The Riddle of the Labyrinth (21 page)

Deciphered, the four words turn out to be these:

Here, “inflection” is a false friend, for as any English speaker knows, there is no suffix
-ger
in the language; nor are there stems
fin-
(in this sense),
wa-, la-
, or
ti-
. The repeated pattern at the ends of these words is mere coincidence and nothing more.

Had Evans truly found inflection in Linear B, or had he been similarly seduced? Though he wrote of his little paradigms, “We have here, surely, good evidences of declension” (that is, inflection on the ends of nouns), his discovery was intuitive and circumstantial, and had to be taken on faith. Kober knew that for the decipherment to advance, someone had to prove conclusively whether or not the Minoan language was inflected. This she set out to do.

If you are studying an unknown language in an unknown script and you see what looks like inflection, how do you prove it actually
is
inflection? For Kober, the answer was to be found in the “patterns of selection and arrangement” that had emerged over the years as she pored over her cards. She had let the tablets speak for themselves and now they were beginning to reward her.

“If a language has inflection, certain signs are bound to appear over and over again in certain positions of the written words, as prefixes [or] suffixes,” she wrote in her 1945 article. “No matter how much these changes may be obscured, the fact that they occur regularly must reveal them, if the amount of material available for analysis is large enough, and the analysis sufficiently intensive.”

As she studied the “Chariot” inscriptions, Kober noticed certain words, and even phrases, that recurred on two or more tablets. Among them was the word
, which appeared on two tablets;
, which appeared on three; and
, found on five. She also identified an entire three-word phrase:
, which was repeated on five different tablets.

Strikingly, Kober also found words
that recurred in similar but not quite identical forms
—forms that differed only in their last few characters. The words seemed to have the same stem but different suffixes. They included
and
(where the stem starts with ); and (where it starts with
);
and
(wher is starts with
); and
and
(where it starts with
).