One Thousand and One Conjectures

Greek arrived in Nubia as a living church language and ended, this conjecture claims, as sacred wallpaper — and the death is measurable as a fossilization ratio: the share of Greek texts that are fixed formulae (the Trisagion, stock epitaph prayers, liturgical…

Institutional Christian Nubia contracted through the fourteenth and fifteenth centuries, yet people in late Dotawo kept writing Old Nubian letters and legal instruments. This conjecture claims documentary writing outlived new liturgical production by roughly a century, because its base was the household…

Nubian scribes had papyrus and later paper for everyday use, yet many land sales and similar instruments at Qasr Ibrim were written on leather. This conjecture says substrate was a deliberate legal technology: permanent instruments — conveyances, manumissions — went onto leather…

Christian Nubia and Christian Ethiopia were neighbours for eight hundred years, both taking their bishops from the Coptic patriarch of Alexandria — yet each received consecrations, texts, and translations via Cairo. This conjecture claims that shared dependence produced a strict hub-and-spoke information…

Late-antique Aksum and early Christian Nubia both wrote monumental Greek at the edge of the Greek world — royal texts, dedications, epitaphs. This conjecture says their Greek is not two independent provincial reflexes of the metropolitan standard but ONE shared regional register:…

Every sheet in a medieval Timbuktu book crossed the Sahara on a camel. Italian watermarks give each sheet a birth date; colophons give the book's completion date; the difference between them is the paper's age-at-use — a direct gauge of trans-Saharan supply-chain…

Sahel manuscripts live on paper in termite country, so books survived by being recopied, not by lasting. This conjecture claims the consequence is a sharp physical horizon: almost no codex in the Timbuktu collections physically predates the mid-sixteenth century even where composition…

Timbuktu's manuscript world has been sampled twice by partly independent digitization efforts — organized library digitization on one side, family- and region-based field projects on the other. Where two independent samples of one underlying population exist, ecology's mark-recapture logic applies to books…

West African languages first took Arabic script — the practice called Ajami — not, on this conjecture, as free-standing literature but as classroom apparatus: interlinear glosses translating the hard words of Arabic law and grammar texts for students in the teaching circle.…

Timbuktu and Djenné sit on the same river system a few hundred kilometres apart, yet this conjecture claims their manuscript cultures are structurally different regimes: Timbuktu's collections formed around scholarly lineages and commercial book production, while Djenné's — as revealed when digitization…

The cities that produced the Sahel's books — riverine, humid, wealthy, repeatedly fought over — are the worst places for paper to survive; remote desert-edge villages, poor in scholars but dry and quiet, are the best. This conjecture claims manuscript survival in…

Sahel manuscripts carry study licences (ijāzas) and audition certificates naming teacher-student chains — a self-recorded academic network embedded in the books themselves. The standard picture routes Sahel learning through metropolitan gateways: Fez, Cairo, the pilgrimage road. This conjecture claims the pre-1600 chains…

The distinctive barnāwī script of Bornu Qurʾans preserves letterforms and orthographic habits that the western Maghreb itself abandoned centuries earlier — a periphery keeping the metropole's discarded past, the way island dialects keep old pronunciations. This conjecture makes the conservatism quantitative and…

What the world calls 'the Timbuktu manuscripts' has been filtered twice: families show curators their prestigious books, and cataloguers privilege named scholarly works over unglamorous talismans, amulet grids, and divination sheets. This conjecture claims the everyday-esoteric stratum was the VOLUME MAJORITY of…

The classic Sahel book is unbound — loose bifolia stacked in a leather wrapper — a format usually explained by binding-material scarcity. This conjecture says it is instead an instructional technology: unbound quires could be dealt out around a teaching circle so…

Ethiopian manuscripts teem with colophons; Sahel manuscripts are famously reticent about their makers. This conjecture says colophon density is an economic signature, not a cultural temperament: books held as institutional endowments (church property, waqf) need internal provenance records — who gave what,…

The surviving record seems to say Ethiopia wrote early and the Sahel wrote late: Ethiopian parchment books survive from the first millennium, Sahel books barely from before 1600. This conjecture claims much of that famous asymmetry is a substrate artifact: parchment in…

Sahel life is metered by the rains: farming in the wet months, caravans, travel, and teaching in the dry. Copying a book needs stable paper, workable ink, daylight leisure, and often a borrowed exemplar newly arrived by caravan — all dry-season goods.…

The Swahili coast demonstrably wrote before 1500 — carved Arabic epitaphs and mosque inscriptions, coin legends, and early Portuguese descriptions of correspondence all attest it — yet its surviving manuscripts begin only in the later eighteenth century. This conjecture claims the gap…

Writing a Bantu language in Arabic script is lossy: the letters underdetermine Swahili's vowels and several consonants, so a prose sentence admits many readings. Verse repairs the channel: the fixed syllable counts and rhymes of the utendi forms constrain decoding, so meter…

Scribes copy habits as well as words, and orthographic conventions outlive the practice that created them. This conjecture claims the eighteenth- and nineteenth-century Swahili manuscripts preserve fossil spellings — systematic Arabic-script conventions for Swahili sounds that make no sense within the copyists'…

The pre-1500 Swahili coast survives epigraphically — tombstones and mosque inscriptions naming patrician families at Kilwa, Mogadishu, and along the whole seaboard — while its books survive only from centuries later. If coastal literacy was continuous and family-borne, the two records should…

Africa's two great paper-borne manuscript cultures faced opposite seas: the Sahel imported Mediterranean paper southward across the desert, while the Swahili coast sat on the monsoon circuit to the Gulf and Gujarat. This conjecture claims their books are literally made of different…

The surviving Swahili manuscript record concentrates spectacularly in a few northern localities — above all the Lamu archipelago — although pre-1500 epigraphy shows literate settlement along three thousand kilometres of coast. This conjecture claims the corpus geography is an artifact of family-archive…

When the Tibetan empire took Dunhuang in 786, it inherited a Chinese town full of Chinese scrolls, and its clerks and monks needed to write Tibetan immediately. This conjecture joins the Tibetan occupation to the humble habit of reusing old paper, and…

Tibetan forts at Mazar-tagh and Miran left hundreds of wooden slips recording grain, soldiers, and mail, while Dunhuang, a paper town, left almost none on wood. Join substrate choice to paper logistics: the wood-to-paper ratio of administrative writing is a map of…

The Sogdian Ancient Letters of c. 313 CE — merchant mail abandoned in a watchtower near Dunhuang — are usually read as a lucky window. Join them instead to a statistician's commonplace: lost mail is a sample of a mail stream, and…

The bilingual Sino-Tibetan treaty pillar of 821/822 in Lhasa is famous as a monument; read it instead as the only surviving printout of a lost diplomatic paper trail. Treaties are drafted, exchanged, sworn, and archived on perishable media before anyone cuts stone,…

The earliest stele of the first Turkic khaganate (Bugut, later sixth century) speaks Sogdian; the classic Orkhon monuments of the second khaganate speak Turkic in runes; the Uyghur empire's Karabalgasun inscription brings Sogdian back among three languages. Join these familiar stones into…

Dunhuang Cave 17 was a deliberate deposit inside a shrine; the Turfan finds come largely from ruins, dumps, and graves. Join the two most famous manuscript hoards of Inner Asia into a single instrument: the two sites sampled broadly similar oasis document…

At Dunhuang the same Buddhism circulates as Chinese scrolls and as Tibetan-style loose-leaf pothi, and the split is usually filed under ethnic habit. This conjecture reassigns it to use-mode: pothi serves communal recitation and random access, since any leaf can face any…

Under Tibetan rule, Dunhuang ran one of the earliest well-documented mass book-production projects: thousands of copies of the same sutras, written largely by conscripted local scribes whose names, and their editors' names, survive in colophons. Join industrial quality control to those colophons:…

At Shatial on the Upper Indus, hundreds of short Sogdian rock inscriptions — mostly a bare name and patronymic — mark a river crossing on the road toward India. Read the cliff not as graffiti but as a passenger manifest accumulated over…

When an oasis switches language — Sogdian giving way to Uyghur at Turfan, Khotanese fading around Khotan and Dunhuang — the switch is usually dated from religious literature. This conjecture says religion is the wrong clock: legal genres conserve a receding language…

Dunhuang preserved bilingual word-lists and phrasebooks — Tibetan-Chinese, Khotanese pairs, and others — usually treated as monastic study aids. This conjecture reassigns most of them to the checkpoint and the counting-house: working glossaries are tools of whoever must process strangers, so their…

The Turkic runic script of the Orkhon stelae is the steppe's most famous writing, and it is natural to assume the khaganates administered themselves in it. This conjecture says the runes were a monumental and mantic craft — a script for the…

Chinese bureaucratic culture validated documents with vermilion office seals, and Central Asian regimes kept sealing in the Chinese manner even when the documents were no longer in Chinese. The conjecture: authority technologies diffuse ahead of, and independently of, language — a Tibetan,…

Old Tibetan contracts were validated with finger-joint measures and personal marks rather than signatures. Join this to the Chinese contract tradition at Dunhuang, which had its own native validation habits: if legal validation follows the court that will enforce the deal rather…

Uyghur script at Turfan splits over time into a formal book hand and an increasingly rapid cursive. Join palaeography to institutional history: cursive is not decay but a bureaucratization index, because only a society generating routine paperwork — receipts, orders, tax notes…

The Tangut state of Xixia embraced printing early and massively, and the dead city of Khara-Khoto preserved the result. Join printing economics to the Buddhist merit economy: printing pays where demand is for many identical copies, and in a merit-driven society that…

Mongol imperial documents follow a rigid formulaic skeleton — heaven-invocation, the ruler's word, addressee, injunction — that appeared across Eurasia within a generation of the conquests. Join this to the Uyghur clerks the Mongols famously hired: the skeleton should be visible, slot…

The Tibetan Buddhist canon survives in two great transmission lines, Tshalpa and Thempangma, both attested centuries after the Tibetan translations sealed at Dunhuang. Join the cave to the canon stemmatically: if the later Kanjur descends from imperial-period exemplars rather than from re-translation,…

In the early ninth century the Tibetan empire standardized Buddhist translation vocabulary by decree, issuing an authorized dictionary of equivalents. Join a philologist's observation to an archivist's need: a vocabulary imposed at a datable moment is a stratigraphic layer, so undated Tibetan…

The Tibetan army's wooden slips at desert forts record the daily metabolism of an occupation: grain in, arrows out, men present, letters relayed. Join quartermaster arithmetic to source criticism: administrative output per garrison-year is roughly conserved across pre-modern armies, so surviving slip…

Bilingual manuscripts are usually catalogued by their languages; this conjecture claims the geometry is the message. On leaves from Dunhuang and Turfan carrying two languages, position encodes hierarchy: the liturgically senior language holds the primary text block, while the community's spoken language…

The Sogdians had no empire, yet a Sogdian slave-sale contract drawn up at Turfan in 639 reads like a formulary product, warranty clauses and witness structure included. Join contract law to network theory: a stateless trading diaspora must standardize its legal instruments…

Turgesh and other steppe coinages kept Sogdian legends long after courts and chanceries had moved on. Join numismatics to network economics: a coin's script must be readable across the whole trading network, so monetary script is locked in by network externality and…

Tibetan rule at Dunhuang ended in 848, yet locals went on using the Tibetan alphabet — including to write the Chinese language phonetically. Join imperial history to the classroom: an alphabet, once seeded, survives on its learnability rather than on its army,…

The most multilingual genre on the Silk Road was not scripture but divination: dice oracles and omen manuals exist in Old Turkic (the Irk Bitig), in Tibetan, and in Chinese, with recognizably shared mechanics. Join genre economics to translation history: divination is…

Cave 17 at Dunhuang held a striking quantity of Khotanese material, though Khotan lies weeks of travel away. Join the cave's chronology to Khotan's catastrophe: if manuscript deposits act as refuges for endangered textual communities, the Khotanese material at Dunhuang should cluster…