Women, gender and the ancient economy: towards a feminist economic history of the ancient Greek world

Abstract

This article asks why women are ignored in debates about ancient economies and suggests a way forward. It argues that women performed a wide variety of diverse economic activities, though this is not particularly discernible from the scholarly literature, which mostly casts them as patrons or prostitutes and, despite the household being a basic economic unit to which women contributed, generally considers economic actors as male by default. However, by drawing on feminist economics, social history and gender studies, it is possible to reframe women’s varied activities in ways that acknowledge their labour, spotlight female agency, challenge the (gendered) categories of analysis and discourses that are predominantly used within ancient history, and recentre questions relating to the structures of inequality created by ancient economies. Three case studies explore some of the problems and raise new questions: Z3, a building in the Kerameikos the function of which is debated, the contribution of tax-farmers to sacrifices on Kos and the water supply in Athens. That is, this article argues that examining how ancient economies were gendered is a profitable way to think about both economic history and gender history.

I. Introduction

Where are the women in research on the ancient economy? Their relative invisibility can only be partially explained by the surviving sources given that the last 50 years have seen a great deal of work on women and gender in other fields of ancient history. Routinely in accounts of ancient economies from Moses Finley to New Institutionalism, women’s experience is either ignored or viewed in limited terms and, as a result, economic actors are conceptualized as being male by default. ¹ This article argues that this assumption skews our understanding of the complex and changing economic systems of the Classical and Hellenistic periods and suggests a recalibration of how we approach ancient economies with a view to highlighting the contribution that women made within them. This is not simply an argument for a greater recognition of women’s lives within ancient economic history (though this is in itself important); through the employment of a feminist theoretical framework we can suggest new ways to consider how ancient economies worked and who they worked for.

I argue here that examining how ancient economies were gendered is a profitable way to think about both economic history and gender history. In the first instance, it allows us to interrogate the ideological discourses that devalue, for example, the labour of women and to think critically about how we identify economic actors, economic agency and inequality in the ancient Greek world. In the second instance, it provides a mechanism for assessing how gender roles were constructed in different contexts and with reference to interacting combinations of other forms of social distinction. This allows us to see the contours of gender as a category of analysis and to see, for example, when sexual difference was a critical factor for operating in the world and those times when it was not. ²

This article therefore proceeds as follows: sections II–IV examine how the economic categories we have hitherto employed have painted women out of our analyses of ancient Greek economies and argues that this has systematically devalued the role of women as economic actors. Sections V–VII sketch a new path for a feminist economic history of the ancient Greek world using two case studies: tax-farmers’ concessions in second-century Kos and water infrastructure in late sixth- to late fourth-century Athens.

II. Ancient economies: the institutionalist turn

Recent trends in the study of ancient economies have provided a new climate in which to view the economic history of the ancient Greek world and New Institutional Economics (NIE) has been influential in shaping this approach. ³ The central idea of NIE is now well-known amongst ancient historians: institutions (both legal frameworks and social norms) form key constraints on economic actors. ⁴ It is undeniable that this focus has revitalized the field, providing a theoretical framework through which to ask questions about both the performance of ancient economies and the structures that shaped them. One of the key conclusions of this research is that the ancient Mediterranean world saw a period of economic growth and cultural efflorescence; it has also enabled new questions to be posed about trade, markets, property rights and labour specialization, to name a few examples.

Be that as it may, NIE-influenced economic history has not been without criticism. Developed as a response to neoclassical economics in which so-called ‘rational’ actors often appeared so abstracted that they were atomized from their social context, it has appealed to ancient economic historians as a tool. Critics argue, however, that it nevertheless betrays ideological assumptions that (intentionally or otherwise) centre neoclassical economics and/or neoliberalism. This places too much emphasis on economic growth instead of on the systems of inequality (such as slavery) that are created through economic activity and, insofar as ancient historians have employed it to date, is top-down in approach. ⁵ We might not be surprised, then, that women and questions about gender appear infrequently. While it is clear that the editors of, for example, the Cambridge Economic History of the Greco-Roman World view ‘household and gender’ as a ‘determinant of economic performance’ and Saller’s chapter in that volume clearly forms a starting point, it is not expanded upon in detail in the volume or in much subsequent scholarship. ⁶

Even on its own terms, however, NIE ought to be particularly amenable to questions about gender. Indeed, feminist economists for the past 30 years have been discussing the need to focus on ‘the economic advantages and disadvantages, entitlements and penalties associated with … gendered roles in the economy’ ⁷ and to ‘illuminate the ways in which assumptions about gender influence virtually all aspects of economic reasoning’. ⁸ However, there is relatively little consideration of how gender might have shaped economic phenomena in many of the important works that have rethought aspects of ancient economies. ⁹ The exceptions are few: Edward Cohen examines how ideals of masculinity restricted men’s economic opportunities, ¹⁰ Barbara Tsakirgis explores the contribution of women to textile manufacture, ¹¹ Edward Harris and Daniel Jew highlight the intersections between women, household and marketplace, ¹² but most recent discussions that deal with women or gender in relation to economic matters are not particularly located within, nor obviously influenced by, NIE. To quote a leading feminist scholar: ‘the world looks different when you look at all the people, not just half of them’. ¹³

One place in which gender history and economic history almost touch is among historians and archaeologists interested in textile production and trade. Many of these have raised questions about gendered divisions of labour, ¹⁴ the shared language of textile production (constructed as a ‘feminine’ activity) and other systems of thought or practice (constructed as ‘masculine’ activities: mathematics, shipbuilding or even poetry), ¹⁵ as well as the materiality, sociality or role of textile production within systems of value. ¹⁶ However, although some of the scholars interested in textiles use NIE paradigms, ¹⁷ none of those who write about gender have done so explicitly nor have they yet fed their analyses into ‘big picture’ discussions of ancient Greek economies; few situate their work in dialogue with, or as a critique of, NIE and only a few NIE practitioners have focused (to date) on textiles. ¹⁸

This is not to say that no historians interested in women or gender ask questions of an economic nature. The work of Sarah Pomeroy and Lin Foxhall stands out here as being pioneering in this regard and Violaine Sebillotte Cuchet has recently argued persuasively for linking economic and gender history. ¹⁹ But the large volume of research on gender in antiquity exists for the most part in an entirely separate subfield governed by different questions and concerns, often drawing on different source traditions than those who work on ancient economies. ²⁰ Much of this is focused on literary studies. While drawing useful conclusions about ideologies and discourses, it is often difficult to see how these cultural histories have a reach beyond the symbolic realm to the hard surfaces of economic life and few attempts are made to connect these spheres.

The problem is that, for the most part, our methodologies have been developed to examine the ideologies of gender roles rather than the realities of life on the ground. We have known since the 1990s, for example, that there was a disconnect between the ideology of female seclusion and the realities of life, and that class, status and age shaped gendered behavioural ideals. ²¹ Women were shown to have frequently worked outside the household, for example in the fields, markets, inns and workshops of the ancient Greek world, even if the empirical evidence for this does not survive in the quantities that it does for men. Similarly, women owned, controlled and disposed of property in greater numbers and in more contexts than the readers of the scholarly literature of the 1970s would have imagined. The picture emphasized here was very much one of male domination and control, of women with limited agency. ²²

The frame through which we view our surviving evidence is therefore crucial. Sebillotte Cuchet’s important recent study suggests seeing women as stakeholders in ancient economies by examining their involvement in economic transactions, emphasizing the different kinds of agency with which women operated and being sensitive to the discursive context(s) of the source material that provides our information. ²³ This is fundamental. In addition, we also need to pay attention to how the categories of analysis we employ shape our understanding of economic processes and phenomena.

In fact, this was a point raised by Pomeroy in 1995 when she called for greater consideration of the role of women within the ancient economy. She was concerned not only with trying to see beyond the gaps and biases of the extant evidence, but with how historians conceptualized ancient economies through the categories of analysis they employed: ‘the categories that have been applied by economic historians are relevant almost exclusively to men’. ²⁴ Since Pomeroy wrote this, ancient economic history has indeed been reconceptualized. A key aim of the institutionalist turn was to broaden focus from production and consumption onto transaction costs, human capital, institutions and demography. But transaction costs are fundamentally shaped by discourses of power which promoted and sustained innumerable inequalities. Human capital frequently meant slave labour which often drew on trafficked women and children and reconceptualized their gender in the process. ²⁵ Gendered expectations and practices affect almost every aspect of demography to some degree or other, from fertility, which is shaped by cultural expectations of marriage and childbirth, to mortality, to the nutritional differences between men and women that can be inferred by the study of some skeletal samples. ²⁶ That is, Pomeroy’s main criticism can also be applied today: although the categories of economic analysis have changed, they remain gendered categories.

Similar reservations can be raised about other aspects of economic activity: the structure of property rights routinely favoured men over women, the role of the state in the ancient world almost always excluded women from access to deliberative and judicial channels of power, technologies physically shaped women’s bodies and demanded their frequent, time-consuming labour. ²⁷ To use these as economic categories without acknowledging that they are not neutral is problematic. ²⁸ We must, therefore, think carefully about how the categories we employ shape our understanding of economic practices and how they in themselves construct notions of gender (or gender difference), in combination with other forms of social distinction, in multiple ways and contexts.

As the recent feminist turn in ancient political history has reincorporated women into the body politic, it is time for economic historians to think carefully about how gender is constructed within, and through, the ancient source material, as well as in the models we use to interpret this. Because gender is so intricately intertwined with other relations of power, such as class, status or belonging, and because these are created by, and lived through, access to resources, this is (or ought to be) very much a part of the framework of ancient economic history.

III. Evidence and models

The difficulty, of course, is that our evidence, which in empirical terms is weighted more towards the lives of men than women, is filtered through layers of ideological discourse that are highly complex to unpick. ²⁹ Nevertheless, it is fundamental that, at least as a first step, we recognize women as active agents in ancient economies in diverse ways and different contexts. As Sebillotte Cuchet notes, ‘there were, in Ancient Greek societies, as many situations for women as for men’, ³⁰ but we are not good at recognizing these except in the most broad-brush ways. Indeed, the surviving source material has been considered a constraining factor in this regard. While it is certainly true that there is less extant evidence for individual women than for individual men in the ancient world, the fact is that there is a general reluctance among scholars to see women as being economically active outside of fairly restricted contexts, mostly patronage and prostitution. ³¹ This is despite the fact that there is much evidence for women owning property and disposing of it (land, slaves, businesses), ³² working inside and outside the home, ³³ loaning and borrowing, ³⁴ trading, ³⁵ producing goods, ³⁶ consuming them, providing services, ³⁷ donating money to cities, sanctuaries and voluntary associations across the Greek world. ³⁸ That is, in many economic activities that are known to us, across a wide variety of economic ‘sectors’, we can find some trace of women involved, despite the fact that much of our source material frames women’s economic activities in particular ideological ways that suggest that women’s roles were limited.

By identifying the diverse economic roles of women of different classes and statuses, therefore, and taking seriously the household as an economic unit which devised multiple strategies to survive and flourish, we ought to be able to see different ways in which women contributed to a family’s income or well-being. ³⁹ This, in turn, allows us to avoid relegating the household to a separate sphere of activity, divorced from discussions of ancient economies, to investigate how women, as economic actors, linked different forms and scales of economic activity and to explore how women’s agency operated on multiple different levels. As recent research on citizenship has shown, gender markers are not always clear-cut and our assumptions about women’s behaviour and political roles have been revised as a result. ⁴⁰

The question here, then, is not whether the women we see in these contexts were exceptional in some way, but why the source material devalues those activities, often to the point of silence, and how we read through these ‘gaps’. The difficulty is compounded because the devaluation of women’s economic activity is not just a feature of the ancient source material; it is also reflected in some of the scholarship. ⁴¹ This can be seen, in particular, in scholarship which examines female labour: in certain contexts (prostitution) women feature prominently, whereas in others (textile production), until recently, they have not. We can use the debate over a building in the Kerameikos to highlight the consequences of this.

IV. Where are the women? Bau Z and the whorification of women’s labour

Located in the Kerameikos in Athens, Bau Z, dated in its third phase to the third quarter of the fourth century (Z3), is an unusual building architecturally that has generated a fair bit of controversy with regard to its interpretation (fig. 1). Relatively large numbers of loom weights (as well as a handful of spindle whorls and needles) were found in two out of five of its archaeological layers, which has led scholars to suggest that textile production occurred there. ⁴² However, it has also been identified as a brothel/inn, either as its primary or secondary function. ⁴³

Fig. 1. Plan of Bau Z3, Kerameikos. Drawing by Thanasi Papapostolou based on Knigge (2005) supplement 5 and Ault (2016) fig. 4.4.

The brothel argument is made most strongly by Hermann Lind, James Davidson and Bradley Ault, although it is a fairly widely (but not universally) accepted interpretation. ⁴⁴ However, it is a conclusion reached without taking full account of the literary discourses that paint certain kinds of women’s work as disreputable. ⁴⁵ Lind, for example, linked the remains to Isaios 6, a law-court speech from the 360s in which the relatives of Euktemon contest his estate, part of which includes a brothel in the Kerameikos. A key part of the argument the speaker makes here is that the children Euktemon said were his legitimate sons, and therefore in line to inherit instead of him, were, in fact, not. Their mother, whom the speaker calls Alke (though the sons name her as Kallippe), is portrayed in the speech as a freedwoman prostitute who managed the brothel (or rather, euphemistically, the sunoikia) and where Euktemon, as a vulnerable old man, spent too much of his time and fell under her spell. The depiction of the activities that took place in the building, for example, Euktemon dining with Alke (6.21), its location near the postern gate (6.20) and the mention of oikēmata where Alke allegedly worked, albeit not here but in a building Euktemon had previously owned in the Piraeus (6.19), were linked to the archaeological finds of Bau Z and used as a template for interpreting them. The small personal items of jewellery and religious figurines were seen therefore as the property of foreign slaves (the paidiskai of Euktemon’s Piraeus sunoikia, 6.19), the coins were evidence for the selling of sex and the supposedly small rooms, along with the keys found in the building, suggested places for clients to hire those slaves or rooms in an inn. ⁴⁶

Despite observing that no phase of the building existed at the time of the speech, and ignoring the rhetorical strategies that denigrate Alke as a dangerous woman, Lind was nevertheless tempted to suggest a continuity of function as a brothel between Z2 (ca. 420–400) and Z3 (ca. 340–300). ⁴⁷ Similarly, Davidson, as part of an argument that examined the ‘spinning hetaira’ motif (see below), viewed the building as a brothel in all its phases, where slave women toiled on the loom by day and serviced customers by night. ⁴⁸ Ault likewise retrograded the sex work back to Z1 (ca. 430–420) and Z2 (ca. 420–400) and emphasized the links between a wide range of the finds and prostitution: from pottery (drinking and tableware, also lamps) to the small rooms linked with the oikēmata, where literary sources tell us prostitutes could be found, ⁴⁹ to washing facilities, to the presence of finds depicting Aphrodite. ⁵⁰

The problem, however, is that none of these indicators is particularly secure and therefore the grounds for identifying the occupants of this building as full- or part-time prostitutes are weak. As Kathleen Lynch demonstrates, more drinking and dining vessels were found in the context of a private house (through the clear-up debris excavated from well J2:4) than in Z3, where they would be expected if this were a brothel/inn. ⁵¹ The rooms that are described by Ault as ‘uniform and [having a] cell-like quality’ (thereby painting a picture of poverty and desperation) are not notably smaller, nor, as far as we can tell from the foundations alone, more ‘cell-like’ than in other Athenian dwellings, either private houses or workshops, and therefore cannot obviously be described as prostitution-related oikēmata. ⁵² The keys which Lind suggested were for doors to rooms are actually for strongboxes. ⁵³ The reputation of the Kerameikos as a kind of ‘red-light district’ does little to identify the function of any specific building (and in any case misunderstands the use of space in Athens). ⁵⁴ The purported interest of the buildings’ users in Aphrodite, as the goddess of love and sex, is not strong, nor is her association with prostitutes particularly sound, given the multitude of roles she plays within Greek religion. ⁵⁵ Other evidence that has been used to build the case for a brothel can similarly be easily questioned.

While it is true that women were predominantly associated with weaving, and many loom weights were found here, these women’s ‘foreignness’ is also overplayed in the scholarship, thereby over-emphasizing their supposed servile or outsider status. ⁵⁶ This attribution is made on the basis of various items of worship and jewellery that have been interpreted as having connections to ‘foreign’ gods, but these represent, for the most part, Kybele or the Mother of the Gods, who was hardly a ‘foreign’ deity in the late fourth century. ⁵⁷ Indeed, she had multiple sanctuaries in Athens (one of which was in the Agora, the civic heart of the city) for more than a century before Z3 was built. ⁵⁸ The ‘foreign’ jewellery includes a medallion that probably depicts Aphrodite (although in what form is debated) and some Punic or Phoenician beads. ⁵⁹ Needless to say, jewellery is a highly portable item and a frequently traded one. Identifying ethnicity or servile status through such objects in a highly connected Mediterranean world is a enterprise fraught with danger, as recent research has ably demonstrated. ⁶⁰

The identification of this building as a place where prostitutes worked therefore ultimately rests simply on the presence of women in the building, women who are (perhaps wrongly) interpreted as being foreign. This is highly problematic. That there is much greater discussion of this building in scholarly discussions of sex work than of textile production reveals some underlying assumptions about women’s labour: that is, there is a disturbing inability to imagine women being economically active in contexts where their labour is not sexual. It seems unlikely, however, that sex work accounted for as large a part of ancient economies, or indeed for most women as large a part of daily life, as, say, the various tasks of textile production did.

One reason why there are few works that take seriously the interpretation of the building as a site of textile production is that this was seen as ‘women’s work’, and therefore it was rather ignored in 20th-century accounts of economic activity, as Mary Harlow and Marie-Louise Nosch show:

Aligning textile production with domesticity and female virtue conceals both the necessity of domestic production in some circumstances and situates textile production in a seemingly unimportant socio-economic category which makes economists refrain from considering its role and value in the ancient economy. This, however, is paradoxically a vision invented in the late 20th century by a professional academic community, who were distant from the experience of practical life and knowledge of the modes and methods of textile production. ⁶¹

Only recently has it begun to be included in anything more than technical or specialist studies and its place within economic systems noted, meaning that the labour of women, as the predominant producers of textiles, contributed to the wealth of families, cities or regions through production and trade. ⁶² Because the debate in Athens was bound up in discussions of the ‘spinning hetaira’, that is, images of women spinning, interpreted as representations of prostitutes because they were interacting with men holding purses, it has been difficult for scholars to disassociate textile work and prostitution. ⁶³ The large numbers of talasiourgoi (wool-workers) recorded on the phialai exeleutherikai inscriptions seemed to reinforce such suggestions: why else would over 80 per cent of the freedwomen attested here be designated in this way if it were not a euphemism for something else? ⁶⁴ The lack of corroborating (literary) evidence for the market, rather than domestic, orientation of textile production was seen to be significant, ⁶⁵ albeit without seriously questioning why that evidence might not exist within the world of elite discourse. From here, the link between textile production and sexual labour seemed assured: women worked wool, women were prostitutes; therefore women worked wool when not working as prostitutes and the archaeological finds of Bau Z were reconciled with reconstructions of sexual labour from other sources. ⁶⁶

The decision to interpret this building as a brothel, and its inhabitants as full- or part-time prostitutes, rather than (for example) as a dwelling of some description where textiles were produced, ⁶⁷ is one, therefore, that is shaped by both ancient and modern discourses that present women’s work as disreputable or unimportant. It may be the case that some of the users of this building sold sex, but the crucial point here is that we cannot tell that from the evidence that survives. It is a choice made by historians to imagine their labour as sexual labour and it is one that has consequences. As I have argued elsewhere, by writing the non-sexual labour of women out of the history of the ancient economy we merely reproduce the gendered discourses of antiquity rather than critiquing them. ⁶⁸

V. A feminist economics approach: provisioning, well-being and economic agency

Where to go from here? How can we build on our understanding of gender and gendered discourses in a way that allows us to investigate ancient economies and vice versa? I suggest we develop an approach that draws explicitly on feminist economics, gender studies and social history. By focusing on the economic agency of women (not just their sexual labour), we can see both how gender is constructed in certain specific contexts, as well as raise some important questions about our understanding of ancient economies.

Key here are the concepts of provisioning and well-being. Developed within the field of feminist economics, provisioning focuses on the interlocking processes that combine production, consumption, social reproduction, distribution and exchange that provide what is culturally defined as the necessities of life. These include the provision of basic needs, but also the things needed to thrive (i.e. have a culturally appropriate standard of living) for those for whom relationships of responsibility exist. ⁶⁹ These relationships are obviously found within the household (and in families and kinship groups more widely), for example, the strong ties that exist between parents and children, slave-owners and the enslaved, the young and the old, but also between members of the broader community (neighbours, the deme/village, the polis, communities of religious worshippers, etc.). How these social dynamics shape, and are shaped by, access to resources is a major question here.

In the ancient Greek world the household was the basic economic unit. It is within the household that the daily activities took place that provided sustenance, or were otherwise necessary (i.e. food preparation, household cleaning, fetching water, disposing of human waste, caring for children, the sick and the elderly, etc.). These activities required regular and unending attention and were embedded within, and negotiated by, social relationships. This makes provisioning a particularly suitable concept to work with, not least because ancient authors themselves recognize the centrality of the oikos. It also emphasizes the importance of writing economic histories that do not ignore the household or the domestic labour that took place within it, as well as underscoring the need not to silo the household into a conceptually separate space, set apart from other economic processes. However, we can only say so much if we stay within the world-view of an Aristotle or a Xenophon; instead it is necessary to interrogate the inequalities that arose from the considerable daily work undertaken by members of households (enslaved and free) to feed, shelter and clothe themselves. That is, who gets to perform these tasks and how this is negotiated is both a result of, and also demonstrates, the dynamics of power within families and communities and reveals much about economic systems on the micro-scale. It also sheds light on the inequalities that arise on the macro-scale: wealth inequalities within and between households and polities, systems of slavery and dependent labour that uphold production and consumption and drive trade, the effects of warfare, natural disaster, or resource distribution in the creation of human mobility, to name but a few examples.

It is here, however, that we can see the importance of thinking about women as active economic agents in diverse ways because women’s multiple interactions within economic systems are crucial for understanding how those systems function. Compare an enslaved woman coming into a household with the free women of that household. Transported perhaps across the Aegean, likely as a result of warfare, kidnap, poverty or some other act of violence, her fate, as for the captured women of Troy, is a cause for lamentation. She is put to work, tending to the domestic duties of the household, perhaps on the loom or in other aspects of textile production, perhaps fetching water or grinding grain, cleaning, preparing food, running errands or looking after children: the myriad jobs that are needed for the household to function. The Greek literary tradition portrays the violence, bitter sadness and fears of such a moment. ⁷⁰

For the free women of the household, her arrival is, in contrast, a moment of joy, an extra pair of hands, a relief from their own daily grind and a marker of their wealth and status. ⁷¹ The enslaved woman arrives as a result of long-distance systems of exchange, mobility and distribution, but also as a result of the demand for domestic labour. Her work ensures the productivity of the household, provides for their daily consumption needs, contributes to, and potentially increases, their income and wealth. If she is freed, it may be with terms and conditions attached, structurally constraining her mobility and ability to live independently, and institutionalizing the continuance of violence against her. ⁷² Embedded within a feminist economic history, therefore, is the acknowledgement that economic processes and transactions are themselves shaped by, and shape in turn, various social dynamics (gender, class, status, ethnicity, etc.), and this, in turn, allows us potentially to link micro- and macro-scales of analysis. ⁷³

Critics might argue that this is not the stuff of economic history. But what defines ancient economies? Drawing on economic sociology, a key frame of reference for many ancient historians over the past couple of decades, we might answer: ‘the complex of activities concerned with the production, distribution, exchange, and consumption of scarce goods and services’. ⁷⁴ As Paula England and Nancy Folbre point out (and as has become abundantly clear during the public health crisis of 2020), this includes work in the household, making meals and childcare. ⁷⁵ This was overwhelmingly the work of women and the enslaved in the ancient world. Conceptualizing ancient economies only from the point of view of top-down institutional structures (such as regulations, legal structures, property rights, state formation) is only going to take us so far. This ought to be supplemented with an approach that seeks to interrogate, as far as is possible, the micro-decisions of everyday life and more explicitly seeks to acknowledge the uneven distribution of our source material.

To draw out some possibilities we can use a couple of examples. My main focus here is on women’s labour, but of course this is only one aspect of ancient economic activity. The aim is to show, first, that we can include women in discussions of economic processes, although this often means reading between the lines of various forms of evidence, and second, that by doing so we can not only say something about ancient economic behaviour, but also see how, through economic processes, gender norms were created, defined and negotiated and systems of value established.

VI. Hidden women

IG XII.4 293 (= Syll. ³ 1000) is an inscription from Kos (mid- to late second century BCE) that records the sacrificial contributions required in connection with a festival to Poseidon and the nymphs Kos and Rhodos. ⁷⁶ Various groups of people who are involved in shipping are recorded here: the trierarch and nauarchs as well as ship-haulers (neōlkoi), oar-makers (kōpoxustai) and merchants, as are a number of tax-farmers who provide sacrifices and feasts for the festival. Aside from religious information, the inscription reveals a network of economic agents active on Kos whose financial success was tapped by the sanctuary for the well-being of the city as a whole. ⁷⁷ We can use this inscription, therefore, to raise questions about the places within ancient economies where women ought to be found, if only we looked for them, and see the kinds of work that was recognized as being economically necessary. That is, we see here how discourses about women’s labour are constructed in our sources and raise questions about how we might consider these in practice.

Of particular interest here are the tax-farmers who have concessions on a wide range of items from agricultural produce and farm animals, rent from houses, various retail traders (of saltfish, incense, pulses, etc.), infrastructure related to the fishing industry (watchtowers etc.), hired labourers, prostitutes and slaves. ⁷⁸ Fishing, shipping and agriculture are prominent here and demonstrate the close connections between these religious and economic networks.

Women do not play a prominent role within this inscription, at least in direct terms, even though we know that they were important within the religious economy of the island. ⁷⁹ Hetairai are mentioned, however, as one of the groups who are taxed (line 5). ⁸⁰ This concession is grouped together with that for a specialist Koan wine, wood, barley meal and rent, although it is unclear why these appear as a unit. ⁸¹

Another group of women are female slaves, who are recorded together with vineyard workers (ἀμπϵλοστα<τ>ϵύντων καὶ τῶν γυναικϵίων σωμάτων, line 9). Predictably, these enslaved women are sometimes considered to be prostitutes, even though hetairai appear in the inscription just a few lines before and form a distinct group from whom tax is collected. ⁸² This is based on the assumption (again) that female labour equates to sexual labour, particularly when it involves enslaved women. Again, we cannot tell whether the enslaved women mentioned here ever sold sex (willingly or through coercive or violent means), but the fact that they are grouped together with vineyard workers suggests that they were primarily agricultural labourers, or that they were considered as valuable as, or generating similar levels of revenue to, the vine specialists, perhaps with specialist skills of their own. ⁸³ This need not be prostitution.

Indeed, women’s work is behind the production or consumption of many of the agricultural products and market-trading activities taxed here: women presumably ate the foodstuffs, drank the wine, spun the yarn, cultivated the gardens and lived in the rented houses, thereby generating revenue for the tax-collectors to collect. Elsewhere in the Greek world, women are attested as incense-sellers (libanotopōlides), as market traders (kapēlides) who would have sold goods such as the pulses and saltfish noted here. ⁸⁴ They are attested as sellers of wood (xulopōlides) ⁸⁵ and bread (artopōlides), ⁸⁶ made from grain that was ground predominantly by women; ⁸⁷ women are also attested as landladies, ⁸⁸ and as pimps. ⁸⁹ It would not be surprising therefore that women were involved in these activities on Kos, but they are often disguised by linguistic factors (the use of the masculine plural for mixed-gender groups makes women invisible) and epigraphic practices that limit women advertising themselves by occupation. ⁹⁰

Moreover, all of the activities recorded in this inscription require workers to be fed and clothed, work that predominantly fell to women, free and enslaved, within households. That is, even though men appear prominently in this inscription, it does not take much to see the networks of women (free and enslaved) on whose work the tax-collectors’ revenue was based. That is, through identifying the economic activities of women we can start to isolate how processes of provisioning at the everyday level are intricately intertwined with aspects of the economy which are more visible.

Indeed, it is the labour of women that provided Kos with one of its most desirable luxury products: silk. It is possible, although we do not need to assume (as Susan Sherwin-White does) that the gunaikeia sōmata were slaves working in the silk industry; ⁹¹ it is clear from Aristotle’s description of silk-worm larvae that women, and women’s knowledge, were central to silk production here: he states outright that it is women who work the fibres from the cocoon, and a woman named Pamphile was regarded as the inventor of the technology. ⁹² The involvement of women in the manufacture of luxury textiles on Kos is also attested by a purple dye-seller (porphuropōlis) from a funerary inscription of the first century CE, when Koan silk and purple garments were at the height of their popularity in Rome. ⁹³ Purple garments were also worn by the priest of Nike (IG XII.4 330.8–9) and that of Zeus Alseios (IG XII.4 328.16–17) on Kos, demonstrating both an internal and external demand for such items. That is, production and consumption are intricately intertwined and women’s labour, which comes to light easily when we frame it in terms of provisioning, lies behind much of this economic activity. Painting women back into the picture therefore allows us to see that their work was economically necessary not just to household well-being, but to Kos and the regional economy more broadly.

The inscription from Kos, therefore, draws our attention to the limited way that ancient institutions recognized the economic activities of women in comparison with those of men. That this is not the whole story is also clear, however. In Kos, women are well represented in epidosis inscriptions, for example at the sanctuary of Aphrodite Pandemos and Pontia, highlighting their wealth and attesting to their ability to dispose of property, control assets, develop social capital, their commitment to familial and communal well-being through religious dedication and their personal piety. ⁹⁴ That sanctuaries served as nodes for various kinds of broader economic activity suggests that women were very much embedded in these economic networks. ⁹⁵

VII. Gendered work

Recognizing diverse forms of women’s labour within our highly skewed source material is clearly desirable, therefore, but we should also consider how this labour intersects with other forms of economic activity. Consideration of the water supply of Greek cities provides a good example of some of the questions that can be raised. Most research on water supply, however, is related to the public organization of water provision and its technologies, the euergetistic mechanisms of infrastructure provision or the discourses surrounding women and water in various contexts, all of which neglect the fact that water collection is also an activity necessary for survival, health and well-being that needs to be performed by someone, usually every day. ⁹⁶

As in the inscription from Kos, we can see here another context in which women’s labour is usually overlooked, presumably because our sources predominantly construct water collection as women’s work. There are two key questions that arise: (i) how did infrastructure projects, like investments in the water supply, affect the day-to-day lives of households? And (ii) how did they create, reinforce or resist gendered norms of behaviour or patterns of labour?

Some of the gendered assumptions, and social hierarchies, embedded in water collection are revealed by Euripides’ Electra:

O black night, nurse of the golden starts! In you, carrying this vessel poised on my head, I go to fetch water from a stream—I do this not from need but to show the gods Aegisthus’ outrage against me—and utter my laments to the wide heaven for my father to hear! (54–59, tr. Kovacs 1998)

These are the first words Electra speaks in the play. Immediately after she appears on stage, Electra uses water collection as the key indicator of status as she complains bitterly about her new circumstances to her farmer husband: now she has to toil to fetch water when, before her expulsion from the royal household, her elite position ensured that she did not. For her, like other indicators of her new-found, self-inflicted ‘poverty’ (shorn hair, functional dress), water collection is a source of shame, an activity performed under cover of darkness. Fetching water here, then, symbolizes her new social position and her precipitous fall from grace. The speech positions her looking back to her old life where high-quality water was no doubt a given (flowing water from a river or a spring was considered the best quality, sweet and clear), ⁹⁷ and looking forward to her new life as a farmer’s wife, where this task is central to her daily routine. ⁹⁸

It is also clear in the play that this activity, despite requiring women to go outside, was perceived as being part of the care that they routinely provided to support the work of men in the fields. ⁹⁹ Her protests, that she fetches water not from need (chreia), but to get back at Aegisthus and Clytemnestra, somewhat bemuse her husband, who tells her that her ponos is unnecessary, before relenting and letting her get on with it. Fetching water, therefore, very much lay at the intersection of discourses about gender, wealth, status and labour.

That fetching water was gendered as women’s work is also clear from the iconographic evidence, particularly scenes of women at the fountain house, popular on Athenian pottery in the late Archaic and early Classical periods. Men appear sometimes in these scenes (see below), but they are vastly outnumbered by women (see, for example, fig. 2). However, in concentrating on how these images construct idealized and idealizing notions of gender, scholars have missed the opportunity to consider how these discourses shaped lived experience. ¹⁰⁰ Such readings deliberately set themselves apart from quotidian concerns; the depictions here are cultural constructs, not actual women whose bodies were transformed by carrying heavy loads. Here, because of the post-structuralist emphasis on the symbolic as opposed to the material world, the ahistorical connective tissue instead of the lived historical experience, the lives of actual women are painted out of the picture in much the same way we have seen in scholarship of the ancient economy. Ironically, for all the emphasis on gender roles, women are stripped of historical agency by these readings; they become the passive objects of artistic production existing solely within a discursive field. ¹⁰¹ But this evidence does, in fact, raise questions about how gendered discourses were experienced ‘on the ground’ in a historical context, and how, as in Euripides’ Electra, they shaped, and were shaped by, social relations, status concerns and wealth.

Fig. 2. Black figure hydria with women at the fountain house, ca. 520–500, Vulci. British Museum BM334. ©The Trustees of the British Museum.

First, scenes of men collecting water are not as common, and when they do occur, it is satyrs, labourers/slaves or youths, often in athletic contexts, who perform this activity (fig. 3; here the labourer is an old man, identified by the Corpus Vasorum Antiquorum as a slave). ¹⁰² This is not the work of the idealized citizen man. Men obviously were capable of using fountains and drawing water from wells though probably did so primarily for themselves (for example, while travelling, or to sate their thirst) or in athletic contexts and not, typically, for the household as a whole. Citizen men, identified iconographically by their mantles, sometimes appear in water-collection scenes, but they never themselves actually fill the jugs; instead they appear as bystanders or as harassers of women at the fountain house (fig. 4). ¹⁰³ These scenes certainly contain elements of fantasy and myth, and as such should not be read as depicting real life in an uncomplicated way, but they do also seem to reveal something about the social expectations placed on both women and men: that is, fetching water was a gendered activity. ¹⁰⁴ Note, in this context, the ‘feminine’ activities (grinding grain and fetching water) that the philosopher Kleanthes is supposed to have performed for pay after arriving in Athens with only four drachmas (albeit two centuries later than the fountain-house scenes). His metic status, and poverty, is important here, of course, and these activities are not presented by Diogenes Laertius as in any way honourable. ¹⁰⁵ In comparison, the water-carrying role of women is ritualized in various religious contexts, for example the participation of (female) hudrophoroi in processions at Demeter sanctuaries across the Greek world. ¹⁰⁶

Fig. 3. Black figure pelike with old man, satyr and women fetching water from a cistern or well, ca. 525–475. Berlin Staatliche Museen, Antikensammlung, 3228. Photo Johannes Laurentius / Art Resource, NY.

Fig. 4. Red Figure pelike attributed to the Pig Painter, ca. 475–450, depicting mantle-wearing men accosting women at fountain. Detroit Institute of Arts, Founders Society Purchase, General Membership Fund, 63.13.

The fact that water collection is so strongly associated with women, and such essential work for human survival, highlights one way in which gender was implicated in providing the necessities of life and supporting well-being. That this was part of daily household work for women, which undergirded all other activities, suggests that it ought not to be seen as marginal. In fact, it overlapped with other activities: food preparation, cleaning, animal husbandry, the watering of plants, pottery manufacture, metallurgy, etc. No activity was possible for long without a supply of water.

This no doubt explains why a great deal of investment went into the water supply of Greek cities, in terms of both public infrastructure (pipes, aqueducts, fountains, etc.) and household supply (cisterns, wells, drains, etc.). Typically viewed in terms of public organization or technological change, we might also ask how these projects affected households, created or shifted demands on women’s labour, or shaped those discourses about women and water outlined above. Women’s lives, it seems, were considerably affected by these investment decisions.

In Athens, for example, investment in the water supply of households went hand in hand with investment in public sources of water. There are two peaks of activity that are archaeologically detectable: first, in the late Archaic period, when numerous wells in and around the Agora have a definite period of use between ca. 600 and the Persian invasions and innovations in waterproof mortars allowed for more effective cisterns (precisely at the time of the public investment in fountain houses and the aqueducts and pipelines that supplied these); and second, in the second half of the fourth century, when cisterns, in particular, became more prevalent in households, similarly coinciding with investment in public infrastructure. ¹⁰⁷

How would this have shaped the lives of those who collected the water? It seems reasonable to suggest that as investment in the water supply increased, the time spent on this task might have decreased, that is, the time spent walking to, or waiting at, a public water source was reduced. In this sense, it might be argued that women’s work was recognized as being socially valuable (prompting investment in women’s time) and the time spent collecting water could be redeployed in other tasks. ¹⁰⁸ So, for example, as public sources of water became more common across the city, or as more households sunk their own wells or installed their own cisterns, this may have cut down the time needed to draw water at home, or walk to the well or fountain house. That this was a time-consuming and physical task can be seen on the bodies of EIA women, which certainly show the strain of this activity: a number of female skeletons from the Agora excavations show well-developed upper-body muscles, flattening of the skull at the bregma and/or vertebrae damage caused by habitually carrying heavy objects on their heads. ¹⁰⁹ However, no similar studies have yet been published (to my knowledge) on later skeletal material in Athens. ¹¹⁰

In this model, we might speculate on how the demands on women’s labour changed. If water was readily available through the well or cistern at home, did women venture out less? Did they divert their labour to other household tasks such as food preparation? Were they accorded more leisure? Did they spend more time selling goods in the market, making textiles or setting up businesses washing for others? Were women and their families able to prosper because they spent less of their time daily fetching water?

Or was the time saved by having a well or cistern at home reincorporated into women’s daily activities? Did the household consume more water? It is certainly the case that some sources of water were perceived as being more suitable for certain tasks than others, even if this must have varied according to location or season, and we cannot be precise on the details. Hippocrates, for example, suggests that well water was the lowest quality, followed by cisterns, followed by springs, whereas Vitruvius suggests the opposite for Augustan Athens. ¹¹¹ Not all water needed to be potable to be useful, of course, and different water sources were likely used for different household functions (drinking, cooking, washing, watering plants, flushing drains, etc.). ¹¹² Visiting a public well or fountain might have remained a part of the routine of the household, an opportunity for sociability as well as providing a different kind of water, even when a source was available at home. This would mean, then, that household investment in water resources had a minimal effect on women’s day-to-day work.

Indeed, the women collecting water in Lysistrata rise at dawn and still find a crowd of other women at the fountain house, jostling for position to fill their jugs. ¹¹³ Plutarch mentions a Solonian law that seemingly encouraged the use of public wells by those who lived within a hippikon (approximately 740 m) of one, which suggests that Athens was well supplied in this regard. ¹¹⁴ However, although a number of wells dated to the first half of the sixth century have been found around the Agora, most are thought to have been in households and few public wells have, in fact, been identified in the city. ¹¹⁵ Those that have all date to the fifth century or later. ¹¹⁶ Perhaps more common was the use of a neighbour’s well, although Plutarch makes clear that this was only to be granted for a limited daily amount if a decent effort had been made to dig a well for oneself and no water was found. ¹¹⁷

In fact, we can see occasional examples of households sharing water resources. The mid-fourth-century remodelling of House C, in which a well and cistern was built in room 12 (in addition to the existing courtyard well), provides a possible example: the orientation of the room is towards the street rather than the rest of the house (fig. 5). ¹¹⁸ This suggests easy accessibility to water for those not living in the building. ¹¹⁹ This kind of arrangement may well have been common and must have produced cooperation between neighbours at the same time as reinforcing social hierarchies.

Fig. 5. Plan of Houses D and C, ‘Street of the Marble Workers’, Athens. Drawing by Thanasi Papapostolou based on American School of Classical Studies at Athens: Agora Excavations 2012.54.0559.

Investment in public and household water infrastructure, therefore, only likely reduced the labour of a small section of sub-elite women. The most elite probably did not perform this task frequently in any case (Electra’s shock at having to fetch her own water is no doubt due to the fact that she would have had slaves to perform this task for her previously) and poorer women would have continued to rely on public water sources or neighbours with wells and cisterns.

Indeed, research in the global South shows that proximity to a water source shapes collection and consumption: too long a journey and household consumption is reduced, thereby reducing household well-being. ¹²⁰ It also shows that there are gendered constraints on who uses public water sources and when. ¹²¹ As Lisa Nevett has argued, the urban topography of Athens was shaped by the movement of women through the streets at different times of the day to men for activities such as water collection. ¹²² This may well be to avoid the perception (or reality) of harassment, gossip or accusations of impropriety, all of which form part of the discourses surrounding women and water. The public wells that archaeologists are able to locate are in high-traffic areas of the city, which points to conveniently accessible locations, but all of the places where such wells can reasonably be suggested to have been (gymnasia, bathhouses, cross-roads and city gates) were also constructed as places where only ‘disreputable’ women were found. ¹²³ These discourses likely shaped when, and how frequently, women collected water (that is, more trips at worse times of the day), or perhaps even meant that they avoided these areas all together.

Investment in water infrastructure, then, is not simply a matter of technological change, public organization or euergetism, but is intricately connected with household provisioning. Because water collection was predominantly seen as women’s work, it fundamentally shaped the daily activities of women and households, in some cases perhaps freeing up time for other activities, in other cases reinforcing gender norms and other forms of social distinction. But whatever impact these investment decisions had on individual households, they reveal a fundamental aspect of women’s labour: it was a crucial part of ancient life and central to the well-being and prosperity of households and cities.

VIII. Towards a feminist economic history

What, then, constitutes a feminist economic history of the ancient Greek world? First, it is clear that women were active agents in ancient economies in a variety of ways and we should recognize them as such. That means that we need to incorporate women’s diverse experiences into all aspects of ancient economic history, think about different ways in which female agency operated, not over-emphasize the role of sex work within the complexities of female labour or assume that women operated without any competencies under the compulsion of men. We can do this, first, by actively highlighting these roles and showing previously held assumptions to be unsafe, and, second, by rethinking the categories of analysis we use to talk about the ancient economy, recognizing that they are not neutral, but are in themselves sites of gendered discourses, ancient and modern.

Reorienting our discussions to consider questions of provisioning is one way we can achieve this because it directs attention to economic processes as well as outcomes, the social relationships in which economic activity is embedded and entangled, as well as the power dynamics and structures of constraint that shape lives in multiple, complex ways. This also allows questions to be raised about how ideals of gender combine with other indicators of social difference to create, reinforce and maintain wealth and control access to resources. Central here is a focus on agency, the processes that create and maintain inequality, and questions of well-being as critical ways of contextualizing the performance of ancient economies (to put it in NIE terms). ¹²⁴

That is, not only does this provide an important means of situating economic activity within a social and cultural context, but also acknowledges that this context is itself shaped by, and constructed through, ideas and ideals of class, gender, status and belonging. Viewing ancient economies in this way not only provides a different lens through which to interpret how resources were accumulated, produced, controlled, distributed or exploited, but also to think across spaces of exchange (local, regional, interregional) and scales of observation (micro, macro), because it views economic activity as a social process grounded in, shaping and being shaped by social relations and dynamics of power. This highlights the structures of inequality that were deeply embedded within ancient economies and encourages us to look for the diverse experiences of different groups: rich and poor, free and unfree, men and women, etc. In some contexts, for example, we might view the work of female market traders as women peddling household surpluses out of necessity. ¹²⁵ But these women also act as vital connecting agents between interregional trade networks, selling imported goods to local consumers. That is, they operate within an Aegean-wide network, linking the scales of local and ‘global’.

A feminist economic history is, therefore, not one that simply focuses on women, but one that draws on these observations to think more critically about the determining factors in the structures and performance of ancient economies. To do so will not only broaden our understanding of the multitude of roles women played, the inequalities they both suffered and promoted through the economic strategies they employed, but also provide a way to consider how critical they were to economic processes. Provisioning intersects with public economies, technological change and trade networks and provides a way to rethink these aspects of economic life. If ancient economic history is concerned with how resources are accessed, distributed and exploited, this very much includes recognizing the multiple roles of women in these processes.