The Staatsbibliothek in Hamburg holds a modest collection of Arabic papyri, acquired between 1910 and 1912 through the Deutsches Papyruskartell at the initiative of C.H. Becker, a papyrologist and Arabist who was then a professor at the city’s Colonial Institute.

Tillier and Vanthieghem 2024: 7–8.

Of the 135 pieces mounted under glass by the skilled restorer H. Ibscher—many more remain in the institution’s archives—roughly half were published by A. Dietrich in two volumes, released in 1937 and 1955, respectively.

Dietrich 1955.

Dietrich also authored an article featuring a fascinating Mamluk-era marriage contract written on silk.

Dietrich 1952.

With the exception of the recent publication of 14 bifolios of papyrus containing nearly the entire text of Surah II of the Qur’an

This text was the subject of an edition and study in Tillier and Vanthieghem 2024.

—commonly known as the Surah of the Cow—the collection has received limited attention from papyrologists.

E. Grob appears to be the only one who conducted on-site research to examine some of the letters discussed in her dissertation, for which she also reproduces images. See, in particular, Grob 2010: 165, 167, 175, 180 et 190.

During a visit to Hamburg in March 2010, I had the opportunity to examine the unpublished materials in this collection and uncovered several particularly noteworthy pieces, some of which are now in press.

P.Hamb.Ar. Inv. 42 (amān), P.Hamb.Ar. Inv. 77 (fiscal register) and P.Hamb.Ar. Inv. 83 (amān) will appear in N. Vanthieghem, “Letters of protection (amāns) from the Fatimid period,” and in Tillier and Vanthieghem 2025.

This article presents the edition of two debt acknowledgments from the Fatimid period preserved in this collection (P.Hamb.Ar. Inv. 80 and 81). Determining their provenance would have been impossible were it not for one of the witnesses, a certain ǦJamāʿa, whose family name is not mentioned. In document 2, he is identified as the preacher (ḫkhaṭīb) of Ṭuṭūn, a village in southern Fayyum located on the Baḥr Tanabṭawayh, where other pieces from Hamburg’s collection were also found.

On the provenance of the Hambourg papyrus collection, see Tillier and Vanthieghem 2024: 8–9. Beyond the documents cited in this book as originating from Ṭuṭūn, the documents P. Hamb. Inv. 42, 77, and 83 can now also be noted as coming from the same village.

Layout, Script, and Formulary

The two documents exhibit similar formal and formulary features. From a formal perspective, both are written on sheets of what appears to be laid paper, identifiable by faint parallel laid lines running vertically relative to the text. These sheets were cut from an old — most likely the same — accounting ledger. Traces of writing in the left and right margins of the debt acknowledgments suggest that the original ledger was likely written on both recto and verso, at least in part. This implies that the entries were initially recorded in an oblong kind of codex of the daftar type, a format well-documented in Fatimid accounting practices, but still insufficiently studied and in need of systematic scholarly analysis. The reused ledger appears to have recorded receipts and expenses of grains, legumes, and other agricultural produce from a village and its surrounding area (nāḥiya), though the village name is not specified—given the mention of the village of Ṭuṭūn in document 2, it is possible that the surrounding area refers to this location. The preserved portion of this ledger lists in sequence beans (fūl), wheat (qamḥ), and barley (shaʿīr), all quantified in sacks (tillīs). These items were either distributed as seed advances (taqwiya), a matter that will be explored later in this article, or sold (mabīʿ).

An edition of the remaining portions of this ledger will be published separately in a forthcoming issue of this journal.

The main body of both documents was written in black ink by the same hand. This handwriting, marked by frequent abusive ligatures and graphic simplifications, reflects the work of a skilled and experienced scribe. In contrast, the two witness statements were added by a different hand, that of Nahār b. ǦJamāʿa, the preacher ( ḫkhaṭīb) of Ṭuṭūn. This second hand, equally skilled as the main scribe, used ink of a different composition, which has since faded to a reddish hue. Notably, while the primary scribe employed diacritical marks sparingly, the second hand omitted them altogether.

The formularies of the two documents are nearly identical. Following the religious invocation (basmala), the text begins with the verb aqarra, commonly used at the start of debt acknowledgments from this period. This is followed by the debtor’s name and a formula stating that the debtor is of sound mind and body (fī siḥḥat ʿaqli-hi wa-badni-hi), legally competent (jawāz amrihi), and acting freely and without coercion (ghayr mukrah wa-lā mujbar). The text then specifies the name of the creditor using the formula an ʿalayhi wa-ʿinda-ha wa-qibali-hi wa-fī khāliṣ mālihi wa-dhimmatihi li-fulān (“is indebted and liable, from his own resources and under his responsibility, to”). The nature of the obligation is detailed with the preposition min, followed by the amount in irdabbs. It is declared binding (daynan thābitan wa-ḥaqqan wājiban lāziman), with an explicit acknowledgment made in the presence of witnesses (iʿtarafa lahu bi-dhālika ʿinda shuhūd hādhā l-wathīqa). Further, the individual commits not to evade repayment in any way (ʿan al-khurūj min al-dayn). The acknowledgment is made explicitly “under the guarantee” of a third party (bi-ḍamān fulān), a detail that needs further discussion. The document concludes by noting that witnesses were called to testify (ashhada ʿalā nafsihi), with the date also specified. Finally, the two witness statements follow, each beginning with the customary shahida (“He testified”) followed by the witness’ name.

The Nature of the Debt

The debt acknowledgments concern, in one case, two-thirds of an irdabb of barsīm seeds (1), and in the other, one and one-third irdabbs of julbān seeds (2). Barsīm (Trifolium alexandrinum), commonly referred to in papyri as qurṭ (a term borrowed from the Greek χόρτος), is an herbaceous plant often mistranslated as "clover," though it bears no resemblance to the European variety. Instead, it is a type of alfalfa, which historically—and still today—served as the primary forage for working animals like donkeys, camels, horses, and mules, as well as for larger livestock such as cows and buffaloes.

For more details on barsīm, see Fairchild 1902.

According to Ibn Mammāti (d. 606/1209), alfalfa could be sown at various times of the year, between Bāba (September) and Hatūr (January), depending on the region and the variety.

Ibn Mammāṭī, Kitāb qawānīn al-dawāwīn: 262; Cooper 1974: 117.

However, sowing could not begin until after the Nile floodwaters receded, a process that was not complete in the Fayyum region until late November or even December.

For details on the receding of waters in the Fayyum following the inundation, see Tillier and Vanthieghem 2020: 357–359.

Once the water had receded, farmers in the Fayyum plowed the land before sowing, unlike in other parts of the Nile Valley where seeds were sown directly onto the silt deposited by the flood.

Girard 1882: 42.

Approximately 2½ waybas of seeds were needed to sow a feddān of land.

Ibn Mammāṭī, Kitāb qawānīn al-dawāwīn: 262; Cooper 1974: 117.

Depending on the region and climatic conditions, alfalfa could be harvested two to three times a year, or even four times in particularly humid areas. Under ideal conditions, harvesting occurred roughly every thirty days. The first or first two cuts provided green forage for livestock, while the final cut was left to dry in the field and then transported to threshing areas for seed production.

Girard 1882: 43–44.

During the four-month growing season, depending on weather conditions, a single feddān of barsīm could sustain between one and four oxen for nearly a month,

Girard 1882: 43–44. The author notes that in the Minūfiyya province, two oxen could graze on ¹⁄ ₂₄ of a feddān per day, while in Tanta, the same pair required ½ a feddān for a month.

or produce 44,005 kilograms of green forage over three to four cuts.

Fairchild 1902: 14 states that the average yield of forage in 1902 was 28 tons per acre (4047 m2), which means ± 44,05 tons per feddān (6368 m2)

The final cut typically yielded between two and three irdabbs of barsīm seeds.

Ibn Mammāṭī, Kitāb qawānīn al-dawāwīn: 263; Cooper 1974:117 and Girard 1882: 43.

Julbān, or grass pea (Lathyrus sativus L.), is a legume whose peas were used for human consumption, while the stalks served as animal fodder.

Girard 1882: 45.

It also had medicinal properties.

Ibn El-Beïthar, Traité des Simples Ibn al-Beithar: 358–359.

Ibn Mammāti notes that it was sown between Hatūr (November) and Kayhak (December), alongside other legumes such as lentils (ʿads) and chickpeas (ḥimmis).

Ibn Mammāṭī, Kitāb qawānīn al-dawāwīn; 260–261; Cooper 1974: 116.

Greek documents from the Fayyum suggest that sowing could extend into January and even February.

Rathbone 1991: 260–262.

To sow a feddān, between four waybas and one irdabb of seeds were required,

Ibn Mammāṭī, Kitāb qawānīn al-dawāwīn: 260–261; Cooper 1974: 116.

though Girard specifies that two-thirds of an irdabb could suffice.

Girard 1882: 45.

After sixty days, julbān could be harvested as forage. Its peas matured after approximately one hundred days

Girard 1882: 45.

and were harvested in Barmūda (April), according to Ibn Mammāti.

Ibn Mammāṭī, Kitāb qawānīn al-dawāwīn: 261; ; Cooper 1974: 116.

A feddān of grass pea could yield about ten irdabbs of peas and ten to fifteen camel-loads of green forage,

Girard 1882: 45.

equivalent to 1800–2700 kilograms.

The average camel load, estimated at 180 kg, is used here as a reference. See Adams 2007: 49–52.

Beyond their vital role as fodder crops and in human nutrition, these two species were integral to agricultural cycles. Sown on fallow or nutrient-depleted lands, particularly those affected by salinization, they played a key role in soil regeneration. Land cultivated with these crops, known as bāq, was highly valued for its proven ability to restore fertility for growing wheat and barley.

Cooper 1974b: 94 and 97–101.

The Context and Mechanism of the Debt

The two acknowledgements were issued in the name of Ḥarīr b. Muḥammad, who admits to having incurred a debt involving quantities of seed to a certain Muḥammad b. Ayyūb. In both cases, the transaction is carried out explicitly “with the guarantee” (bi-ḍamān) of an individual identified as ʿUmar b. Aḥmad. On the surface, the arrangement appears straightforward: Ḥarīr b. Muḥammad owes a debt in seed to his creditor, and ʿUmar b. Aḥmad guarantees repayment from his own assets should Ḥarīr default.

Such acknowledgements of debt for small amounts of grain or seed are common in Arabic documentary material.

For examples, see in particular Chrest.Khoury 34–36; CPR XXVI 19; P.Cair.Arab. II 106–112; P.RagibQalamun 1.

Most editors interpret them as grain debts or debts related to agricultural produce, though they rarely clarify the precise nature or purpose of the debt. Werner Diem has suggested viewing these as micro-loans mainly of cereals between acquaintances, similar to those mentioned in private letters—where it is not unusual to find someone asking a correspondent to send a small amount of wheat.

Diem 2006: 47–49.

More recently, L. Bondioli has offered a compelling hypothesis that these documents may represent forward contracts, whereby a merchant would purchase a future agricultural yield still in the field, in exchange for which the cultivator committed to delivering the crop after the harvest.

Bondioli 2021: 106–135.

Such arrangements are well attested in papyrological sources, particularly for commodities such as flax, grapes, and wood, though they typically involve large quantities.

On advance sales in general, see the overview by Rāġib 1982: 7–12 along with Diem 2006.

It therefore seems unlikely that merchants would have devoted their time to purchasing trivial amounts in advance—especially for something like clover (barsīm) or grass pea (julbān) seeds, which held little market value.

In a forthcoming book, Mathieu Tillier and I argue that these so-called microcredits or forward sales are, in reality, debt acknowledgements issued by the fiscal administration in connection with seed advances distributed by state agents or their representatives to cultivators (muzāriʿūn) to enable them to sow their fields.

Tillier and Vanthieghem 2025.

These advances, known as taqwiya (pl. taqāwī), were typically allocated shortly before the sowing season and, in principle, were to be repaid within the year. In practice, however, repayments were often significantly delayed—if made at all—leaving the burden of the debt to fall on the heirs. During the period under consideration, the distribution of seeds in villages was typically handled by the khāzin (“cashier”), who was in charge of the public granaries. Alternatively, it could be undertaken through the intermediary of a ḍāmin—usually a wealthy individual from the village or nearby, who was granted a tax farm (ḍamān) allowing him to collect taxes on behalf of the state from a particular village over a set period.

In the case of the two debt acknowledgements presented here, Ḥarīr b. Muḥammad appears to have been a cultivator (muzāriʿ). As the recipient of such seed advances, he would have been able to sow approximately 1.59 feddāns (c. 10,093 m²) of clover (barsīm), and 2 feddāns (12,696 m²) of grass pea (julbān).

The calculation for barsīm is as follows: 2½ waybas (approximately 0.42 irdabb) of seed are required to sow one feddan. The ⅔ irdabb received by Ḥarīr b. Muḥammad would thus have allowed him to cultivate approximately 1.59 feddans. As for grass pea, since ⅔ irdabb of seed is sufficient to sow one feddan (see above), the 1⅓ irdabb (i.e., 4/3 irdabb) mentioned in document 2 would have been enough to sow 2 feddans.

However, if we accept the model of state-distributed seed advances outlined above, what then were the respective roles of Muḥammad b. Ayyūb and ʿUmar b. Aḥmad in this transaction? The phrase bi-ḍamān associated with the latter strongly suggests that ʿUmar b. Aḥmad was the ḍāmin—the tax farmer—for the village of Ṭuṭūn that year. As for Muḥammad b. Ayyūb, he was most likely the local representative of the ḍāmin within the village.

Editions

1. Acknowledgement of Debt for Clover (barsīm) Seed Advances.

An oblong fragment of beige paper, preserved in its full height and width. The top and bottom margins are relatively narrow; the lateral margins are even tighter, with the text written almost up to the right and left edges. The document comprises fifteen lines, penned—as in the following document—by two different hands. The first hand wrote the main body of the text (lines 1–11) in black ink, while the second hand added two testimonial subscriptions (lines 12–15) in ink that has since browned with age. Both hands use a cursive script characterized by numerous ligatures. Some letters are marked with diacritical dots.