Women’s Libraries and “Women’s Books”, 1729–1830

Abstract

During the eighteenth century, as the practice of selling books at auction spread throughout Europe, moving beyond its original Dutch origins to the rest of the continent, an increasing number of female-owned libraries were also sold at auction. These sales were accompanied by printed catalogues that, together with widespread discourses about women’s reading, helped construct new images of the female reader. During the same period, a new ideal type of female-gendered library was conceptualized, the bibliothèque choisie, that foregrounded personal reading taste and belles-lettres, and was sometimes associated with historical, real female readers. This essay explores the historical reality of the discursively imagined woman reader, using bibliometric tools to examine a corpus of auction catalogues of smaller and medium-sized female-owned libraries or bibliothèques choisies sold in France, the Dutch Republic and British Isles between 1729 and 1830. Adopting a geographically comparative perspective, I study the historical evolution of these female-owned libraries, focusing particularly on women as readers of devotional literature and of novels. I demonstrate that while women’s libraries did differ perceptibly from men’s libraries, these differences were sometimes subtler than inherited narratives might suggest, and require rich contextualization taking into account multiple geographic, social, and material factors.

In a powerful fantasy of female intellectual autonomy, her epistolary novel Het land, in brieven (The countryside, in letters) (1788), Dutch author Elisabeth Maria Post took as protagonist a single, young woman of independent means, Emilia. Every day, after her prayers and household chores, Emilia retired to her “book cabinet” (boekvertrekje), in which she had “placed with forethought all that can lead to gravity and true wisdom”.

Surveying her domain, she described her bookcase’s contents to her intimate friend and correspondent Eufrozyne:

My choice of books clearly shows my repugnance for strictness carried too far. Above all I cherish the word of God; as well as those writings that help clarify it, or assist my progress in the practice of faith and the moral duties of a Christian; especially when those subjects are presented with a taste, gracefulness, and noble fire that encourage a similar elevation. But not only these works, History too, and Physics attract me, as they provide thousands of examples of Divine wisdom in organizing all creatures, and especially the fates of humankind, filling me with wonder at God’s greatness and the vanity of His creations. The poetic style I admire especially, both in sad and solemn, and in cheerful dress, when preaching as well as onstage. And would I scorn those pictures of human morals that awaken only noble passions, pleasantly moving and improving the soul, merely because the world is so corrupt that such virtue finds no example in it? Would I scorn a Clarissa, a Grandison, a Sofia, a Burgheim and other such writings, merely because they are novels? (Post 1987, 21–22)¹

With its combination of Bibles, devotional works, history and popular scientific works, plays, and novels, Emilia’s bookcase offers a suggestive picture of how a library might shape female identities and subjectivities. It also provides a tantalizing—albeit fictional—glimpse into female reading at the end of a decisive period in the history of women’s relationship with books. Not only did the eighteenth century witness increased book ownership across all strata of society. Women, particularly, enjoyed greater access to books, as new forms of literary sociability, lending networks, and female bookish engagements expanded, culminating finally in a massive influx of women authors into the literary field at century’s end (Hesse 2001; Pearson 1999).

Increased production and demand for books, in turn, underwrote new forms of distribution. These included the sale of books at auction, with the second-hand book trade possibly accounting for up to half of the books circulating in eighteenth-century Europe. As the practice of auctioning books spread, moving from the Dutch Republic to the rest of the continent, a growing number of female-owned libraries were also sold at auction. Auctions were accompanied by printed catalogues that, by foregrounding the figure of the woman collector, helped construct new images of female readership. These printed auction catalogues are thus a unique source not only to explore the historical reality of the fictional female libraries described in novels such as Post’s Het land. They also provide source material for a comparison of the contents of eighteenth-century women’s libraries with men’s, in an exploration of how, if at all, they might have functioned as a “room of one’s own”, to reference Virginia Woolf’s resonant image of female intellectual autonomy, which Emilia’s “book cabinet” in Het land already presaged in the 1780s.

The MEDIATE Database

Sophisticated comparisons of female and male-owned eighteenth-century libraries are now rendered feasible by a database developed for the European Research Council-funded MEDIATE project (Measuring Enlightenment: Disseminating Ideas, Authors, and Texts in Europe, 1665–1830) at Radboud University (The Netherlands).² This project sought to model the circulation of books in western Europe during a period overlapping with the Enlightenment movement, by creating a database holding structured data extracted from several hundred printed catalogues of private libraries primarily sold at auction in the Dutch Republic, British Isles and France between 1665 and 1830. Including data harvested from 600 individual collections (the dataset’s basic conceptual unit, rather than the catalogue, which sometimes described multiple collections), and representing over half a million books, the MEDIATE dataset was assembled to ensure comparable numbers, temporal distribution and sizes of collections from these three regions. The database additionally includes ten Italian collections: this small number reflects the fact that a robust library auctioning culture failed to develop in eighteenth-century Italy. Further enhancing cross-national comparability of the data, the database focuses on catalogues listing fewer than 1,000 lots—with one lot corresponding mostly, but not always, to a single book title (including multi-year runs of periodicals or multi-volume works counted by cataloguer-booksellers as a single title). These catalogues thus represent, size-wise, the smaller 50% of the corpus of extant private library auction catalogues.³ Their smaller size increases the likelihood, first, that they describe real collections, rather than including substantial numbers of books surreptitiously added by booksellers from their unwanted stock.⁴ Secondly, they might be supposed to document reading preferences not only of the best-known bibliophiles and intellectuals, but also of more obscure, middling collectors, including even in some cases artisans and working-class readers.

In an earlier iteration of the project, I hypothesized that these smaller and mid-size libraries reflected an eighteenth-century literary field that could anachronistically be described as “middlebrow”—a concept that has, significantly, often been gendered female (Holmes 2018; Rubin 1992). Associated with the “middling” classes, and belonging neither to the elite culture of the high Enlightenment, nor the popular literature of chapbooks and bibliothèque bleue, “middlebrow” works mirrored the aspirations of a more modest category of readers, who were increasingly gaining entrance to the literary field during the eighteenth century. Smaller collections, according to this hypothesis, were also more likely to belong to female owners, whose engagements with print culture—beyond the public interventions of a few, well-known authors—remains poorly understood. Hence, in creating the MEDIATE corpus, project members sought to source as many catalogues of female-owned collections as possible. The database presently includes 45 sales catalogues, dating from 1674 to 1829, describing 44 collections owned at least in part by a woman.⁵ Two of the catalogues—those of celebrated actor David Garrick and his Austrian-born wife, dancer Eva Maria Veigel, sold in London in 1823, and that of Amsterdam-based patrician Johan Schimmelpenninck and his wife Johanna Engelina Gülcher, sold in 1829—described books jointly accumulated by a husband-wife couple. As their book ownership is impossible to separate, I include these two catalogues in the statistical analyses below both in the male and in the female collection counts. The female-owned libraries are, on average, smaller than male-owned ones: 578 books, versus 824 books in men’s libraries. The database includes, finally, six anonymous collections whose owner’s gender is unknown, and a single collection—that of the Italian Frangipani family, sold in Rome in January 1787—ascribed to a family rather than an individual owner.

Printed Female-owned Library Catalogues: Methodological Issues

The transcriptions and data extracted from the 45 printed sales catalogues of female-owned libraries in the MEDIATE database would, on the face of it, seem to present an ideal dataset to test hypotheses about eighteenth-century women’s book ownership. But auction catalogues of female-owned libraries are far from straightforward to use. As has been extensively documented by book historians, cataloguer-booksellers who drew up auction catalogues tended to focus on the most costly books, sometimes omitting altogether the most well-worn—i.e. the most well-read—and therefore hard-to-sell items. Conversely, booksellers sometimes added extraneous material to catalogues, typically unsold titles that had been gathering dust in their own storerooms (Blom et al. 2020). More troublingly for modern-day historians, like other genres of printed ephemera, library sales catalogues survive in limited numbers: only an estimated 10–20% of all auction catalogues of private libraries printed before 1800 have been preserved in at least a single copy (Jagersma 2021). Those proportions appear lower still for female-owned libraries, with a survival rate possibly of only 6%, making preservation bias a real issue in studying this material (Rozendaal forthcoming). Thus, catalogues of female-owned collections that have been preserved are often of the best-known collectors, and may present idiosyncrasies of their own. For example, the two catalogues prepared by auctioneers George Squibb in 1816 and John Broster in 1823 for the sale of the library of bluestocking Hester Lynch-Piozzi (formerly Lynch-Thrale) were ostensibly conceived as documents of special value to literary historians, as the sellers foregrounded Piozzi’s association with two famous male writers of her day, James Boswell and Samuel Johnson, specifying which books had been annotated or gifted to Lynch-Piozzi by the two men, and transcribing verbatim their marginal annotations. This, then, is not a transparent list of the books in a “typical” woman reader’s library, but rather, a performative document seeking to enhance literary reputations, and to profit from their marketing.

A thornier issue is that of comparability—a crucial issue, given the MEDIATE project’s bibliometric, transnational approach. Indeed, the sheer diversity of the corpus of extant printed library catalogues makes across-the-board generalizations hazardous. This holds especially for the more restricted corpus of catalogues of female-owned collections, since smaller numbers impede economies of scale that might flatten out statistical outliers. Although the MEDIATE team initially sought to transcribe all extant catalogues of collections owned at least in part by a female owner, it became evident that differences between them were so significant that studying the corpus as a whole would entail potentially insurmountable methodological challenges. Besides the 45 catalogues of 44 female-owned collections now in the database, the team sourced an additional 25 catalogues of female-owned libraries. These 70 catalogues show a dizzying array of sizes, dates of publication, and geographical provenance. In size alone, they range from 27 to 5,764 book items.⁶ Even in analyses restricted to a single national context, such size differences would make it difficult to pinpoint similarities between, for instance, the small collections of mostly devotional books (sometimes fewer than one hundred items) typical of northern French provincial libraries, and the thousands of volumes recorded in the libraries of great French aristocrats such as Jeanne-Baptiste, Comtesse de Verrue (3,452 items), sold in Paris in 1737, or the 5,715-item library of Louis XV’s mistress, Madame de Pompadour, sold in Paris in 1765. Given such differences, any attempt to identify such a thing as a “typical” female library would appear foolhardy. This is why, in the present analysis, I restrict myself to the smaller, more easily comparable corpus of female-owned libraries currently assembled in the MEDIATE database.

Besides issues generic to early modern private library auction catalogues, catalogues of female-owned libraries present further specificities. These include, notably, difficulties of attribution, as female-owned collections were sometimes auctioned anonymously; in these cases, knowledge regarding the owner comes from handwritten annotations or archival evidence, typically the result of serendipitous discoveries. Thus, it is only on the basis of a handwritten annotation on the copy preserved in the University of Amsterdam library that we were able to attribute the catalogue of an anonymous collection sold in Leiden in 1799 to Maria Suzanna Barnaart, the widow of society poet Jan Pauluszoon Markon (Montoya and Jagersma 2018).⁷ In other cases, female and male-owned collections were sold together, for example in the case of married couples, making it difficult to distinguish books acquired by each individual spouse.⁸ Conversely, libraries sold under the name of a male owner often included books acquired both by male and female family members. This last set of practices points to a fundamental feature of eighteenth-century libraries, namely that were rarely “private” in the modern sense of the term, but functioned instead as communal resources, within a family, within owners’ professional networks, and within the community at large (Williams 2017). Printed auction catalogues of private libraries were ambivalent documents, both public, performative displays of the owner’s social identity and public advertisements of goods for sale, and documents offering a glimpse into the private lives of their owners, with books on one’s shelves commonly perceived as revealing owners’ innermost being (Wall 1997). Because of this uneasy mix of publicity and intimacy, it could be problematic for women to associate themselves with this type of publication. When dealing with female-owned libraries, in short, there is no such thing as a “clean” catalogue, free of extraneous considerations, describing only books owned by a female collector, and providing a straightforward picture of the collection she built up and all the books she read over a lifetime. Making the catalogues speak, then, presents unique challenges.

Finally, despite the fact that the MEDIATE dataset of 600 smaller and mid-sized collections was drawn up to ensure maximum comparability between national contexts, periods, and classes of library owners, perfect equivalence between all variables is impossible to attain, given national differences in auction practices, preservation biases, and past collectors’ and booksellers’ personal choices. Thus, there remain important differences between the 551 male-owned collections and the 44 female-owned collections in the database that need to be taken into account in analyses of women’s book ownership patterns. The first is the geographic distribution of the collections, summarized in Table 12.1.

Table 12.1 shows that while Dutch collections are over-represented overall, in the corpus of female collections, different proportions obtain. French women collectors are over-represented, with French women responsible for 50%, or half of all female-owned libraries. French predominance reflects proportions in the total corpus of extant catalogues of female-owned libraries, i.e. also including catalogues not currently in the MEDIATE database. French female-owned libraries are followed by Dutch female-owned collections, making up 29.5% of the female total. British female collectors represent 20.5% of the total population of female collections, while Italian women are completely absent—although it bears noting that four of the ten Italian collections are anonymous or described as belonging to a family, rather than an individual owner. The typical “female library”, then, at least the one publicly, commercially showcased in eighteenth-century printed auction catalogues, appears to have been largely a French affair.

Table 12.1 Geographic distribution of libraries in the MEDIATE database

Discrepancies obtain, further, in the temporal distribution of male and female collections, as female-owned libraries tended overwhelmingly to cluster towards the end of the eighteenth century, contrasting with the more even spread of male libraries sold at auction (Table 12.2).

Table 12.2 Temporal distribution of libraries in the MEDIATE database

The earliest catalogue in the database dates from 1674, when the 126-book library of Neeltje Verelst (née Van Tienen), the widow of Alphen bailiff Adriaen Verelst, was put up for sale in Rotterdam. The last catalogue dates from 1829, the date of sale of the collection of Johanna Engelina Gülcher. While only a single female-owned collection was sold before the 1720s, starting in 1729, with the Paris sale of the library of the famous memorialist (and niece of Madame de Maintenon) Marthe-Marguerite de Caylus, numbers begin to rise. The sale of female-owned libraries reached a high point in the 1790s, with nine catalogues, or 20.4% of the total number of female libraries, sold in that decade alone. This concentration may be attributed both to the wave of sales that took place as French aristocrats sought to flee the country, and to the influx of French women onto the literary field that also marked the revolutionary decades, presumably making female interventions in the public sphere more common than before (Hesse 2001). By contrast, male catalogues are more evenly distributed across the century. The differences in temporal distribution between male and female catalogues mean that, just as the “typical” female library was a French one, it was also one put up for sale in the final decades of the eighteenth century, and that therefore—given the time lag of auction catalogues, drawn up mostly after the owner’s death—represented women’s book acquisitions in the 1750s through 1770s. For practical purposes, this uneven temporal distribution means that in order to establish a corpus comparable to the male one, the starting date for the female corpus needs to be set later than the MEDIATE dataset starting-point of 1665. In the analyses below, I therefore take 1729 as a starting date, as after this date catalogues of female-owned collections appeared regularly, at a rate of at least two per decade.

Female Libraries and the Bibliothèque choisie

The question whether there was, indeed, such a thing as a specifically female-gendered library in the eighteenth century, and whether women generally read differently than men—different kinds of books, for different purposes—has been answered variously by modern historians. Sabine Heissler, for example, has posited that early modern women’s libraries fell into three distinct categories: first, small collections devoted primarily to devotional works, sometimes supplemented by a handful of practical, domestic books; secondly, the libraries of learned women noted in their day for their erudition or contribution to scholarly debate; and finally, sizeable libraries, put together by aristocratic women, which assembled large numbers of literary works or belles-lettres, in particular novels, sometimes authored by women (Heissler 1998). But eighteenth-century commentators, too, interrogated relations between libraries and gender. Accompanying the rise in public sales of female-owned libraries, a new discourse took shape describing an ideal type of library, the so-called bibliothèque choisie. Breaking with the older model of the large scholar’s library or bibliothèque universelle, this new kind of library did not primarily reflect professional needs, but instead foregrounded personal reading taste and belles-lettres, i.e. works of poetry, plays, and novels read not in the practice of a profession, but for pleasure. Because it did not aim for comprehensive coverage of all fields, this kind of library was considerably smaller than the great scholars’ libraries: a few hundred titles, not thousands, with this difference sometimes denoted in English by the terms “closet” or “cabinet”, suggesting just a handful of books, rather than the official “library”. Ideas about the bibliothèque choisie were popularized in review journals such as Amsterdam-based journalist Jean Le Clerc’s Bibliothèque choisie pour servir de suite à la Bibliothèque universelle (1703–1708), and underwrote the rise of a new genre of advice literature targeting readers new to the literary field (Montoya 2021a). Taken together, these attributes meant that the bibliothèque choisie and the English “closet” were increasingly also conceptualized as specifically female-gendered spaces—not altogether unlike the fictional library described by Elisabeth Maria Post, with its sprinkling of titles in different genres, but focusing especially on poetry, plays and novels.

In a few cases, the bibliothèque choisie could be identified with real, historical women readers. One such woman was Maria Leti, the Italian-born daughter of popular historical author Gregorio Leti, and wife of Jean Le Clerc (Montoya 2009). Maria Leti assisted her husband in his scholarly work, as evidenced by her mention in letters of correspondents including John Locke and Pierre Bayle. With her sister Susanne, she translated her father’s works into French. When Maria Leti’s small personal library was put up for auction in Amsterdam in 1735 alongside her husband’s, differences between them mirrored differences between the scholarly, male-gendered bibliothèque universelle (Le Clerc) and the more modern, female-gendered bibliothèque choisie (Leti). Le Clerc’s library was considerably larger than his wife’s, numbering 2,756 lots (excluding manuscripts), versus his wife’s 677 lots (703 books). The content of the two libraries, classified according to the eighteenth-century Parisian booksellers’ subject classification system, also differed significantly. Le Clerc’s catalogue listed relatively more books categorized as “theology”, “law” and “arts and sciences”, while Maria Leti’s library on the contrary foregrounded history and belles-lettres, which made up 61.4% of its content, versus 34.2% in Le Clerc’s library (Table 12.3).

Table 12.3 Subject categorization of books in Jean Le Clerc and Maria Leti’s libraries (1735)

While Maria Leti and Jean Le Clerc’s library catalogues provide an exceptional opportunity to compare a male and female library assembled in the same years and cultural context, the MEDIATE database allows historians to expand this comparison by drawing on a larger dataset—with one caveat. As data-enrichment work is still ongoing, not all collections have undergone subject-categorization of their books, which remains uneven even within individual catalogues. The proportion of uncategorized books is highest in male-owned collections: 68.27%. In female-owned collections, the proportion of uncategorized books is 36.28%. In the seven collections whose owner’s gender is unknown or corporate (the Frangipani family library), 47.69% of the books remain uncategorized. Notwithstanding this data health warning, a global analysis reveals proportions that are comparable to those in the Le Clerc-Leti catalogues (Table 12.4).

Table 12.4 Subject categorization in female, male and anonymous collections, 1729–1830

Similarly to proportions in the Le Clerc-Leti catalogues, men’s libraries contain relatively more books categorized as “theology”, “law” and “arts and sciences” than do women’s. This follows from the male practice of professions from which women were barred, respectively in ecclesiastic functions, as magistrates and government officials, and as medical doctors (medical works accounting for many of the books labelled “arts and sciences”). Female libraries, by contrast, because of their focus on reading for pleasure, report larger proportions of books of history and belles-lettres. Additionally, women’s catalogues list larger proportions of books subject-categorized as “unknown”. This is because female catalogues were often less bibliographically precise than male ones, more frequently listing bundles of unspecified books, described for example as “trente-un volumes dépareillez de divers Ouvrages in douze” (in the 1729 catalogue of Caylus’s books) or “Fifty-two miscellaneous” (in the 1783 catalogue of the library of one “Mrs. Armstrong”). Anonymous or corporately owned libraries, finally, display subject-category proportions closer to the female-owned collections than to the male ones. However, their number is too small—six, four of them Italian—to warrant far-ranging statistical conclusions.

From Subject Categories to Individual Authors

Let us turn now from broad subject categorizations to the twenty individual authors reported in the largest number of female-owned and male-owned libraries (Table 12.5). In keeping with inconsistent author attributions in the catalogues, the term “author” is understood capaciously here, including named and unnamed authors, translators, editors, illustrators, and corporate authors such as the Catholic Church or literary societies. When books by a given author are present in the same number of libraries, their relative ranking is determined by the number of individual books reported in the catalogues: the higher the number of copies of their works, the higher the author ranking.

Table 12.5 Authors reported in female- and male-owned libraries, 1729–1830

There is some overlap between the two lists: twelve authors or designations (in the case of the Bible or unspecified books) appear on both. But there are also differences. First, frequencies overall are lower in the female collections, most likely a consequence of the smaller size of women’s libraries. Beyond this, differences emerge in the reading matter consumed by the two groups. Eight authors appear on the male list but not on the female one: Homer, Terence, Erasmus, Grotius, Juvenal, Seneca, Plutarch, and Pufendorf. Five of these fall primarily into the category belles-lettres, two in the category “law”, and one in “history”. Most are authors from classical antiquity, who were central to male schooling, and may hence reflect, at least in part, left-over schoolbooks from an earlier phase in their owners’ lives. More generally, they embody the continuing relevance of classical learning in eighteenth-century culture (Montoya 2021b). Thus, women’s lack of access to formal, Latin-language schooling appreciably impacted their choice of reading matter.

Mutatis mutandis, eight authors appear among the women’s top twenty, but not among the men’s: Cervantes, Boileau-Despréaux, Molière, the Catholic Church, La Fontaine, Thomas à Kempis, Pierre Corneille, and Richard Steele. Six fall into the category of belles-lettres, while the remaining two can be subsumed under “theology”. This latter finding points to an interesting phenomenon: while men’s libraries report larger proportions of works of theology than women’s, in female-owned collections some individual religious titles significantly outperform their standing in male collections. Thomas à Kempis Imitatione Christi, for example, figures in 46.5% of the female-owned collections, but only 34.6% of the male-owned ones. At the other end, among non-religious works, is the higher standing of Voltaire in women’s libraries, 74.4% versus 62.1% in men’s libraries. Voltaire’s top-ranking works are his Œuvres complètes (in 34.8% of women’s libraries, but only 22.6% of men’s), followed by his Henriade (25.5% and 20.9%) and Histoire de Charles XII (23.2% and 17.4%).

In order to enable a more fine-grained appreciation of the individual books behind blanket designations such as “Voltaire” or “Catholic Church”, the MEDIATE database also provides some basic matching of books to Virtual International Authority File (VIAF) Work records. Given both the labour-intensive and hence slow pace of VIAF Work matching, as well as the inconsistencies and incompleteness of VIAF itself, so far only 14% of the books have been matched to VIAF Works. Nonetheless, some patterns have already begun to emerge (Table 12.6).

Table 12.6 VIAF Works reported in female- and male-owned libraries, 1729–1830

The list of top-ranked authors in male-owned collections, once again, is made up overwhelmingly of classical authors, whom most male readers would have first encountered during their school days. Fourteen, or 70% of the top twenty male-owned titles were classical authors. By contrast, only two, or 10% of the top authors in female libraries dated back to classical antiquity. Instead, over half of the books in women’s libraries—eleven, or 55%—dated from after 1500, versus three in male-owned libraries. Five authors who made the author top (Table 12.5) fail to surface in the top VIAF Works ranking—Voltaire, Cicero, Erasmus, Pierre Corneille, Grotius, and Pufendorf—indicating that these authors’ impact was due not to one or two successful titles, but to a prolific output in multiple genres.¹² Conversely, a number of single-title successes emerge in women’s collections that were obscured by an author count alone, like Mézeray’s Abrégé chronologique de l’histoire de France (1667–1668) and Montaigne’s Essais (1580–1595). Religious service books now figure prominently among the top-ranked female-owned titles: both missals and books of hours make an appearance, but remain absent on the male list. Finally, for the first time a female author breaks through to the top twenty: Marie de Rabutin-Chantal, marquise de Sévigné, whose Lettres (1725) rank sixteenth of titles reported in the largest number of female-owned libraries.

“Women’s Books”: Devotional Works and Novels

Thus far, our analysis of eighteenth-century private libraries tends to confirm narratives about women as readers primarily of religious books and novels. However, the slightly skewed composition of the corpus needs to be considered. Catalogues of female-owned libraries, as noted above, were produced more often in France than elsewhere. Auctions of women’s collections were also concentrated in the final decades of the eighteenth century. So might Tables 12.5 and 12.6 simply reflect French reading taste at a particular moment in time? One way to test this hypothesis is to limit the analysis to French libraries from the second half of the century (Table 12.7).

Table 12.7 VIAF works reported in female- and male-owned French libraries, 1750–1830

Clearly, differences between female and male book ownership persist even when geographic variables are eliminated, as only 50% of the works on the female and male lists overlap. Table 12.7 also sheds light on women’s religious book ownership. Service and devotional books now account for three of the female top ten titles, but still fail to appear on the male list. At the same time, the Bible’s more prominent place in male libraries may seem surprising, given the evidence of women’s engagement with religious reading material. Complete editions of the Bible are reported in 80.9% of the male libraries, but do not appear at all in the female top—although editions of the New Testament do figure in 55.5% of women’s libraries. While religious books were among women’s preferred reading, then, not all religious books did equally well in their libraries. But the apparent absence of Bibles also tells us something about the nature of these female library catalogues. The seven collections that do not report a Bible are—with a single, English exception—all French. This may initially seem to confirm tired old clichés about Catholic readers’ infrequent reading of the Bible. But a closer examination reveals that some of the French libraries that did not list a Bible did have a strong devotional content. The small library of Marie-Marguerite-Joseph Imbert (124 items), sold in Lille in 1756, lists no Bible but does list several Catholic service books, breviaries, and a copy of Thomas à Kempis’ Imitatione Christi—as well as multiple titles described simply as devotional works. Similarly, the even smaller library of Marie Angélique Desurmont (91 items), also sold in Lille, in 1778, contains no Bible but does report several breviaries, unidentified “paquets”, and again, a copy of the Imitatione Christi. Might it be possible that if these catalogues did not list a Bible, it could be because the owner had literally read it to pieces?¹³

Besides being associated with religious reading, women were also perceived to be avid consumers of novels, both by eighteenth-century contemporaries and modern critics: according to a long-standing trope, “women were the primary consumers of novels” (Diaconoff 2005, 7). The trope of the female novel reader reappeared regularly in eighteenth-century texts, including novels authored by women writers (Pearson 1999). Charlotte Lennox’s The Female Quixote (1752), for example, portrays the protagonist Arabella, the eponymous “female Quixote”, as a prototypical female reader of Baroque novels or “Romances” by French authors like Georges and Madeleine de Scudéry, and Gautier de Costes de La Calprenède:

From her earliest Youth she had discovered a Fondness for Reading, which extremely delighted the Marquis; he permitted her therefore the Use of his Library, in which, unfortunately for her, were great Store of Romances, and, what was still more unfortunate, not in the original French, but very bad Translations. […] the deceased Marchioness had purchased these Books to soften a Solitude which she found very disagreeable; and, after her Death, the Marquis removed them from her Closet into his Library. (Lennox 2008, 7)

Suggesting a qualitative difference between two kinds of library, the female-gendered “Closet” and the male-gendered “Library”, the narrator underlines the feminisation both of the “book closet” and the novel. As in the course of the eighteenth century, novels moved from older “Romance” formats to more realistic conventions, literary critics and cultural commentators continued to evoke this special affinity between women readers and novels. In the visual arts, a memorable series of paintings and drawings by Joshua Reynolds portrayed his niece Theophila Palmer engrossed in Samuel Richardson’s Clarissa Harlowe (1747–1748). Women readers—and in their footsteps, women writers—responded with particular feeling to Jean-Jacques Rousseau’s La Nouvelle Héloïse (1761). So it comes as no surprise that in the late eighteenth-century novel Het land, the only four books referenced by title in Emilia’s “book cabinet” were also novels: Richardson’s Clarissa and Sir Charles Grandison (1753–1754), Johann Timotheus Hermes’ Sophiens Reise von Memem nach Sachsen (1769–1773) and Johann Martin Miller’s Geschichte Karls von Burgheim und Emiliens von Rosenau (1778–1779). Taking all these titles as a starting point, let us then examine the historical basis of this trope of the novel-consuming female reader (Table 12.8).

Table 12.8 Selected novels in female- and male-owned libraries, 1729–1830

While Renaissance classics like Torquato Tasso’s Gerusalemme liberata (1581), Ludovico Ariosto’s Orlando furioso (1516) and Miguel de Cervantes’s Don Quixote (1605–1615), are reported in similar frequencies in male and female libraries, the numbers for other novels are higher overall—sometimes significantly so—in female collections. The romances of Scudéry and La Calprenède so avidly consumed by Lennox’s Arabella are reported up to three times more frequently in women’s collections than in men’s. A few modern novels stand out as runaway bestsellers in women’s libraries. They include Marie-Madeleine de Lafayette’s Princesse de Clèves (1678), reported four times more often in female-owned than in male-owned collections, and Richardson’s Clarissa and Rousseau’s Nouvelle Héloïse, which figure up to three times as often in women’s libraries than men’s. Only a very small handful of novels manage to do better in male-owned collections than in female ones, and differences are minimal at best: François Rabelais’s Œuvres, reported in 29.0% of male libraries, versus 27.9% of female ones; John Barclay’s political allegory Argenis (18.4% and 16.3%), and a few eighteenth-century titles whose absolute numbers, however, are too low to be statistically significant. In short, the stereotype of the novel-reading woman reader, contrasted to male readers who supposedly consumed more serious material—or were too busy exercising roles in public life to read for pleasure—does appear to have some historical basis, although differentiation needs to be made between specific titles, as some novels did appeal particularly to male readers.

Women’s Libraries as a Space of Female Intimacy

If novels and devotional works, then, were the reading-matter of choice of eighteenth-century women, are there other books that appeared exclusively in female-owned libraries? One candidate might be books that explicitly targeted a female audience. Works aimed at female readers can be found in both male- and female-owned collections. However, they do tend to occur more frequently in the latter. Francesco Algarotti’s Il Newtonianismo per le dame (1737) is reported in five female-owned collections, and in fifteen male-owned collections, or respectively in 12.8% of the female-owned and 3.9% of the male-owned collections whose catalogues were published in 1737 or later. By contrast, a title that sought to popularize Newton’s ideas, but without explicitly targeting a female readership, Voltaire’s Eléments de la philosophie de Newton (1738) is reported in 7.7% of the female and 7.7% of the male catalogues from its date of publication. But on the other hand, the female-authored but not female-aimed Institutions de physique (1740), a work of popular Newtonianism by Gabrielle-Emilie du Châtelet, is reported in 5.1% of the female-owned but only 2.9% of the male-owned collections from its date of publication.

This last example points to another possibility, namely that women library-owners might have shown a marked preference for works by other women. Although studies addressing earlier periods have produced inconclusive outcomes (Coolahan and Empey 2016), eighteenth-century library auction catalogues suggest that by this date, there was a movement, reinforced by critical discourses around the bibliothèque choisie and women’s “closets”, towards new, distinctly female-gendered types of library. The MEDIATE data demonstrates that eighteenth-century women did, indeed, own works by other women more often than did their male counterparts. In female-owned libraries sold in the period 1729–1830, books by female authors represent 5.6% of the total, versus 1.9% in male-owned collections and 2.3% in anonymous or family libraries. In other words, although overall percentages remain low, female-owned collections report over twice as many books by female authors as do male-owned collections. This holds, significantly, for all periods. In the decisive decades 1771–1800, for example, during which works by women authors increasingly became available, percentages rise even more, to 7.2% in female-owned collections, while the proportion in male-owned libraries lags even further behind, at barely 2.2%. And just as novels consistently did better in women’s libraries than in men’s, so too did women authors consistently achieve greater success in female-owned collections than in men’s (Table 12.9).

Table 12.9 Female authors reported in female- and male-owned libraries, 1729–1830

The lone female author who was more successful among male audiences—although only very slightly, as she is reported in 40.3% of male-owned libraries, versus 39.5% of female-owned ones—is translator Anne Dacier. Rather than any particular engagement with female authorship, this simply appears to reflect, yet again, male collectors’ predilection for the authors of classical antiquity, whom Dacier translated. Five women authors appear on the female list but not on the male one: the memorialists Sévigné, Françoise d’Aubigné, marquise de Maintenon, Françoise Bertaut de Motteville, and the novelists Marie-Catherine de Villedieu and Marie-Angélique de Gomez. Conversely, five authors appear in the male top but not the female one; however, in all cases, their frequency in female-owned libraries is actually higher—despite their lower relative ranking—than in male-owned libraries. At the same time, the female partiality towards works by female memorialists appears noteworthy, as it may suggest broader engagements, on the part of women library owners, with issues of female subjectivity, female self-fashioning, and women’s place in history—in short, with the public performance of new, varied female identities.

There emerges, then, a final meaning that could be given to the female-gendered library, as a “room of one’s own” and shared female space. Libraries could function as a communal female space by gathering together the works of women writers. But they could also become meeting-places for communities of women readers. As Gerda Lerner has posited, “in the modern world, clusters of learned women […] appear in the form of supportive networks of female friends, which I will call ‘affinitive clusters’. In the seventeenth and eighteenth centuries, female readership constitutes such affinitive clusters” (Lerner 1993, 226). Eighteenth-century libraries, as noted, could rarely be considered “private” or solely a reflection of “individual” reading choices, as they were embedded in larger family and professional networks, and hence became loci of various interpretive communities. In Elisabeth Maria Post’s Het land, the female protagonist Emilia sought to share her library or “book cabinet” with her female correspondent, Eufrozyne, turning the library into a space of collective, female intimacy. By reading together, new forms of female agency and solidarity could be forged.

In a few cases, there is evidence that real, historical affinitive clusters also emerged around the libraries described in eighteenth-century auction catalogues. Through the correspondence of Jean Le Clerc with John Locke, for example, the figure emerges of another female intellectual, the philosopher Damaris Cudworth Masham, daughter of Cambridge Platonist and theologian Ralph Cudworth, who used her bookish exchanges with her male colleagues to establish ties with a female counterpart in Amsterdam, Maria Leti. These networks can be traced through correspondences, but also through the libraries that linked these figures together, and were described in a series of catalogues: the multiple manuscript catalogues that John Locke made of his books up until his death in 1704, the 1691 printed auction catalogue of the library of Ralph Cudworth, and the catalogue of Maria Leti’s books printed in 1735.

Whereas emerging female networks in the early eighteenth-century Republic of Letters were still largely virtual, they had acquired more consistency by the end of the century, as evidenced by the auction catalogues of female-owned libraries, such as the 1816 catalogue of Penllyn Library, housed in Penllyn Castle in Glamorgan, Wales. This 709-item library had been assembled by Emilia Gwinnett (1741–1807), the youngest daughter of clergyman Samuel Gwinnett of Gloucestershire, and sister of the more famous Button Gwinnett, who after emigrating to the American colonies, became one of the signatories of the American Declaration of Independence. Probably sometime after 1775, Emilia Gwinnett settled in Wales, where she moved into Penllyn Castle together with its owner, the widow Lady Louisa Barbara Vernon. Upon her death in 1786, Louisa Vernon left the castle and library to Emilia Gwinnett, disinheriting her own daughter to do so. This is a first indication that the relationship between the two women was an especially intimate one. Further evidence is provided by Emilia Gwinnett’s request in her will that she “be buried […] as near the remains of the late Lady Barbara Vernon as may be”¹⁴—a detail that recalls another, more famous female couple, that likewise settled in Wales in 1780, creating a library around which various literary networks converged, the “ladies of Llangolen” Eleanor Butler and Sarah Ponsonby. The traces left by Emilia Gwinnet and Louisa Vernon of the life they shared at Penllyn Castle, with a library at its very centre, are exemplary of the many functions libraries could play as spaces of female intimacy—albeit in this case, one made public after their deaths by the sale of their library. Female libraries thus brought together the works of female authors and shaped new female relationships to books, and new relations between female readers, in a complex interplay between fact and fiction, material culture and critical discourse.

With the library of Emilia Gwinnett, then, we have moved decisively from literary accounts of women’s books and female intimacies in Elisabeth Maria Post’s fictional “book cabinet”, to historical evidence about women’s actual book ownership. Bibliographically speaking, eighteenth-century female-owned libraries tended to lean towards ideals embodied by the bibliothèque choisie, focusing on belles-lettres and reading for pleasure rather than for the professions. But these libraries also shaped new, bookish female identities, meshing public performances of reading with more private subjectivities. Printed sale catalogues provide unique source material to explore these hybrid identities; coupled with modern digital tools such as the MEDIATE database, they allow historians for the first time to draw large-scale comparisons between female and male book culture. But we are only at the very beginning of attempting to tease out some of the insights that this rich, under-explored source may still provide about eighteenth-century women’s myriad engagements with the printed word, and how “the woman reader” was historically constructed.