Why We Still Cannot Read Etruscan

In 1964, excavators at Pyrgi, the port town of the Etruscan city of Caere, found three sheets of gold leaf rolled up and buried beside a temple. Two were inscribed in Etruscan. One was in Phoenician.

This was, on the face of it, the thing everybody had been waiting for. A bilingual. The Rosetta Stone had cracked Egyptian; here was the same gift for Etruscan, in gold, from around 500 BCE.

It did help. It just didn't do what everyone hoped, and understanding why is the whole story of this language.

It was never a code

First, clear away the usual framing. Etruscan is not a mysterious script awaiting a genius.

The Etruscans borrowed their alphabet from Greek traders in the eighth century BCE, and it stayed recognisably Greek. Anyone who reads Greek letters can sound out an Etruscan inscription after about ten minutes of instruction. Scholars have been able to do this for centuries.

So we can read every surviving Etruscan text aloud, confidently, including the long ones. We know the phonology. We know a great deal of the grammar: the case endings, how nouns decline, how verbs are marked, how the language builds a sentence.

What we lack is a dictionary. And a dictionary is a much harder thing to reconstruct than an alphabet, because an alphabet has about twenty-six answers and a vocabulary has thousands.

Nothing to translate from

Most ancient languages are cracked with a cousin. Discover that a word looks like the Latin, Greek and Sanskrit words for the same thing, and you have a meaning and a sound law thrown in. Indo-European scholarship is an enormous machine for exactly this, and it works.

Etruscan isn't Indo-European. It's not Semitic either, or Uralic, or anything else with a large attested family.

The best current view, set out by Helmut Rix, groups it with two other languages into a small family usually called Tyrsenian: Etruscan, Raetic from the eastern Alps, and Lemnian, known from a handful of inscriptions on the Aegean island of Lemnos. The grammatical resemblances are real and reasonably convincing.

The trouble is arithmetic. Raetic and Lemnian survive in tiny quantities, and the three languages appear to have split early enough that shared vocabulary is thin. Having relatives helps confirm what kind of language Etruscan is. It barely helps at all with what any given word means.

Thirteen thousand inscriptions, one sentence

Here is the part that surprises people, because the raw number sounds so promising.

Around twelve or thirteen thousand Etruscan inscriptions survive. That's a substantial corpus, more text than we have for plenty of languages we understand perfectly well.

Except that the overwhelming majority are epitaphs and dedications, and they are one or two lines long. A name. A father's name. A mother's family. An age at death. Sometimes an office held. Then the next tomb, doing the same thing again.

Thousands of examples of the same handful of formulae. It's a magnificent resource for names, family structure and grammatical endings, and we know an extraordinary amount about how Etruscans were named and who they were related to. For vocabulary it's close to worthless, because it uses roughly the same forty words over and over.

Continuous prose amounts to four documents. The Liber Linteus, at about 1,200 words, preserved as mummy wrapping. The Tabula Capuana, a tile with around three hundred. The Tabula Cortonensis, a bronze tablet found in 1992 carrying about two hundred. The Cippus Perusinus, forty-six lines on a boundary stone.

That is the entire literature. Put together, it comes to something like two thousand words of running text: ten minutes' reading, against the millions of words that survive in Greek.

And of those four, three are administrative or legal and one is a liturgical calendar. Nobody left us a story, and the one man who wrote twenty books about them lost every word of it.

Why the bilingual wasn't enough

Which brings us back to the gold sheets.

The Pyrgi texts commemorate the same event: a dedication by a ruler of Caere to a goddess, Uni in Etruscan and Astarte in Phoenician. The Phoenician version can be read straightforwardly.

But the two are not a translation in the sense the Rosetta Stone is. They're parallel compositions, each phrased naturally in its own language, of different lengths, saying overlapping rather than identical things. You cannot line them up word by word and read the meanings off.

What they gave was a genuine, useful handful: confirmation of some divine names, some vocabulary around dedication and time, a few grammatical points. A good day's work for a field where progress is measured in individual words.

They did not give a lexicon, because there wasn't one to give.

Two dice, six words, still arguing

If you want a single object that captures the state of the field, it's a pair of ivory dice found near Tuscania.

Instead of pips, the faces carry words: mach, thu, zal, huth, ci, śa.

So we have, indisputably, the Etruscan words for one to six. We've had them for a long time. And there is still no complete agreement about which word is which number, because the argument has to be settled by inference: from the convention that opposite faces sum to seven, from number-words embedded in ages on tombstones, from the order they appear in elsewhere.

Six words. Certain meaning as a set. Decades of debate over the assignment.

This is what the earlier centuries of the field got wrong, badly and repeatedly. Scholars tried to force Etruscan into Hebrew, Albanian, Hungarian, Basque and most of the rest of Eurasia, and each attempt produced confident translations that dissolved on contact with the next inscription. The method that actually works is slower and less satisfying: study Etruscan against itself, watch what varies with what, and accept a word at a time.

It is also why there won't be a breakthrough moment. The realistic path is new long texts coming out of the ground (the Cortona tablet in 1992 was worth more than a decade of theorising) and each one moves the dictionary along by a few dozen entries.

We can pronounce a religion's entire surviving liturgy. We just can't say what it asked anyone to do.