Some of the most thrilling detective stories in history involve no crime at all, only marks on stone and clay that nobody could read. Dr Helena Marsh, who has spent twenty years studying ancient writing systems, talked to us about how scholars cracked some of the world's most baffling scripts, and why a few still refuse to give up their secrets.
Interviewer: Where does a decipherment usually begin? Helena Marsh: Almost always with a question of scale. If a script has around twenty-five to thirty signs, it's probably an alphabet. If it has eighty or ninety, it's likely to be a syllabary, where each sign stands for a syllable such as 'ka' or 'to'. Hundreds of signs suggest a system that includes signs for whole words. Simply counting the signs tells you a surprising amount before you've read a single word.
Interviewer: The Rosetta Stone is surely the most famous example. Helena Marsh: It is, and deservedly so. It was found in 1799 near the town of Rashid, which Europeans called Rosetta, in Egypt. What made it so valuable was that it carries the same official decree in three scripts: Egyptian hieroglyphs, a later Egyptian script called Demotic, and ancient Greek. Since scholars could read the Greek, they had a bilingual key, or rather a trilingual one. Even so, it took more than twenty years. The English scholar Thomas Young made important early progress, but it was the Frenchman Jean-François Champollion who announced the breakthrough in 1822.
Interviewer: Why did it take so long? Helena Marsh: Because of a mistaken assumption. For centuries, hieroglyphs were thought to be purely symbolic: a picture of a bird stood for an idea, not a sound. Champollion's great insight was that many signs represented sounds, and that the system mixed sound signs and meaning signs. Once that misconception was abandoned, progress was rapid. It's a lesson I give my students: sometimes the hardest obstacle isn't the evidence but what you expect to find.
Interviewer: Not every script came with a convenient translation, though. Helena Marsh: No, and that's what makes Linear B so remarkable. It was found on clay tablets in Crete and mainland Greece, and there was no bilingual text at all. The American scholar Alice Kober did painstaking groundwork in the 1940s, analysing how words changed their endings. Then, in 1952, Michael Ventris, a young British architect who had been fascinated by the script since he was a schoolboy, showed that it recorded an early form of Greek. Most experts had been convinced it couldn't be Greek, so his result came as a genuine shock.
Interviewer: And the Maya script? Helena Marsh: That's a more collaborative story. Maya glyphs were deciphered gradually during the second half of the twentieth century. The Russian scholar Yuri Knorozov argued in the 1950s that the script used sound signs, an idea that was dismissed by many at first, and Tatiana Proskouriakoff showed that many inscriptions recorded the lives of real rulers. Today, scholars can read most known Maya texts.
Interviewer: Which scripts remain unread? Helena Marsh: Several. Linear A, the older Cretan script, the Indus script of ancient South Asia and the rongorongo of Easter Island are the best known. The problem is usually not a lack of cleverness but a lack of material. The surviving Indus inscriptions are mostly very short, often just a few signs on a seal. Without longer texts, or a bilingual inscription, any proposed solution is hard to verify. Interviewer: Will they ever be cracked? Helena Marsh: I'd never say never. Computers now help us spot patterns far faster than we could by hand, but a computer can't tell you what a word means. That still requires intuition, patience and, very often, a lucky discovery in the ground.