Palmer L.R. The Greek Language

Подождите немного. Документ загружается.

< previous page page_198 next page >

Page 198

the official language of the State, its organs and institutions, the Church, post-elementary education, the

Universities and a great part of the Press.

Demotic is the standard spoken language of Greece today, in that it is the natural form of everyday expression used

by all classes, regardless of differences of education. No less important is that since the beginning of the nineteenth

century it has increasingly served the needs of creative literature. Here history has repeated itself in that poetry

preceded prose: demotic was used at the beginning of the nineteenth century by Solomos and his school in the

literary renaissance of the Ionian islands, whereas the first work of prose in demotic appeared in 1888. Political

events likewise favoured the adoption of a national language based on the southern group of dialects because the

Peloponnese was the gathering ground for the revolutionary fighters, and between 1821 and 1833, when Athens

became the capital of the resurrected Greek state, there developed 'A new common language based on the dialects

of the Peloponnese but with a good many Ionian features. Peloponnesian Greek was well fitted to be the basis of a

national language. Without either the radical phonetic changes of northern Greek or the archaic features of Cretan

or Cypriot, it was easily understood by all Greeks, and was sufficiently close to the language of late Byzantine and

post-Byzantine vernacular literature to be acceptable to all as a common tongue.'1 The triumph of demotic has

recently been officially acknowledged. By a decision of the Greek Ministry of Education the 'katharevousa' has

ceased by law to be the official language of the Greek state; 'demotic' has been taught from September 1976 in all

Greek schools.

Yet for all the radical changes outlined in this chapter, the natural spoken language of Greece today is still

recognizably the same as that of Homer. It remains a highly archaic language of unmistakably Indo-European

structure in its nominal and verbal morphology.

1 R. Browning, 105-6 (1969).

< previous page page_198 next page >

< previous page page_199 next page >

Page 199

PART II

COMPARATIVE-HISTORICAL GRAMMAR

< previous page page_199 next page >

< previous page page_201 next page >

Page 201

VII

Writing and Pronunciation

The comparative method in linguistics rests on a simple fundamental fact: all speech consists of the use of vocal

sounds to signal messages, and the connection between the vocable, the speech 'expression', and the meaning it

conveys is arbitrary. Consequently, if we observe far-reaching resemblances between two languages, the fact of

arbitrariness compels us to rule out the possibility of coincidence, that is independent invention, and to conclude

that there must be some historical connection. This may, of course, be of the most superficial kind, for words may

easily be borrowed from one language into another, like potato and tomato. Of greater importance for the

comparatist are the resemblances of grammatical structure, about which a brief word may be said. In linguistic

descriptions the speech corpus, taken as an adequate sample of the speech habits of a given community, is broken

down into a finite number of minimal meaningful elements, S(ound)-M(eaning) units, the so-called 'monemes', the

combination of which into more complex units at different levels by stated rules will generate all the conceivable

utterances in the speech community in question. A small proportion of these monemes have structural functions

within the utterances, and the part of the description which specifies these is called the grammar. Such structural

devices are rarely borrowed, and it is on resemblances of this kind that the comparatist lays greatest stress in

linguistic genetics.

Fundamental, then, for the comparatist is the establishment of the sound-meaning units in the languages to be

compared. With contemporary speech the sound aspect is directly observed. But for the most part the historian of

language must have recourse to written records. He cannot make a direct comparison, say, of the Greek alphabetic,

or Linear-B syllabic word-forms with others written in the devanagari of India. To make his comparison of S-M

forms and so establish a basis for his deductions he must deduce the spoken expression from the written

characrers: he must seek to convert the graphemes into

< previous page page_201 next page >

< previous page page_202 next page >

Page 202

phonemes. The first task of the historian of Greek must, therefore, be to present the problems that arise in

determining the pronunciation of the Greek alphabetic texts. The Linear B script discussed in Chapter II, though

historically earlier, is scientifically secondary: the values of the syllabic characters are arrived at by correlation with

those of the later alphabetic script.

The whole series of Greek letters from alpha to tau was taken over from a North Semitic alphabet.1 From the

tabulation (Fig. 6) it emerges that the shapes of the characters are basically the same, their names are virtually the

same (with inevitable adaptations), and the order is the same (with the exception of the sibilants, see below).

Another peculiarity is that at least the first line in archaic inscriptions runs from right to left. In North Semitic

inscriptions the text was written in every line from right to left, but in archaic Greek practice the even lines run in

the opposite direction to the odd lines. This method of writing is known as boustrophedon 'as the ox turns'. Such

texts may also begin from left to right with the even lines right to left.

Boustrophedon is common to all the early Greek local alphabets and is one of the common divergencies from

North Semitic practice, a fact which indicates that the Greek alphabet originated in a comparatively restricted area.

Another change, and this was a notable advance towards the goal of perfect phonemic notation, was the use of

certain characters denoting Semitic consonants but superfluous for Greek, to render the Greek vowels. 'Alep, he'

and 'ayin were used for a, e and o. Yôd, which denoted the semivowel [j], a sound absent from alphabetic Greek

except as a glide sound for Mycenaean (see p. 42), was used for the vowel [i], a value which it occasionally also

had in Semitic. Waw was given two forms: one (upsilon), taken over from a cursive Phoenician script, was used

for the vowel

; the other (digamma) was preserved at its correct alphabetic place in a number of Greek dialects

(see pp. 226-8).

1 The Greeks themselves described the letters of the alphabet as 'Phoenician'. Supporting evidence

has been sought in the expression

(infinitive) 'to write' and the corresponding agent noun

which occur in an early fifth-century inscription from Crete of a man proposed for the office as 'recorder'

and scribe, but another suggestion is that the verb and the derived noun refer to writing in red letters.

< previous page page_202 next page >

< previous page page_203 next page >

Page 203

N. Sem.

Greek Sem. Greek Greek

AlphabetsAlphabetsName Name Value

'alep [a]

bêt [b]

gimel [g]

dalet [d]

he' [e]

waw [w]

zayin [dz], [zd]

het [h],

tet

yód [i]

kap [k]

lamed [l]

mem [m]

nûn [n]

samek [ks]

'ayin [o]

pe' [p]

qóp [q]

lé, ló [r]

in [s]

taw [t]

[u]

Fig. 6. The Greek Alphabet.

(1) Later

(6) The velar plosive used in many

local scripts before o and u

(2) Also

(7) Perhaps from samek

(3) Corresponds to sa

dê (8) Later

(4)Perhaps corresponds to in

(5)Later

< previous page page_203 next page >

< previous page page_204 next page >

Page 204

This Greek proto-alphabet was further enriched by letters for the aspirated consonants

and for ph and kh

respectively, parallel to

for th. This alphabet, consisting of 22 letters, was used (with the omission of digamma

which had no application in Attic-Ionic, see p. 226) by the Athenians down to the end of the fifth century B.C. It

used

for all varieties of e, while was used similarly for all varieties of o. still stood for the aspirate but on

occasion was used syllabically for he. In 403 B.C., under the archonship of Eucleides, the more evolved East Ionic

(Milesian) alphabet was adopted at the instigation of the reformer Archinos. This alphabet distinguished

(e) from

(h) and (o) from W (w). It had also introduced the signs (x) and (y) for the clusters ks and ps,1 possibly

in parallel with

in its original value ds (see below).

In determining the values of the characters of the Greek alphabet a variety of clues is used. First we have the

modern use of its many descendants, notably the Roman alphabet and its derivatives. Pronunciation evolves, of

course, while orthography remains unchanged. However, if the modern testimony is uniform, this gives a prima

facie value for the earliest stage which can be tested. Thus there is little doubt that the ancient values a [p], t [t] and

k [k] did not differ to any great extent from their present-day pronunciation. Then we have descriptions and

comments by contemporary commentators on linguistic matters. Further, hints may be derived from onomatopoeia

(as in the sheep's cry

or in the word cf. Lat. mugire, English mooing for the lowing of cattle) and

puns. Word borrowings from one language into another enable us to make cross-checks and to draw on other

traditions. Such traditions may make specially valuable contributions when the alphabet is adopted or adapted for

other dialects or languages. Finally, spelling variations give glimpses of phonological changes.

Vowels

a in MnG has the value [a] and that this is true of OG is indicated by the history of the corresponding Latin a and

by the

1 The Western alphabets gave the value ks to and kh to y, while ps was rendered by the digraph FS

< previous page page_204 next page >

< previous page page_205 next page >

Page 205

Greek loan-words in other languages. There is no evidence that the long vowel had a different quality (as it has in

English [a:] v.

e was used in the archaic alphabets for all the mid-front vowels, whether short or long. Many scholars hold that the

short vowel was a closed variety

, but this does not necessarily apply to the corresponding long vowel (see

below on h).

o in MnG is the mid-raised back vowel

of a rather open quality, which is higher than the vowel of English got.

This was probably also the value in OG, but as with E many scholars hold that the ancient value of the short vowel

was of a closed character

. For further discussion see below on w.

is today pronounced [i] but it originally stood for the high back rounded vowel [u]. In Attic-Ionic this sound was

fronted to [y] (see pp. 206f.). On the Boeotian use to render , see p. 118.

Among the long vowels the mid-raised front and mid-raised back representatives require discussion. (h) in early

Attic inscriptions (and in many local alphabets) stood for the glottal fricative (the aspirate), but in Eastern Ionic,

which had lost the aspirate (psilosis), the letter was used for a long mid-low front vowel

. This sound was

distinct from the high mid-front long vowel which came to be represented by

. This digraph stood in the earliest

period for a diphthong, while e denoted 1, the short vowel, 2, its lengthened forms, e.g.

= < es-nai

(see p. 264), the product of 'compensatory lengthening' and the contraction (p. 239) of e + e, e. g.

< , 3,

inherited long e, e.g.

and 4, the sound that developed in Attic-Ionic from inherited a (pp. 62f.), e.g. .

That this uniform writing concealed two varieties of long mid-raised front vowels is shown by their different

representations after the introduction of the Ionic alphabet.

(h) was now used for inherited e and the product of a,

and this indicated that the sound was of an open character (see also above on the onomatopoeic ). This is

supported by the fact that some varieties of island Ionic use (h) for the product of long a and keep it distinct from

inherited e, for which the notation is e, e.g. < kasigneta. This suggests a very open pronunciation

for h. The

< previous page page_205 next page >

< previous page page_206 next page >

Page 206

diphthong [ei], on the other hand, had become monophthongized by the fifth century with the value

The

sequence

thus became a digraph with this monophthongal value and it was used also for the products of the

contraction of e + e.

The closed quality of

appears also in its historical development: in the late fourth century B.C. there are

occasional spelling alternations between

and i, and these are the first pointers to the later development of to

[i:]. In this general 'itacistic' tendency

long preceded h. Transcriptions of Greek words in Oriental languages and

also in Latin indicate that h. retained its mid articulation until well into the Christian era. In Attic the first spelling

alternations between h. and i appear in the middle of the second century A.D..

From the fact that lengthened e is written

with a closer (mid-high) quality than h. it has been deduced that e,

too, had this closer articulation. This is open to doubt, for it is common for long and short vowels to differ in

quality.

The mid vowels of the back series closely parallel the history of the o-series. W, a character invented by the

Ionians about 600 B.C., had an open value

, as is indicated by the onomatopoeic for the braying of

the donkey and

for the cawing of the crow, and by the fact that the Ionic contraction of o + a is written w

The contraction of o + o in the earliest inscriptions is written o and

was reserved for the diphthong [ou]. During

the fifth century came to be used also for the products of the contraction and it would appear that by then

had become monophthongized to , paralleling [ei] > When the Boeotians adopted the Attic alphabet about

35º B.C., their use of

to represent their u-vowels ( in Attic having the value [y]) indicates that by that date

had been raised to [u:]. For structural reasons it has been argued that this change may have occurred much earlier

than the mid fourth century. If the Attic change of[u] to [y] took place in the period of the seventh to sixth centuries

B.C., there would have been an empty slot in the highest position of the back series, and the tendency towards

equality of acoustic distance would have favoured the change of the series

The diphthongs 1, and 2, , and were originally pronounced in accordance with the spelling, i.e.

there

< previous page page_206 next page >

< previous page page_207 next page >

Page 207

was a glide from the vowel indicated by the first letter to the high position front or back respectively. For Attic the

point must be made that the second element of the second series was not fronted to [y]. For the later development

of the

to a fricative [v] and other changes in the diphthongs, see p. 178.

Greek also inherited long diphthongs from IE and others were created within Greek itself by contraction:

In the course of time these diphthongs were eliminated either by shortening (p. 215) or by the loss of the second

element (p. 177). The modern practice of writing the iota 'subscript'

goes back to Byzantine times.

Consonantsk

In the system of the plosive phonemes Greek used a triple opposition based on the two features of voice (V) and

aspiration (A). Thus a, t, k were structurally (-V-A), b, d, y ( + V - A) and j, qc (-V + A). There was no series of

voiced aspirates (+ V + A) like the Sanskrit bh, dh, gh. In MnG jqand c represent the voiceless fricatives [f], [q]

and [c], but in classical times they were still plosives. That this was still so at the time of their first contact with

the Romans is shown by the transcriptions like Pilipus for

dracuma for and tus from .

Dionysius Thrax classifies the voiceless consonants

into three subclasses, I, smooth 1 , 2, shaggy

1 and 'media' . The pseudo-Aristotle De Audibilibus (c. 200 B.C.) applies the term to

sounds 'in which we expel the breath simultaneously with the sounds', an apt description of an aspirate. That the

sounds were plosives is also indicated by the use of

in early alphabets lacking the letters jc In Crete the feature

of aspiration was simply ignored, but in Thera and Melos the digraphs ah, kh were used, which led to the

redundant spelling qh for q. Yet another clue is afforded by the dissimilation of the aspirates ('Grassmann's Law',

pp. 230 if.):

for , cf. the alternation and . Again, in instances of expressive

gemination

1 The opposing terms and are used respectively of land covered with, or bare of, vegetation and

of animals 'furry' or 'hairy' as contrasted with ' smooth' or ' hairless'.

< previous page page_207 next page >

< previous page page_208 next page >

Page 208

a doubling of j is written aj that of q as

. Akin to this is the Greek loan , from the

Prakrit sakkhara. In transliterations of demotic Egyptian words it is significant that the demotic letters are retained

for the fricatives f and h and that j and c correspond to the digraphs ph and kh. In Coptic writing, too, j, q and c

correspond to voiceless plosives, and in the Armenian alphabet they render the aspirates p', t' and k'.

Finally, in syntagms like prepositions and nouns, if there is elision and the noun has an initial aspirate, the final

consonant of the preposition is written j

or q . Similarly, the forms and alternate

before words beginning with smooth and rough breathing respectively.

That the sounds classified by Dionysius Thrax as 'media' were voiced is clearly indicated first by their clustering: b,

d and y may combine in clusters

and the same is true of p, t, m, j, q and c (e.g.

) This rule extends also to sandhi phenomena, as indicated by spellings

like . Thus there are two mutually exclusive cluster-classes, and the distinction is evidently one of voice.

This is consistent with the above-quoted statement that the 'media' are intermediate between the voiceless stops and

the aspirated stops in respect to aspiration. In English, too, b, d and g are followed by considerably less emission of

breath than p, t, k, which are phonetically aspirated plosives. Further evidence in support of the feature of voicing

comes from Greek loan-words and transcriptions in Latin (barbarus, draco) and the transcriptions of Greek names

on Indian coins (Arkhebiyasa =

, Diyamedusa for ). For the later development of the voiced

plosives to voiced fricatives (b, d and g in Modern Greek have the values [v], [d] and [g]) see p. 178.

The pronunciation of the nasals m and v and the liquids l and r offers few difficulties. The articulation of labial and

dental nasals is succinctly described by Dionysius of Halicarnassus. The velar nasal [h] was a positional variant1

found before velar stops and it was written in the standard orthography as

though in earlier

1 The few cases where there is a distinction between [gg] and [hg] hardly justify granting/h/phonernic

status.

< previous page page_208 next page >