Carpenter, Rhys_Early Ionian Writing_AJPh, 56, 4_1935_291-301

8/9/2019 Carpenter, Rhys_Early Ionian Writing_AJPh, 56, 4_1935_291-301

1/12

Early Ionian WritingAuthor(s): Rhys CarpenterSource: The American Journal of Philology, Vol. 56, No. 4 (1935), pp. 291-301Published by: The Johns Hopkins University PressStable URL: http://www.jstor.org/stable/289967 .Accessed: 22/02/2015 22:37

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

The Johns Hopkins University Press is collaborating with JSTOR to digitize, preserve and extend access to The American Journal of Philology.

http://www.jstor.org

This content downloaded from 1 32.248.9.8 on Sun, 22 Feb 201 5 22:37:20 PMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=jhuphttp://www.jstor.org/stable/289967?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/289967?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=jhup


2/12

EARLY IONIAN WRITING.

Although everyone is aware that the Ionic alphabet employedspecial symbols for long E and long 0, it is apparently notcommon knowledge that the approximate date of introductionof these symbols can be ascertained. The present paper arguesthat no student of early Ionian literature, and particularly nostudent of Homer, can afford to ignore the seemingly trivialhistory of these two letters.

As the Ionicalphabetic symbols

for X and o were not con-temporary inventions, they must be treated separately.

The Northwest Semitic (presumably Phoenician) parentscript employed a symbol, heth, from which Greek B was imme-diately derived, and gave to this the value of an aspirate. Boththe symbol and its approximate value were accepted and pre-served by practically all the Greek communities when writingwas introduced into the Hellenic world. But in the east of thearea in which the Ionic dialect

prevailed,on the Asia Minor

coast and its immediately adjacent islands, no isolable aspirate,no true " h " existed. Hence while the Semitic letter heth wasprobably called heta quite correctly by most of the Greeks, justas beth was called beta, it could only have been called 'eta bythese Ionians; and on the normal acrophonic principle thesymbol, which could not stand for "h"' in Ionia, should haveobtained the value of an "e ". As the Greek ear could dis-tinguish at least two qualities of the E-vowel in the

spokenlanguage, it was inevitable that the symbol E should be assignedto one and B to the other of these.

We may infer, in passing, that the Phoenician alphabet didnot enter Greece exclusively through Ionia, since otherwise hetaby becoming 'eta would have been lost as a symbol for " h" tothe rest of the Greek-speaking world. We may also infer thatthe supplementary (so-called non-Phoenician) symbols P andX were of Ionic invention, since the Ionians, like the otherGreeks, could distinguish two qualities of " p " and two quali-ties of "k ", but unlike the other Greeks (who could indicatethese sounds by P and rp, and by K and KB) they could notwrite these sounds in terms of their " h "-less alphabet. Hencea special symbol had to be invented for " p' " in distinction tonormal pi and another for " kEi" in distinction to normal "I ".

291


http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp


3/12

RHYS CARPENTER.

Since only the psilotic lonians would have been thus embar-

rassed, it must be they who invented this solution.It is tempting to push inference one step farther to concludethat the "East Greek" usage, being Ionian, must be primaryand hence older than the "West Greek." In that case, the"West Greek," and Latin, usage of the symbol " X " must havebeen derived from the combination Ic's ( X) with the i droppedas superfluous by a people who had no other use or knowledgeof the X-symbol. Having thus shut themselves off from usingthe

symbol" X "

as k', but desiring to improve onK

B, these" West Greeks " seem to have seized on the next available Ionicsymbol (Y) for their purpose. But they thereby deprivedthemselves of any symbol for the combination "p's" and werethus obliged to continue in the old way, writing (P', unless, likeOzolian Lokris, they chose to invent something entirely new forthe purpose. The apparently inexplicable differences betweenthe " East Greek " and the " West Greek" alphabets, familiarfrom

every epigraphical handbook, are thus not inexplicableat all.This series of inferences cannot be invalidated ab initio by

supposing that the Ionian cities originally accepted the Semiticvalue "h" for the heta symbol and only later sloughed it off.Since the alphabet does not appear in Greece until the finalphase of the Geometric period (the late VIIIth century B. C.),and since no one will be anxious to maintain that the AsiaticIonic dialect became

psiloticat so late a

period of its develop-ment, it follows that the Eastern Ionic cities could never haveused B as "h" for the simple reason that none of their wordscontained that sound.

Hence no Asiatic Ionic 1 inscription or manuscript, no matterwhat its date, can ever have employed the letter B or H as " h."

The converse dogma, that every Asiatic Ionic inscription ormanuscript, no matter what its date, must employ the specificsymbol B or H for

longE, is

equallytrue.

Superficiallyit

might be urged that the practice could have been introduced into

The evidence of inscriptions shows that "psiloticism," or theabsence of the aspirate, was characteristic of Miletus, Ephesus, Clazo-menae, Phocaea, Teos, Chios, Erythrae, and Samos; its existence is alsoimplied for Halicarnassus (Bechtel, Die griechischen Dialekte, III, pp.36-37).

292




4/12


Ionian writing at some later (though, of course, still very early)

date,since the lonians

mighthave

arbitrarily adopted theirneighbors' "IIH as a vowel at any period. But this argumenttoo is faulty; for the very existence of a special symbol for longas opposed to short E implies a luxury, not a need. Preciselybecause they had an extra symbol on their hands when thealphabet first came to them, the lonians were faced with anoriginal necessity either of discarding the superfluous symbolaltogether or else finding some use for it. This dilemma wouldnot

recur,once a method of

writinghad been established. Hence

the symbol E would either have been entirely missing fromAsiatic Ionic writing or from the start have had its value as eta.In agreement with this is the observation that no other Greekalphabet ever felt the need of inventing a special symbol forlong E (though in course of time they were all persuaded toaccept the Ionian symbol already in use).

The date of the introduction of the symbol for long E intoAsiatic Ionic writing is therefore the same as the introductionof the alphabet itself into Ionia. Even if the exact date of thisevent be disputed, the important conclusion still holds: no AsiaticIonic manuscript can ever have written 3 or H for "h" orfailed to distinguish between long E and short.

After this v-symbol (8) had been in use for several genera-tions, its form was abruptly changed to H. The reason for thechange is unknown; but the date of the event can be determinedvery closely. There is a considerable accumulation of archaic(sixth century B. C.) statuary from Milesian and Samian terri-tory, and nearly a dozen of these pieces bear dedicatory inscrip-tions.2 As the statues can be placed in chronological sequenceby their style, the inscriptions can be arranged in historicalorder. Fragments of the reliefs of column drums from theArtemisium at Ephesus, with remains of the dedicatory inscrip-tion of Croesus of Lydia, introduce a fixed point into this series,further strengthened by more or less certain historical identifica-tions of some of the dedicators of the other statuary, notably theequation of the Aeakes of one of the Samian dedications 3 with

2 The Milesian material: Pryce, Catalogue of Sculpture in the Dept.of Greek and Roman Antiquities of the British Museum, I, 1: B 273,B 278, B 281, with facsimiles of the inscriptions Figs. 166, 168, 170, 172.The Samian material: Buschor, Alt-Samische Standbilder (1934).

8 Athenische Mitteilungen, XXXI (1906), pp. 152-153.

293




5/12

RHYS CARPENTER.

the Aiakes of Herodotus.4 In addition, the progress of sculp-

tural archaeology has been sufficient to lend very considerablecredibility to Buschor's attempts to assign absolute dates to theSamian statues on purely stylistic grounds. Combining thisevidence, it becomes apparent that the open form of H can betraced back in Ionia to about 570 B. C., before which the closedform B is alone in use.5 With either form the value of thesymbol is, of course, invariably y.

The case for Q is entirely different. The specific symbol canbe traced back

onlyso far as the first

quarterof the sixth

centuryB. C. In the few Ionic writings earlier than 570 B. C. we findthe symbol "0 " used indifferently for the short and the longform of the vowel. The distinction between e and V is thereforeprimary but accidental, while the distinction between o and wis subsequent and deliberate, due to reasoning by analogy.Hence H has its immemorial (Semitic) position in the standardGreek alphabet while Q as a newcomer must trail along at thevery end of the

series,after the three Ionic

symbols4, X, 4',

which are already present in our earliest inscriptions.In the famous Abu-Simbel graffiti of the Greek mercenaries

of Psammetichus II we possess a dated document of 589 B. C.which I venture to reproduce once again in Figure 1, since allof its epigraphic content is seldom squeezed out of it. The datewas finally fixed by the discovery of the sarcophagus of Pedi-samtawi (the "Potasimto" of line 4) with the pertinentinformation that he was a leader of mercenaries for the

second,not the first, Psammetichus.6 Although there are among thewriters of these inscriptions natives of such purely Ionic citiesas Teos and Colophon, and though the author of the chiefmemorial also employs the Ionic script, there is no use of Q.We can only conclude that a separate symbol for long 0 was notin current use in 589 B. C.

But the inscribed potsherds from Naucratis show abundant

4 Herodotus, II, 182; III, 39; etc. The Thales, son of Orion (Brit.Mus. Cat. Sc., I, 1, B 281) can hardly be the famous philosopher, sincethe latter is called son of Examyas by Diogenes Laertius.

6 Cf. the study of this problem in Ath. Mitt., XXXI (1906), pp. 154 ff.by Curtius who, with the somewhat less extensive material then avail.able, reached much the same conclusion for the date.

6M. N. Tod, A Selection of Greek Historical Inscriptions, No. 4.

294




6/12


use of the new symbol. Here it seems to have escaped even the

archaeologists that the most important collection of these Nau-cratite inscriptions can be dated to a perfectly definite span ofyears. Herodotus writes:

Amasis gave to the Greeks who arrived in Egypt the town ofNaucratis to inhabit, while to those who did not choose to livethere but to carry on oversea trade he gave locations for settingup altars and precincts to the gods. .. . And the Aeginetans

B5AVTNoSNeo5MToo\40A1PNT

IYAMA/'i)at^9TnErl) ?N* ASi^^?WTIXolT?wDoArlNtaON" 04 K10k0yrLP EIVI ?TAM?ftN.1 \hor osooS+ E noroTA IMTTO irvT oS ^5S

[rSs\o v JSo,,

P$.N Ao18 Bl0V oNoW O uhgrl tQQ fc\ON/lot

Fig. 1. Inscriptions at Abu-Simbel (from Roehl's Imagines).

founded one of their own apart, a precinct of Zeus, and theSamians another one, of Hera, and the Milesians one of Apollo(II, 178).

In the days when it was still fashionable to be wiser than oursources, it was repeatedly pointed out that excavation hadestablished that the Greek settlement dated back to the reign ofPsammetichus I (651-610 B. C.), that the foundation date couldnot be later than 625 B. C., and that Herodotus must accordinglyhave been wrong in crediting its foundation to Amasis (570-526B. C.). Yet the obvious truth of the matter is that Herodotus

295




7/12

RHYS CARPENTER.

does not say that Naucratis was founded by Amasis, but that this

kingsettled certain Greeks there and

gaveothers the

necessaryground for constructing sanctuaries. The Milesians dedicatedtheir precinct to Apollo; and this very plot of ground was found,excavated, and indisputably identified by Flinders Petrie in1884.7 Within the precinct there were scattered remains of thearchitecture of two successive temples, and a trench was uncov-ered yielding masses of broken vases with dedicatory inscriptionsto the Milesian Apollo. The technique of archaeological dis-

coverywas at that time in the

making.Hence

Petrie,and after

him E. A. Gardner, could come to the conclusion that the firsttemple dated from the seventh century B. C., the second wascontemporary with the Erechtheum at Athens and hence fromthe end of the fifth century B. C., while the vases in the trenchformed a stratified deposit in precise and datable sequence. Allthree of these conclusions were erroneous.

The first temple actually dates 8 from the early years ofAmasis'

reign (i. e.,570-560

B.C.),the second

templedates9

just after the Persian occupation of Egypt (i. e., 520-510 B. C.),and the contents of the trench are a miscellaneous dump ofdiscarded dedications all thrown in and buried at one time.'0The earlier temple therefore exactly confirms Herodotus, who inturn makes it very unlikely that there was any Apollo precinctand hence any buriable dedication earlier than 569 B. C. Thedate of the second temple shows that it was the Persians whodestroyed the first temple,-as of course they would have

done,since the Greeks had come into Egypt both as traders dependenton Egyptian protection and as mercenaries paid for keeping thearmed foreigner out of the country and so must have espousedthe native cause against the Persian.11 The situation at Nau-cratis is precisely parallel to that at Athens fifty years later,when the Persians sacked the Acropolis and the Atheniansreturning thereafter gave their ruined dedications pious burial

7Egypt Exploration Fund, Third Memoir = Naukratis: Part I.s Brit. Mus. Cat. Sc., I, 1 (Pryce), p. 171.

'Ibid., p. 176, with the reference to Dinsmoor's original observationof the true date.

10It is impossible to imagine how a stratified trenchful of dedicationscould have formed in front of a temple.

11As Herodotus expressly says,-ol e7rKovpol oi roV Alyv7Triov (so.Amasis) 6ovres dav8pes EXXJ\ves e Kai' Kapes (III, 11).

296




8/12


9/12

RHYS CARPENTER.

Before tracing the literary implications of these observations

I wish to revert to the Abu-Simbel inscriptions for some furtherinformation on early writing.The third graffito in the series (Figure 1) can only be

transliterated:

T?JyAXe0o'' 'ypae? o 'IaAro . . .

even though this reading leads to the strange assumption thatthe soldier Telephos used one and the same symbol for L andfor G and another symbol (8) indifferently for long E and forII. Such a state of affairs is possible only where a script hasbeen hybridised. No alphabet would deliberately start out witha pair of indistinguishable symbols for unrelated sounds. InAthens, in the fifth century, precisely the same confusionreigned. If we find the symbol A in an Attic inscription ofthat period we must look at once whether we find I or Felsewhere in the writing. If L exists, A must be read " G";while if r exists, A must be read "L". The former followsthe old native Attic, the latter the imported Ionic tradition.Both cannot co-exist for long: the newcomer must oust hisrival or himself succumb. Similarly at Athens H may representthe aspirate or it may represent 7; that it should representboth or either of these indiscriminately would be an unen-durable and therefore unenduring condition. This parallel willgive us the clue to the trouble besetting Telephos the Ialysianat Abu-Simbel. Obviously the use of E as D is in the Ionictradition; hence B for "h" must be the older native Rhodianusage, just as A for G is attested in Ionic while the samesign is used for L in old-Rhodian, as inscriptions prove.13 Thereis then no puzzle here.

But philologists have been badly puzzled by a phenomenonoccurring in another of the Abu-Simbel graffiti (not shown inFigure 1), where some speaker of Doric has unmistakablywritten HEAau where grammar and phonology demand jhaaw(rov oTprpaTov). Similarly HEt has turned up for jdL, "I am,"on a grave inscription in Thera.14 There is only one reasonableor even possible explanation of this phenomenon; yet I am notaware that it has been given, though Gustav Meyer in his

13Att. Mitt., XVI (1891), p. 113; Roehl, I. G. A., no. 473.14 I. G., XII, 3, no. 769: Appwvos regi.

298




10/12


Griechische Grammatik (3rd ed., ? 4862) correctly saw that thecombination

HE must somehow be read as v. His advice hasbeen little heeded by epigraphists.A Rhodian or Theran or any other Greek whose alphabet was

passing through a period of Ionicisation could scarcely fail tobe restive under the awkward ambiguity involving such crucialletters as H and E. The Rhodian could perhaps do nothing toindicate whether he meant L or G when he wrote A/; but hecould do something to show whether B was to be read as H oras E: if he meant it for E he could

appendan " E " after

thesymbol,-a sort of determinative to fix it in its novel vocalicinstead of traditional aspirate function. Hence BE is to beread " i. e. r." B EXace s therefore not Acacreor 7XeAcAroryet 7Xaaoe,but just what it ought to be, ;Xaae. And 83EuL s notEL nor 'tu nor yet Ut, but just plain ,t, "I am." Thephilologist may rest in peace; the epigraphist should neverhave broken so rudely on his musings.

This hasadmittedly

been adigression.

But it is intended toshow that the classical scholar may come to some very curiousconclusions if his epigraphical knowledge or assistance be faulty.And so it is with the main subject of this paper, the use of Eand 0 in Ionic; for even the great Wilamowitz could write 15with that note of half-stifled fury and superior scorn which weretypical of the times quite as much as the man:

Denn die Mehrdeutigkeit der Zeichen E und 0, die fur das

Ionische genau so gilt wie fur das Attische (was die IUnbelehr-baren freilich immer noch nicht wissen, die bei solchen Variantenim Homertexte fortfahren von Atticismen zu faseln), ....

I can only bow humbly to the storm and admit my Unbelehr-barkeit. I believe that it is now self-evident that there musthave been an -} in Ionic from the start and certain that therewas an o from about 575 B. C. Even those who do not yet seefit to agree that there was no alphabet at all in Ionia before

725 B. C., no extensive use of writing before 650 B. C., and noadequate supply of paper and ink, and hence no books, beforethe sixth century,16-even these must make allowance for the

15 Sappho und Simonides, p. 84.21 So Th. Birt, Abriss des antiken Buchwesens, in Muller's Handbuch

d. klass. Altertumswissenschaft, I, 3 (1913), p. 277: "Das 6. Jht. hat

299




11/12

RHYS CARPENTER.

history of these two symbols. We must all say that if our

current text of Homer shows misunderstandings and misspellingswhich are rooted in the ambiguity of o, ov, and o, these mistakes,in order to be of Ionian origin, must antedate the year 570 B. C.;whereas if there are other mistakes rooted in the ambiguity ofc, et, and v, these cannot be based on an Asiatic Ionic text at all.It is well known that the ancient editors and commentators ofHomer thought that they had detected such mistakes in thetraditional text and thought that they knew the reason fortheir

occurrence, giving almost precisely the same explanationwhich has just been offered. They blamed the errors on thexerwayp;a,aLievot, the scribes who wrongly transliterated their O'sand E's because they were copying from a text which made nodistinction between o, ov, and o on the one hand and E,a, and 'qon the other. The following table exhibits typical instances ofthe various " metagraphic " mistakes which ancient or modernscholars have claimed to have detected:

Type of Ionic (oral) first written later copied intoerror: tradition: down as: our texts as:

?t foro B'P TO cteto? o Locrea H-f v^ej Puer lit Ky7o6V0S r"*t HAu

6L for V ^KVu/Evos A^|VI %Or Seirv AfrtVoset for ve /r IO/ A O .1) for ?e ove (P)oi

o IVrHoIov

i/y o?ov for C WY dS o0 , O ouy oVs

ou for o Vo,ros v0o50 YVoUro.s

O for ov H4deaou


12/12

EARLY IONIAN WRITING. 301

It is obvious whither these observations must tend-to a

categorical defense and unescapable acceptance of the ancienttradition that the orally transmitted text of Homer was firstreduced to writing in Athens in the sixth century B. C. But Ido not wish to argue that matter here, being content to haveshown that E and O may be fundamental factors for EarlyIonian literature and that the facts about them have to beunderstood. It is the scholar who merely brushes such thingsaside who is truly unbelehrbar.

RHYS CARPENTER.BRYN MAWR COLLEGE.

Date post:	01-Jun-2018
Category:	Documents
Upload:	the-gathering
View:	215 times
Download:	0 times

Carpenter, Rhys_Early Ionian Writing_AJPh, 56, 4_1935_291-301

Documents