Experli

Mathematics

The thousand-year road trip of the digits 0–9

Published July 5, 2026

# The thousand-year road trip of the digits 0–9

Look at any number on this page. The notation you are effortlessly parsing — ten digits, value by position, zero holding empty places — is the single most successful piece of intellectual technology in human history, the only writing system every literate person on Earth now shares. It came from India, and its journey to your screen is documented, stage by stage, in inscriptions and manuscripts. The trip took about a thousand years.

Stage one: out of India, by ~600 CE

The system's Indian origin story is told elsewhere in this corpus — [the number-name ladder of the Vedas](/c/30ed9b53-ddb7-5087-820e-341220291d7b), [the dot-zero of the Bakhshali manuscript](/c/a606b1f4-a934-5031-887e-0968fad00c24), [Brahmagupta's rules making zero a number](/c/d10352e6-dc8c-58ee-bbbf-9b1489efdc9e). By the sixth century the place-value system was in working use across the subcontinent — and beyond it. Datta and Singh's *History of Hindu Mathematics* (1938), the standard source-book, anchors the early spread on epigraphy, including a Cambodian inscription of 605 CE using place-value notation:

> "…by the end of the 6th century A.D., the knowledge of the system > had spread over an area roughly of the size of Europe." > > — Datta & Singh, *History of Hindu Mathematics*, Vol. I (1938)

Precision matters here, so: the 605 CE Cambodian evidence is place-value *word-numeral* notation (numbers spelled with value-bearing words in positional order); the region's earliest surviving *digit* zero is the famous Khmer "605" (Śaka era — 683 CE) inscription. Both are the system; the word-form and digit-form traveled together. Southeast Asia was writing Indian-style numbers while Europe's largest library still counted in Roman numerals.

Stage two: Baghdad, 8th century

"In Arabia," Datta and Singh record, "the new system was introduced in the eighth century" — the Sanskrit astronomical treatises arrived at the Abbasid court around 770 CE and were translated as the *Zīj al-Sindhind*. In the early 800s, al-Khwārizmī of Baghdad wrote the book that named the subject: his treatise on reckoning with the Hindu numerals, lost in Arabic, survives in Latin as *Dixit Algorizmi* — "Thus spoke al-Khwārizmī." His Latinized name became the word **algorithm**; the system itself became "Arabic numerals" in European mouths, a mislabel the Arab mathematicians themselves never used — their own name was *ḥisāb al-hind*, Hindu reckoning.

Honesty about the relay: the Islamic world was no passive courier. Arab and Persian mathematicians extended the system — al-Uqlīdisī worked out decimal fractions in the 10th century, centuries before Europe — and it was their commercial arithmetic, not raw Indian texts, that Europe eventually copied. Even in Arabia adoption was slow: "it did not come into common use," Datta and Singh note, "until five or six hundred years later." Notation changes at the speed of habit, not of proof.

Stage three: Europe, resisting until the 16th century

Europe met the numerals repeatedly — Gerbert of Aurillac (the future Pope Sylvester II) used them on counting-board apices around 980; Fibonacci's *Liber Abaci* (1202) opened by announcing "the nine figures of the Indians" and demonstrating what merchants could do with them. And still, adoption crawled. Florence banned the new figures in contracts in 1299 (too easy to falsify, said the bankers). Account books, university records, and church registers clung to Roman numerals for centuries. Datta and Singh's chosen witness is wonderfully mundane:

> "It was exceptional for common people to use the new system before > the sixteenth century… Calendars of 1557–96 have generally Roman > numerals."

Popular almanacs — the mass media of the day — were still printing dates as MDXCVI while Kepler was a student. A full millennium after the Cambodian inscription, the system was still fighting for the European kitchen table.

Why it won anyway

Positional notation wins because arithmetic in it is an algorithm: digit-by-digit routines a clerk can execute without understanding — the same property that lets silicon execute them today. Every rival notation (Roman additive, Greek alphabetic, Chinese rod-numeral hybrids had their own strengths) required either tables or a calculating device; the Indian system *is* the calculating device, flattened onto paper. That's what Fibonacci was selling, what the Florentine bankers feared, and what finally made the almanacs switch.

The fair summary the claim insists on: invented in India, proven in Sanskrit astronomy, industrialized in Arabic commerce, and adopted by Europe last and loudest — Europe then carried it worldwide and named it after the middleman. Every stage is documented; no stage was fast. The digits crossed Asia before they crossed the Alps, and the almanac-buying public of Elizabethan England was, numerically speaking, the last place on the route still saying no.

---

Sources

- [Datta & Singh, *History of Hindu Mathematics: A Source Book*, 1938](https://archive.org/details/wg143) — Vol. I, Numeral Notation: the 605 CE Cambodian epigraphy and spread estimate; Arabia and Europe adoption timelines; the 1557–96 almanac evidence. - Plofker, *Mathematics in India*, 2009 — transmission to the Islamic world (secondary synthesis, for context). - Chrisomalis, *Numerical Notation: A Comparative History*, 2010 — the comparative and European-adoption detail (Florence 1299, Gerbert; referenced for context only).

Related claims

- [The Bakhshali manuscript's zero dot (~3rd–4th c. CE)](/c/a606b1f4-a934-5031-887e-0968fad00c24) - [Brahmagupta writes the rules for zero (628 CE)](/c/d10352e6-dc8c-58ee-bbbf-9b1489efdc9e) - [The Yajurveda's ladder of tens (~1200–800 BCE)](/c/30ed9b53-ddb7-5087-820e-341220291d7b)

References

  1. [1]Hoernle's 1888 study of the Bakhshali manuscript (a birch-bark mathematical text discovered in 1881 near Peshawar) describes the dot serving two roles: as a placeholder for an unknown quantity (analogous to modern x) AND as a fundamental digit in the decimal place-value system — the zero. The manuscript was carbon-dated by Oxford's Bodleian Library in 2017 to between 224 and 383 CE, making it the earliest extant evidence of place-value zero in any tradition. Source: On the Bakshali Manuscript (T1)
  2. [2]The Yajurveda Saṁhitā (Vājasaneyī xvii.2, c. 1200–800 BCE) lists thirteen decimal denominations — eka (1) through parārdha (10¹²) — each ten times the preceding; the same list recurs in the Taittirīya Saṁhitā. Datta & Singh (1938) contrast this with Greek terminology, which stopped at the myriad (10⁴), and Roman, at mille (10³). Named decuple ranks are a documented Vedic-era feature of Sanskrit, many centuries before written place-value numerals. Source: History of Hindu Mathematics — A Source Book (T1)
  3. [3]The decimal place-value system originated in India and spread in documented stages: epigraphic evidence (including a 605 CE Cambodian inscription) shows coverage "roughly of the size of Europe" by the end of the 6th century; Arab mathematicians adopted it in the 8th century; common European use came only around the 16th — popular almanacs of 1557–96 still print Roman numerals. Datta & Singh (1938) document each stage from primary evidence. Source: History of Hindu Mathematics — A Source Book (T1)
  4. [4]Brahmasphutasiddhanta XVIII.19 (628 CE) gives explicit rules for arithmetic with zero as a number: addition, subtraction, multiplication, square root. Brahmagupta also writes a rule for division by zero — getting it wrong (treats x/0 as finite) but pioneering the question itself. Bhaskara II refined the rule ~500 years later (1150 CE) treating 1/0 as khahara — closer to modern infinity-as-limit. Source: Algebra, with Arithmetic and Mensuration, from the Sanscrit of Brahmegupta and Bháscara (T1)