what you were reading

The page counted before anyone spelled it out.

The field you have just handled is a modelled reconstruction of a tribute page of the kind bound into the Codex Mendoza — a manuscript painted in the Aztec pictorial tradition in early colonial New Spain, and held today in the Bodleian Library in Oxford.

Before an alphabet touches such a page, it is already fully readable. Place signs run down one edge and along the foot; the goods owed stand in registers above them; and quantity is not written but repeated. A dot is one. A flag is twenty. A plume is four hundred. A pouch is eight thousand. You count with your eyes, and the size, direction and grouping of things carry as much sense as any word.

the vigesimal signs, drawn here in our own hand

the hands

The painters are not named.

Nahuatl calls them tlacuiloque — those who paint‑write; one of them is a tlacuilo. The word does not separate painting from writing, because in this system the separation does not hold. Colour, placement, repetition, scale and the direction a figure faces are all grammar.

They were at work a generation after the fall of Tenochtitlan, in a city under Spanish rule, trained in a tradition that had not stopped. What remains of them is not a signature but the conduct of the hand: the weight of a line, the ordering of registers, the decision about what is worth repeating and what can be assumed.

Naming any of them here would be an invention. We can say only that they were skilled, that they were Indigenous, and that the manuscript is unthinkable without them.

the second system

Then another hand wrote into the margins.

The Spanish text was added inside the same object: short glosses set against the painted signs, and longer prose facing them. It names, dates, totals and explains — and in doing so it re‑routes the eye. Alphabetic writing runs in lines; it wants a beginning and an end. The painted page does not.

Once a number is spelled, you stop counting the plumes.

Both systems are working. Neither is decoration. But they ask for different readers, and where they meet, one of them is doing the arranging.

On the manuscript's last page the Spanish commentator excuses the roughness of his work, explaining that the interpretation was given to him only shortly before the fleet sailed. That remark is in the manuscript; it is paraphrased here rather than quoted, and it tells us that the writing was hurried and second‑hand — an explanation received, not a knowledge held. documented

what the manuscript holds

Three parts, bound as one.

1. part iA history of the Mexica rulers and the towns brought under Tenochtitlan — reigns and conquests, told in signs.
2. part iiA tribute record: which towns delivered what, and how much. It stands very close to the Matrícula de Tributos in Mexico City — either copied from it, or drawn from a source both share.
3. part iiiA life, year by year: birth, naming, schooling, work, marriage, punishment, office, old age. No pre‑conquest prototype for this part is known; it appears to have been composed for this manuscript, with a European reader in view.

what is not known

The manuscript is surer than its story.

It is called the Codex Mendoza after Antonio de Mendoza, first viceroy of New Spain, on the traditional account that he had it made for the emperor. That account is old and widely repeated; a document of the commission has not been found.

Its route to Europe is documented in parts. The name of André Thevet, cosmographer to the French crown, is written in the manuscript, dated 1553. It later belonged to Richard Hakluyt and then to Samuel Purchas, who published engraved copies of it in 1625. It reached the Bodleian Library with John Selden's books in 1659. The frequently told story of a capture at sea by French corsairs does not rest on the same kind of evidence.

Not known: the painters' names. The exact year. Who ordered it. Who wrote the Spanish. The honest shape of this page is a documented object with an argued history.

the boundary of the reconstruction

What we made, and what we did not.

• The leaf you handled is drawn here. It traces no existing folio and copies no published facsimile.
• The place signs are schematic hill glyphs. No town is named, because naming one would manufacture evidence.
• The Spanish words, figures and script blocks are modelled. They show where annotation sits and what it does to reading — not what any annotator wrote.
• The counting signs follow the numeration described in the published scholarship; their arrangement on this page is our own composition.
• The reading routes are an interpretation made visible so it can be argued with. They are not a claim about how any particular reader's eye moved.
• The colours are chosen to sit with the manuscript's known palette. They are not a pigment analysis.

documented

attested in the manuscript itself or in the archival record.

interpreted

a scholarly reading, well supported but a reading.

disputed

actively argued, or resting on tradition rather than record.

reconstructed

made for this station to be encountered, not found.

where this comes from

Read further, and check us.

1. Bodleian Library, University of Oxford. Codex Mendoza. MS. Arch. Selden. A. 1. Digitised in full in Digital Bodleian.
2. Frances F. Berdan & Patricia Rieff Anawalt. The Codex Mendoza, 4 vols. Berkeley: University of California Press, 1992. (Includes H. B. Nicholson, “The History of the Codex Mendoza”, vol. 1.)
3. Frances F. Berdan & Patricia Rieff Anawalt. The Essential Codex Mendoza. Berkeley: University of California Press, 1997.
4. Elizabeth Hill Boone. Stories in Red and Black: Pictorial Histories of the Aztecs and Mixtecs. Austin: University of Texas Press, 2000.
5. Jorge Gómez Tejada, ed. Códice Mendoza: nuevas perspectivas. Quito: USFQ Press, 2019.
6. Instituto Nacional de Antropología e Historia (Mexico) with the Bodleian Libraries. Códice Mendoza, digital interactive edition, 2015.
7. Matrícula de Tributos. Biblioteca Nacional de Antropología e Historia, Mexico City — the comparison for part two.