Th e Greek Its Origins, Evolution and Impact on the Development of in Europe Nikolaos Pantelidis

1. Introduction. ) Linear A, an as yet undeciphered syllabic script, in an unknown pre-. Th e Τhe is certainly not the fi rst script texts preserved can be dated to the fi rst half of the on European soil. Yet it is the fi rst “true” alphabet 2nd millennium BC, are mostly incised on clay tab- in the , in the sense that it is the lets and come mainly from the island of . fi rst script, in which every sound in a word is rep- ) . A syllabic script (containing also a resented graphically by a distinct symbol, contrary wealth of ideograms in parallel and complementary to earlier pictographic or syllabic (or quasi-syllabic) use to the syllabograms) used in Greece during the writing systems, which were in use at least till the second half of the 2nd millennium BC. Although it end of the 2nd millenium BC in the Middle East, bears strong similarities in the forms of its symbols Asia Minor, as well as in Greece itself. to the Linear A script, the exact relationship be- Th e Greek Alphabet had a great impact on the tween the two is still a matter of dispute. Th e fi nds development of script in the European continent are mostly inscribed clay tablets from the archives and the whole world. All writing systems nowa- of the palaces in the centres of the Mycenaean ci- days in use in nearly the whole of Europe (i.e. the vilisation in Crete and mainland Greece. Th e texts various versions of the Cyrillic and the Latin al- belong to 14th-13th centuries BC and are in their phabets) can ultimately be traced back to it. Th e overwhelming majority short and contain mostly Greek Alphabet itself stands at the end of a long lists of personnel, livestock, various commodities line of descent from several centuries older writing and artefacts such as olive oil, honey, wool, furni- systems and forms part of a complex family tree1. ture etc. Th e script, deciphered in 1952 by the Eng- Furthermore, as far as the graphic representation of lish architect Michael Ventris and the Cambridge the Greek language itself is concerned, it replaced classicist John Chadwick, was shown to represent an earlier syllabic system, the Linear B (see below a very archaic form of the Greek language, more § 2). Th e extinction (of some) of the other earlier archaic in many respects than the language of the systems (see § 2) in use in Middle/Late Bronze- early Greek epic poet Homer. It pushed back the Age Greece, perhaps went hand in hand with the beginning of the recorded history of the Greek extinction of the pre-Greek languages they most language by several centuries. Th e script was highly probably served. unsuitable for the representation of Greek, since it lacked important phonemic distinctions, such as 2. Earlier writing systems in Greece vowel length distinctions (e.g. /a/ vs. /a:/, the dis- and Cyprus. tinction between the liquids /r/ and /l/, or distinc- tions of manner of articulation, as in the velar stops Th e following scripts were in use during the 2nd /k/ : /g/ : /kh/ (voiceless : voiced : aspirated). millenium BC in Greece and Cyprus, prior to the d) Cyprominoan script (15th – 12th centuries introduction of the alphabet: BC). A syllabic script related to the syllabaries of a) Hieroglyphic, an undeciphered pictographic Greece (Linear A and B) and the . system in use in Crete and other islands of the Aege- It records a pre-Greek language of Cyprus. an in the fi rst half of the 2nd millenium BC (not to be e) Cypriot syllabary. A syllabic script, related to confused with the homonymous Egyptian script). Linear B, but more elaborate than the latter, in use in Cyprus for the representation of the Greek dialect of th th 1 See Coulmas 2003, ch.6. Cyprus from the 11 century to the 3 century BC,

216 SCRIPTURA MUNDI deciphered in 1871. Th e island was settled by Greek runs from right to left , just as was the case with the populations from the aft er the fall of various Semitic scripts, including the Phoenician. the Mycenaean civilisation in mainland Greece in e) authors were aware of the the 12th century BC. Th e abandonment of the syl- Phoenician provenance of the alphabet. Th e histo- labary went hand in hand with the disappearance of rian Herodotus for instance records an old myth the Cypriot dialect from written discourse. according to which the mythical Phoenician king Kadmos introduced the letters in Greece. Th is 3. Th e Greek Alphabet. particular myth may have a nucleus of historical reality, yet refers to a king, who, if at all existent, 3.1. Th e adoption of the Phoenician script by the should have lived in the period of the Mycenaean Greeks. civilisation (16th-12th century BC), rather too early to have introduced the alphabet in Greece. Th e Th e Greek alphabet belongs to the phonograph- term phoinikēia (grammata) ‘Phoenician (letters)’ ic type of scripts (Coulmas 2003:111), i.e. their used to refer to the alphabet with regard to its prov- graphic symbols represent individual sounds (pho- enance is also securely attested in ancient Greek in- nemes). A symbol can of course be shared by more scriptions. than one sound with common features, as was the case in the earliest Greek alphabet (e.g. no length Th ere are a number of issues concerning the distinctions in vowels, see below § 3.2). process of adaptation of the Phoenician script: Th e beginnings of the Greek alphabet lie in the a) Th e date of creation of the alphabet on the Geometric Period (the so-called “Dark Ages”) of basis of the Phoenician script. Th ere can be little ancient Greek history (11th-8th century BC). It doubt that the alphabet was adopted (and adapted) is generally agreed that the Greek alphabet is the between the collapse of the Mycenaean civilisation result of the adaptation of the north Semitic Phoe- in the 12th century BC which was accompanied by nician abjad-type2 of script (quasi-syllabic), i.e. a the abandonment of Linear B, and 800 BC, since script which, as other scripts of the ancient Semitic the earliest, though very fragmentary and meager, peoples, consisted only of symbols for consonan- Greek alphabetic epigraphical remains (some of tal phonemes. Th ere is abundant evidence for the them abecedaria) may belong to the fi rst half of the Phoenician provenance of the Greek alphabet: 8th century BC or even somewhat earlier4. Th e pe- a) Th e majority of the Greek names of the sym- riod 1200-800 BC, the so-called “Dark Ages”, saw bols have no Greek etymology but can be traced a substantial increase in the trading contacts of the back to the Semitic appellatives, by which the let- Greeks with the Phoenicians in the Eastern Medi- ters were originally named according to the ac- terranean. Th e fi rst “fully” preserved Greek - rophonic principle, e.g. alpha (ˀaleph ‘ox’, [ ˀ ] = betic inscription comes from and can be symbol of the glottal stop in the International Pho- securely dated to the middle or to the second half netic Alphabet), (bet() ‘house’, symbol of the of the 8th century BC. It is the famous Dipylon-in- voiced bilabial stop /b/), (gim()el ‘camel’, scription on a vase. Th e inscription runs from right symbol for the voiced velar stop /g/ ) etc.3 to left , as did the Phoenician script, but a number b) Th e order of the letters in the Greek alphabet of the letters more or less diff er in shape from their refl ects the order of the respective Phoenician ones. Phoenician originals. Most scholars would agree c) Th e shapes of the letters on the earliest pre- that (the end of ) the 9th century BC is a higly pos- served Greek epigraphic fi nds are very similar to sible date of adaptation (or at least a terminus ante the respective Phoenician ones. quem), although the issue is still debated and alter- d) Th e direction of the script. In the earliest pre- native datings cannot be altogether excluded. served inscriptions in Greek alphabet, the script b) Th e place of adaptation of the Phoenician script / creation of the Greek alphabet. Many lo- 2 On the term abjad see Coulmas 2003:113. cations have been proposed on diff erent grounds 3 On the history of the letter names in Semitic and Greek see Willi 2008. 4 Woodard 2010:44.

THE GREEK ALPHABET 217 as the possible places of creation of the Greek al- received their vocalic value in much the same way phabet. Th e range covers a wide geographic space, or on the basis of acoustical similarities of the vow- from Al Mina in Syria, where an important Greek els with the Phoenician consonants6. trading station was founded during the “Dark Th e earliest alphabet probably ended in Ages”, Cyprus (Woodard 2010:41-43) on the basis (). It also lacked symbols for the aspirated la- of the spelling habits of the Cypriot syllabary, the bial stop (/ph/) and the aspirated velar stop (/kh/), island of , the island of Crete (where one and there were no signs for the consonant clusters of the most archaic Greek was used) and /ks/ and /ps/. But in most later local alphabets the Greek island of , which had also early symbols for /ph/ and /kh/ were created. Finally, no extensive contacts to the eastern Mediterranean vowel length distinctions were made. and especially the Phoenicians. Th e matter is still Here are the symbols of the earliest Greek al- disputed. phabet in their “modern” form, with their Greek c) Was the alphabet of single or multiple ori- names and the phonemes they originally represent- gin, i.e. was it created at a single time at a specifi c ed: unique place, from where it spread to the Greek Α alpha = /a/, /a:/ world? Th is is an also as yet unresolved matter. B beta = /b/ d) Th e exact processes by which the original al- Γ gamma = /g/ phabet (if it was the result of a single adaptation Δ = /d/ process) split into the numerous types of alphabets E later name “e psilon” ‘plain’, i.e. of Archaic and Classical Greece5. ‘unaspirated e’ = /e/, /e:/ Ϝ 3.2. Th e structure of the earliest Greek alphabet. Η (h) = /h/ Th e earliest developments. Θ = /th/ I = /i/, /i:/ Th e Phoenician script, which itself existed in sever- K = /k/ al variants, contained 22 symbols which, as already Λ = /l/ mentioned, represented only consonantal sounds. M my = /m/ Of course not all letters retained their original con- N ny = /n/ sonantal values aft er the adoption by the Greeks. Ο = /o/, /o:/, later o micron ‘short o’ Some (as e.g. theta <Θ>) were used to denote Q = /k/, the ancestor of Latin sounds similar to the Phoenician ones or interpre- Π = /p/ table as similar by the Greek adaptors. Ρ = /r/ Th e major innovation of the Greek adaptors of Σ = //, in some alphabets the letter the Phoenician script is the change of symbols for was in use (see below) consonants absent from the Greek phonological T tau = /t/ system to letters for vowels. Th is was accomplished Y = /u/, later y psilon on the basis of acrophony. For instance, the name of the fi rst letter of the Phoenician script began Th e Phoenician script also contained an Y- with a glottal stop [ˀ] which was exactly the pho- shaped symbol called (phonetic value: [w]), netic value of . Th e glottal stop was followed which split to two diff erent Greek letters: by the vowel [a] ([ˀa-]). Th e sound [ ˀ ] was absent Th e <Υ> (y psilon), whose original Greek pho- from the phonological system of the Greek lan- netic value was [u]7, and the <Ϝ> digamma (= guage of that time. Th e adaptors interpreted the “double gamma”, because of its shape), whose Greek name as beginning with a plain /a/, thus endow- phonetic value was that of a semivowel [w]. Th e di- ing the letter with the value of the fi rst –according gamma was not used in the Ionic-speaking regions to their interpretation- sound of the name. Other vocalic (originally consonantal) letters might have 6 Powell 1991:42-46, Woodard 2010:34-35. 7 In the Ionic group of this sound was 5 See for instance Woodard 1997. fronted to /y/.

218 SCRIPTURA MUNDI of the Greek world, because in the Ionic group of 3.3. Direction of the script. Ancient Greek dialects, the sound [w] disappeared before the time of the earliest (surviving) Greek Th e Phoenician script ran from right to left , and inscriptions. was eventually placed right aft er this was the original direction of script in the tau whereas <Ϝ> digamma retained the origi- Greek alphabet. Th e direction changed later to ‘left nal position of waw aft er . → right’ as it is today. Th e Phoenician script contained also a sym- In several regions the boustrophedon-system bol called qop(h) which represented a consonant emerged at an early date. Th e Greek word boustro- acoustically and articulatorily close to [k] (voiceless phedon is an adverb meaning ‘taking turns like the uvular stop [q] or other?). was transferred to oxen (during ploughing)’. Th e name was given be- Greek as qoppa (<Ϙ>), but the distinction between cause the script changed direction in every line of the two Phoenician consonants was unknown in the text, ‘right → left ’, in the next line ‘left → right’, Greek. Th us the qoppa came to represent the same then ‘right → left ’ again and so on. Th is system was consonantal phoneme as kappa (), namely abandoned relatively early, although there are re- /k/, albeit in diff erent contexts (e.g. before other gions, as e.g. Crete, where boustrophedon was cur- consonants, before back vowels). Th e qoppa (<Ϙ>) rent until the 5th century9. survived in several local Greek alphabets (in Ath- Th ere were no lower case letters and in the vast ens until the middle of the 6th century BC), but majority of cases no word division. Th e texts were gradually fell out of use everywhere till the 5th cen- written continuously, a fact which oft en creates tury BC. In later periods it was used as a numerical problems in the interpretation of the epigraphic symbol for ‘90’. It is also the direct ancestor of the texts, especially those written in poorly recorded letter of the . dialects. Th ere were also no accent marks. Th e Phoenician script also contained four sym- bols for sibilants or aff ricates with a sibilant com- 3.4. Later developments. Th e local Greek ponent8: samek(h), shin, ṣade, zayin. alphabets. Th e samek(h) is continued in some Greek alpha- bets as <Ξ> () (Greek phonetic value: [ks]), but Th e Greek alphabet from the time of the earliest al- is absent form others (see below § 3.4). phabetic inscriptions appears split into several local Shin was a W-shaped Phoenician symbol of the variants. Th e German scholar A. Kirchhoff divided palatoalveolar sibilant [ ʃ ] (the initial consonant the local Greek alphabets into three major groups. of Engl. she). In most Greek alphabets the letter Th e classifi cation is based on the overall structure was adopted with the name sigma ‘hissing, hissing of every alphabet (their closeness to the earliest sound’ (<Σ>) as the graphic representation of the alphabet) and the shapes of the letters. Kirchhoff sibilant /s/. For the same (?) consonant in some displayed the distribution of the three major types Greek alphabets the letter ṣade (possible Phoeni- of alphabets with a diff erent colour on a map in cian phonetic value: [ts] or something similar) was the third edition of his book on the history of the adopted as san, its shape very much resembling the Greek alphabet (1877), and it is by the names of shape of . Eventually only sigma would sur- those colours that the diff erent types of alphabets vive as the representation of [s]. are known since then: Zayin is continued in the Greek alphabet as (a) Th e “green” alphabets. Th is most archaic (). Th e exact phonetic value of the latter group of local alphabets was in use in Crete and is highly disputed, although it seems to have origi- the Cycladic Aegean islands of Th era and Melos. nally represented a voiced aff ricate ([ʣ] or [ʤ]) in It is generally believed to refl ect the oldest state of Greek. the Greek alphabet. It lacked symbols for the aspi- Further letters such as <Φ> (, phonetic rated labial stop (/ph/) and the aspirated velar stop value: [ph]) were created aft er the adoption of the (/kh/), and there were no signs for the consonant Phoenician script (see below § 3.4). clusters /ks/ and /ps/ (the digraphs <ΚΣ> and

8 Powell 1991:46-48, Woodard 2010:31-33. 9 On boustrophedon see Jeff ery 1961:43-50.

THE GREEK ALPHABET 219 <ΠΣ> were used instead). It further did not make = /e/, /e:/, /ε:/ any length distinctions in vowels. = /o/, /o:/, /ɔ:/ For the representation of /ph/ and /kh/ various strategies were employed: In Crete the aspiration Th e older Attic (Athenian) alphabet: was omitted altogether and the signs for the respec- A B Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ Ν Ο (Ϙ) Ρ Σ Τ Υ Φ Χ tive unaspirated stops were employed (pi and kappa or qoppa, <Π>, <Κ>, <Ϙ>), whereas in the other 4.2. Th e Ionic alphabet. Structure and spread. islands digraphs consisting of <Π> <Κ> <Ϙ> + the letter eta, which was originally the symbol for As mentioned above, the Ionic cities of the Ae- /h/, were in use: <ΠΗ>, <ΚΗ>, <ϘΗ>. gean coast of Asia Minor used an alphabet of the (b) Th e “blue” alphabets, also known as “east- “blue” type. Furthermore this alphabet contained ern” alphabets. Th ese contained symbols for the symbols for /ks/ and /ps/. Th e symbol for /ps/ ( aspirated labial stop (/ph/) and the aspirated velar <Ψ>) was created in aft er the adoption of the stop (/kh/), namely phi (<Φ>) and () re- Phoenician script. A number of alphabets in Asia spectively. Th e alphabets of the Ionic colonies of Minor (mainly Ionia) and elsewhere also possessed Asia Minor (cities of Miletos, Ephesos etc.) and a letter of non-Phoenician origin which is believed the older Attic (Athenian) alphabet belonged to to be continued in much later times as the symbol this group, among others. which was named and employed only as a (c) Th e “red” alphabets, also known as “west- numerical symbol for ‘900’ ( ϡ ). Th e shape of this ern” alphabets. Th ese include signs for /ph/ and / additional letter was reminiscent of tau (ᛏ, Ⲧ) and it kh/ and cross-shaped symbols (X, + etc.) for the was mostly used as a symbol for what was written as consonant cluster /ks/, but most of the them no (in ) or <ΣΣ> in other dialects, symbol for the cluster /ps/. Th e cities of the Greek e.g, te<Ⲧ>ares ‘four’ (Ionic Greek te<ΣΣ>ares, At- island of Euboea (north of Athens) among others tic teares) etc. Its use was abandoned in the used a “red” alphabet. Th ey founded colonies on 5th century BC.10 the coast of South (e.g. in the vicinity Another important development concerns the of today’s Naples) from which the alphabet spread letter eta (earlier , ), originally the symbol to the Etruscans and from the latter to the Romans, for “aspiration” (/h/), a sound appearing almost something which explains the fact that the symbol exclusively at the beginning of the word. In the for /ks/ has the shape X in the Latin alphabet. Th e dialects of Ionia (closely related and belonging to “blue” alphabets instead used cross-shaped letters the same group as the dialects of Athens and the (called chi) for /kh/. Th e Latin alphabet also re- island of Euboea) the phoneme /h/ was lost early. tained the use of as a symbol for /h/. Th is change left the letter without a phonetic value. Th e Ionians re-used to represent the 4. Th e way to the unifi cation of script in long open e (symbol in the International Phonetic the Ancient Greek world. Alphabet: [ε:]). In a last step a letter also for long open o (symbol in the International Phonetic Al- 4.1. Th e Attic alphabet (the alphabet of Athens). phabet: [ɔ:]) was created, the o mega <Ω> (= “long o”) rendering the Ionic alphabet symmetrical in Th e alphabet of Athens belonged to the “blue” this respect: group. It had symbols for /ph/ and /kh/ (phi <Φ> and chi ) respectively and lacked symbols for A B Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ Ν Ξ Ο Π Ρ Σ Τ Υ /ks/ and /ps/ using digraphs instead. In its older Φ Χ Ψ Ω form it preserved the letter qoppa (see above) till the 6th century BC. It did not make any length or From Ionia, more specifi cally the city of Mile- height distinctions in the vowels: Th e letters tos, the alphabet (and the orthographic habit of and could be used for short, closed long and not marking /h/, a sound which was not lost in the open long e and o respectively. Th at is, and represented three vocalic phonemes each: 10 On sampi see Jeff ery 1961:38-39, Buck 18, 349

220 SCRIPTURA MUNDI Attic dialect until much later) fi rst spread to Ath- 6. Alphabets derived from the Greek ens, where it was introduced by an offi cial decree in alphabet in Europe. 403/402 BC and, during the course of the 4th cen- tury BC, to the rest of the Greek-speaking world, Th e Greek alphabet had a substantial infl uence on replacing all former local alphabets. Th is is exactly the development of writing in Europe, as well as the alphabet (the capital letters) which is still in use outside Europe proper (Asia Minor, Armenia etc.): today for the representation of the Greek language. A Greek alphabet of the “red” type (see above), in use in ancient Greek colonies of , 5. Further developments: Th e accent was adopted by the Etruscans and from them by marks and the spiritus lenis / spiritus the Romans, being thus the ancestor of the Latin asper. Th e minuscule. alphabet. Th e scripts of numerous other peoples of ancient Italy (Oscans, Umbrians etc.) can also ulti- Th ere are clear testimonies in ancient sources as mately be traced back to the Greek alphabet. Some well as other evidence that Ancient Greek had a Oscan inscriptions are also written in the Greek pitch accent, whereas has a stress alphabet. accent, as do most modern European languages. A number of Gaulish inscriptions from south- Many otherwise homophonous forms were dis- ern Gaul have also been found written in the Greek tinguished by the diff erent accent (e.g. ḗ = ‘or’ vs. alphabet.  = ‘he said’). As already mentioned, the Ancient Later, in the 4th century AD, the Gothic bishop Greek writing system did not have accent marks or Wulfi la developed an alphabet and spelling system symbols for /h/ (the spiritus asper) in the beginning based on the Greek model for writing down his of the word or its absence. Th e accent marks fi rst Gothic translation of the New Testament for the appear in the Hellenistic period (around the end Gothic Germanic peoples living along the Danube of the 3rd century BC) in papyri containing older river. works of poetry. Th eir original purpose was to help Last but not least, the Greek alphabet formed distinguish between forms that had become ho- the basis for the creation of one of the most widely mophonous due to major sound changes that had used scripts in Europe and beyond, the Cyrillic al- begun to take place already in the late Classical (4th phabet, created in the 9th century AD for the pur- century BC) and extended into the Hellenistic pe- poses of Christian missions to the Slavs. Th e Cyril- riod (aft er 300 BC). Aft er passing through various lic alphabet is nowadays used by all Slavic peoples stages, the accent marks, as well as the spiritus asper who belong to the Eastern Orthodox Church or < ‘ > (marking the former presence of /h/) and the follow the Orthodox rite. It has also probably been spiritus lenis < ’ > (marking the former absence of the basis of the creation of the other old and nowa- /h/) became a regular part of the Greek writing days obsolete Slavic alphabet, the Glagolitic. system much later, in the 10th century AD11. Th e accent marks, as well as the spiritus asper and the spiritus lenis, were abolished in 1982. Selected bibliography Th e Greek minuscule script appears for the fi rst time in Byzantine manuscripts of the 9th century. It Th ere is a vast amount of literature on the Greek also contained a number of ligatures, one of them, alphabet. Here only a small and by no means repre- the <ς>, very much resembling the fi nal sentative selection can be presented. Th e fi rst two <Σ> in shape. It was a ligature of <ΣΤ> and was in works do not directly concern the Greek alphabet. use from medieval up to modern times, also having the numerical value ‘6’. Allen, S. W. 1968. Vox Graeca, a Guide to the Pro- nunciation of Classical Greek. Cambridge: Cam- bridge University Press. Buck, C. D. 1955. Th e Greek Dialects. Chicago: 11 On the accentual system of Ancient Greek and the accent Th e University of Chicago Press. marking system see Allen 1968:106-124. Coulmas, F. 2003. Writing Systems. Cambridge:

THE GREEK ALPHABET 221 Cambridge University Press. Semitic letter names as a chapter in the history Jeff ery, L. H. 1961. Th e Local Scripts of Archaic of the alphabet”. Classical Quarterly 58: 401– Greece: a Study of the Origin of the Greek Alpha- 423. bet and its Development fr om the eighth to the Woodard, R. D. 1997. Greek Writing fr om Knossos fi ft h centuries BC. Oxford: Clarendon Press. to Homer, a Linguistic Interpretation of the Ori- Kirchhoff , A. 1877. Studien zur Geschichte des gin of the Greek Alphabet and the Continuity of griechischen Alphabets (3rd edition). Berlin: Ancient Greek Literacy. New York/Oxford: Ox- Ferd. Dümmlers Verlagsbuchhandlung. ford University Press. Powell, B. B. 1991. Homer and the Origin of the Woodard, R. D. 2010. “Phoinikēia grammata: an Greek Alphabet. Cambridge: Cambridge Uni- alphabet for the Greek language”. In: Bakker, versity Press. Egbert J. (ed.), A companion to the ancient Greek Willi, A. 2008. “Cows, houses, hooks: the Graeco- language. Oxford: Blackwell. 25-46.

222 SCRIPTURA MUNDI