CHAPTER 5

Guilds and Women

A [master’s] daughter . . . may not be a master unless she has been a master’s wife . . . because girls were leaving their fathers and mothers, and beginning to practise their craft, and taking on apprentices, and doing nothing other than leading debauched lives.

—Beltmakers’ guild ordinance, Paris, c. 1270

Be it ordained . . . that no woman or maid weave any worsted etamines or says, for that they be not of sufficient power to work the said worsteds as they ought to be wrought.

—Worsted-weavers’ guild ordinance, Norwich, 1511

Brewing is a learned art and given to men alone.

—Brewers’ guild petition against widowed brewer, Munich, 1599

Women are incapable of lifting the weights involved in our trade.

—Wool-weavers’ guild statute, Rome, 1758

A woman left to herself cannot be in a position to continue on her own to practise a trade, to which only a man, and only after much prior knowledge and examination, has been found eligible.

—Combined guilds of Vienna, complaint against widows running workshops, 1808

How did guilds treat women? This question is important. For one thing, discrimination against women redistributes resources from females to males, reducing the well-being of half the population. In addition, evidence from modern developing economies provides strong reason to think that, independently of the distributional reasons to be concerned about it, discrimination against women inflicts large economic costs on society as a whole. As we shall see in the final section of this chapter, preventing females from investing in their own human capital not only reduces women’s own earnings but lowers per capita GDP.¹ Likewise, excluding women from the labour force as workers, business-owners, and entrepreneurs not only prevents women from supporting themselves and their families, but also slows economic growth.²

The developing economies of medieval and early modern Europe had characteristics that make it even more important to examine how guilds affected women. Across wide swathes of the European continent, women married late, many never married, fertility was low, families were small, mortality was high, and widowhood was early and lasting. This meant that females often had to earn a living outside the household.³ Against this background, constraints on female human capital investment, labour-force participation, and entrepreneurship had an even stronger effect on women’s well-being. Women’s physical endowments made them on average less productive in agriculture but more productive in crafts and commerce, so institutional constraints on women in the secondary and tertiary sectors, where guilds chiefly ruled, were particularly important. Finally, many scholars argue that women’s position was positively associated with pre-modern economic performance.⁴ For all these reasons, how guilds affected women is central to assessing how they affected the whole economy.

This chapter finds that guilds treated females and males in fundamentally different ways. Most guilds restricted women’s training, excluding them from apprenticeship and journeymanship on the basis of gender alone. Most guilds used the same justification to prevent women from becoming masters, and thus from being business owners and self-employed entrepreneurs. Guilds usually allowed a master’s widow to continue the family workshop, but only if she satisfied certain conditions and limited her business in various ways. During the lifetime of the male master, many guilds restricted the work of his wife, daughters, female relatives, maidservants, and female employees. Guilds thus used their entry barriers and market powers to restrict females’ training, work, and entrepreneurship.

Some scholars have nonetheless claimed that guilds did not actually harm women. Some base this idea on the fact that certain guilds were all-female, mixed-sex, or granted widows’ rights.⁵ Others contend that guilds, in restricting women’s work and training, simply reflected the biological realities that made it efficient for women to specialize in domestic work.⁶ The few girls who wanted to learn occupational skills could do so within the family or use other training institutions, it is argued, and the few women who wanted to practise crafts and trades could work in the informal sector.⁷ Furthermore, according to this line of scholarship, guild discrimination against women did not matter because it was only a reflection of wider cultural attitudes which would have had the same economic impact without guilds.⁸ Even if guilds did limit the options of a few independent females, the argument continues, it did not harm wider economic performance, based as it was on the “family economy” in which women shared in the business lives of husbands and fathers.⁹ For all these reasons, according to this view, there was no incompatibility between a good economic position for women and a good economic position for guilds.¹⁰

These arguments raise five fundamental issues. First, did guilds in fact create a good economic position for women, despite or even because of gender-based rules? Second, is it biology rather than institutions that determines economic behaviour and performance? Third, can good black markets replace bad formal institutions? Fourth, are cultural attitudes, not institutions, the key to economic performance? Finally, does women’s status truly matter for the economy as a whole?

This chapter addresses these questions by looking at the way guilds treated women as independent entrepreneurs and business owners, as widows of deceased masters, as masters’ wives and daughters, as apprentices and journeywomen, as servants, and as freelance workers. As we shall see, women were not altogether excluded from the world of the guilds. Indeed, they did nearly every type of guild work permitted to them—and many kinds of work they were forbidden. But seeing women doing guilded work does not mean that guild constraints were absent or ineffectual. It has become popular to focus on examples of women who managed to learn and practise guilded occupations, and to view them as demonstrating that female agency was so strong that institutional constraints didn’t matter—that “there is no essential contradiction between women and guilds”.¹¹ It is undoubtedly true that there were individual women who managed to achieve eminence in their vocations, including guilded ones, and to deny this would be to paint a false portrait of pre-modern economies. But to set these cases in an accurate perspective, we need to examine how the economy and its institutions worked in aggregate, for everybody. This chapter seeks to understand how guilds affected not just a few fortunate individuals, but the vast mass of females who struggled to make a living for themselves and their families in economies where much work was shaped by guild rules.

A GENDER TYPOLOGY OF GUILDS

Almost every guild regulated women’s work. Table 5.1 presents a schematic typology of European guilds according to how they treated females. The columns comprise the five categories proposed by Ariadne Schmidt in her gender typology of guilds.¹² The rows show measures of how each guild type treated women in different roles.

Type 1 are all-female guilds, in which women comprised all the masters, journeymen, apprentices, and wage workers (though not usually the guild officers). Type 2 are mixed-sex guilds, in which both females and males could participate in all roles (except usually as guild officers). Such guilds had two variants: in the first, male and female masters exercised the same occupation; in the second, their work differed (e.g., male tailors, female seamstresses). Type 3 are guilds with widows’ rights, in which a guild member’s widow could continue operating the workshop using her deceased husband’s license, but could not be a master in her own right, a journeyman, an apprentice, or a guild officer. A rare variant of this type granted similar rights to daughters. Type 4 are guilds that granted no widows’ rights while still letting women do some jobs in the occupation. Type 5, finally, are guilds that excluded all participation by women.

Type 3 was the most widespread variety, a guild that let a widow continue to operate her deceased husband’s workshop, usually with limitations, but not be a master in her own right. Types 1 and 2, in which women could be masters in their own right, were rare: gender was a visible and culturally acceptable criterion for limiting entry, and most guilds used it. Types 4 and 5, where women could do few or no forms of work, were also rare: women were productive in most occupations, and guild masters profited by employing them, especially since it was easy to use guild regulations to cap women’s wages or squeeze them out if they posed too great a competitive threat.

TABLE 5.1: Gender Typology of European Guilds
	Type 1	Type 2	Type 3	Type 4	Type 5
	All masters are female	Masters can be either male or female	No female masters, but widows can continue workshop	No female masters, no widows’ rights, some female workers	No female masters, no widows’ rights, no female workers
Prevalence in Europe
No. observations in guilds database	55	343	numerous	52	45
Estimated % of total European guilds	0.06%	0.38%	c. 99%	< 0.06%	< 0.06%
Summary assessment of prevalence	very rare	rare	dominant	very rare	very rare
Proportion of females among:
Guild officers	very few	very few	< 1%	0%	0%
Independent masters	100%	c. 22%	0%	0%	0%
Masters’ widow(er)s running workshop	a few	a few	8%	0%	0%
Journeymen	100%	minority	0%	0%	0%
Apprentices	100%	4–13%	very few	0%	0%
Wage or piece-rate workers	majority	some	some	some	0%
Master’s spouse may work	usually	usually	usually	seldom	never
Master’s daughters may work	usually	usually	sometimes	seldom	never
Master’s maidservants may work	usually	usually	seldom	seldom	never
Master’s other female relatives may work	rarely	rarely	rarely	never	never
Females outside masters’ households may work	never	never	never	never	never
Source: Typology based on Schmidt 2009, 173. Percentages based on analyses in this chapter.

This gender typology has limitations. There were mixed types, such as guilds that let masters’ widows run workshops (as in Type 3) but simultaneously banned some or all female wage-workers (as in Type 5). There were also finer gradations. Within Types 1 through 4, as we shall see, some guilds let wives, daughters, and maidservants participate in the craft workshop, many restricted them, and some banned them entirely. Some guilds shifted from one type to another, responding to a downturn by ejecting females, then turning a blind eye to women’s work during a boom.¹³ But this typology, though simplified, provides a stylized framework which can be fleshed out with observations from the guilds database and used to explore the questions posed in the introduction to the chapter.

The first proposition it helps us explore is the idea that the European guild system did not actually restrict women’s economic activities since it included all-female and mixed-sex guilds in which female masters could freely operate workshops.¹⁴ The guilds database contains 3,495 observations of guilds’ treatment of women, 2,503 qualitative cases and 992 quantitative ones. The top panel of Table 5.1 shows what these reveal about the prevalence of different types of guild. Examples of all five categories in the gender typology can certainly be found. But all-female and mixed-sex guilds (Types 1 and 2) were extremely uncommon. Examining these very unusual guild types can shed a small amount of light on how guilds treated women, but only in so far as exceptions illuminate a rule.

All-Female Guilds

All-female guilds existed, but were vanishingly rare. The guilds database covers 23 European societies for over eight and a half centuries, between c. 1000 and c. 1860, but contains just 55 observations of exclusively female guilds.¹⁵ Historians have fully investigated the documentary sources for a number of European societies—England, Hungary, Italy, Norway, the Southern Netherlands, and the German territories of Bremen, Hamburg and Württemberg—and definitively concluded that they never had even one all-female guild; for other societies, historians have not fully investigated the sources, but have also found no evidence of any all-female guild.

Indeed, as Table 5.2 shows, all-female guilds have been found in only eight European societies: France, Germany, Poland, the Northern Netherlands, Switzerland, Spain, Ottoman Thessaly (modern Greece), and Denmark. France was home to 60 per cent of the all-female guilds ever recorded, significantly more than its 16.5 per cent share of the overall guilds database.¹⁶ Outside France, Germany, and Poland, only ten all-female guilds have ever been recorded. In the countries that had them, all-female guilds clustered in just a few places: Paris contained one-third of the all-female guilds in France, Cologne two-thirds of those in Germany, Zürich all of those in Switzerland.

TABLE 5.2: All-Female Guilds by Country, Sector, and Time-Period, 1268–1789
	Medieval only	Medieval and early modern	Early modern only	Whole period
				no.	%
Country:
France	6	7	20	33	60.0hh
Germany	3	2	1	6	10.9ll
Poland	1	0	5	6	10.9hh
N. Netherlands	0	0	4	4	7.3s
Spain	1	0	1	2	3.6s
Switzerland	1	0	1	2	3.6s
Greece	0	0	1	1	1.8hh
Denmark	0	0	1	1	1.8s
Sector:
Clothing	3	3	14	20	36.4
Textile	8	3	4	15	27.3
Textile/retail	0	3	3	6	10.9
Food/retail	0	0	5	5	9.1
Retail	1	0	2	3	5.5
Miscellaneous	0	0	6	6	10.9
Total no.	12	9	34	55	100.0
Total %	21.8	16.4	61.8	100.0
Significance	s	hh	l
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. Time-period: period during which each guild is mentioned in sources. Poland: although not formally guilds, these 6 guild-like associations are included since they had monopoly privileges embodied in statutes; see Karpinski 1989, 288. England and Italy have zero observations in this table, significantly lower (at the 0.05 level) than their percentage of observations in the guilds database overall. The Southern Netherlands has zero observations, significantly lower (at the 0.10 level) than its percentage of observations in the guilds database overall. For all other countries and centuries not represented in this table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Qualitative guilds database: 55 observations of all-female guilds.

Even in these hotspots, all-female guilds were a very small minority, as Table 5.3 shows. Their share of all guilds ranged from 1 to 2 per cent in Nuremberg and Madrid to between 9 and 10 per cent in Kraków and Cologne. Even in Paris, the acknowledged European capital of all-female guilds, they made up only 3 to 4 per cent of the total at any time. The average across all ten towns in Table 5.3 was only 5.4 per cent. And these were towns that were unusual in having at least one all-female guild.

TABLE 5.3: All-Female Guilds as a Proportion of Total in Cities with All-Female Guilds, 1268–1789
Place	Country	Period	Total guilds	No. all-female guilds	% all-female guilds
Lyon	France	1789	72	5	6.9
Paris	France	1268	100	4	4.0
Paris	France	c. 1750	120	4	3.3
Rouen	France	18th century	70	5	7.1
Cologne	Germany	15th century	45	4	8.9
Nuremberg	Germany	1597	95	1	1.1
Amsterdam	N. Netherlands	1579	16	1	6.3
Gouda	N. Netherlands	1661–4	27	1	3.7
Kraków	Poland	16th century	50–55	5	9.1–10.0
Madrid	Spain	17th century	58–100	1	1.0–1.7
Zürich	Switzerland	14th century	13	1	7.7
Mean					5.4
Note: Kraków: the 5 associations of small-scale female retailers were not officially guilds, so it is unclear whether they were included in, or additional to, the estimate of “approximately 50 guilds” in sixteenth-century Kraków. Source: Quantitative guilds database: 11 observations of all-female guilds as % of total guilds.

To set these figures in context, Table 5.4 presents available estimates of the number of guilds in existence in the “long eighteenth century” (c. 1700–c. 1810) in France, Italy, Germany, the Northern Netherlands, the Southern Netherlands, and Spain. Together, these six societies account for 77 per cent of observations in the guilds database and 63 per cent of the population of the 23 societies in the guilds database at this period.¹⁷ The estimates of guild numbers in Table 5.4 vary in their derivation and their quality, and for France, Germany and Spain are based on contemporary reports which may be inaccurate. But for all their weaknesses, they are the best available in the literature and yield a figure of between 78,000 and 98,000 guilds for the six societies taken together. The 45 all-female guilds observed in these societies throughout their entire history thus amount to less than one-tenth of one per cent of their guilds. Even for the Northern Netherlands, with its precise modern counts of total guild numbers and its legendarily advantaged status for women, all-female guilds comprised less than half of one per cent of guilds. In a quantitative perspective, female guilds were of vanishingly small importance.

TABLE 5.4: All-Female and Mixed-Sex Guilds as Percentage of Total Guilds, Six European Countries
			All-female guilds		Mixed-sex guilds
Country	Period	Total guilds	no.	%	no.	%
France	c. 1750	20,000	33	0.17	64	0.32a
Germany	1810	30,000–50,000	6	0.02	147	0.49b
Italy	1700	998	0	0.00	35	3.51c
N. Netherlands	1784	1,078	4	0.37	28	2.60d
S. Netherlands	1784	836	0	0.00	24	2.87d
Spain	1787–8	25,581	2	0.01	3	0.01e
Total 6 countries	18th c.	78,493–98,493	45	0.06	301	0.38
Notes: Percentages calculated using lower-bound estimate (30,000 for Germany and 78,493 for all 6 countries), so as to obtain a maximum estimate of the percentage of all-female and mixed-sex guilds. Sources: aForbonnais 1758, I:478. bKluge 2007, 62; Germany boundaries of 1810. cMocarelli 2008, 162, 165 (Table 2); covers cities with over 10,000 inhabitants in at least 3 of the 6 dates 1300, 1400, 1500, 1600, 1700, and 1800. dDe Munck, Lourens and Lucassen 2006, 37 (Table 2.1): calculated as total formed from 1100 to 1784, minus total disappeared before 1784; covers towns over 2,500 inhabitants. eWilliams 1907, 3:253; Castile and Aragon only.

Up to now, historians knew that all-female guilds existed, but had no precise numbers. This led some to conclude that all-female guilds, like other forms of female independence, were prevalent in the Middle Ages and only died out after c. 1500, consonant with the idea of early modern “guild decadence”.¹⁸ Now that we have numbers, we can test this conjecture. The medieval period accounts for just 20 per cent of all-female guilds in Table 5.2, not significantly different from its 27 per cent share of the overall guilds database. The medieval period was no golden age for all-female guilds.

A surprising number of the few all-female guilds that have ever been recorded were very short-lived. There were four all-female guilds in Paris in 1268, but they vanished within a century.¹⁹ There were four in late medieval Cologne, but they were gone by the early sixteenth century.²⁰ The single Danish all-female guild survived for just six years (from 1534 to 1540).²¹

All-female guilds clustered in a very few sectors, as Table 5.2 shows. Clothing accounted for over one-third of the total, textiles for over one-quarter, and retailing for another quarter. Only six all-female guilds existed outside these categories, and of these three were in occupations related to personal adornment, clothing, or retailing. Only the guilds of the female barge-pullers in Utrecht and Lyon, and the guild of the Trikala soap-makers, lie outside this pattern.

All-female guilds were typically formed either by the powerful and wealthy practitioners of luxury trades, or by poor and powerless practitioners of trades no males cared to practise. In medieval Cologne, for instance, the all-female guilds comprised women from wealthy merchant families, were established during a guild takeover of town government after 1397, and were sustained by their members’ strong links with the stratum that formed the post-revolt Cologne political elite.²² On the other end of the spectrum were low-status occupations such as the veil-makers of fifteenth-century Nördlingen,²³ the fish-sellers of sixteenth-century Malmö,²⁴ and the all-female guilds of eighteenth-century Rouen in which average taxable income was less than half that in male guilds.²⁵ Whether their members were rich or poor, the vast majority of all-female guilds had male officers, and the very few female guild officers lacked the powers of their male counterparts.²⁶

Finally, all-female guilds granted a good economic position only to women who were members of the guild. All-female guilds, like all-male ones, imposed entry barriers, manipulated markets in their members’ favour, and persecuted women (and men) who tried to work illicitly in violation of guild privileges. In 1504, for instance, the all-female old-clothes-drapers’ guild in Rouen prosecuted Perette Guelle because “even though she is not a mistress or sworn worker of the trade in Rouen she attempted every day to meddle and put to work in the said trade a girl under her, a thing she cannot and should not do”.²⁷ In 1530, the same guild penalized three named females and “other women of this town” who wanted to go on sewing and selling linen clothing without being members of the guild.²⁸ In 1565 it prosecuted and fined a non-guilded woman for selling old cloth when “she is not a mistress of their trade nor is she a sworn reseller”.²⁹ The all-female guild of the Rouen new-linen-drapers likewise mounted repeated prosecutions against illicit production and sale by women who were not guild members. In 1566, when two poor women were discovered illicitly selling offcuts of linen in violation of the guild’s exclusive privileges, the guild let them off the money fine “having regard to the poverty of these women”, but warned that re-offending would bring them a whipping.³⁰ All-female guilds did not create economic benefits for all women any more than all-male guilds created benefits for all men. Both types of guild secured exclusive privileges and artificially high cartel profits for their own members, at the expense of outsiders.

All-female guilds were thus extremely rare. They comprised a tiny percentage of all guilds even in towns where they existed, and a vanishingly small share of guilds overall. They were restricted to a narrow range of sectors and occupations. They were seldom administered by women themselves and did not give female producers the degree of control over their industries that other types of guild gave male producers. Against this background, they cannot be adduced in support of the view that the European guild system was open to female entrepreneurs and business-owners. There is no sign that all-female guilds reflected or created a wider economic environment that was favourable to women.

Mixed-Sex Guilds

What about the Type 2 guilds that let both women and men be masters on their own account? Establishing the prevalence of mixed-sex guilds is difficult because documents recording females in guilds often fail to provide enough information to ascertain that women were truly independent masters. For one thing, guilds are sometimes defined as mixed-sex if they record the possibility of female apprentices, even if there is no information about whether girls received full craft training or were allowed to become masters later.³¹ Second, guilds are sometimes defined as mixed-sex if they recorded the possibility of a master’s daughter following her father’s occupation, even if there is no information about whether daughters merely inherited the workshop premises, enjoyed social and religious but not economic rights, or lost their rights on marriage (in which case they were a minority variant of widows’-rights guilds).³² Third, guilds are sometimes defined as mixed-sex if they record the possibility of females doing particular craft tasks or if they use particular occupational designations in the feminine form, even if such activities were auxiliary tasks, involved retailing employers’ wares, and did not imply mastership.³³ Fourth, guilds are sometimes defined as mixed-sex if they recorded the possibility of married women holding guild membership, even if there is no information about whether such women had to be married to guild masters, or enjoyed the social and religious benefits of guild membership but no economic rights.³⁴ Fifth, guilds are sometimes defined as mixed-sex if they record the possibility of “mistresses” as well as “masters”, even though the term “mistress” (and its cognates in other European languages) could mean not just an independent female master of the guild, but also a master’s wife whether or not she was active in the craft, or (in its German form) a master’s widow (in which case the guild was merely a normal Type 3 guild).³⁵ Sixth, guilds are often defined as mixed-sex if they recorded “guild sisters” alongside “guild brothers”, even if there is no information about whether these “sisters” were masters’ widows, wives, or daughters, and whether they enjoyed economic as well as socioreligious membership.³⁶ Seventh, guilds are often defined as mixed-sex if their ordinances record female pronouns alongside male ones, even when they provide no information about whether the females in question actually shared in the economic rights of guild membership or whether the pronouns had simply been recopied from an earlier ordinance or a different guild. Eighth, guilds are sometimes defined as mixed-sex if documents use names or occupational designations in the feminine form, even though in languages such as German such terms were used not just to refer to independent female practitioners of an occupation, but to masters’ wives who worked alongside their husbands, masters’ widows who exercised widows’ rights, or women whose surname derived from an occupation even if no-one in the family practised it.³⁷ Ninth, guilds are sometimes defined as mixed-sex if they record females obtaining an exception or partial permit to practise an activity otherwise monopolized by an all-male guild.³⁸ Finally, guilds are sometimes defined as mixed-sex if they recorded the possibility of female members who practised separate or more limited work than the male members of the guild: selling but not making bread,³⁹ making veils but not other fabrics,⁴⁰ embroidering or seamstressing but not tailoring.⁴¹

The qualitative guilds database contains 343 observations of guilds that have been identified as mixed-sex, by any scholar on any grounds. Of these, nearly 17 per cent are guilds for which all that is known is that they gave some rights to masters’ daughters (11 per cent), wives (5 per cent), or female relatives (1 per cent), which means that they are not sufficiently demonstrated to be mixed-sex guilds. A further 39 per cent consist of cases in which the guild might be mixed-sex, but the evidence is quite ambiguous. Another 6 per cent comprise guilds that were probably mixed-sex, although there is some uncertainty about the status of female members. Only 38 per cent of the observations refer to guilds that were definitely mixed-sex, following Ariadne Schmidt’s generous definition in including guilds which allowed female members to practise one occupation and male members another, rather than Kurt Wesoly’s which requires female masters to have practised the same occupation as male ones. Less than half of the 343 observations in the guilds database, therefore, show some certainty of being “true” mixed-sex guilds.⁴² Yet even if we were to take all 343 observations as definitive mixed-sex guilds, they would still account for a very small proportion of all guilds: just 0.4 per cent, according to the figures in Table 5.4.

Like all-female guilds, mixed-sex guilds were clustered in particular places. Table 5.5 shows a number of European cities for which we know both the total number of guilds and the number that were mixed-sex. Since most places had no mixed-sex guilds, the cities in Table 5.5 are ones with an unusually high number of them. Even in those places, however, mixed-sex guilds were in a small minority. The highest proportions were in medieval Paris and Ghent, where mixed-sex guilds made up 10 to 12 per cent of all guilds. The lowest were in early modern Paris and Lyon with between 1 and 2 per cent. Even in this special sample, consisting of towns in which mixed-sex guilds were unusually numerous, they comprised on average just 7 per cent of all guilds.

Mixed-sex guilds were also clustered in particular societies, as Table 5.6 shows. The over-representation of Germany results from the fact that Helmut Wachendorf identified numerous mixed-sex guilds in late medieval and early modern Germany; Kurt Wesoly later established that in many cases, Wachendorf’s identification was either over-optimistic or wholly unjustified.⁴³ If the 114 German observations deriving from Wachendorf were to be excluded, Germany would be significantly under-represented in Table 5.6 relative to its share of the guilds database. The Northern Netherlands is also over-represented, with 8 per cent of observations of mixed-sex guilds compared to only 5.5 per cent of observations in the guilds database; this small but statistically significant difference chimes with other research studies suggesting that Dutch women enjoyed an unusually good economic position.⁴⁴

TABLE 5.5: Mixed-Sex Guilds as a Proportion of Total Guilds, Six European Cities, 1268–1789
Place	Country	Period	Total guilds	No. mixed-sex guilds	% mixed-sex guilds
Paris	France	1268	100	10	10.0
Paris	France	end of 17th century	124	2	1.6
Rouen	France	early modern period	112	10	8.9
Lyon	France	1789	72	1	1.4
Cologne	Germany	15th century	45	4	8.9
Ghent	S. Netherlands	14th century	61	7	11.5
Mean					7.0
Source: Quantitative guilds database: 6 studies of guild samples showing mixed-sex guilds as percentage of total guilds.

The observations in Table 5.6 span the 636 years from 1226 to 1862 but are concentrated in the medieval period, which accounts for 58 per cent of observations, significantly higher than its 27 per cent share of the overall guilds database. To some extent this too reflects Wachendorf’s over-optimistic identification of mixed-sex guilds in the late medieval period in which he specialized. But even if the 110 medieval observations derived from Wachendorf are excluded, the medieval period accounts for 38 per cent of the cases in Table 5.6, significantly more than its share of the overall guilds database. This is consistent with qualitative evidence suggesting independent female mastership was increasingly restricted during the early modern period, even in strongholds of female guild membership such as Paris and Cologne.

TABLE 5.6: Mixed-Sex Guilds by Country, Sector, and Time-Period, 1226–1862
	Medieval only	Medieval and early modern	Early modern only	Whole period
				no.	%
Country:
Denmark	–	1	2	3	0.9s
England	13	5	13	31	9.0s
France	27	1	36	64	18.7s
Germany	129	8	10	147	42.9hh
Hungary	1	–	–	1	0.3s
Italy	9	1	25	35	10.2ll
N. Netherlands	1	–	27	28	8.2hh
S. Netherlands	11	1	12	24	7.0s
Spain	2	–	1	3	0.9ll
Switzerland	6	–	1	7	2.0s
Sector:
Textile	48	4	34	86	25.1
Textile/retail	7	–	7	14	4.1
Clothing	30	3	24	57	16.6
Clothing/retail	2	–	1	3	0.9
Pure retail	14	5	28	47	13.7
Leather	25	–	2	27	7.9
Food	22	–	2	24	7.0
Food/retail	2	1	12	15	4.4
Medical	14	1	3	18	5.2
Metal	10	1	4	15	4.4
Miscellaneous	25	2	9	36	10.5
Miscellaneous/retail	–	–	1	1	0.3
Subset involving retail	25	6	49	80	23.3
Total no.	199	17	127	343	100.0
Total %	58.0	5.0	37.0	100.0
Significance	hh	hh	ll
Notes: Austria, Bohemia, Poland, and Sweden have zero observations, significantly lower (at the 0.05 level) than their percentage of observations in the guilds database overall. Bulgaria has zero observations, significantly lower (at the 0.10 level) than its percentage of observations in the guilds database overall. For all other countries not represented in this table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Qualitative guilds database: 343 observations of mixed-sex guilds.

The 343 mixed-sex guilds in Table 5.6 include a wider range of occupations than the 55 all-female guilds in Table 5.2, but the sectoral pattern is similar. Clothing and textiles are the largest sectors, comprising nearly half of mixed-sex guilds and three-quarters of all-female ones. Food-related occupations make up a similar share, around 10 per cent, of both mixed-sex and all-female guilds. Retailing comprises 22 per cent of both guild types. But mixed-sex guilds could also be found in leather, metal, and medical occupations, as well as in a wide array of miscellaneous activities. This is not surprising, since being a member of an existing mixed-sex guild merely required a woman to be able to do the work, whereas establishing an all-female guild also demanded a special sociopolitical context.

TABLE 5.7: Percentage of Female Masters in Mixed-Sex Guilds, 1244–1862
Country	Medieval	Early modern	Whole period	No. obs.
England	0.4	18.1	13.1	7
France	87.1	40.6	42.1	31
Germany	17.0	7.5	14.1	36
Italy	32.7	39.2	37.8	14
N. Netherlands	–	39.8	39.8	11
S. Netherlands	–	17.0	17.0	2
Switzerland	22.8	–	22.8	3
All countries	20.0	32.8	28.6	104
No. obs.	34	70	104
Notes: “Female masters” includes masters’ widows, masters’ wives, and masters’ daughters, alongside independent female masters. For 16 of the 104 observations, different categories of female are listed separately; across these 16 observations, there was an average of 14.5% independent female masters and 7.1% masters’ widows. Assuming 7.1% masters’ widows in the larger sample of 104 would imply that they had an average of 21.5% independent female masters. Source: Quantitative guilds database: 104 observations of percentage of female masters in mixed-sex guilds.

To what extent did mixed-sex guilds open the guild system to women? One measure is what share of their membership was female. Table 5.7 shows 104 observations of mixed-sex guilds for which we have records of the total percentage of female masters—i.e., independent ones plus masters’ widows. The average across all 104 observations is 28.6 per cent. For 16 of the observations, the documents distinguish different categories of female master, yielding 14.5 per cent independent female masters and 7.1 per cent masters’ widows; the latter figure is credibly close to the 8 per cent of masters’ widows observed in the large sample of widows’-rights guilds discussed in Table 5.8 below. Assuming that in Table 5.7 the larger sample of 104 mixed-sex guilds also had an average of 7.1 per cent masters’ widows, the remaining 21.5 per cent would have been independent female masters. This share is not tiny, but neither is it substantial. Mixed-sex guilds were certainly not characterized by equality between the sexes, even in numbers.

Nor did mixed-sex guilds typically grant female masters equal economic rights. A number of mixed-sex guilds forbade female masters to keep apprentices, or limited them to training girls.⁴⁵ A number of mixed-sex guilds limited female masters’ working space or equipment, as in sixteenth-century Chartres where the tailors’ guild forbade its one female master to keep an open workshop,⁴⁶ or in eighteenth-century Rome where the silk-weavers’ guild restricted the number of looms a female master could have.⁴⁷ Mixed-sex guilds often limited the goods and services female masters could provide: a number of weavers’ guilds restricted female masters to certain fabrics;⁴⁸ a number of tailors’ guilds limited women to remaking old clothes, sewing underwear, or making garments for women and children;⁴⁹ the Cologne fishmongers’ guild barred female masters from cutting or weighing fish for sale.⁵⁰ Guilds described such restrictions as essential for protecting male masters, as in fifteenth-century Strasbourg when the weavers’ guild barred female masters from making narrow cloths, “because the majority of the male weavers have to earn their living making narrow cloths”.⁵¹

Mixed-sex guilds also typically excluded female masters from guild decision-making, barring them from holding office,⁵² voting on guild issues,⁵³ speaking at assemblies,⁵⁴ attending assemblies at all,⁵⁵ or entering the guild house even on festive occasions.⁵⁶ A mixed-sex guild might let female masters do the work, but it seldom let them have a say on how the guild regulated that work.

Guilds with Widows’ Rights

Given the extreme rarity of all-female and mixed-sex guilds, women’s treatment in the European guild system must be assessed primarily on the basis of what is known of mainstream guilds. As Table 5.1 shows, the overwhelming majority of guilds were those of Type 3, which did not let women be masters on their own account, but did let a master’s widow continue to operate the workshop using her deceased husband’s guild license. Indeed, the fact that most guilds granted rights to masters’ widows is sometimes used to support the view that “women had considerable rights and a strong legal position in the guild system”.⁵⁷

TABLE 5.8: Widows as Percentage of Masters in Widows’-Rights Guilds, 1284–1862
Country	Medieval mean	Early modern mean	Whole period
			mean	no.	%
Austria	–	6.1	6.1	7	1.7s
England	8.6	35.0	15.2	4	1.0ll
France	2.3	8.7	7.6	52	12.6ll
Germany	5.3	9.9	9.3	165	39.9hh
Italy	–	3.7	3.7	17	4.1ll
N. Netherlands	0.7	6.4	6.1	40	9.7hh
Spain	15.8	7.6	13.1	3	0.7ll
Sweden	–	7.3	7.3	126	30.4hh
Total mean	5.2	8.3	8.0	414	100.0
Total no.	40	374	414
Total %	9.7	90.3	100.0
Significance	ll	hh
Notes: Bohemia, Bulgaria, Poland, the S. Netherlands, and Switzerland have zero observations, significantly lower (at the 0.05 level) than their percentage of observations in the guilds database overall. For all other countries not represented in this table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Quantitative guilds database: 414 observations of widows as percent of masters in widows’-rights guilds.

To explore this idea, a good starting point is to see how many widows there actually were in guilds. The quantitative database contains 414 observations of the share of widows in widows’-rights guilds. As Table 5.8 shows, these 414 observations reflect the situation in central, northwest, and Nordic Europe to a greater extent than other parts of the continent and are heavily weighted towards the early modern period, because sources were more plentiful after 1500. Nonetheless, they shed light on eight societies across nearly six centuries, and are drawn from over 150 different occupations. This large data compilation shows that widows comprised just 8 per cent of masters on average and that 40 per cent of these guilds had no practising widows at all, despite their formal guarantee of widows’ rights.

TABLE 5.9: Percentage of Female-Headed Households in European Towns, 1354–1861
Country	Medieval mean female headship	Early modern mean female headship	Whole period
			mean female headship	no. obs.	column %
Austria	–	19.5	19.5	5	2.4
Denmark	–	14.9	14.9	3	1.5
England	–	18.2	18.2	44	21.5
France	–	17.5	17.5	53	25.9
Germany	15.7	18.1	17.3	72	35.1
Greece	–	22.9	22.9	1	0.5
Italy	19.3	15.2	16.1	9	4.4
N. Netherlands	–	17.8	17.8	7	3.4
Norway	–	36.4	36.4	2	1.0
S. Netherlands	–	22.3	22.3	6	2.9
Spain	–	25.3	25.3	3	1.5
Total mean	16.0	18.3	18.0	205	100.0
Total no.	25	180	205
Total %	12.2	87.8	100.0
Source: Quantitative guilds database: 205 observations of percentage of female-headed households in towns.

It might be thought that the low percentage of widows among guild masters simply reflects a demographic regime in which few households were headed by widows. But this cannot be the explanation. Households headed by widows were much more common. The quantitative database contains 205 observations of the percentage of urban households headed by females—almost all of them widows—in eleven European societies spanning the five centuries from 1354 to 1861.⁵⁸ As Table 5.9 shows, female-headed households comprised an average of 18 per cent of the total, 16 per cent in the medieval period, and 18.3 per cent in the early modern period. For seven of the eleven European societies in Table 5.9, female headship lay in the 15 to 20 per cent range; the four exceptions were ones with very few observations.

This raises a question. On average, widows accounted for 18 per cent of household heads in European towns, but only 8 per cent of guild masters. This implies that many masters’ widows did not actually exercise their right to continue the workshop. Why? Investigating this question helps clarify the effect of guilds on women’s economic opportunities.

CONDITIONS ON WIDOWS’ RIGHTS

Widows were an embarrassing paradox inside the guild system. First, guild mastership brought sociopolitical status as well as economic rights. As the eighteenth-century German jurist Adrian Beier put it, “Ultimately the guild is a public institution; mastership rights are linked to the performance of services for the state, such as surveillance and monitoring, for which women are not suitable.”⁵⁹ Second, widows almost never underwent formal apprenticeship or guild examination.⁶⁰ Yet without such guild training widows practised the craft successfully—too successfully in the eyes of many male masters. Guilds typically dealt with this paradox by imposing conditions on whether a widow could put into practice her right to continue the workshop.

The qualitative database contains 312 observations of guilds imposing conditions on widows putting their rights into practice. These are shown in Table 5.10. They are drawn from over 120 occupations, ranging from low-skilled hucksters to high-skilled goldsmiths, from light crafts, such as mending old clothes, to heavy ones like making swords. Guild conditions on widows’ practice thus do not reflect assumptions that women were incapable of skilled or heavy work, but extended across the occupational spectrum. As Table 5.10 shows, Denmark, Estonia, Germany, and Norway are significantly over-represented among such cases relative to their share of the overall guilds database, while England, France, Italy, and the two parts of the Netherlands are significantly under-represented. Nonetheless, the practice was sufficiently widespread that guilds can be observed imposing conditions on widows’ rights in 15 different European societies. The observations in Table 5.10 span the five and a half centuries from 1268 to after 1800, with the medieval period significantly over-represented relative to its share of the wider guilds database, casting doubt on the widely held view that the Middle Ages was a golden age for female guild practice and that gender-related restrictions were mainly imposed by degenerate early modern guilds.

A first set of conditions, comprising over 40 per cent of the observations in Table 5.10, related to remarriage. Most guilds let a widow continue to practise only if she refrained from remarrying. If she remarried a man of a different occupation, she lost her right to continue the workshop; if she remarried a man of the same occupation, it was under his guild license that the workshop would continue.⁶¹ In very rare cases (just 9 of the observations in Table 5.10), a widow could go on working after remarriage if she complied with restrictive conditions such as paying an extra fee, not employing apprentices, not training her new husband, or not training her children; it was even more extraordinary for a widow to take her occupation with her into a new marriage without conditions. To place these remarriage conditions in context, there is no known case of a guild cancelling a widower’s license if he remarried a woman from outside the guild. In the 1260s the Paris poulterers’ guild justified this asymmetry by saying that “the man is not in the lordship of the wife but the wife is in the lordship of the man”, although the scepticism this evoked even at the time is shown by a marginal note retorting, “non pas tous jours” (“not always”).⁶² The result of this asymmetry was that a widow trying to continue the family workshop was disadvantaged compared to a widower in the same position, since a household normally relied on the work of at least two adults and was handicapped by having only one.

TABLE 5.10: Conditions on Widows Taking Up Guild Rights, 1268 to Early Nineteenth Century
	Remarriage	Time after widowhood	Age	Having children	Having son	Paying fee	Reputation	Approval	Total
									no.	%
Country:
Austria	–	3	–	–	–	1	–	1	5	1.6s
Denmark	8	6	–	–	–	–	2	1	17	5.4hh
England	6	1	–	–	–	–	–	–	7	2.2ll
Estonia	–	16	–	9	–	–	–	–	25	8.0hh
France	24	1	–	–	–	–	4	–	29	9.3ll
Germany	40	65	4	12	21	1	8	3	154	49.4hh
Hungary	1	–	–	–	–	–	–	–	1	0.3s
Italy	15	2	–	1	2	1	–	–	21	6.7ll
N. Netherlands	3	–	–	–	–	–	–	–	3	1.0ll
Norway	5	1	–	–	–	–	2	2	10	3.2hh
Poland	1	–	–	–	–	–	–	–	1	0.3l
S. Netherlands	1	–	–	–	–	1	–	–	2	0.6ll
Spain	17	3	–	2	1	–	–	–	23	7.4s
Sweden	–	4	–	–	–	–	–	–	4	1.3s
Switzerland	10	–	–	–	–	–	–	–	10	3.2s
Period:
Medieval	48	60	1	12	16	–	2	2	141	45.2hh
Early modern	83	42	3	12	8	4	14	5	171	54.8ll
Total no.	131	102	4	24	24	4	16	7	312	100.0
Total %	42.0	32.7	1.3	7.7	7.7	1.3	5.1	2.2	100.0
Notes: Bohemia has zero observations, significantly lower (at the 0.05 level) than its percentage of observations in the guilds database overall. Bulgaria has zero observations, significantly lower (at the 0.10 level) than its percentage of observations in the guilds database overall. For all other countries not represented in this table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Qualitative guilds database: 312 observations of guilds imposing conditions on widows’ practice.

Remarriage conditions on widows’ rights were enforced in practice, as shown by the real-life conflicts that gave rise to many of the observations in Table 5.10. In Paris in 1399, for instance, the candlemakers’ guild forbade a master’s widow to continue working, and only allowed her to resume work after an appeal to a public court and on condition that she did not do wage-work for other masters, did not provide labourers to work for others, did not accept new apprentices, and, if she remarried, did not train her new husband or any joint offspring.⁶³ Other guilds were much more severe, as in Bottwar in 1711 when a shopkeeper’s widow married a dyer, causing the shopkeepers’ guild to close down her shop and confiscate her wares,⁶⁴ or in Rome in 1775 when a tailor’s widow tried to go on working after remarrying and was formally prosecuted by the tailors’ guild.⁶⁵

A second condition guilds imposed, comprising one-third of the observations in Table 5.10, was a limit on the time-period during which a widow could continue the workshop after her husband’s death. The most frequent duration was a year (or a year and a day), but it could be as long as 4 years or as short as 4 weeks. In about half of all cases, the time-limit was absolute, while in the other half it was relaxed if the widow had children, employed a (male) journeyman, restricted the goods and services she produced, or satisfied other conditions.

Time-limits on widows’ rights did not just appear in law, but were enforced in practice, as shown by the real-life situations giving rise to a number of the observations in Table 5.10. In 1780, for instance, the Rome gold-leafers’ guild tried to limit a widow to the statutory six-months’ practice, whereupon she appealed on the grounds that she had conducted the craft alongside her husband for over thirty years; even Papal intervention, however, sufficed only to extend her right to work for two further years.⁶⁶ Contemporaries openly acknowledged that guilds time-limits harmed widows, as in the 1730s when the Prussian state abolished all such time-limits in order to disencumber the welfare system of craftsmen’s widows.⁶⁷

Guilds also imposed demographic conditions on widows’ permission to continue the family workshop, as emerges from 16 per cent of cases in Table 5.10. A widow’s guild rights were, as Josef Ehmer emphasizes, a strictly temporary arrangement based on the assumption that “sooner or later, the woman would pass on the trade to a man who would be recognized as a master and full member of the guild”.⁶⁸ Some guilds embedded this view in their ordinances, with rules that made a widow’s rights conditional on her demographic characteristics, including how old she was, whether she had offspring, and above all whether she had a son.

Demographic conditions on widows’ rights were, as we have seen, not just imposed on paper but enforced in practice. They affected women’s life chances. A widow who was regarded as too young could have her rights abrogated for that reason alone, as in sixteenth-century Augsburg where the goldsmiths’ guild challenged the right of Joachim Nitzel’s widow to continue operating the family workshop on the grounds that “by reason of her youth, Nitzlin was to be classed as a future wife not as a widow”.⁶⁹ A widow without offspring could be forced to struggle for the right to continue working, as in Venice in 1631 when the impoverished widow of a blacksmith claimed the right to continue operating the workshop, “as has been observed on other occasions”, but the guild rejected her claim because she was childless; the widow only secured the right to go on working after incurring the costs of a legal appeal.⁷⁰

Guilds also imposed other, miscellaneous conditions on widows’ rights, comprising about 9 per cent of the cases in Table 5.10. Some made a widow’s permission to continue the workshop conditional on her paying an additional fee, above and beyond those already paid by her deceased husband when he was admitted. In Rome in 1757, for instance, the weavers’ guild allowed a master’s license to pass to his son automatically but required a formal application and a fee before it would transfer the license to the master’s widow.⁷¹ A number of guilds made a widow’s rights to work dependent on her demonstrating a good reputation, most often defined in sexual terms but sometimes more generally. In late medieval Halberstadt, the bakers’ guild forbade one master’s widow to continue to work because she had “brought a child into the world at the wrong time”.⁷² And many guilds subjected a widow’s right to continue operating the workshop to the judgment and favour of the male guild masters or the town councilors. In fourteenth-century Berlin, for instance, the old-clothes-menders’ guild made widows’ rights “solely dependent on approval by the guild members”.⁷³ Such catch-all reputation and approval clauses inevitably involved discretion, opening them to arbitrary deployment by guilds whose male members wished to limit female competition.

LIMITATIONS ON WHAT WIDOWS COULD DO

Even if a widow could fulfill all the conditions required to take up guild rights, she still often faced limits on what she could do. The qualitative database contains 262 observations of guilds imposing business restrictions on widows that they did not impose on male masters. They are drawn from guilds in over 80 different occupations, refuting the idea that guild restrictions simply reflected the natural order according to which women were incapable of certain activities: the observations range from unskilled occupations such as vegetable-selling to skilled ones such as practising as an apothecary, from light crafts such as embroidery to heavy ones such as carpentry.

As Table 5.11 shows, Estonia, Finland, Germany, Norway, and Switzerland are significantly over-represented relative to their share of the overall guilds database, resembling Table 5.10 in revealing a particular intensity of restrictions in the Scandinavian and German-speaking world. Conversely, England and the two parts of the Netherlands are significantly under-represented, again resembling Table 5.10 in suggesting a lower intensity of restrictions in the north Atlantic economies. Nonetheless, the practice was sufficiently widespread that guilds can be observed imposing conditions on widows’ rights in 14 different European societies.

The observations in Table 5.11 span the 561 years from 1268 to 1829. While conditions on widows taking up rights in Table 5.10 are weighted towards the Middle Ages, restrictions on widows’ business practice in Table 5.11 are proportionately balanced between the medieval and early modern periods. In neither case does the chronological distribution support the idea of a medieval “golden age” for women and guilds.

TABLE 5.11: Guild Limits on What Practising Widows Could Do, 1268–1829

Keeping apprentices

Working without journeymen

Journeymen numbers

Keeping advanced journeymen

Workshop size

Equipment

Raw materials

Market access

Type of work

“Half-rights”

Guild decision-making

Total

no.

%

Country:

Austria

4

4

–

–

–

–

–

–

1

–

1

10

3.8s

Denmark

–

–

3

–

–

–

–

–

–

–

–

3

1.1s

England

1

–

1

–

–

–

–

–

1

–

3

6

2.3ll

Estonia

–

2

–

–

–

–

–

2

–

–

–

4

1.5hh

Finland

–

2

–

–

–

–

–

–

–

–

–

2

0.8hh

France

17

5

3

–

2

–

–

–

2

–

9

38

14.5s

Germany

23

43

11

8

–

3

–

4

10

2

27

131

50.0hh

Italy

9

5

–

–

–

–

1

–

–

–

18

33

12.6s

N. Netherlands

–

2

–

–

–

–

–

–

–

–

3

5

1.9ll

Norway

–

5

–

–

–

–

–

–

–

–

–

5

1.9hh

Poland

–

–

–

–

–

–

–

–

–

–

1

1

0.4s

Scotland

–

1

–

–

–

–

–

–

–

–

–

1

0.4s

Spain

5

5

–

–

–

–

1

–

–

–

–

11

4.2s

Switzerland

10

1

–

–

–

–

–

–

1

–

–

12

4.6hh

Period:

Medieval

8

32

1

–

–

–

–

4

10

1

10

66

25.2s

Early modern

61

43

17

8

2

3

2

2

5

1

52

196

74.8s

Total no.

69

75

18

8

2

3

2

6

15

2

62

262

100.0

Total %

26.3

28.6

6.9

3.1

0.8

1.1

0.8

2.3

5.7

0.8

23.7

100.0

Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. The Southern Netherlands and Sweden have zero observations, significantly lower (at the 0.05 level) than their percentage of observations in the guilds database. Bohemia and Bulgaria have zero observations, significantly lower (at the 0.10 level) than its percentage of observations in the guilds database. For all other countries not represented in this table, their percentage of observations (zero) is not significantly lower than in the guilds database. Source: Qualitative guilds database: 262 observations of guild restrictions on what practising widows were alllowed to do.

About one-quarter of the observations in Table 5.11 show guilds limiting widows’ access to apprentices, the cheapest source of labour.⁷⁴ Only a few European guilds allowed widows to employ apprentices freely, and nearly all were in England, where the customs of cities such as London and York explicitly allowed wives to train apprentices and widows to retain them.⁷⁵ Most European guilds, by contrast, forbade widows to keep apprentices, although a few let them train their own sons or finish training existing apprentices, but not take on new ones. Some guilds, such as those of the Barcelona silk-workers, prevented widows from actually keeping apprentices, even though the guild ordinance had no rule against it.⁷⁶ Counter to the claim that guild restrictions were easily circumvented, guilds can be observed enforcing these prohibitions. In 1399, for instance, the Paris candlemakers’ guild only allowed a widow to continue working on condition that she did not employ apprentices.⁷⁷ In eighteenth-century Augsburg, a careful study of the apprenticeship admissions of seven separate guilds found that rules against widows keeping apprentices were universally enforced.⁷⁸ In nineteenth-century Vienna, the haberdashers’ guild prohibited widows from retaining existing apprentices, with the result that one in five haberdasher apprentices had to finish training in a different workshop.⁷⁹

As discussed in the introduction, some have argued that guild restrictions on females just reflected the natural order according to which women’s domestic responsibilities prevented them from learning vocational skills. So, it might be argued, guilds had to forbid widows to keep apprentices because women lacked the skills to teach the craft. Some guilds indeed used this argument.⁸⁰ But then why, in the same occupation, did some guilds forbid the practice and others permit it? Silk-weavers’ guilds completely barred widows from employing apprentices in early modern Barcelona, forbade it with some exceptions in Seville, Lyon, and Geneva, permitted it under certain conditions in Grenada, and permitted it unconditionally in Vienna.⁸¹ Tailors’ guilds completely barred widows from keeping apprentices in Augsburg and Würzburg, barred it with some exceptions in Turin, barred taking new ones but permitted retaining existing ones in Paris, and permitted it freely in London.⁸² Virtually all guilds in early modern Bristol, Oxford, London, and York permitted widows to employ apprentices, while all guilds in early modern Chartres allowed them only to finish training existing apprentices, and in early modern Vienna all guilds except the silk-weavers banned it completely.⁸³ Preventing widows from employing apprentices thus cannot be viewed as serving objective training standards.

It did, however, serve the purpose of reducing widows’ ability to compete with male masters. As Marta Vicente points out for early modern Barcelona, being prevented from employing apprentices blocked widows’ access to the cheapest form of labour.⁸⁴ Employing an apprentice was typically less than half as costly as paying a journeyman.⁸⁵ In seventeenth-century Germany, widows explained in petitions that they were compelled to give up their workshops because “the small amount of business they could do with no apprentices or journeymen was not enough to pay their guild dues”.⁸⁶ Lack of access to cheap apprentice labour decreased a widow’s ability to stay in business even when she had the legal right to do so.

A second limit many guilds placed on widows’ business practice, accounting for 29 per cent of the cases in Table 5.11, took the opposite track. Many guilds only permitted a widow to practise the occupation at all if she employed a journeyman or some other male guild member to supervise. A particularly strict variant of this rule was to bar the widow from doing core craft tasks (cutting cloth as a tailor, slaughtering animals as a butcher, doing any work “with her own hands”) and to require her instead to employ a male guild worker. Guilds not only included these regulations in their ordinances but, again, implemented them in practice. A particularly striking example is provided by the Hildesheim wigmakers who in 1724 conducted a concerted campaign against a particular widow of a deceased colleague, on the grounds that she was violating the requirement that she employ male guild labour and was instead “conducting her full work like an honourable master sitting in the guild”. Claiming that “such a woman would not be suitable for any honourable guild in the whole world and would by no means be tolerated in such a guild”, a number of male masters threatened to mount a physical attack on her workshop and the guild prosecuted her before the town magistrates. Ultimately, the guild made her business life so difficult that she ceased to operate as an independent widow and entered into a new marriage.⁸⁷

It might be argued that widows had to be prevented from running workshops without male supervision because they themselves had not undergone formal guild apprenticeships.⁸⁸ But doubt is cast on this idea by the absence of observations in Table 5.11 of guilds imposing this requirement in England or the Southern Netherlands, economies with sophisticated industries requiring considerable skill. Conversely, in Germany, where such restrictions were widespread, a number of widows are recorded pointing out that their own skills exceeded those of the journeymen the guild obliged them to employ. In sixteenth-century Augsburg, for instance, one butcher’s widow demanded vainly to be allowed to work on her own and “dispense with what she felt to be incompetent hired help”.⁸⁹ So it is quite problematic to view guild restrictions on widows as reflecting objective skill requirements.

Rather, such restrictions generated benefits for male guild members. First, they compelled widows to create jobs for male guild members, as in eighteenth-century Finland where almost all guilds obliged widows to employ journeymen and “enforced this rule very strictly as it ensured employment for candidates for the position of master craftsman”.⁹⁰ Second, these requirements reduced widows’ capacity to compete with male guild members. Journeymen were so costly that in many occupations only a tiny minority of masters could afford them: in eighteenth-century Wildberg, 91 per cent of male masters in traditional crafts and 97 per cent in proto-industrial worsted-weaving employed no journeymen;⁹¹ even in wealthy Augsburg, 40 per cent of guild masters worked without journeymen.⁹² Employing a journeyman was thus exceptional, and imposing this costly burden on a widow disadvantaged her compared to male masters. Thus in sixteenth-century Augsburg the nearly universal requirement that a widow employ a journeyman made it impossible for many widows to continue practising the craft,⁹³ and in early modern Nantes only a minority of widows entitled to continue guild workshops did so since “a journeyman’s wages could outweigh the profits to be gained from his work.”⁹⁴ In eighteenth-century Turin, a female button-maker appealed against a regulation forbidding her from working unless she used “the name of a certified worker, whom she much pay exorbitantly . . . making it impossible for the poor supplicant to provide for her large family when calamity strikes”.⁹⁵

Guilds also restricted widows’ access to inputs. Some limited the number or seniority of journeymen a widow could employ, and they enforced such rules sufficiently to cause business hardship. In seventeenth-century German towns, for instance, one reason widows gave for not being able to continue guild workshops even when they were legally entitled to do so was that caps on their workforce decreased the viability of their business.⁹⁶ In Augsburg in 1731, 17 weavers’ widows went so far as to petition for permission to employ journeymen “like the male masters”.⁹⁷ Other guilds limited the size of widows’ workshops, the amount of equipment they could use, or the volume of raw materials they could process. Not infrequently, guilds limited widows’ market access, as in 1555 when the Augsburg retailers’ and tanners’ guilds reduced competition for male masters by excluding widows from selling wares in the St Ulrich’s market on Mondays.⁹⁸

Other guilds limited the kinds of work widows were allowed to undertake. In Frankfurt in 1410, the fishmongers’ guild forbade widows to purchase fish.⁹⁹ In a number of medieval German and Swiss towns, tailors’ guilds restricted widows to making garments out of just a few cheap fabrics.¹⁰⁰ In Augsburg in the 1530s, the butchers’ guild restricted widows to low-profit work such as sausage-making.¹⁰¹ In Lyon in 1540, the barber-surgeons’ guild restricted widows to shaving beards and healing simple wounds.¹⁰² In eighteenth-century Nantes, the butchers’ guild limited women to trading in offal, heads, and feet.¹⁰³ Many early modern printers’ guilds let widows run the workshop “but not to pull the press, which was man’s work”.¹⁰⁴ Inevitably, such restrictions made widows’ businesses less viable.

Finally, guilds limited widows’ participation in collective decision-making, sociability, and public representation, as shown by 24 per cent of the cases in Table 5.11.¹⁰⁵ Guilds, as we shall see in Chapter 9, held periodic assemblies to decide on guild rules and policy, elected officers to implement these rules and policies, voted on regulations and lobbying, held banquets to consolidate multi-stranded ties linking work and sociability, and organized processions, demonstrations, and other forms of public display to demonstrate their importance to society. Guilds almost always restricted women from holding guild office, frequently prohibited them from attending guild assemblies, and when they allowed them to be present barred them from casting a vote or sharing the guild meal. The Kirchdorf-Micheldorf scythe-smiths’ guild, for instance, had no objection to widows running smithies or making scythes with their own hands, but excluded them from the midday table at the annual assembly, requiring them to eat by themselves at a different time.¹⁰⁶

These findings help explain the wide gap between the estimated 18 per cent widowed household headship and the roughly 8 per cent widowed guild mastership observed earlier. Many widows gave up the trade. Some were directly forced out by guild conditions, as we saw with widows whose shops were closed by the guild because they were childless, sonless, too young, disliked by male guild masters, or wished to remarry. Other widows left the trade because guilds denied them cheap apprentices, made them hire costly journeymen, restricted their workshops, equipment, and raw materials, forbade them to do particular kinds of work, and denied them a voice in guild decisions.¹⁰⁷

Such restrictions may also have contributed to widows’ relative poverty. In Augsburg in 1601, for instance, widows made up 15 per cent of all weavers, but only 5 per cent of prosperous workshops with employees.¹⁰⁸ In early modern Nuremberg, widowed female ironworkers had much lower taxable wealth than male ones, over two-thirds of them lying in the lowest tax bracket.¹⁰⁹ In early modern Dijon, only 2 to 4 per cent of artisan households were headed by widows, and their taxable wealth was much lower than that of their male-headed counterparts.¹¹⁰

This is not to say that widows’ rights were valueless. Women who had been married to guild masters were clearly better off in guilds that granted even conditional and limited widows’ rights than in those that wholly banned females as masters—Types 4 and 5 in Table 5.1.

Guilds That Excluded Females from Mastership

Only a few guilds denied widows all rights. Guild masters sought to draw a balance between providing for their own widows while suppressing competition from those of other men. So most ordinary male guilds preferred not to exclude widows completely but still sought to limit their ability to compete.

However, there were a few guilds which did exclude widows. The qualitative database contains 52 observations of guilds which granted no widows’ rights though they did let other women work at the occupation. These Type 4 guilds thus account for about 0.06% of all European guilds, about the same percentage as the all-female guilds of Type 1.

As Table 5.12 shows, observations of non-widows’-rights guilds span the five and a half centuries from 1247 to 1797. The medieval period accounts for 38 per cent of cases, compared to only 27 per cent in the guilds database, providing no support for the idea that guild limits on women were an early modern development. Consistent with previous findings, England and the Southern Netherlands are under-represented while Germany is over-represented, although so too is the Northern Netherlands, surprisingly given the otherwise comparatively favourable economic status of Dutch women.

The occupational distribution of these guilds with no widows’ rights decisively refutes the view, held by some scholars, that guilds simply reflected women’s putative lack of strength and skill. As Table 5.12 shows, just 21 per cent of guilds without widows’ rights were in heavy crafts (sword-making, carpentry, bricklaying) while 56 per cent were in light and sedentary occupations (tailoring, weaving, selling beer). Likewise, just 19 per cent were in highly skilled occupations (apothecaries, goldsmiths, barber-surgeons), while 69 per cent were ordinary crafts (beltmakers, tailors, wooden-shoe-makers), and 12 per cent in unskilled occupations (cheesemongers, carters, gardeners).

A number of the occupations that excluded widows in Table 5.12 were organized into mixed-sex or all-female guilds in other places, casting further doubt on the idea that guilds’ treatment of women merely reflected women’s technical limitations. Barge-operators and silk-weavers, for instance, barred widows in some places and organized all-female guilds in others. Dyers, tailors, barber-surgeons, goldsmiths, hatters, belt-makers, and cheesemongers prohibited widows in some places but in others had guilds of mixed sex. The mixed-sex goldsmiths’ guild of early modern London, where both apprenticeship and mastership were open to females,¹¹¹ would have been surprised to learn from the goldsmiths’ guild of sixteenth-century Augsburg that “women understand nothing of the goldsmiths’ work, and do not know when jewels, gold or silver are good and true, for the goldsmiths’ craft is not considered to be the least but the greatest craft”.¹¹² Likewise, the mixed-sex cheesemongers’ guild of sixteenth-century Gouda, in which 16 per cent of incoming masters were female,¹¹³ would have been astonished to learn from the exclusively male cheesemongers’ guild of sixteenth-century Genoa that widows must be barred from continuing to operate businesses on the grounds that women were “not considered to be legal persons, nor suitable to be in charge of shops”.¹¹⁴

TABLE 5.12: Guilds with No Widows’ Rights, 1247–1797
	Widows’ rights prohibited	Widows’ rights abolished	Widows’ rights not mentioned	Total
				no.	%
Country:
France	5	1	–	6	11.5s
Germany	8	2	18	28	53.8hh
Italy	6	–	–	6	11.5s
N. Netherlands	5	4	–	9	17.3hh
Spain	1	–	–	1	1.9s
Switzerland	–	2	–	2	3.8s
Period:
Medieval	3	1	16	20	38.5h
Early modern	22	8	3	32	61.5l
Physical demands:
Heavy	8	–	3	11	21.2
Medium	4	4	4	12	23.1
Light	13	5	11	29	55.8
Skill requirements:
Skilled	8	2	–	10	19.2
Medium	13	7	16	36	69.2
Unskilled	4	–	2	6	11.5
Total no.	25	9	18	52	100.0
Total %	48.1	17.3	34.6	100.0
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. England has zero observations, significantly lower (at the 0.05 level) than its percentage of observations in the guilds database. The S. Netherlands has zero observations, significantly lower (at the 0.10 level) than its percentage of observations in the guilds database. For all other countries not represented in this table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Qualitative guilds database: 52 observations of guilds prohibiting, abolishing, or not mentioning widows’ rights.

Guilds Excluding All Female Participation

Finally, there were guilds—Type 5 in Table 5.1—that banned work by females in any position, even as male masters’ family members or employees. The qualitative database contains 45 observations of such guilds, which thus comprised about 0.06 per cent of all guilds, approximately the same percentage as all-female and non-widows’-rights guilds. Although Table 5.13 shows that such guilds can be observed for over five centuries, from 1257 to 1759, the medieval period accounts for 42 per cent of cases, significantly more than its 27 per cent share of the overall guilds database, again casting doubt on the idea that guild discrimination against women was an early modern innovation. Consistent with previous findings, the Southern Netherlands is under-represented and Germany is over-represented. But Italy is also under-represented and the Northern Netherlands over-represented, complicating the picture. The very small numbers involved caution against placing weight on cross-country comparisons.

Just over half the cases in Table 5.13 comprised guilds that excluded all females from any work in the occupation, as did the medieval Paris Saracen-carpet-makers¹¹⁵ and the early modern Gouda beer-carriers and waggoners.¹¹⁶ A further one-third of cases involved guilds excluding females from core tasks. The medieval Paris fullers agreed not to let a female touch a cloth before it was sheared;¹¹⁷ the medieval Frankfurt fishmongers did not let female workers do the fish buying,¹¹⁸ the sixteenth-century Lyon silk-weavers forbade women to work at the loom;¹¹⁹ the eighteenth-century Gouda pipe-makers forbade females to form pipe-heads.¹²⁰ Another 7 per cent of cases involved excluding women from most tasks, as in fifteenth-century Danzig, where the cabinetmakers banned females from all work except dyeing, varnishing, and sawing.¹²¹ A further 7 per cent of cases involved guilds forbidding females to carry out certain important craft tasks, as when the sixteenth-century Poitiers candlemakers forbade female workers to sell candles.¹²² Finally, there were isolated cases of guild bans imposed with minor exceptions, as when the fifteenth-century Bristol weavers banned all females except those already working at the craft,¹²³ or when the tawers and pursers in sixteenth-century Frankfurt an der Oder cleansed workshops of all females but maidservants.¹²⁴

TABLE 5.13: Guilds that Prohibited Work by All Females, 1257–1759
	All females	All females	All females	All females	All but existing workers	All but servants	Total
	all work	core work	most work	some work	all work	all work	no.	%
Country:
England	1	–	–	–	1	–	2	4.4s
France	2	3	–	2	–	–	7	15.6s
Germany	14	9	3	1	–	1	28	62.2hh
Italy	–	1	–	–	–	–	1	2.2ll
N. Netherlands	5	1	–	–	–	–	6	13.3hh
Spain	1	–	–	–	–	–	1	2.2s
Period:
Medieval	7	7	3	1	1	–	19	42.2hh
Early modern	16	7	–	2	–	1	26	57.8ll
Physical demands:
Heavy	3	3	2	–	–	–	8	17.8
Medium	3	5	–	–	–	–	8	17.8
Light	17	6	1	3	1	1	29	64.4
Skill requirements:
Skilled	2	1	–	–	–	–	3	6.7
Medium	18	11	2	2	1	1	35	77.8
Unskilled	3	2	1	1	–	–	7	15.6
Total no.	23	14	3	3	1	1	45	100.0
Total %	51.1	31.1	6.7	6.7	2.2	2.2	100.0
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. The S. Netherlands has zero observations, significantly lower (at the 0.10 level) than its percentage of observations in the guilds database. For all other countries not represented in this table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Qualitative guilds database: 45 observations of guilds prohibiting work by all females (with minor exceptions).

Again, the occupational patterns in Table 5.13 cast doubt on the notion that guild exclusion of female workers reflected technical requirements. Only 18 per cent of such guilds were in physically heavy crafts such as armour-making or carpentry, while 64 per cent were in light occupations such as weaving, candle-making, or selling fish. A majority (78 per cent) were in ordinary crafts requiring medium levels of skill, with only 7 per cent in skilled occupations such as medicine, and 16 per cent in unskilled ones such as carding wool or roasting meat. When a guild imposed such extreme bans on female workers, it was actuated not by the technical requirements of the occupation, but by other factors. The seventeenth-century Freiberg braid makers excluded females because of pressure from guild journeymen.¹²⁵ Eighteenth-century Augsburg ribbon-weavers excluded females because other German guilds were ostracising them for working with women.¹²⁶ And almost all early modern German hatters’ guilds excluded females because of pressure from hatters’ guilds in other German towns.¹²⁷ It was institutional pressure, not technical demands, that led guilds to the extreme step of banning female workers completely.

APPRENTICESHIP AND WOMEN

Up to now we have focused on whether guilds let women be masters. But masters were the aristocracy of craft guilds: they owned the businesses, they employed the workers. Below this privileged tip lay a vast iceberg of dependent workers—apprentices, journeymen, other servants, family members, and freelancers such as spinners—who vastly outnumbered the masters.¹²⁸

Let’s start with apprentices. Studies of modern developing economies hold that human capital investment plays a key role in economic growth, and investing in human capital through apprenticeship is widely regarded as central to the idea of a guild. How did guilds deal with women’s human capital?

TABLE 5.14: Guilds Excluding Females from Apprenticeship, 1261–1780
	All guilds in country	All guilds in region	All guilds in city	Individual guild: total exclusion	Individual guild: minor exceptions	Total
						No.	%
Country:
Austria	–	–	–	1	–	1	1.4s
Bohemia	–	–	–	1	–	1	1.4s
Denmark	2	–	–	–	–	2	2.8hh
England	–	–	–	3	–	3	4.2s
Finland	1	–	–	–	–	1	1.4hh
France	–	1	–	1	10	12	16.9s
Germany	1	1	2	16	5	25	35.2s
Italy	2	–	8	2	–	12	16.9s
N. Netherlands	–	–	–	1	–	1	1.4s
Spain	–	–	1	2	2	5	7.0s
Sweden	1	–	–	–	–	1	1.4s
Switzerland	–	–	–	4	3	7	9.9hh
Period:
Medieval	2	1	4	4	2	13	18.3l
Early modern	5	1	7	27	18	58	81.7h
Total no.	7	2	11	31	20	71	100.0
Total %	9.9	2.8	15.5	43.7	28.2	100.0
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. The S. Netherlands has zero observations, significantly lower (at the 0.05 level) than its percentage of observations in the guilds database. For all other countries not represented in this table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Qualitative guilds database: 71 observations of guild restrictions on female apprenticeship.

Had guilds been concerned to guarantee skills, they should have let females get training as well as males. But most of them did the opposite, as emerges from 71 observations in the qualitative database which show guilds restricting female apprenticeship. As Table 5.14 shows, these cases come from 12 different societies, but reveal a familiar pattern in which central Europe and Scandinavia tend to be over-represented, while the North Atlantic societies are under-represented. Of the 20 cases of blanket prohibitions against female apprentices by all guilds in a particular society, region, or city, 40 per cent are in Scandinavia and Germany. Many central European guilds did not even bother to publish a rule, but took for granted that girls would not be admitted to apprenticeship, as with the Wildberg worsted-weavers’ guild which stated no rule against female apprentices but enrolled no girls among the 1,258 apprentices registered between 1598 and 1760.¹²⁹

Observations of guilds blocking girls’ apprenticeship span over five centuries from 1261 to 1780, but the early modern period is over-represented compared to the medieval period (though only at a borderline level of statistical significance).¹³⁰ This arises partly from the formation of new early modern guilds, whose first step was often to exclude females from apprenticeship, as occurred in 1555 when the London silk-weavers came under the control of the Weavers’ Company and immediately prohibited training girls.¹³¹ It also arises from the many observations between c. 1450 and c. 1700 of existing guilds shifting from permitting female apprentices to banning them. Guilds enforced these new restrictions in order to reduce competition for their male members at the expense of young women who evinced a strong desire to invest in their own human capital. In 1696, for instance, despite petitions to the contrary, the Geneva gold-and-silk-cloth-weavers’ guild put a stop to female apprenticeship because “girls could feed themselves more cheaply than men, thereby depressing wages and reducing the prestige of the trade”.¹³²

As such cases illustrate, girls wanted training and got it unless the guild blocked them. That is, female apprentices existed. How many were there, and how did they come into being? Many studies have emphasized how tiny the numbers of female apprentices in European societies were, especially after c. 1450.¹³³ But others point to large percentages of female apprentices in particular cases, concluding that the guild system was compatible with high human capital investment for females.¹³⁴ Can quantitative findings help assess these incommensurate views?

Quantifying female apprenticeship is difficult, not least because guilds often banned it. Nonetheless, the quantitative database contains 157 observations of the share of females in a group of apprentices. The compilation is far from representative. For one thing, it is necessarily restricted to unusual situations in which female apprentices were permitted. For another, England comprises nearly 80 per cent of the cases—only five other societies are represented at all. Furthermore, the early modern period accounts for over 90 per cent of the observations, significantly more than its share of the overall guilds database. But although these 157 observations are biased towards observations of societies and guilds that were unusually favourable to females, they nonetheless serve better than scattered examples for assessing whether the European guild system was compatible with high human capital investment for females.

Guilds were only one of a variety of institutions that arranged and recorded European apprenticeships, so Table 5.15 breaks down the sample according to the source of the observation.¹³⁵ Distinguishing between guild and non-guild apprenticeships clarifies the debate between optimists and pessimists on female apprenticeship, particularly in the English context. In England, female apprenticeships were widespread. But the reason for this, it turns out, was that non-guild apprenticeships were widespread.¹³⁶ In the English samples in Table 5.15, girls made up 31 per cent of parish or charity (i.e., non-guild) apprentices but just 2 to 5 per cent of guild apprentices. In France, girls made up 72 per cent in the single sample of charity apprentices, but less than 12 per cent in the nine notarial samples combining guild and non-guild apprenticeships. The five observations for Italy are highly variegated, ranging from 78 per cent female apprentices in the mixed-sex (but mainly female) Florentine silk-throwers’ guild in 1663, to 59 per cent of the non-guild apprenticeships organized by a Turin charity in 1721, to an average of less than 13 per cent in three notarial samples of combined guild and non-guild apprentices between the thirteenth and the sixteenth century. The seven observations for the mixed-sex (but highly feminized) Viennese silk-weavers’ guild show 34 per cent female apprentices in the three decades before guild abolition (1830–60), rising to 42 per cent in the four decades after abolition (1860–1900).

The high proportion of females in some of the observations in Table 5.15 suggests that females had a strong demand for apprenticeships as a means to invest in their own human capital. This demand was satisfied in two specific contexts. The first was in mixed-sex guilds, particularly those with a large share of female masters. But mixed-sex guilds were very rare, as we have seen, comprising less than half of one per cent of all European guilds. The second context in which girls could get into apprenticeships was outside the guild system. Across all 157 observations in Table 5.15, just 14.6 per cent of apprentices were female. But this masks a big divide: girls accounted for nearly one-third of apprentices outside guilds but just 5 to 6 per cent of those inside the guild system. This supports the pessimist view: European guilds were not favourable to investment in women’s human capital.

TABLE 5.15: Proportion of Female Apprentices in Guild and Non-Guild Apprenticeships, Thirteenth to Nineteenth Century
	Guild apprenticeships: guild records		Guild apprenticeships: other documents		Mixed guild & non-guild apprenticeships: notarial records		Non-guild apprenticeships: charity & parish		Non-guild apprenticeships: official registers after guild abolition		Total
	mean	no.	mean	no.	mean	no.	mean	no.	mean	no.	mean	no.	col. %
Country:
Austria	34.2	3	–	–	–	–	–	–	47.4	5	42.5	8	5.1s
England	2.1	73	4.8	12	–	–	30.5	39	–	–	11.3	124	79.0hh
France	–	–	–	–	11.9	9	72.0	1	–	–	17.9	10	6.4ll
Italy	78.3	1	–	–	12.6	3	59.2	1	–	–	35.1	5	3.2ll
S. Netherlands	27.5	4	–	–	–	–	–	–	–	–	27.5	4	2.5ll
Switzerland	13.5	6	–	–	–	–	–	–	–	–	13.5	6	3.8s
Period:
Medieval	18.4	6	–	–	12.6	7	–	–	–	–	15.3	13	8.3ll
Medieval & early modern	–	–	–	–	4.7	1	–	–	–	–	4.7	1	0.6ll
Early modern	5.1	81	4.8	12	13.0	4	32.2	41	47.4	5	14.6	143	91.1hh
Total	6.0	87	4.8	12	12.1	12	32.2	41	47.4	5	14.6	157	100.0
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. Germany, the N. Netherlands, and Spain have zero observations, significantly lower (at the 0.05 level) than their percentage of observations in the guilds database. Sweden has zero observations, significantly lower (at the 0.10 level) than its percentage of observations in the guilds database. For all other countries not represented in this table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Quantitative guilds database: 157 observations of share of female apprentices.

TABLE 5.16: Guilds and Journeywomen, 1261–1754
	Journeywomen allowed by:		Journeywomen exist:			Journeywomen are prohibited by guild:		Total
	mixed-sex guild	female guild	in no cases	rarely, no future	very rarely	in all work	in most work	no.	%
Country:
Bohemia	–	–	–	–	–	1	–	1	3.3s
England	–	–	1	–	2	–	–	3	10.0s
France	3	–	–	1	2	–	1	7	23.3s
Germany	2	3	–	–	–	4	–	9	30.0s
Italy	2	–	–	–	–	–	–	2	6.7s
Spain	–	–	–	6	–	–	–	6	20.0hh
Switzerland	1	–	–	–	–	1	–	2	6.7h
Period:
Medieval	4	3	–	4	2	1	–	14	46.7hh
Early modern	4	–	1	3	2	5	1	16	53.3ll
Total no.	8	3	1	7	4	6	1	30	100.0
Total %	26.7	10.0	3.3	23.3	13.3	20.0	3.3	100.0
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. For all countries not represented in this table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Qualitative guilds database: 30 observations of guilds mentioning female journeymanship.

JOURNEYWOMEN

The classic guild career path involved progressing from apprenticeship to journeymanship, and thence to mastership. Did guilds let women follow this career? Few guilds even mentioned journeywomen, as reflected in the fact that the guilds database contains just 30 references to them across eight centuries and 23 societies.

As Table 5.16 shows, observations of journeywomen fall into three groups. In about 37 per cent of the very few cases we have at all, a guild formally referred to the status of journeywomen; unsurprisingly, these were in either mixed-sex guilds (28 per cent) or all-female ones (10 per cent). What is more surprising is that there are so few such observations. Even guilds that had female masters often made no reference to journeywomen.

Another 40 per cent of the small number of observations consist of references to whether journeywomen actually existed. In 14 per cent of cases they existed but were very rare, in 24 per cent they were not only rare but forbidden to proceed to mastership, and in 3 per cent they were never observed, even though there was no regulation forbidding their existence. In fourteenth-century Montpellier, for instance, girls comprised over a quarter of the apprentices in notarized contracts but hardly ever appeared among journeymen, and even then were not allowed to progress to mastership.¹³⁷

In the final 23 per cent of the small number of cases, guilds formally placed limits on journeywomen, as in 1561 when the Lyon silk guild largely abolished female apprenticeship and strictly limited the activities of females who had graduated to the journeywoman stage,¹³⁸ or in 1592 when the Geneva trimmings-makers’ guild barred girls who had completed apprenticeships from becoming journeywomen.¹³⁹ Such restrictions were enforced in practice, as in eighteenth-century Augsburg where a master’s daughter who had long been carrying out skilled work for her father was prosecuted by the painters’ guild as soon as she tried to get a job as a journeywoman (Gesellin).¹⁴⁰

Journeywomen were very rare at any period, but were increasingly restricted after c. 1500. This accounts for the fact that of the scant references to journeywomen, the medieval period accounts for 47 per cent, significantly more than its share of the overall guilds database. The contexts in which journeywomen were mentioned also changed: before 1500, only 7 per cent of observations involved complete prohibitions or the nonexistence of journeywomen, significantly more than the 44 per cent after 1500. Journeywomen were rare at all times, but to the extent that they appear at all, they were significantly more often banned or restricted by guilds after 1500 than before.

GUILDS AND MASTERS’ WIVES

As mentioned in the introduction, some scholars have argued that guilds were not harmful to women since they granted them a good economic position as members of masters’ households. According to this view, guilds did not even discriminate against females, since they also made male work conditional on membership in a master’s household.¹⁴¹

Guilds certainly limited work by both sexes outside masters’ households. But many placed additional limits on women inside masters’ households. Guild masters sought to draw a balance between enjoying the benefits of work by their own wives, daughters, maidservants, and other female household members, while suppressing competition from those of other masters. In addition, guild masters often faced opposition from male journeymen, apprentices, and poorer fellow masters who believed that guilded work by female household members took away their jobs. So most guilds did not completely forbid the work of female family members, but sought instead to deal with complaints by male journeymen, apprentices, and poor masters, sometimes by confirming that female family members were allowed to assist the master, sometimes by agreeing to restrict or ban such participation.

This even applied to masters’ wives, as emerges from 74 observations in the qualitative database shown in Table 5.17. Half the observations are of guilds that explicitly permitted masters’ wives either to be masters of the guild (if it was a mixed-sex guild) or to work freely at all aspects of the occupation (if it was a normal all-male guild). The 1541 Leiden dyers, for instance, declared that “[i]t is always to be understood that a man and his wife who have given the aforesaid oath shall be considered a single person.”¹⁴² This is not surprising, since a guild master benefited greatly if his wife could work unrestrictedly. Most central and north-western European societies were characterized by small, nuclear-family households, where the only other adult worker was a man’s wife: offspring were either too young to work as productively as adult workers or had already left home, and servants were expensive. Yet, when a guild explicitly declared that the master’s wife was allowed to work, this suggested that her right to do so had been questioned and the guild felt the need to clarify it formally.

The other half of observations in Table 5.17 involve restrictions or prohibitions on the work of masters’ wives. In 27 per cent of cases, guilds restricted but did not wholly ban it, as in 1478 when the Leiden fullers’ guild forbade masters’ wives to deliver or collect cloths, carry out textile work for clients, or do hand finishing,¹⁴³ or in 1579 when the Zwickau tanners’ guild prohibited masters’ wives from purchasing raw hides in the market, except when their husbands were ill or it was market day.¹⁴⁴ In 22 per cent of cases, guilds completely outlawed work by masters’ wives, as in 1592 when the Nuremberg roofers’ guild forbade wives to work on roofs or mix mortar,¹⁴⁵ or in 1696 when the Freiberg braid-makers’ guild banned wives from the workshop (though the government later repealed the ban).¹⁴⁶ In most cases the prohibition was absolute, though there were occasional exceptions—if the husband was ill, for example, or the wife had been apprenticed within the guild.

TABLE 5.17: Guilds and Masters’ Wives, 1271–1777
	Guild permits wife’s work:		Guild restricts wife’s work:		Guild prohibits wife’s work:			Total
	as master	as wife	wholly	unless emergency	wholly	unless emergency	unless apprenticed	no.	%
Country:
Denmark	2	2	–	–	–	–	–	4	5.4hh
England	7	3	1		2	–	1	14	18.9hh
France	4	–	3	1	–	–	–	8	10.8s
Germany	13	4	11	–	9	2	–	39	52.7hh
N. Netherlands	1	1	–	–	2	–	–	4	5.4s
S. Netherlands	–	–	2	–	–	–	–	2	2.7s
Spain	–	–	1	–	–	–	–	1	1.4l
Switzerland	1	–	1	–	–	–	–	2	2.7s
Period:
Medieval	19	7	11	–	5	2	1	45	60.8hh
Early modern	9	3	8	1	8	–	–	29	39.2ll
Total no.	28	10	19	1	13	2	1	74	100.0
Total %	37.8	13.5	25.7	1.4	17.6	2.7	1.4	100.0
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. Italy has zero observations, significantly lower (at the 0.05 level) than its percentage of observations in the guilds database. For all other countries not represented in this table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Qualitative guilds database: 74 observations of guilds’ regulation of wives’ work.

Here again, guild restrictions on wives’ work did not just exist on paper but were implemented in practice. In the 1490s, for example, three Leiden fullers were fined for letting their wives work.¹⁴⁷ During the sixteenth century the York bakers’ guild prosecuted masters who sent their wives to buy wheat.¹⁴⁸ During the eighteenth century various Barcelona guilds objected to masters’ wives working while their husbands were temporarily out of town.¹⁴⁹ In London in 1769, a member of the London weavers’ guild who allowed his wife to weave at the loom had his house invaded at 1:00 am by six fellow weavers who demanded “where the B——h his wife was, saying they would murder her directly, and they would cut his ears off, if he did not come up and show them where the work was”. Even after the husband sought to placate them by promising that “they might depend upon it, she should make no more of it”, the men searched out the family loom and cut away the “two coloured flowered sattin” his wife had been weaving.¹⁵⁰

The geographical and chronological patterns in Table 5.17 suggest that guild concerns about masters’ wives were significantly more salient in Germany, England, and Denmark than in other European societies, particularly Italy and Spain. Such concerns were also significantly more acute in the medieval period than the early modern period, though also significantly more often resolved in a permissive rather than a restrictive spirit. That many guilds perceived the need to regulate work even by women who were married masters of these very guilds casts doubt on the argument that the European guild system granted females a good economic position as long as they worked inside masters’ households.¹⁵¹

GUILDS AND MASTERS’ DAUGHTERS

The position of masters’ daughters is also sometimes adduced to support the view that guilds did not discriminate against females. S. R. Epstein goes so far as to question the whole idea that guilds restricted women’s participation, justifying his view by stating that masters’ daughters “did earn income from craft work”.¹⁵² Many guilds certainly tolerated craft work by masters’ daughters. But many others restricted it, excluding daughters from apprenticeship, prohibiting them from inheriting a father’s guild license, and even banning them from working in the workshop during a father’s lifetime.

TABLE 5.18: Guilds and Masters’ Daughters, 1261–1771
	Guild permits daughter’s work:		Guild restricts daughter’s work:		Guild completely prohibits daughter’s work:			Total
	as guild member	as daughter	as guild member	as daughter	as guild member	as daughter	unless apprenticed	no.	%
Country:
England	–	–	–	–	–	4	1	5	5.7s
France	5	–	–	–	–	2	–	7	8.0ll
Germany	42	12	–	6	4	10	–	74	84.1hh
Italy	–	–	–	–	–	1	–	1	1.1ll
S. Netherlands	–	–	1	–	–	–	–	1	1.1ll
Period:
Medieval	42	–	1	4	2	4	1	57	64.8hh
Early modern	5	12	–	2	2	13	–	31	35.2ll
Total no.	47	12	1	6	4	17	1	88	100.0
Total %	53.4	13.6	1.1	6.8	4.5	19.3	1.1	100.0
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. The N. Netherlands and Spain have zero observations, significantly lower (at the 0.05 level) than their percentage of observations in the guilds database. Austria has zero observations, significantly lower (at the 0.10 level) than its percentage of observations in the guilds database. For all other countries not represented in this table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Qualitative guilds database: 88 observations of guilds’ regulation of daughters’ work.

The qualitative database contains 88 observations of guilds’ treatment of masters’ daughters, shown in Table 5.18. These observations span over five centuries from 1211 to 1771, but the medieval period is significantly over-represented, testifying to a particularly acute concern to clarify daughters’ role in medieval guilds. The cases come from only five European societies, with Germany accounting for 84 per cent, significantly and substantially higher than its 28 per cent share of the guilds database, while France, Italy, Spain, and the two parts of the Netherlands are significantly under-represented. The intense concern about this issue in German guilds makes it clear that references to masters’ daughters in guild records can be interpreted in different ways. Medieval German guilds frequently granted partial rights to masters’ daughters, which permitted them to work while they were still unmarried; this could be regarded as a positive indicator, except that it also suggests that the status of masters’ daughters was a matter for concern which guilds needed to clarify. By contrast, the many occurrences of this issue among German guilds in the early modern period are unambiguously negative, forbidding masters’ daughters to work at all.

Guilds sought to balance the benefits to masters of having a daughter’s assistance against the competition daughters posed to journeymen, apprentices, and poorer guild masters. Thus about half the observations in Table 5.18 are of guilds that granted full or partial rights to masters’ daughters, usually permission to practice the occupation so long as they remained in the unmarried state—a variant of the widows’ rights permitted by Type 3 guilds. Almost all of these cases are from medieval Germany, plus a handful from early modern France.

Another 14 per cent of cases involve guilds allowing daughters to work at an occupation, but only in their role as masters’ offspring, not as independent producers. German guilds were often driven to spell out the status of masters’ daughters because of pressure to exclude them. These pressures were described clearly in 1771 by the foreman of the Hochberg linen-weavers’ guild, which regulated urban and rural weavers alike:

[N]owadays, when a master sets his bodily daughter at the loom, although this is rare, he cannot be told not to do so, and also the journeyman who works alongside her is not penalized anywhere. But if the master were to use another female in addition, one who did not belong to his household, this would absolutely not be put up with, the journeyman would go off on the tramp—or, as soon as he came into another guild jurisdiction, he would infallibly have to let himself be severely punished for this, and leniency here in this country would not help the journeyman outside afterwards. . . . [I]f the journeymen and nationals of these parts wished to complete their tramping years outside, such leniency might put them in a very awkward position, especially in Imperial Cities.¹⁵³

Permitting work by daughters as well as wives brought considerable benefits to the small nuclear-family households that characterized most central, western, and northern European societies. Nonetheless, the remaining 33 per cent of observations in Table 5.18 involve restrictions or prohibitions on work by masters’ daughters. In 8 per cent of cases, guilds did not ban daughters’ work completely, but limited what they could do, as in 1427, when the Munich butchers’ guild forbade masters’ daughters “to stand in the butcher’s bench and sell meat”,¹⁵⁴ or in 1607 when the Frankfurt cord-makers’ guild yielded to journeyman pressure by limiting masters’ daughters to “finishing small tasks”.¹⁵⁵ But in 25 per cent of cases, guilds yielded to the anxieties of male guild members by completely banning daughters from the workshop, as did the York weavers’ guild in 1578¹⁵⁶ and the Rome wool-weavers’ guild in 1758.¹⁵⁷

Like other limits on women’s work, restrictions on daughters were enforced in practice. In 1605, for example, the Memmingen glaziers’ guild fined a master for letting his daughter carry out glazing tasks outside the household,¹⁵⁸ and in the 1720s, the Augsburg ribbon-makers’ guild banned daughters’ work because its journeymen were being ostracized in other cities for working alongside females.¹⁵⁹ So visceral was the resentment by guild journeymen against women’s work that it passed into popular culture. There is an English ballad from the 1720s in which a journeyman weaver ascribes his “want of Work” to masters employing female family members.¹⁶⁰ Daughters were the next-most-tolerated females in the guild system, after wives and widows, but even they were vulnerable to attack by male guild members, particularly journeymen, who viewed them as a threat.

GUILDS AND FEMALE SERVANTS

Given how guilds restricted masters’ wives and daughters, it is not surprising that they were even harsher when it came to female servants. Co-resident members of the master’s biological family were at least regarded as potentially “zunftfähig”, the German legal term for someone for whom guilded work was in principle “befitting”, at least under a master’s authority. But female servants, although they lived in masters’ households, were increasingly defined as “zunftunfähig”, persons for whom guilded work was in principle “unbefitting”. This was the same status guilds and jurists ascribed to persons to whom guilded work was supposed to be unbefitting because they were Jews, gypsies, bastards, members of other religions, cultures or ethnicities, or “dishonourable”, as we saw in Chapter 3.¹⁶¹

This does not mean that female servants were ignored by guilds. On the contrary, the qualitative database contains 78 observations of guilds’ attempts to regulate the work of female servants between 1344 and 1771, shown in Table 5.19. About 29 per cent date from the medieval period, almost identical to the medieval share of the wider guilds database. However, the sample shows a familiar geographical clustering, with Germany accounting for 73 per cent of cases, significantly higher than its 28 per cent of observations in the overall guilds database. England is also over-represented, with 18 per cent of observations, compared to only 9.6 per cent of the guilds database, a smaller but statistically significant difference. Conversely, France, Italy, the Northern Netherlands, and the Southern Netherlands are under-represented. Although, therefore, guild restrictions on female servants’ work are observed in six European societies, they cluster predominantly in Germany, whose guilds show unusually acute concern over this issue, almost entirely aimed at restriction.

TABLE 5.19: Guilds and Female Servants, 1344–1771
	Permitted to work by all-female guild	Permitted to work by normal guild	Restricted in their work	Forbidden to work	Forbidden to work or be trained	Forbidden to work unless apprenticed	Total
							no.	%
Country:
England	–	–	–	13	–	1	14	17.9hh
France	–	–	3	–	–	–	3	3.8ll
Germany	2	1	19	30	3	–	55	70.5hh
Italy	–	–	–	2	–	–	2	2.6ll
Spain	–	–	–	1	–	–	1	1.3l
Switzerland	–	–	1	2	–	–	3	3.8s
Period:
Medieval	2	1	5	14	–	1	23	29.5s
Early modern	–	–	18	34	3	–	55	70.5s
Total no.	2	1	23	48	3	1	78	100.0
Total %	2.6	1.3	29.5	61.5	3.8	1.3	100.0
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. l = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. The N. Netherlands and the S. Netherlands have zero observations, significantly lower (at the 0.05 level) than their percentage of observations in the guilds database. For all other countries not represented in this table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Qualitative guilds database: 78 observations of guilds’ regulation of female servants’ work.

In contrast to the treatment of masters’ wives and daughters, it was very rare for guilds to permit female servants to do guilded work. Of the three such cases in Table 5.19, two are in all-female guilds, where one might expect female servants to be permitted to work. This does not mean that female servants were inactive in guilded occupations, however, as is demonstrated by the 96 per cent of observations in Table 5.19 in which they were clearly working sufficiently to motivate guilds to restrict them. In about 30 per cent of cases, guilds did not wholly ban their work, but instead forbade them to do certain tasks, as in 1716 when the Augsburg gold-beaters’ guild ordered that henceforth gold-inlaying should be reserved for journeymen or masters’ wives and daughters, “because they are fit [fähig] for craft privileges”, but should be forbidden to maidservants because they “are unfit [unfähig] for craft privileges”.¹⁶² But in 65 per cent of cases, guilds banned female servants outright from any participation in the occupation, often in response to pressure from male journeymen and masters who regarded female servants as a threat to their monopoly over paid guild work. The reasoning behind such prohibitions was explained by the Augsburg bookbinders’ guild in 1720, when it banned female relatives and servants from the workshop on the grounds that “once they are released from service, they settle down secretly in corners and continue this work they have cribbed, and in so doing cause no little loss and damage to our profession”.¹⁶³

Guild restrictions on female servants’ work were no mere formalities but were, here again, enforced in practice. In Nuremberg in the 1530s, for instance, the journeymen glovers objected to the activities of female servants in the workshops; they communicated their concerns to glovers’ guilds in other cities such as Strasbourg; the Strasbourg glovers’ guild persuaded their city council to write to the Nuremberg council demanding that female servants be forbidden to work at the craft; ultimately, the Nuremberg city council and glovers’ guild responded by banning female servants from glovers’ workshops.¹⁶⁴ Hatters’ guilds were even harsher towards female servants, to such a degree that after c. 1500 German hatter journeymen ceased to include the Northern Netherlands in their tramping route, because Dutch hatters still employed female servants in the workshop.¹⁶⁵ In 1649, the Frankfurt hatters’ guild boycotted journeymen from Fulda where female servants were still allowed to assist in hatting, whereupon the ruling Abbot of Fulda ordered Fulda’s hatters to stop employing females in any capacity.¹⁶⁶

What about other female household members? Such women were treated much like female servants, even when they were close relatives of the master. Around 1560, for instance, the Strasbourg belt-maker Hans Kranicher violated guild rules by letting his stepdaughter work at the bench. The guild and town council responded by forbidding Kranicher ever to let another female relative work for him and ordering journeymen to ostracize his workshop for the next two years. All five belt-maker journeymen then working in Strasbourg left the city and refused to return, belt-maker journeymen from other German towns boycotted Strasbourg, belt-makers’ guilds throughout Germany became involved in the conflict, and even after the passage of ten years and a multi-guild agreement negotiated in Frankfurt, no belt-maker journeyman in the Empire was willing to risk defilement by working in Strasbourg.¹⁶⁷

GUILDS AND INDEPENDENT FEMALES

Guilds treated independent women even more severely, as emerges from 64 observations in the qualitative database, spanning the five centuries from 1321 to 1822. These are shown in Table 5.20. Contrary to the idea that guild restrictions on women were an early modern innovation, the medieval period accounts for 52 per cent of cases, significantly more than its 27 per cent share of the overall guilds database. Although the observations come from 10 different European societies, France is significantly under-represented and, as with so many other guild restrictions, Germany is significantly over-represented.

None of the observations in Table 5.20 show guilds adopting a permissive attitude towards independent females. The only distinction is between guilds that merely restricted such women’s work (5 per cent of cases), those that prohibited it completely (78 per cent), and those that went so far as to attack it through legal proceedings or violent action (17 per cent). The target also differed, with about half of the cases (47 per cent) involving independent females who competed with guild masters, over one-third (36 per cent) involving women who competed with male journeymen and apprentices, and the final 17 per cent involving women who competed with both masters and journeymen.

As guild control of an occupation tightened, the work of independent women tended to contract. Creation of a new guild often led to the ejection of independent female producers, as with the establishment of the Leiden woollen-cloth-finishers’ guild in 1508,¹⁶⁸ the Malmö drapers’ guild in 1526,¹⁶⁹ the Württemberg Black Forest worsted-weavers’ guilds in the 1580s and 1590s,¹⁷⁰ the Madrid trimmings-makers’ guild in 1600,¹⁷¹ the Bristol tobacco-pipe-makers’ guild in 1652,¹⁷² and the Turin button-makers’ guild in 1737.¹⁷³ The establishment of a guild could have a quantitatively important impact on female industrial activity: one-fifth of all tobacco-pipe-makers in Bristol were female before a new guild ejected women from the craft in 1652;¹⁷⁴ Turin had a large body of female button-makers with their own apprentices before a new guild closed down their businesses in 1737.¹⁷⁵

TABLE 5.20: Guilds and Independent Females, 1321–1822
	Guild restricts independent females as:		Guild prohibits independent females as:		Guild attacks independent females as:			Total
	independent producers	employees	independent producers	employees	free-lance workers	producers in adjacent crafts	factory workers	no.	%
Country:
Denmark	–	–	1	–	–	–	–	1	1.6s
England	–	–	3	1	–	2	–	6	9.4s
France	–	–	–	1	–	2	–	3	4.7ll
Germany	1	1	11	13	7	2	1	36	56.3hh
Italy	–	–	1	4	2	1	–	8	12.5s
N. Netherlands	–	–	2	–	–	–	–	2	3.1s
Scotland	–	–	–	–	–	1	–	1	1.6hh
S. Netherlands	–	–	–	1	1	–	–	2	3.1s
Spain	–	1	–	–	–	1	–	2	3.1s
Switzerland	–	–	1	–	1	1	–	3	4.7s
Period:
Medieval	1	1	11	13	6	1	–	33	51.6hh
Early modern	–	1	8	7	5	9	1	31	48.4ll
Total no.	1	2	19	20	11	10	1	64	100.0
Total %	1.6	3.1	29.7	31.3	17.2	15.6	1.6	100.0
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. For all countries not represented in this table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Qualitative guilds database: 64 observations of guilds’ regulation of independent females’ work.

Where guilds already existed, bans on females arose from male guild workers’ concerns about competition. Complaints by journeymen led to the exclusion of females from dyeing by the Ghent blue-dyers in 1402,¹⁷⁶ from wool-combing by the Wildberg worsted-weavers in the 1740s,¹⁷⁷ and from silk weaving by the Lyon silk-workers between 1751 and 1786.¹⁷⁸ Petitions from poor guild masters led to the exclusion of females from casual labouring jobs by the Lüneburg carpenters and masons in 1563,¹⁷⁹ and from weft-making by the Württemberg worsted-weavers in 1611.¹⁸⁰ In eighteenth-century Barcelona, the silk-twisters’ guild banned giving work to independent females “as long as male silk twisters were unemployed”.¹⁸¹

Guilds did not just restrict such women in their ordinances, but harassed them in practice, as shown by the many conflicts underlying the observations in Table 5.20. These included formal complaints by guilds to the authorities,¹⁸² desperate petitions and appeals by the women,¹⁸³ guild fines against women and their employers,¹⁸⁴ prosecutions in municipal courts,¹⁸⁵ correspondence between guilds of different cities,¹⁸⁶ the blacklisting of journeymen who worked alongside such women,¹⁸⁷ violent attacks by guild members against workshops that employed women,¹⁸⁸ and the disappearance of entire groups of female producers from the written record.¹⁸⁹

Spinners were treated differently, and some scholars use their example to argue that guilds were not inimical to women’s work.¹⁹⁰ Tens of thousands of freelance female spinners were indeed the backbone of the textile sector, the largest industry in pre-modern Europe. But the fact that women were allowed to work as spinners does not mean that guild bans on females in other craft jobs did no harm. Institutional rules that segment labour markets by excluding workers from particular jobs reduce the wages which excluded workers would earn in other sectors by reducing their outside options. In many places spinning became the only craft work available to women outside guild masters’ households, forcing them to accept the low wages offered in the sole sector open to them.¹⁹¹

Some guilds went even further and capped spinners’ pay, as in fifteenth-century Lüneburg¹⁹² and seventeenth-century Württemberg.¹⁹³ Similar initiatives by weavers’ guilds in early modern Thuringia were only blocked by state and merchant intervention.¹⁹⁴ In eighteenth-century Spain, likewise, women were allowed to weave ribbons but were obliged to sell them to male guild masters, “who paid them a very low price, and took all the profit”. Only in the 1770s when new regulations allowed women to work for themselves rather than for guild masters did the earnings of the female ribbon-weavers rise: “now women can sell their work, or work for themselves, [so] they ask for a fair compensation”.¹⁹⁵

DID THIS MATTER?

These empirical findings on how guilds treated women make it possible to answer the five questions posed at the beginning of this chapter. Did the guild system provide generously for women via all-female guilds, mixed-sex guilds, and widows’ rights? Did guild restrictions on women’s income-earning work have much impact, given that physical weakness and childbearing might have made females unproductive outside the household anyway? Did the few women who wanted to learn vocational skills and earn their livelihoods in crafts and trades simply do so informally, evading guild rules or using different institutional mechanisms? Did guilds merely reflect invariant cultural norms about women’s proper role in the economy, which would have prevailed regardless of the institutional framework? Even if guilds did harm some individual women, how much impact did this have on the wider economy, based as it was on the gainful employment of men with families to support?

Did the Guild System Provide Well for Women?

Did the guild system actually provide well for women, as some have argued, through all-female guilds, mixed-sex guilds, and widows’ rights in male guilds?¹⁹⁶ The evidence in this chapter suggests that this view is too optimistic.

All-female guilds were vanishing rare. Across Europe as a whole, only 55 have ever been identified, comprising less than one tenth of one per cent of all guilds. They were only observed in eight European societies; and in two of these there was only one all-female guild, ever. Even in societies where all-female guilds existed, they were to be found only in a handful of cities, typically very large ones. And in those few cities where they existed, they made up only 5 to 6 per cent of all guilds. Many were short-lived, arising in unusual political and institutional circumstances and then disappearing after a few years when conditions changed. Virtually every all-female guild was run by male officials. The limits the guild system placed on females are evidenced even by the exceptional 55 all-female guilds recorded in the literature, given their tiny numbers, short lifespans, narrow prevalence, and lack of female decision-making.¹⁹⁷

Mixed-sex guilds, though more widespread than all-female ones, were also extremely rare. They were nonexistent across vast swathes of the European guild system, and accounted for less than half of one per cent of all guilds. Their membership was still predominantly male, with an average of just 22 per cent independent female masters and another 7 per cent masters’ widows. They often discriminated against their female members, constraining their right to keep apprentices, limiting their workshops and equipment, requiring them to apply for special licenses, restricting them to certain types of work, forbidding them to hold guild office, and excluding them from guild assemblies, voting, and sociability.

All-female and mixed-sex guilds certainly offered women much better opportunities than did all-male guilds. But together they comprised less than half of 1 per cent of all guilds in pre-modern Europe. They do not support the argument that the guild system provided well for women.

All-male guilds which granted some rights to masters’ widows were the dominant European guild type. But rights for masters’ widows did not necessarily imply a strong position for women in general. Widows’ rights did not benefit the other women who might want to work at any particular occupation. They did not even benefit widows of men of other guilds in the town, as in 1791 when the Mainz retailers’ guild prohibited a guilded shoemaker’s widow from opening a shop to sell English porcelain.¹⁹⁸ Even widows of guild members were not usually treated the same as male masters. Taking up their guild rights depended on fulfilling conditions not required of male masters: refraining from remarrying, observing time-limits, having children, having a son, being the right age, paying extra fees, proving “honourable” behaviour, or retaining the favour of existing guild members. Even if a widow satisfied all of the conditions, she often faced limits on keeping apprentices, the size of her workforce, the amount of equipment and raw materials allowed her, her type of output, and her participation in guild activities; and on top of that, she could be obliged to operate with expensive journeymen or under male supervision. Guild conditions and restrictions made it difficult for many masters’ widows to stay in business, as is shown by widows’ low share in guild membership (8 per cent) relative to their household headship rates (18 per cent). The empirical findings on widows’-rights guilds, just as for all-female and mixed-sex guilds, cast doubt on the idea that the European guild system provided well for women.

What about Biology?

What about the view that guilds just treated women in accordance with their natural, biologically determined role? In a pre-mechanical era, it might be argued, physical strength determined economic roles and women’s natural weakness made them too unproductive to specialize in income-earning work. In a pre-contraceptive society, the argument continues, women’s reproductive roles prevented them from working outside the home and made human capital investment unrewarding, further reducing women’s productivity. The guild system therefore did not discriminate against females, the argument concludes, but simply reflected the natural order of things. According to this view, it did not matter that guilds restricted females’ work and training, because “women were mostly restricted to activities learned informally at home”.¹⁹⁹ The evidence in this chapter shows that this view is problematic.

REPRODUCTIVE AND DoMESTIC RESPONSIBILITIES

Some guilds indeed adduced women’s reproductive and domestic roles as an excuse to restrict their work. In 1290, the Paris carpet-makers’ guild forbade women to make carpets, claiming it might harm them in pregnancy.²⁰⁰ In 1611, the Württemberg worsted-weavers’ guild forbade girls to make and sell wefts, “in order that such daughters shall be kept to other and necessary domestic tasks and business”.²⁰¹ In 1624, the Frankfurt stonemasons’ guild forbade a widow to continue the family workshop, adjuring her to earn a living by “taking care of expectant mothers or some other appropriately female occupation”.²⁰² An eighteenth-century Spanish writer censured women’s work in retailing on the grounds that it made them neglect “all the work that keeps them within their houses”.²⁰³ In 1787, the Santander authorities forbade women to produce and sell tallow, on the grounds that it made them abandon

the assistance and care of their respective fathers, husbands, and children . . . and even implying the loss of their children, and the possible break-up of their marriages, and that the day-labourer husband or artisan lacks the relief and rest which should comfort his labours and which he should find in the domestic industriousness of this wife and daughters.²⁰⁴

In 1803, the Berlin tailors’ guild banned women’s work on the grounds that

[m]arried women must be maintained by their husbands, know housekeeping, and care for and educate their children. The unmarried ones may work as domestics (for which there is no lack of opportunity locally) or engage in other feminine occupations outside regular manufacturing.²⁰⁵

Guilds themselves thus originated the arguments used by some modern scholars to justify guild restrictions.²⁰⁶

The evidence, however, casts doubt on such arguments. Marriage and reproduction were not universal for European females. In most northern, western, and central European societies, women started to work in their teens and married only in their mid to late twenties. Between 10 and 20 per cent never married at all, and at any one time more than half of all females of prime working age were not married.²⁰⁷ In some parts of southern and eastern Europe, a woman might typically marry in her early twenties, but often to a much older man, so she ended up supporting a household for many years during an elderly husband’s decline or after his death.²⁰⁸ These demographic realities gave women strong incentives to learn vocational skills, as their typical life-cycle would include long phases of income-earning work outside the household. Even women with children were more productive in crafts and commerce, which could often be practised at home and in combination with childcare, than in the farming, labouring, and burden-bearing jobs they had to take if guilds excluded them.

This can be seen from quantitative findings on income-earning activity by married and widowed women in pre-modern European towns. The observations in Table 5.21 come from five cities in four different societies in northwest and southern Europe at various dates between 1690 and 1810. All but one refer to women who were currently or previously married, and hence more likely to be specializing in household production than income-earning work. In the one sample of women of mixed marital status (at least half of whom were married), 90 per cent reported occupations; across the four samples of widowed and deserted females, an average of 80 per cent reported occupations; across the six samples of married women, the average was 49 per cent. The two samples of women supporting children show 70 per cent with jobs among the married and 72 per cent among the widowed and deserted.

Nor do married women appear merely as “helping” with their husbands’ occupations. In London between 1695 and 1725, less than 10 per cent of married women who reported employment were working with their husbands.²⁰⁹ In the Turin census of 1802, only 48 per cent of married couples were recorded with both spouses practising the same occupation; among couples applying for poor relief in 1802, it was only 36.5 per cent; among couples marrying between 1803 and 1814, it was only 25 per cent.²¹⁰

Women themselves also stated, and revealed by their actions, their preference for practicing crafts and trades rather than doing domestic work. In Cordova in 1499, Juana Fernández declared firmly that she was a household-linen-seller “who states that this is her occupation, without her said husband”.²¹¹ In Barcelona in 1717, the 70-year-old baker Joan Cusso declared, “I could not support myself if my wife, though old, would not work so hard to feed the two of us”.²¹² In Turin in 1769, the married Angela Maria Negro declared “that she [had] always, for the past twenty-two years up to this day, exercised the craft of making silk and gold buttons, enabling her to support herself and her family, composed of eight children, a responsibility that the Father is unable to assume because he is too poor, earning his living as a servant”.²¹³ In 1776, the female linen-drapers of Paris petitioned against the abolition of their guild on the grounds that the linen trade was, for a woman, “the only one where she was obliged neither to rent herself to a greedy entrepreneur nor to submit to a tyrannical associate, disguised under the appellation of ‘husband’.”²¹⁴

TABLE 5.21: Percentage of Women Living from Income-Earning Occupations, 1690–1810
				% living from income-earning occupations among women of:
Locality	Country	Date	Sample of women	mixed marital status	currently married status	widowed/deserted status
Turin	Italy	1690	French residents in Turin (50% married)	90	–	–
London	England	1695–1725	witnesses in 3 London courts	–	60	–
London	England	1695–1725	witnesses in 3 London courts	–	–	85
Turin	Italy	1762–92	women with children applying for poor relief	–	70	–
Turin	Italy	1762–92	women with children applying for poor relief	–	–	72
Antwerp	S. Netherlands	1796	women in city census	–	38	–
Antwerp	S. Netherlands	1796	women in city census	–	–	77
Bologna	Italy	1796	women in welfare-related census	–	63	–
Turin	Italy	1802	women in city census	–	22	–
Turin	Italy	1802	women in city census	–	–	85
Tilburg	N. Netherlands	1810	women in population register	–	40	–
Average				90	49	80
Note: For London 1695–1725, figure is sum of those wholly and partly maintained by income-earning work. For married women: 33% wholly, 27% partly; for widows: 73% wholly, 12% partly. Source: Quantitative guilds database: 11 observations of the proportion of ever-married females reporting income-earning livelihoods.

Given the strong desire evinced by many married women to work at crafts and trades rather than staying at home, it is unsurprising that widows, deserted wives, and spinsters displayed the same tendency. In Barcelona in 1710, the deserted Theresa Salvat supported herself by running the violin-cord-making workshop of her fugitive husband.²¹⁵ In Barcelona in 1716, the widowed Margarita Mensa described how her work as a guilded carter enabled her to support three great-grandchildren, one granddaughter, and “other relatives”.²¹⁶ In 1799, when the Barcelona carters’ guild sought to prevent a master’s widow from practising, she petitioned the crown to make the guild let her work, so that she could support her children.²¹⁷ In 1822, when the Santander authorities told the unmarried Josefa del Río that she should stop working as a fruiterer and instead “occupy herself by going into service”, she replied “that she did not want to be a servant, but to be a fruit seller”.²¹⁸

Where women were allowed to do guilded work, they were recorded as producing a non-trivial share of output. In fourteenth-century York, where the weavers’ guild admitted females to apprenticeship and independent practice, one-quarter of the cloth woven in the city was produced by females.²¹⁹ In eighteenth-century Lyon, although the silk guild did not admit females to training or independent practice, “it was commonly understood that wives of masters produced one-quarter of the workshop’s output under the license of the male head of household”.²²⁰

Women thus did not agree that they were naturally destined for domestic roles and should therefore accept guild restrictions. They often stated explicitly and revealed by their actions that they preferred income-earning work. Quantitative findings confirm that half of married women and four-fifths of widows maintained themselves through working at occupations outside the household. Married women did not merely assist their husbands but pursued separate occupations. Women’s own choices show that they thought they would be productive at the craft and service occupations governed by guilds. By preventing women from legally practising crafts and trades, therefore, guilds were not reflecting “natural” gender roles but rather manipulating markets in their members’ favour at the expense of women with families to support.

LACK OF SKILL?

Another variant of the “biological” argument is that when guilds excluded women, they were simply reflecting the fact that interruptions due to motherhood, or the expectation of such interruptions, made women unable or unwilling to invest in their human capital, since they would not be able to practice any occupation intensively enough to repay the costs of learning.²²¹

Again, this echoes arguments guilds themselves used to suppress female competition. In 1540, the Lyon barber-surgeons’ guild restricted widows to shaving beards and healing wounds, claiming they were insufficiently skilled for more complicated work.²²² In 1566, the Augsburg goldsmiths’ guild declared a total ban on females on the grounds that “women understand nothing of the goldsmiths’ work”.²²³ In 1599, the Munich brewers’ guild petitioned to prevent a widow from selling the beer she made on the grounds that “brewing is a learned art and given to men alone”.²²⁴ Sixteenth-century German hosiers’ guilds prohibited women from using the newly introduced knitting frame on the grounds that it was so complicated that only men could learn to operate it, although in reality the frame made knitting easier; by 1600 the guilds had extended the prohibition, claiming that all knitting was beyond the capacity of females to learn.²²⁵ In seventeenth-century Barcelona, the silk-makers’ guild argued that women should not keep workshops or even sell silk products “because they did not have proper masters’ training”.²²⁶

Empirical findings reveal these arguments to be false. Table 5.22 shows five occupations from which at least one European guild barred women on the grounds that females could not learn the requisite skills. For each of the occupations, the qualitative database attests to the existence of multiple guilds elsewhere, in which women ran workshops as widows or, in some cases, independent masters in all-female or mixed-sex guilds.

In silk-weaving, for instance, despite the claim of the seventeenth-century Barcelona guild that females lacked the skill even to sell silk wares in the market, there were no fewer than 22 guilds elsewhere that permitted women to operate workshops. Most were widows’-rights guilds, of course, but some were among the few all-female or mixed-sex guilds in Europe. One was in Cologne, whose city authorities wrote formally to the city of Antwerp in 1498 to explain “that in our city the craft of silk-making is carried on mainly by women and very rarely by men, with the result that the women are much more knowledgeable about the trade than are the men”.²²⁷ In Bologna, so dominant were females in the mixed-sex guild of the silk-weavers that they constituted over 80 per cent of the mastership in 1796.²²⁸ Even the existence of one such guild, let alone 22, refutes the claim that women lacked the skill to make and sell silk.

TABLE 5.22: Women Running Workshops in Occupations Guilds Claim Are Too Skilled for Females, 1261–1797
	Barber-surgeons	Brewers	Goldsmiths	Hosiers	Silk-weavers	Total
						no.	%
Country:
Austria	–	–	–	–	1	1	1.3s
Denmark	–	–	1	–	–	1	1.9s
England	1	1	1	–	2	5	9.6s
France	3	1	–	–	2	6	11.5s
Germany	3	5	5	–	1	14	26.9s
Italy	3	–	1	2	10	16	30.8hh
N. Netherlands	–	1	–	–	–	1	1.9s
Spain	–	–	1	–	5	6	11.5s
Switzerland	–	–	1	–	1	2	3.8s
Period:
Medieval	1	4	5	–	6	16	30.8s
Early modern	9	4	5	2	16	36	69.2s
Total no.	10	8	10	2	22	52	100.0
Total %	19.2	15.4	19.2	3.8	42.3	100.0
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. The S. Netherlands has zero observations, significantly lower (at the 0.10 level) than its percentage of observations in the guilds database. For all other countries not represented in this table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Qualitative guilds database: 52 observations of women masters in occupations some guilds claim are too skilled for females to practise.

The same is true of the goldsmiths’ craft. The Augsburg goldsmiths’ guild might claim that women “understand nothing of the goldsmiths’ work”, but at least 10 other goldsmiths’ guilds in Europe allowed widows to continue to operate the family workshop, girls to be formally apprenticed, or women to be masters. In Cordova in 1477, a female jeweller called Leonor González described in her will the numerous elaborate ornaments she had made out of precious metals and the large fortune she had amassed through her work.²²⁹ In the workshops of the Augsburg goldsmiths themselves, the demanding arts of making clock cases, clock-keys, clock-chains, and filigree work from gold wire were regarded as the specialties of the masters’ daughters because of the delicacy and skill required.²³⁰

Similarly, the Lyon barber-surgeons might claim that women lacked the skill to practise medicine, but at least 10 barber-surgeons’ guilds in other European cities let widows continue the family business. In Chateaubriand in 1732, when a guilded surgeon sued a woman for practising medicine, she presented numerous attestations from “distinguished people” whom she had cured, including one patient whose gangrenous leg she had healed after a male surgeon planned to amputate it.²³¹ In Frankfurt am Main in 1745, the surgeons’ guild petitioned against 25 non-guilded “encroachers”, 12 of them female.²³² The willingness of patients to patronize these female medical practitioners suggests that the parties with the most at stake did not regard women as lacking expertise.

Women and their families were willing to devote resources to training young women in craft skills, evidently believing that females could learn and practise them productively. Medieval Spanish maidservants’ contracts often specified craft training, as in 1465 when a Cordova couple put their 10-year-old daughter into service with a female weaver to teach her the craft and give her a loom when she was trained.²³³ In 1521, a Malaga embroiderer agreed to train a girl in “making embroideries and worked ribbons, wide and narrow” for a year, in return for a training fee of 6,000 maravedis (roughly 1,250 Euros in 2011).²³⁴ Contracts often stated explicitly that a female apprentice should be trained well enough to practise the craft, as in 1470 when a Cordova mason apprenticed his daughter to a female weaver who was to show her the craft in such a fashion that she “emerges from this [training] well prepared to exercise it [the craft]”,²³⁵ or in 1502 when a Malaga weaver specified that a master was to teach his daughter “the craft of weaving stuffs, so that at the end of the said period of time she will be able to work with him”.²³⁶ Women did not regard their domestic roles as rendering vocational training useless. In 1471 a married woman in Cordova paid a Jew from Seville to train her in the making of ceruse, a costly cosmetic of white lead.²³⁷ In Augsburg in the early 1570s, Barbara Karg described how she and her mother between them had more than 60 years of experience in healing the sick, and how her mother had received formal instruction from “a learned medico”.²³⁸

Contemporaries acknowledged women’s possession of craft skills. In 1608, a Rome apothecary sued the guild officers for inspecting his workshop so invasively that his wife miscarried, depriving him of her essential work in the shop; he claimed to have trained her in the special arts of the confectioner, and formally valued her work at 200 scudi annually.²³⁹ In 1750, an Augsburg bookbinder declared that he valued the work of his wife “as absolutely equal to that of a journeymen”—i.e., an adult male who had completed a guild apprenticeship.²⁴⁰ In 1769, when six members of the London weavers’ guild violently invaded the workshop of Daniel Clark and accused his wife Elizabeth of illegitimately weaving at the loom, she testified that “I did it, they were offended because it was a work too good for a woman to have a hand in.”²⁴¹ In 1781, a Lyon official wrote that, “in every profession where strength is not used as a necessary agent and first mover, it is well known that women perform the craft with more address, assiduity, and perfection than men”.²⁴² Such evidence casts doubt on the idea that women had no use for craft skills and that guild barriers therefore made no difference.

WEAKNESS?

Physical weakness was another “natural” characteristic guilds frequently used to justify restrictions on women, as in 1511 when the Norwich worsted-weavers’ guild excluded female practitioners because of “their lack of strength, resulting in inferior products”,²⁴³ or in 1758 when the Rome wool-weavers’ guild banned women because they “are incapable of lifting the weights involved in our trade”.²⁴⁴

Women’s physical endowments certainly affected their productivity relative to males. Research on twentieth-century populations suggests that the ratio of female to male physical capacity lies in the 0.8–1.0 range for activities, such as running and sit-ups, which involve endurance and moving one’s own body weight, but around 0.5 for activities requiring a one-off effort at, for example, lifting an outside weight. Research on nineteenth-century Britain and America finds that in agriculture the ratio of female to male physical productivity lay in the range 0.61–0.67.²⁴⁵ This implies that females would have been less productive at agriculture and labouring than at crafts and trades. This was the precise opposite of the gender specialization that resulted when guilds excluded females.

Within guilded activities, physiological endowments were only one of several determinants of women’s work. Holding other factors constant, upper-body weakness would create incentives for women to specialize in crafts involving light raw materials, light wares, services, and commercial interactions, and it seems likely that these considerations influenced the clustering of females in textiles, clothing, and retailing which we observed earlier in this chapter. But other factors were seldom held constant. Scarcity of male labour, low female wages, poorer alternative female jobs, better alternative male jobs, the possession of occupation-specific skills, ownership of complementary equipment, the goodwill associated with an existing family business, and many other variables created countervailing incentives for women to engage in heavy work.²⁴⁶

In fact, this is what we observe. The qualitative database contains 355 observations of women engaging in physically arduous activities. As Table 5.23 shows, these observations derive from 16 societies and cover seven centuries, testifying to the pervasiveness of hard physical labour by females.

About 12 per cent of the observations in Table 5.23 involve women and girls carrying out heavy farm work and general labouring, both as family members and as paid workers. They show females ploughing, reaping, raking, threshing, winnowing, carrying thatch, carrying turves, mowing grass, carrying hay, loading manure, turning grain in granaries, stacking peat in the bogs, filling bags with peat, carrying earth, manuring vineyards, harvesting grapes, and carrying drinking water for cattle. Against this background of women’s involvement in heavy farm work and general labouring, it is difficult to argue that they were too weak to do similarly arduous work in crafts and trades.

TABLE 5.23: Females Doing Heavy Work, Twelfth to Nineteenth Century
	Doing heavy farmwork or general labouring	Running the workshop in a heavy craft	Doing the work in a heavy craft	Total
				no.	%
Country:
Austria	1	5	5	11	3.1s
Denmark	–	1	–	1	0.3s
England	18	64	45	127	35.6hh
Estonia	–	6	–	6	1.7hh
Finland	–	–	1	1	0.3s
France	–	11	9	20	5.6ll
Germany	11	56	48	115	32.2h
Hungary	–	–	1	1	0.3s
Italy	–	16	1	17	4.8ll
N. Netherlands	4	13	7	24	6.7s
Norway	1	–	2	3	0.8hh
S. Netherlands	–	1	2	3	0.8ll
Scotland	–	–	1	1	0.3s
Spain	–	5	–	5	1.4ll
Sweden	1	2	1	4	1.1s
Switzerland	6	9	3	18	5.0hh
Period:
Medieval	9	70	44	123	34.5hh
Early modern	33	119	82	234	65.5ll
Total no.	42	189	126	357	100.0
Total %	11.8	52.9	35.3	100.0
Notes: s = not significantly different from percentage of observations in guilds database at 0.10 level. ll = significantly lower than percentage of observations in guilds database at 0.05 level. l = significantly lower than percentage of observations in guilds database at 0.10 level. hh = significantly higher than percentage of observations in guilds database at 0.05 level. h = significantly higher than percentage of observations in guilds database at 0.10 level. Bohemia, Bulgaria and Poland have zero observations, significantly lower (at the 0.05 level) than their percentage of observations in the guilds database. For all other countries not represented in this table, their percentage of observations (zero) is not significantly lower than in the guilds database overall. Source: Qualitative guilds database: 357 observations of females carrying out heavy work.

Unsurprisingly, therefore, over half the observations (53 per cent) involve women operating workshops in heavy crafts. Women ran heavy craft businesses as masters in the few all-female or mixed-sex guilds, as masters’ widows in most ordinary guilds, and as independent businesswomen in suburbs and villages outside guild jurisdiction. We see women operating businesses as anvil-makers, blacksmiths, brass-founders, brewers, bricklayers, brick-makers, cabinetmakers, carpenters, chimney-sweeps, coopers, coppersmiths, crossbow-makers, cutlers, distillers, dyers, founders, fullers, glassblowers, glaziers, goldbeaters, goldsmiths, gunsmiths, hammer-makers, harness-makers, iron-grill workers, iron-makers, ironmongers, joiners, knife-smiths, lime-diggers, masons, metal forgers, millers, nailers, peat-carriers, pewterers, plasterers and limers, plumbers, porters, roofers, saddlers, scythe-smiths, shipbuilders, slaters, smiths, stonemasons, sword-makers, tanners, tilers, tool-makers, turners, wheelwrights, and many more. Of course, not all of these women did the heavy work themselves, since they could employ offspring or servants. But that in itself shows why women’s physical capacities could not justify guilds’ excluding them: the family and the market provided strong labour for women to hire. That women ran workshops in so many heavy crafts shows that what was needed was not so much physical strength as practical knowledge and business experience. These examples do not necessarily mean that it was common for women to run workshops in heavy crafts, but the fact that they did so at all casts doubt on the claim that women’s physical capacities justify guilds in excluding them.

Finally, over one-third of observations in Table 5.23 show women actually carrying out the physical tasks involved in heavy industrial activities. Thus we observe women toiling in the workshops of armourers, blacksmiths, brewers, brick-makers, bronze-casters, cabinetmakers, carpenters, chain-smiths, coopers, coppersmiths, farriers, founders, fountain-masters, fullers, glassmakers, goldbeaters, goldsmiths, harness-makers, iron-goods-makers, joiners, knife-smiths, lime-brewers, nailers, pewterers, roofers, saddlers, shipwrights, silversmiths, smiths, tilers, wallers, wheelwrights, whitesmiths, and writing-table-makers. In the textile sector, we see women dragging linen in the bleaching-fields, carrying loads of raw wool, and transporting yarn and cloth on foot from one town to another. In the building trades, we see them carrying earth, sand and mortar, operating cranes, cutting stone for road-mending, working treadmills, carrying and laying bricks, constructing walls, carrying timber, working in quarries, and carrying roofing tiles. In mining, we observe women dressing metal, rinsing lead, doing stamping work, and transporting wood, coal, ore, and salt. It is evident, therefore, that women had the physical capacities to carry out the tasks in heavy crafts and trades. Again, these examples do not necessarily imply that it was common for women to carry out heavy labour—although in certain times and places, it evidently was. But the fact that they did so at all shows how problematic is the argument that the physical demands of industry were sufficient to justify guilds’ excluding women.

Guild restrictions on women’s work indeed sometimes pushed females into physically more arduous activities. In 1454, for instance, the Danzig joiners’ and cabinetmakers’ guild ruled that women were to be excluded from all workshop activities except for painting, varnishing, and “using the saws”—one of the physically most demanding tasks.²⁴⁷ In 1463, the Nuremberg roofers’ guild ruled that women and girls could be employed to carry heavy loads such as bricks, but not to do light work such as stirring mortar.²⁴⁸ In 1699, the Nuremberg guild of buckle-makers, chain-smiths, and jewellery-smiths forbade females to do finer work and restricted them to wielding the hammer in the smithy, by far the heaviest job.²⁴⁹

The evidence in this section casts a sobering light on claims that pre-modern females were “mostly restricted to domestic activities”. Many women had the desire and capacity to learn occupational skills and use them to earn income, upper-body weakness and reproductive responsibilities notwithstanding. The behaviour of male guild masters reveals that they also regarded women as productive—in fact, too productive—in guilded trades. Throughout medieval and early modern Europe, as this chapter has shown, females without guild licenses were treated as dangerous competitors by male journeymen and masters. Guild harassment of women who worked in violation of guild rules provides perhaps the best testimony of all to women’s ability to learn guilded occupations and practise them productively. The corollary is that when guilds limited women’s work, they were not merely confirming the natural order but in fact were preventing females from allocating their time to those activities that would be most beneficial to them—and, as we shall see shortly, to the economy at large.

Female Agency

What about female agency – the idea that women easily circumvented guild restrictions? As we have seen, women did manage to learn vocational skills and to work at nearly every guilded occupation they were permitted to practise—and many they were not. Some take these findings as testimony to the power of female agency over any institutional rules imposed by guilds.²⁵⁰

Alternative Training Institutions

A first strand of argument is that guild obstacles to female apprenticeship did not matter because girls could get training using other institutional mechanisms. Some scholars have argued, for instance, that females obtained training “first of all [through] the family”.²⁵¹ If that did not suffice, the argument continues, girls had a lavish array of other options, including apprenticeships in all-female and mixed-sex guilds, charitable schools, female religious communities, and non-guilded state manufactories.²⁵²

Does the fact that women got training inside the family mean that being excluded from guild apprenticeship had no negative effects? Even scholars who are enthusiastic about familial training for young women acknowledge that for young men “the family should not be overrated as a source of training”.²⁵³ This makes sense: a major advantage of non-familial training institutions is that they enable people to match their talents with training that will maximize their happiness and productivity, rather than limiting them to the training available in their family of origin. Barring girls from seeking training outside their families cannot have made them better off.

Girls did sometimes enter apprenticeships, as we have seen, but they did this largely outside the guild system. All-female and mixed-sex guilds were extremely rare, and thus provided only a few training niches for girls. Ordinary all-male guilds seldom admitted female apprentices, either banning them outright or taking for granted that they would not put themselves forward to be excluded. Charitable and other non-guild apprenticeships provided an alternative in some societies, particularly England, France, and Italy. But even then, as we saw in Table 5.15, girls secured less than one-third of apprenticeship niches. In most European societies, non-guild apprenticeships were extremely rare and reserved mainly for boys: in over 7,000 pages of surviving church court records for two Württemberg communities between 1646 and 1800, for instance, charity apprenticeships were arranged exclusively for boys and just one informal, private apprenticeship for a girl was mentioned (in sewing).²⁵⁴

What about other sources of training? Alternative training institutions for girls are hard to find. In late medieval Provençal towns, “formal education was restricted to young males. Parallel, informal, institutions may have existed for females, but they are not visible from our sources.”²⁵⁵ In early modern France, the state and church created alternative training institutions for girls precisely because most of the guild system excluded females, but even then the number of training niches remained tiny.²⁵⁶ The alternative training institutions available to females often served girls poorly. The Civic Orphanage in Amsterdam, for example, taught girls to knit and sew in-house while placing boys in apprenticeships with masters.²⁵⁷ Scholars who emphasize alternative training institutions for girls themselves admit that “training by orphanages and similar institutions, as well as academies, seems to have been marginal, in purely quantitative terms, next to the sheer numbers apprenticed by guild masters”.²⁵⁸

Girls sometimes found that guilds hindered even their access to alternative training institutions. In the 1770s, for instance, the Valencia guild of rope-makers, braid-makers, and button-makers strongly opposed the establishment of a school to teach girls “to make buttons or any other craft appropriate to their sex and womanly strengths” for fear it would generate dangerous competitors for male guild members.²⁵⁹ In Württemberg in 1799, when a Teinach cotton manufactory offered to provide girls with a 14-day apprenticeship and then employ them as cotton-spinners, the guild-dominated Wildberg town council objected on the grounds that “the persons here who are capable of such work can earn their livings from spinning wool, which cannot be diminished without disadvantaging the worsted-weavers’ craft”.²⁶⁰ In 1830, the Görlitz tailors’ guild mounted a legal challenge to a sewing-school that taught only seven pupils, on the grounds that such sewing-schools were “illegal establishments of female tailor workshops”.²⁶¹

If alternative training institutions had provided a satisfactory substitute for guild apprenticeships, then girls and their families should not have perceived guilds as an obstacle. Instead, we observe them complaining vociferously about guild restrictions on female apprenticeship. In 1609, for instance, a Madrid braid-maker’s widow lamented loudly when the guild prosecuted her for teaching the craft to her daughter.²⁶² In Geneva in 1696, 15 gold-and-silk-cloth weavers objected strongly when their guild prohibited them from providing training to girls,²⁶³ and in 1700 a Geneva watchmaker protested when his guild forbade him to train his daughter.²⁶⁴ In Augsburg in 1723, the divorced wife of a painter complained when the painters’ guild forbade her to train the daughter of a fellow master.²⁶⁵ If girls could so easily have learned a craft through alternative institutions, why should they, their families, and the craftsmen who wanted to train them have regarded guild restrictions as objectionable constraints?

It is thus unlikely that alternative institutions—whether the family, all-female and mixed-sex guilds, charitable bodies, or spinning and sewing schools—remotely compensated for the exclusion of females from guild apprenticeships. If guild apprenticeships were the best way for boys to obtain vocational training, then why bar them to girls? Conversely, if alternative training institutions were good enough for females, then guild apprenticeships cannot have been essential for males. The fact that some females got training using alternative mechanisms does not mean that women’s opportunities were unharmed by guild restrictions.

HOW GOOD AN ALTERNATIVE IS THE INFORMAL SECTOR?

Female agency is also supposed to have neutralized guild restrictions on women working at crafts and trades, due to the benign properties of the “informal sector”. Even though guilds legally barred females from many activities, it is claimed, women easily circumvented these restrictions by moving into “informal” work—i.e., illegal activity in the black market, which provided rich opportunities to enterprising females. Evidence of women working illicitly is taken as evidence that “there is no essential contradiction between women and guilds”.²⁶⁶

It is true that guild regulations could never be enforced perfectly, and that people restricted by guilds sometimes circumvented the hindrances by working illegally. But the fact that people can be seen making certain choices does not imply that institutional restrictions on those choices have no effect. If a woman made a choice that violated guild rules, she faced the risk of being penalized. This risk did not have to be 100 per cent in order to have a non-zero expected value. Consider the case of a seventeenth-century German servant girl who wove at the loom, counter to guild regulations. We know from the documentary sources that she or her employer faced a fine of three Gulden, equivalent to one year’s salary for a maidservant.²⁶⁷ Even if there was only a 25 per cent chance of being caught, the expected cost of her illegal work was one quarter of a year’s salary—which for some girls must have exceeded the expected benefit. On the margin, some girls would refrain from working illegally and some employers would hesitate to give them work. The same theoretical reasoning applied to an independent unmarried woman “combing wool like a journeyman”, a master’s widow trying to keep her workshop open despite having remarried, or a guild couple illegally training an orphan girl as an apprentice. All carried penalties of fines, confiscation, or social opprobrium. And for all, as archival records show, there was at least some risk of detection. As a result, the expected cost of doing these things was non-zero, and there would therefore be some marginal women and their potential employers who would refrain from making that choice, while others would go ahead. Only if the penalty or the risk of detection for violating guild restrictions were zero would no one’s choices be affected. The fact that some people can be observed making a particular choice does not imply, therefore, that the institutional rules governing that choice had no effect.²⁶⁸

Empirically, too, the evidence in this chapter has shown that women did not circumvent guild regulations without cost. Women sought to operate as independent masters in crafts and trades where there were mixed-sex guilds, but often faced restrictions on their apprentices, workshops, equipment, work permits, products, services, legal privileges, and participation in guild decision-making. Widows sought to continue family workshops but many were unable to satisfy conditions relating to remarriage, time-limits, age, having children to support, having a son to take over, special fees, and reputation clauses. Even those who met the conditions faced limitations on their workforce, workshop, equipment, raw materials, market access, products, services, and say in guild affairs. Masters’ daughters (and many other girls), facing a non-trivial possibility of spinsterhood, early widowhood, or marriage with an unproductive husband, therefore wanted to invest in their own human capital via craft apprenticeship, but were excluded by guilds or not permitted to progress to journeymanship or mastership. Guild regulations were seldom perfectly enforced, but circumventing them was risky. Not all women, especially the poor, could afford to take those risks.

As for the informal sector, certainly it offered excluded groups such as women opportunities better than those provided by formal institutions such as guilds.²⁶⁹ But we should not romanticize the shadow economy. As Dora Dumont found in her work on early modern Italy, women working informally in contravention of guild regulations enjoyed greater flexibility than guild membership allowed, but this does not mean that guilds were good for women. Rather, it “runs the risk of glorifying undesirable circumstances. Even though the apparently powerless managed to exercise control within adverse circumstances, it does not follow that those circumstances were preferable.”²⁷⁰ On the contrary, after the Bologna guilds were abolished in 1796, women eagerly welcomed their access to formal markets—too much so, in the view of officials who in 1803 bemoaned “the insubordination of women in the years following the abolition of the guilds”.²⁷¹ Formal markets in which women could work legally gave them, among other things, security enough to be insubordinate.

Culture Rather Than Institutions?

In a wider perspective, however, it might be argued that women were oppressed because of patriarchal cultural attitudes, to which guilds were irrelevant. Some scholars go so far as to dismiss the idea that guilds limited women’s economic options, arguing that “pre-modern gender discrimination was not invented by, and certainly not restricted to, guilds”.²⁷² Guilds’ ill-treatment of female workers, it is claimed, may have been “merely reinforcing other, possibly more significant social mechanisms”.²⁷³ In any case, the argument continues, “exclusion of women and minorities was a generalized social, religious and cultural phenomenon . . . the absence of guilds did not automatically improve a minority’s lot”.²⁷⁴ This amounts to the claim that institutions do not fundamentally made any difference, since economic outcomes are driven by invariant cultural norms.

However, it hardly makes sense to dismiss the harm caused by A on the grounds that harm was also caused by B. The fact that other pre-modern social mechanisms were also used to restrict women’s work does not mean that it did not matter when guilds did so. Empirically, too, there is little support for the idea that institutional instruments play no role in operationalizing cultural norms. As Francesca Trivellato found for early modern Venice,

guilds did not simply mirror the patriarchal, anti-Semitic, and xenophobic values that imbued society at the time. They also existed precisely to enforce these values, and to maintain and manage those inequalities that were perceived to be “natural” at the time, including those between the genders. Craft guilds were thus essential, not an accessory to the socio-economic fibre.²⁷⁵

Patriarchal beliefs may have been widespread, but they were much more economically effective where institutional mechanisms enabled holders of these beliefs to give them practical expression. The consequence was that, in Judith Brown’s assessment, there was “an inverse relationship between the ability of guilds to regulate economic activity and the extent of female participation in the labor force”.²⁷⁶

Conversely, women moved into almost any economic activity as soon as guild regulations loosened.²⁷⁷ In occupations where guild rules were more liberal, such as retailing, female participation was significantly higher than in those where guilds were more restrictive.²⁷⁸ In societies with similar cultures, such as the Northern and Southern Netherlands, women enjoyed different opportunities in major occupations such as the clothing trades precisely because guilds in the Northern Netherlands had less political influence than those in the Southern Netherlands where male-dominated tailors’ guilds enjoyed “a better position to defend themselves against encroachments on their monopoly”.²⁷⁹

Institutions did matter, as they enabled individuals to organize themselves to work together, for good or ill. Few would argue, for instance, that because beliefs in the virtues of high-quality craft wares and skilled training were universal in pre-modern Europe it did not matter what institutions were available to ensure that such quality and training levels were achieved in practice. Even if patriarchal beliefs were universal, the institutions to implement these beliefs varied. Strong guilds gave male masters the institutional capacity to act collectively to exclude women and to penalize free-riding by individual masters, some of whom—as revealed preferences show—otherwise chose to train female apprentices, employ female workers, and purchase from female suppliers.

Craft guilds’ exclusion of women may not have been an expression of cultural attitudes at all. Guilds provided institutional mechanisms that enabled individual men, regardless of their cultural attitudes, to cooperate to restrict women’s work in order to reduce market competition.²⁸⁰ Gender may often have been just an excuse. In the end, male guild members may not really have cared which gender an outsider was: the main point was to fasten on some trait that could justify restricting competition and maintaining rents for insiders.²⁸¹

Did Guild Constraints on Women Matter for the Wider Economy?

But how much did this matter in the great scheme of things? Guilds’ treatment of females was clearly unfortunate for the women involved, but many would argue that the performance of the economy at large was more important. Guilds may simply have redistributed resources from females to males, but left the efficiency of the economy unharmed.

Doubt is cast on this idea by the research on modern developing societies mentioned at the beginning of this chapter, which has consistently found that excluding females from training, jobs, and business ownership harms the wider economy. For one thing, excluding females from human capital investment causes economic harm. A UN study, for instance, estimated that a 1 per cent increase in female secondary school enrolment in 2000–2004 would have increased per annum GDP growth in the Asia and Pacific region by 0.23 per cent, an estimate paralleled in studies of other developing economies.²⁸² Second, excluding females from the labour force harms developing economies: the same UN study estimated that if the female labour force participation rate in India in 2000–2004 had been the same as in the United States, India’s GDP would have increased by 4.2 per cent a year and its GDP growth rate by 1.08 percentage points a year. Achieving US female labour force participation rates would have increased the Malaysian GDP growth rate by 0.77 of a percentage point and the Indonesian rate by 0.56 of a percentage point.²⁸³ Finally, preventing females from becoming business owners and self-employed entrepreneurs slows economic growth. A 2016 study by Cuberes and Teignier quantified the effects of “frictions” (i.e., instititutional constraints) that keep women from becoming the owners of businesses or self-employed entrepreneurs. Analyzing 106 non-OECD countries in the latest year for which ILO data were available in 2016, the study found that constraints on women’s ability to become business owners and self-employed entrepreneurs reduced national income by 5 per cent in Latin America, 6 per cent in Sub-Saharan Africa, 7 to 8 per cent in Central Asia, the Middle East, North Africa, and East Asia, and 9 to 10 per cent in South Asia. When they added to these the constraints limiting women’s ability to participate in the labour force as employees, the reduction in national income rose to 10 per cent in Central Asia, 12 per cent in Sub-Saharan Africa, 16 per cent in East Asia, 17 per cent in Latin America, 25 per cent in South Asia, and 38 per cent in the Middle East and North Africa.²⁸⁴

This chapter has shown that guilds restricted women’s economic role in precisely these three ways. Most guilds hindered girls’ admission to apprenticeship, restricting their ability to invest in their own vocational human capital. Most guilds restricted women’s ability to participate in the labour force in those industrial and service occupations which guilds regulated. Many guilds restricted women’s ability to continue existing family businesses and most prevented women from operating as independent entrepreneurs. Studies of modern LDCs suggest strongly that imposing such restrictions on women’s access to training, employment, and business ownership not only reduces women’s welfare but damages the performance of the economy as a whole. The limits guilds placed on the economic choices of women, therefore, inflicted harm on the women themselves, on the families they supported, and almost certainly on the economy at large.

Epigraph sources: Beltmakers’ guild ordinance, Depping 1837, 236; Worsted-weavers’ guild ordinance, Hudson and Tingey 1906, II:378 (spelling modernized); Brewers’ guild petition, Wiesner 1995, 89; Wool-weavers’ guild statute, Groppi 2002, 43; Combined guilds of Vienna, complaint, Steidl 2000, 325.

¹ Knowles et al. 2002; United Nations Economic and Social Commission for Asia and the Pacific 2007, 105–106.

² United Nations Economic and Social Commission for Asia and the Pacific 2007, 104–105; Cuberes and Teignier 2016, 21 (Table 5).

³ Willen 1984, 204–205; Ogilvie 2003, ch. 2.

⁴ Ogilvie 2003, ch. 7; Dennison and Ogilvie 2014, 672–76; Gruber and Szołtysek 2016.

⁵ Clark 1919, 196; Wachendorf 1934; Crowston 2001, 160; Crowston 2008, 21–22, 25–26, 34–39, 44; Schmidt 2009, 183.

⁶ Epstein 1998, 687; Epstein and Prak 2008, 10.

⁷ Epstein 1998, 689; Epstein and Prak 2008, 10–11; Crowston 2008, 22, 28–34, 44; Reith 2008, 130, 136; Schmidt 2009, 184.

⁸ Epstein and Prak 2008, 10; Schmidt 2009, 183–85.

⁹ Clark 1919, 196; Crowston 2001, 160; Crowston 2008, 39–43; Schmidt 2009, 182.

¹⁰ As argued recently by Crowston 2008, 22, 44.

¹¹ Crowston 2006, 28.

¹² Schmidt 2009, 173.

¹³ Kowaleski and Bennett 1989, 479–80.

¹⁴ Clark 1919, 196; Wachendorf 1934; Crowston 2001, 160; Crowston 2008, 21–22, 25–26, 34–39, 44; Schmidt 2009, 183.

¹⁵ The all-female “guilds” shown in Table 5.2 include six associations of female stall-holders in fifteenth- and sixteenth-century Poznań and Kraków, which had monopoly privileges but were not formally guilds (see Karpinski 1990, 288). Deceulaer and Panhuysen 2000, 97–98, record that in 1741 the seamstresses of Mons regarded themselves as having an independent guild despite being denied guild rights by the authorities, lacking formal privileges, and being officially subject to the male tailors’ guild; formally, therefore, this organization was a variant of a mixed-sex guild of tailors and seamstresses and it is accordingly not included in Table 5.2.

¹⁶ As a reminder, throughout this book, “significant” means that the null hypothesis of no difference cannot be rejected at the 0.05 level.

¹⁷ Population estimates from http://www.ggdc.net/maddison/maddison-project/home.htm.

¹⁸ On medieval all-female guilds see Wachendorf 1934; on the medieval period as a “golden age” for women see Clark 1919.

¹⁹ Jacobsen 1998, 61.

²⁰ Wensky 1982, 636–41; Jacobsen 1998, 61–62 with n. 13; Howell 1986, 132–33.

²¹ Jacobsen 1983, 11.

²² Howell 1986, 130–31.

²³ Uitz 1998, 51.

²⁴ Jacobsen 1983, 11.

²⁵ Hafter 2007, 221 (fig. 16).

²⁶ Jacobsen 1998, 61–62; Kowaleski and Bennett 1989, 482–83; Wensky 1982, 636, 641; Howell 1986, 129–30; Wesoly 1980, 91 n. 118; Schmidt 2009, 173.

²⁷ Broomhall 2008, 206.

²⁸ Broomhall 2008, 205–206.

²⁹ Broomhall 2008, 206–207.

³⁰ Broomhall 2008, 207.

³¹ Wesoly 1980, 94–95.

³² Wesoly 1980, 92; Rippmann 1984, 334–35; Schötz 1997, 162; Schmidt 2009, 175.

³³ Wesoly 1980, 102–103, 108–110.

³⁴ Roper 1985, 52–53; Wesoly 1980, 99; Lesemann 1994, 57; Bennett 1992, 160; Kowaleski and Bennett 1989, 478, 487; Dixon 1895, 226; Wright 1985, 106–107; Geering 1886, 95–97.

³⁵ Bennett 1992, 160; Kowaleski and Bennett 1989, 478, 487; Dixon 1895, 226; Wright 1985, 106–107; Wesoly 1980, 91.

³⁶ Kowaleski and Bennett 1989, 477–78.

³⁷ As pointed out in Roper 1985, 18–19; Wesoly 1980, 92.

³⁸ Schmidt 2009, 174–75.

³⁹ Schmidt 2009, 175.

⁴⁰ Wesoly 1980, 92

⁴¹ Wesoly 1980, 105–106.

⁴² This particularly applies to the 120 observations for Germany and Switzerland that derive from the classic work of Wachendorf (1934), which accounts for 35 per cent of the 343 observations. Wesoly 1980, 89–94, investigated a number of the guilds identified by Wachendorf as mixed-sex, and concluded that the categorization had often been optimistic and sometimes completely mistaken.

⁴³ Wachendorf 1934; Wesoly 1980, 89–94.

⁴⁴ For surveys, see Bosch 1962, 324–27, 333, 347; Schama 1987, 402–404, 407–12; De Vries and Van der Woude 1997, 598–601; Dekker 1998, 167, 171, 176; Ogilvie 2003, 344–46.

⁴⁵ Davies 1994, 261–62 (London); Locklin 2007, 54 (Rennes); Schmölz-Häberlein 2004, 100 (Hochberg); Bog 1968, 76 (Windsheim); Rodocanachi 1894, 2:148 (Rome); Zucca Micheletto 2014, 153 with n. 92 (Turin).

⁴⁶ Aclocque 1917, 51.

⁴⁷ Groppi 2002, 42.

⁴⁸ Wiesner 1986, 175–76; Uitz 1991, 96; Groppi 2002, 42.

⁴⁹ Howell 1986, 135, 243 n. 45; Aclocque 1917, 51; Schmidt 2009, 181; Locklin 2007, 49.

⁵⁰ Howell 1986, 135, 243 with n. 56.

⁵¹ Quoted in Wiesner 1986, 175.

⁵² On England, Kowaleski and Bennett 1989, 477–78; Bennett 1992, 160. On France, Locklin 2007, 49; Gibson 1981, 88, 94. On Italy, Dumont 1998, 19; Sabbatini 1998, 237; Groppi 1998, 384; Trivellato 2008, 214. On Spain, Córdoba de la Llave 1993, 111.

⁵³ Trivellato 2008, 214.

⁵⁴ Groppi 1998, 384.

⁵⁵ On France, Locklin 2007, 54. On Italy, Zucca Micheletto 2014, 179.

⁵⁶ Kluge 2007, 135.

⁵⁷ Clark 1919, 196; Ehmer 2001 [Family], 196 (quotation).

⁵⁸ It thus provides an urban counterpart to the study of 278 European villages between 1427 and 1895 in Ogilvie and Edwards 2000, 970–71, which found average rural female headship of 13.8% across the period (17.5% in the fifteenth century, 13.5% in the sixteenth, 15.0% in the seventeenth, 13.5% in the eighteenth, and 13.0% in the nineteenth).

⁵⁹ Beier 1717, 35ff. Steidl 1998, 121, shows that Beier’s characterization of the guild master’s state-delegated functions of “surveillance and monitoring” (Wachen und Gaffen) referred to regulation, public order, and moral surveillance, particularly over male workers.

⁶⁰ See, e.g., Wiesner 1986, 162.

⁶¹ Ogilvie 2003, 259; Kowaleski and Bennett 1989, 479; Stannek 1998, 101; Sogner and Sandvik 1990, 635; Musgrave 1997, 158.

⁶² Depping 1837, 179 n. 1.

⁶³ Shahar 1983, 198.

⁶⁴ Ogilvie 2003, 264.

⁶⁵ Groppi 2002, 44.

⁶⁶ Groppi 2002, 44.

⁶⁷ Kluge 2007, 139 n. 656.

⁶⁸ Ehmer 2001 [Family], 196.

⁶⁹ Roper 1989, 53.

⁷⁰ Shaw 2006, 174–75.

⁷¹ Groppi 2002, 43.

⁷² Uitz 1998, 49.

⁷³ Uitz 1991, 97.

⁷⁴ Ogilvie 2003, 260; Stuart 1999, 213.

⁷⁵ On London, see Davies 1994, 192, 262; Rappaport 1989, 41. On York, see Lipson 1915, I:318; Palliser 1972, 100.

⁷⁶ Solà 2016, para. 20.

⁷⁷ Shahar 1983, 198.

⁷⁸ Werkstetter 2004, 164, 168.

⁷⁹ Steidl 2016, para. 24.

⁸⁰ Shahar 1983, 198.

⁸¹ Steidl 2007, 148 (Vienna); Davis 1982, 68 (Lyon); Zucca Micheletto 2014, 142–43, 145 (Turin); Solà 2016, para. 20 (Barcelona); Williams 1907, 3:58 (Granada); Perry 1990, 17 (Seville); Monter 1980, 203 (Geneva).

⁸² Davies 1994, 192, 262 (London); Lanza 2007, 89 (Paris); Werkstetter 2004, 164, 168 (Augsburg); Wachendorf 1934, 70 (Würzburg); Zucca Micheletto 2014, 153 with n. 93 (Turin).

⁸³ Steidl 2007, 148 (Vienna); Ben-Amos 1991 [Women], 239 (Bristol); Prior 1985, 103–109 (Oxford); Lipson 1915, I:318 (York); Palliser 1972, 100 (York); Aclocque 1917, 49 (Chartres).

⁸⁴ Vicente 1996, 132

⁸⁵ Nicholas 1995, 1123.

⁸⁶ Wiesner 1986, 163.

⁸⁷ Lesemann 1994, 63.

⁸⁸ Córdoba de la Llave 1993, 112 n. 45; Manninen 1984, 160.

⁸⁹ Roper 1989, 52, 63.

⁹⁰ Manninen 1984, 160.

⁹¹ Ogilvie 2003, 260.

⁹² Werkstetter 2004, 178.

⁹³ Roper 1985, 77; Roper 1989, 52.

⁹⁴ Musgrave 1997, 164–65.

⁹⁵ Zucca Micheletto 2013, 241 (quotation), 243; this petitioner was a married woman whose husband was too poor to support her, but the guild also barred a master’s widow from operating a workshop unless she paid a male worker certified by the guild.

⁹⁶ Wiesner 1986, 163.

⁹⁷ Werkstetter 2004, 169 n. 35.

⁹⁸ Roper 1985, 62–63

⁹⁹ Wachendorf 1934, 102–103.

¹⁰⁰ Wachendorf 1934, 68.

¹⁰¹ Roper 1989, 50.

¹⁰² Davis 1982, 69.

¹⁰³ Musgrave 1997, 157.

¹⁰⁴ Hufton 1984, 364–65.

¹⁰⁵ Weisser 1780, 183; Roper 1989, 40, 49; Kowaleski and Bennett 1989, 478–79; Dambruyne 1998, 53.

¹⁰⁶ Fischer 1966, 66.

¹⁰⁷ Jacobsen 1983, 9.

¹⁰⁸ Clasen 1981, 23.

¹⁰⁹ Stahlschmidt 1971, 186–87.

¹¹⁰ Farr 1988, 83, 105.

¹¹¹ Hill 1989, 92.

¹¹² Roper 1989, 46 (quotation), 50.

¹¹³ Dambruyne 1998, 53 (Table 5).

¹¹⁴ Massa 1998, 257 on Genoa; see Van den Heuvel 2008 on women’s prevalence in the retail sector.

¹¹⁵ Depping 1837, 408–10.

¹¹⁶ Schmidt 2009, 176, 187 n. 39.

¹¹⁷ Depping 1837, 398.

¹¹⁸ Wachendorf 1934, 102–103.

¹¹⁹ Hafter 1995, 48–52.

¹²⁰ Van Nederveen Meerkerk and Schmidt 2008, 725.

¹²¹ Wachendorf 1934, 80.

¹²² Boissonnade 1900, II:71.

¹²³ Lipson 1915, I:317; Bennett 1992, 159; Karras 2004, 96.

¹²⁴ Uitz 1991, 98.

¹²⁵ Stannek 1998, 103 n. 8.

¹²⁶ Pfister 2004 [Craft Guilds], 301.

¹²⁷ Wachendorf 1934, 66; Wiesner 1986, 167; Wesoly 1980, 113.

¹²⁸ Ehmer 2001 [Artisans], 817–18; Ehmer 2005, 70–71.

¹²⁹ Ogilvie 1997, 176.

¹³⁰ Medieval observations account for 18.3 per cent of Table 5.14, compared to 27 per cent of the guilds database overall, a difference significant at the 0.092 level.

¹³¹ Kowaleski and Bennett 1989, 486 n. 34.

¹³² Monter 1980, 202–203.

¹³³ Shahar 1990, 518; Kammeier-Nebel 1996, 85; Wiesner 1996, 93; Mottu-Weber 1990, 349–50; Prior 1994, 138; Mendelson and Crawford 1998, 329; Roberts 1985, 142; Bennett 1992, 160; Kowaleski and Bennett 1989, 476–77; Schlenkrich 1998, 114.

¹³⁴ Crowston 2008, esp. 30–34.

¹³⁵ On the variety of institutional mechanisms generating and recording female apprenticeships in England alone, see Simonton 1988, esp. ch. 5.

¹³⁶ Hill 1989, 88–89.

¹³⁷ Béghin 1996, 8.

¹³⁸ Davis 1982, 68.

¹³⁹ Mottu-Weber 1990, 348.

¹⁴⁰ Werkstetter 2004, 172–73.

¹⁴¹ Clark 1919, 196; Crowston 2001, 160; Crowston 2008, 39–43; Epstein 2008, 167; Schmidt 2009, 182.

¹⁴² Howell 1986, 74.

¹⁴³ Howell 1986, 73, 91.

¹⁴⁴ Stannek 1998, 103 n. 8.

¹⁴⁵ Schöller 1994, 319.

¹⁴⁶ Stannek 1998, 103.

¹⁴⁷ Howell 1986, 73, 91.

¹⁴⁸ Willen 1984, 210.

¹⁴⁹ Vicente 1996, 132.

¹⁵⁰ Anishanslin 2016, 155–56.

¹⁵¹ Clark 1919, 196; Crowston 2001, 160; Crowston 2008, 39–43; Epstein 2008, 167; Schmidt 2009, 182.

¹⁵² Epstein 2008, 167.

¹⁵³ Quoted in Schmölz-Häberlein 2004, 100.

¹⁵⁴ Wachendorf 1934, 53.

¹⁵⁵ Wiesner 1986, 166.

¹⁵⁶ Willen 1984, 215.

¹⁵⁷ Groppi 2002, 43.

¹⁵⁸ Wiesner 1986, 155.

¹⁵⁹ Werkstetter 2004, 171.

¹⁶⁰ Anishanslin 2016, 156–57.

¹⁶¹ Werkstetter 2004, 165 with n. 16.

¹⁶² Quoted in Werkstetter 2004, 165 n. 16.

¹⁶³ Quoted in Werkstetter 2004, 173.

¹⁶⁴ Wiesner 1986, 167.

¹⁶⁵ Stuart 1999, 214.

¹⁶⁶ Wiesner 1986, 167.

¹⁶⁷ Wesoly 1980, 112–13; Wiesner 1986, 168.

¹⁶⁸ Howell 1986, 73.

¹⁶⁹ Jacobsen 1983, 7.

¹⁷⁰ Troeltsch 1897, 10–11; Ogilvie 1997, 90.

¹⁷¹ Zanardi 2013, 143.

¹⁷² Ben-Amos [Women] 1991, 241.

¹⁷³ Cerutti 2010, 574, 578.

¹⁷⁴ Ben-Amos 1991 [Women], 241.

¹⁷⁵ Cerutti 2010, 574, 578.

¹⁷⁶ Hutton 2011, 132.

¹⁷⁷ Ogilvie 2003, 133.

¹⁷⁸ Hafter 1995, 56–57.

¹⁷⁹ Wachendorf 1934, 88.

¹⁸⁰ Ogilvie 2003, 132.

¹⁸¹ La Force 1965, 100.

¹⁸² Wachendorf 1934, 57–58; Simon-Muscheid 1990, 383.

¹⁸³ Stannek 1998, 103; Cerutti 2010, 574, 578.

¹⁸⁴ Ogilvie 2003, 133.

¹⁸⁵ Shaw 2006, 88.

¹⁸⁶ Wiesner 1986, 167.

¹⁸⁷ Schmölz-Häberlein 2004, 100.

¹⁸⁸ Kriedte 1993, 267; Solà 2016, para. 5.

¹⁸⁹ Solà 2016, para. 5.

¹⁹⁰ Epstein 2008, 167.

¹⁹¹ Ogilvie 1990; Ogilvie 2003, ch. 6–7; Ogilvie 2004 [Guilds]; Ogilvie 2004 [How]; Ogilvie 2004 [Women]; Ogilvie 2006.

¹⁹² Bodemann 1883, XLVI.

¹⁹³ Ogilvie 1990; Ogilvie 2003, ch. 6–7; Ogilvie 2004 [Guilds]; Ogilvie 2004 [How]; Ogilvie 2004 [Women]; Ogilvie 2006.

¹⁹⁴ See, e.g., Finkenwirth 1910, 13, 15–16, 20, 33.

¹⁹⁵ Quoted in Sarasúa 1998, 363.

¹⁹⁶ Clark 1919, 196; Wachendorf 1934; Crowston 2001, 160; Crowston 2008, 21–22, 25–26, 34–39, 44; Schmidt 2009, 183; Ehmer 2001 [Family], 196.

¹⁹⁷ Herlihy 1990; Wensky 1980, 30–186.

¹⁹⁸ Garner 2013, 135.

¹⁹⁹ Epstein 1998, 687 n. 10.

²⁰⁰ Depping 1837, 409.

²⁰¹ Quoted in Troeltsch 1897, 446.

²⁰² Quoted in Wiesner 1986, 162.

²⁰³ Quoted in Sarasúa 1998, 361.

²⁰⁴ Quoted in Sarasúa 1998, 362.

²⁰⁵ Quoted in Quataert 1985, 1134.

²⁰⁶ Epstein 1998, 687 n. 10.

²⁰⁷ Hajnal 1965; Dennison and Ogilvie 2014; Ogilvie 2003, ch. 2.

²⁰⁸ Vicente 1996, 132.

²⁰⁹ Earle 1989, 338.

²¹⁰ Zucca Micheletto 2014, 179.

²¹¹ Quoted in Córdoba de la Llave 1993, 104 n. 5.

²¹² Quoted in Vicente 1996, 132, 138 n. 32.

²¹³ Quoted in Zucca Micheletto 2013, 241.

²¹⁴ Quoted in Coffin 1996, 37.

²¹⁵ Vicente 1996, 132, 137 n. 28.

²¹⁶ Quoted in Vicente 1996, 138 n. 31.

²¹⁷ Hunt 2009, 178–79.

²¹⁸ Quoted in Sarasúa 1998, 362.

²¹⁹ Lipson 1915, I:317.

²²⁰ Hafter 2007, 66 (quotation), 133.

²²¹ Epstein 1996, 687 n. 10; Epstein 2008, 167; Epstein and Prak 2008, 10–11.

²²² Davis 1982, 69.

²²³ Roper 1989, 46 (quotation), 50.

²²⁴ Quoted in Wiesner 1995, 89.

²²⁵ Wiesner 1985, 96.

²²⁶ Quoted in Vicente 1996, 131, 137 n. 27.

²²⁷ Quoted in Wachendorf 1934, 46; and Wensky 1982, 639.

²²⁸ Dumont 1998, 6.

²²⁹ Córdoba de la Llave 1993, 110 with n. 34.

²³⁰ Rathke-Köhl 1964, 69.

²³¹ Quoted in Locklin 2007, 60.

²³² Brandt 2003, 12.

²³³ Córdoba de la Llave 1993, 105–106.

²³⁴ Córdoba de la Llave 1993, 106. Currency conversion (1 maravedi in 1556–98 = 0.1875 Euros in 2011) from http://elsexagenario.blogspot.co.uk/2011/12/curiosidades-del-siglo-xvi.html.

²³⁵ Quoted in Córdoba de la Llave 1993, 107.

²³⁶ Quoted in Córdoba de la Llave 1993, 106.

²³⁷ Córdoba de la Llave 1993, 107 with n. 17.

²³⁸ Quoted in Roper 1985, 59.

²³⁹ Cohen 2007, 493.

²⁴⁰ Quoted in Werkstetter 2004, 166.

²⁴¹ Quoted in Anishanslin 2016, 155–56.

²⁴² Quoted in Hafter 2007, 200.

²⁴³ Quoted in Goldberg 1992, 333.

²⁴⁴ Groppi 2002, 43 (quotation); Dumont 1998, 6.

²⁴⁵ For a more detailed discussion of these issues, in the context of early modern European women’s work and male-female wage ratios, see Ogilvie 2003, 289–90.

²⁴⁶ As emerges strikingly from data on female participation in paid physical labour in Sweden between 1550 and 1759, analyzed in Gary 2017.

²⁴⁷ Wachendorf 1934, 80.

²⁴⁸ Wachendorf 1934, 87.

²⁴⁹ Wachendorf 1934, 85.

²⁵⁰ Crowston 2006, 28.

²⁵¹ Epstein and Prak 2008, 10 (quotation), 11.

²⁵² Crowston 2006, 28.

²⁵³ Epstein and Prak 2008, 10.

²⁵⁴ Ogilvie 2003, 99.

²⁵⁵ Bednarski and Courtemanche 2009, 114.

²⁵⁶ Crowston 2006, 14–18.

²⁵⁷ McCants 1997, 70–88.

²⁵⁸ Epstein and Prak 2008, 11.

²⁵⁹ Quoted in MacKay 2006, 165 n. 5.

²⁶⁰ Pfarrarchiv Wildberg, KKP Vol. VIII (1791–1806), fol. 106v–1074, 17.1.1799.

²⁶¹ Quoted in Quataert 1985, 1137.

²⁶² MacKay 2006, 64.

²⁶³ Monter 1980, 202–203.

²⁶⁴ Monter 1980, 203–204.

²⁶⁵ Werkstetter 2001, 315–16; Werkstetter 2004, 172 n. 52.

²⁶⁶ Crowston 2006, 28.

²⁶⁷ Ogilvie 2003, 132.

²⁶⁸ For further exploration of this issue, see Ogilvie 2014 [Choices], 277–305.

²⁶⁹ Vicente 1996, 133; Ogilvie 2003, 308.

²⁷⁰ Dumont 1998, 21.

²⁷¹ Dumont 1998, 24 n. 34.

²⁷² Epstein 2008, 167.

²⁷³ Epstein and Prak 2008, 10.

²⁷⁴ Grafe 2009, 81.

²⁷⁵ Trivellato 2006, 169.

²⁷⁶ Brown 1986, 212; for a similar assessment, see Crossick 1997, 14.

²⁷⁷ Ogilvie 2003; Smith 2006.

²⁷⁸ Van den Heuvel 2006; Van den Heuvel 2007.

²⁷⁹ Deceulaer and Panhuyusen 2000, 100.

²⁸⁰ Musgrave 1997, 153. For a similar analysis, see Vicente 1996, 130.

²⁸¹ See Brandt 2003, 8 with n. 12, and 11–12, on this dynamic at work in guilds’ exclusionary activities in early modern Frankfurt guilds.

²⁸² Knowles et al. 2002; United Nations Economic and Social Commission for Asia and the Pacific 2007, 105–106.

²⁸³ United Nations Economic and Social Commission for Asia and the Pacific 2007, 104–105.

²⁸⁴ Cuberes and Teignier 2016, 21 (Table 5).