Skip to main content

Full text of "The Language of Tragedy and Its Relation to Old Attic"

See other formats


STOP 



Early Journal Content on JSTOR, Free to Anyone in the World 

This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in 
the world by JSTOR. 

Known as the Early Journal Content, this set of works include research articles, news, letters, and other 
writings published in more than 200 of the oldest leading academic journals. The works date from the 
mid-seventeenth to the early twentieth centuries. 

We encourage people to read and share the Early Journal Content openly and to tell others that this 
resource exists. People may post this content online or redistribute in any way for non-commercial 
purposes. 

Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early- 
journal-content . 



JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people 
discover, use, and build upon a wide range of content through a powerful research and teaching 
platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit 
organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please 
contact support@jstor.org. 



IV.— THE LANGUAGE OF TRAGEDY AND ITS 
RELATION TO OLD ATTIC. 

Numerous treatises on the dialect of tragedy have appeared 
during the last half-century, but some phases of the question are 
still inadequately treated and some important truths do not 
appear to have received sufficient recognition. It is the purpose 
of this discussion to supplement the results already obtained 
by a presentation of some facts not yet noted, and to endeavor 
to establish from the facts at hand the proper deductions con- 
cerning the subject as a whole. Two parts of the tragic diction, 
the forms and vocabulary, have been here considered, and both 
have been viewed from a comparative standpoint. The com- 
parative table of Ionic and Doric forms in the trimeter of the 
three tragedians throws, it is thought, some new light upon the 
following points: the origin of the alien forms and the in- 
debtedness of Athenian to Dorian tragedy, the influence of the 
Doric and Ionic writers at Athens upon Athenian tragedy, and 
the fallacy that the use of un- Attic forms was due to the demands 
of metre. The comparative statistic of Attic and un-Attic words 
in the early Ionic and Attic inscriptions, and in the tragic, Ionic 
and Attic writers shows quite clearly the relation of the tragic 
diction to Ionic as well as Attic; while the study of the vocabulary 
of Aeschylus demonstrates plainly the relation which the diction 
of the dialogue bears to that of the chorus, the reasons for the 
disposition of words in the dialogue or chorus, and the explanation 
of the striking difference between the structure of the tragic 
trimeter and that of comedy. 

The Epic or Ionic coloring of the dialogue is well known and 
has been the occasion of extensive comment. 1 The presence of 
Doric forms also has been noted, but so far their occurrence has 
received comparatively little attention, although these forms are 

1 The best discussions of the dialectic forms are by Gerth, Curt. Stud. I, 2, 
193 ff., and by Smyth, Ionic Dialect, 74 ff. '. the most convenient compilations 
of the forms and their occurrence are by Gerth, 1. c, Koster, Studia tragico- 
Homerica (1891), Franklin, Traces of Epic Influence in Aeschylus (1895), 
Wittekind, Sermo Sophocleus (1895), Uhlmann, Propriet. sermonis Aesch. 
(1881-1892). Some earlier works of value are mentioned by Smyth, 1. c, p. 74. 



286 



AMERICAN JOURNAL OF PHILOLOGY. 



sufficiently numerous to impart a decided Doric tinge to the 
trimeter. 1 This may readily be seen by a comparison of some 
of the Doric forms with a like number of so-called Epic-Ionic 
forms. 



Ionic 




Forms. 2 A 


tesci 


^zivos 




i^EKUvaatv 


I 


el\lcr(ra 




tlvtKa 


2 


dv 




flOVVOS 




yovpara 




8ovpi\r]TTTOS 




&0vpi7TT]KT0S 


I 


Sovpews 




i^vvovpos 


I 


povvoiKa 


I 


Qprit 




QprjKrj 


I 


QprjKios 


2 


KaTaifSaTrjs 


I 


KkaieaKov (?) 


I 


Total 


II 



Soph. 
IO 



I 

13 

i 
i 



Eur. 
2 

6 

4 



27 



13 

7 

15 

i 

53 



Doric 
Forms. 

J A@ava 
yapopos 

ya7TQTOS 

yurofxos 
yanovos 
ya&ovarj (em.) 
yaixovziv 

yamdov (em.) 

dapos 
8atos 

tKUTt 

j. 

VULOS 

vapa 



Aesch. Soph. Eur. 
461 

2 

3 

1 



vapos 
onaBos 



Kvvayos 

Total 



o 
3 

7 
3 
2 
1 
1 
3 

35 



2 

1 

2 
1 

17 



30 

1 
S 

10 
6 

5 
6S 



From this table, which contains most of the so-called Ionic 
forms, it appears that the Doric forms are largely in excess of 
the Ionic, that Aeschylus is most Doric and Sophocles most 
Ionic. No adequate treatment of this Doric element has as yet 
been given, so that an analysis of these Doricisms, together with 
an explanation of their presence and of the excess in Aeschylus, 
seems to be required. This Doric element may be divided into 
five classes : 

1 Barlen, De vocali a pro v in trim, discusses most fully the Doric element, 
but he has restricted his work to such forms as show long a. A number of 
other Doric forms have been overlooked by all writers on the subject. 

8 It has been difficult at times to determine how often a given form occurs, 
as the editors by no means agree. However, it is a trivial matter whether 
Aeschylus has n or 12 Ionicisms and 32 or 34 Doricisms; the important 
thing is to know whether the Doric forms are much more numerous than the 
Ionic forms or not. 

3 0pjjf and congeners, which constitute the bulk of the Ionicisms in Eurip., 
are mainly in Hecuba (11 times) and the spurious (?) Rhesus (18 times). 



LANGUAGE OF TRAGED Y IN RELA TION TO OLD A TTIC. 287 

a) Local forms 1 (mainly Sicilian) as ydpopos, paa66s (cf. Kiihner- 

BlasS, I p. 157), dppoc (Ah. 385), pe\\d> (Ah. 389), pa (ch.), d8e\(pe6s 

Sept. 576 (Ah. 71, 123, Gerth, 1. c, 232), Brjv (Ah. 384), Zaoopai 

El. 818, eacrercu, OC IIl8, ptaaos, Se6(T(rvTos, dopvvo-oov, eaavdrj (Ah. 

99 on forms with -0-0--), n-d/«nu (Sicilian acc'd to Wilamowitz, H. F. 

II 229) ; local words, as daxibapos, \iTpo(TK.6iros (Ah. 391), TlaXiKoi, 

Korrafios (cf. Weber, Anacreontea 87), \dra£ Soph, fr., fiaao-dpa 
(cf. Et. M. 190, 58, Hesych. s. v.); and some terms apparently 
Spartan or Cretan, as epSopayerds (Gerth, 1. c, 265), tyopos (ch.), 

kchtis (possibly KaatyvtjTos and KamyvrjTais), oliceve and dye'Xat. 

b) Military or hunting terms, as Xo^uyot, Kvvdyot, nwdyla, vohayos, 

avynvvdyds, and Xo^uyeV^s. 

c) Forms which have the Doric 5 as yd- compounds, the forms 
above in the Doric column, and &p6s, Xdds- (cf. Kretschmer, KZ 
XXXI, 290), vfios (Ah. 51) npiopos (cf. Wackernagel, KZ 
XXVII, 263, XXVIII, 132), &a\<k (Horn. Pr,\6s), |3Sre (Kiihner- 

olasS, II, 380), iKerddoKO?, 7roipuTG>p, iroivdcrdpeaSa, poddytvrjs, av&d<rop t 
6oiv(t(r6p€(r6a, Ooivdraip, BotvaTrjpiov, papeprtia, Tropnapa, TTpoairopTTuTds 
(ch.), iropnaaop (Kiih.-BlasS, § 238, 3), Xeopro&apayv, diftdpos, dyt)(ri\dos 
IT., and ireirupevGi* 

d) Scattered forms, many of which are found in other dialects, 
as -aa-, - P a- (Kiihner-Blass, 1, 147), <riv, £w6s (Arcad. and Argive), 

ni&oiKos Aesch. fr., nori, rot, eyav Aescll. Suppl. 740, o-(p<f, a(ptv, a<f>6s, 

TTToXit and compounds, Ktnp (so in Pindar but k%> in Homer), 
■noraivios (Phot.), ojmirra (Laconian inscr.), aUv, aUl, generally 
emended to del (Ah. 378-9), x«'peo-<rt (Ah. 229), n-dXeoy (Ah. 236), 
noioq, deiSo), deiKTjs, deipco (Ah. I93)> vdov, 7tvp7Tv6ov, SidirXoop, KnWippuov, 
opeo-Koov (Ah. I94), ira>\eipepai (Ah. 214), eKpvcpSev (All. 317), 

1 The references Ah. are to Ahrens, De Dialecto Dorica, on which I have 
relied mainly for the forms of the Doric dialect. In this article no attempt 
has been made to determine the dialect of any form for myself. 

- The form oikt/oc in Solon cannot be taken as authoritative. As Smyth, 
1. c. 68, observes," an oiKEOf might readily have been transcribed okr/of because 
this word was antiquated even in Solon's time". It is very probable that this 
law of Solon was taken from Crete, as the word is .Cretan (Gortyna code), and 
the form oiaeog is the Doric genitive (cf. Ah. 233, 236), so that neither word 
nor form can rightly be considered either Ionic or Attic. 

3 Of these terms dykXcu and o'ikcvc were common in Crete, ayiTitu and K&aig 
in everyday use at Sparta. As designations of the whole body of Spartan 
youth they were prominent constitutional terms of the most powerful state of 
Greece, and doubtless as well known at Athens as "ephoroi" or "satrap": 
naotc, like £0opoc, in tragedy can be drawn only from Sparta. 



288 AMERICAN JOURNAL OF PHILOLOGY. 

Xpioeov (Ah. 121, 194)1 X a ^ Keou i fipfTta (Ah. 235)> ^rjpios, (piaios 

(Ah. 231), iatav (Ah. 317), etc. 

e) Apocope and possibly some cases of psilosis (Ah. 353, 

36-7)- 

The assignment to the Doric dialect of such of the above forms 
as occur in the epos is due to reasons presented below. Of these 
Doric forms, those of class a owe their appearance in tragedy 
partly to Aeschylus' residence in Sicily, partly to subject-matter, 1 
but mainly, I believe, to dithyrambic or choral poetry. Those 
of class b, on account of the military supremacy of the Spartans, 2 
may have been in vogue in Attic as early as the time of Thespis, 
but the statement of Phrynichus, o< p.iv rpayiKol iroirjTiu . . . Sapi- 

£ov<ri, to rj els a /leTandeirrfS, Kvpayos, oi 8c ' Adrjvaioi . . . tu rj <pv\aTTOV<nv, 

olov KwriyiTTjs, would indicate that the tragedians made use of these 
well-known forms solely to give, by means of the 5, a Doric tinge 
to the trimeter. Most of the forms of class c appear to have 
been chosen for no other reason than to give a Doric coloring. 
In the use of these forms the "necessities of metre" manifestly 
had no influence, for the Doric and Ionic or Attic forms were 
metrically equivalent. It may then reasonably be supposed that 
the same aesthetic reason prompted the use of the other forms 
of this and those of the remaining classes. The query which 
naturally arises here, why the Doric forms were used in preference 
to the Attic, finds its answer in the reply to the question why 
Doric u appears in the chorus. 3 The presence of so many forms 
can hardly have been due to Aeschylus' residence in Sicily, nor 
to possible association with Pindar. Some support is given to 
this view by the excess of Doric forms in Aeschylus, but it is 
reasonably certain that the rules for tragic composition were 
formulated long before he left Athens, and it is not at all 
probable that the effect upon Aeschylus of his stay among the 
Dorians could have given a permanent coloring to such an 
artistic production as tragedy. 4 The Doric tinge of the trimeter 

1 The influence of the subject-matter is seen also in the use of Egyptian 
j3apic in the Supplices, ridpa in the Persae, and in the new words of the 
Bacchae, ftdicxevois, jiaKXfviia, jjaicxevaifio^, etc. Cf. Hermann, Opusc. II, 101-2, 
on the foreign air of the Persae. 

2 Cf. Lobeck, Phrynichus 430, Hoffmann, Gr. Dial. Ill, 308. 

3 " In den Chorgesangen tier Tragodien ein stilvoll abgedampfter Dorismos 
an die alten Zusammenhange mit der dorischen Lyrik mahnte". G. Meyer, 
Gr. Gr 3 . p. 3- 

4 The strongest Sicilian influence upon tragedy is probably that of Stesich- 
orus. " Erzahlte er (Stech.) die Mythen in lyrischen Versmassen und liess 



LANGUAGE OF TRAGEDY IN RELATION TO OLD ATTIC. 289 

must be explained, I believe, in the same way as the Doric 
coloring of the chorus, viz., as a royalty to a Dorian invention. 
Some form of the dialogue seems to have existed in the dithy- 
ramb. Suidas, S. V. 'Apiap, Says : Xeyerat . . . opopAaai. to qdopevop 
viro tov \opov Kai aarvpovs *lo~€PtyK(tv enfitrpa Xiyovras. 10 the Same 

effect is Aristotle's statement, Poet. c. 4 : <a\ 17 p.er(Tpayq>&ia) dn6 
ray i£apx6vTav tov St0vpap.@ov. Whatever the origin of the dithy- 
ramb "its elevation to the rank of artistic poetry" as Haigh, 
Tragic Drama, p. 16, notes, was due to the Dorians. " The 
Dorian stamp is upon all choral poetry in its language, rhythm, 
and metre" 1 , and there is no reason to suppose that the dithy- 
ramb formed an exception to this rule. The connection between 
the Doric dithyramb and Attic tragedy seems clearly established, 2 
and Arion 3 as well as Epigenes were by common tradition 
counted as the first of the tragic poets. The common statement 
that " in Greek literature different kinds of composition adhertd 
generally to the dialect in which they started " (Rutherford, 
New Phryn. Introd., p. 3) is exemplified in tragedy, and the Doric 
color of the choral odes is generally ascribed to the imitation of 
a Doric original. In my opinion Doric 5 must have the same 
significance in the dialogue as in the chorus, and the Doric tinge 
is an acknowledgment by the Athenians of their obligation to a 
Doric creation. It is not reasonable to suppose that the Attic 
tragedians, if they had introduced the dialogue into tragedy, 
would have given such a Doric cast to the speech of epic 
characters and to the iambic, that is, Ionic metre. 

Since, then, it is generally conceded that Athenian tragedy is 
a development of Dorian originals, it is surprising that -<ro--, o-vp, 
jrori and other forms of class d should have been called Ionic. 
These were native Doric forms. As such they would of course 
appear in choral poetry, and are no more epic in the dithyramb 
than in the Doric prose inscriptions. Why, then, should they be 
called Ionic in the Attic dithyramb or in the choral (Doric) odes 

sie von Ch8ren an den religiosen Volksfesten vortragen ... so dass (seine 
Gedichte) auch in Attica vielverbreitet und namentlich von den Tragikern 
vielbentitzt wurden". Christ, Gr. Literat. (1890), p. 136. 

1 Smyth, Greek Melic Poets, xxiii. 

'Cf. Haigh, 1. c, 25. Schaefer, De Dorismi in trag. Gr. usu, p. 3, says: 
" Ceterum non potest dubitare quin tragici in eligendis Dorismis aliquatenus 
dithyrambicorum secuti sint auctoritatem ". 

3 On Arion see Christ, Gr. Lit. (1890) 135, Haigh, I.e., l6ff., and Paully- 
Wissowa, s. v. 
20 



29O AMERICAN JOURNAL OF PHILOLOGY. 

of tragedy ? That some of these forms also appear in Homer 
does not make them any less Doric. This agreement enabled 
them to serve a double purpose, in that they are appropriate to 
the heroic characters of tragedy, and at the same time show a 
deference on the part of the tragedians to Doric originals. All 
the alien forms of tragedy, as Gerth, 1. c, p. 268, has shown, are 
found in Doric choral poetry. Many of these alienisms never 
appear in Homer. Here is agreement with the one and dis- 
agreement with the other. If the epic influence upon tragedy 
is so strong that ttotL and -<r<r- cannot be called Doric, how comes 
it that the tragedians in striking opposition to Homeric usage 
employ fia\6v, ekoti, &uos, yurd/xof, etc.? Or why the frequent use 
of Doric vlv 1 to the exclusion of epic fuV? These usages show 
clearly that to the tragedians Doric was the important dialect. 2 
According to the law of Greek literature already mentioned 
Athenian tragedy should show some of the characteristic forms of 
the Dorian originals. It is not credible then that the tragedians 
who had received 5, 0-0-, wort, uw, etc. in choral poetry, should 
retain of these only 5 as a mark of indebtedness to the Dorians, 
reject by some occult means the other Doric forms, and then 
import the same forms from Ionia. In these professed imitations 
or developments of Dorian originals it is not necessary to prove 
that the borrowed forms are Doric. The burden of proof rests 
with those who hold that in such a tragic form as 8dkacr<ras the 
ending may be Doric but the sibilants are Ionic. In all the 
works on the dialect of tragedy I have failed to find any reason 
for such a view. So long as 5 stands in the chorus, so long 
should -<r<r-, a-vp, etc. remain as memorials of the Doric origin of 
tragedy. 3 This applies also to apocope which is a mark of Aeolic 
and Doric as contrasted with Ionic and Attic. Kirchhoff, accord- 
ing to Smyth, Ionic Dialect, 273 n. 1, believes that the instances 

1 Smyth, Ionic Dialect, p. 445, n. 2, says viv was doubtless Old Attic, for 
what reason I do not know. The form occurs about 65 times in the trimeter 
of Sophocles. 

! If it be asked why Jelvof and /iovvos or Bpy!; appear in tragedy, it may be 
said in reply first, that these forms are found in the Doric lyrics and may 
rightly appear in tragedy as imitated forms ; and secondly that Doric occasion- 
ally shows u and ov. Cf. Ahrens, 190, Keiv6s,5eiviadae, etc. 

3 Kirchhoff, acc'd to Smyth, p. 306, holds that the presence of 00 is due to 
textual corruption, and Fick, BB XIV 253, would change oa to tt in Attic 
writers. If these views are correct, all other Doric forms should be eliminated 
from tragedy. 



LANGUAGE OF TRAGEDY IN RELATION TO OLD ATTIC. 29I 

of apocope in the Attic poets are survivals of a period when Attic 
had not yet developed an artificial objection to its presence. 
Rather is it a mark of the dithyrambic poetry, the practices 
of which poetry, tragedy was free, or was forced, to imitate, 
and to some extent an evidence of the long residence of Aeschylus 
among a people who used it freely. There is no reason for 
regarding it in tragedy as an imitation of Homeric rather than of 
Doric usage. 

The extent and importance of the Doric element in tragedy 
is thus made apparent, and it appears, I believe, that the Doric 
coloring of the dialogue was designed to be but little less than 
that of the chorus. Koehler, De Dorismi . . . apud Aesch. 
necessitudine, p. 4, finds that some parts of the choral odes were 
composed without a single Doric form. In Ag. 40-102 there 
are, according to Haigh, Tragic Drama, 367, only four Doricisms, 
and in the parodos of the Antigone only four in 27 lines. But 
the Doric of the chorus is in a way differentiated from that of 
the dialogue. In the latter the Doric tinge is obtained by the 
use of Doric words and of 5 for ij in the stem, whereas the chorus 
exhibits the Doric cast most strikingly in the ending. The 
aesthetic effect in the dialogue was obtained in more limited 
ways, and this restriction seems to have been due to the use of 
the iambic metre. The iambics of the new Ionic, approaching 
as they did so closely to prose expression, could not tolerate so 
many of the forms of Dorian poetry as could the metres of the 
choral odes. 1 ij was a characteristic feature of the Ionic iambics, 
a a mark of the dithyrambic dialogue. The old race strife which 
broke forth was settled by a compromise, the use of Attic 5 after 
f, 1 and P , with an occasional form of Doric poetry for aesthetic 
effect. 

In addition to the Doric forms, tragedy contains a few cases of 
psilosis, some twenty cases of the omission of the augment, 2 some 
half dozen Aeolisms, a few doubtful Ionisms as above and some 
fifty forms called poetic, 3 which can hardly be assigned to any 
dialect. Gerth, Curt. Stud. I 2, 268, as seen above, noted that all 
the alien forms of tragedy are found in choral poetry. This coin- 

1 On the rhythms of the dithyramb cf. Smyth, Melic Poets, p. lv. 

*Cf. Lautensach, Gram. Stud, zu den Gr. Tragikern undKomikera(i8o,o,)l8l. 

3 These are mainly verb forms as wfrvw, /lifivu, nine aorists like duaffeiv, 
avaax^Buv, <f>doc, (jtaeafdpog and incai, fjSi, I6e, ala, yala, a<j>dg, rijiiv, iro?J.6<;, 
aire, etc. 



29 2 A ME RICA AT JO URKA L OF PHIL OLOGY. 

cidence explains satisfactorily the use of these few Ionic and 
Aeolic forms by the tragedians. It is true that these forms are 
quite appropriate to the heroic characters of tragedy, but the 
presence of such personages merely rendered more acceptable the 
continuance in tragedy of some characteristics of the Dorian 
models, that is, the use of an occasional epic form. 1 It is held 
also that plagiarism and the Ionic source of the iambic metre 
affected the tragic diction. Quotations from the Ionic poets 2 
occasioned at most the use of a few poetic words or forms, and the 
iambic metre merely served to restrain the use of the Doric forms. 
Rutherford, as we have seen above (p. 289) rightly remarks 
that "in Greek literature different kinds of composition had a 
tendency generally to adhere to the dialect in which they started", 
and Smyth, Ionic Dialect, p. 69, (also Wittekind, 1. c, p. 3) suggests 
that the Ionic source of the iambic metre should then have given 
an Ionic coloring to the dialogue. That the rule however was 
not inviolable is shown by the fact that there is no Ionic coloring 
in the dialogue of comedy. Here the characters employ the 
dialect appropriate to them, and neither metrical considerations 
nor earlier models avail to force inappropriate speech upon any 
character. For this reason the iambics of comedy do not show 
an Ionic coloring or a Doric cast. The tragedians however ad- 
hered to the practices of choral poetry, putting, strange to say, 
Doric forms into the mouths of epic personages, because they 
found already well-established a Doric-epic speech. This form of 
speech which was eminently appropriate in such poetry as the 
Stesichorean versions of epic tales, precursors of tragedy, had 
become familiar at Athens, and was retained in the Attic dithy- 
ramb and tragedies without objection on the score of propriety. 
In no other way is it possible to explain the use of Doric forms 
by the epic characters of the dialogue. Where comedy, as in the 
chorus or the paratragedic parts, had precedents, there also was 

l "Stesichorus created a High-Doric dialect by combining epic with Doric". 
(Smyth, Melic Poets, p. 258). The epic element in lyric poetry is generally 
admitted, so that a discussion here is unnecessary. Cf. Brugmann, Gr. Gr. 
(1900) p. 19, and the works there cited : G. Meyer 3 , p. 3: Farnell Greek Lyric 
Poetry, p. 77. On the union of epic and lyric elements in the drama see 
Zarncke, Die Entsteh.derGr. Literaturenspracten, pp. 5-8, Azelius.De Assim. 
syntact. apud Soph. (1897) p. I; and on the influence of Stesichoruson tragedy, 
Christ, Gr. Lit. p. 136, Mahaffy, Gr. Lit., I 1, 225, Smyth, Melic Poets, p. 258. 

s Cf. Clemens, Alex. Stromat. VI, p. 738 (Pott), Donaldson, Theatre, pp. 56, 
59 n. 1, Verrall, JHS I, 260, II 179. 



LANGUAGE OF TRAGED Y IN RELA TION TO OLD A TTTC. 293 

adherence to earlier practices. Another thing which may have 
affected Athenian poetry was the presence at Athens of Anacreon, 
Simonides and Lasus at the very time when tragedy was being 
developed on Attic soil. Gorgias seems to have influenced 
strongly Attic oratory 1 at a time when Athens was the literary 
center of Greece, and it is quite possible that at an earlier time 
when the literary and political position of Athens was insigni- 
ficant, the presence of such a poet as Anacreon would have ap- 
preciably affected Attic literature, had it not been that his 
influence was restricted and ultimately eclipsed by the popu- 
larity of the Dorian lyric, possibly by the favor shown the 
choral songs of Simonides and Lasus. However this may be, 
Dorian poetry prevailed. Anacreon left Athens but the Dorian 
remained. The presence of the latter can hardly have been with- 
out effect upon Attic poetry. Lasus was at Athens during Aes- 
chylus' infancy and down to the time when the latter began his 
poetical career. By reason of his innovations in music, 2 his posi- 
tion at the court and his nationality, his commanding position in 
the dithyramb — even being called its inventor, 3 and his personal 
association with the first tragedians, he was, I believe, a deter- 
mining factor in the form of Attic representations of Doric inven- 
tions. How he affected the language of tragedy cannot now be 
determined, but it is probable that his presence and activity in the 
dithyramb tended to preserve the Doric element of the earlier 
models. Such influence as the Ionic poetry may have exerted 
upon the early tragedians was further minimized by the subse- 
quent removal of Aeschylus to Dorian territory. The Ionic ele- 
ment in Sophocles, as seen above, is much greater than in Aes- 
chylus, and this is quite possibly due to the fact that when the 
former was subjected to Ionic influence — by association with 
Herodotus, 4 — there was no counteracting influence. Thus early 
did Dorian supremacy assert itself at Athens, and it is singular 
that the forms of literature in which the Athenians became espe- 

1 Cf. Wilamowitz, Phil. Untersch., VII 312, Maass, Hermes, XXII 566 f. 
Blass, Att. Bered. I 56 f. Nieschke, De Thuc. Antiph. disc, takes, and I 
believe properly, an opposing view. 

2 Cf. Plutarch, De Musica, 29. 

3 Schol. to the Birds, 1403. For the facts concerning Lasus see Christ, Gk. 
Lit. (1890) p. 157, Smyth, Melic Poets, p. 299. 

4 Cf. OC 337 and Hdt. II 35 (Cf. Jebb, ad loc), Ant. 905 and Hdt. Ill 119 
etc., Haigh, Trag. Drama, 136, n. 2. 



294 A MERICAN JO URNAL OF PHIL OLOGY. 

cially pre-eminent, the drama and oratory, should have been 
developments of Dorian originals. 

These circumstances explain that foreign air in the tragic 
diction which was noticed by Herodian who says: n-epl 'Op6oypa<f>ias 

497 v^-v • ovSev flavp-atTTOj*, ol yap rpaytKol TvonqTiKais Xe^eaiv duidaoi 

KexpwSai. But Rutherford in the Introduction to the New 
Phrynichus, pp. 3-4, maintains that " the basis of the language of 
tragedy is the Attic of the time when tragedy sprang into life . . . 
It must however be remembered that the tragic poetry of Athens 
contained words, expressions and metaphors which it would be 
ridiculous to employ in other species of composition or in the 
course of ordinary conversation " 1 . It is difficult to speak soberly 
of the last statement, which seems however to be rather a careless 
expression than an accurate representation of his position. The 
fallacy of Rutherford's theory is very clearly shown by the testi- 
mony of the surviving monuments of Old Attic. This matter 
has received most recently a thorough treatment by Smyth, Ionic 
Dialect, p. 66 f., who finds that not only was old Attic free from 
the characteristic Ionic forms 17 after e, t, p, ov for o, ev for ov, « for e, 
but that such forms as £eiW and p.ovvos are pure Ionic and never 
existed in Old Attic at any period -. The inscription from Sigeum 
which is little later than 600 B. C. shows the two dialects were dif- 
ferentiated at that early period, and is a severe criticism of the 
theory that Attic speech two generations later was still Ionic 
in form and vocabulary. But Haigh, Tragic Drama, p. 365, 
regards Rutherford's view as beyond dispute, so that further 
consideration of this theory seems demanded. Rutherford could 
hardly have put forth this view, if he had known the extent of 
the foreign element in tragedy. In Aeschylus alone there are, I 
estimate, about 5800 poetic or so-called Old Attic words. It is 
not possible that so many words survived for centuries in the 
colloquial Attic, but as soon as they were extensively employed 
in tragedy — and certain parts of comedy too — as soon as litera- 

1 If the language of tragedy is to throw any light on Old Attic, it must be 
reliable, viz., pure, otherwise it will be impossible to distinguish Old Attic 
forms from Epic and Doric importations, or from Aeschylean and Euripidean 
inventions. Haigh, 1. c, p. 366, attempts to select epic and Old Attic words, 
but there is of course no basis for such selection. 

2 Un-Attic forms of foreign names and in metrical inscriptions require, of 
course, no discussion. On the foreign element in Athens see Plato, Crat. 406, 
Lys. 223 A, and for a general discussion of this element, Kretschmer, KZ 
XXIX 38 iff. 



LANGUAGE OF TRAGEDY IN RELA TION TO OLD A TTIC. 29S 

ture developed they forthwith dropped out of use. Such a theory 
is wholly at variance with the principles of the conservation of 
language, and requires evidence of the most convincing kind to 
substantiate it. On the contrary, Rutherford offers only the 
statement that the language of tragedy has many points of resem- 
blance to the language of Herodotus and Hippocrates, and there- 
fore is Old Attic 1 . If this reasoning is valid, one may prove in 
the same way that the Doric element in the chorus is a survival 
in Attic of the language in vogue when Attic and Doric had not 
yet separated 2 , or that the numerous Attic forms in Herodotus 
are survivals from the period when Ionic had not yet separated, 
from Attic. It is necessary first to prove that these alien forms 
could not have crept into tragedy or Ionic in any other way. 
But Herodotus and Hippocrates are rather questionable authori- 
ties for determining matters of dialect. The ancients noticed 
their use of ovofxara yXoxro-n/iimKa 3 . Both were Dorians who trav- 
eled extensively, both are said to have been in Athens, and the 
works of both have a decided Attic coloring. Rutherford refers 
indifferently to both the spurious and the genuine works of Hip- 
pocrates, although it is impossible to see how the usages of cer- 
tain divers medical writings of Hellenistic or Roman times are 
of any value in determining the character of 6th century Attic. 
Herodotus was called the Thurian by Aristotle 4 and Pliny says 
the history was composed at Thurii. Naturally there would be 
some resemblances to tragic diction in this Qovptas ('Aminos) X070- 
jtoio's, but these are in a measure signs of Athenian residence and 
association with the tragedians. The theory of Rutherford then 
is based largely on the diction of a tragedian who lived for a time 
and wrote some of his works among Dorians, and on that of a 

1 Schulhof, Attic, Ionic and Tragic, pp. 1, 6, 7, holds that the basis of the 
language of tragedy is to be found in Archilochus and Simonides of Amorgos. 

2 The review of Old Attic by Smyth, 1. c, also refutes clearly Baden's claim 
that Old Attic abounded with forms showing Doric a. Barlen's theory is more 
objectionable than Rutherford's, for the view of the latter is based on the fact 
that Ionic and Attic were at no very early date one dialect, whereas Attic and 
Doric separated long before Attic and Ionic became distinct dialects. 

3 Cf. Rutherford, p. 161, and on the epic element in Hdt. Mure, Gr. Lit. IV 
App. q, Hoffmann, III 186, and the works and grammarians cited by Smyth, 
80. 91. 

4 Cf. Christ, Gr. Lit. (1898) p. 328. The title " Thurian " indicates not so 
much long residence, as Strabo holds, but renown acquired while a resident 
there. A man is called after the scene of his achievements, not after the 
scenes of his inactive life. 



296 AMERICAN JOURNAL OF PHILOLOGY. 

Dorian who lived a long time and possibly revised his whole his- 
tory among Athenians. The value of deductions from such 
facts is apparent. 

This matter of foreign residence and its effect upon purity of 
speech is of importance in the study of ancient authors. The 
Greeks themselves in passing judgment upon the dialect of any 
one took into consideration foreign residence, and a glance at the 
Greek writers shows the need of such action. Lysias, Isocrates, 
Isaeus, Aristophanes wrote among Athenians and their speech is 
comparatively pure. The Greek of the men who traveled or 
wrote among foreigners is markedly irregular and impure. 
Xenophon and Aristotle need but a mention. Thucydides and 
Andocides present many divergences from the Attic norm ; Hero- 
dotus, Hippocrates and Solon 1 present a hopeless confusion as to 
dialect. Campbell in the Introduction to his Plato notes that the 
poetic element increases in the later dialogues when the philoso- 
pher is more exposed to the influence of Magna Graecia. The 
language of Xenophon differs in many respects from the Attic 
norm, and Rutherford very properly refuses to accept as authori- 
tative Attic all his usages. The same treatment should certainly 
be applied to all writers who transgress in the same way and for 
the same reason. It is apparent then that not every word and 
form on Attic soil are to be accepted unconditionally as old Attic, 
and those who would on the strength of a few isolated forms recon- 
struct old Attic are rearing an inverted pyramid, and that too on a 
very insecure apex. 

It remains to speak of one more theory concerning the 
occurrence of alien forms in Greek tragedy, viz., the influence 
of metrical considerations. Smyth, Ionic Dialect, p. 76 says 
that "the necessities of the trimeter, not the requirements of 
emphasis, decided the question as to whether the Ionic or the 
Attic form should be admitted." Jebb, OR 1418, expresses the 
same sentiment. This theory would refute Rutherford's view of 
the source of the alien forms of tragedy, but the influence of the 
metre seems to be greatly exaggerated. Dialectic forms and 
words are not employed in the dialogue of comedy, although the 

1 Thucydides and Solon are possibly affected also by earlier works in their 
respective lines of composition. The occurrence of ac in Thuc. and 
Antiphon may be ascribed to the influence of the usages of Ionic historical 
writers or to the Sicilian rhetoric, but I believe that the foreign residence 
of the one and the anti-democratic tendencies of the other made them more 
ready to accept the usages which prevailed in all Greek literature. 



LANGUAGE OF TRAGEDY IN RELATION TO OLD ATTIC. 297 



metre was the same. But one poetic form, the ending -\u08a is 
ever admitted into comedy. Furthermore the presence of yap.6por, 

yaro/iot, Sapos, eKart, /35\oV, jfivptv, (fidcvvoc = (pativoc, and many Similar 

forms is sufficient evidence that metrical demands were not 
the determining factor in the selection of forms, yii- is used 
in the dialogue but yij- in the chorus. Editors would not 
tolerate this variation for a moment were it not for the testimony 
of the grammarians. |«»-or and poivos, ata and yam are used for 
the same reason as yapopos and yaropoc, Sapos and fia\6i>. There 
were then no exigencies of metre which were superior to aesthetic 
considerations, and the diction of tragedy is not determined by 
metrical necessities, but results from a free selection of forms which 
show at once an appreciation of the need of suiting the speech 
to the character, and above all a knowledge of indebtedness to 
foreign originals. 

No less striking than the results obtained by a comparison of 
the Doric and Ionic forms are those obtained by a comparative 
study of the tragic vocabulary. Some attention has been given 
to the vocabulary of a few Greek writers, but the results have 
been relatively of little value because of lack of comparison with 
others. The abnormal character of the tragic diction may readily 
be seen by a comparison with that of other writers. In every 
author the words considered are taken in alphabetical order. 

Of 500 words from Aeschylus 383 are un- Attic" or 76 per cent. 
112 " " Antiphon 10 " " " 9 

120 " " Hipponax 26 " " " 22 

100 " " Thucydides 22 " " " 22 

96 " " Archil, iamb. 23 " " " 24 

143 " " " any metre 44 " " " 31 

100 " " Herodotus 26 " " " 26 

400 " " Attic Inscriptions prior to 445 B. C. 10 

are rare or 2}4 per cent. 
123 words from Ionic Inscriptions prior to 445 B. c. 10 are 

un-Attic or 9 per cent. 

From this table it appears that the vocabulary of the old Ionic 
inscriptions was very similar to that of old and classical Attic. 

•By un-Attic words is meant here poetic, rare or dialectic words. It is not 
likely that any two persons would arrive at exactly the same results in this 
work, but I believe that the relative results, which is the important thing, 
would agree. By far the greater number of words can readily be assigned to 
the poetic or prose column, so that there is not much chance for disagreement. 



298 AMERICAN JOURNAL OF PHILOLOGY. 

Moreover Hipponax, Herodotus and Archilochus 1 (in his iambics) 
are about as close to the speech of Lysias as is the latter's contem- 
porary, Thucydides. This is the more surprising as Archilochus 
was called 'o^npiKWTaros. 2 Herodotus by reason of the scope 
of his history would naturally employ many strange terms, while 
Hipponax, many of whose fragments are preserved by the lexico- 
graphers, merely on account of some peculiar form, was noted for 
his use of unusual words, and the delvers after curiosities in 
language found in him a rich field. 3 It is clear then that Ionic, 
although clearly separated from Attic, is quite close to this dialect 
but in a very different way, however, than Rutherford imagined. 
In the following poets a greater divergence from the Attic norm 
is shown. 

Of 139 words from Anacreon 62areun-Atticor44percent. 

119 " " Simonides Am. 50 " " " 42 " 

105 " " Homer 80 " " " 76 " 

Anacreon's iambic poetry has perished, so that little knowledge 
of Ionic is to be gained from his works in this metre 4 . Simonides' 
fragments show that he paraphrased or copied rather loosely the 
Hesiodic epic, and this explains the large number of poetic words 
in his poems. Of the 80 un-Attic words in Homer 44 are epic, 
29 are poetic and 7 appear rarely in prose, so that Homer may be 
called essentially non-Ionic as well as un-Attic. 

The peculiar character of the tragic diction is shown very clear- 
ly by these comparisons. It is in respect to vocabulary about as 
far removed from Attic as is Homer, and apparently is as unlike 
Ionic as it is unlike Attic. Exclusive of proper names the ap- 
proximate number of words in Aeschylus is 7700. At the same 
ratio as above about 5800 words would be poetic. It is manifest 
that these words could not, as Rutherford asserts, have been 
drawn from the Attic of Thespis' time. Three causes seem to 
have combined to differentiate the speech of Aeschylus from that 
of the Attic and Ionic writers, his residence abroad, 5 his peculiar 

1 This study of Archilochus' vocabulary appears to substantiate Smyth's 
view of the epic element in elegy as against Hoffmann, who holds that the epic 
element in the earlier elegists, Archilochus and Callinus, is altogether absent. 

s Cf. Laeger, De vet. epicorum studio in Archil. . . . Hippon. reliquis con- 
spicuo(i885)p. 7; Deuticke, Archil. Parioquid in graec. litteris sit tribuendum. 

3 Cf. Bergk, Gr. Lit. II 330, and Smyth, Ionic Dialect, p. 46. 

4 For his vocabulary see Weber, Anacreontea (1895). 

5 On this point cf. Macrobius, Sat. V 19, 17; Athen. XV 89; Eust. Od., 
p [872 Plut. de exil. 13. 



LANGUAGE OF TRAGEDY IN RELATION TO OLD ATTIC. 299 

genius, 1 and the usages of dithyrambic poetry. The first cause is 
of least importance, and has been considered above. The second 
cause brought out those "horse-crested", " high-paced" words, 
and inspired Aristophanes to call Aeschylus ayptmows . . . ko/m-o- 
<l>aKiKopprjfiav, but although this gave a certain unique character 
to his speech, it was relatively of little importance even in his case, 
and is merely an incident in the tragic diction as a whole. The 
great part of the foreign element is undoubtedly due to the third 
cause. Aristotle, Rhet. Ill 3, says that compound words were the 
especial prerogative of the dithyramb, 5 and Aeschylus appears to 
have utilized the privilege to the fullest extent. Todt, De Aesch. 
vocabulorum inventore, finds 1100 new words in the extant 
plays, possibly 14,000 such words in the complete works. Verily 
Aeschylus was a Trouper 1S1W ovoparaiv* The life of these words 
was for the most part coextensive with that of their creator, but 
the precedent had been established, and the liberties taken with 
the Greek speech were continued by the later dramatists. Over 
1000 words appear for the first time in Sophocles, 4 and about 850 
in Euripides. 

The relation which the diction of the dialogue tragedy bears to 
that of the chorus is shown very clearly by the following com- 
parison. Of 500 words considered 383 are poetic, and of these 
un-Attic words 134 are found in the dialogue only, 160 in the 
chorus only, 60 in both dialogue and chorus and 29 in fragments. 
The total number of poetic words in the dialogue is 194, and in 
the chorus 220 with the 29 from the fragments to be divided pro 
rata. So evenly distributed are the un-Attic words that it is clear 
the tragedians did not observe any distinction in the diction ol 

1 Cf. Vita of Aeschylus, Dionysius 2, 10, Epicharmus, Schol. Eum. 626, 
Quintil. X, I, 66. 

2 These seem to abound in lyric poetry. "As an inventor of striking 
compounds Stesichorus is the precursor of Pindar". Smyth, Melic Poets, 
p. 258. 

3 Homer in 706 Teubner pages employs about 6700 words (estimate); 
Aeschylus in 276 Teubner pages employs about 7700 words. 

4 On the diction of Sophocles cf. schol. ad. OC 1648; Diog. Laert. IV 3. 7; 
Kuenstler, De vocibus apud Soph, primum obviis, and similar titles by Schind- 
ler, Kriebitzsch, Schulz, Kotsmich, Jasper, Juris ; Ludewig.De dictionis Soph, 
ubertate, and similar title by Schmidt ; and Campbell Intr. to his Sophocles. 
For Euripides see Schirlitz, De serm. trag. per Eurip. incrementis, Funck, De 
praep. ... in Eurip. probato, Curt. Stud. IX 113, Mommsen, Lehre von den 
Gr. Praep. (1895) p. 76. Many barbarian words also appear in the tragedians, 
as KivanriQ, fiayig, oivdov, etc. 



3°o 



AMERICAN JOURNAL OF PHILOLOGY. 



the choral and spoken parts. This fact is also demonstrated by 
a table showing the disposition of synonyms 1 in Aeschylus. 
Attention is desired here especially to the location of words 
with the metrical scheme <-> — , and •-> >~> — . 



Prose 






Poetic 






Prose 






Poetic 






Words. 


Dial. 


. Ch. 


Words. 


Dial. 


. Ch. 


Words. 


Dial. 


Ch. 


Words. 


Dial. 


Ch. 


ayaddc 


3 


'5 


huB/Jx; 


5 


5 


dKdAovOoc 








biraiiq 


1 





oo<f>6(; 


12 


2 


ksSvoq 


5 


9 


BepaTTOv 


o 





OTTaUV 


3 





atoftfct -£? 


8 


2 


irqrvfios 





i 


Oepdiraiva 





o 


dp.q>inoAo(; 








aifievdfc, '£$ 


4 


2 


IrvfioQ 





4 




















va/uepryg 


i 





iravv 





3 


7rdy%v 


I 





aAr/d&s 


3 


I 


CTV/iUC 





3 


TTavToq 


6 


3 














irrjrvfiug 


2 


o 


ijldriOV 


o 





rpdpog 


4 


I 


jUtAaf 


6 


II 


fieMyx<fto<; 


4 


i 














(fifAavoc) 






(only in gen. 


Dat. 


) IKZTTjg 


S 


b 


etficrios 


4 











k€/mcv6c 


4 


5 














addvaroc 





5 


aipdirof 


"j 





avudev 


4 


3 


dvitcadsv 





3 


kfypiepog 


3 





£<p9j/u.epiog 





i 


Kara) 


5 


2 


cvepffe 


3 


2 


irorafios 


6 


8 


peog 


3 


i 


netyaAy 


i 





Kdpa 


12 


9 


ptvfia 


2 


2 


pfadpOV 


2 


i 








Kpdg 


5 


2 


pof) 


2 





vdfia 


2 











Kpara 
(Soph.) 


6 


3 


poof 


O 


I 














Kpara, ra 
(Soph.) 


i 


i 


yevea 


O 


4 


yhva 


6 


3 


dldvfiog 





6 


(5/7ToAof 





i 


yovy 


I 


I 


yevtOAov 


3 


i 


dtnAovg 


8 


6 


Sifuiot 


o 


i 


y6vo<; 


2 


I 


rkavup-a 


i 





SiceoQ 


2 


i 














(firvfia 


i 





StTrAdfJcoc 














dfift^dAo)^ 





I 


dp<f>l~A£K-UQ 


i 











7ri?M> 


often 


a[t<piA6yG)c 





I 








elfii 


often 


reAedo) 





4 



It is apparent that words were not assigned to the chorus by 
reason of their poetical nature, nor to the dialogue because of their 
prosaic character. Metre determined the placing of the words, 
but was not, as seen above in the case of the forms, the cause of 
their introduction into tragedy. A well-known rule of early 
iambic structure is here revealed, that the anapaest was abhorred 
by the trimeter. Here is to be obtained also, I believe, the ex- 
planation of the great difference between the structure of the tragic 
trimeter and that of comedy. 2 In Aristophanes there is one 

' I have followed Schmidt's Synonymik der Gr. und Lat. Sprachen, and 
Dindorf's Lex. Aeschyleum. Absolute exactness in this sort of work is of 
course impossible, but it is hoped that the above words are in the main 
equivalent. 

5 The ancients distinguished four kinds of iambic trimeter, rpayiKdg, ko/iuk6s, 
aarvptKoc, ml . . . iafipiKdc. Cf. Christ, Metrik p. 340. 



LANG UA GE OF TRA GED Y IN RELA TION TO OLD A TTIC. 30 1 

anapaest to every 2j£ verses; in Aeschylus there are 53 anapaests 
to some 4500 iambic verses, 1 and here the anapaests are mainly 
due to proper names. 2 The difference in the structure of the 
iambics of tragedy and comedy was due to the rules concerning 
dialect and vocabulary. To tragedy was allotted freedom in 
diction, but the strict construction of early iambic verse was 
enforced. Comedy was restricted to Attic forms and words, but 
as compensation for purity of diction, freedom of metrical con- 
struction was allowed. Epic words are freely used by Archilochus, 
but there is more strictness in regard to resolution and the use of 
anapaests. The Attic element is greater in Euripides and so is 
the number of anapaests. Metrical license then was the price of 
dialectic purity. 

It remains to treat of the vocabulary of the Old Attic speech. 
Keil, Die Solonische Verfassung, p. 59, greatly errs in holding that 
there was a very great difference between 6th and 5th century 
Attic. His contention is based on the speech of Aeschylus which 
he says was grounded in 6th century Attic, the authentic remains 
of the laws of Solon, and the inscriptional words Upovpyoivrts, 

(aKopovs? ovdos, lnv€vetr6ai, Baav; CIA IV 3, 18, p. I38; Baiav imf&aktiv, 
Srjfiov irKrjdvovrot, CIA I 57 ; Si^o^n'a, CIA I, I ; ajrdjra£, I 286, 288 ; 
ov& in a otSe tpyif, IV I 27 a ; cVian/raro, firuxpdtprts, II 948; (about 

300 b. a). The first reason is a mere opinion unsupported 
by facts, and has been shown to be without foundation. The 
other reasons will be considered in reverse order. cVicty-nro 
and inuxpdi'vTts by reason of their date cannot be tolerated 
in this discussion. How the use of a word about 300 B. c. 
proves that it became obsolete about 500 B. C, or indicates 
the character of 6th century Attic it is impossible to see. But 
admitting that Keil's notion about the other words is correct, — it 
certainly is not,— even then his theory of 6th century Attic is 
clearly erroneous. It is a fact commonly known that in every 
author there occur a certain number of rare or uncommon words. 
In five plays of Aeschylus there are 488 words found nowhere 
else, 4 approximately 700 such words in the extant plays, or about 
9 per cent of his vocabulary. In Thucydides 3 per cent, in 

1 Cf. Rumpel, Philol. XXVIII 610, XXV 540.. 

* 'Aya/ie/ivuv six times, XlvXddiK three times, etc. 

sjaropof was probably brought into Attic along with some foreign cult- 
In Nikander, Alexiph. 217, the term is used of the devotee of Cybele. On 
foreign deities at Athens see Muller's Hdb. V 3, p. 127-8. 

* Cf. Mitchell, Frogs, 802. 



302 AMERICAN JOURNAL OF PHILOLOGY. 

Antiphon 2 per cent, and in Anacreon 10 per cent of the words 
are not used by other writers. 1 The words mentioned by Keil 
form about 2 percent of the words in the prose inscriptions prior 
to 445 b. c. Allowing from 2 per cent to 10 per cent as a legit- 
imate number of odd words for each author or set of writings, 
there is not much left to indicate any remarkable change in the 
Attic speech. But Keil's interpretation of the isolated occurrence 
of these words is arbitrary and erroneous. In the consideration 
of these terms it must not be forgotten that new words came into 
Attic after Draco's time because of the troubled economic and 
political conditions, the influx of foreigners, of foreign literature 
and religious observances. The words quoted by Keil are all 
found after 500 B. C, and it is not clear why they are regarded 
as a relic of former rather than evidence of new usage. All that 
can reasonably be inferred in this matter is this : Whenever a 

word of Common meaning as anoira^ = avpirav, Upovpyovvres — Uptvs, 

Baav = (rjfiiav or rifiav, is not found in epic, new Ionic or Doric, 
the presumption is that it belongs to the new speech. Such 
words as ovdos, iiweveadai " to bake in an oven ", because of their 
meaning are necessarily uncommon in literature, and cannot 
therefore be said to be old words which became obsolete shortly 
after the 6th century. Specifically the facts about these words 

are as follows : dmav, (JaKopovr, iirvtvarBai, Upovpyovvres, ajroVa^ and 

hiX<>p.T]via are found only in the Attic inscriptions or lexicographers, 
and therefore are to be regarded as new forms in Attic. The 
same is true of the additional forms not quoted by Keil, 

irapaifiaTris, CIA I 5 I| 6\ei£a>v, I, I, rpnroav (ioapy(ov I, 5- Of ^" e 

other words <?g») is found in Homer twice, possibly in Archilochus 
once (em.), in the covenant between Oeanthia and Chaleion, 
and is apparently an unpopular word in all the dialects, although 
from its meaning it might be expected to appear frequently. 
7r\r)d{>«> is common in the later Attic and is manifestly a new 
word. The expression oiS' en-ei oiSe ipya in the oath which the 
Chalcidians swear to the Athenians is possibly a Chalcidian 
formula added to the regular Attic ovrt t*x v V °" Te P1X c "'y oiSepta 
(cf. Thuc. V 47, 2 ; 47, 10), inserted as a sop to the Ionic covenanters 
in imitation of the epos or for additional sanctity. No one will 
deny that each century witnessed some changes in the Attic 
speech, but it is unwarranted to assert that all rare 5th century 

'I have considered here 100 words in alphabetical order from each author. 
2 Brugmann, Gr. Gr. (1900) p. 569 regards b'Aeifav as a late form. 



LANGUAGE OF TRAGEDY IN RELA TION TO OLD ATTIC. 303, 

inscriptional forms are a survival from hoary antiquity, in common 
use in the 6th century, bu{ obsolete soon afterwards. So far as 
can be judged the majority of them are new words. 

In Lysias several of the old laws are quoted to show that 
some words were formerly in use that are not found in later 

times. These are iiriopKrjtiavTa = opoaavra, clitiKKu, Spacricafciv, nokoZvrai 

otKijoy, noSoKuKKrj, wecpaofie'iiov, <n-a<rtf»oi> = ' money at interest '. Nearly 
all of these words are confined to Solon, whose language is 
properly subject to suspicion. The laws are the works of a 
traveler who spent the years of his youth in foreign parts in 
contact with tradespeople where a foreign idiom is most 
noticeable, and peculiarities of dialect best acquired. There 
can be no doubt that if Solon was away from Athens in early 
life and again in later years, his speech would have a foreign 
air. It is significant that Solon himself recognized this truth, 
his statement on this point being preserved in fragment 32. 2 
That he borrowed laws from abroad is conclusively shown by 
the statement of Herodotus, II 177 : 26\a»> 6 'A<Va!os- Xa/3a>y «{ 

AlyviTTov tovtov top vopov ' \6r)vaLoi<ri ZdcTo' Tcj> eKUfoi is aid xpeavrai. 

Of the terms quoted by Lysias which are found elsewhere than 
in Solon ttoJUo/m' meaning " to upturn with the plow " (polare 
agros of Ennius) in the active occurs in Hesiod, Op. 4 60 : as 
meaning "to walk" is found only in tragedy, so that it can 
hardly be called an old Ionic-Attic word. oUeis, which appears 
in Homer and Sophocles, is a common term in the Gortyna code, 
and its presence in Solon is probably due to the adoption of a 
Cretan law. The most notable Solonian words are amoves and 
Kvpfius. The former is probably an Attic but its fellow is 
demonstrably an imported word. Theophrastus derived it an-o 
tS>v KprjrtKav KopvfidvTwv, and Apollodorus also says it was an 
invention of the Corybantes. 3 These priests were the Phrygian 
devotees of Rhea in distinction from the Cretan Curetes and 

1 The laws of Charondas were written in verse and said to have been 
chanted in Athens. Plutarch, Solon III mentions the claim that Solon's laws 
also were composed in verse. 

'Headlam, JHS XIII 50-69 finds some notable similarities between the 
procedure of the Gortyna code and Draco's law of murder, which was said to 
have come from Crete. But the law of Draco preserved on the irparog dfwv 
CIA I 61 is very like later Attic. On the old term nufjaKpirat cf. Gilbert, Gr. 
Const. Ant. (trans. 1895), p. 114 note. 

»Cf. also Kvpftic = icvppaoia, the distinctive Asiatic hat, and Cretan Kvppavmc. 



304 AMERICAN JOURNAL OF PHILOLOGY. 

the Idaean Dactyli. 1 The term Kvpfrs may have been ob- 
tained by Solon from Crete, but its origin appears clearly to be in 
the East. 2 It is quite possible that the failure of the Athenians to 
understand the meanings of Solon's statutes may have been due 
in part to the number of foreign expressions in them. Solon then 
cannot be taken as a criterion of old Attic. In Demosthenes, 630, 

28 Occurs an un- Attic word: to de prjb' atrouiav, fir) xpw aTa npaTTfoScu' 
to. yap anoiva avopafav oi naXaioi. Plato's Use of anoiva, Laws IX 

862 C, Rep. Ill 393 E has no weight. Rutherford 1. c. overlooks 
the fact that the word in the Republic is in a paraphrase of the 
Chryses' incident of the Iliad, and the use in the Laws, in a 
conversation between a Spartan, a Cretan and Athenian £«W is 
hardly more reliable. The Ionic status of the word is equally 
uncertain. Herodotus uses it twice, but this may be due to his 
Dorian extraction and he is by no means a model of dialectic 
purity. Pindar employs the term frequently, and this indicates 
the source from which the word got into tragedy and the laws. 
Philochorus, fr. 94 (Miiller, FHG I 399) has preserved an old law 
derived possibly from Cleisthenes : rovs Se <j>paropas <jrdi/ny«r 

bi^aBai Kai roiis opy(S>vas mil tovs 6p.oyakaKTas, ovs yevvrjras Kahovfifv. 

ofioyaXam-fs is manifestly a "poetic" term, probably of late forma- 
tion, which was unsuccessfully introduced for the older and also 
later common term yfwijrai. Rutherford, Phrynichus, p. 24, in his 
discussion of opycav and kindred forms, completely misconstrues the 
facts. The words are post-Homeric, exceedingly rare except in 
4th century Attic, where they are used by Lysias, Plato and 
often by Isaeus, who wrote a speech entitled lipos 'Opytavas. 3 
Yet Rutherford finds in this evidence that the forms were 
a survival from Ionic-Attic, were in common use in Attic 
before 500 b. c, and became obsolete shortly afterwards. This 
is a fhgrant misinterpretation of the facts. But in spite of 
their Attic use the words are probably no more Attic than 
Renascence is native English. The verb opyidfa represents the 
Oriental type of worship and the forms were doubtless im- 

1 It is significant that the laws on the Kvp(3si( at Athens were set up in the 
sanctuary of the Phrygian goddess. 

'Despite Busolt, Gr. Gesch. II (r8gs) 291. On the etymology of the word cf. 
Roscher, Lex. Mythol. 2. p. 1607, and Wilamowitz, Aristotles und Athen, 
p. 45 Anm. 

3 In the 4th century a private cult of the Orgeones existed by which Cybele 
was honored in conformance with Phrygian observances. Cf. Curt., Das 
Metroon in Athen, p. gff. ; Roscher, Lex., Myth. 2, p. 1655. 



LANGUAGE OF TRAGED Y IN RELA TION TO OLD A TTIC. 305 

ported along with the orgiastic cults of Bacchus and Aphrodite 
from the East with whom they were often associated. These 
things show very clearly the doubtful character of the strange 
terms in the old laws. It is impossible here to consider each law 
which has been preserved. Scores of references to the early 
laws have been collected by Telfy, Corpus Juris Attici, and Schel- 
ling, De Solonis legibus apud oratores Atticos. These laws 
exhibit regularly common Attic forms and words. Schelling, p. 
6 holds that obsolete words and dialectic forms were expunged 
from the laws because " nulla in iis flexio reperitur, quae formis 
dialectae Ionicae aut antiquioris Atticae similis sit ; occurrunt 
ubique recentioris dialecti declinationes. Et quod ad ipsa voca- 
bula attinet, vix unum aut alterum paullo obsoleto invenire quis 
possit ". But the bilingual inscription from Sigeum which dates 
from Solon's time, and the other prose inscriptions of the same 
early period, effectually dispose of the notion that the laws of 
Solon, if written in current Attic, must contain a mass of peculiar 
forms and words. 

This survey of the language of tragedy and its relation to old 
Attic seems to establish the following conclusions : The alien 
forms are drawn from Doric poetry. The large number of un- 
Attic words is due in part to the adoption of the vocabulary of 
the dithyramb, and in part to the formation of new words after 
the manner of dithyrambic poetry. The presence of the large 
number of Doric forms in the diafogue seems to substantiate the 
statements above quoted that the Dorians had developed some 
form of dialogue before tragedy was cultivated on Attic soil. 
The diction of the dialogue is essentially the same as that of the 
chorus, the slight difference being due to a little greater restric- 
tion in the use of alienisms in the former on account of the use 
of the iambic metre. Words are assigned to the chorus or tri- 
meter for metrical reasons, but alien forms are not introduced 
into tragedy metri gratia. Moreover the tragic diction is far 
removed from Ionic as this appears in Hipponax and the early 
inscriptions, and this divergence indicates also that the alienisms 
were drawn from another source. In short the language of trag- 
edy commemorates the influence of the Dorian genius upon 
Athenian literature. 

The Jambs Millikin University, JAMES DENNISON ROGERS. 

Decatur, III. 



81