The Riddle of the Labyrinth (40 page)

A handful of Cyriot characters looked like Linear B signs, a resemblance first pointed out in 1927 by the scholar A. E. Cowley:

The Cypriot syllabary was a millennium younger than Linear B, and, as investigators knew, a lot can happen to a script in a thousand years. Similar-looking scripts often record extremely dissimilar languages: English, Hungarian, and Vietnamese, for instance, are all written in versions of the Roman alphabet. Even in related languages, identical characters can have entirely different sound-values. In German, a close relative of English, the letter
w
is pronounced “v,” while the letter
v
is pronounced “f.” One has only to compare the American and German pronunciations of the word
Volkswagen
to take the point.

However tenuous the connection, the Cypriot script offered the only external clue the Linear B decipherers had to work with, and none could resist exploring it. In the early 1940s, Kober had tried plugging Cypriot sound-values into the Knossos inscriptions but, as she later wrote, “had no results.” Evans, too, had tried to exploit the clue. On one fragmentary tablet in particular, it seemed to rear up seductively:

The fragment belonged to a tablet inventorying the horses of Knossos, and it appeared to count horses of two kinds: those with manes (shown at top center and bottom right), which were evidently full-grown horses, and those without (top left, top right, bottom center), evidently young horses. On the intact portion of the tablet, each maneless horse is preceded by a two-character Linear B word:
.

Evans tried substituting Cypriot values for those characters. What he got was “po-lo,” which looked an awful lot like
pōlos
, the Classical Greek word for “foal.” (The English word
foal
is a cognate, as is the name of the sport polo.) But championing Minoan supremacy to the last, he rejected the reading as mere coincidence—though one certain to be seized upon, as he wrote in a testy footnote in
The Palace of Minos
, by “those who believe that the Minoan Cretans were a Greek-speaking people.”

Ventris, who also believed that Linear B recorded a non-Greek language, had remained suitably wary of this Cypriot clue. But in May 1952, after revisiting his earlier idea about Cretan place-names, he allowed himself to follow it. As an experiment, he plugged certain Cypriot sound-values into his third grid, shown on the next page.

He started by unpacking
, which he had previously rejected as “Amnisos,” the Classical Greek name of the port of Knossos. From his analysis of the “pure vowel” signs, he was reasonably certain that
stood for “a.” He next turned to the Cypriot sign
, “na.” If the Linear B sign
had the same value, then he could insert “na” into the grid where Row C8 and Column V5 intersected—in the cell whose “phonetic coordinates”

were the consonant “n” and the vowel “a.” (On his grid, Ventris draws the character as
, an acceptable variant form.) Simplified, the relevant portion of his grid now looked like this: