3 THE ORIGINS OF ALPHABETIC WRITING

Transcription

1 3 THE ORIGINS OF ALPHABETIC WRITING What abides through the transformations of letter shapes and their abstraction from their visible form, as described in Chapter 2, is the set of 26 letters that now constitute the English alphabet. Although that set has now been constant within living memory, and most of the letters have been constant for over 2,000 years, at least in their capital or upper-case forms, there is nothing immutable about the selection of signs involved. Their accidental, arbitrary character has already been remarked on, but the way in which they evolved is nevertheless important for the light it sheds on how the spelling of English functions today. Their evolution also provides, incidentally, a revealing commentary on the course of human history and civilization, especially as manifested in their earliest periods in the region of the Mediterranean and south-west Asia. The first writing Forms of writing were devised independently in several parts of the world, but earliest in south-west Asia (the area traditionally described, from a European perspective, as the Near East) and Egypt, and later in China and in pre-columbian America. The purpose of writing was to permit language (thoughts, utterances, etc) to be recorded beyond the capacity of the individual memory to retain it. The writing tradition and culture to which Modern English is heir can be traced back, 1

2 across quite radical breaks at several points, over 5,000 years, to that earliest writing in the south-west Asia. Writing developed only gradually from earlier forms of graphic representation, such as the purely pictorial depiction of things and simple tokens or marks used for counting. The gradualness of this process would make it impossible to pinpoint a precise beginning for writing, even if full evidence were available. It is, however, clear that the earliest forms of writing, covering well over the first 1,000 years of that south-west Asian tradition, were not alphabetic, if we define an alphabet as a small set of simple signs designed to systematically represent the sounds of a spoken language. Despite the greater simplicity of its underlying concept, the alphabet was intellectually more advanced than its precursors, and built on accumulated experience of more complex, yet less systematic, scripts. The broad setting for the invention of the alphabet was the flowering of early civilizations in the Fertile Crescent that stretched from the Persian Gulf in the east, westwards to the eastern Mediterranean, and thence southwards to the lower Nile in Egypt. The timespan of the main writing traditions of those civilizations can be roughly delimited as the period from around 3000BC to the beginning of the Christian era. In that period and that region, two very different scripts arose, the splendour of whose remains has assured them an impressive place in the annals of humankind. Cuneiform, written by impressing wedge-shaped marks (Latin cuneus wedge ) into tablets of soft clay, is generally thought to be the first of these scripts, 2

3 both because its oldest surviving examples 1, showing its evolution from primitive counting tokens, are of a somewhat earlier date, and because some scholars believe it may have influenced the invention of the second. The use of cuneiform centred on Mesopotamia, between the lower course of the rivers Tigris and Euphrates in what is today Iraq, where alluvial clay is abundant; but cuneiform writing came to be adopted over a much wider area. The second of those two early scripts was Egyptian hieroglyphics (from Greek for sacred carvings ), which is famous for the formalized pictorial quality of its characters. It was typically written, or drawn, with reed brush-pen and ink on scrolls of paper-like papyrus, whose raw material, the pith of reeds, was readily provided by the delta of the Nile. For all the differences between their graphic forms, both cuneiform and hieroglyphics are also famed for their monumental inscriptions on stone, and there are some striking similarities in their inner structure as writing systems. The political and military power of the kingdoms that used these writing systems ensured that they remained supreme in their regions of origin at least until the conquests of Alexander around 330BC and the subsequent dominance of the alphabet, whether in its Aramaic, Greek or Roman version. Viewed from the inevitably foreshortened perspective of the non-specialist modern observer, the history of the Fertile Crescent appears a continuing, fluctuating struggle for supremacy between kingdoms and peoples, involving at 1 In his draft of this chapter, Christopher Upward noted various places where illustrations would be helpful. However, since this material is now not being presented in a book but on an online website, most of these illustrations are not being included. For those who wish further information about cuneiform writing or hieroglyphics or other writing systems mentioned in this chapter, much more information and many more illustrations are available online than could be included here. 3

4 various times (to list only the best known) Egyptians in the south-west, Hittites and Greeks in the north-west, Medes and Persians in the north-east, and Babylonians and Assyrians in the centre. Their civilizations rose, fell and sometimes rose again, often in direct contact with each other through wars and peace treaties, migrations and deportations, and sometimes subjugation, enslavement and destruction. The Judaeo-Christian Old Testament testifies vividly to the experiences of a minor Semitic people, the Hebrews, tossed between the great powers of those days and anxious above all to ensure its own survival. According to the Book of Genesis, their patriarch, Abram (his name later changed to Abraham see Genesis 17.5) hailed from Ur ( Ur of the Chaldees ) in Mesopotamia, where cuneiform originated, but to escape famine he travelled to Egypt, the homeland of hieroglyphics. His migrations, if the stories are to be taken as having a genuine historical basis, coincided in both place and time (c. 1700BC) more or less with the oldest samples of alphabetic writing yet discovered. His descendants are said both to have benefited (under Joseph) from the rule of the Egyptian pharaohs and suffered from it (leading to the exodus under the leadership of Moses, whose tablets 2 are the first reference to writing in the Bible), while later generations of Hebrews were taken into captivity back in the land of cuneiform, Mesopotamian Babylon, in the early 6th century BC. Pictures, sounds and meanings in Egyptian hieroglyphics Although many of the letter forms of the alphabet had their source in Egyptian 2 According to Exodus 24.12, God wrote the laws he gave to the Israelites on two stone tablets. 4

5 hieroglyphics, the alphabet as such originated in the zone between the heartlands of hieroglyphics and cuneiform. One may well surmise that contact with the two ancient systems is likely to have encouraged innovation. Writing systems in general are commonly observed to have originated in the drawing of pictures to represent visible objects. This was true, for instance, of both cuneiform and Chinese characters. The pictorial element in the outward form of classic Egyptian hieroglyphics is visually perhaps more powerful than in any other writing system, and large numbers of its signs can be recognized today, though simplified and stylized, as depicting people, animals, birds, plants, buildings, furniture, tools, weapons and actions. A truly pictographic writing system needs a very large number of signs (in principle one for every concept to be written down). The number in general use at any one time in Egyptian hieroglyphics was between 500 and 1,000, although over 6,000 have been listed from texts of all periods. Nevertheless, to classify hieroglyphics just as a pictographic writing system is seriously misleading, and many scholars firmly describe it as actually phonographic (i.e. having signs standing for sounds rather than for the objects they depict). Even its earliest surviving examples (from around 3000BC) show it already equipped to indicate major features of pronunciation. Furthermore, because the classic picture signs had to be drawn with such skill and care, for practical everyday purposes more convenient versions were created which show that the pictorial aspect was not essential to the system at all. The Egyptian language at an early stage developed a more cursive, simplified equivalent to hieroglyphics proper, known as hieratic, while much later there arose the radically different demotic, until finally, especially after the arrival of Christianity in Egypt, an alphabetic 5

6 system, Coptic, was introduced, which used mainly Greek letters. It was indeed the discovery of a trilingual inscription, the Rosetta stone, using three of different writing systems (Greek, hieroglyphics and demotic), that allowed the essential breakthrough to the decipherment of hieroglyphics early in the 19th century. Hieroglyphic signs could have various functions. Some were actually used in a pictographic mode to represent the object they depicted, but the phonographic mode, by which the same signs could also stand for sounds, was more important in the representation of words. Then there was a third mode, by which the signs could serve to indicate the meaning (they are then called determinatives ) or pronunciation (in which case they are called phonetic complements ) of another, usually preceding, sign. The function of such determinatives and phonetic complements is of great potential importance for any writing system whose signs contain a significant degree of ambiguity, as can be illustrated from the spelling of English. Thus the second A in stationary has a function like that of a determinative, telling the informed reader that the word has a different meaning from stationery with E; and the H after C in chap has the function of a phonetic complement, showing that the C does not have the same sound-value as in cap. Some hieroglyphic signs could be used for all three of these very different functions, a fact that superficially, at least made for a high level of ambiguity in the system and immensely complicated the task of its 19th century decipherers. For literate native speakers of Old Egyptian, however, the degree of ambiguity would have been less, as the different modes of representation could serve to disambiguate known homophones (i.e. different words with the same pronunciation) and the native speakers much greater familiarity with the context of what was written and of the 6

7 language itself would have further assisted their reading and comprehension. The rebus principle, with consonants only The way in which hieroglyphic signs could represent sounds as well as depict objects was the starting point for the invention of the alphabet. They did so by means of the rebus principle. A rebus is typically a picture-sign that depicts an object, but is specially used to represent the sound of the object s name rather than the object itself. Thus, using an imaginary example from English, pictures of a bee and a leaf could stand for the two syllables heard in the word belief (which would then be a double rebus), rather than suggesting a meaning to do with bees and leaves; an authentic example in English is the group of letters IOU, which we interpret not as single letters (as we would read ABC), nor as a complex vowel sound, but as the syllables of the words I owe you. So in hieroglyphics a sign showing a bird, the swallow, whose name contained the consonants wr, could also serve to represent the same consonants occurring in a quite different word, for instance in wr meaning big. For the picture of a swallow to be used both for the meaning swallow and for big was, of course, ambiguous, and so another sign would be placed after it as a determinative or as a phonetic complement, to indicate how the sign should be interpreted. A characteristic of Egyptian hieroglyphics is that the script was normally written without indicating vowels. Thus it may be that wr meaning big was pronounced with different vowels from wr meaning swallow. Literate speakers of Ancient Egyptian obviously pronounced the words they read complete with the appropriate vowel sounds which they knew from speech, but modern scholars still 7

8 have only a partial idea of what those vowels may have been. Furthermore, the signs as used to represent consonant sounds were of three types. About four dozen stood for a string of three consonants and are known as triconsonantals; for example, a single sign stood for the string nfr. About twice that number of signs stood for two consonants, and are known biconsonantals, like wr. And twenty-six stood for a single consonant (uniconsonantals). There is clearly a certain parallel between uniconsonantals and alphabetic consonant letters, but there are also important differences. Firstly, the uniconsonantals would necessarily have implied a vowel as well. Secondly, there were typically several possible ways to write the sounds of a word using hieroglyphics, whereas the alphabet was originally designed to give each sound a unique written form. That is why an alphabet can often make do with just 20 to 30 letters, while Egyptian hieroglyphics used hundreds of signs. The uniconsonantal signs worked by the so-called acrophonic principle, which means that the consonant sound they stood for was the initial sound of the word whose picture they also represented. So a pictogram showing a human mouth firstly stood for the meaning mouth, which was pronounced /q/ followed by some vowel or other; but secondly, it was also used to spell the consonant /q/ occurring in quite different words. Scholars disagree as to whether the uniconsonantals constitute an Egyptian alphabet, it being argued against this view that since each uniconsonantal was pronounced with a following vowel, the sign should be construed as syllabic rather than purely consonantal. Such differences of view appear to be at least partly a matter of semantics, that is, their protagonists do not disagree about the nature of the system itself, but only about how to describe it. 8

9 Picture-signs become letters The power of Egypt under the pharaohs meant that other peoples of the region came to know the hieroglyphic writing system. In the 2nd millennium BC, Semitic peoples in Sinai, Palestine and further north in Syria began borrowing and adapting the device of uniconsonantal acrophonic signs for their own languages, in the process developing what can be considered the first real alphabet. The evidence for how this happened is fragmentary, and scholars differ as to the correct way to interpret it, but one example will suffice at this stage to show the kind of process that appears to have been involved. A common hieroglyphic sign stood for a house, representing it pictographically by the rectangular ground-plan of its walls with a gap on the lower side for a doorway. The sign was a biconsonantal, as the Egyptian word for house contained the consonants /o/ and /q/ (plus a vowel or two), and the same sign could stand for the verb to go since it contained the same consonants. What is perhaps the earliest example of alphabetic writing yet discovered, found in Sinai and dated to around 1700BC, used a sign of similar shape, but gave it the sole function of representing a single consonant, in this case /a/ (as deduced from decipherment of the initial letter of the name of the goddess Ba alat 3 ), which bore no phonetic relation to the Egyptian /o/ and /q/ at all. The reason why the sign was allocated the new sound-value /a/ was that the Semitic root for house begins with that sound. So what the creators of this new system had done was to take a hieroglyphic sign and its meaning from Egyptian, but apply the acrophonic 3 For more on the decipherment of the name of Ba alat, see Florian Coulmas The Writing Systems of the World (Oxford 1989: Basil Blackwell), pp

10 principle to a differently pronounced word having the same meaning in their own language, and use the sign as a standard representation for its initial consonant sound. The sign itself no longer conveyed the meaning house, nor could it be used for any other purpose than to indicate the one sound /a/. Analysis of the other signs used in texts found in Sinai (reinforced by later finds in places further north) showed in several cases a parallel procedure: a sign originating in hieroglyphics was reapplied to stand for the initial consonant sound of the equivalent Semitic, not Egyptian, word but the sign now stood only for that sound, and no longer signified the object depicted. A link between the sign and the object it had originally depicted was nevertheless retained, in that the letter was named after the object. This link has persisted in Hebrew to the present day: the name for the second letter of the Hebrew alphabet, beth, still means house (as in the place-names Bethel House of God, Bethlehem House of Bread ). The sign was thus no longer a pictogram, but solely a phonogram, and standing as it did consistently for a single consonant sound, we can for the first time call it not just a sign, but a letter in the true alphabetic sense. In these Proto-Sinaitic letters, the alphabetic principle of unambiguous sound-symbol correspondence for single consonants had been established. Prototype letter-shapes The alphabet used in the texts found in Sinai and further north included the first known prototypes at least for the English letters A, B, C, D, E, F, H, K, L, M, N, O, R, S, T, though all would undergo radical changes before beginning to resemble their modern forms. Some of the sound-values have also changed radically, but others 10

11 have remained constant ever since. Pictographic origins, letter-names and acrophonic sound-values have been identified as follows 4, though not all are accepted by all scholars. The letter A derived from a hieroglyphic pictogram for an ox 5, depicting a roughly triangular head with horns pointing up. The Proto-Sinaitic name for an ox was alpu, which therefore also became the name of the letter. In Semitic languages this letter generally represented, as it still does in Hebrew and Arabic, the sound of a glottal stop (/>/), which is regarded as a consonant and conventionally transliterated (i.e. when Arabic or Hebrew words are spelt in the Roman alphabet) by an apostrophe, as in alpu, or in Qur an, the name of the Islamic sacred text (also spelt Koran). The prototype for B, as described above, depicted a house (bētu) and stood for /a/, though it has been suggested by Martin Bernal 6 that the letter-shape B perhaps did not derive from Semitic bet at all. English C originated as a sign standing for a /f/ sound. That sign appears first to have depicted a bent stick, perhaps something like a boomerang, but later took the name gimel, meaning camel. English D originated in a pictogram of a fish, and represented the sound /c/; it became simplified to the shape of a triangle in Phoenician, with the name daleth. E (lying on its back) originated in the figure of a man with uplifted arms and 4 W. V. Davies Egyptian Hieroglyphics, in J. T. Hooker (ed.) Reading the Past (London 1990: British Museum Publications) pp Here again, any reader who wants further information about Proto-Sinaitic letterforms is advised to perform an online search for the various websites where such information is readily available. 6 Cadmean Letters (Winona Lake 1990: Eisenbrauns), pp

12 was pronounced as a breathier [g]-like consonant. English F originated as a letter pronounced like /v/ (which by various routes ultimately also gave rise to the letters U, V, W, Y), and seems first to have depicted some kind of hook or peg, called wawwu. The other letters, except O, are presumed to have been pronounced roughly as in English. H: on its side with two or three vertical bars, perhaps representing a fence. K: the palm of a hand with fingers, pronounced kappu. L: a vertical line with a hook at the top, depicting a goad or crook, called lamdu. M: an extended zigzag, representing rippling water and called perhaps mayyuma. N: a shorter zigzag, standing for a snake, called nahashu. O: originated in hieroglyphics as a vivid almond-shaped eye with pupil, but slightly simplified in Proto-Sinaitic; it too was a consonant, a pharyngeal fricative or a variety of glottal stop transliterated in Roman by a reversed apostrophe, as in the letter name ēnu. R: the profile of a human head, called ra shu in Proto-Sinaitic. S: originally lying on its back like a cursive W, apparently deriving from a hieroglyphic sign for a lotus pool, with plants growing vertically from the surface of the water; but its later Semitic name meant tooth. T: a cross, derived from a hieroglyphic sign showing crossed planks; in Proto- Sinaitic it was called tawwu, which meant a mark. 12

13 From prototype to Roman: the tortuous development of B So fragmentary is the evidence for the development of the alphabet through the next millennium that not merely its details but recently even its main features have been subject to conflicting interpretations. Through most of the 20th century, the generally accepted (but here necessarily much simplified) account suggested a linear transmission of the alphabet, from Egyptian hieroglyphics to Proto-Sinaitic, from Proto-Sinaitic to Phoenician, from Phoenician to Greek, from Greek to Etruscan, and from there finally reaching its destination (from our point of view), which was Roman. There are, of course, no teleological implications intended in such a perspective: not merely would the idea of the Roman alphabet as the preordained, supreme achievement of the whole process be philosophically questionable indeed, as will be repeatedly shown in the History, the Roman alphabet is not a perfect creation but a far-reaching re-evaluation of the ancient inscriptional evidence by Martin Bernal in his already mentioned Cadmean Letters suggests a more complex model. Instead of a single line of development from one stage of civilization to the next higher stage, he proposes successive waves of transmission in all directions from the Proto-Sinaitic source over the best part of a thousand years, with intermediate currents and eddies further complicating the picture. The case of the letter B illustrates the greater complexity of Bernal s account. We start with the original rectangular Egyptian hieroglyph representing a house, /o-q/: 13

14 This hieroglyphic was broader than it was high, the gap in the lower side representing the doorway in the ground-plan of a house. As we have seen, this hieroglyphic sign did not have the same sound-value or simple function as the alphabetic letter that arose from it, standing as it did both for the word house and for the sound sequence /o-q/ in different words. The hieroglyphic shape underwent a slight adjustment in Proto-Sinaitic, presumably making it easier to write (if less symmetrically pleasing): We may note that whereas the hieroglyphic rectangle with its central doorway required five strokes to write it, the Proto-Sinaitic form required only four. But, as already described, the key change lay in the function of the symbol: it now represented only the sound /a/, as the initial consonant of the Semitic word for house, which was bet. We may speculate that greater ease of writing may similarly account for the further adaptation the shape underwent by the time it reached Phoenician (c BC): Here the four strokes were reduced to three, the letter being turned clockwise through ninety degrees and the line-end for the upper door-post bent inward until it touched the back wall of the house ; the changed angle and curved upright 14

15 would also seem more convenient for cursive script. It is when we come to the Greek form of the letter that Bernal s account differs significantly from the traditional one. According to the latter, Greek adopted most of the Phoenician alphabet around the 8th century BC, or perhaps slightly earlier, after centuries of illiteracy (at least since the demise of the Mycenaean Linear B script around 1200BC), but somehow, shortly after the Greeks adopted the alphabet, they are supposed to have invented several new letters, and different Greek dialects further to have found themselves using different sets of letters, with some letters standing for different sounds in different dialects. For Bernal this account is too problematic to be convincing, and for him only a much longer timespan can explain the state of the Greek alphabet as found in the earliest surviving inscriptions in the various dialects. The shape of the letter B, so dissimilar to its alleged Phoenician antecedent provides part of his evidence. Early surviving inscriptions (none older than the 8th century BC) show a wide variety of forms, some based on the Phoenician, others derived from the voiceless equivalent of B which was pi, but, in areas with the oldest writing traditions, also the form B itself or its mirror-image with the bowls to the left of the upright. This letter-shape, he suggests, derives from a Semitic M, which accords with what is known of the phonetic development of the sound /a/ in archaic Greek. Bernal is thus suggesting that at least elements of the Greek alphabet can be traced back centuries earlier than the previously presumed date of borrowing from Phoenician around the 8th century BC, indeed back beyond 1400BC. But there must have been some later re-organization of the Greek alphabet, since, even if the letter-shape B does not derive from Phoenician bet, all 15

16 its other aspects (its name beta, its position as the second letter in the alphabet, and its sound-value /a/) do correspond to the Phoenician alphabetic tradition. We cannot here pursue Bernal s interpretation of the origins of the other letters of the Greek alphabet, but it should at least be made clear that he uses the same techniques of the widest possible analysis of early alphabetic forms across the ancient Mediterranean and south-west Asian world, combined with analysis of linguistic data drawn also from the widest possible range of sources, to question several other long-held beliefs about the origins of the Greek alphabet. He shows for instance that the Greeks could well have derived all but one of their vowel letters and all of their supposed new consonant letters from existing sources, and that there are no grounds for attributing the invention of those letters to some vague concept of Greek genius. In the light of Bernal s theories, the following account of the further development of the alphabet in the hands of the Greeks, Etruscans and Romans will attempt to tone down the previously accepted view of a straightforward linear transmission from one civilization to the next. The final stage in the evolution of modern Greco-Roman B involved its final horizontal reversal from a form with two bowls on the left and a vertical line on the right into its mirror image B. The reason for this reversal, which affected many letters of the Greek alphabet, will be explained later in this chapter. With the shape B, a form had been achieved which was to survive for 2,500 years through Classical Greek and Latin down to present-day English. The Cyrillic (Russian) version of the letter is somewhat reminiscent of earlier forms, the capital Б being like an inverted, squared-off version of the Phoenician letter; whether or not 16

17 Phoenician actually inspired Cyrillic here, an alternative to B was required in order to distinguish the Cyrillic sign for /a/ from the sign for /u/, which was written B (Greek B had come to be pronounced /u/ by the time the Cyrillic alphabet was devised). Thus we see a single sign in one system splitting into two separate signs in another system (further examples of such letter-splitting are seen at a later date in English I/J and U/V). Modern B bears no resemblance to the original hieroglyphic ground-plan of a house, and indeed if Bernal is correct there was in fact no continuity of letter-shape. But what has remained constant ever since the Proto- Sinaitic form (other than in modern Greek and Russian) is the alphabetic soundvalue /a/, and its position in the alphabet. Other Proto-Sinaitic letters were subject to similar metamorphoses, as they evolved through the stages from hieroglyphics to the English alphabet. Experimenting with alphabetic writing Sinai was not the only location where examples of early alphabetic writing have been found from the mid-2nd millennium BC. Further north, up the eastern littoral of the Mediterranean, several other fragmentary finds (Proto-Canaanite) demonstrate that related Semitic peoples were also experimenting with ways to write their related languages using alphabetic letters. Their attempts showed sufficient similarity to the Proto-Sinaitic letters to indicate at least a parallel pedigree, and perhaps even a direct descent. Over the ensuing centuries it is clear that an evolution and consolidation took place, until by 1000BC the Phoenician alphabet appears in a reasonably full, mature form, which it has been possible to 17

18 decipher much more fully than the Proto-Sinaitic and Proto-Canaanite fragments. These early alphabetic scripts are referred to collectively as West (or sometimes North) Semitic, as opposed to the East Semitic cuneiform writing of Mesopotamia. One of these alphabetic scripts is of particular interest, although it took the alphabet no further. Ugaritic dated from around the 14th to 12th centuries BC, and might be described as a mongrel system, as it used the cuneiform writing technique of impressing configurations of wedge-shaped marks in clay, but at the same time it had at most 30 signs, which were not used as cuneiform signs otherwise normally were but mostly corresponded to the letters of the emergent alphabet. Indeed, with a little imagination, one may even detect some resemblance between the first alphabetic letters and various Ugaritic signs. Not merely has decipherment shown common sound-values between most Ugaritic signs and letters of the alphabet, but a surviving listing of the signs orders them in more or less the same sequence as in the letter-alphabet. The Ugaritic script included extra letters specific to its phonological needs, over and above those of the Proto-Canaanite alphabets, and it has been suggested that Ugaritic may therefore foreshadow the slightly longer Arabic alphabet of modern times (though it is not a direct ancestor). At all events, Ugaritic testifies to the intensity of scriptorial experiment being conducted at that time in the West Semitic speech communities. The branching alphabetic tree In the following section the Phoenician writing system will be looked at more carefully as the major next step in the line of descent that was to lead on eventually to the Roman and ultimately the English alphabet. (It should be stressed here that 18

19 although the Greeks themselves believed their letters came from Phoenician, the precise timing and line of descent have been disputed, especially by Martin Bernal 7, with the suggestion that the date of first borrowing may have been closer to 1500BC and that other Semitic alphabets may have played an important part in the creation of the Greek alphabet.) The Phoenician alphabet was by no means the only system to have flourished around 1000BC. Another branch, Aramaic, which probably derived from Phoenician, was to prove of considerable long-term significance too. To begin with, it displaced cuneiform in the lands of the Persian Empire further east, and then pursued a long line of development, leading on by separate sub-branches to two scripts of continuing importance in the modern world, Hebrew and Arabic. Although these share a distant common ancestry with the Graeco-Roman line (their first letters are respectively aleph, elif, and thus unmistakably cognate with Phoenician and with Greek alpha), precisely how and when the early divergence from Phoenician occurred is not certain. These modern descendants of the Aramaic line, Hebrew and Arabic, have long lost any obvious graphic resemblance to each other, but there are otherwise clear links between them: they share many word roots (e.g. Hebrew shalom, Arabic salaam peace ), both are written from right to left, and both tend to spell out consonants, but not vowels, in full (a feature which, as we have seen, can be traced all the way back to Egyptian hieroglyphics). 7 Martin Bernal Black Athena: The Afroasiatic Roots of Classical Civilization (London 1987: Free Association Books) Vol.1, pp

20 Egyptian hieroglyphics Proto-Sinaitic Proto-Canaanite Phoenician Punic West Semitic??? South Semitic (Ethiopian) Amharic Aramaic Greek Etruscan Coptic Runic Roman English Cyrillic (Russian) Hebrew Arabic From Egyptian Hieroglyphics to English Major/minor, definite/doubtful lines of descent. The Mediterranean cradle of western writing The Phoenicians and their script Especially during the period from about 1000BC to 500BC, the Phoenicians were active as traders and colonizers throughout the Mediterranean and beyond, starting from such centres as Tyre and Sidon on the coast of modern Lebanon. Particularly significant for the development of the alphabet were their numerous contacts with the Greeks and with peoples of the Italian peninsula. Their script continued in use for well over 1,000 years, its last recorded inscriptions dating 20

21 from the early Christian era. Their best known colony was Carthage (in modern Tunisia, reputedly founded in 814BC), whose power grew after the Phoenician homeland passed under Assyrian control in the 7th century BC and subsequently under Persian control; Phoenician script then experienced a kind of second edition as Carthaginian Punic. Surviving examples of Phoenician script are not very numerous, far less so than examples of its eastern relative, Aramaic, but carving on the stone sarcophagus of King Ahiram from Byblos (c. 1050BC) shows its early form, and later examples (4th century BC) show some tendency to more cursive writing, with most letters given a long, sweeping descender. The Phoenician alphabet had 22 letters, strictly speaking all of them consonants, although three of them could serve as vowel letters in certain circumstances. Texts were regularly written from right to left. Both the dominance of consonants and the right left orientation are characteristics that have passed into modern Hebrew and Arabic writing. The letters, with rough English equivalents, were as follows: 21

22 aleph beth gimel daleth he waw zain ox house camel door hook weapon A Z B C/G D E F/U/W heth thet yod kap lamed mem nun arm palm of hand goad water fish H Th I K L M N samek ayin pe sade qop resh shin taw fulcrum eye mouth ape head tooth mark S O P Ts Q R Sh T Phoenician letters with names, meanings and English equivalents Compared with earlier Semitic lettering, the letter shapes here 8 have acquired something of the clarity and simplicity we associate with Roman letters, but they still lack the latter s geometric quality. Verticals and horizontals are not yet exactly at right angles, parallel lines are not yet of exactly equal length, and circles are not yet exactly centred. To the modern eye the general impression is of rather casually handwritten capitals. When we examine the letters individually, we still do not see any (with the possible exception of O) that have unmistakably the same design as 8 Drawn after the inscription on the Moabite stone, late 9th century BC, in Douglas C. McMurtrie The Book (London 1989: Bracken Books; first edition 1943), p. 33. The table has not reproduced entirely satisfactorily, but has been left as CU drew it in his draft of this chapter. The Z under aleph should of course be at the righthand end of the row, under zain. 22

23 their Roman equivalents. One or two (A, K) have roughly the same shape, but are tilted at strange angles. A few (E, M, N, Q) have such long tails as not to be readily recognizable. Some (H, I, T) have extra arms, while others have arms or bowls missing (B, F, P, R), and yet others look strangely distorted, with some Roman curves written straight and Roman straights written curved. Beside these differences in letter shape, there were differences in the inventory of letters, which was by no means yet exactly that of the Roman alphabet. The third letter had the value /f/, not the /j/ of C, there are three letters with roughly the value /r/, rendered above as S, TS, SH (sounds which the Semitic languages variously distinguish), Z is in the middle of the alphabet and not at the end, there is a letter standing for an aspirated T, rather imprecisely rendered as English TH above, there is no G or J, and the English tail-enders U, V, W, X, Y are missing or else disguised earlier in the alphabetic order. As well as lacking dedicated vowel letters, the Semitic alphabets differed in another important respect from the subsequent Roman or English versions: each consonant sound (apart from some overlap with the three S signs) had its own dedicated letter. There is thus none of the ambiguity inherent in English over ways of spelling the sound /R/ (compare fashion, passion, ration), or interpreting the sound-values represented by CH (compare chief, chef, chemist). However, we should not be surprised that the Phoenicians managed to devise so straightforward a system for distinguishing and representing their consonant sounds: the Semitic alphabet was specifically designed for that purpose, whereas the Roman alphabet was not designed to represent the sounds of a modern European language such as English. The absence of full vowel letters in the Phoenician alphabet can be traced 23

24 back to the earliest beginnings of the system: the letters were originally designed to stand, acrophonically, for the initial letters of the words they depicted, and in Semitic languages words do not normally begin with vowels, so in the natural course of events no letters arose to represent vowels. A consequence of the acrophonic origins of the letters is that reciting the Phoenician alphabet means reciting not the sounds which the letters represent, but the words from which the sounds were taken. In English we are used to most of the letter names deriving simply from the sounds the letters stand for, and reciting the alphabet means essentially reciting sounds relating to letters. To appreciate the mnemonics of the Phoenician alphabet, by contrast, one needs to imagine reciting the English alphabet not as ay, bee, cee, dee, ee, etc, but as apple, ball, cat, dog, elf, etc. From Phoenician to Greek The next step in the long story of alphabetic evolution towards the English alphabet was the adaptation of the Semitic alphabet by the early Greeks to meet the needs of their language. The Greeks themselves believed their letters came from the Phoenicians (indeed they called them Phoenician letters ) before the Trojan War; but the earliest reference to their use is much later, citing the recording of the names of the victors in the first Olympiad in 776BC, and the oldest surviving text was inscribed about 50 years after that. This discrepancy of many centuries between the traditional Greek account and the hard archaeological evidence is reflected in scholarly debates about the probable date of the Greek acquisition of the alphabet. Scholars who restrict themselves to archaeological evidence suggest 24

25 the Greeks may have acquired their letters as late as the 8th century BC (perhaps via the Greek settlement of Al Mina in Syria 9 ), while of those who take a broader historical view, some argue for a Phoenician source perhaps 300 years earlier, or an Aramaic source, or even dissemination by Semitic peoples some 500 years earlier still from a pre-phoenician Canaanite source 10. At all events, the Greek alphabet, especially in its earliest forms, clearly has strong links with the Semitic alphabet. Letters of the alphabet can be classified in four ways, by their names, their shapes, their sound-values and their position in the alphabet, and in all four respects there is a clear affinity between most Greek letters and their Phoenician counterparts, although there is confusion in some cases, some have changed sound-values, and some Greek letters have no Phoenician equivalent. The transmission of this alphabet to the Greeks occurred in a complex, untidy and incompletely understood manner. The Greeks were geographically scattered among widely separated, independent city states, and spoke a variety of dialects. It is not surprising that the alphabet was not adopted by all the Greek communities as a uniform, fixed set of letters, but developed marked variations in the course of its dissemination. For at least three centuries there were numerous local scripts but no agreed standard: some scripts took slightly different selections of letters from the Semitic source, and some included additional letters; some scripts gave the same letter a different shape, or represented the same sound by a 9 L. H. Jeffery The Local Scripts of Archaic Greece (Oxford 1961: Clarendon; revised edition with supplement by A W Johnston, 1990). 10 Bernal Black Athena: The Afroasiatic Roots of Classical Civilization, Vol. 1 The Fabrication of Ancient Greece , pp. 16,

26 different letter. Broadly speaking, three types of alphabet have been defined, sometimes named green, blue and red after the colouring of a 19th-century map showing their distribution 11. It was not until the year 403BC that an agreement was reached in Athens to standardize letter forms and values for Greek writing everywhere along the lines of the blue version of the alphabet. This delay in standardization was to have a profound effect on the subsequent development of the alphabet in western Europe, as we will see below. First, however, we must examine the most obvious and general changes brought about by the introduction of the alphabet to Greek. Reversed orientation of lines and letters One of the most striking developments in letter-shape that distinguished writing in Phoenician and early Greek from the later standardized Greek involved horizontal reversals (i.e. letters rewritten as mirror images of their previous forms), and was associated with a reversal of the direction of writing. Although Phoenician and much other Semitic writing (including modern Hebrew and Arabic) was and still is written from right to left, surviving texts suggest that the early Greeks did not normally follow this practice when writing successive lines of text. (Indeed, one of the arguments against a late, direct transmission of the alphabet from Phoenician to Greek is based on this point: it is observed that early Canaanite writing varied in its orientation, which suggests a Canaanite influence on early Greek writing 12, and we may also note that cuneiform was written from left to right.) In early Greek 11 L. Threatte The Greek Alphabet, in Peter T. Daniels & William Bright (eds.) The World s Writing Systems (New York 1996: Oxford University Press Inc.), p Bernal, Black Athena p

27 texts, the first line normally starts on the right, but the second line begins at the end where the previous line finished, so that successive lines were written in alternate directions. This practice is known as boustrophedon, literally as the ox turns (i.e. the direction of writing alternated with each line of script, just as an ox ploughing a field turns at the end of each furrow and ploughs in the opposite direction to that of the previous furrow), and it persisted until as late as the 5th century BC in Crete. Readers and writers did not thus have to zigzag, as in modern writing, from the end of one line to the opposite end of the next, but moved from one line-end to begin the next immediately below. This process entailed a similar alternation in the direction letters faced, with, for instance, reverse ( retrograde ) Ǝ on the 1st line, but the later standard E on line 2 (early literate Greeks must have been adept at mirror reading and writing). The result was that the standard Phoenician orientation of both letters and text ceased to be standard in Greek, indeed early Greek did not have a standard orientation. But by about the 5th century BC, continuous text was normally written from left to right, the direction that almost all western alphabetic writing has followed ever since. At the same time, asymmetrical letter forms such as E usually dropped their previous Semitic orientation (this explains the already-mentioned reversal of B between early Greek and Greco-Roman). The left-to-right direction of writing may have a double advantage for right-handed writers: what has just been written is less likely to be hidden by the hand moving across the writing surface; and in using ink, there may be less danger of smudging. But any idea that the left-to-right direction is inherently superior may simply reflect a Eurocentric outlook, and we should reflect 27

28 on Herodotus s words 13 : In writing or calculating, instead of going, like the Greeks, from left to right, the Egyptians go from right to left and obstinately maintain that theirs is the dexterous method, ours being left-handed and awkward. As examples of this reversal of orientation, compare for example, the lettershapes for he, kappa and resh in the table of the Phoenician alphabet with the equivalent Greek letter-forms Ε, Κ, Ρ for epsilon, kappa and rho of the Greek alphabet. Note also the re-orientation of aleph to alpha Α and lamed to lambda Λ. The introduction of vowel letters in Greek We have observed that the letters of the Semitic alphabets were, strictly speaking, all consonants (although some could have a secondary function indicating vowels). In particular, the aleph and ayin, although ancestors of our A and O, stood for consonant sounds articulated in the throat (respectively, a glottal stop and a voiced pharyngeal fricative or perhaps pharyngealized glottal stop, indicated by the reversed apostrophes, < > and < >, before the two letter names), and not for vowels. The Greeks, however, identified vowel sounds in their language and felt the need to represent them with appropriate letters. For this purpose they adopted various Phoenician letters which stood for consonant sounds that did not occur, and therefore did not need to be spelt, in Greek. So Greek alpha (Α) no longer represented the glottal stop of its precursor in Phoenician, but an [`] sound; epsilon (E) was not a breathy Phoenician H, but typically a short E-sound; eta (H) in some accents at first represented a [g] sound (and was called heta) but after 13 Herodotus The Histories, translated by Aubrey de Sélincourt (Harmondsworth 1954: Penguin Classics), p

29 standardization came to stand for a longer E-sound; and Phoenician ayin typically came to represent a Greek short O. Two further letters already had a subsidiary function as vowels in Semitic writing, and phonetically they share some characteristics of both consonants and vowels, their sounds being often described as semivowels rather than as plain consonants in English. One was the Phoenician waw, which as a consonant stood roughly for [v], but could also serve to represent the sound of long [t9] (a similar dual function is hinted at in English, whose W is called double-u ; contrast also the dual values of U in English suede, suet). Some local Greek scripts for a time kept a letter (digamma, F) derived from Phoenician waw in its original position after E, to represent a labial sound somewhere between [e] and [w], but it did not survive the standardization of written Greek. More importantly, the later standard Greek alphabet developed its /x/ vowel, upsilon (Υ), from that same Phoenician waw, placing it after Τ, after all the directly Phoenician-derived letters. The other Phoenician consonant that could also be used as a vowel was yod, which basically stood for the sound of the English consonant Y, [i], as in yes. But rather as Y can alternate with the vowel letter I in English in the spelling laniard/lanyard, so yod could effectively also serve as an I-glide and hence as a short I, [H]. It was this function to which Greek applied the Phoenician yod, calling it iota and leaving its position in the alphabet unchanged. The significance of these new Greek vowel letters in the development towards the Roman and eventually the English alphabet was that there were now clearly recognizable letters Α, Ε, Ι, Ο with roughly the sound-values that these letters were 29

30 from then on to keep in most languages (though with major variations in modern English). Modern U and Y had not yet emerged so distinctively, in that Greek Υ was the source from which the Roman alphabet was still to develop its U (and some 1,000 years later, its V) as well as the directly Greek-derived Y. The standardized Classical Greek alphabet included two further vowel letters which were not transmitted as vowels to Roman. One was eta (Η) which had been switched from its Phoenician consonantal value, and the other was omega (Ω). The latter development has been attributed to speakers of the Ionian dialect, who around 600BC opened a gap at the bottom of the circular Ο which they henceforth called o-micron small O, as opposed to the new letter o-mega big O. The purpose of the two additional Greek vowel letters, eta and omega, was to distinguish long and short values of the vowels E and O, which earlier local Greek scripts had mostly not bothered to spell differently. But although early Greek had alternative long and short sound-values for all its vowel letters (modern Greek no longer makes such distinctions of quantity), it did not create alternative letter forms for long A, I or Y. And although Latin similarly had long and short values for its vowels, it did not distinguish any of them with separate letters, not even adopting the facility offered by Classical Greek of using different letters for short and long E, O with epsilon/eta and omicron/omega respectively. We may here interrupt our history of the alphabet to speculate what the consequences might have been for English spelling if the Roman alphabet had taken over these extra vowel letters from Greek. It may be that the two identically spelt Modern English verb forms to read and he read would be distinguished as to rηd, he red; or maybe the different vowel sounds of weep, wept would be 30