The Riddle of the Labyrinth (10 page)

Did each pair show variants of the same sign, or different signs entirely? There was no swift way of telling. One of the American investigators, the Brooklyn College classics professor Alice Kober, spent years agonizing over the symbol
, and whether it was the same or different when it had just one horizontal crossbar instead of two, as the scribes sometimes wrote it.

Nor were the Linear B decipherers working from nice, neat typefaces, either. (Today, there are several digital fonts available for Linear B; the one used in this book is known as Aegean.) Clay is splendid for stamping, as the Sumerian cuneiform tablets plainly show. It is a less good medium for inscription. Wet clay exerts drag, and as a stylus moves through it, it leaves burrs in its wake, distorting the look of the characters. If you are at all doubtful, try writing your name in a block of Plasticine with a straight pin taped to a stick.

There were perhaps seventy scribes at Knossos, and, as with any group of writers, some simply had better penmanship than others. To make matters worse, whenever the Cretan scribes made errors—provided they caught them in time, for dried clay is the original read-only medium—they “erased” them by smearing the clay with a finger (or the flat edge of a wooden stylus) and overwriting the correct sign. This often rendered the new sign blurry and, to modern eyes, hard to interpret.

Evans, at least, had the advantage of being able to see the Knossos tablets firsthand and make careful drawings of the characters. Not so for the other scholars, who had to make do with the few, poor-quality photographs Evans made available. In the end, the task of compiling a definitive list of Linear B signs was so difficult that despite years of work by scholars on both sides of the Atlantic, it was not completed until 1951.

At the same time the decipherer is counting the signs, he must also determine the direction in which the script is written. Most Western scripts run from left to right. Other scripts, like Phoenician, Hebrew, and Arabic, flow from right to left. Still others, like Chinese and Japanese, are written from top to bottom.

Scripts can run in stranger ways. In its earliest days, the Greek alphabet, like its mother, the Phoenician, was written right to left. Eventually, it settled into its familiar left-to-right order—an order that the alphabets descended from it share today. But for a time in between, Greek writing literally flopped back and forth within a single document, running right to left and left to right in alternate lines. The result, a curving thread of script that calls to mind a furrowed field, is known as boustrophedon, Greek for “as the ox turns.” The following inscription, written in a version of the Greek alphabet used in the sixth century B.C., is an example of boustrophedon writing. Notice, too, that in the “backward” lines, the direction of each letter is also reversed:

Even odder-looking is reverse boustrophedon, in which every other line is written upside-down. (The Germans call this style
Schlangenschrift
, or “snake writing.”) Rongorongo is written like this, suggesting that the reader had to turn the tablet 180 degrees at the start of each line.

There are several ways to determine the direction of an unknown script. Punctuation helps, of course, but not all writing systems use punctuation. The most effective test is to scan the writing surface for “white space”—places where the text has filled out a line incompletely. If our alien were studying this book, the unfilled spaces at the right-hand ends of many paragraphs would strongly suggest that the Roman alphabet runs left to right. Knowing the script direction gives him essential information about English spelling: He now knows, for example, that the word just before the colon in this sentence is
spelling
and not
gnilleps
. Such distinctions are crucial to decipherment.

As Evans examined the tablets, he made use of similar clues. On large tablets, long texts were broken up into paragraphs, which often ended with empty space on the right-hand side. He knew then that Linear B ran from left to right.

The decipherer's next task is to find word breaks. Many writing systems mark divisions between words with spaces or by other means. But some systems do not. Boustrophedon writing, as in the early Greek example above, arranges characters in one long, sinuous line, with no spaces between words.

A script with word divisions gives the decipherer a running start. Imagine trying to solve a cryptogram in which all the characters are run together, like this:

The script is a modern version of the Dancing Men cipher, invented by Sir Arthur Conan Doyle for “The Adventure of the Dancing Men,” a Sherlock Holmes story from 1903. The secret messages in Conan Doyle's original story required only seventeen discrete characters; the designers of the several Dancing Men fonts available today have created the rest.

As it appears above, the cipher is one continuous stream, a mass of black symbols with no breaks of any kind. If this were an ancient script, its decipherer would be at a deep disadvantage: Without knowing where words start and end, he has no ready way to tell whether the system is logographic, syllabic, or alphabetic, much less what the language might be—or even if the strange symbols are writing at all.

Now look at the same text, this time written with the symbols as Conan Doyle meant them to be used:

At first glance, the two blocks of text look alike. But closer inspection reveals that some figures in the second cipher are holding tiny flags. In the cipher on which “The Adventure of the Dancing Men” hinges, Conan Doyle used these flag-bearing figures to signal word breaks. Holmes, being Holmes, picked up on the flags immediately and cracked the code.

With the word breaks marked, the second text gives a decipherer far more to work with. He now knows that the cipher contains a one-letter word (the character
), which, assuming the text is in English, is almost certainly “a” or “I.” The text also contains a three-letter word,
, which is a good candidate for “the.” (Both texts above read, “The Dancing Men is a cipher invented by Sir Arthur Conan Doyle.”)

In this respect, too, Evans was lucky. On the tablets, the Linear B scribes had thoughtfully separated groups of signs with small vertical tick marks, as in this six-word line, part of a longer inscription: