How do you count to ten on a Roman's fingers?

Medieval manual

In 725, the monk Bede the Venerable opened his treatise On the Reckoning of Time (De temporum ratione) with an unexpected chapter: De computo vel loquela digitorum — “On the reckoning or speech of the fingers”. It was a working tool: clergy computing the date of Easter needed figures to hand, literally.

Bede's system divides each hand into two fields. On the left: the last three fingers show units (1–9), thumb and index show tens (10–90). On the right: thumb and index show hundreds (100–900), the last three fingers show thousands (1,000–9,000). Two hands thus reach 9,999.

Bede, 725

Bede takes the system past the wrist: from 10,000 to 90,000 the left hand touches chest, navel, thigh; from 100,000 to 900,000 the same on the right; and for one million the hands join. But a line must be drawn here: there is no ancient evidence for the body positions — we know them from Bede alone, and other medieval manuals give different instructions. The historian J. Hilton Turner explicitly warns against projecting them into antiquity.

Bede writes in 725. So how do we know the Romans did anything similar? From three independent kinds of evidence: objects, texts and the manual itself — fitted together by historians.

Roman artefact

1. The objects. Roman-period ivory tesserae show a hand gesture on one face and the numeral on the reverse. Turn the pieces below: 4, 7, 12. The gestures correspond closely to the later instructions — with small exceptions debated by scholars (Turner flags 10 and 30; other studies, 6).

These are the Roman pieces. Bede gives the rest of the instructions.

Roman text

2. The texts. Ancient authors mention finger numbers as something readers already know — evidence of a convention, not of every finger bend:

Medieval manual Reconstruction

3. The manual — and the reconstruction. No ancient author left us one complete manual: early medieval texts allow the whole system to be reconstructed, and the archaeological and literary pieces show strong continuity between the ancient and medieval traditions. So the modern synthesis of reference puts the conclusion (Williams & Williams, 1995) — “strong continuity”, not “every Bede gesture archaeologically proved in Rome”.

Between Bede and print, finger reckoning passes through merchants' hands. Fibonacci recommends it in Liber abbaci (1202) to those who want to keep large sums in memory, and the abbacus tradition carries it on to Pacioli's famous printed chart in the Summa de arithmetica (Venice, 1494, fol. 36v) — the first printed book with a finger-numeral chart.

But Pacioli's chart is not Bede photocopied. It belongs to a later version of the tradition, transmitted through medieval commercial arithmetic — and on the right hand the fields are reversed against Bede: hundreds on the last three fingers, thousands on thumb and index. Scholars attribute the rearrangement to Pacioli, on the logic of “small fingers, small numbers”. The first three gestures (1, 2, 3) stay recognisably the same.

An honest note: the first version of this page placed hundreds and thousands exactly as Pacioli does, while calling it Bede. Each scheme is now in its place — and the counter above follows Bede's configuration.

Hands leave numbers and become words. And here the wrong question is “how many Latin names survived?” — because the Romans themselves had no single list: Isidore of Seville (7th century) gives pollex, index or salutaris, “the shameless one”, anularis or medicinalis, auricularis — while classical Latin also said medius or minimus. The good question is: what exactly survived — the word or the idea?

• Directly inherited — the Latin word lives in Romanian (digitus → deget).
• Formed in Romanian — a new label from inherited material (inel → inelar).
• Learned borrowing — the Latin form re-enters late, through books (index → index).
• Same idea, different word — the naming principle survives without the word (index ↔ arătător).

And a coda: from the same Latin family, by the learned international route, Romanian also has digital (DEX: from Fr. digital; DN: cf. Lat. digitalis). Deget and digital are one Latin root living at opposite ends of the vocabulary — from fingers to digits to computers. The link is lexical, not a direct historical cause.

The comparison that matters follows the route: by which road each name reached its present form.

The French case deserves a pause: index looks ultra-Latin, but the Académie française dates it to the 16th century as a borrowing from Latin. Romanian arătător looks not Latin at all — but it keeps almost exactly the naming logic of index. The language that looks less Latin here preserves the Roman logic more transparently.

Hands change their vocabulary more easily than their logic: pointing, middle, ring, smallness — the same human observations keep bearing new names.