Following the Mediterranean Paper Trail: A Study of European Paper in Late Medieval Cairo (c. 1350–1600)

AS IS WELL KNOWN, PAPERMAKING TECHNOLOGY first spread from China and Central Asia into the Middle East during the eighth century CE. In subsequent centuries, it moved across Islamicate lands and entered western Europe via Spain. By 1300, Italian papermakers had devised a new method of laid-paper production using metallic wire moulds. This technique allowed the introduction of watermarks to Italian paper, distinguishing it from the non-watermark paper still widespread in the Middle East and North Africa.¹ It is also well known that watermark paper began to reach Middle Eastern markets from Europe during the fourteenth and fifteenth centuries, eventually supplanting other paper in the region.² Many paper manuscripts produced in this period are extant from Middle Eastern communities, and the physical features of their paper can inform the material history of both Europe and the Ottoman Empire. However, despite its great potential for more in-depth analysis, scholars of European codicology generally do not take this evidence into account when investigating the diaspora of watermark paper. Moreover, the past surveys of watermarks in Middle Eastern manuscript collections have been necessarily limited in geographic and temporal scope.³ These issues stem from the disciplinary barriers between European and Middle Eastern history: most historians of Europe are not trained to work with Arabic, Persian, Hebrew, or Turkish manuscripts, nor are scholars of Middle Eastern manuscripts trained in the history of European papermaking and codicology.

This article seeks to cross those disciplinary boundaries while examining watermark paper in the ‘Cairo Genizah’ manuscript collections at the Cambridge University Library. Mainly dating between 1100 and 1897, this corpus provides a continuous cross-section of Egyptian-Jewish literary activity during the entire transition to watermark paper, including samples of paper stocks used in Cairo, Alexandria, Rosetta, Guadalajara, Híjar, Lisbon, Faro, Salonika, Constantinople, and Jerusalem, among other places. As we shall see, Genizah data about these stocks can supplement the study of paper in Egypt, Italy, Iberia, and the Ottoman Empire between 1350 and 1600. It also presents new opportunities for collaboration between European and Middle Eastern historians that will enhance the current under standing of papermaking and trade around the Mediterranean Sea.

The corpus and methodological concerns

The Taylor-Schechter Collection, the Mosseri Collection, and the Lewis-Gibson Collection at Cambridge University Library contain close to 200,000 fragments of manuscripts and printed texts produced between 600 and 1900 CE. Nearly all these fragments come from what is known as the ‘Cairo Genizah’. A ‘genizah’ is a hidden space, often in a synagogue or cemetery, where Jewish communities deposit old, damaged sacred texts to prevent them from being profaned. The ‘Cairo Genizah’ refers to a specific corpus of texts that Cairene Jews preserved in this way, mainly within the Ben Ezra Synagogue of Old Cairo, for most of the second millennium. This genizah is significant because, in addition to ‘sacred’ texts such as bibles and prayer books, the Ben Ezra community also deposited thousands of non-religious texts from practically every genre they could access. Due to the actions of numerous individuals involved in the nineteenth-century Egyptian manuscripts trade, the Cambridge collections contain the majority of extant Cairo Genizah manuscripts.⁴ Other institutions, most notably the Bodleian Library (Oxford), the British Library (London), the Russian National Library (St Petersburg), the John Rylands Library (Manchester), and the Jewish Theological Seminary (New York), hold smaller collections. This article focuses on Cambridge material because it was physically accessible for examination by the authors, but our methodology could be expanded to these other Genizah collections.⁵

This is the first time that a study on the spread of watermark paper has been attempted with Cairo Genizah material. In the last 125 years, nearly all Genizah researchers have been scholars of Hebrew, Judaeo-Arabic, Middle Eastern History, and medieval Judaism. Their work has focused on the ‘peak’ period of the Cairo Genizah between the eleventh and thirteenth centuries. Accordingly, many later items on watermark paper remain uncatalogued, and many Genizah catalogues do not document the physical features of paper at all. It is thus difficult to conduct even a basic survey that represents the breadth of watermark paper stocks present among Genizah fragments.

The authors utilized the little data already available for the Cambridge collections to assemble a corpus for studying the emergence of watermark paper in medieval Egypt. This was a pilot, scoping project. Rather than surveying thousands of fragments to find every piece of watermark paper, we first focused on manuscripts that were internally dated to establish a chronology. Our initial source for identifying such manuscripts was Pierre Delbes’ Les documents datés de la Geniza du Caire. We examined Delbes’ list of dated documents for the period between 1250 and 1600 to confirm (or some times correct) his given dates and identify the types of paper. Most of these manuscripts do not have extant watermarks, but they do clearly depict an Egyptian transition to watermark paper during the fifteenth and sixteenth centuries (see Appendix below). We supplemented Delbes’ list with the limited number of Genizah manuscripts that were already identified as ‘watermarked’ in the Cambridge University Digital Library catalogue.⁶ Further data on paper dated between 1480 and 1600 came from Posegay’s previous survey of printed material in Genizah collections.⁷ Within the resulting corpus of approximately 300 classmarks, we recorded around 115 extant watermarks.⁸ This subset does not include all the watermark paper present in the Cambridge Genizah collections, and this number of surviving water marks does not reflect the proportion of watermarked paper in the entire Cairo Genizah. In fact, we estimate that less than ten percent of the watermarks in Cambridge Genizah collections have been documented to any degree. Nevertheless, our survey uncovered significant new data for the historical distribution of paper around the Mediterranean Sea during the late medieval period.

The current physical condition of Genizah material further complicates the issues of a limited corpus. Most Genizah manuscripts are fragmentary, scraped, stained, water-damaged, or worm-eaten, and, at least in Cambridge, nearly all have been removed from their original quires for ‘conservation’.⁹ The conserved fragments, such as they are, now reside within plastic folders that cannot always be easily measured or transversely lit. Most of these folders are large, cumbersome, and contain as many as several hundred unrelated fragments that must all be requested together from Cambridge University Library’s manuscripts reading room.¹⁰

Finally, our designation of ‘watermark paper’ indicates rag paper produced using metallic wire moulds that could imprint watermarks, ultimately descended from thirteenth-century Italian methods. Watermark paper can be identified by the density and regularity of its laid and chain lines even if no watermarked image is present. It is thus possible to differentiate ‘watermark paper’, almost exclusively produced in Europe, from the ‘non-watermark’ rag paper more common in the Middle East and North Africa.¹¹ This study assumes that all or nearly all non-watermark paper found in the Cairo Genizah was made in Islamicate lands.

Watermark paper in the Cairo Genizah

Every manuscript in our sample copied before c. 1375 is non-watermark paper. These items comprise all manner of genres found throughout Genizah history, including poetry, literary works, legal contracts, personal letters, prayer books, bibles, and court records. Some of these records are related to activity at the Egyptian court of Joshua Nagid (d. 1355),¹² the great-grandson of Maimonides and the nagid (‘prince, leader’) of the Egyptian Jewish community from 1334 to 1355. Joshua’s son, David, succeeded him and served intermittently as nagid until his death around 1415.¹³ At least two documents related to his tenure also survive in the Genizah, including a record of a real estate transaction dated to 1409.¹⁴ Despite the spread of new papermarking technologies in western Europe at this time, all of Joshua and David’s documents are still non-watermark paper. From the same period come a pair of documents dated 1359 and 1387, respectively.¹⁵ They are also on non-watermark paper, most likely produced in the Middle East.

These few fourteenth-century manuscripts reveal an issue with the Cairo Genizah when using it to trace the spread of paper. The corpus contains material that Cairene Jews deposited continuously from the eleventh century until the end of the nineteenth century, but the greatest concentration of deposits came before the second half of the fourteenth century. There is then a lull in manuscript deposits, followed by a second, smaller peak in the sixteenth century, caused in part by large numbers of Jewish refugees who fled Spain and Portugal in the 1490s.¹⁶ Aforementioned documents notwithstanding, the reduction in Genizah activity means that there are relatively few items that we can be sure were in Egypt between about 1355 and 1492. Our corpus does include a court record and a Jewish ethical text dated, respectively, to 1458/9 and 1481, but both are on non-watermark paper.¹⁷

Some of the earliest watermark paper in the Genizah that was used for writing in the Middle East is connected to Nathan Sholal, one of the last Egyptian nagids. He served from 1484 to 1502 and, notably, lived in Italy and Jerusalem before settling in Egypt.¹⁸ One legal document that mentions him is from 1489, written in Egypt but watermarked with the top half of a hand and small five-pointed star/flower.¹⁹ Another record is dated 1495/6, but this one is non-watermark paper, suggesting an ongoing transition in the late fifteenth-century Cairene paper market. There are also two letters sent to Sholal from Jerusalem, both on watermark paper.²⁰ The earliest dated document in our corpus unrelated to Sholal is a matchmaking agreement signed in Milig, a city in the southern Nile Delta, in 1492.²¹ For this one, another partial watermark survives: the top half of a hand with a large six-pointed flower (Fig. 1). It seems to belong with a family of watermarks widely attested in Spain during the 1490s. Whether or not a Spanish papermaker produced it, it almost certainly came from Europe before ending up in Egypt.²²

The Cairo Genizah also preserves numerous manuscripts that, while not internally dated, have known watermarks that allow us to estimate their period of copying. The following tables include some examples of water-mark paper from our corpus, aiming to demonstrate new geographical data for the diffusion of European paper stocks for several different types of texts. Each table provides the classmark, measurements of watermarks in millimetres, comparable watermarks as traced from published albums, and comparative dates as listed by those albums.²³ These ‘comparanda’, practically all from non-Genizah manuscripts, offer locations where paper with a given watermark has been attested. Those locations are not necessarily where that paper was produced. The earliest watermark in our corpus is a letter ‘R’ with a simple cross that appears in a Hebrew medical treatise (Fig. 2). It resembles paper previously documented in Genoa (1377) and Paris (1382).²⁴ Many other Genizah watermarks are datable to the fifteenth century (Table 1).

It is likely that many of these manuscripts were copied in Iberia and came to Egypt with Jewish refugees after 1492. They are primarily from large codices of medical and biblical texts copied in ‘Sephardic’ scripts,²⁶ typical of Jews in the western Mediterranean. These types of books would have held considerable value over multiple decades, which might explain their survival from the early or mid-fifteenth century until after the Iberian exile. It could then be assumed that much of their paper was first used in Spain or Portugal, ultimately sent there from Italy and France.²⁷ That said, some Genizah manuscripts do indicate that a small quantity of blank European-made watermark paper was already reaching Egypt in the middle of the fifteenth century. For example, T-S A26.190 (c. 1440, see Table 1) is from a fairly small Bible codex written in a more eastern Hebrew script style, suggesting it was copied in the Middle East. Another example is a note related to a dowry negotiation, also written in an eastern style, with a partial pinecone watermark that may match Briquet 12437 (Susa 1459, Pinerolo 1462).²⁸ Finally, T-S AS 180.199, which is an Arabic medical prescription, has a partial key watermark with a broken chain line matching a mould used in Fabriano in 1488.²⁹

In addition to manuscripts, the Cambridge Genizah collections contain hundreds of Iberian incunable fragments with known publication dates and locations. Some of these incunables have identifiable watermarks, enabling us to use Genizah material as a resource for studying the distribution of paper stocks within Europe itself. Some of this paper even appears from European cities where it has not previously been documented. For example, there are numerous imprints of Talmud tractates that Solomon ibn Alkabe? printed in Guadalajara around 1480. We have recorded at least three different watermarks among them, including a hand with mounted crown (Fig. 3; measuring 85 × 25 mm) similar to marks known from Barcelona (1475) and Burgos (1475).³⁰ Other incunables are listed in Table 2.

Like the above manuscripts, most of these printed books probably came to Egypt with refugees fleeing the Iberian exile,³⁵ although it is also possible that some reached the Middle East via book traders shortly before 1492.³⁶

In subsequent decades, Iberian Jewish refugees established new printing presses in Ottoman territory, especially in Salonika and Constantinople.³⁷ Their greatest local competitors were the famous Soncino family, led by Eliezer and Gershom Soncino, who themselves had left Italy in search of more favourable conditions for Jewish printers.³⁸ Many pages of sixteenth-century Ottoman Hebrew books thus survive in Genizah collections. We can use them to identify paper stocks in Salonika and Constantinople at the time of their printing (Table 3).

One interesting example is the earliest known Passover Haggadah printed with illustrations.³⁹ Its publisher and location are not known for certain, but Eva Frojmovic has argued that it was produced in Constantinople around 1506. She bases this conclusion on the Haggadah’s watermark (Fig. 4), which matches the paper in a separate book that the Ibn Nahmias brothers printed that year in Constantinople.⁴⁰ Frojmovic finds no record of this water mark (a hand with spread fingers and an unusually rounded cuff) in Briquet, and the present authors have been unable to identify it in any catalogue. It is possible that this paper was produced by an as yet undocumented manufacturer using watermark papermaking methods in Constantinople, although the Ibn Nahmias brothers also used watermarked paper from southeastern France around the same time.⁴¹

It is here that we must carefully distinguish between the spread of paper and the spread of Italian paper-marking technology between Europe and the Middle East. It is generally accepted, both in Genizah studies and in Islamic codicology, that the introduction of European-made watermark paper to Middle Eastern markets disrupted and eventually eliminated local paper-making methods.⁴⁶ By about 1600, Europe supplied most Middle Eastern paper markets, and older non-watermark paper persisted mainly in peripheral regions. The large number of European crown-star-crescent, tre lune, and anchor watermarks that appear in the Ottoman Empire during the sixteenth and seventeenth centuries supports this conclusion.⁴⁷ It is only later, with rare exceptions,⁴⁸ that Middle Eastern papermakers began employing European techniques to create their own watermarks.⁴⁹

The assumption is that all, or nearly all, watermark paper extant from the Middle East from around 1350 to 1600 was physically produced in Europe. However, we should be cautious before drawing this conclusion for the Genizah corpus. The Jewish printers who left Iberia and Italy to settle in the Ottoman Empire would have been accustomed to working with watermark paper. Their eastward migration would have contributed to the increased demand for Italian paper in Ottoman territories, just as earlier centuries saw Italian paper sold to bookmakers in Iberia.⁵⁰ If Iberian and Italian Jewish printers brought their craft across the Mediterranean and established successful businesses in the Ottoman Empire, then papermakers (both Jewish and Muslim) could have done the same. If they produced paper with the techniques they had used in Europe, then we would expect their watermarks to look essentially the same as European ones, at least during the sixteenth century. This situation would mirror the use of Iberian and Italian typography at the early Hebrew printing presses in Salonika and Constantinople.⁵¹ More thorough documentation of paper stocks in Middle Eastern manuscript collections will clarify how much, if any, watermark paper was produced in places like Constantinople, Jerusalem, and Cairo before 1600.

With that in mind, we now return to Egypt, where watermark paper did become increasingly common after 1500. Many Genizah manuscripts are internally dated to this period, and by the end of the sixteenth century a large majority were on watermark paper (see Appendix below).⁵² Some of that paper has further identifiable watermarks. Several of the items in the following table were copied or sent outside of Egypt, but all were found in Cairo during the second half of the nineteenth century (Table 4).

These manuscripts indicate that Cairo and other Egyptian cities had access to various European (especially Italian) watermark paper stocks throughout the sixteenth century. This fact is again not surprising, as trade between Egypt and Venice in particular is well documented in both the Mamluk (pre-1517) and Ottoman (post-1517) periods.⁵⁸ Other Genizah material also attests to this commercial connection, including a significant Egyptian demand for Hebrew books printed in Italy.⁵⁹ Evidently, water-mark paper outcompeted local non-watermark paper, contributing to the decline of older papermaking methods in Egypt. Our survey (see Appendix) confirms that by the second half of the sixteenth century, almost all paper used in Cairo was watermark paper, most likely imported from Europe.⁶⁰

There is, however, literary evidence that a group of people known as ‘papermakers’ or ‘stationers’ (warrāqīn in Arabic, sing. warrāq) existed in Cairo beyond 1600. Anthony T. Quickel notes several Arabic chronicles that refer to Cairene warrāqīn ‘just before and after the Ottoman takeover of Egypt’ and up through the eighteenth century.⁶¹ These warrāqīn provided various book-related services, including copying, binding, and selling, but the extent to which they manufactured paper in Cairo (both during the Ottoman period and in earlier centuries) is not precisely known.⁶² If, as our data supports (see Appendix), watermark paper supplanted other Egyptian paper during the early Ottoman period, then a Cairene warrāq of the late sixteenth century was likely more of a paper-seller than a paper-maker. If so, then most paper in sixteenth-century Egypt was produced by Italian and French papermakers, shipped by Venetian merchants, purchased by mainly Muslim warrāqīn in Alexandria and Cairo, and sold to local communities—including Jewish communities—for use in writing. It is not impossible that some sixteenth-century Cairenes adopted European papermaking techniques, thus maintaining a local supply of paper, but if they did, then it appears they did not also differentiate that paper with new watermarks. If it exists, any paper produced in this way would be indistinguishable from European-made paper in our survey.

Conclusion

This examination of paper in late medieval Cairo shows that the Cambridge Genizah collections are an untapped resource for the study of papermaking around the Mediterranean basin between 1350 and 1600. They contain undocumented samples of European paper stocks used in Egypt during the transition from non-watermark to watermark paper in the Middle East, as well as watermark paper used for letter-writing in Jerusalem and Hebrew printing in the Ottoman Empire. We can thus suggest that the first watermark paper to reach the Cairene Jewish community most likely arrived in books produced in the Iberian peninsula during the late fourteenth or early fifteenth century. The earliest watermark paper that Cairene Jews used for writing in Egypt appears to have arrived in the middle of the fifteenth century, but it was not until the late sixteenth century that it replaced almost all other paper in our corpus. It is also likely that the widespread adoption of European paper stocks in Egyptian Jewish communities is connected to the forcible expulsion of Jews and Muslims from Iberia in 1492 and 1497. Refugees of these events settled in Egypt and the Ottoman Empire, bringing with them books produced in Europe on watermark paper. Especially notable are Iberian Hebrew printers who (re-) established their presses in Salonika and Constantinople.

These conclusions offer paths for future collaboration between Genizah scholars and European codicologists to investigate paper-trade networks and the diffusion of European papermaking techniques in the Ottoman Empire. Furthermore, as we have seen, Genizah collections can even provide new data for the distribution of European paper stocks in Spain and Portugal prior to the Iberian exile. Although we have not examined them here, there are also numerous Italian imprints in the Genizah—up through the middle of the eighteenth century—which could be used to further chart the trade and production of paper in Italy. From the evidence collected thus far, we hypothesize that further studies of the Cairo Genizah corpus will reveal what paper stocks were available to Jewish scribes and printers in fifteenth-century Iberia, sixteenth-century Ottoman cities, and members of practically every level of Egyptian-Jewish society between 1400 and 1600.

This research was supported by a Leverhulme Trust Early Career Fellowship and the Cambridge University Library’s Research & Collection Programme through the ‘Thinking Paper’ Project.

Footnotes

APPENDIX

CHRONOLOGY OF DATED PAPER MANUSCRIPTS IN CAMBRIDGE GENIZAH COLLECTIONS (1250–1685)

This table is not an exhaustive list of all dated paper manuscripts in Cambridge Genizah collections; see the introduction above for a discussion of the authors’ corpus. Every item below was found in Cairo.