In late September 1662, Margaret Johnson, the wife of a butcher from Handley, Cheshire, sold drink to some of her neighbours, she ‘having perry at her house’. Thomas Stockton, a 16-year-old servant, drank until he ran out of money, promising to pay Margaret back later. On St Luke’s Day, 18 October, he returned, offering to ‘bring her some apples’ in compensation that night. Margaret accepted, but for ‘fear of her husband’, she had Thomas place the apples outside her house in the gorse-cote, a type of shed, about an hour before midnight. One week later, Margaret rose early on Saturday morning on 25 October about ‘two hours before day’ by her reckoning, or ‘about four of the clock’ according to her husband, Ralph. She set about various chores, driving away the swine ‘which were at some straw’ and looking to the horses. Informing her husband that the horses were missing from their yard, she then went into the village to light a candle ‘about an inch longer than her middle finger’, so that they might ‘cut the meat down and ready it for the market’. Upon her return with the light, Ralph went looking for his horses, being ‘absent about half an hour’. Margaret took advantage of his absence to move her apples from the gorse-cote ‘into a loft above stairs in the said house, in a basket at twice or thrice carrying’. Ralph returned from his search for the horses empty-handed, but ‘it being by this time daylight’, his wife was able to point out their location under a neighbour’s ash tree. Ralph then ‘got his horses up and loaded’, and ‘about sunrising’ the couple started the seven-and-a-half-mile journey to Chester market. While away, constables searched their house for goods stolen from the parish minister the night prior. They found the apples, and though these were unrelated to the crime at hand, Thomas Stockton would later confess to stealing them from his master.Footnote 1
The rich depositions from this quarter sessions theft case reveal much about time-telling conventions and the seasonal, calendrical, weekly, and daily rhythms of work in early modern England. The darkness of early morning did not keep the Johnsons from their chores, but the lack of light and its provision through candles or daybreak clearly shaped their approach to their tasks. Such a busy early morning was probably not their daily norm, but rather due to weekly trends, as the married couple prepared and set out for a Saturday market in Chester. Three separate deponents described how work took place ‘upon St Luke’s day’, a holy day when labour was ostensibly prohibited. Margaret’s implicit involvement in perry and cider production and sale, like the Johnsons’ butchery activities and trip to market, speaks to seasonal cycles of labour: the autumnal harvest and storage of fruit, and commencement of the winter slaughtering season, respectively. Moreover, Margaret’s implied engagement in cider production, apparently hidden from her husband, and the married couples’ division of Saturday-morning labour, demonstrate how temporal patterns of work could differ according to gender.
By drawing on cases such as this the work-task data allows for an exploration of the hourly, daily, monthly, and seasonal experience of work. In doing so, this chapter brings together two overlapping subjects central to the history of work – seasonality, and the working year/day – which have not received equal amounts of historiographical attention. Scholars have long recognised the highly seasonal nature of work in agrarian societies and its variance according to farming type and region.Footnote 2 Yet in other respects, the seasonality of labour has been taken for granted. Relatively few studies have considered its potential variance according to gender, occupation, status, or type of employment. Exceptions have been limited largely to urban contexts, construction, or the agricultural sector, with the seasonality of many other facets of rural labour and economy – like commerce, carework, or transport – left unexplored.Footnote 3 The common theme across this literature is a perception, or assumption, that work was intense during the summer, and lax during the winter – a pattern usually attributed to weather and the availability of daylight.Footnote 4 While the work-task data certainly does not disprove the winter-summer dichotomy, it enables a more nuanced appraisal of winter work and its contribution to the economy than proxy measurements like marriage seasonality or sources like farming and building accounts typically allow.
In contrast to seasonality, the early modern working year, week, and day have been central to various historiographical debates. For economic historians, estimating the number of days and hours worked in a year is crucial to building long-run wage series, and these in turn underpin narratives about changing standards of living, and the development of industrial and ‘industrious’ revolutions. More than just a key variable in calculations, the number of annual work days and hours lies at the heart of such narratives, many of which contend that worktime gradually increased in the early modern period, spurring economic change.Footnote 5 Part of the ‘industrious revolution’ theory of Jan De Vries, for example, postulates that workers gradually gave up holidays and leisure time in the latter half of the seventeenth century to increase their purchasing power for consumer goods.Footnote 6 Likewise, Hans-Joachim Voth’s study of depositional evidence from eighteenth-century London suggests a decline in the observance of St Monday and religious holidays resulted in more annual work hours for industrialisation.Footnote 7
Yet, aside from notable exceptions like Voth’s time-use study, little empirical evidence has been offered to back up these economic grand narratives. Social and cultural historians’ theories of time-use and its regimentation during societal transitions of reformation or modernisation have been built on similarly shaky ground. Most famously, E. P. Thompson argued that the rise of industrial capitalism in the eighteenth century fundamentally altered time-consciousness and discipline within the English workforce, with strict factory regimes based on clock-time supplanting the casual irregularity of traditional work patterns.Footnote 8 Although historians have since challenged and refined Thompson’s argument, the assumption that premodern rural time regimes lacked sophistication, consistency, and clock awareness has lingered. In contrast, the work-task data shows women and men of early modern rural England were well-versed in clock-time and structured their workdays and weeks in strikingly regular ways.Footnote 9 In addition to providing robust empirical evidence for working hours and days that most other studies have lacked, depositional evidence can move the debates forward to consider variability across gender, occupation, employment arrangement, and sector.
4.1 Recording Time
This chapter addresses these issues in the literature, illuminating the common experience of rural ‘worktime’ to a greater extent than has been possible before. It does so in three sections, covering seasonality and the labours of the months; the working week and holidays; and the hours and phases of the working day. Such analysis is possible because, as the opening anecdote illustrates, deponents in the quarter sessions and church courts tended to situate their activities within rich timescapes, while coroners’ inquests usually noted the date and time of accidental deaths. Month or date observations were most commonly linked to festivals – ‘a week after hocktide’ – and calendrical units – ‘on or about Thursday sevennight last past’. Somewhat less frequently they were related to the weather – ‘about the snow time about twelve tide’ – or lifecycle events – ‘the day of the christening of the said Mr Windham his child’. Witnesses might also use work itself as a touchstone, demonstrating labour’s influence upon senses of time. References to cropping activities abound in the observations – ‘haymaking time’, ‘latter end of wheat harvest’ – while animal husbandry likewise left a mark – ‘at sheep shearing time’. Commercial events such as ‘fair time’ and ‘market day’, or contracts of service and employment, also structured the timescape, while the day itself could be delineated in terms of clock-time, natural or fire light, phases of the day, and social events such as ‘before evening prayer’.Footnote 10 As shown in Table 4.1, these observations can be aggregated into work-task subsamples according to different increments of time.
Table 4.1 Long description
The table presents a comparative analysis of labor documentation across different historical record sources. The first column identifies temporal categories, followed by five data columns showing: the percentage of tasks recorded in quarter sessions, church courts, and coroners’ reports, total task counts for each category, and their percentage representation within the complete dataset. The data reveals how task documentation varies across different archival sources and temporal classifications.
For overall sample, the corresponding values are 56.5, 32.2, 11.3, 9650, and 100.
For month, the corresponding values are 62.1, 21.5, 16.4, 6559, and 68.
For weekday, the corresponding values are 63., 12.9, 24.1, 4331, and 44.9.
For calendar date, the corresponding values are 59.7, 9.3, 31.1, 3361, and 34.8.
For phase of day, the corresponding values are 58.2, 16.7, 25.1, 2236, and 23.2.
For hour, the corresponding values are 43.1, 12.3, 44.6, 1256, and 13.
a This count includes ‘harvest time’ data. This relates to tasks for which no specific month was given but were instead described as occurring during ‘harvest time’. These tasks have been assigned specific months based on the distribution of harvest tasks for which there is month data, as explained in Appendix C. Without these additions there are 6,137 tasks with month data.
These incidental references to work, when systematically collected from depositions, hold many advantages over other historical sources of time-use and seasonality. They are closer to realities on the ground than proxy measurements based on marriage patterns or unemployment rates, or aspirational descriptions in labour legislation and literature. And they capture a wider range of tasks and experiences, especially those of women and servants, than financial accounts usually afford. Nevertheless, they do have limitations linked to the proclivities of respective courts and sources. As Table 4.1 shows, some time-use information, such as months and weekdays, was much more readily reported than others, and some sources were more likely to record certain types of information than others. This means that the subsamples of data used for analysis throughout this chapter draw on different proportions of material from each of the respective courts. That said, the proportions of each category of work observed in each subsample do not diverge significantly from the overall dataset, with the exception of very low levels of commerce reported for hours and phases of the day.
More potentially consequential are the ways in which crime, dating and time-telling conventions, and the judicial process might shape results in ways unconnected to actual rhythms of work. For example, certain hours, days, or months could be more prone to crime than others. Yet as explained in Section 1.2, all activities in the work-task dataset have been assigned an ‘information status’ based on how linked or ‘integral’ the task was to the court case at hand. Filtering out integral cases allows us to mitigate the temporality of crime as and when necessary: it decreases the number of late-night tasks in our sample; suggests there was little to no weekly pattern to crime; and points to an intense concentration of theft in November and especially December.Footnote 11 Accounting for time-telling conventions and the judicial process in our results – especially with regard to the monthly data – is a trickier business. To take the first issue, nearly 400 tasks in the dataset were described as happening in ‘harvest time’, rather than a specific month. This obviously creates problems for seasonality analysis based on the monthly distribution of tasks. Because it was customary to date autumnal activities in this manner – indicative of the cultural significance of the harvest – leaving this data out would seriously underrepresent tasks in that season, and especially the harvest fieldwork so central to early modern society.Footnote 12 Thus, we have assigned all undated ‘harvest time’ activities to specific months (largely July–September) based on the monthly distribution of harvest-related and ‘harvest time’ tasks for which we do have precise monthly data.Footnote 13
The seasonality of the judicial process and deposition-taking presents an even more complicated problem. The majority of our month data derives from the quarter sessions, a court which, as the name implies, met seasonally: after Easter, Midsummer/Trinity, Michaelmas, and Christmas/Epiphany. While every effort was made to survey an equal number of session rolls across sample years, in the end depositions were more numerous for some quarterly sessions and particular months than others.Footnote 14 Across the whole year, these deviations were not extreme, but the potential effect on a time-use study becomes more cogent, given that the most overrepresented month, December, had nearly twice as many depositions as the least represented month, February; the equivalent of a random-hour recall survey making twice as many calls in one month as another.Footnote 15 Reasons for this might include discrepancies in session roll survival, a seasonality to legal business, and – as the December depositions suggest – criminogenic seasons.Footnote 16 Monthly weightings have therefore been applied to the quarter sessions tasks to create the effect of a more even distribution of sampled depositions across all months.Footnote 17
Figure 4.1 shows the effect of applying adjustments – ‘harvest time’ additions, monthly weights, excluding integral, the female multiplier – to the monthly distribution of tasks. These adjustments, which have been applied to the seasonality analysis below unless otherwise noted, do not aim to ‘fix’ the data or present an infallible statistical picture, but they do account for various factors that might otherwise obscure or significantly distort seasonal patterns in the observed work tasks. We remain cautious about reading these results as precise indicators of overall work intensity in different months due to the sampling issues outlined above, but it is possible to make some meaningful observations about how the distribution of tasks varied through the seasons. Moreover, the great strength of this data is that it allows for comparisons of time-use across different types of work and groups of people. The rest of the chapter does just that, proceeding inward from the seasonal and monthly experience, through the pattern of the week, down to the phases and hours of the day.
Seasonality of labour: an index of monthly tasks showing data adjustments.
Notes: 100 = monthly average. All Adjustments = Integral excluded, harvest additions, monthly weights, and female multiplier (x2.36). This differs from the standard multiplier as it is designed to give an equal number of male and female tasks with weighted monthly data attached.
Figure 4.1 Long description
The table presents a comparative analysis of labor documentation across different historical record sources. The first column identifies temporal categories, followed by five data columns showing: the percentage of tasks recorded in quarter sessions, church courts, and coroners' reports, total task counts for each category, and their percentage representation within the complete dataset. The data reveals how task documentation varies across different archival sources and temporal classifications.
For overall sample, the corresponding values are 56.5, 32.2, 11.3, 9650, and 100.
For month, the corresponding values are 62.1, 21.5, 16.4, 6559, and 68.
For weekday, the corresponding values are 63., 12.9, 24.1, 4331, and 44.9.
For calendar date, the corresponding values are 59.7, 9.3, 31.1, 3361, and 34.8.
For phase of day, the corresponding values are 58.2, 16.7, 25.1, 2236, and 23.2.
For hour, the corresponding values are 43.1, 12.3, 44.6, 1256, and 13.
4.2 Seasonality of Work
What was the shape of the working year in early modern England and in what ways did it vary? The adjusted data in Figure 4.1 provides the best approximation of the frequency of work tasks across the months and quarters of a typical year. The highest concentration of tasks was observed in the summer quarter (July–September), the season of hay and cereal harvests, while the winter quarter (January–March) saw the lowest levels of activity. These trends fall in line with the oft-remarked discrepancy in labour and employment between winter and summer. The individual months, however, show variation within this general pattern. During the winter half of the year, for example, activity rose to notable peaks, nearing the annual monthly average in December, February, and March, with January exhibiting the only pronounced slump in work levels.Footnote 18
More revealing than overall task numbers are the monthly distributions within particular work categories and subcategories. Figure 4.2 highlights those sectors with a pronounced peak in the summer half of the year, when as we would expect agricultural labour was at its most concentrated. Indeed, fieldwork, gathering food, and farm transport tasks overwhelmingly took place in the summer quarter, with over 70 per cent of tasks in each category recorded during that period as workers mowed and made hay in July, reaped and gleaned corn in August and September, and spent all three months hauling in such harvests to storage. Yet other types of work contributed to the late summer peak in activity as well. A total of 30 per cent of observed craft and construction tasks took place between July and September, with building work, at 56 per cent, and groundworks, at 53 per cent – rather than general craftwork – proving particularly seasonal.Footnote 19
Summer work: indexed monthly distribution of agriculture, construction, and transport tasks.
Notes: 100 = monthly average. Harvest additions and monthly weights applied; Integral excluded; F adjusted (x2.36). This differs from the standard multiplier as it is designed to give an equal number of male and female tasks with weighted monthly data attached.
Figure 4.2 Long description
The vertical axis is labeled index and ranges from 25 to 250 in increments of 25. The horizontal axis presents the months of the year. Three lines are marked for agriculture and land, crafts and construction, and transport. The approximate values are as follows. For agriculture and land region, January, 42, February, 68, March, 62, April, 70, May, 100, June, 120, July, 150, August, 240, September, 176, October, 70, November, 46, and December, 52. For crafts and construction region, January, 68, February, 112, March, 66, April, 78, May, 126, June, 118, July, 140, August, 100, September, 110, October, 108, November, 88, and December, 86. For transport region, January, 76, February, 98, March, 92, April, 82, May, 108, June, 124, July, 102, August, 112, September, 150, October, 82, November, 72, and December, 112. All lines follow an increasing trend with major fluctuations.
Transport activity also reached its peak in the summer. Movement of goods was above average or at its height during the harvest quarter, especially when done by cart or horse, with 42 per cent and 47 per cent of such tasks respectively taking place in this period, as when John Blydes drove a cart ‘laden with wheat to the Stokesby ferry’ in Norfolk in 1594.Footnote 20 At the same time, it clearly depressed labour in other sectors. August’s repertoire of tasks suggests that 52 per cent of people’s time was devoted to agriculture during this prime harvest month, while most other work categories simultaneously fell around or below the average monthly distribution of 8.3 per cent.Footnote 21 Commercial activity in particular was at its lowest in August, while even the busy construction sector concentrated its summer work in July and September rather than August.Footnote 22 The participation of craftsmen in seasonal agricultural labour is explored further at the end of this section, but it is clear that priority was given to the harvest over other types of work, something dictated in the labour laws but rarely demonstrated empirically.Footnote 23
While fieldwork and gathering were concentrated in the summer quarter, other tasks connected to the land – like fishing and hunting, digging marl/earth, and weeding – clustered in the spring. In contrast, animal husbandry represented a more constant and evenly spread tax on time and labour, appearing as one of the top subcategories in every single month of the year. But here as well activity was at its most concentrated in the spring and summer months, particularly May through July, and even more so when we take the harvesting of hay fodder into account. Much of this peak derived from the care and movement of sheep and cattle. The droving of livestock was at its height in the spring quarter, at 33 per cent, as people moved animals, especially sheep, from winter to summer pasturage. Sheep washing and shearing was overwhelmingly concentrated in the months of May and June – 83 per cent of tasks – while cow milking was more evenly spread throughout the summer half of the year, with 75 per cent of tasks in this period and just 25 per cent in winter. Of course, work on the land did not entirely cease as the weather turned cold. Autumn and winter quarters were characterised by ploughing and sowing and otherwise preparing the ground, and foddering of livestock reached its height in the barren months from January to March. Wood husbandry was likewise most heavily focused in the winter quarter, while people fetched wood and faggots for fuel most often during the three-month period from December to February, protecting their households against the cold. But although agricultural work certainly continued during the winter half of the year, it was a season more notably characterised by other sectors, as Figure 4.3 shows.Footnote 24
Winter work: indexed monthly distribution of carework, commerce, and food processing tasks.
Notes: 100 = monthly average. Harvest additions and monthly weights applied; while carework and commerce exclude integral tasks, food processing reflects raw numbers. This is because a large percentage of recorded food processing tasks (mostly butchery) were integral to criminal cases, as deponents slaughtered sheep under suspicion in the winter. Excluding these would render the food processing sample too small for robust analysis. While the raw pattern certainly reflects the seasonality of crime, and overrepresents sheep slaughtering in particular, it is reasonable to infer that it genuinely reflects seasonal trends in butchery intensity, as culprits responded either to their own needs or to economic demand. F adjusted for food processing (x2.58), for carework and commerce (x2.36). These differ from the standard multiplier as they are designed to give equal numbers of male and female tasks with weighted monthly data attached after filtering.
Figure 4.3 Long description
The vertical axis is labeled index and ranges from 40 to 180 in increments of 20. The horizontal axis presents the months from July to June. Three lines are marked for carework, commerce, and food processing. The approximate values are as follows. For the carework region, July, 98, August, 96, September, 94, October, 88, November, 86, December, 94, January, 62, February, 138, March, 176, April, 88, May, 74, and June, 102. For the commerce region, July, 112, August, 72, September, 72, October, 116, November, 110, December, 154, January, 84, February, 78, March, 102, April, 94, May, 98, and June, 94. For the food processing region, July, 52, August, 68, September, 98, October, 130, November, 108, December, 96, January, 136, February, 86, March, 130, April, 96, May, 116, and June, 48. All three lines follow a major fluctuating trend throughout the months.
Food processing in particular had an inverse relationship to summertime agricultural work, as people prepared and stored harvest produce for winter consumption. Threshing and winnowing at 70 per cent of tasks, milling at 67 per cent, and storage and preservation at 58 per cent were all at their height during the coldest half of the year. Butchery was at its busiest in the autumn quarter, at 38 per cent of tasks, when, as we saw from the Johnsons at the beginning of this chapter, October kicked off the winter slaughtering season. Like the harvest, this had knock-on effects in other sectors such as transport. Droving saw one of its peaks in the autumn quarter, as livestock was moved for exchange in meat markets. In November 1674, for example, the butcher William Cubbech bought a heifer in Setchey market, Norfolk, which he then drove homeward and slaughtered for sale at Lynn market in the first week of December.Footnote 25 This vignette highlights how food processing interconnected with commerce, another major wintertime sector.
Figure 4.3 conveys the relative aseasonality of commerce in its monthly distribution, aside from a modest trough in August and a very notable peak in December. Figure 4.4 looks more closely at the largest commerce subcategories, buying and selling, according to the types of goods transacted. This suggests that livestock purchases were only one factor in the commercial peak in December, and in autumn more broadly. Even when accounting for the potential distortions of December crime, the autumn quarter saw the highest proportion of commercial exchanges for most good types, and those transactions were of higher value on average.Footnote 26 These annual patterns of buying and selling derive partly from the seasonality of certain commercial spaces and events, like fairs and markets. Fair activities in the work-task dataset fall neatly in line with the monthly distribution of fairs recorded in Harrison’s Description of England: there were very few in the winter quarter, 9 per cent, with most concentrated in spring, 31 per cent; summer, 33 per cent; and early autumn, 27 per cent.Footnote 27 Contrastingly, marketplace activity, while of course taking place weekly throughout the year, was much more intense during the autumn, 34 per cent, and winter, 32 per cent. Again, December played a leading role in this trend, accounting for 16 per cent of the market tasks, against the 8.3 per cent monthly average.
Seasonality of buying and selling: monthly distribution by good type.
Notes: Harvest additions and monthly weights applied; Integral included: the chart uses raw numbers for more robust analysis, since the shape and pattern of the commercial year remain largely the same even when integral tasks are excluded. F adjusted (x2.58). This differs from the standard multiplier as it is designed to give an equal number of male and female tasks with weighted monthly data attached. ‘Other’ includes books, tools, furniture, housewares, iron, fuel, silverware, and timber.
Figure 4.4 Long description
The vertical axis is labeled number of tasks and ranges from 0 to 250 in increments of 50. The horizontal axis is labeled with months from January to December. Each bar is divided into eight task categories. Food and drink, livestock and horses, grain, textiles and wool, clothing, other, and unclear, shown with different shades and patterns. Food and drink is the category with the highest number of tasks from January to April, July, August, October, November, and December. Grain has the highest count in May and June, while livestock and horses lead in September. December has the highest total number of tasks, and August has the lowest. The categories with the lowest counts in most months are clothing and other.
This commercial dominance of December, when buyers and sellers contracted 13 per cent of their business,Footnote 28 which was over 3 per cent more than the next busiest month, warrants further scrutiny. All major categories of goods were bought or sold at rates above the monthly average, but textiles, livestock, grain, and food were in particularly high demand. As we have already seen, the livestock and grain purchases were linked to the seasonality of food processing, with households preparing for the long winter months ahead. But while some of the food and drink was undoubtedly salted or stored to ward off future dearth, much of it was intended ‘for provision of [the] house in the holy days’, as one Richard Hull of Thornton, Lancashire, put it in 1636.Footnote 29 Hull claimed that his father-in-law had procured and sent ‘turkey … wheat flour puddings and fresh beef’ as gifts, and others were similarly forthright in connecting transactions explicitly to Christmastide celebrations. The servant Winifrid Oliver of St Mary Bourne in Hampshire bought ‘turkeys against Christmas’ in 1630 at her dame’s behest, while butcher Louis Spratly sold flesh from his stall in Southwark ‘upon … twelfth eve’ in 1597.Footnote 30 The Twelve Days of Christmas thus left their mark upon annual work patterns, not only in the sheer amount of food and drink transactions during the two months of December and January – over 29 per cent of the annual total – or the level of marketplace and financial activity but of course in the large proportion of labour devoted to processing and preparing such food for feasting. In terms of butchery, at 28 per cent of tasks, milling at 24 per cent, carrying goods at 24 per cent, and food preparation at 20 per cent, these two months were manic, with averages well above the two-month norm of 16.6 per cent, something witnesses not infrequently ascribed to the festive period. The miller of a water corn mill in Hulme, Cheshire, for example, explained in 1622 that he never slept in his workplace, except from ‘about a fortnight before Christmas because of that time there is much grinding’.Footnote 31
In light of this, it seems likely that December’s high rates of theft, and of depositions, derived in part from the intense social pressures to feast, celebrate, and give appropriately during the holy season, alongside other crime factors like decreased daylight, fewer wage work opportunities, and winter privation.Footnote 32 Witnesses sometimes admitted to the desperate theft and slaughter of animals at Christmas to ‘serve their great need’, or because they ‘wanted money to buy meat withall’, while seasonal gift-giving was an oft-used excuse for ill-gotten gains.Footnote 33 Poverty, crime, and plenty thus all came to a head at the turn of the New Year. The cultural importance of Yuletide in early modern England is well documented, but less is known about the impact of the festive season upon work and the economy. Certainly, the evidence here suggests that Christmas’ position at the centre of the commercial year is not unique to modern industrial society. Moreover, and like today, the early modern festival had wide-ranging effects on other economic sectors.
Carework is another work category in which the supposedly lax winter season was in fact a busy one. Even allowing for the smaller size of the subsample, Figure 4.3 demonstrates a significant seasonal trend in care tasks, with an intense peak concentrated in February and March. This primarily derives from work in midwifery, childcare, and healthcare, and correlates with prevailing demographic patterns of the time. Over a third of midwifery tasks, 35 per cent, were observed in just two months, February and March, more than double the two-month average and falling neatly in line with the pronounced peak in monthly recorded baptisms for England in the period 1540 to 1700.Footnote 34 Nursing and infant-care duties increased in line with this annual seasonal baby boom. If these transitional months from winter to spring were a time of new life, they were also a season of loss. Parish registers show a marked increase in burials starting in December and rising steadily to a crest in March and April, when the cumulative deprivations and diseases of winter exacted their final toll.Footnote 35 It is perhaps unsurprising then to find 37 per cent of healthcare tasks, attending to the sick and dying, in the three months of December, February, and March. Mary Davis, a covenant servant from Chelvey, Somerset, tended upon her master ‘during the time of his sickness’ and death ‘in the month of December’ in 1669, while Joanna Clark, a married woman of Waterford, Hertfordshire, was hired to care for a sick woman ‘upon Shrove Tuesday’ in 1596.Footnote 36 The picture of carework that emerges from this material is one of steady monthly labour punctuated by an exacting season when the symbolic clash of winter and spring – death and life – must have felt all too real.
Although the summer harvest was undoubtedly the greatest drain on time and energy relative to other seasons, the findings discussed here complicate the traditional narrative of an idle or lax winter. Between commerce, carework, and food processing and provision, men and women had plenty to keep them occupied in the cold months, and for several sectors this season was the busiest time of the year. But to be more precise, many historians characterise winter as a slack season specifically in terms of wage work and paid employment. It may be that our more inclusive definition of work emphasises winter work, where traditional approaches have overlooked it. Commerce, for instance, was not usually waged work, though it of course generated income. Nevertheless, the seasonal distribution of ‘for another’ work, done for someone outside of the household, as well as the proportions of tasks that were done ‘for another’ each quarter, shown in Figure 4.5, suggest the importance of work for pay in the winter months.
‘For another’ vs overall work: indexed monthly distribution of tasks.
Notes: 100 = monthly average. Harvest additions and monthly weights applied; Integral excluded; F adjusted (x2.58). This differs from the standard multiplier as it is designed to give an equal number of male and female tasks with weighted monthly data attached.
Figure 4.5 Long description
The line graph presents index versus months. The vertical axis is labeled index and ranges from 40 to 180 in increments of 20. The horizontal axis presents the months of the year. Two lines are marked for, for another and overall. The approximate values are as follows. For the, for another line, January, 62, February, 116, March, 110, April, 70, May, 86, June, 90, July, 114, August, 166, September, 144, October, 84, November, 58, and December, 88. For the overall line, January, 70, February, 94, March, 90, April, 80, May, 102, June, 108, July, 120, August, 140, September, 132, October, 90, November, 76, and December, 92. Both lines follow an increasing trend with major fluctuations.
In simple terms of distribution, there was an even more pronounced gulf between August and January for ‘for another’ work than for work overall, as the harvest generated maximum opportunity for waged labour. Beyond the harvest, however, more ‘for another’ tasks were observed in both the winter and autumn quarters than in the spring. Indeed, when we consider the percentage of quarterly tasks done for another, we find 48 per cent from January to March, above the 46 per cent for the harvest quarter.Footnote 37 Much of this winter work done for others fell in the categories of transport, food processing, carework, and housework; the latter three were sectors where women played an important role. Indeed, gender could shape seasonal patterns of work in significant ways. Yet to what extent and in what manner did men’s and women’s work overlap over the course of the year? At least in terms of the monthly distribution of tasks, Figure 4.6 shows relatively little difference between the genders. The key distinction is that more tasks were recorded for women in the first half of the year, with the opposite true in the second. Women’s tasks substantially outnumber men’s from February to April – 56 per cent against 44 per cent – when carework was at its height, while men’s share of observed work tasks notably increased in August and September – 54 per cent against 46 per cent – when they played a bigger role in the grain harvest. Both genders participated in the end-of-year commercial boom, but men did 57 per cent of commerce activities in December, compared to the typical 50:50 commercial split between the genders. This partly sprang from men’s involvement in larger transactions like livestock purchases, whereas buying and selling food and drink in the month of Christmas was more evenly divided between the genders.
Gendered seasonality: indexed monthly distribution of male and female tasks.
Notes: 100 = monthly average. Harvest additions and monthly weights applied; Integral excluded; F adjusted (x2.58). This differs from the standard multiplier as it is designed to give an equal number of male and female tasks with weighted monthly data attached.
Figure 4.6 Long description
The line graph presents index versus months. The vertical axis is labeled index and ranges from 40 to 160 in increments of 20. The horizontal axis presents the months of the year. Two lines are marked for females and males. The approximate values are as follows. For the female line, January, 58, February, 114, March, 96, April, 88, May, 102, June, 108, July, 126, August, 126, September, 122, October, 88, November, 78, and December, 84. For the male line, January, 72, February, 74, March, 86, April, 76, May, 100, June, 102, July, 118, August, 152, September, 142, October, 96, November, 80, and December, 102. Both lines follow an increasing trend with major fluctuations.
Table 4.2 turns from task distribution to consider specific types of work. The highly gendered nature of the working year emerges in some predictable ways. Nonetheless, there were seasons when women and men worked in the same sector, highlighted by the italicised categories and subcategories in the table. Although the gender division of labour could still be stark within these broader subcategories, as discussed more fully in Chapter 6, men and women did work alongside one-another, particularly in summertime agriculture like May/June sheep shearing, June weeding, July haymaking, and August/September reaping.
Table 4.2 Long description
The table presents a seasonal breakdown of labor distribution by gender. The first column details women’s primary tasks with participation percentages and counts, the middle column depicts quarters, and the third column lists men’s primary tasks with equivalent metrics. The data reveals how agricultural and domestic work patterns shift seasonally between genders.
For January to March, women’s corresponding dataset is housework, 30%, care work, 24%, midwifery, 32, laundry, 28, food and drink provision, 22, childcare, 16, buy, 16, carry goods, 16 and for men the corresponding dataset is agriculture, 34%, transport, 18%, animal husbandry, 70, fieldwork, 52, wood husbandry, 34, buy, 29,
For April to June, women’s corresponding dataset is housework, 25%, agriculture, 23%, animal husbandry, 28, collecting water, 27, childcare, 26, food and drink provision, 23 and for men the corresponding dataset is agriculture, 37%, Transport, 20%, animal husbandry, 121, carting, 69, hunting and fishing, 48, financial, 41.
For July to September, women’s corresponding dataset is agriculture, 33%, housework, 24%, fieldwork, 64, collecting water, 34, food and drink provision, 31, gathering food, 24 and for men the corresponding dataset is agriculture, 52%, Transport, 15%, fieldwork, 286, farm transport, 157, animal husbandry, 71, carting, 71, buildings, 58.
For October to December, women’s corresponding dataset is housework, 28%, Carework, 18%, Commerce, 18%, food and drink provision, 28, buy, 24, carry goods, 21, laundry, 19, and for men the corresponding dataset is agriculture, 28%, commerce, 17%, transport, 17%, animal husbandry, 73, sell, 59, buy, 48, carry goods, 43.
Notes: Harvest additions and monthly weights applied; Integral excluded. Overlap in seasonal work between the two genders is italicised. Top two work categories (given as a repertoire percentage) and top four subcategories (no. of tasks) are shown.
Contrastingly, the winter quarter stands out as the season when the gender division of labour was most pronounced. With a relatively reduced workload, men focused on animal husbandry, ploughing, wood husbandry, and threshing. Women were particularly busy, partly due to the boom in carework discussed above but also because of the demands – and opportunities – of work in the housework category. Food provision, whether for household consumption or sale, was particularly intensive from December to February, when wives and female servants shouldered the brunt of festive responsibilities. Unsurprisingly, light and fire provision was a greater priority in the autumn and winter quarters, when women warded their households against the cold and dark. Lower temperatures and sunlight levels may have shaped the striking seasonality of laundry as well. Observed laundry work was at its ebb in the three months from November to January, when just 15 per cent of annual laundry tasks took place. A sudden and dramatic laundry peak in February and March, with 34 per cent of all tasks in just those two months combined, coupled with a smaller one in October, with 10 per cent of tasks, hints that women prepared in advance for the difficulties of the darkest months and then dealt with a washing backlog after them.Footnote 38
The pre-eminence of housework, alongside carework, during the autumn and winter quarters supports Amanda Flather’s argument that ‘women spent most of the winter working in and around the house and yard, cooking, cleaning, and caring for children’.Footnote 39 Yet when we view our locational and seasonal data together, the spatial dynamics to the gender division of labour become more complex. Women were indeed more likely to work within or around their own household in the winter half of the year, when 37 per cent of their tasks took place there as opposed to just 22 per cent in summer.Footnote 40 During the darkest months, from November to January, housework took up the largest share of these labours, representing 51 per cent of their repertoire of work at home, followed by spinning, carding, and other textile work. But women still performed the majority of their wintertime tasks outside their own household. Notably, 72 per cent of winter carework took place away from the home, as did over 40 per cent of housework activities, with women venturing forth for water, light, and laundry. Nor were women’s forays spatially confined to their own parishes: 21 per cent of women’s work tasks took them outside their home parish, accounting for nearly 40 per cent of all inter-parish work travel. During the winter half of the year, it was certainly true that women stayed closer to home, but the difference was not drastic, with 18 per cent of tasks outside the parish against 23 per cent in summer. Moreover, women’s out-of-parish work was above average in the autumn quarter, at 24 per cent, when they engaged heavily in commerce.Footnote 41
Mixed-gender participation in the important commercial season complicates the narrative that ‘for much of the winter different work in separate places kept contact between husbands and wives to a minimum’.Footnote 42 As the example of the Johnsons at the start of the chapter underlines, husbands and wives could work together to engage in the winter market boom. More generally, our location data shows that men, like women, spent the greatest amount of time working within or around the household during the winter half of the year, at 14 per cent against 9 per cent in summer. This household-based work reached a peak for men in December, at 19 per cent of their tasks, when, as Flather put it, ‘they retreated to the house out of the cold and wet weather for a few weeks to mend tools and perhaps to weave some cloth’.Footnote 43 Men’s household work in December, however, went far beyond craftwork. Their share of work within the household reached an annual high in the final month of the year, with a 40:60 male-to-female split, against an average of 30:70 across the year, an increase that came across almost all categories, including housework and carework. From the perspective of the spatial gender division of labour, December thus represented a winter counterpoint to the harvest months, with men and women more frequently working alongside each other in and near the home, as had been the case outside in the fields during July and August. William Cresey of Cheshunt, Hertfordshire, for example, dressed a slaughtered pig in his mother’s house on St Stephen’s Day 1590, which she then roasted for them to eat. Likewise, Laurence Farlton and his wife, of Newton, Cambridgeshire, pilled hemp together in their house on a December night in 1661.Footnote 44
If gender influenced the seasonality of work, did occupations do the same? Table 4.3 breaks down and compares the quarterly working patterns of agricultural workers with those of artisans and tradesmen. What emerges are two quite distinct working years, which occasionally intersected or overlapped in character, just as we saw with women and men. As might be expected, agriculture and craftwork dominated the working years of the respective occupational groups. But the seasonality of work in adjacent sectors is more interesting. Chapter 2 explored the central importance of tertiary forms of labour like commerce to workers in the primary and especially secondary sectors. These results show that such tertiary engagement often fell along seasonal lines. For artisans and tradesmen, commerce, though of constant importance, became more pronounced in the autumn and winter, when the market for foodstuffs, textiles, and other wares reached its height. For agricultural workers, commerce took up the second-largest share of their time in autumn and spring, the two big quarters for livestock transactions. As we have seen already in this chapter, the summer harvest pulled in labour from diverse groups, like craftsmen, who were not otherwise heavily involved in agriculture. Yet while the fieldwork of tradesmen was largely seasonal, other agricultural activities like animal husbandry complemented their work repertoire throughout the year. Such evidence of by-employment or makeshift economies appears most stark, for both occupational groups, in the spring – the quarter when the overall proportion of ‘for another’ work fell to an annual low at 35 per cent.Footnote 45
Table 4.3 Long description
The table examines labor distribution patterns across agricultural and artisan or trade occupations. The first section provides a time period with seasonal breakdowns, and the second column depicts percentage distributions of primary occupations, a total of 555 tasks. The third column indicates the percentage distributions of artisan or trade occupations, a total of 654 tasks. The data reveals how work emphasis shifts between farming and skilled trades throughout the year, with percentage distributions indicating seasonal labor priorities.
1. Agricultural occupations, 555 tasks:
For Overall, the corresponding values are agriculture, 51% and transport, 16%.
For January to March, the corresponding values are agriculture 52%, transport, 17%, and distribution, 19%.
For April to June, the corresponding values are agriculture, 45%, commerce, 15%, and distribution, 20%.
For July to September, the corresponding values are agriculture, 64%, transport, 16%, and distribution, 38%.
For October to December, the corresponding values are agriculture, 36%, commerce, 19%, and distribution, 24%.
2. Artisan or Trade occupations, 654 tasks:
For Overall, the corresponding values are crafts and construction, 30%, and commerce, 19%.
For January to March, the corresponding values are commerce, 28%, crafts and construction, 27%, and distribution, 20%.
For April to June, the corresponding values are crafts and construction, 32%, management, 15%, commerce, 15%, agriculture, 15%, and distribution 23 %.
For July to September, the corresponding values are crafts and construction, 38%, agriculture 26%, and distribution 32%.
For October to December, the corresponding values are crafts and construction, 26%, commerce 21%, and distribution 25%.
Notes: Harvest additions and monthly weights applied; Integral excluded. Distribution = proportion of tasks per quarter. Overlap in seasonal work between the two occupational groups is italicised.
a Agricultural = yeomen, husbandmen, and agricultural trades.
b Artisan/Trade = artisans, commercial trades, and transport trades.
In contrast to husbandmen and artisans or tradesmen, the work experience of male servants and labourers showed relatively little variation through the year. It was dominated by agriculture and transport work, with little evidence of market activity outside the autumn quarter. That being said, some specific tasks did vary with the seasons, and often clearly extended from their particular employment arrangement. Labourers’ tasks were more seasonal: threshing and wood husbandry in autumn and winter, carting and the digging of marl or earth in the spring, and harvest fieldwork and farm transport in the summer. Male servants shared in many of these tasks, but on the whole were more oriented to steady year-round responsibilities on the farms where they lived and worked. Animal husbandry was a constant, while their fieldwork and food processing tasks followed the seasonal cycle of ploughing, sowing, reaping, and threshing. Female servants experienced similar levels of seasonal variation to their male counterparts. Housework and animal husbandry dominated the whole year, but milking especially characterised the summer half, and food and drink provision the winter. Such gendered experiences of the working year emerge even more clearly when we consider the workweek and holidays.
4.3 Working Week and Year
If seasonal and monthly patterns shaped early modern work, so too did the rhythms of the week. Yet as Figure 4.7 suggests, the general distribution of tasks varied relatively little according to weekday for men and women alike. Men observed a fairly cyclical work pattern, with moderately lower midweek activity levels bookended by Monday and Saturday peaks in activity: there was certainly no ‘St Monday’ in evidence here. Women were busy on these two days as well, but their workweek was less straightforwardly structured. On Sunday, both genders performed fewer tasks, but this held greater contrast for men than it did for women. For the latter, just as many or more tasks were recorded on Sunday as on Tuesday or Wednesday. Yet, while the lines between ‘days of rest’ and ‘days of work’ were clearly more blurred for women, men still contributed significantly to Sunday work. Indeed, it is striking quite how much sabbath work the dataset captures overall – only slightly fewer tasks than that observed on Wednesdays. The character of Sunday labour is explored further below, but some of the general causes behind these weekday variations emerge when we break down the data according to work categories.
The gendered workweek: distribution and gender division of tasks per weekday.
Notes: Integral excluded. Removing integral tasks has a negligible effect on weekday distribution overall, but they are excluded here to demonstrate that these patterns are not a consequence of certain days being more criminogenic. F adjusted (x2.67). This differs from the standard multiplier as it is designed to give an equal number of male and female tasks with weekday data attached. Overall daily task total (F adjusted): Mon 737, Tue 536, Wed 539, Thu 683, Fri 634, Sat 818, Sun 521.
Figure 4.7 Long description
The double bar graph displays number of tasks versus categories. The vertical axis is labeled number of tasks and ranges from 0 to 500 in increments of 100. The horizontal axis is labeled with days of the week from Monday to Sunday. Each bar represents task counts for two gender categories, female and male. The data in the format, day, gender, and task count are as follows. Monday, females 47, males 53. Tuesday, females 43, males 57. Wednesday, females 46, males 54. Thursday, females 57, males 43. Friday, females 52, males 48. Saturday, females 50, males 50. Sunday, females 54, and males 46.
As Figure 4.8 shows, agricultural work was distributed evenly throughout the week, including Sundays. Conversely, commerce and transport fell off significantly on the sabbath day. Tasks within the latter two categories reached their height on Saturday, which was far and away the most common market day recorded in our dataset, across all regions.Footnote 46 Women in particular played a leading role in Saturday markets, on that day accounting for 59 per cent of commerce tasks, 56 per cent of management tasks, and performing their largest daily share of transport at 43 per cent. Market days more broadly, as catalysts for commerce and transport, brought variation to the workweek. At the beginning of this chapter, the Johnsons’ preparations for a Saturday market in Chester demonstrated how these weekly events could lengthen the workday, as men and women rose early to travel to market towns and came home late. The cumulative impact of these differing work hours comes through in the data for ‘phases of the day’, discussed in Section 4.4, where Saturday exhibits the highest percentage of combined early morning and morning tasks per weekday, at 43 per cent, against a Monday-to-Friday average of 38 per cent. Not infrequently, deponents dated actions in relation to market days, and in general people seem to have had a keen awareness of when these events took place within their region. Witnesses’ suspicions were sometimes aroused, for example, when individuals transported certain goods to towns when it was ‘not the market day there’.Footnote 47 Some weekdays, like Wednesday, played host to markets less often than others, and this may partly explain the relatively lower number of overall tasks reported on such days in the dataset, as commerce and transport tasks in particular tended to be more common on market days.
The workweek for agriculture, commerce, and transport.
Notes: Integral included as they have a negligible impact on weekday distribution. F adjusted (x2.67). This differs from the standard multiplier as it is designed to give an equal number of male and female tasks with weekday data attached. Task totals per category: agriculture and land (1,445 tasks); commerce (1,165 tasks); transport (862 tasks).
Figure 4.8 Long description
The bar graph displays task percentage versus categories. The vertical axis is labeled percentage of tasks per day and ranges from 0 to 30 in increments of 5. The horizontal axis is labeled with days of the week from Monday to Sunday. Each bar is divided into three task categories. Agriculture and land, commerce, and transport. The data in the format, day, agriculture and land percentage, commerce percentage, and transport percentage, are as follows. Monday, 16%, 15%, and 17%. Tuesday, 13%, 14%, and 14%. Wednesday, 14%, 10%, and 14%. Thursday, 13%, 12%, and 14%. Friday, 15%, 17%, and 16%. Saturday, 14%, 27%, and 19%. Sunday, 13%, 6%, and 8%.
The high level of activity on Mondays, however, is harder to explain. Although it was not a popular market day, Monday saw above-average activity in commerce, and in almost every other work category besides. These findings cast doubt upon the relevance of ‘St Monday’ as a traditional day off work in early modern England.Footnote 48 Certainly, if the phenomenon did exist, it does not seem to have been a rural one.Footnote 49 The strongest empirical evidence for St Monday comes from mid-eighteenth-century London, where Voth found that Sundays and Mondays were the weekdays with the lowest probability of observing people at work. He did not find equivalent evidence in a contemporary northern and more rural sample of depositions, and this complements other evidence – especially literary evidence – that St Monday was an observance of urban artisans.Footnote 50 The work-task data suggests the observance may have been even more limited than this, perhaps extending no further than the metropolis. When we isolate those work tasks in the sample performed in large towns and market towns, it shows that Monday’s work activity was above average even in these urban places.Footnote 51
Furthermore, mapping the workweeks of different occupational groups, as shown in Figure 4.9, demonstrates that artisans, tradesmen, and labourers did not shy from work on Monday, any more so than agricultural workers and servants did. If anything, the opposite seems to have been true, with Monday seeing a flurry of activity across sectors as workers returned to their labours after a mandated day of rest. Moreover, when we survey the data for women and different male occupational groups, no single weekday emerges as an alternative ‘day off’ to St Monday. Certainly, breaking down the data in this way reveals some variation between different workers, although we must allow for the much smaller size of these occupational and gender subsamples. Notably, the number of recorded tasks is lower on Wednesday for artisans and tradesmen, and on Tuesday and Thursday for labourers. Overall, however, these findings suggest a regular six-day workweek prevailed across most sectors, with only a moderate dip in activity across most sectors on Sunday.
Occupational workweeks (male).
Notes: Integral included as they have a negligible impact on weekday distribution. Agricultural = yeomen, husbandmen, and agricultural trades (478 tasks); craft/trade = artisans, commercial trades, and transport trades (636 tasks); labourers (373 tasks); male servants (222 tasks).
Figure 4.9 Long description
The bar graph displays task percentage versus categories. The vertical axis is labeled percentage of tasks per day and ranges from 0 to 25 in increments of 5. The horizontal axis is labeled with days of the week from Monday to Sunday. Each bar is divided into four occupational categories. Agricultural, N = 478 tasks, Labourers, N = 373 tasks, Craft or Trade, N = 636 tasks, and Male Servants, N = 222 tasks. The data in the format, day, agricultural percentage, labourers percentage, craft or trade percentage, and male servants percentage, are as follows. Monday, 18%, 17%, 16%, and 14%. Tuesday, 14%, 18%, 13%, and 15%. Wednesday, 16%, 17%, 14%, and 18%. Thursday, 17%, 13%, 15%, and 9%. Friday, 12%, 16%, 17%, and 11%. Saturday, 11%, 19%, 18%, and 17%. Sunday, 9%, 10%, 11%, and 15%. All values are approximated.
Economic and social historians’ interests in St Monday and the length of the workweek derive from a broader concern with the number of days in the working year, and associated levels of income and ‘industriousness’. For example, historians have often assumed a five-day workweek existed in preindustrial England when calculating annual incomes and standards of living from wage series. Likewise, Voth’s findings of St Monday and its decline in London over the latter half of the eighteenth century fold into broader arguments that an increase in workdays during the early industrial period – due in part to a longer workweek – led to greater outputs. Yet grand narratives of economic change over time like these largely trade on a strict dichotomy between workdays and ‘non-work’ or leisure days, which the work-task data simply does not support.Footnote 52 The most obvious case in point can be found in the results for Sunday, the quintessential day of rest.
From the early medieval period through successive reformations, both canon and secular law decreed that the Lord’s Day should be free from ‘abusive pursuit of any servile occupations’.Footnote 53 Yet even on this day, when all but the most indispensable forms of work were ostensibly forbidden by law and custom, the work-task data shows that men and women did a substantial number of tasks. Certainly, they did more work than the ‘nil’ assumed in the calculations of ‘workdays’ which underpin most wage series. That being said, the types of work and forms of employment found on Sundays remain important factors to consider. Our inclusive definition of work, for example, may capture more sabbath activity than studies devoted solely to waged labour. Yet when we look at the percentage of tasks done ‘for another’ – those activities most likely to be remunerated – Sunday’s proportion, at 36 per cent, is above the daily average of 32 per cent, and second only to Monday’s of 37 per cent. These results suggest more remunerated or employed labour on the sabbath than other approaches to the history of work typically assume. One possible reason for this is that the work-task approach captures the work experiences of women and servants in much more detail than wage accounts.
To take servants first, we have already seen from Figure 4.9 that in the distribution of tasks across the week that male servants carried out, an above-average proportion of their tasks occurred on Sunday, at 15 per cent. Specifically, out of the four occupational groups displayed in Figure 4.9, male servants shouldered the lion’s share of agricultural and transport work on the Lord’s Day.Footnote 54 And so we find that William Bend, a servant from Parson Drove, Cambridgeshire, ‘went to water two mares at the pond’ on a ‘Sunday morning’ in 1664, and that Jacob Jackson of Hurworth, Durham, delivered his master’s sheep after marking their ears on a Sunday in 1603.Footnote 55 Such evidence suggests that workers on annual contracts benefited less from the day of rest than those with more freedom and agency in their work schedules. Of course, this technically flew in the face of contemporary Sabbatarian commentary, which emphasised the responsibility of householders to furnish Sunday rest to those under their supervision, including even animals, ‘so that they might have bodily refreshment and time for spiritual exercises’.Footnote 56 Servants and apprentices sometimes held their masters to account for dereliction of this duty, as in 1650 when shoemaker’s apprentice Benjamin Hooper of Wells, Somerset, claimed he ran away ‘because his master did make him work upon the sabbath days’ in cleaning shoes, ‘packing of wares’, and running the shop.Footnote 57
To judge from sabbath-breaking prosecutions in the church courts, Hooper’s experience was shared by many servants.Footnote 58 Yet the proportion of all Sunday ‘for another’ tasks undertaken by servants was comparable to the proportion of ‘for another’ tasks they undertook on weekdays, so other forms of ‘for another’ work continued on this day too. The small number of examples of ‘explicitly paid’ sabbath labour are telling here, and do reflect Sunday’s status as an important day of recreation. Thomas Westcott and William Philipps, for example, played the violin and cittern ‘for a little money’ late on Sunday night in 1691 at ‘the house of Hanna Dyer who sold beer or ale’ in Newton Poppleford in Devon.Footnote 59 Likewise, in 1637 Alice Alredd of Atherton, Lancashire, was paid in ready money and linens for serving ‘so much drink and victuals as came to nine pence’ to three men who came into her house ‘upon Sunday’.Footnote 60 These examples hint at what is perhaps the most notable characteristic of ‘for another’ work on Sundays: that women did the majority of it, at 62 per cent. And while young servants bore the brunt of Sunday labour among male occupations, it seems married women took the lead here, performing a larger share of ‘for another’ female tasks compared to other weekdays. The relationship between gender, status, and sabbath work comes into sharper focus when turning to the categories of tasks most commonly performed on the day.
Table 4.4 compares Sunday’s work repertoire to a Monday-to-Friday average, highlighting the distinct nature of sabbath activity. Sunday work was more characterised by housework and carework than other weekdays, with crafts and construction, commerce, and transport activity all subdued. Although the former two categories were still dominated by women, men performed a larger share of such tasks on the Lord’s Day than they did during the week. Agricultural work made up just as much of the repertoire on Sundays as it did Monday through Friday, but on the Lord’s Day tasks overwhelmingly centred on animal husbandry, with a fairly even gender division of labour. In an illustrative example from Heacham, Norfolk, in 1624, John Orrman ‘in the forenoon before divine service … went to his cows which were feeding in a clay pit … and there kept by a girl of his whom he sent away to church’. He stayed in the pit until noon, alongside Ann Elvun and Margery Eran ‘who were there also feeding of their cattle’, after which he then drove his cows homeward.Footnote 61
Table 4.4 Long description
The table indicates weekday labor tasks and their corresponding statistical values. The first column lists task categories, followed by two columns on Monday to Friday average, and Sunday. There are two subcategories for each column that is repertoire in percentage and percentage by F. The data is arranged below:
1. Dataset for Monday to Friday average:
For Agriculture and land, the corresponding values are 28.7 and 32.7.
For Carework, the corresponding values are 10.9 and 86.8.
For Commerce, the corresponding values are 11.2 and 48.3.
For Crafts and construction, the corresponding values are 7.9 and 22.
For Food processing, the corresponding values are 3.7 and 34.7.
For Housework, the corresponding values are 17.2 and 87.8.
For Management, the corresponding values are 4.1 and 39.9.
For Transport, the corresponding values are 15.1 and 28.7.
For Other, the corresponding values are 1.2 and 37.3.
For Total, the corresponding values are 100.1 and 49.
For Total tasks, the corresponding values are 3129 and 49.
2. Dataset for Sunday:
For agriculture and land, the corresponding values are 28.1 and 41.3.
For carework, the corresponding values are 19.4 and 79.2.
For commerce, the corresponding values are 6.2 and 59.8.
For crafts and construction, the corresponding values are 5 and 18.7.
For food processing, the corresponding values are 1.5 and 61.8.
For housework, the corresponding values are 18.8 and 81.6.
For management, the corresponding values are 9.7 and 38.5.
For transport, the corresponding values are 9.7 and 19.1.
For other tasks, the corresponding values are 1.7 and 54.8.
For Total, the corresponding values are 100.1 and 54.4.
For Total tasks, the corresponding values are 521 and 54.4.
Notes: Integral excluded; F adjusted (x2.42). This differs from the standard multiplier as it is designed to give an equal number of male and female tasks with weekday data attached. Saturday has been excluded from the Monday-to-Friday average because the high proportion of market-day commerce makes it distinct from the other days of the week.
For many men like John Orrman, the tending of livestock was the principal Sunday task, with a notable absence of the fieldwork, carting, farm transport, and other physical labour, which otherwise dominated the male workweek. Artisans generally seem to have set down tools as well, with the occasional exception of building work. Roofing was not an uncommon task on Sunday, and it was a dangerous one at that to judge from several coroner’s reports. On a Sunday in 1527, for example, the labourer John Webbe met his untimely end while ‘thatching roof at the house of John Colyns’ in West Lavington, Wiltshire.Footnote 62 Thatchers, alongside carpenters, masons, and other builders, were also sometimes called upon to inspect damaged church structures on Sunday, occasionally quite far from their home parish.Footnote 63 This illustrates the willingness of parsons, churchwardens, and other church officials to engage workers in ostensibly prohibited sabbath work. Of course, some Sabbatarian commentators did make allowance for labour deemed necessary to the preservation of life and property, which might include building work.Footnote 64
Compared to men, women must have noticed relatively little difference between Sundays and other weekdays, in terms of the character of their work and time-use. Caring for children, milking, or tending to livestock, and providing the household with food and water, were constant responsibilities, many of which fell into the ‘preservation of life’ category. Nevertheless, women’s work patterns did shift in a few notable ways on their ‘day of rest’. For one, laundry activities fell to a weekly low, likely due to the public and arduous nature of the work. Food and drink provision, on the other hand, rose to a high. As touched on above, this may reflect the recreational demands and money-making opportunities of the Lord’s Day, over and above standard household subsistence. Indeed, what little commerce did take place on the sabbath appears to have passed principally through the hands of women engaged in selling or buying food and drink. On a Sunday in November 1597, for instance, Mawde Leighe of Sidbury in Devon sold cheese to Joan Lugge, while in July 1632 Alice Baylie, the wife of a labourer, sold drink from her house in Poole Keynes in Wiltshire.Footnote 65 Thus, just as women picked up the slack left by reduced waged work for men in winter, they found ways to generate income and employment on the one weekday when labour was technically forbidden.
In contrast to prosecutions from church and criminal courts, which tend to capture the most egregious and publicly visible examples of profaning the Lord’s Day with ‘servile labour’, our depositional evidence suggests a common attitude towards sabbath work, which was respectful yet pragmatic. Those activities most strictly condemned by authorities from the medieval period onwards – trading, travel, carrying burdens, craftwork – were mostly avoided, but this left an array of tasks that often stretched the definition of ‘necessary’. What seems to have been readily understood by ordinary women and men as the chief Sabbatarian duty was that work should not disrupt or conflict with divine service. The innkeeper William Hellycar of Williton in Somerset, for example, refused to accompany Samuel Crustman, a weaver, in hunting fowl one Sunday in 1638 because ‘it would be too far to go unto the said ground and return again at evening prayer’.Footnote 66 When witnesses did admit to Sunday labours, they might clarify that these happened ‘after prayers in the afternoon’, or ‘a little before evening prayer’. John Houlton went to a common near Heacham, Norfolk, on a Sunday morning in 1624 to retrieve his grass-filled cart, but upon hearing the ringing of a bell he perceived it to be ‘church time and so came away to church’.Footnote 67 These examples speak to a common belief that work should not sully worship but also a view that it was the sabbath service rather than the day itself which took priority. Thus, the modest decrease in work tasks observed on Sundays likely reflects those hours taken up with worship, and no doubt some recreation, rather than day-long abstentions from work.
The key conclusion is that men and women carefully honoured the sabbath, as they understood it, while also pursuing what work they could. Remaining completely idle one day out of every week was likely neither affordable nor realistic. Crucially, this balanced approach embraced opportunities for income generation and paid employment, especially for women, and this has far-reaching implications for our understandings of the early modern economy in England. For one, it underlines the systematic underestimation of women and servants’ contributions in the literature. More broadly, it complicates those narratives and models which hinge upon the length of the working year and its change over time, for these often assume that every Sunday and holy day was a ‘day off’. Indeed, our evidence for activities done upon holy days further undermines such binary understandings. Using a subsample of the work-task dataset for which we have full calendar dates, Table 4.5 displays the repertoire of work tasks done on those holy days (excluding Sundays) prescribed in the Book of Common Prayer of 1561. Publication of the latter capped off a series of reforms, begun in 1536, which culled dozens of traditional feast days from the calendar. Twenty-seven holy days, in addition to Sundays, were left upon which all ‘lawful bodily labour’ should be set aside for prayer and worship. As the core festivals of the church throughout our period, they provide a conservative but serviceable list to test the extent to which holy days were observed as days off work: a question central to theories of economic, and of course religious, change.Footnote 68
Table 4.5 Long description
The table indicates weekday labor tasks and their corresponding statistical values. The first column lists task categories, followed by two columns on workdays, and book of common prayer holy days. There are two subcategories for each column that is repertoire in percentage and percentage by F. The data is arranged below:
1. Dataset for Workdays:
For agriculture and land, the corresponding values are 27.8 and 29.
For carework, the corresponding values are 10.6 and 87.5.
For commerce, the corresponding values are 10 and 52.8.
For crafts and construction, the corresponding values are 8.4 and 17.1.
For food processing, the corresponding values are 3.8 and 37.7.
For housework, the corresponding values are 18.4 and 88.4.
For management, the corresponding values are 3.9 and 44.
For transport, the corresponding values are 15.9 and 31.6.
For other, the corresponding values are 1.3 and 48.1.
For total, the corresponding values are 100.1 and 49.
For total tasks, the corresponding values are 2958 and 49.
2. Dataset for book of common prayer holy days:
For agriculture and land, the corresponding values are 24.5 and 45.1.
For carework, the corresponding values are 17.4 and 92.6.
For commerce, the corresponding values are 17.3 and 29.5.
For crafts and construction, the corresponding values are 2.6 and 0.
For food processing, the corresponding values are 4.3 and 39.8.
For housework, the corresponding values are 15.4 and 93.8.
For management, the corresponding values are 4.3 and 39.8.
For transport, the corresponding values are 12.8 and 39.8.
For other tasks, the corresponding values are 1.5 and 57.
For total, the corresponding values are 100.1 and 56.1.
For total tasks, the corresponding values are 312 and 56.1.
Notes: Integral excluded; F adjusted (x2.65). This differs from the standard multiplier as it is designed to give an equal number of male and female tasks with calendar date data attached. Workdays = tasks with calendar date data which did not occur on Sunday or a holy day. Holy day count excludes Sunday tasks.
While we must allow for a much smaller sample size where holy days are concerned, the difference in work repertoires between the types of calendar days is striking. Most obviously, commerce took up a much larger share of time on holy days compared to workdays, and especially compared to the Sunday repertoire presented in Table 4.4, where it was just 6.2 per cent. This relates partly to the close connection between fairs and traditional feast days, for at least 36 per cent of holy day commercial activities were associated with these major economic events.Footnote 69 Thus, Margaret Slater sold linens at a fair in Preston ‘upon the feast day of Simon and Jude’ in 1626, while the glover Francis Richardson ran a stall ‘on Easter Monday … at Oxborough Fair’ in Norfolk in 1627.Footnote 70 Yet even when we filter out ‘fairtime’ activities, commerce appears more pronounced on feast days, likely due to the celebratory demands they occasioned.
Commerce aside, the holy day and sabbath repertoires appear largely similar in terms of work categories and gender division of labour: housework and carework carried on while craftwork and transport abated, so that women were less likely than men to have a day off. Indeed, Thomas Tusser warned huswives in particular to ‘forget not the feasts that belong to the plough’. He tasked them, alongside their servant maids, with providing ‘flesh [and] corn … wafers and cakes … good cheer and welcome’ to their employees on certain festive occasions, alongside twice weekly roast beef dinners on Sundays and Thursdays.Footnote 71 The only other pronounced difference between Sundays and holy days was the nature of agricultural work. Although it generally did not cease on either type of ‘day off’, on holy days agricultural work was characterised by a wider array of activity than the animal husbandry which dominated the sabbath.Footnote 72 One might therefore witness William Carter driving a cart within a Topcroft farm in Norfolk on St Stephen’s Day 1571, or John Whitehead and James Booth ploughing a field in Idle, Yorkshire, on Lady Day 1696.Footnote 73
Overall, this evidence suggests that holy days fell somewhere between Sundays and workdays in the estimations of working men and women, perhaps reflecting contemporary religious debates among church and state authorities. Should holy days be disregarded entirely, treated as equal to the sabbath, or divided into tiers with Sundays and the major feasts of Christ on top?Footnote 74 The work-task data suggests that the latter tier system prevailed in popular practice. Sorting holy day tasks into feast day categories, as in Table 4.6, shows that work was most prevalent on busy quarter days, and least during the chief festivals of Christ’s life, Christmastide, and Eastertide.
Table 4.6 Long description
The table presents an analysis of labor patterns during religious holy days. The first column lists holy day categories, followed by three data columns, percentage of tasks performed during each holy day, percentage of annual holy days each period represents, and the percentage difference between these distributions. The data reveals variations in work activity during different religious observances.
For Christmastide, the corresponding values are 14.0, 17.2, and negative 3.3.
For Eastertide, the corresponding values are 6.5, 10.3, and negative 3.8.
For Whitsuntide, the corresponding values are 13.7, 13.8, and negative 0.1.
For Saints’ days, the corresponding values are 36.3, 34.5, and 1.8.
For Quarter days, the corresponding values are 29.5, 24.1, and 5.3.
For total, the corresponding values are 100, 99.9, and 0.1.
For Total tasks and days, the corresponding values are 336 tasks, 29 days, and not applicable.
Note: Book of Common Prayer (1561) Holy days. Christmas Day is placed in ‘Quarter Days’ alongside Lady Day, Midsummer, and Michaelmas, and cross-quarter days Candlemas, May Day, and All Saints. Christmastide includes the Feasts of Stephen, John the Evangelist, Holy Innocents, New Year, Epiphany; Eastertide includes Easter Sunday, Monday, Tuesday; Whitsuntide includes Ascension, Whit Sunday, Monday, Tuesday; Saints’ Days includes Feasts of Mathias, Mark, Peter and Paul, James, Bartholomew, Matthew, Luke, Simon and Jude, Andrew, and Thomas. Holy day count includes Sunday tasks when the Sunday fell on a holy day.
Certainly, further research is needed into how such weekly and annual work patterns may have shifted over the course of two centuries of intense religious and economic change. Yet already these sabbath and holy day findings suggest a more nuanced relationship between work and traditional days of worship and leisure in early modern England than historians have typically allowed. Rather than a strict dichotomy between ‘workday’ and ‘day off’, there was a spectrum, within which men and women deployed different labour strategies for Sundays, holy days, and workdays. These strategies rarely resulted in a completely idle day off, and the extent to which they did was dependent on factors like gender, occupation, or age and marital status. Any attempt to measure the working week or year must grapple with these realities, just as it must account for the length and character of the working day itself.
4.4 Working Day
If the work-task data challenges ideas about the length and character of the early modern workweek and year, it also confronts assumptions about the rhythms of the premodern working day. Whilst E.P. Thompson and Jacques Le Goff characterised preindustrial agrarian work rhythms as erratic and irregular, more recent studies of working hours in the early modern period have found that regular workdays, with consistent start and end times, were widely observed.Footnote 75 Voth’s influential study of increased industriousness in eighteenth-century London argued that the number of days worked went up, but that the length of the working day did not; and Hailwood’s recent study of sixteenth-and seventeenth-century rural working hours argued that a regular working day was in place long before the arrival of industrial modernity.Footnote 76 This section builds on Hailwood’s analysis of the south-west data, adding to it the north and east samples. The increased number of observations of work tasks taking place at specific hours, such as ‘at two of the clock in the afternoon’ or ‘between three and four of the clock in the morning’; and within broader phases of the day, such as ‘in the morning early’, ‘in the evening’, allows for an examination not only of the distribution of work tasks across the day but of how this varied by gender, the seasons, types of work, and occupational groupings.
The basic outline of the working day can be seen in Figure 4.10. This plots the numbers of tasks recorded in the database for each hour of the day, excluding integral tasks, which were often linked to crimes, and therefore overrepresent night-time activity – and adjusted to give an equal total of male and female work tasks, resulting in a total of 1,487 tasks. The overall shape looks, in most respects, like a recognisably modern, ‘classic’ working day. There were the first stirrings of activity between five and eight in the morning, before a pronounced peak of busyness between eight and midday. The next two hours show a slackening of pace, which coincided with taking dinner, the main meal of the day in early modern England. The afternoon, from two through to six, saw further peak hours. Levels of activity then dropped, but here we can see one of the key findings that has been overlooked in previous studies: the importance of the evening as a time of work. Indeed, the hours between 6 pm and 9 pm witnessed lower levels of work than during the morning or afternoon peaks, but they were not a time when all the day’s labour had come to an end.
Tasks per hour.
Notes: Integral excluded; F adjusted (x2.58). This differs from the standard multiplier as it is designed to give an equal number of male and female tasks with hourly data attached.
Figure 4.10 Long description
The vertical axis is labeled number of tasks and ranges from 0 to 160 in increments of 20. The horizontal axis is labeled hour of the day and ranges from 1 to 24 in increments of 1. The approximate number of tasks for each hour is as follows. Hour 1 is 10, 2 is 8, 3 is 14, 4 is 24, 5 is 42, 6 is 42, 7 is 44, 8 is 92, 9 is 68, 10 is 114, 11 is 86, 12 is 60, 13 is 68, 14 is 104, 15 is 150, 16 is 104, 17 is 78, 18 is 62, 19 is 64, 20 is 64, 21 is 46, 22 is 42, 23 is 40, and 24 is 28.
The line between ‘worktime’ and ‘non-worktime’ thus was a blurry one. As with Sundays and holy days, this pattern is partly a result of the work-task methodology recording women’s work, both paid and unpaid. As Figures 4.11 and 4.12 show, women’s work was less confined to the ‘peak’ hours of the workday than was men’s. For women, with the exception of a spike between 2 pm and 4 pm, work tasks were fairly evenly spread between 8 am and 9 pm, and there were also more tasks recorded in the two or three hours either side of this window than was the case for men. That said, Figure 4.12 shows that men also undertook a significant amount of ‘off-peak’ work tasks but in a lower proportion relative to ‘peak’ working hours.
Tasks per hour by gender (female).
Notes: Integral excluded; F adjusted (x2.58). This differs from the standard multiplier as it is designed to give an equal number of male and female tasks with hourly data attached.
Figure 4.11 Long description
The vertical axis is labeled number of tasks and ranges from 0 to 80 in increments of 10. The horizontal axis is labeled hour of the day and ranges from 1 to 24 in increments of 1. The approximate number of tasks for each hour is as follows. Hour 1 is 6, 2 is 5, 3 is 10, 4 is 18, 5 is 26, 6 is 26, 7 is 20, 8 is 46, 9 is 34, 10 is 42, 11 is 34, 12 is 32, 13 is 28, 14 is 54, 15 is 74, 16 is 46, 17 is 38, 18 is 24, 19 is 42, 20 is 38, 21 is 26, 22 is 26, 23 is 16, and 24 is 12.
Tasks per hour by gender (male).
Notes: Integral excluded.
Figure 4.12 Long description
The vertical axis is labeled number of tasks and ranges from 0 to 90 in increments of 10. The horizontal axis is labeled hour of the day and ranges from 1 to 24 in increments of 1. The approximate number of tasks for each hour is as follows. Hour 1 is 6, 2 is 4, 3 is 6, 4 is 8, 5 is 16, 6 is 16, 7 is 24, 8 is 46, 9 is 36, 10 is 74, 11 is 52, 12 is 26, 13 is 42, 14 is 52, 15 is 76, 16 is 58, 17 is 40, 18 is 38, 19 is 18, 20 is 28, 21 is 20, 22 is 24, 23 is 16, and 24 is 16.
It has been suggested that seasonality would have had a significant impact on the length of the working day in preindustrial society, with shorter daylight hours restricting the window in which to work.Footnote 77 As the above analysis has demonstrated, work tasks undertaken did change with the seasons, but the extent to which work rhythms did may well have been overestimated previously. By splitting the work-task data into two broad seasons – April to September; October to March – Figures 4.13 and 4.14 make it is clear that work activity did tend to start earlier in the summer, but there was not an earlier cessation to labour in the winter months; in fact, winter evenings were a period of considerable activity.
Tasks per hour in summer.
Figure 4.13 Long description
The vertical axis is labeled number of tasks and ranges from 0 to 90 in increments of 10. The horizontal axis is labeled hour of the day and ranges from 1 to 24 in increments of 1. The approximate number of tasks for each hour is as follows. Hour 1 is 6, 2 is 2, 3 is 6, 4 is 2, 5 is 22, 6 is 28, 7 is 32, 8 is 58, 9 is 40, 10 is 66, 11 is 54, 12 is 38, 13 is 38, 14 is 62, 15 is 76, 16 is 58, 17 is 46, 18 is 34, 19 is 28, 20 is 22, 21 is 24, 22 is 20, 23 is 16, and 24 is 10.
Tasks per hour in winter.
Notes: Integral excluded; F adjusted (x2.58). This differs from the standard multiplier as it is designed to give an equal number of male and female tasks with hourly data attached.
Figure 4.14 Long description
The vertical axis is labeled number of tasks and ranges from 0 to 70 in increments of 10. The horizontal axis is labeled hour of the day and ranges from 1 to 24 in increments of 1. The approximate number of tasks for each hour is as follows. Hour 1 is 2, 2 is 5, 3 is 10, 4 is 20, 5 is 12, 6 is 14, 7 is 10, 8 is 32, 9 is 30, 10 is 48, 11 is 30, 12 is 22, 13 is 28, 14 is 38, 15 is 66, 16 is 46, 17 is 32, 18 is 26, 19 is 36, 20 is 38, 21 is 16, 22 is 18, 23 is 20, and 24 is 19.
As Hailwood has previously argued, historians have overstated the extent to which early modern workers were dependent on good light levels to complete their work tasks. The example of the Johnsons at the start of the chapter shows how workers often made do with the light of a fire, a candle, or even the stars and moon, as well as relying on touch and feel.Footnote 78 But taking a closer look at what types of work were taking place during these evenings, and during other phases of the day, gives a better understanding of the long tail to the working day in early modern England. The number of tasks available for such an analysis can be increased by including not only those instances where a specific hour was recorded but also those where deponents referenced a broader ‘phase of the day’. Using instances where deponents deployed both an hour and a phase, such as ‘at seven of the clock in the evening’, it is possible to determine which hour periods each phase statement generally referred to. Whilst there was some variation in usage, five key phases emerged from this analysis, and Table 4.7 shows the distribution of 2,313 work tasks across them. The pattern is similar to the hours analysis: the majority of work, at 62.3 per cent, is concentrated between the ‘peak’ work periods of morning and afternoon, 6 am to 6 pm, but a significant proportion also took place during ‘off-peak’ hours, especially during the evening.
Table 4.7 Long description
The table presents the temporal distribution of daily tasks across five time periods. The first column lists time segments that are early morning to night, followed by two data columns, the number of tasks recorded and their percentage share of total daily activities. The data reveals how work patterns fluctuate throughout the day.
For early morning, 3 to 6 a m, the corresponding values are 130 and 5.6.
For morning, 6 to 12 a m, the corresponding values are 748 and 32.3.
For afternoon, 12 to 6 p m, the corresponding values are 694 and 30.
For evening, 6 to 8 p m, the corresponding values are 343 and 14.8.
For night, 8 p m to 3 a m, the corresponding values are 396 and 17.2.
For total, the corresponding values are 2,311 and 99.9.
Notes: Integral excluded; F adjusted (x2.22). Differs from standard multiplier as it is designed to give an equal number of male and female tasks with phase-of-day data attached.
What tasks took place when? In the summer months, the most common subcategory of evening task was animal husbandry, with 24 tasks, much of which involved checking on or moving animals that were out to pasture. In September of 1699, William Rowsell and his wife went ‘in the evening’ from North Curry, Somerset, to ‘a moor called West Sedgemoor’ and ‘fetched out thence six ewe sheep and one weather sheep and drove them home’.Footnote 79 Other common tasks included fieldwork, with 20 tasks, and milking, with 16, which could be undertaken in the longer and warmer evenings. In the winter none of those activities featured prominently. At that time of year the top subcategory was carry goods, with 16 tasks, which included bringing home goods from market, taking wheat to the mill, and collecting hay and coals from nearby stores. These tasks are a reminder that the movement of goods was a time-consuming activity at all times of year, and was also one that could be undertaken in low light when necessary. Other activities that needed completing all year round also feature in the top five winter evening subcategories: locking doors and gates, food and drink provision, collecting water, and light and fire provision. The Cheshire servant Alice Smith would have known only too well that these housework tasks were not put on hold in the dark days of winter. When a man and woman arrived to lodge at her dame’s Bowdon alehouse, one January evening in 1662, they first required food to be prepared, and then Smith did ‘light a candle’ for them, showed them to a bed chamber, and when they called for drink a short while later ‘she went and brought them some’.Footnote 80 Similar tasks would have been carried out on evenings in all seasons across England’s alehouses and domestic homes alike.
Whereas activities such as droving, carrying goods, and housework explain the busy evening work period, it is more surprising to find that as many as 17.2 per cent of tasks took place during the night. This is not simply a function of using depositional material that could often relate to nocturnal wrongdoing, as this number excludes integral activities that were closely linked to criminal activities. A closer look at subcategories is helpful here too and demonstrates that there were in fact many legitimate forms of work that could take place at night. Food and drink provision constitutes the largest subcategory, with 46 tasks. Much of this involved the preparation of food and serving of drinks to paying customers in the alehouse world of commercialised housework, where publicans clearly did not adhere to the 9 pm closing time stipulated in legislation from the period.Footnote 81 Animal husbandry, with 38 tasks, featured prominently, especially feeding. John King, a yeoman of Bishopstone, Wiltshire, was by no means unusual to be up at night ‘about 12 of the clock to feed his horses’ in October of 1661.Footnote 82 He did so as he had ‘a journey the next morning’, and this was often a reason to make sure a horse was well fed at night: the Norfolk woolcomber John Thacker borrowed a horse in 1679 and was given by its owner ‘a lock of barley to give his horse that night before he did take his journey’ the following morning.Footnote 83 Journeys to fetch or carry goods were another common activity that took place at night. For those transporting yarn, clothes, and victuals on foot to and from suppliers, producers and consumers, an early start or a late finish to the working day was often required to fit in these time-consuming travels. Healthcare had a strong nocturnal dimension to it too. Attending the sick, dressing wounds, or giving medicine were all tasks that could be required at any time of the day or night, and very frequently the small hours were a time when those on their sickbed were especially vulnerable.Footnote 84 The prominence of both evening and night-time work in our data is not a ‘fiction of the archive’ created by using depositional material; rather, it demonstrates that when a broad definition of work is adopted, it becomes clear that the temporal rhythms required by many everyday tasks did not neatly align with daylight hours.
This was true for some forms of work more than others, and by grouping the various phases of the day into ‘peak’ (6 am–6 pm) and ‘off peak’ (6 pm–6 am) hours, it is possible to compare the temporal characteristics of different work categories, as shown in Table 4.8. Work most closely associated with craftsmen and artisans – crafts and construction, and the food processing of butchers, maltsters, and millers – had the highest concentration of work tasks within peak hours. In many respects these workers are the closest to modern industrial workers, often keeping set hours in their workshops so customers would know when and where to find them. But categories such as housework, and in particular carework, were – as has been suggested above – much less bound to those peak hours. These patterns can also be identified by looking at the distribution between peak and off-peak work for broad occupational groupings, as in Table 4.9.
Table 4.8 Long description
The table analyzes temporal work patterns across different occupational categories. The first column lists labor types, followed by three data columns, total task counts, percentage of tasks performed during peak daytime hours from 6 a m to 6 p m, and percentage during off-peak night time hours from 6 p m to 6 a m.
For Agriculture and land, the corresponding values are 725, 67.1, and 32.9.
For Care work, the corresponding values are 230, 40, and 60.
For Commerce, the corresponding values are 141, 65.8, and 34.2.
For Crafts and construction, the corresponding values are 174, 77.1, and 22.9.
For Food processing, the corresponding values are 86, 81.9, and 18.1.
For Housework, the corresponding values are 504, 58.3, and 41.7.
For Management, the corresponding values are 86, 58.4, and 41.6.
For Transport, the corresponding values are 336, 62, and 38.
For Other, the corresponding values are 29, 49.4, and 50.6.
For Total, the corresponding values are 2,311, 62.4, and 37.6.
Notes: Integral excluded; F adjusted (x2.22). This differs from the standard multiplier as it is designed to give an equal number of male and female tasks with phase-of-day data attached.
Table 4.9 Long description
The table examines temporal work distributions across gendered occupational categories. The first column lists specific job types with gender markers, male or female, followed by task counts and their percentage allocation between peak daytime at 6 a m to 6 p m and off-peak night time at 6 p m to 6 a m work period. The data reveals distinct timing patterns across different labor roles and genders.
For agricultural occupations, male, the corresponding dataset is 167, 56.9, and 43.1.
For craft or trade occupations, male, the corresponding dataset is 208, 60.6, and 39.4.
For labourers, male, the corresponding dataset is 163, 73.6, and 26.4.
For servants, male, the corresponding dataset is 98, 67.3, and 32.7.
For servants, female, the corresponding dataset is 109, 45.9, and 54.1.
Notes: Integral excluded; F adjusted (x2.22). This differs from the standard multiplier as it is designed to give an equal number of male and female tasks with phase-of-day data attached.
The experiences of male and female servants were very different: whilst the former may have had a certain degree of leisure time after the peak hours of the day, for the latter this was often when the working day was at its most demanding. Craftsmen and tradesmen had a stronger concentration of ‘peak’ hours than agriculturalists, but the most striking figure here is that of labourers, who have the highest percentage of ‘peak’ work. If husbandmen had long hours of fieldwork and animal husbandry contributing to their off-peak work burden – more so than craftsmen – it may well be the case that labourers, with a lack of land and animals, had fewer tasks to attend to outside of the hours during which they could find paid day work. That said, paid work should not be conflated with work undertaken during the ‘peak’ hours of the day. For men, ‘for another’ work was concentrated in peak hours, at 62 per cent, but this still left 38 per cent of work done for others outside the household taking place in the evening and at night. For women, ‘for another’ work was actually more likely to take place during ‘off-peak’ hours, at 54 per cent, than during peak hours, at 46 per cent, reflecting the fact that housework and carework – even when taking place in the evening or at night – were very often undertaken by women for pay.Footnote 85
Taken together, the work-task data shows that the rural working day was not entirely irregular or erratic. There were very clear ‘peak’ working hours during which the majority of work tasks took place. But there was not one uniform temporal pattern for all types of work and workers; men, and especially those engaged in artisanal or labouring work, performed a higher percentage of their work during ‘peak’ hours, whereas the labour of women was spread more evenly around the clock. Once again, these findings demonstrate the importance of an inclusive definition of work, and of decentring the experience of male artisans when understanding the full range of experiences of work in early modern England. What does appear to have been a unifying experience though is the undertaking of long hours of work. Even for groups of workers with a high concentration of their tasks in ‘peak’ hours, significant levels of activity in the evening and during the night were recorded. Even when all housework and carework are excluded from the analysis, and the overall totals of female tasks are not adjusted, adjustments that would bring our data more into line with the definition of ‘labour force participation’, the proportion of tasks undertaken during ‘peak’ hours only rises to 69 per cent, as opposed to 62 per cent in the more inclusive and adjusted data used above. In short, undertaking work tasks during ‘off-peak’ hours was common practice.
This has implications for debates about rising work intensity in our period. The work-task data, as discussed in Section 1.4, does not lend itself to the straightforward analysis of change over time within the period, and attempts to do so on this issue reveal no clear patterns. But there does not appear to have been significant scope for early modern rural workers to abandon a ‘leisure preference’ in order to work longer hours, as the working day was already long and arduous. That is not to say that a shift to different types of work might not have been a factor, of course, but there is little evidence in our data that rural workers lacked day-to-day industriousness in the first place.
4.5 Conclusion
Early modern work hinged upon seasonal, weekly, and daily rhythms, yet robust empirical evidence for these has often been lacking in histories of labour and economic change. The work-task approach goes some way in filling this gap, providing a mixture of reassuring results, novel findings, and overt challenges to received wisdom on the topic. The monthly distribution of tasks demonstrates familiar patterns of seasonality, with intense work in summer characterised particularly by harvest fieldwork, carting, and construction. Yet winter was not a season of idleness or markedly reduced work hours, as has often been reflexively assumed. Indeed, it was the major season of activity for certain sectors of the economy like carework, which responded to the excess of births and deaths in late winter, and commerce, which pivoted around the Christmas festival. The fact that women played key roles in such winter labour, often income-generating or paid, may explain its relative neglect in a literature that has prioritised the experience of the male wage worker. Certainly, winter commerce was a seasonal activity shared between men and women, and across occupations, in a working year often bifurcated according to identity. More well-known seasonal tasks such as spring weeding, June sheep shearing, and July haymaking also brought the genders together, but the grain harvest was unrivalled in mobilising men, women, children, husbandmen, and artisans alike.
While the seasonality of labour varied substantially according to sector, gender, and occupation, the working week did not. Men of most occupations and women observed fairly regular weekly schedules. Surprisingly, considering the emphasis placed on St Monday in the literature, Monday was the busiest day in terms of observed tasks. The other end of the week saw the commercial dominance of Saturday, the favoured weekday for markets in all three regions of our sample. Points of difference emerge, more intriguingly, around Sunday labour. Rather than a pronounced ‘day of rest’, the data reveals only a moderate decline in activity, comparable to other midweek days. This drop came in those sectors most targeted by Sabbatarian restrictions, like commerce, transport, or craftwork, while church-going commitments and recreation reduced the number of hours available for work tasks. Yet much sabbath work was done for others outside the household, suggesting that paid labour did not cease on the Lord’s Day. Nor did it cease on holy days. Such festivals fell somewhere between Sundays and ‘workdays’ in terms of task repertoires, being especially flush with commercial activity linked to fairs and conviviality.
The working year which emerges from these findings is a complex spectrum rather than a binary model of workdays and ‘days off’. The latter has underwritten many narratives of economic change, predicated upon longer workweeks and workdays, or fewer holidays. But such a formulation fundamentally misconstrues how early modern worktime was navigated in practice, undercounting the economic contributions of certain groups in the process. It was servants and women, for example, who were more likely to work on Sundays, holy days, and during off-peak hours, though they were not alone in doing so. Such blurred boundaries between work and leisure time do not reflect a lax or erratic work regime, as some histories have painted the preindustrial economy. Instead, the work-task approach has revealed a relatively stable early modern working day which, like the workweek, was long, and altered minimally across sectors and seasons. These results reflect an inclusive approach to the definition of work, which arguably gets much closer to reality than a fixation on day labouring for wages. The varied experiences thus captured allow a more grounded understanding of the nexus between work, industriousness, and the rhythms of everyday life in early modern England.
Open access