Hailwood 2020; Ogle 2015, 73), but its logic is not based on the use uniform durations measured by clocks, but instead on days being divided.

This logic that differs from clock time can be found even in the sources Thompson cites. For instance, Thompson referred to the farming and account books of Henry Best from 1641. While much of the work Best documented is represented in terms of task, that is not true of all the work he discussed. Best documented seasonal variations—the workday in September began at seven o’clock instead of five (1857, 52). With variations in the length of the workday, there were also variations in wages. For instance, Best paid threshers 6 d. a day from the harvest (August-September) until the wheat and rye were threshed (around November), and between then and Candlemas (in the calendar this is February 2, but England was still using the Julian calendar, so in 1641 this would have been equivalent to our February 12), they were paid 4 d. a day. From Candlemas until Lady Day (March 25, equivalent to our April 4) they were paid 5 d. a day, and from then until the cutting of hay, 6 d. a day (1857, 140). The variation in wages was based on the changes in the duration of daylight. In addition, to determine wages, Best advised employers to note exactly when the break at noon ended (1857, 32). In contrast with sunrise and sunset, noon remains constant throughout the year. In Best’s pre-industrial context, there was a relationship between wage and time worked, but since the unit was the day, and the daylight hours varied in length with the season, wages changed with the season, and noon seemed to be his anchor for determining those wages.

The widespread distribution of cultural knowledge about time and work across different forms of labor that predate the Industrial Revolution suggests that such ideas were common and that factories were not innovators in these ideas. Jonathan Martineau (2015, 93) points out that the spread of clock time pre-dates its hegemony. In addition to Best’s discussion of agricultural labor that Thompson cites, I shall draw upon examples from construction trades, the Statue of Artificers of 1563, and mining.

The first example is from 14th century documents about the construction of York Minster Cathedral. In the Fabric Rolls of York Minster are two discussions of the workday—one from 1352 and one from 1370.

By 1352, there were clocks in York. St. Mary’s Abbey, located just outside of the city walls, recorded the use of a clock in 1324 (Craster and Thornton 1934, 73). Archaeological evidence of an early 14th century clockmaker working near York Minster has been recovered in the form of a seal of Robert the Horologer (Hall 2008; Rogers and Hall 2007). This was found in strata dating to around 1300, a period when King Edward I, who had contracted clockmakers for Westminster, was spending time in York.

Despite evidence of the presence of clocks in 14th century York, the 1352 document made no mention of them. Instead, it discussed a variety of ways of knowing the time that vary with the season and amount of daylight. Throughout the year, work began when there was sufficient light and ended when there was not enough light to work. There were a variety of other time cues that varied with the time of year: the tolling of the bell associated with the Virgin Mary, the concept of a “space of time,” the vicars leaving their dining hall, the vespers bell, and the curfew bell (York Minster 1859:172; see also Birth 2012, 55–66). Some of these time indications, like “space of time,” are presently obscure. A space of time refers to the relationship of duration to the time it takes to walk a specified distance. Indeed, one of the less frequently used definitions of the word “mile” is the time it takes to walk a mile—such is an example of a space of time. Some of these indications of time were more landmarks of passing from one period of the day to another than they are markers of uniform units of time, such as bells struck at the end of Masses. Even so, the 1352 document concludes with a discussion of how wages were docked for those who arrived late, quit work early, or took breaks that were too long (1859, 172)—a clear indication of a conscious linking of time, labor, and wages.

The 1370 discussion of the workday in the Fabric Rolls of York Minster has many similarities with the 1352 discussion, but this document also mentioned the use of a clock to determine midday, but not to measure duration. One might wonder why the clock is only used for this purpose, but if one considers that the one time during daylight hours that clock time agrees with the seasonally variable canonical/temporal hours is solar noon, then the reference to the clock at midday becomes clearer. It is also strikingly parallel to Best’s advice of almost 300 years later to note solar noon to determine wages. Canonical/temporal hours divided the duration of daylight into 12 equal segments, but because of differences in the duration of daylight with the season, these hours varied in length based on the time of year, but throughout the year, noon was when the Sun was at its zenith. The clock at noon thus served as an anchor for both clock hours and canonical/temporal hours and would have been particularly useful on cloudy days.

These two documents from York Minster are similar in two critical respects. First, they both discuss seasonal variation in the workday. This is consistent with practices throughout England, and consistent with winter workdays receiving a day wage less than that of summer workdays that was also found in Best’s account. This relationship of wage to length of workday indicates that the cognitive connection of time and wage existed long before the Industrial Revolution. The second is the emphasis on wages being subject to reduction if laborers are not working when they should. Again, this is evidence of a link between wages and time.

The Statue of Artificers of 1563 discussed the relationship of wages, work, and time. This act set wage caps for a variety of occupations. Section IX refers to “All artificers and labourers being hired for wages by the day or week” and goes on to discuss how they must report to work before five in the morning between March and September, and work until “betwixt 7 and 8 of the clock at night.” Between September and March, workers had to work from “the spring of the day in the morning until the night of the same day.” This section concluded stating that a worker “forfeit(s) one penny for every hour’s absence to be deducted out of his wages” (Bland et al., 1919, 327). Not only does this document match the idea of the seasonal variation of labor found in the Fabric Rolls of York Minster, but also the practice of deducting wages for time not worked.

Lastly, there is even evidence of shift-work represented in terms of clock hours. In his 1556 treatise on the mining of metals, Georg Agricola described work in terms of clock hours, not seasonally variable temporal hours:

Diei & noctis horae quatour & viginti, divisae sunt in tres operas, quaeque autem opera est septem horarum. Tres vero reliquae horae, inter operas interjectae & tanquam mediae sunt, quibus operarii accedunt ad fodinas, aut ab eis discedunt [Day and night are twenty four hours: they are divided into three shifts, each shift is seven hours. The three remaining hours are placed between shifts so that the miners can come and go]. (Agricola 1556, 69, author's own translation)

This shows that were seasonal cycles of production were not relevant, uniform units of time were adopted to represent work. In effect, that in the pre-industrial period, there were multiple ways of relating time and wages.

Thompson suggested for the pre-clock period that “the computation [of wages] is difficult, and dependent upon many variables. Clearly, a straightforward time-measurement was more convenient.” (1967, 61). Yet the evidence suggests that there was no trouble reckoning time in terms of day work, that clock signals were used in at least some contexts dating back to the 14^th century, and that different logics were used for labor by daylight versus labor in mines. It is possible to reckon time without relying on a clock to measure it. Viewing pre-clock periods through the lens of clock-based timekeeping makes it difficult to understand how seasonally variable temporal hours could be used to determine wages, because such reckoning is based on a temporal-cultural logic completely different from the clock.

Work was commonly related to the temporal hours that varied in length depending on the seasonal changes in the duration of daylight. In this system, the primary unit of work is the day, and wages vary based on the length of daylight. The documentary evidence emphasizes a process of deducting from wages for time not worked, not a logic of paying only for the time worked. While these two logics produce the same results, they reach those results in different ways. In a daywork system, the workday extends from sunrise to sunset with the duration of daylight varying with the season. Wages are set by the season, and if a worker misses, say, two seasonally variable hours out of the 12 hours of daylight, then the wage is reduced by 1/6^th, which is equivalent to two of the temporal hours for that season of the year. For instance, in the rebuilding of Knaresbrough Castle, John Roundell and Robert Hill were instructed to “pay each Labourer his Wages proportionably, according to the hours he shall labour at the said Work; and deduct out of his said Wages for so many hours as hee shall neglect” (Lambert 1649, 78 emphasis added). In industrial wage labor, work is compensated by adding up the hours worked not the subtraction from a day wage of the hours not worked.

The contrast between the construction of York Minster and Agricola’s discussion of mining is not just between two different occupations at two different times. It is also a contrast between work that was subject to seasonal variations in daylight versus work that was not. Whenever the workday is defined by the period of daylight, it will also vary with latitude and season. With their representation of uniform units of time, clocks become useful tools for measuring labor when labor is divorced from place and season, but clocks are not sufficient for this to take place. This raises another contrast between mining and construction (and agriculture), mining produces at much the same rate throughout the year, and in the case of metals, there is demand for the product throughout the year. The conditions of year-round production foster the conditions for year-round uniform labor rhythms, but as Marx points out in volume 2 of Capital (1993) such production requires capital to sustain it.

Capital circulation and wage payments

Industrial production does not have the same level of seasonal variability that agricultural production has. With the emergence of shifts in factories, the length of the workday became uniform throughout the year, which, as Dohrn-van Rossum has noted (1996, 289), made new demands on a workforce used to seasonally variable labor. While seasonal variation in wages was tied to seasonal variations in the duration of daylight, in an agricultural economy, such wage variations also reflected seasonal variations in the circulation of capital. A difference between agriculture and industrialism is the ability of industrialism to generate a stable supply of commodities that could convert into stable capital and consequently wage flows provided there was a year-round demand for the commodities.

This creates tension between seasonal variation in production and an emerging financial and factory system that tended to homogenize the workday throughout the year. Thompson’s focus on factories, particularly that of Josiah Wedgwood, sheds light on labor, but it can also shed light on the relationship of manufacturing to the global circulation of commodities and capital. In particular, it reveals the role of finance and ideas of time in the transformation of labor from seasonally variable patterns to year-round production. Wedgwood’s factory was a particularly important example that combined changes in systems of capital circulation, changes and variations in clock technology, and changes in the relationship of time and labor.

Wedgwood was an innovator in accounting and industrial management (McKendrick 1961, 1970). Wedgwood’s factory discipline included fixed hours of work (McKendrick 1961, 38). Related to these fixed hours, the factory used a bell to signal the start of the workday and employed a primitive clocking-in system (McKendrick 1961, 41). Wedgwood’s wage accounts marked when workers arrived, the duration of the meal break, and when workers departed work (McKendrick 1961, 42). These factors give a false impression that Wedgwood was using clock time to represent labor in a fully modern sense. Such was not the case, however.

The clock used by Etruria, Wedgwood’s factory complex in Stoke-on-Trent in the UK, was not a conventional one, but a variety of clock known as an engine or mill clock. Such clocks had two dials. One dial was driven solely by the pendulum and measured mean time. The other dial was linked to an engine or mill and would only measure time when the engine or mill was turning (Rooney 2021, 163–164). Referring to an 1834 factory inspection report, Frederick Engels describes how these clocks worked. He recounted how in one factory there were two clocks:

an ordinary one and a machine clock, which registered the revolutions of the main shaft; if the machinery went slowly working-hours were measured by the machine clock until the number of revolutions due in the twelve hours was reached; if work went well, so that the number was reached before the usual working-hours were ended, the operatives were forced to toil on to the end of the twelfth hour. (Engels 1998, 272–273)

In 1815, Samuel Ogden described such clocks as “annexed” to the factory steam engine so that on these clocks “the time of which, for the right speed of the engine, is calculated to common time” (1815, 29). This allowed the employer to take into account the rate of work to determine the length of the workday. The workday was not set by an abstract duration that remained the same regardless of the pace of production, but a clock time that slowed and lengthened the workday when production slowed. In fact, time-books in mills kept track of the time the mill was working versus the time the workers were at work (see HM Factories Inspectorate, 1842, 49–50).

In the case of the mill clock at Etruria, there were two dials, with the dial tied to the mill’s machinery used by Wedgwood’s managers to measure the duration of work. It was placed in the engine house in Wedgwood’s Etruria complex and its second dial was inscribed “engine time.” The engine-driven dial is described as stopping when the engine stopped (Barnard 1920, 22–23). Etruria’s clock was made by John Whitehurst, whose company would make such clocks for other mills and factories into the 19^th century (Hughes and Craven 1998, 206; Craven 2015, 230).

Thompson mentions factories that kept time differently from a conventional clock by stopping the pendulum, but engine clocks, such as the one that Wedgwood used, did this mechanically. An editorial to the Leeds Mercury compared the practice Thompson mentions to engine clocks: the manipulation of the duration of labor was achieved

by means of a clock with two faces; one of them in the counting house, the hands of which (connected of course with those of the mill clock).” (Anonymous, 1969, 165)

In 1841, when the Select Committee on the Act for the Regulation of Mills and Factories looked into such clocks, they learned that these clocks were an impediment to the enforcement of existing labor laws. In one transcription of testimony

A superintendent is asked, ‘Do you find much difficulty in consequence of the hour of the mill clock not agreeing with the standard time in the county?—Yes: it throws a great impediment in the way of detecting offenses.’ (1841, 28)

Using such technology, Wedgwood’s factory measured the actual time workers labored, not a truly abstract labor time. There was not a single hegemonic clock time, but different sorts of clocks that measured time and labor in different ways. In a way, such clocks mark a transition from the sort of labor Thompson imagined as task-oriented and labor represented in terms of homogeneous, uniform clock time.

In another innovation, Wedgwood maintained large stocks of his product ready for sale throughout the year and fostered credit for the wholesale and retail trades (Weatherill 1982, 249–251). In addition, rather than engaging in direct sales to retailers spread throughout the world, Wedgwood relied on middlemen purchasing from his warehouses (Weatherill 1986). Jacob Price broadly describes how this system worked: the warehouse acted as a wholesaler who sold to merchants on credit (1980, 141). Wedgwood not only produced pottery but acted as what Price calls a “capital mobilizer” fostering trade (1980, 142).

Wedgwood’s product was marketed as “Queen’s Ware” although archaeologists typically refer to it today as “creamware.” This commodity’s distribution spanned many different locations, climates, and rhythms of seasonal production. In 1770, Arthur Young noted that the rapid expansion of Wedgwood’s factory related to “large quantities” of creamware being “exported to Germany, Ireland, Holland, Russia, Spain, the East Indies, and much to America” (1770, 306 emphasis in original).

The distribution of Wedgwood’s products indicates the emergence of a multi-regional capitalist system that included much more than dishware. In A Guide to the Artifacts of Colonial America, Noël Hume describes Wedgwood’s creamware as commonly found throughout the Americas (1970, 125); George L. Miller describes creamware as the dominant ceramic in the late 18^th-century world market (1984, 3); Neil Ewins indicates that 1.2 million ceramic vessels were exported to North America in 1770 alone (1997, 6).

Late 18^th-century accounts corroborate the data provided by archaeologists. For instance, the French traveler Barthélemy Faujas de Saint-Fond wrote of creamware

in traveling from Paris to Petersburg, from Amsterdam to the furthest part of Sweden, and from Dunkirk to the extremity of the south of France, one is served at every inn with English ware. Spain, Portugal, and Italy are supplied, and vessels are loaded with it for the East and West Indies and the continent of America. (Faujas de Saint-Fond, 1799, 97–98)

Wedgwood’s product spanned class differences and even divides between enslaved and free. The British pro-slavery politician Bryan Edwards wrote in his The History, Civil and Commercial, of the British Colonies in the West Indies that tradesmen and domestic slaves in their homes “display a shelf or two of plates and dishes of queen’s or Staffordshire ware.” (1793: 127).

In his 1770 book, Young also discussed the wages paid at the potteries, but there is a notable difference from the pre-industrial discussions of wages: there is no reference to seasonal variations (Young, 1770, 308–309). While Wedgwood’s creamware spanned consumers’ different rhythms of production, the wages associated with the production of creamware did not vary with the season.

The distribution of Wedgwood’s products raises the issue of the origin of the capital that sustained Etruria’s year-round production with its year-round homogenization of labor. In the 18^th century British economy, the sugar and slave trades were major generators of capital that funded industrial production in England (see Mintz 1986). Since sugar plantations were a major generator of capital during this period, their financial relationship to Wedgwood’s year-round production illuminates how credit allowed for capital to circulate even as some commodities were tied to seasonal production.

The trade in enslaved Africans, the sugar trade, the North American staples trade, and Wedgwood’s crockery all had different rhythms of production and distribution which were smoothed over by credit. The production of provisions in North America to feed and supply West Indian sugar plantations did not coincide with the production of sugar in the Caribbean, and the production of sugar did not coincide with financial cycles in Great Britain. The time it takes sugar cane to mature varies with the method of cultivation. The practice of ratooning sped up production, but often decreased yields. Planting fresh cane increased yield but slowed production. The judicious planter would stagger planting to ensure ongoing yield. James Grainger (1766) poem The Sugar-cane, explained:

While rolls the Sun from Aries to the Bull,

Following the imagery of the zodiac, Grainger recommended planting from late March through May so that the canes ripen in August to September and can be shipped before the height of hurricane season in order to arrive in Britain by October. These verses are a bit deceptive in that they seem to imply that canes take 6 months to mature, when really 14–18 months is more the norm. The supply of food to these plantations from North America was tied to a 12 month cycle. The movement of shipping of supplies for the plantations, enslaved Africans, and sugar was restricted during hurricane season from July to September (Mulcahy 2004; Pares 1956, 1960). Wedgwood’s production of creamware knew no such seasonal constraints. With sugar cane requiring such long growing periods but the need for supplies was year-round, the Jamaican sugar planter Edward Long noted that credit was necessary to allow enough money to circulate to permit “the business of planting” (Long 1774, 530).

The rhythms of sugar production and shipping affected the rhythm of commercial transactions in the West Indies (SD Smith 2009, 620), including the slave trade, with the highest prices for slaves occurring right before and during the harvest season (Burnard and Morgan 2001, 215). The period of Wedgwood’s expansion of production in the late 18^th century was a time when a variety of credit mechanisms were developed to deal with the hazards of transport and timing (Pearson and Richardson 2008), but in this context Wedgwood offered a commodity that knew no seasonal restrictions at a moment when credit overcame seasonal production cycles. Thus, Wedgwood’s creamware circulated through a variety of places with vastly different agricultural and business cycles—a circulation enabled by credit that could be issued before the sale of the harvest.

The preference for using credit over specie was also driven by several other factors. One factor was that due to wear, coins would weigh less than their face value indicated. European nations also limited the circulation of specie in the colonies. A period accountant known for books on interest and credit, John Hewitt, observed in 1755 that “a Remittance of Money may be more speedily, more conveniently, and more safely made in Bills of Exchange, than in Specie” (1755, 3).

Therefore, the move to uniform units of time that did not vary with day length that Thompson documents for labor was paralleled by the move to a more uniform movement of capital than was previously permitted by agricultural production and small-scale finance. Wedgwood’s success at both year-round production and taking advantage of systems of finance demonstrated a capability of converting the seasonal generation of capital in agriculture, such as sugar, into a year-round capital flow by means of the year-round production and sale of pottery. The year-round homogenization of labor in Wedgwood’s factory was the result of year-round production to meet year-round demand enabled by systems of credit.

Between his distinctive clock, his uniform level of production, the distribution of his product, and his use of credit to encourage that distribution, Wedgwood can be viewed as a model for how there was a profound shift from seasonality in production to year-round uniformity. Year-round production was driven by year-round demand enabled by credit. Wedgwood had developed a product for which there was global demand transcending seasonal agricultural cycles on multiple continents. While this explains the uniformity of labor time and work discipline at Etruria, it still leaves unanswered why workers would accept any clock as a means of measuring labor.

Thompson’s working class

Thompson documents workers’ acceptance of clock time (1967, 85), but his explanation echoes Max Weber’s The Protestant Ethic and the Spirit of Capitalism (1958). Thompson attributes the adoption of work discipline tied to the clock as part of “the evolution of the Puritan ethic” (1967, 87). There is another explanation that is directly tied to the transition from seasonally variable day work to uniform hourly wage work.

In 1783, Wedgwood published a pamphlet: An Address to the Workmen in the Pottery, on the Subject of Entering into the Service of Foreign Manufacturers (1783). His plea that workers stay employed with British manufacturers rather than being lured to work for non-English factories existed because of the mobility of his workers—their skill was not tied to a particular locality. This suggests a change from the days of the 1563 Statue of Artificers, which sought to limit the movement of workers by binding them to their employers for a year at a time (Minchinton 1972, 11, 13; Tawney 1972, 43).

In his The Making of the English Working Class, Thompson documented how the working class formed as the labor requirements for agriculture changed. He saw the shifting of public, common-use lands to private hands by enclosure as a key element in the formation of the working class: “Enclosure, indeed, was the culmination of a long secular process by which men’s customary relations to the agrarian means of production were undermined” (1980, 239).

Building on the literature on industrialization, Thompson observed, “It is notorious that in the early stages of industrialization, the growing towns attracted uprooted and migrant labour of all types” (1980, 275). Thompson’s emphasis on enclosure generating a displacement of workers differed from J. D. Chamber’s thesis that enclosure’s effect on shifting workers away from agriculture was exaggerated. Chambers claimed the population of agricultural laborers increased during this period, and that enclosure attracted labor “from far and wide” to develop previously undeveloped land (1953, 332–333). Chambers noted, “Even the proletarianized labourers continued to remain on the soil in increasing numbers in most areas until the 1830’s and in some parts to the 1840’s” (1953, 338). Still, while there was disagreement between Thompson and Chambers on the effect of enclosure on labor migration to towns and cities, there was agreement on the process of a move from subsistence labor to wage labor.

A narrow focus on the importance of enclosure, misses a larger process that was going on relevant to timekeeping. This is not a shift from agriculture to industry; it is a shift from a labor force tied to a particular locality and its cycles to an itinerant, wage-earning labor force. The year-round availability of capital increased a year-round demand for labor and encouraged workers to take advantage of the year-round wage opportunities rather than be tied to a specific locale with its lower winter wages. This is not a matter of either agriculture or industry, but workers seeking employment in both as wage opportunities presented themselves, much as could be seen in late 20^thcentury Trinidad in the relationship of factories and farming (Birth 1999). Wolfgram Fischer suggests an additional factor—those seeking to develop an inexpensive workforce “planted the new industry where a rural proletariat already existed, or where smallholdings which did not support a family throughout the year prevailed” (1973, 161). By means of this, agricultural and industrial work opportunities co-existed. In addition, unlike in agriculture, in factories there were no lower winter-time wages.

Wages that did not vary with the season seems to have attracted many workers, but around the time of the emergence of factories the workforce became more mobile than before. Data indicate that during the late medieval period workers shifted occupations frequently, but that with the exception of a few trades, like construction, workers did not travel far (Penn and Dyer 1990, 363). With the growing numbers of landless wage workers, and improvements in transportation the distances workers traveled increased. Wedgwood’s factory is an example. It was located in the Midlands and was linked to the interior and the coast by a canal system that Wedgwood worked hard to have constructed (Bentley et al. 1765; Davies 1980).

As long as wages were tied to the concept of a day’s work, wages varied along a north-south axis (see Clark 2001, 485)—they were higher where the duration of daylight was greater. It is likely that it was workers moving to different locations with different durations of daylight that prompted Adam Smith’s observation that “at all times and places” equal durations of work need to be valued the same (2000, 54). A workforce consisting of workers from different places with different durations of daylight made the use of day wages impractical. As long as wage laborers were not moving about in large numbers or great distances, such differences meant little. But, with the massive displacement of labor that Thompson documents, such variation complicated employers and laborers finding agreed upon standards for wages for day work.

Thompson suggested that as the new clock-based time system was “imposed,” that “the workers begin to fight, not against time, but about it” (1967, 85). Yet the evidence he offers contradicts the claim of time consciousness being imposed—it documents employers confiscating workers’ watches and demanding shifts longer than the clock indicated (1967, 86). This evidence makes more sense with a population of workers who think about the relationship of wages and time but dispute the factory’s clock. So when Thompson demonstrated that the workers eventually adopted clock time as a means of representing labor, it was not because of some Puritan ethic, as he argued, but they were doing so in their own interest to have their wages represented in terms they could accept.

The removal of the length of day from the representation of wages also meant that wages could be conceptualized in terms of the duration worked rather than the duration of daylight minus the duration not worked, which was the approach used with day wages. This is where the clock played a critical role. The clock allowed the use of hours of uniform length to represent labor time rather than the old reliance on days of variable length. With itinerant laborers coming from far-flung latitudes, representing labor in terms of clock hours would negate any disputes about a day’s wage—it was no longer a wage determined by a day’s work, which might vary considerably from Newcastle to London, but a wage determined by an hour’s work. Wedgwood’s engine clock sought to document the actual time worked. It can be viewed as a transitional clock in a labor process that had not yet fully abstracted what Marx called “socially average labor time” (2000, 59–60)—the way in which labor time came to homogenize the diversity of work and workers through the imagination of the “average” worker working at the “average” rate of production.

Thus, to Thompson’s link between the clock and labor must be added the effects of year-round capital circulation and landless itinerant labor. It was these factors that raised the value of uniformity in timekeeping thereby cultivating the application of clock time to represent labor. Initially, the uniformity was not abstract clock time, but a clock time directly tied to production activities. This was the purpose of the engine clock—to measure labor only when the machinery was being used. It was only with labor laws that this practice was ended.

Conclusion

Vanessa Ogle argues that the changes Thompson documents are more about the temporality of capitalist modernity than the internalization of time-discipline (2019, 326).

The processes of production and exchange combined with itinerant labor moved the representation of wages away from local and seasonal day rates to uniform hourly rates. This process was not just about timekeeping as a means of alienating workers from their labor, but as LeFebvre alluded to in his study of the Middle Ages (2014, 84, 92–93), it also alienated labor from seasonal variations and from the location of work. This was an alienation from production, seasonality, and place.

Thompson was correct in emphasizing the importance of the clock as a means of representing labor in the Industrial Revolution, but there are factors that he did not in the transition to capitalism, such as the different time conceptions involved in task work and day work, the relationship of day work to the amount of daylight tied to specific locations, seasonal variations in the annual circulation of capital, and the specific clocks factories were using to represent labor time. What the clock allowed was uniformity across latitudes and seasons in a period where there were increasing patterns of production and consumption that were uniform across space and season. Ironically, what informs this new reading of Thompson’s thesis are aspects of 18^th century capitalism that he was known for emphasizing in other works, such as his Making of the English Working Class (1967). Still, one point that runs through much of Thompson’s work is that the emerging working class acted rationally (Harrison 1986, 135). The adoption of uniform clock time to represent work by those whose labor had become alienated from place and season fits this general tenor of Thompson’s work better than his emphasis on the evolution of a Puritan ethic in an age of secularization.

Thompson’s argument was about a transition from an agrarian economy to an industrial one, but the transition he identified is much more significant. When one takes into account the global web of exchange and finance in relationship to time, work, and the movement of workers, then the clock does not appear to be a tool that facilitated industrialization, but a tool which facilitated globalization. The uniformity of units of labor achieved with clock time that ignored place and seasonality allowed the conceptualization of what Marx called socially average labor time (2000, 59–60). Such uniformity alienated from local rhythms and place are two fundamental features of the global economy (Postone 1993, 193–194; Virilio 2000).

Barbara Adam observes, “(w)ith clock-time the tie with temporality, contextuality and variation has been severed and contingent rhythmicity replaced with a decontextualized, invariable machine-time” (2006, 123). Usually, the emphasis in this narrative is on how the clock creates a consciousness of time measurement, but Adam’s emphasis on decontextualization and invariability is more salient. Ideas of duration in relationship to activity existed before the Industrial Revolution, but they were tied to location and time of year. What the clock achieved was an alienation of workers away from the rhythms of their locale and season thereby allowing the conceptualizing and accounting of relationships of wages and work between any location at any time. It would eventually allow the wage comparison of workers doing the same work in different parts of the world that is important in global capitalism—a comparison engaged in by both workers seeking locations with better pay, and employers seeking locations with lower labor costs.