Who Invented the Alphabet:
The Semites or the Greeks?
I would make the startling suggestion that the alphabet was invented by a single human being, who created this remarkable technology to record the Greek hexameters of the poet we call Homer.
Certainly everyone agrees that the invention of the alphabet made possible the development of philosophy, science and democracy, some of the finest achievements in the history of human culture. But who invented the alphabet? Was it really the Semitic-speaking Phoenicians, as many of us learned in grammar school? Or was it actually the Greeks, to whom the Phoenicians supposedly passed it?
I dont believe the Phoenicians actually had an alphabet. The alphabet was a Greek invention. I would even make the startling suggestion that the alphabet was invented by a single human being, who created this remarkable technology to record the Greek hexameters of the poet we call Homer.
Writing itself was an extraordinary invention, let alone the alphabet. By preserving speech and thought, writing radically transforms every society that possesses it. Think about how improbable it was that someone should discover how to represent speech sounds graphically at all; it took hundreds of thousands of years. Yet it did happenfor the first time, as far as we know, in Sumer around 3400 B.C. But how did it happen?
Let us imagine one scenario. Suppose I am involved in a financial transaction, and as proof of purchase of sale I want to indicateor writemy name, Barry. Perhaps scorpions are a special interest of mine. I can therefore draw a scorpion to symbolize me. (One of the first Egyptian rulers in the late fourth millennium B.C. did exactly that to indicate his name, King Scorpion.) So now and henceforth the image of the scorpion will signify me and my name, Barry. Of course, nothing about this image tells you that the drawing of the scorpion indicates Barry or should be pronounced Barry. After all, scorpion is not Barry; and the scorpion is already referred to in speech by another sound (scorpios if you speek Greek, for example, or akrab if you speak Hebrew).
The seemingly simple discovery, however, that graphic marks can represent the sounds of human speechwhat scholars call the phonological principlewas a momentous achievement. Writing, to give a brief definition, is nothing but a system of graphic marks with a conventional reference. By conventional, I mean that we agree among ourselves, quite arbitrarily, that a certain sign will represent a specific sound, object or idea. This loose definition of writing is not restricted to the recording of speech, which we call lexigraphic writing; it also includes mathematical and musical notation, as well as the bewildering forest of symbols that decorates our computer screenswhat we call semasiographic writing. The discovery of the phonological principle made lexigraphic writing possible.
It is important to keep in mind that lexigraphic writing differs from speech. Writing is material; speech is not. Writing is potentially eternal; speech is evanescent. The elements of speech, in a way, come from nature. Our physical apparatus determines the sounds we can make. Graphic marks, however, have a history and, as we have already begun to see, are human and entirely arbitrary.
In short, speech is a human faculty; writing is a technology. Speech separates man from the animals; writing separates prehistoric man from historic man.
Lets go on with my name. To use the symbol of the scorpion to represent Barry is completely arbitrary. It is related only to the fact that I am fond of scorpions, for one reason or another. To put the matter a bit differently, the scorpion as a designation for Barry has no phonetic value. Can we represent my name with signs that do have phonetic value? The first part of my name, Bar, sounds like the word bear; so lets use the image of a bear to designate the sound bar. The second part of my name, reµ, sounds something like the word reedso well use the image of a reed to signify the sound reµ. Now I have two ways to write my name:
The first sign is called a logogram, or word sign, because the scorpion stands for me, Barry. The second two signs, the bear plus the reed, are syllabograms because they stand for syllables. The bear symbol and the reed symbol are assigned necessary sounds, easy to remember because the sound is what we use to refer to the objects suggested by the shapes of the symbols. This procedure, in writing, is called the rebus principle, a Latin word meaning from the things.
With the rebus principle, I have just created a relatively efficient system for bringing within the sphere of writing the enormous resources for symbolic expression that speech can offer. Note the difference between the arbitrary limited use of the logographic scorpion and the flexibility and variations offered by the syllabographic bear and reed. Chinese writing (from about 1500 B.C.), which is largely logographic and has tens of thousands of characters (an extremely inefficient system), never systematically incorporated the rebus principle, but it appears full-blown in Mesopotamian cuneiform, made up of wedges impressed in clay, and Egyptian hieroglyphics, which consists of stylized pictures. These and other forms of ancient writings are called logosyllabic because the writing is made up of logograms (signs that stand for words but have no necessary sound attached) and syllabograms (signs that stand for syllables).
Logosyllabic writing nonetheless requires hundreds of signs to convey meaning within the context of any particular language. Egyptian hieroglyphics uses nearly 750 different signs, for instance, and Mesopotamian cuneiform has a stock of about 600 signs. In actual practice, logosyllabic forms of writing are depressingly clumsy. (The Assyriologist Jean Bottéro called cuneiform a hellish script.) To us, they are interesting above all because they mark steps on the evolutionary path leading to the alphabet.
Structural change never takes place within a writing tradition. Someone standing outside a tradition, however, is in an objective position to make changes. Just this happened in the first half of the second millennium B.C., when Cretans in the Mediterranean and Semites in Syria/Palestine invented systems of writing that had no logograms, but were entirely phonetic. The earliest remnants of these experiments in the Levant are the Proto-Canaanite inscriptions from Palestine (about 1700 B.C.) and the Proto-Sinaitic inscriptions from the Egyptian desert (about 1550 B.C.). We cannot read these inscriptions (in spite of imaginative suggestions), but by about 1400 B.C. the prototype for the later Phoenician, Aramaic, Hebrew and Arabic scripts had come into existence, as proven by the discovery of signs of similar value, in a similar order, found on tablets at Ugarit in Syria; the Ugaritic signs were made by impressing a reed into clay and look superficially like Mesopotamian cuneiform. We call this script West Semitic, clearly attested by 1000 B.C.
The 22 or so signs of the West Semitic signaries (or sign systems) encode the consonants of Semitic speech remarkably well. By convention, we alphabet-users know that the West Semitic signary contains consonants, but no vowels. For example, the word for king in this signary is written with three signs whose consonantal values are M, L and K. But how does one pronounce MLK? Melek? Or Molek, Malik, Milok, Mulak, Melka, Muliki, Molka or Emlak? If you speak Phoenician (or some other West Semitic language, such as Hebrew), you will know (Malek); otherwise, you wont.
Nevertheless, this West Semitic signary is a graphic system of signs of remarkable simplicity and utility (see chart, below). For the first time, graphic signs were organized into a fixed order, and their values were cleverly encoded in a name attached to each sign. The fixed order and the attached names made the system easy to learn, unlike their complex logographic forebears.
This West Semitic system of signs is often called an alphabet. A book by the eminent epigrapher Joseph Naveh, Early History of the Alphabet (Brill, 1982), for example, is devoted only to West Semitic inscriptions. Even some hieroglyphic signs have been called alphabetic. Although some signs in Egyptian hieroglyphics stand for words (logograms) and unpronounced signifiers called determinatives (which identify the kind of word, such as a divine-name or place-name), other hieroglyphic signs seem to stand for one, two or three consonants. Scholars long ago noted that the 24 or so signs in Egyptian hieroglyphics that stand for a single consonant closely resemble the signs of West Semitic writing. For this reason, Sir Alan Gardiner in his still-standard Egyptian Grammar (Clarendon, 1927) described this system of signs as alphabetic. But Egyptians never viewed these signs as a group to be differentiated from other signs. They never used these signs to represent a continuous text.
So was Egyptian writing really partly alphabetic? And does the West Semitic system of what we loosely call consonants in fact constitute an alphabet? Or are these descriptions biased by our own prejudgments as alphabet-users?
In a seminal book, A Study of Writing
(University of Chicago, 1952), University of Chicago Assyriologist I.J. Gelb argued that the traditional term alphabet is inappropriate for describing West Semitic writing (and the 24 Egyptian uniconsonantal signs as well).1
The first and only alphabet was the Greek alphabet, he wrote, in which the signs designate phonemes.
Phonemes are the atoms of spoken language, the smallest parts in the stream of sound that distinguish one utterance from another. They are to be distinguished from whole parts of speech, such as words, or parts of words, such as syllables. The word Port differs from Fort, for example, in that it begins with a different sound. The alphabetic signs P and F, that is, represent different phonemes.
For Gelb and those, like me, who follow him, the uniconsonantal signs in Egyptian hieroglyphics and in the West Semitic sign system are not really letters of an alphabet; they are in fact syllabic signs consisting of a consonant and a variable vowel.
To explain: Try to pronounce the letter T. You probably said tuh, or something like that. Tuh, however, is not the sound of T; it is the sound of T plus a vowel appended to the sound of T. Evidently one cannot pronounce the sign T by itself, for it must be taken in combination with a vowel. In real alphabetic writing, on the other hand, the signs represent phonemes and not syllables or words. Alphabetic writing goes inside the structure of the spoken language. That, at least, is the theory of it.
According to Gelb, West Semitic signs only appear to represent phonemes. Really they are syllabic signs that imply but do not specify vowels. Thus each sign represents a multiplicity of possible values. In Phoenician, K may in theory have the value ka, ke, ki, ko, ku or, at the end of words, k. In the Greek alphabet, by contrast,K has only the value k (which you cannot pronounce by itself). The two systems of signs function in completely different ways, and if the Greek alphabet is indeed an alphabet, as everyone agrees, then West Semitic is something else.
Gelbs study of the history of writing, which considered internal structure rather than superficial resemblances of form, has been enormously influential. Following his lead, many scholars now speak of the West Semitic syllabary. Those who balk at the term but agree that West Semitic writing functions differently than the Greek alphabet does, like to call the West Semitic characters consonantal signs, implying that each West Semitic sign does stand for a phoneme. But since phonemes do not exist as real segments in speech, this is not very likely. West Semitic writing, then, was a great inventionone had to learn only about 22 signs to record speech in writingbut it was not based on an alphabet. Only a native speaker could resolve the enormous phonetic ambiguities that arose in reading Phoenician, Edomite, Hebrew or any other West Semitic language. Greek alphabetic writing, on the other hand, can be pronounced without the readers knowing what the sounds mean. That, as we shall see, is an important key to identifying the historical birth of this extraordinary invention we call the alphabet.
But first we should ask whether this invention was a unique event or something that happened repeatedly. In short, did the shift from West Semitic writing to alphabetic writing take place once, or many times?
I believe that the Greek alphabet was invented by a single person at a single time. The evidence: Every surviving example of early Greek alphabetic writing contains the same arbitrary structural alterations made to the West Semitic characters that served as the model for the Greek alphabet. This suggests a single source, for if there were multiple sources we would expect to find many different alterations.
For convenience, I call this supreme inventor of the Greek alphabet the Adapter. The Adapter chose five signs from the West Semitic syllabary to use as vowel sounds, as reflected in every early Greek alphabetic inscription. Both the number of signs (five) and the particular signs chosen are arbitrary. Ancient Greek has many more than five vowel sounds; indeed, in later Greek inscriptions, seven vowel signs are employed, and there could have been more.2
The Adapters assignment of vocalic (that is, vowel) qualities to five West Semitic signs often makes phonetic sense. For example, imagine that a Greek had learned from a Phoenician trader that the first sound of a sign name in the Phoenician West Semitic syllabary encodes the signs phonetic value: The sign name beth encodes the consonant B plus a variable vowel. Some West Semitic consonantal sounds do not exist in Greek, however, so the Adapter heard a vowel instead of the unfamiliar Phoenician consonant. For example, the West Semitic signary begins with aleph, a consonantal constriction of the throat (a so-called glottal stop.) When the Adapter heard this pronounced, he may well have perceived it as the vowel A and thus used this sign to signify that vowel. Some vowel signs chosen by the Adapter, however, are completely arbitrary. Thus, the West Semitic ayin, which stood for a similar kind of consonantal constriction of the throat, unknown in Indo-European languages, was chosen as the sign for the vowel O.
The Adapter also added three signs of his own invention, the so-called supplementals F C Y, which seem to have represented consonants accompanied by a strong gust of air (called aspiration).
It may even be possible to pinpoint when and where this alphabetic revolution occurred. The earliest examples of Greek alphabetic writing date from about 775 to 750 B.C. Based on the theory that so useful a technology would very quickly begin to leave traces of its existence, it is reasonable to suggest that the Adapter probably worked in the period just preceding the earliest surviving examples.
Almost all early alphabetic Greek inscriptions seem to have some relationship to Euboea, an island in the Aegean, located just east of mainland Greece. Euboea contains an extraordinary archaeological site called Lefkandi, where ambitious architecture and rich burials testify to its wealth and importance at a time when the rest of Greece was poor. Eastern imports prove that Lefkandi maintained relations with the East throughout the dark period in Greece, between the end of the Bronze Age (c. 1200 B.C.) and the age of Homer (c. 800 B.C.). Several of the earliest alphabetic Greek inscriptions come from this site.
Euboean traders almost certainly ventured to the west and to the east, probably carrying the alphabet with them. Euboean pottery has been found on Cyprus, where there was also a permanent Phoenician settlement (at Kition). Here Euboeans may have seen the Phoenician West Semitic syllabary, which served as the model for the Adapter.3
The Euboeans also had trading quarters at a Phoenician site now called Al Mina, on the Mediterranean coast of Syria.
In the West, the first Greek colony was established around 800 B.C. on the island of Pithekoussai, modern Ischia, in the Bay of Naples off Italys west coast. The Lefkandians from Euboea, seeking local iron ores, may well have participated in the establishment of Pithekoussai. As with the Lefkandian site, some of the earliest alphabetic Greek inscriptions come from the Pithekoussai cemetery.
In Latium, just north of Pithekoussai in the vicinity of Rome, what may well be the earliest dated alphabetic Greek inscription was recently unearthed.4
Because it was excavated stratigraphically, it can be quite securely dated to about 775 B.C. The inscription, part of a word or name scratched on the side of a one-handled flask, consists of only five letters, but it is evidence of how quickly the new invention spread: It probably spread to Latium from settlers at the new colony on Pithekoussai.
If the East supplied a script that could be adapted for writing down Greek, the West supplied the earliest alphabetic Greek inscriptionsand Euboean traders traveled to both places. It therefore seems likely that the Adapter moved in this Euboean economic and social ambit.
Should we conclude that the invention was based on the commercial needs of this society? I dont think so. There is sufficient evidence, especially from two early inscriptions, to show that the inspiration was poeticmore specifically, Homeric. The Greek alphabet was invented because of the desire to record epic verse.
In 1871, a short inscription was found in Athens, which always maintained close relations with neighboring Euboea. The inscription was scratched from right to left along the shoulder of an Attic jug apparently given out as a prize in an athletic contest. It is called the Dipylon jug because it was found near the citys Dipylon Gate. The inscription dates to about 740 B.C. and consists of a line of perfect Greek hexametera
and 11 additional signs that are difficult to interpret: Whoever of all the dancers now dances most friskily
Evidently, the second line completed the thoughthe will get this prize, or something to that effect.
A second inscription takes us back to Pithekoussai off the west coast of Italy. This inscription is contemporaneous with the bit of poetry on the Dipylon jug. But it is scratched in three lines on a clay cup:
(1) I am the delicious cup of Nestor.
(2) Whoever drinks from this cup, straightway that man
(3) the desire of beautiful-crowned Aphrodite will seize.
The inscription seems to be a humorous relic from a drinking-party game called skolion (zigzag): In playing the game, a partygoer would compose a cliff-hanger line of poetry, then pass a myrtle branch zigzag to another guest, who would then be challenged to cap or finish the thought. The first line of the inscription, I am the delicious cup of Nestor, parodies a practice of scratching ones name on a cup. In this inscription, the cup is supposed to belong to the famous warrior Nestor, whose cup of gold is described in the Iliad (XI.6327). Here, however, it is a humble clay cup. The second player, to whom the cup is passed, not only meets the challenge but goes one better (line 2) by parodying in dactylic hexameter another kind of cup inscription, in which something like Whoever steals me will go blind is scratched on the cup. But again the thought is not complete. So the cup is passed to a third player, who adds line 3, predicting in hexameter a pleasant sexual experience.
If I am right in this analysis, the Euboeans at Pithekoussai were familiar enough with the Iliad in the eighth century B.C. to parody it and expect someone to laugh. In this Cup of Nestor inscription we have not only one of the oldest examples of alphabetic writing, but Europes first literary allusion.
We can only be amazed at the sophisticated level of expression in the oldest Greek inscriptions, from a time when we would expect simple expressionunless alphabetic writing was invented specifically to record epic song. The Greek alphabet, as we shall see, was the first technology capable of doing this. On this basis I, and now others, have connected the invention of the alphabet with a seemingly unrelated problem, the Homeric Question.5
Recent research into poetic oral composition among nonliterate peoples has established powerful reasons for believing that Homer was an oral poet, an unlettered bard who composed his songs in performance, using a special rhythmical language. Oral poets like Homer were entertainers whose spontaneous songs, ever alive on their lips, were variations on familiar stories and events. As a result, these poems were never sung twice in the same way. There was no standard text, no correct version to preserve.
That we have a fixed text for the Iliad and the Odyssey means that, although Homer was an oral poet, he must have lived at a time when alphabetic writing was available. Was the person I have been calling the Adapter the same figure who wrote down Homers poems, in an age when writing was so novel that the bard would have been unaware of its mechanics?
Both archaeological and internal evidence suggest that Homer lived and sang in the period between 800 and 750 B.C., the same time as the invention of the Greek alphabet (about 800 B.C.). Why did the first writing system capable of recording Homers complex rhythms appear just at the moment when they were in fact recorded? Was it just a coincidence?
Another fact: The earliest known war in Greece was fought on a windy plain on the island of Euboea, perhaps in the ninth or eighth century B.C. One of Homers songs tells of warfare on a windy plain. Another tells of a man, Odysseus, who returns from that war across trackless seas to his native Ithaca, a port located between Italy and Euboea. Was this, too, merely coincidental? Perhaps. But now lets take a look at the internal evidence.
It is possible to notate ordinary Greek in a syllabaryit was done twice, in Cretan Linear B script and in the related Cypriote syllabary. It is also done in the Hebrew press in Israel to this day. But you have to be a Hebrew speaker or a Greek speaker to know how to pronounce the words. Without the vowels, it is impossible to know. While this may have been adequate for ordinary Greek, spoken by Greeks, it is not possible to recreate the sounds of a Greek hexameter poem from a syllabic notation wherein signs included only indeterminate vowels. There are no native speakers of Greek hexameters, which are made up of a hodgepodge of grammatical and dialectic forms that do not occur in ordinary spoken Greek: These curious forms seem to have entered the tradition in order to facilitate the hexametric rhythm.
To record the Greek hexameters of Homer, a system of writing was needed that did more than simply remind a native speaker of words in a language he already understood. The new sign system we call an alphabet told the Greek speakerand perhaps the non-Greek speaker as wellhow to pronounce the words in lines of hexameter unpronounceable from signs that did not specify vowels.
My guess is that the Adapter was a prosperous Euboean engaged in international trade, wealthy enough to buy plenty of expensive papyrus and engage the cooperation of a great poet. Homers poems as recorded are so long and so complex that they could not have been part of the poets ordinary oral repertoire. Homer probably dictated these poems slowly over a period of weeks, thus making them a collaborative effort between a master of an ancient tradition of oral verse-making (Homer) and a secretary capable of recording hexameters in a new technology.
I realize that this explanation flies in the face of expectations that writing always served economic interests, and that alphabetic writing was bound to appear eventually. But these expectations are rooted in our own biases as long-time users of the alphabet. It is the invention of the alphabet that has made us who we are.
Alphabets and Scripts