Introduction
In 1949, Nabia Abbott published the editio princeps of the earliest Thousand Nights text alongside an exhaustive decipherment of the damaged paper bifolium acquired two years prior by the Oriental Institute from an Egyptian antiquities dealer, cataloged today as OIM 17618 at the University of Chicago’s Institute for the Study of Ancient Cultures (ISAC). The title inscription reads Kitāb fīhi ḥadīth alf Laylah (كتاب فيه حديث ألف ليلة, “A Book Containing Stories from a Thousand Nights”), this ninth-century leaf remains our earliest physical witness to the Thousand and One Nights (Alf Layla wa-Layla / “ألف ليلة وليلة”) in Arabic.
Before Abbott’s decipherment, scholars assumed the written transmission of the Nights began half a millennium later. Textbook consensus placed the text’s written emergence no earlier than the fourteenth or fifteenth century, anchored by Antoine Galland’s early eighteenth-century foundational French translation (Mahdi 1984, 1–10; Marzolph and van Leeuwen 2004, 1–5; Grotzfeld 2006, 112–115). OIM 17618 radically altered that timeline. It proved that bound, commercial paper copies of the tales were already circulating in the Abbasid Caliphate during the ninth century (Abbott 1949, 129–131; Bloom 2001, 58–60).
Yet OIM 17618 is not a neat literary codex. As cataloged in the Advanced Papyrological Information System (APIS Record 8241), it survives as a conjoined bifolium of light-brown linen paper measuring 24.2 × 13 cm. It consists of an originally blank flyleaf (pages 1–2) and the opening narrative folio (pages 3–4) of an otherwise lost volume (Abbott 1949, 129–130; APIS 2021; ISAC 2021). Crucially, the literary narrative contains neither a colophon nor a date (Abbott 1949, 130). Its chronology must instead be reconstructed from early Islamic paper economics. Because high-grade linen rag paper remained a scarce, costly commodity in the ninth century, scribes and merchants routinely repurposed damaged volumes as scrap (Bloom 2001, 58-60).
In her foundational study, Abbott found that OIM 17618 accumulated six distinct historical layers, formed through decades of recycling (Abbott 1949, 130-133; APIS 2021):
- The Primary Narrative (early-to-mid 9th century CE): The title page on page 3, “Kitāb fīhi ḥadīth alf laylah,” and the opening frame narrative on page 4, written in an informal Abbasid cursive book hand (Abbott 1949, 130, 133).
- Devotional Phrases and Pen Trials: Pious invocations, including basmala formulas and praises to God, scribbled across the flyleaf margins (Abbott 1949, 130-131).
- Private Commercial Draft: A draft of a business letter regarding financial remittances and travel to Antioch, inscribed on page 1 (Abbott 1949, 131).
- The Dated Promissory Note (Ṣafar 266 AH / October 879 CE): An administrative debt acknowledgment naming Aḥmad ibn Maḥfūẓ (Abbott 1949, 131–132; APIS 2021). This administrative record provides a decisive terminus ante quem for the artifact: because a clerk drafted this legal receipt onto the discarded flyleaf in the autumn of 879 CE, the underlying literary text was copied, read, and retired before that date (Abbott 1949, 131–132; Bloom 2001, 59).
- Legal Attestations and Signatures: Informal witness notations clustered around the Maḥfūẓ receipt (Abbott 1949, 132).
- Marginal Verse and Aphorisms: Casual poetic jottings added along the outer edges before the bifolium was buried in Egyptian soil (Abbott 1949, 132-133).
Figure 1: AI-assisted interface mockup demonstrating segmentation of ninth-century bifolium OIM 17618, showing metric overlay rulers, codicological strata layers, and metadata panel. Courtesy of ISAC. (AI-assisted digital workflow mockup; see disclosure).
Far from an erased palimpsest, OIM 17618 is an accretive document generated by paper recycling (Abbott 1949, 130; Kraus 2024, 126–128). The manuscript descended so rapidly from a bound storytelling book into administrative scratch paper proves that the ninth-century Nights circulated in commercial, vernacular reading circles far removed from elite scribal patronage (Bloom 2001, 58–60; Hirschler 2022, 22–24).
Historical Discussion: Arabic and Persian Traditions
The narrative core of The Thousand and One Nights represents a rich confluence of Indo-Persian folk traditions and early Islamic urban storytelling (Mahdi 1984, 11–15; Chraïbi 2004, 15–20). Its overarching frame story derives from a Sasanian Persian work, Hazār Afsān (هزار افسان, “Thousand Tales”), translated into Arabic during the classical translation movement in eighth-century Baghdad (Abbott 1949, 133; Marzolph and van Leeuwen 2004, 2).
Tenth-century historical records provide direct evidence of this transmission. The historian al-Mas’udi (d. 956 CE) notes in his Murūj al-Dhahab (“Meadows of Gold”) that the Hazār Afsān was translated from Persian into Arabic under the popular title Alf Layla (ألف ليلة, “A Thousand Nights”) (al-Mas’udi 1966; Chraïbi 2004, 22; Mahdi 1984, 13). Decades later, the Baghdadi bibliographer Ibn al-Nadīm (d. c. 995/998 CE) provided valuable insight into the Hazār Afsān in his Kitāb al-Fihrist, as a collection of entertaining evening tales. Central to these stories is a king who executes his brides until the vizier’s daughter, Shahrazād (شهرزاد), cleverly preserves her life by leaving her tales unfinished by dawn (Ibn al-Nadīm 1970, 713–715; Marzolph and van Leeuwen 2004, 2). Although these tenth-century Arabic accounts confirm the cultural currency of this frame story in Baghdad, some elite commentators categorized such narratives as informal evening diversion (samar / سمر), lacking high literary prestige. In this context, the material value of OIM 17618 provides tangible evidence that while court intellectuals looked down on the tales, ordinary readers bought, read, and discarded them in everyday book markets (Green 2013, 4).
The title on OIM 17618 preserves the archaic heading “Kitāb fīhi ḥadīth alf laylah,” translating to “A book retaining stories from a Thousand Nights.” The expanded canonical title Alf Layla wa-Layla, or “A Thousand and One Nights,” does not appear in the documentary record until three centuries later, when Goitein identified it in a Cairo Genizah book list dated 1150 CE. (Goitein 1958, 301-302). The Genizah document bridges the early truncated title with the expanded recensions of the late Middle Ages.
OIM 17618 also captures a transitional phase in Arabic onomastics. On page 4 (verso), the frame tale names the primary female characters as Dīnāzād (دينازاد) and Shīrāzād (شيرازاد) (Abbott 1949, 133; APIS 2021). These spellings retain archaic Middle Persian phonetic forms (Abbott 1949, 133–134; Mahdi 1984, 14–16). Only in later Syrian and Egyptian manuscripts were these names standardized into the familiar Arabic forms Dīnārzād (دينارزاد) and Shahrzād (شهرزاد). In the ninth century, the assimilation of Persian names into Arabic prose remained phonologically fluid (Reynolds 2006).
Figure 2: Detail of OIM 17618 (Page 4, verso) in early Abbasid cursive script, preserving the archaic name Dīnāzād (قالت دينازاد, line 2). Courtesy of ISAC / APIS.
Comparing OIM 17618 with the Galland Manuscript (Bibliothèque nationale de France MS Ar. 3609-3611) illustrates this ongoing crystallization of textual and material forms over time. OIM 17618 features an early Abbasid cursive book hand, known as proto-naskhī, typical of early secular paper. Marked by rounded ligatures, variable letter proportions, and sporadic diacritical pointing, reflecting the informal marketplace circulation of vernacular entertainment tales rather than the formal elegance associated with institutional or calligraphic contexts. By contrast, Heinz Grotzfeld’s paleographical and watermarking studies firmly anchor the Syrian Galland codex to a standardized book hand of the fourteenth century (c. 1300–1350 CE). This five-century gap between the two manuscripts shows the transition from informal storytelling notebooks to formalized codicological recensions (Abbott 1949, 135; Grotzfeld 2006, 112-115; Mahdi 1984, 25).
Figure 3: Comparative layout of ninth-century bifolium OIM 17618 (left) alongside the fourteenth-century Syrian Galland manuscript (BnF Arabe 3609, right), showing computational alignment vectors and paper substrate details. (AI-assisted comparative composite; see disclosure).
Digital Humanities Infrastructure and Future Research Horizons
The Institute for the Study of Ancient Cultures (ISAC) openly hosted high-resolution digital surrogates of the Chicago bifolium, indexed in the Advanced Papyrological Information System under Record 8241 (ISAC 2021; APIS 2021). These digital assets are available through the International Image Interoperability Framework (IIIF) standard, enabling non-invasive examination of ink degradation, substrate fibers, and cursive ductus directly in web browsers. This approach lets researchers study these fragile features without risking damage to the original documents (Stutzmann and Tarte 2017). Loading the IIIF manifest into open-source workspaces like Mirador further facilitates side-by-side paleographical comparison with later recensions, such as the fourteenth-century Syrian Galland codex (BnF Ar. 3609–3611).
Beyond basic visual inspection, however, OIM 17618 exposes the methodological boundaries of automated text recognition. Off-the-shelf Handwritten Text Recognition (HTR) engines like Kraken and platforms like eScriptorium cannot accurately transcribe this particular leaf. Supervised neural models require substantial training data, whereas OIM 17618 contains only four damaged pages distributed across six distinct hands. Moreover, early cursive writing (رسم) often lacks systematic vocalization and diacritical dots and vowels, leaving many letterforms ambiguous (Fedeli 2020, 152). Automated baseline detectors also fail when encountering the upside-down, perpendicular, and diagonal marginalia that intersect the text blocks (Kiessling et al. 2019).
Figure 4: Conceptual UI mockup showing manual polygon segmentation isolating primary narrative text (teal) from secondary administrative marginalia (amber).
The prominent computational value lies in spatial stratigraphy and comparative formulaic analysis. Within eScriptorium, researchers can set aside automated text detection and draw manual polygon region masks, isolating the literary narrative from secondary administrative scribbles. To decipher the administrative strata, specifically the damaged promissory note of Ṣafar 266 AH (October 879 CE), scholars can turn to the Arabic Papyrology Database (APD) (Abbott 1949, 131–132; APD 2024). By querying the APD’s corpus of ninth-century Egyptian documents for parallel legal debt acknowledgments (iqrār bi-dayn) and witness formulas, researchers can reconstruct damaged or unpointed readings with philological precision (Abbott 1949, 131–132). Digital tools make their most valuable contribution not when attempting to replace human transcription with machine learning, but when providing structured environments for spatial mapping and formulaic corroboration (Mann 2020, 250; Rudy 2010, 31).
Conclusion
The discovery of the Oriental Institute Museum fragment, OIM 17618, significantly redefines our interpretation of The Arabian Nights. It proves that this collection did not originate as a fixed monumental canon, but circulated as an open, vernacular storytelling tradition. Moreover, it remains an indispensable benchmark for early Islamic codicology, paper history, and Arabic literary studies. Broadening comparative research to include a wide range of early Abbasid papyrological and documentary sources will further clarify how everyday scribal habits shaped the tales before they crystallized into medieval Syrian and Egyptian recensions.
Distinguishing between existing institutional digital infrastructure and proposed computational workflows clarifies the key role of digital humanities in manuscript studies. The University of Chicago’s IIIF repository ensures democratic access to digital surrogates, pairing manual polygon segmentation in eScriptorium with formulaic searches in the Arabic Papyrology Database establishes a workable method for untangling the leaf’s complex scribal history (Kiessling et al. 2019; APD 2024). This technological integration demonstrates that digital tools achieve their greatest utility when they support, rather than supplant, disciplined material and philological autopsy (Stutzmann and Tarte 2017; Mahony 2019, 193). Far from serving merely as an isolated literary relic, the OIM 17618 bifolium emerges through digital fragmentology as an accretive material archive: a ninth-century paper leaf that records the life cycle of medieval literature, passing from secular entertainment codex to recycled administrative scrap (Abbott 1949, 130–133; Bloom 2001, 58–60).
References
Nabia Abbott, “A Ninth-Century Fragment of the ‘Thousand Nights’ New Light on the Early History of the Arabian Nights,” Journal of Near Eastern Studies 8, No. 3 (1949): 129–64. https://doi.org/10.1086/370926.
‘Ali ibn al-Husayn al-Mas’udi, Murūj al-Dhahab wa-Ma’ādin al-Jawhar, edited by Charles Pellat (Beirut: Université Libanaise, 1966).
APD (Arabic Papyrology Database), directed by Andreas Kaplony (Munich: Ludwig-Maximilians-Universität München, 2024). https://www.apd.gwi.uni-muenchen.de/.
APIS (Advanced Papyrological Information System), “OIM 17618: A Volume That Contained Selected Tales from The Book of the Tales of a Thousand Nights” (Chicago: The Oriental Institute / APIS Record 8241, 2021). https://papyri.info/apis/chicago.apis.8241.
Jonathan M. Bloom, Paper Before Print: The History and Impact of Paper in the Islamic World (New Haven: Yale University Press, 2001). https://doi.org/10.37862/aaeportal.00217.
Aboubakr Chraïbi, ed., Les Mille et Une Nuits en partage (Paris: Éditions Philippe Picquier, 2004).
Alba Fedeli, “Digital Humanities and Qurʾanic Manuscript Studies: New Perspectives and Challenges for Collaborative Space and Plural Views,” Journal of College of Sharia & Islamic Studies 38, No. 1 (2020): 147–58. https://doi.org/10.29117/jcsis.2020.0260.
Shelomo Dov Goitein, “The Oldest Documentary Evidence for the Title Alf Laila Wa-Laila.” Journal of the American Oriental Society 78, No. 4 (1958): 301–302. https://doi.org/10.2307/595794.
Jack Green, “Behind the Scene at the Oriental Institute Museum,” The Oriental Institute of the University of Chicago (2013).
Heinz Grotzfeld, “The Age of the Galland Manuscript of the Nights,” in The Arabian Nights Reader, edited by Ulrich Marzolph (Detroit: Wayne State University Press, 2006), 105–21.
Konrad Hirschler, “Saleroom Fiction versus Provenance,” Journal of Islamic Manuscripts 13, No. 1 (2022): 1–54. https://doi.org/10.1163/1878464x-01301001.
Muhammad ibn Ishaq Ibn al-Nadīm, Kitāb al-Fihrist, translated by Bayard Dodge (New York: Columbia University Press, 1970).
ISAC (Institute for the Study of Ancient Cultures), “Manuscript Fragment E17618 (OIM 17618),” Online Cultural Heritage Database (Chicago: University of Chicago, 2021). https://isac-idb.uchicago.edu/id/750c9ec5-4d25-42d8-91f8-dd3c5b24b9c5.
Benjamin Kiessling, Robin Tissot, Peter Stokes, and Daniel Stökl Ben Ezra, “eScriptorium: An Open Source Platform for Historical Document Analysis,” Proceedings of the International Conference on Document Analysis and Recognition Workshops 2 (2019): 19–24. https://doi.org/10.1109/ICDARW.2019.10032.
Sven Kraus, “Fragmentology and Graph Technology: A Case Study of Frankfurt Fragm. Lat. II 11,” Digital Philology 13, No. 1 (2024): 124–32. https://doi.org/10.1353/dph.2024.a926889.
Muhsin Mahdi, ed., The Thousand and One Nights (Alf Layla wa-Layla): From the Earliest Known Sources. 2 vols (Leiden: Brill, 1984). https://doi.org/10.1163/9789004539433.
Simon Mahony, “Analysis of Ancient and Medieval Texts and Manuscripts: Digital Approaches,” Variants 14 (2019): 191–93. https://doi.org/10.4000/variants.778.
Joshua L. Mann, “Paratextuality and Digitally Mediated Sacred Texts,” in From Scrolls to Scrolling: Sacred Texts, Materiality, and Dynamic Media Cultures, edited by Bradford A. Anderson (Berlin: De Gruyter, 2020), 247–66. https://doi.org/10.1515/9783110634440-011.
Ulrich Marzolph and Richard van Leeuwen, The Arabian Nights Encyclopedia (Santa Barbara, CA: ABC-CLIO, 2004). https://doi.org/10.5040/9798400613944.
Dwight F. Reynolds, “A Thousand and One Nights: A History of the Text and Its Reception,” in The Cambridge History of Arabic Literature: Arabic Literature in the Post-Classical Period, edited by Roger Allen and D. S. Richards (Cambridge: Cambridge University Press, 2006), 270–91. https://doi.org/10.1017/chol9780521771603.014.
Kathryn M. Rudy, “Dirty Books: Quantifying Patterns of Use in Medieval Manuscripts Using a Densitometer,” Journal of Historians of Netherlandish Art 2, No. 1–2 (2010). https://doi.org/10.5092/jhna.2010.2.1.1.
Dominique Stutzmann and S. Tarte, “Paléographie numérique / Digital Palaeography | Fondation des Treilles,” in HAL (Le Centre pour la Communication Scientifique Directe), Centre National de la Recherche Scientifique, January 17 (2017). https://hal.science/hal-01855261.
AI disclosure:
Figure 1. ChatGPT (GPT-4o / DALL-E 3, OpenAI) was used to generate the interface mockup. The author defined the UI parameters, validated the polygon segmentation masks against the codicological strata, and edited the final output. (https://chatgpt.com/s/m_6ad24cd31a3c81919de1fb093b4d4a64).
Figure 3. ChatGPT (GPT-4o / DALL-E 3, OpenAI) was used to generate the comparative visual layout. The author supplied authentic manuscript facsimiles as reference inputs, cross-examined the output for paleographical orientation, and edited the final composite. (https://chatgpt.com/s/m_6ad24f88c7248191b4b9db6a3cf37ed0).
