On Sat, 02 Mar 2002 12:37:30 +0800, Andrei Markine <
andrey@...>
wrote:
>At 3/1/2002 11:48 PM +0100, Piotr wrote:
>>
>>No, they hesitated between the _letters_ zeta (which originally stood for
>>[zd] rather than [z])
>
>Was it [zd] or [dz]?
This is what W. Sidney Allen says in "Vox Graeca":
(ix) Consonant-groups represented by single symbols
(a) <Z> There is fairly clear evidence that at quite an early period
the symbol ][, later Z, had come to represent the sequence [zd], as is
stated by the grammarians (e.g. Dionysius Thrax, Ars Gramm., p. 14 U;
cf. Dion. Hal., De Comp. xiv, P. 53 UR), rather than [dz] as it is
often pronounced by English classical scholars. Internal indications
of this are seen in the following facts: (1) The combinations Athé:nas
+ de thúras + de (with -de as in oîkónde) are represented by
Athé:naze, thúraze (cf. pp. 45 f.); (2) In most dialects, including
Attic, a nasal is regularly lost before the fricative s; thus, whereas
the n of sun is preserved before the stop d in e.g. súndesmos, it is
lost in sústasis. The same loss is regularly found before z, e.g.
súzux, suzên, and plázo: beside éplagxa, thus indicating that the
sound immediately following the nasal was a fricative and not a stop.
The [zd] value also incidentally adds point to the comic ô: Bdeû
déspota cited by Tzetzes, possibly referring to Aristophanes,
Lysistrata, 940, where the MSS have Zeû.
Prehistorically the combination represented by z derives in some cases
from an Indo-European sd [zd]; thus ózos 'branch' is cognate with
German Ast, deriving from osdos (cf. also Hittite hasd-); ízo: is a
reduplicated present from an original si-sd-o: (from which also
derives Latin si:do), related to the root sed- in the same way as e.g.
mí-mn-o: is related to méno:. But more often z derives from an
original dy or gy -- e.g. in pezós from ped-yos, ázomai beside ágios;
and these original groups must first have developed though an
affricate state, e.g. [dz^] (as in edge) -> [dz] (as in adze)[111]
(cf. Latin medius -> Italian mezzo); so that the presumed
pronunciation of these latter forms with [zd] represents a
metathesis of the fricative and stop elements. However, such
metatheses are of a particularly common type; R.P. wasp, for example,
derives from an earlier and still dialectal waps (cf. Old Prussian
wobse); and the particular change in question is closely paralleled
e.g. in Old Church Slavonic mez^da from Indo-European medhya:; an
intermediate stage must here have been medz^a, which has given Russian
mez^á 'boundary' (Russian méz^du 'between' is a borrowing from 0.C.S.,
being the locative dual of mez^da). A sequence [dz] would in any case
have been peculiarly isolated in Greek when it possessed neither any
other affricates such as [ts] nor an independent /z/ phoneme;[112] in
the sequence [zd], on the other hand, the [z] element would be a
normal voiced variant of the /s/ phoneme as in, for example, Lésbos
(cf, p. 46).
This having been said, it nevertheless remains probable that at the
time when the Semitic alphabet was adopted by Greek the 'zayin' symbol
was at first applied to a still existing affricate type of
combination; for it is difficult to see why a sequence [zd] should not
have been represented by sd instead of by a special sign; whereas,
since voice-assimilation in Greek is normally regressive rather than
progressive,[113] ds would not be a satisfactory representation of
[dz]; it has also been suggested that the affricated combination was
at this early period a single phoneme and so preferably represented by
a single symbol. Similar considerations apply to the Mycenaean Linear
B writing-system, which has a special series of characters
corresponding in part to the z of later dialects, and in part
representing a voiceless sound derived from ky for which an affricate
value of some kind is most probable.
The value of z as an affricate may also have survived in some of the
Greek dialects; in some early Cretan inscriptions we find it used to
represent a voiceless sound (? [ts]) deriving from ty; and forms of
the letter are used with a probable value [ts] in the native Oscan and
Umbrian alphabets. A voiced affricate value seems also to have been
known to late Latin speakers if one may judge from such spellings as
baptidiare for baptizare and conversely zebus for diebus.[114]
However, the metathesis of [dz] to [zd] must have occurred at an early
date in Attic and most other dialects;[115] and the
continuation of the [zd] value up to the 5th and early 4th. century is
indicated by the use of z to represent Iranian zd (e.g. O:romaze:s =
Auramazda in Plato, Artaozos = Artavazda in Xenophon).[116] Later in
the 4 c. we begin to find z replacing s used for Iranian z;[117] and
in Greek inscriptions there begin to be some confusions between z and
s (e.g. anabazmous 329 B.C.; cf. p. 46). This suggests that at some
time in the 4 c. the change to the modern Greek value as [z] was
already taking place; indeed it is probably referred to by Aristotle
(Met., 993 a) when he says that, whereas some people would analyse z
into s + d, others consider it a separate sound which does not
comprise already recognized elements. It has been plausibly suggested
by G. Nagy[118] that this change does not represent a normal phonetic
development but rather a dialectal replacement from the Koine (just as
ss replaced tt). Such a [z] would presumably have arisen from an
earlier [dz], and after short vowels at least the original
quantitative pattern is likely to have been preserved by gemination,
i.e. [zz];[119] this is also indicated by its representation as ss in
the early Latin borrowing massa = máza (cf. VL, P. 46).
The grammarians' statements of the [zd] value are of course of late
date and almost certainly reflect a grammatical tradition rather than
a continuation of this value in current speech.
It remains to mention that in the texts of Lesbian poetry medial z is
replaced by sd (úsdos = ózos etc.; initially also according to the
grammarians), whereas z is used for a result of synizesis in e.g. zà
from [dya] = dià. These spellings almost certainly represent a later
editing, based on the then general value of z, since they are not
found in early Lesbian inscriptions; but they point to the
preservation of the pronunciation [zd] in this dialect after it had
changed to [z(z)] elsewhere; and to the coexistence with it of some
other sound (? [dz] or [z]) of local origin, for which at the
editorial date z was the most appropriate writing.
[111] This is also a probable development for the cases where Greek z
apparently derives from an original y, e.g. zugón = Latin iugum.
[112] Cf. also Allen, Lingua, 7 (1958), p. 121, n. 40 and refs.
[113] I.e. a voiced consonant such as /d/ may account for a voiced
allophone of a preceding but not of a following /s/. Note that, for
example, in the aorist of a verb such as tríbo: (étripsa) it is the
plosive that is assimilated to the following fricative and not vice
versa.
[114] Cf. also M. Leumann, Mél. Marouzeau, pp. 384 ff.
[115] An Attic inscription of c. 480 has toisz(e) for toîsde where sz
is a geminated writing for voiced [zd], parallel to the frequent -sst-
for voiceless [st] in arissto:n for Arísto:n etc. Similarly dikaszoito
on a 6 c. Argive inscription (Threatte, pp. 527, 546).
[116] Attic inscriptions of the 5 c. show variation between single and
double z in forms az(z)eioi, buz(z)antioi, klaz(z)omenioi -- all
referring to places in Asia Minor. Just possibly this is an attempt to
represent an affricate of the type [dz]; a spelling ds would, by
recessive assimilation, be mispronounced as [ts], and 3 as [zd],
whereas a spelling zz [zdzd] would at least include the required
sequence [dz].
[117] The evidence is discussed in detail by M. Vasmer, Izsledovanie v
oblasti drevne-grec^eskoj fonetiki (Moscow, 1914).
[118] Greek Dialects and the transformation of an Indo-European
process, p. 127.
[119] This appears from its regular prosodic value in later verse, as
well as from fairly common use of sz for intervocalic z in Hellenistic
inscriptions (Threatte, p. 547). Gemination will, of course, only have
applied to intervocalic position within the word. Cases are found of
short quantity before initial (but not medial) z in later poetry: this
applies to all the cases mentioned by Maas, $123.
=======================
Miguel Carrasquer Vidal
mcv@...