The Riddle of the Labyrinth Read online

  In the course of the season, Evans’s workers unearthed an exquisite alabaster vase shaped like a triton shell; the high alabaster throne; pieces of many different frescoes, which Evans would arrange to have painstakingly restored; statuary; painted pottery; and the charred remains of graceful wooden columns that, like those at Mycenae, were smooth, round, and downward-tapering, wider at the top than at the bottom. So delighted was Sir John with his son’s discoveries that he sent him 500 British pounds.

  But all this, as Evans’s assistant John Linton Myres would write, was “almost thrown into the shade” by the discovery that season of the inscribed clay tablets, “a discovery which carries back the existence of written documents in the Hellenic lands some seven centuries beyond the first known monuments of the historic Greek writing.”

  On March 30, workmen lifted from the soil of Knossos “a kind of baked clay bar, rather like a stone chisel in shape, though broken at one end, with script on it and what appear to be numerals,” as Evans later wrote. He saw immediately that the linear script on the bar resembled that of the clay “slip” he had been shown in 1895. The slip had been destroyed not long afterward, a casualty of the Cretan insurrection, but Evans had had the foresight to make a copy of it.

  The discovery of writing at Knossos was, Evans later wrote, “the dramatic fulfillment of my most sanguine expectations.” In the weeks that followed, his workmen unearthed more and more tablets, many badly broken but some completely intact. At times, they came upon entire small chambers filled with tablets, the archival record rooms of the Palace of Minos. On April 5, 1900, they encountered a terra-cotta bathtub full of tablets—they had fallen into the tub centuries before, when an upper story gave way. Also in the tub were bits of charred wood, suggesting the tablets had originally been stored in wooden boxes. This was confirmed elsewhere in the palace, when tablets were found next to a set of small bronze hinges—hinges that had held the wooden box lids in place. On May 10, Evans wrote his father that he had “just struck the largest deposit yet, some hundreds of pieces.”

  Like the clay bar Evans first unearthed at the site, most of the Knossos tablets were small and wedge-shaped, between two and seven inches long and a half inch to three inches high. Tapered at the ends, they were clearly designed to fit comfortably in a scribe’s hand. They were made of ordinary local clay, incised with a stylus of some kind.

  In addition, there were some larger, squarer tablets, also made of clay and sometimes fashioned around armatures of straw. Where the smaller tablets had room for just a line or two of writing, perhaps ten or twenty characters in all, the larger tablets held much more. One very large rectangular tablet, more than ten inches high and six inches wide, was incised with twenty-four lines of text. It would come to be known as the “Man” tablet for the column of “man” pictograms () down its right-hand edge.

  The twenty-four-line “Man” tablet from Knossos. The “man” symbol () followed by a numeral, is repeated down its right-hand edge.

  Arthur J. Evans, The Palace of Minos, Volume IV

  The scribes of Knossos were superb bureaucrats. Tablets were arranged by subject in their boxes, the file cabinets of the Bronze Age Aegean. Once filled and closed, each box was secured with a clay sealing indicating its contents. The sealing was impressed with a seal-stone like the ones Evans had encountered on his first trip to Crete. (Long before the Knossos tablets could be read, their subject matter—grain, livestock, chariots, weapons, and the like—could often be gleaned from the pictograms on the seals.)

  Evans would identify three different scripts at Knossos. The first was a hieroglyphic script, the same kind he had seen on the seal-stones and engraved gems. From the depth of the soil in which traces of the script were found, he determined that it was in use from about 2000 to 1650 B.C. But while this script was found often on the clay sealings used to secure boxes of tablets, it was rare on the tablets themselves: In the whole of the palace, Evans came across only a single cache of tablets bearing hieroglyphic inscriptions.

  Other tablets featured what he described as “a new system of linear writing,” which had evolved from the hieroglyphic script in the eighteenth century B.C. This linear script was a “style of writing fundamentally different from that of the hieroglyphic class, and far ahead of it in development. . . . The letters themselves . . . are of a free, upright European character.”

  By 1902, Evans had further distinguished two types of linear script. The first, Linear Script Class A, was used from about 1750 to 1450 B.C. The second, Linear Script Class B, developed out of Linear A toward the end of this period. It was in use until the final destruction of the palace in the early fourteenth century B.C. Evans called the scripts “linear” not because their characters were arrayed in lines, although they were, but because those characters were made by means of linear strokes—a method quite different from the cuneiform writing of ancient Mesopotamia, in which signs were impressed in clay with a wedge-shaped tool. (With its unadorned outline forms, the Phoenician alphabet, which gave rise to nearly all the alphabets of the modern world, was also a linear script. So are its descendants, including our own familiar Roman alphabet.)

  In contrast to Linear B tablets, Linear A tablets, like these, are nearly always unruled.

  Arthur J. Evans, The Palace of Minos, Volume IV

  The great majority of the Knossos tablets were written in Linear B, including the first ones unearthed in the spring of 1900 and the large ones containing many lines of text. Linear B looked similar to Linear A but was by no means identical. The two scripts had many characters in common—among them and —but each also used characters not found in the other. The B script looked neater and more stylized than the A. Most Linear A texts were incised directly onto unruled clay, giving the writing a somewhat scattershot appearance. In contrast, Linear B tablets were nearly always ruled: The text sat on tidy horizontal lines that had been cut into the wet clay before the writing began. “Evidently the tablets were supplied in this state to the clerk, like ruled sheets of paper in a modern business office,” Evans wrote.

  Of the three Cretan scripts Evans discovered—hieroglyphic, Linear A, and Linear B—Linear B stood the best chance of being deciphered. As with any secret code, the more text a decipherer has to work with, the greater the likelihood of solution. The number of Linear B tablets unearthed at Knossos far outstripped any other kind; over time, more than two thousand would be found there. From the beginning, it was Linear Script Class B, used in the twilight days of the Palace of Minos, that held the greatest promise.

  The Linear B tablets have a stark beauty. Some have smooth, charcoal-gray surfaces resembling slate, others are reddish brown, still others are bright orange. (The color depends on the level of oxygen to which they were exposed when the palace burned down.) The incised characters are generally crisp and made with care. They are, as Evans put it, “the work of practised scribes.” On the backs of tablets, those scribes left traces of themselves in the form of fingerprints and even doodles. To look at the tablets even now is to be in the presence of other people—living, thinking, literate people.

  This feeling animates all archaeological decipherment. The pull of an undeciphered ancient script comes not only from the fact that its discoverer cannot read it, but also from the knowledge that once, long ago, someone could. To Evans, the scribes of Knossos were real people who had set down the workings of their Bronze Age world, precisely and deliberately, on pieces of wet clay. Men could read those tablets once. It should be possible, even after thirty centuries, for man to read them again.

  Three-thousand-year-old scribal doodles from Knossos (top) and the Greek mainland (bottom)

  John Chadwick, The Mycenaean World; reprinted with the permission of Cambridge University Press

  “We have here locked up for us materials which may some day enlarge the bounds of history,” Evans’s assistant John L. Myres wrote in 1901. “The problems attaching to the decipherment of these clay records are of enthralling interest.” br />
  And so they would be, for fifty years to come.

  2

  THE VANISHED KEY

  ABOUT FIVE THOUSAND YEARS AGO, man stopped having to remember quite so much. Spoken language had already been in existence for at least fifty thousand years, evolving hot on the heels, evolutionarily speaking, of the dawn of man. But it wasn’t until long afterward that man realized he could set down his language in graphic form, using visual symbols to encode speech and store it for later retrieval. For the first time, people did not have to depend on memory alone to transmit the history, lore, and daily activities of their communities. We call these marvelous storage-and-retrieval systems writing.

  One of the foremost inventions in the history of mankind, writing probably developed independently in several places at around the same time. Before that time, people relied on a range of crude systems, like knotted strings, clay tokens, or notches cut in sticks, to help them count, tally, and remember. Scholars call these proto-writing. But writing proper—a full symbolic system that can record any imaginable text in a community’s language—began only with the advent of Sumerian cuneiform in about 3300 B.C. In a separate though possibly related development, the hieroglyphs of ancient Egypt arose around the same time.

  Writing systems seem to have been rare in the ancient world, and even today, they are something of a linguistic luxury: By some estimates, only about 15 percent of the roughly six thousand languages spoken around the globe have written forms. It is entirely possible to have language without writing. Not so the other way round.

  A writing system is simply a map. It works by taking the sounds of a language and mapping them, singly or in combination, onto designated graphic symbols. There are three types of mapping possible—three ways, that is, in which the sounds of language can be made to meet the eye. Every writing system in the world is one of these three types, or some combination of them.

  The first type, in which a written symbol stands for a whole word (or a whole concept), is called logographic (or ideographic) writing. Chinese writing, with its tens of thousands of characters, each signifying a different word of the language, is the best-known example of a logographic script.

  In the second type, a symbol stands for a single syllable, like ma or pa, bo or do, tam or kam. Examples of syllabic scripts, or syllabaries, as they are also called, include the Japanese kana script. (Japanese writing as a whole is a mixed script; besides syllabic characters, it includes a great many logograms, which like the kana are borrowed from Chinese script.)

  In the third type of writing system, symbols stand for individual sounds. This is an alphabet. We owe the alphabet to the Phoenicians, a Semitic people, who fashioned a letter-by-letter writing system from an earlier Semitic script in about 1000 B.C. As the Phoenicians conceived it, the alphabet had twenty-two characters—consonants only, no vowels. This alphabet was later taken up by the Greeks, who added vowel signs; from the Greeks it passed to the Etruscans and on to the Romans, who gave us the familiar alphabet used to write many Western languages, including English. The Phoenician script and its immediate descendants are the progenitors of nearly all the alphabets used round the world today, from Roman to Cyrillic to Hebrew and Arabic and, quite probably, many of the graceful curvilinear scripts of India.

  Whichever type it is, every writing system operates on the same basic principle, using individual symbols to represent one or more sounds of a language. In logographic systems, a character stands for the whole string of sounds known as a word. A syllabic system bites off smaller chunks, using a character to stand for perhaps two or three sounds. In an alphabetic system, the segment is smaller still, usually just a single sound, as with our l or m or t or v. In this way, every writing system functions as a kind of elementary encoding device, mediating between spoken sound and graphic symbol. But in order for the code to work properly, it must be transparent to all who would use it.

  To know the relationship between the sounds of a language and the written symbols that represent them—to hold the key to the code—is to be able to read that language. As long as there is someone alive who retains the key, the language can be read. But with the passage of time, the key can be lost. Now the link between sound and symbol is broken, and the text becomes as impenetrable as any secret code. That is where decipherment comes in.

  In his book The Story of Archaeological Decipherment, the scholar Maurice Pope elegantly describes the lure of ancient script:

  Decipherments are by far the most glamorous achievements of scholarship. There is a touch of magic about unknown writing, especially when it comes from the remote past, and a corresponding glory is bound to attach itself to the person who first solves its mystery. Moreover a decipherment is not just a mystery solved. It is also a key to further knowledge, opening a treasure-vault of history through which for countless centuries no human mind has wandered. Finally, it may be a dramatic personal triumph.

  When a reader confronts a text, the crucial question is this: What is known about the language of the text and what, correspondingly, is known about the script used to write it? For any written text, this relationship between language and script can take just one of four forms. Diagrammed, they make a tidy four-cell table:

  The upper-left-hand cell, which is blank, represents the most straightforward case. Here, a known language is written with a known script, as in the passage you are now reading, written in English by means of the Roman alphabet. No decipherment is required.

  The other three cases entail decipherments of varying difficulty. In Case I, in the upper-right-hand corner, an unknown script is used to write a known language. That is the case for Rongorongo, an undeciphered script of Easter Island discovered in the 1860s. Dating to the eighteenth century A.D. or earlier, the script was apparently used to write Rapa Nui, a Polynesian language still spoken on the island. But because Rongorongo fell into disuse (and because it lacks many of the internal clues, like word breaks, that help analysts tease a script apart), it is now impossible to tell which symbols in the script correspond to which sounds of the language.

  In Case II, on the lower left of the table, a known script is used to write an unknown language. That happened with Etruscan, an ancient non-Indo-European language of Italy that remains poorly understood to this day. The script used to write Etruscan is known: It derived from the Greek alphabet. As a result, it is still possible to read an Etruscan text aloud, giving each letter its familiar Greek sound-value. But to even the most knowledgeable linguist, the result sounds like gibberish. No one knows what most Etruscan words mean, or how Etruscan grammar worked. Read aloud, an Etruscan text is awash in sound but signifies practically nothing.

  Case III, on the lower right, is the toughest of all. Here, both language and script are unknown. This is the most inhospitable environment for decipherment possible, for it gives the decipherer no outside aid: no familiar script to help sound out the language, no familiar language to help sort the script. Such was the case for Linear B when Evans unearthed it in 1900. The script was a linguistic terra incognita with neither map nor compass at hand.

  IN EVANS’S DAY, the most famous archaeological decipherment in history was that of the Egyptian hieroglyphics. Unlike Linear B, the hieroglyphs of ancient Egypt did not have to be dug up and discovered: They were always there, and they always beguiled. Their decipherment, accomplished in the 1820s, would exert a considerable and ultimately harmful influence on Evans’s approach to Linear B.

  Developed in about 3000 B.C., Egyptian hieroglyphic writing was in use for more than three millennia. After that, with the spread of Christianity, the Egyptian language was written increasingly with the Coptic alphabet, a twenty-four-character script derived from the Greek alphabet. By about 400 A.D., the hieroglyphs had been abandoned entirely. Carved in stone, they would tantalize scholars for centuries to come.

  By the modern era, no one was even certain what language the ancient Egyptians had spoken. Whatever it was—possibly an ancestor of Coptic, an Afro-Asi
atic language later used in the region—it had long since been supplanted by Arabic. Faced with the decipherer’s worst-case scenario, an unknown script writing an unknown language, generations of scholars could do little more than speculate on what the hieroglyphs said.

  In 1799, a key appeared at last with the discovery of the Rosetta stone. The Napoleonic Wars were raging, and one of Napoleon’s campaigns had brought a troop of French soldiers to the Egyptian town of Rashid, known in the West as Rosetta. The soldiers were charged with dismantling an ancient wall there. As they did, they came upon a large black slab set into the wall. Weighing three-quarters of a ton, it was removed to Cairo for study and eventually made its way to the British Museum, where it resides today.

  There were three scripts on the stone. On the bottom was a passage in Greek, which could be read with ease. It was a decree from 196 B.C. describing, as the journalist Simon Singh wrote in The Code Book, “the benefits that the Pharaoh Ptolemy had bestowed upon the people of Egypt, and . . . the honors that the [Egyptian] priests had, in return, piled upon the pharaoh.”

  The other two scripts wrote the ancient Egyptian language. On top was a passage in the familiar ornate hieroglyphs. In the middle was a passage in the style of Egyptian writing known as demotic. Cursive, streamlined, and faster to write than hieroglyphics, the demotic script had been introduced in the seventh century B.C. Where the Egyptians had used hieroglyphs for religious, dedicatory, and other official inscriptions, they used demotic for everyday writing.