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. Classical Philology Vol. VII April, IQI2 No. 2 THE NEW METRIC 1 By Otto Schroeder The new metric, of which Henri Weil 2 remained the chief repre- sentative up to the time of his death, is distinguished from the older metric of Rossbach and Westphal chiefly in these points: First, by more careful discrimination of the Aeolic, that is, syllable-count- ing, choriambic-glyconic-asclepiadean meters, from the so-called logaoedics. Secondly, and here Theodor Bergk 3 and Hermann Usener 4 are the leaders, by the reduction of the main body of logaoedics to the old ascending four-beat proverb verse, with great freedom in the arses, known as the enoplic, as opposed to the anapaestic par- oemiac. Thirdly, in the interpretation of many cola there is greater recognition of the equivocal nature of many externally similar com- binations of syllables, a tendency which in some metricians has led to the renunciation of all interpretation, while it caused the greatest embarrassment to others until at last they came to study the con- text as we do in the interpretations of texts, and since it was a ques- tion of song and dance rhythms, to treat as the determining factors considerations of juxtaposition and opposition, the keynote theme, variations, and the rounding of the period. In the fourth place, an attempt has been made in various quarters to treat rhythmic clauses 1 Translated from the German by Paul Shorey. 2 Etudes de litterature et de rylhmique grecques. Paria 1902. • "Uber das alteate Versmass der Griechen" (1854), Kl. Schriften, II, 392. 4 Allgriechischer Versbau. Bonn 1887. [Classical Philologi VII, April, 1912] 137 138 Otto Schboedeb as speech clauses have been treated for generations, historically, as a result of which metric has at last begun to be a science. For what may be called the philological interpretation of the periods, a little essay of Heinrich Rudolph Ahrens 1 has been of weight. Hermann Usener 2 and Ulrich von Wilamowitz-Moellendorff 3 almost contemporaneously apprehended the historical problem, the one rather from the point of view of natural science, the other from that of the history of art. Westphal's method of investigation and exposition and Ross- bach's kindly, rather than acute, co-operation were from the begin- ning harmless for strong heads. But even before the appearance of the third edition, which was a retrogression in some important matters, as for example, the three-beat pherecrateans and the dac- tylic tripodies, there arose a violent opposition, which for a time was in danger of spilling the baby with the bath, because in just indig- nation at the coquetting with alleged analogies of modern music, it, with an equal disregard of history, undertook to treat the song-and- dance verses of the old songs almost in the same way as the spoken verses of the epos and the dramatic dialogue, or of the Horatian, nay, the most modern, book lyric. Metricians, like artists, it would seem, are very intolerant of one another, yet with a little good will, why should it not be possible at last to come to an understanding ? 1. AEOLICS In the separation of eight-, twelve-, and sixteen-syllabled Aeolics, Henri Weil's position remains unassailable, however much our rhythmic feeling may be opposed to the continuous shifting of the rhythm in the choriambic, 4 * - ~ -. If the philologians who concern themselves primarily or seconda- rily with Greek verses always brought to their task something of the saving grace of rhythmic feeling, our young science would present a very different aspect. i Philologus, XXVII (1868), 581. 2 In the above-mentioned book, 1887. 3 "Ioniker bei den Lyrikern, " in an excursus on Isyllos. Berlin 1886. * Thomas Goodell, Chapters on Greek Metric (1902), 231; Paul Shorey, "Chori- ambic Dimeter," Transactions of the American Philological Association XXXVIII (1908), 58 ff.; and in Class. Philol. V (1910), 515 £f. The New Meteic 139 But no one has a right to be heard in this matter who has not acquired a historic feeling for the art of ancient verse through acquaintance with the facts, and there is no more doubt of the fact of a quadrisyllable metron with both arses in the middle, e'/c irorafiov, than of the fact of another quadrisyllabic metron with a double arsis at the beginning, ifie BeiXdv. There is to be sure a difference between the double arsis of the Ionic and the choriambic: in the Ionic, the two shorts may be contracted, fir) \lav rdicov St po/xeva 6pr)TO'; 'Aprefimp suffices: ch ch ia(=ch) ia ch(ia) ch ia(=ch) ia ia ia Ionics also show traces of an earlier separation of the theses, as in the case of iambi and trochees the outer arsis of every measure may be anceps, irepivaiov -rat iraKatov, which in anapaests and dactyls would be as incomprehensible as it is unexampled in pre-Alexandrian times. 1 There can be no doubt, then, of the actual existence of the choriambic measure, nor yet of its descending-ascending movement, _ w ~ _, by which it is markedly distinguished from the dactylic meter with its two successive descents, ^ - ~ •*• ~. Opposition is directed only against the principle that glyconics and their con- geners should be treated differently from other Greek verses that have now a one-, and now a two-syllabled thesis. It is maintained that a purely syllable-counting art of verse never existed in Greece. One is glad to answer an honest questioner. We may even thank him for an opportunity of growing richer by imparting and learning 1 Vorarbeiten zur griechischen Versgeschichte, Leipzig, 1908, pp. 26, 36; but M^Scios rb fiva/i ' ivl t J 65£ | K^iriypa\fi( KXetras, Theocr. Epigr. xx. Wilam. 140 Otto Schboedeb through teaching. It is claimed that the antistrophic variations, which the assumption of choriambics explains so easily irovrov divo? i(j>^fi€vov 8' avOalper ov ovBev. Eur. Hipp. 739, 748: [o-Ta\da-]aovo-iv e? olBpa -rraTpws Takat- \vat\ [X^ov-]tui Zt/vo? pe\ddpov irapa ko(- [rai<:]. Inasmuch as the number of the syllables in these cases remains the same, some metricians, in opposition to the opponents of syllable counting, have found a syllable-counting system here also. The source of this confusion is probably Hephaestion, with his Alcaic deca-, hendeca-, and dodecasyllabics, and his Sapphic "nines" and "elevens." But the moderns have developed from it a new form of derivative metric. It was possible to arrange "eights," sevens," "sixes," "fives," and "fours" like the pipes of an organ, and so again "nines" and "tens" up to sixteen, and anything that it was possible thus to group by the number of the syllables, Aeolics, enop- lics, syllable-counting verses, and stress verses in wild confusion, it was supposed had been in that fashion deduced from one another. Enough of that. There are countless cases of antistrophic variation where the explanation of a transposition of a dactyl breaks down. The New Metric 141 We are compelled to assume in addition the interchange of iambs and trochees, precisely as in the Artemon strophe, e.g., 7r/3o'?<ly8ai' Spo/ids ef ifi&v = ffe tov 0av6vr' owttot' e>a (Eur. Suppl. 1000, 1023). Rarer, but of convincing evidence, are correspondences between gly conies and pure iambics: 'JSXeidvia, irdpeSpe M.oi-pav ftadvcfjpovcov = e7rel Terpdopounv &6' - dp/idrav £vyol<; (Pind. Nem. vii. 1, 93), where there can be no question of a transposition of a dactyl. The same holds of Anacreon's glyconic %avdf) 8' "EivpvirvXy fie\ei, which there is no rational ground for separating from the Artemon strophe. 1 And there is yet one other passage which, if our reading is correct, outweighs a hundred others, because it admits of only one explana- tion: that, namely, of shifting of quantities between neighboring syllables and not that of a transposition of single feet, that is to say, by the assumption of a process in a fluid element, not of a manipula- tion of clods. An ithyphallic in the antistrophe corresponds to an acephalic pherecratean : — &t, ft a (vei . —Eur. Suppl. 599, 609 (0pdao<; vulgo). After this we need not be surprised to meet ithyphallics in the closest combination with glyconics and pherecrateans : ov irvp ovSk oiTav iirl Belirvov. 2 1 U. v. Wilamowitz Choriambic Dimeter 24. 2 Eupolis, Kolakes (Usener Altgr. Versbau 94); cf. Arist. Thesm. 1020-21; Eur. El. 142-43. 142 Otto Schroeder The at first hesitating, later (in Plautus, e.g.) unlimited, intrusion of real dactyls into Aeolic-choriambic meters forms a separate chapter. That even in Euripides' time no dactyl was heard in genuine Aeolics (glyconics and asclepiadeans) is proved by the testimony of the delicate ear of Aristophanes, who, in Frogs 1323-24, ridicules verses such as: irep 1/3 a W, & rexvov, wXevai, or opcv; rov iroSa rovrov; 6p&, for no other reason save that the accession of a real dactyl (- — , or reversed, ~ -) carried with it the apparent — — , and so the Aeolic became a kind of Alcmanic, like elp-jvai eparai i-irefiav, or Ibycean, like -qpi ptev aX re KuSwwai. Acephalic verses 1 could remain dimeters, which in the majority of cases appears obviously necessary, only if they are treated as decurtata, which would everywhere lead to otherwise unknown, nay monstrous, forms. Acephalism is in Greek a prerogative of the old syllable-counting verses, whether we assume that the accompanying instrument or the chorus-master before the performance or at the beginning of the member marked the octosyllabic movement, or that later, e.g. in Pindar, a preceding colon, closely connected by synapheia, by prolongation of the last note, helped to fill up and complete the time 2 — in any case, it is only and precisely on the assumption of a fixed number of syllables that at the beginning of the colon one or two notes might remain unexpressed in the text provided only the conclusion came out unambiguously. Listen to Choephoroe 319, . aK6ra> do<; avrifioipop, ^dpire; 8' 6/Weo?, or 345 3 . . el yap for' 'L\i§> wpos two? Avkudv, irdrep. 1 For examples see Aesch. Cant. 117; Soph. Cant. 83; Eur. Cant. 188-91; Aristoph. Cant. 96. 2 Neue Jahrbb., 1910, I, 175 (where read " kontrahiert " for "konstruiert"). 3 Here a warning lest we at once take all cola of the form el yhp {it ' 'I\t067) ft 6 \ e (Anacr.) avdr

v iv = 0- e\a yeiro 8' av 1 aar v irvp (Eur.) There is a further apparent contradiction to the principle of syllable-counting in the liberty early practiced in the Aeolic eight of replacing a thesis by two shorts, which has the effect of increas- ing the number of syllables. But this very license in the extent to which it is practiced presents insuperable difficulties to the enoplic interpretation of the verse, as Friedrich Blass has already pointed out. 1 The stress verse has always been very recalcitrant to the resolution of the theses. And especially to find a thesis of two shorts before or after an arsis of two shorts we should have to look far. ^Traprmv iva 7 e v ? eave Murray). And we have probably to go down as far as the Rhesus before we find an Alcaic ten with resolution of the thesis of the dactyl: ^evyvvre KepoSeTa To^a vevpate = p.rjiroT' e? i fi e t iv a fie'pyjriv eiVj;? — Rhes. 33 = 51 1 Fleckeisens N. Jahrb. 133 (1886), 463. The New Metbic 145 If now, as has been made probable elsewhere, 1 dactyls are nothing but enoplics of regular dissyllabic arsis that have become descending, it is obvious that this very fact accounts for the unresolvability of the theses in dactylic sung verses, as in the spoken verses of the epos. And there is no greater conceivable absurdity than to attempt to derive from Aeolic syllable-counting the unresolvability of the theses in the heroic hexameter, which rightly considered is merely the irrationality of the theses maintained throughout. 2 Indeed, com- pared with enoplics, our syllable-counting Aeolics are the liveliest thesis-resolvers. 3 KaXa yepoV aeiao/ieva (Corinna). avair o Xifrfiev op,pa %4p,iv le pdv (ibid. 9, 41, 52, 57). yX&aaa p, e X t r o ? aoorov = iv re/ievei iXa> yets (Paean vi. 59 = 120). & Heyio-To ir o X i Ovyarep (P. viii. 2, 9). ■jravBanoi ttoXii iirl voaov (Soph. Ant. 1141). fiijKert ir a p a ydyy; , "v ov (Soph. El. 855). e7ret aoi 7r oXv0vro<; ael (Eur. Heracl. 777). Among enoplics the ithyphallic and lecythian clausulas, with arsis of one short syllable throughout, earliest and in largest measure admitted a thesis of two shorts: otototoI iroiroi Ba (Aesch.). K?)pe? a v a it X a ktjtoi (Soph.). u7ro Bvolv fitacrTopoiv (Aesch.). Tt? rdBe vefieai<; crrvyel (Aesch.). tI roBe veov a^os; /ieya (Aesch.). It' en 6viov iiirb Bepav (Eur.). 1 Vorarbeiten 32 ff. 2 An attempt recently repeated in a Gottingen doctoral dissertation (1910) in- spired neither by the Muses nor the Graces. 8 I have not made any use, in the enumeration, of those glyconics which in con- sequence of resolutions throughout could be read also as iambics: Uraye ToKidaitpvii aSovdv, Eur. El. 126. 146 Otto Schroeder But a lecythian, which, like the glyconics above cited from Pin- dar, should resolve the first and last thesis, is a thing unheard of even in the most audacious dithyrambic. Resolution of the last thesis in Aeolic "eights" would be in any case wholly excluded if, as even adherents of syllable-counting sometimes seem to admit, choriambic and glyconic dimeters were descending verses with the customary omission of the final arsis, which, however, at the option of the poet could at any moment be reinstated, as in alel 8' iv are^dvoiaiv &t]v (Eur. Here. 677). Against such wild fancies it is enough to quote a word of Wilamowitz: "Si una aut duae syllabae metrum excedunt, hoc demonstrat mensuram huic uersui non conuenire. ' n In plain language, a syllable apparently attached to the glyconic proves a change of rhythm. 2 Accordingly in the passage of the Heracles just cited a descending enoplic of four stresses follows the four Aeolic dimeters. Or, to take a poet who cannot be suspected of want of style, in Pindar's ^a/stTe? 'Opxofievov iraXaiyovcov Miwav eirivicoTroi, (0. xiv. 4), we have not two Aeolic dimeters separated by an intervening syllable, but a glyconic is followed in immediate sequence by an acatalectic paroemiac : just as in the same poem, alevaov dva>cre KvSifMov aed\a>v irrepolai. yairav, a catalectic paroemiac follows a choriambiambikon. Of other combinations with choriambic-glyconic Aeolics no more need be said, for the only issue at present is whether choriambic- Aeolics exist at all. Yet finally we must not forget that the testimony of antiquity on this point is more harmonious than is often the case. With different reasons, and therefore from different sources, we find in Heliodorus and Hephaestion and Aristides Quintilianus an explana- tion of glyconics and their associates that altogether ignores dactyls. Hephaestion 3 in agreement with Heliodorus 4 explains the glyconic and the lesser and greater asclepiadean as antispastic dimeter, 1 Comm. metr., Gottingae (1895), I, 11. 2 For "superfluous" syllables in Ionized enoplics, see below pp. 155 S. 3 Enchiridion, ed. Consbruch, 32 ff. 4 Colometriae Aristophaneae quae supersunt, ed. Thiemann, Halis 1869, passim. The New Metric 147 trimeter, and tetrameter, and Aristides Quintilianus 1 also operates with a quadrisyllabic measure which he calls Ba^elo?, namely -~ ~- /Sa/c^eto? airo rpo^aiov, ~ - /3a/c%eio? airb Idfifiov, but concerns himself solely with the dimeter, the wois avvOeros icarh irepioSov 8m&eKd(Tr)iio<:, as he calls it, of which he enumerates twelve iambo- trochaic variations, in which he is thinking of the most common forms of the Aeolic "eight." Hephaestion, on the other hand, with the aid of TroXvcr^^aTto-^o? (pp. 56 ff.), it would seem, arrives at a doctrine of Aeolic syllable-counting, which if somewhat con- fused is not to be despised, and even reaches a sort of understanding of the pyrrhic beginning. 2 Hephaestion is to be pardoned for conceiving the possibility of the indefinite multiplication of that strange product of syllable-counting that we find notably in the approach or prelude, that is to say, in the first metron of the verse, and which a wag dubbed by the name of avTio-n-aaTos or dis- located foot: how could the Alexandrian grammarian attain to an historical insight into the evolution of Aeolic verses from the plain "eight" to the asclepiadeans ? But H. Weil should not have fol- lowed him; for though in rude antiquity an accidental sequence of antispasts with or without ambiguity of the external syllables, = — -, would be perfectly conceivable, in the artistic verse of the Greeks known to us this chaotic polyschematism rarely appears to extend beyond two, and never beyond four syllables. And while in the choriambus, including that of the glyconic, the collision of arses only makes the colon more flexible and flowing, in Aeolics on the contrary, the clash of theses yields a felt dissonance. Juxta- posed shorts finally even in Aeolic members, and that certainly not merely under the influence of the epic recited verse, have so affected one another as to legitimate hiatus, that is, abbreviation of the final vowel, with which a division into antispasts, oo—l I I (instead of I I ~-), is incompatible. This also disposes of a slight objection to the choriambus in Aeolic verse. To the cases collected by Edward i De musica, ed. Jahn, 25 ff.; cf. F. Blass, N. Jahrbb. 133 (1886), 461; H. Weil, Etudes rythm. 154, 163 ff. 2 Hephaestion 22. 18; 31. 17; 57. 9, 10; 128. 23; 129. 5. 148 Otto Schroeder Clapp, 1 the author of this objection, of hiatus outside of dactylic cola, we may further add from Pindar in tribrachs: avBder o fxai ivopiaov 0. ii. 101 (92) •jroWd /j,o i vir' ayK&pos &icea /3e'\j/ ibid. 91 (83) to, Be Kal av Bpdaiv ifi-irpeirei P. viii. 28 nay, even in a syllable that follows the tribrach : eirero ol, ovBe viv N. iii. 39, and in the pyrrhic-Aeolic beginning: Kal ay dvopos Xinrov 0. ix. 23, Kal orav Ka\a ep-(£ aedev, k.t.X. yXv KvtfiK p ov aftd'xavov, k.t.X. eirei 01) vr\ a ? re k.t.X. 1 Take again the shiftings of the thesis, a thing necessarily foreign to the nature of the stress verse: TrapairXijKTq) j(€pl avyKCLTaKTa 1 * KeXaivoli ghfSeo-iv /3oto, Kal PoTrjpcK iirirovd)p,av vrroTrerpihi to v ove ipmv = ainpov aeipofiev at fidy^ovTat = aXX' 'AyrjarixSpa fie Trjpel = elptjva<; ipaTa*; eirefiav, — these variations are most easily explained by antispastic influence at the beginning and choriambic at the close of the cola. Finally an ascending proverb verse, which had lost its introductory syllable, may have more easily preserved its ascending movement if it made choriambics of, that is, Aeolized, the first two single feet: ova afiepiwv - iroov 6 <5' epii fie'/Ji.T)- vev. This then prepared the way for the entrance of the seven-syllabled pherecratean into the Aeolic system of the choriambic "eight." For there can be no doubt that the bacchic catalexis ( ~ - . * ) in a syllable-counting measure can be understood only as a loan from the proverb verse which was catalectic from the beginning. But we cannot look for success in the attempt to make one of the two measures the original measure and derive the other from it by restriction (monosyllabization) or license (polysyllabization) of the arses. This, we have already seen, in our study of Aeolics with i Vorarbeiten 27 ff., 41 ff. 150 Otto Schroeder their specific peculiarities in the matter of antistrophic variation, acephalism, syllaba anceps, and resolution of the theses. The most essential thing now is to determine also the specific characteristics of the stress verse, or enoplic. 1. 'RpacrpoviSr) XapiXa- e ftpf/fid toi yekoi- op (Archil.) 2. Irj'ie 3>oty3e, vol- Se Tewr' apeo-r'- e'i- r) (Soph.) 3. Set- vol S* aft! eirov- rai Kij/se? avaifKaicr) - toi (Soph.) 4. & fieyaXa (frans, &, •*■ fiarep alcxyvax e/xa? (Soph.) The first of these short verses (la) has a dissyllabic arsis in the second and third places; the fourth arsis is unexpressed. The third (2a) is distinguished from it by the limitation of the dissyllabic arsis to the second place. The fifth (3a) is ambiguous. Among Ionics it would be a dimeter with the first (others would say the second) metron con- tracted to a spondee; among Aeolics, an acephalic pherecratean; among enoplics, we shall meet the same combination of syllables as a verse of three stresses in the so-called 7rcuai>tcr/*o'?. In our passage (O.R. 471) after two anapaestic and before an ithyphallic dimeter, that is to say, between pure, though differently stylized enoplics of four stresses, no one will think of reading it otherwise than in four- stressed enoplic movement, that is, like ^pao-fiovihrj XapiXae, only without arsis accompanying the first two theses. Finally, the seventh (4a) is identical with the second half of the elegiac pentameter, quite frequent among iambic and enoplic dimeters, 1 a standing component of Chalcidic (dactylo-epitrite) periods and in its true nature only a proverb verse, like 'Epaa-/ioviBr) XapiXae, though in this form distinguished from it by latency of the initial arsis and contraction of the two final theses. In all these cases even where only three, or, (as in 3a) two stresses are in sight, it will be natural to German ears to recognize four: 1 It is very instructive to find among pure four-stressed Aeolics, such as irpQyh re itui xa.p6.8pai. (irpdovh re vulgo), Alcman fr. 60, such forms as iropcpvptounv aXk i-ia-s Cod., -cots V. v. Wilamowitz, Comm. Gramm. I T1879], 4), and r]fiia 'ara), with which we may now without hesitation associate the ev^fielre of the Mysteries and the afte ravpe of the women of Elis, then two things become clear: first, the oldest stress verses were catalectic; secondly, whether the initial arsis was expressed or not, 1 the oldest stress verses had ascending movement. This raises a presumption that tetrameters like Zyvos etyaijnv • etrwvvfiia 8' etreicpalveTO fi6pcuv&fievov xaxdv oIko.S'' S.ye in the exodos of the Peace. That the Adonis cry 2 is literally identical with the two-stressed dactylic clausula can- not be denied. Lastly the pentasyllabic iambic may occasionally in the form of an acephalic ithyphallic be called a dimeter. 3 Which of these abstractly possible interpretations is the right one must always be determined as in grammar by the context. But what of dactyls with a prefixed syllable — descending rhythms with ana- crusis, as they are innocently named and believed to be ? There is just one place where extra metrum such a syllable has been found before a series of five four-stressed dactylics, namely, at the beginning of the conclusion of the great commos of the Philoctetes (1203): aXX' 3> ijevoi, ev i- \o|ei>ta?- ayviai. Xap/rrei S' U7TO- fiapp.apvya2<: - 6 %pva6i. — Bacchyl. iii. str. 2, 3. Since now, apart from the two exceptions in the Philoctetes and the Oedipus Coloneus, 2 there exist no debatable cases of a syllable pre- fixed to descending cola, and since moreover of the hundreds of enoplic cola of the form %aip\ & 'l&icdepye, or iambic of the form triya, icddi&, there is not one where only two stresses would be conceivable, we are on firm ground. Anacrusis in the sense of an immovable prefix forming no part of the rhythm is a name for some- thing that does not exist as a general prerogative of descending Greek verse-members. The readiness of modern readers to accept a movable after-syllable outside of the rhythm in ascending verse is due to our habituation to the licenses at the close of the five-foot iambic, 3 for which Greek offers no analogy. Such lines as Heraus in eure Schatten, rege Wipfel, .... Als ob ich sie zum ersten Mai betrate would in Greek inevitably be six-stressed trimeters, and the last syllable through latency of the preceding arsis would become a complete thesis, as in %aipot,aiov (Isthm. i. ep. 4). The critic who sees no problem whatever here may observe with malicious delight the repeated attempts to master the difficulties. We are concerned only with sincere fellow-combatants and submit the course of our thought to their serious considerations. If iambic measures with an after-syllable are found only in Ionized enoplics, the irregularity must be connected with the Ionization of the enoplics. The conclusion is inevitable. Accordingly it will now not be diffi- cult to see 1 glimmering beneath the unexampled thirteen-syllabled trimeter Hvd&vo? kt\. an old paroemiac with iambic basis, as rj JZ.d8p,ov rj ?LirapT(ov Upbv yevos avBpoav, and to perceive the same simple paroemiac also under the second of the above cited dimeters, 2 which is made still more interesting by its agreement with a dimeter with choriambic beginning: 1 That is, beneath the flaunting quasi-iambic oixonXdpots £t6ittcus perhaps a simple didifioitriv twtnrTais. 2 I.e., under the flaunting quasi-iambic %v$ov vtfKi tt\ovtov Kpvtpatov perhaps a simple %v§ov pjyay 6\f3ov istcpv^ev. 156 Otto Schroedeb rj KacrTO/seia) 7J 'Io- Xdov iv a p fioljai viv v/JLvq> Now as is proved by the certain equation: 1 the Ionic rhythm, the most pliant creation of Greek rhythmical fancy, is compelled by means of a momentary acceleration of the tempo 2 to contract the individual time-divisions, both long and short to less than the normal duration maintained before and after. 3 In no other way can we explain consistently with strict preservation of the triple rhythm the presence of trochaics among light Ionics after pure forms or choriambic, 4 after Ionic formation in the so- called anaklomenos or in galliambics, 5 after palimbacchics with or without tribrachs. 6 When the Ionic however had reached this point it required only one step further in order in the mere exuberance of its elasticity, in the anaklomenos, e.g., to accelerate the three sylla- bles following a prolongation of the first long: ji u i im j j ^>e/o' v8a>p, (pep' olvov, & ttcu Whether the anaklomenos is or is not already, as may well be the case, an Ionized enoplic (&vvap,ai icpeKT)v tov Igt6v) the extended •jrvKvmavi is just as conceivable in the pure Ionic: Ikstsv- a> ot/Se, crol Be, which with its elegant transition from heavy to light Ionics may have taken the lead : The nine-syllabled verse now ran : JJ/3! RR J'J l.irapT&v le- pov yevoi avBp&v, the iambic[hyper-dimeter: evSov vefiei "jtXovtov Kpvaiov, the same with choriambic beginning: (^ Kacrropeip fj 'Io-) Xdov ivap- fio^ai viv v/ivqy the iambic hyper-trimeter finally: Hv9&vo6p- fuyye<> vfivoc would be a monstrosity: it is an Aeolic "eight" with antispastic instead of choriambic structure in the center. If now from this digression on the primitive, yet it would seem perpetually recurrent subterfuge "ical avWaftri," we return to our characterization of the two great kinds of verse, it will after all that has been said not be found an easy task to bring Aeolics and enoplics, syllable-counting and stress verses, under one head, however designated. 3. Periods In the ambiguity of the cola we have several times referred to light to be expected from the context; in certain acephala a dimeter sometimes appears necessary (p. 142) and then again only three- fourths of a dimeter (ibid. n. 3) ; the heroic hexameter everywhere in Greek lyric was to count six, and not eight stresses (p. 152); the Alcaic "tens" with apparent anacrusis (Bacchyl. iii. str. 2, 3) turned out to be unimpeachable choriambic trimeters; in a hundred cola of the form x a V'» & 'Eicdepye or aiya, icddi^e not one could be found in which scansion with two theses was admissible; and we kept finding decurtata, that is, members with three visible but four audible stresses. How can we know all that? and how is the con- nection to throw any light upon it ? what in general is the meaning of "connection" in metrical structures? Some metricians 2 see con- nection only in the return of the identical. They are moved to laughter when in the explanation of structure of Greek verses they 1 Vorarbeiten 90-91. ' So Fr. Blass; on this see Vorarbeiten 160-52. The New Metric 159 hear of a variation. "Responsion of what does not correspond," they say, and the gallery applauds. Now repetition of the identical is everywhere rare in Greek lyric, within the cola as within the period. As is true of all music, the life of the rhythm of Greek sung verses resides in the variation, in the diversities of every kind. 1 Whoever disregards this will in the majority of Greek strophes apprehend only the parts, nay, not even the parts. Hence the present situation with regard to the understanding of metrical art. An unhappy chance, or it may be the iron law of human limitations, has willed that in some really eminent philolo- gians the organs needed for the estimate of great rhythmic structures are atrophied. For many years now they have concerned them- selves with the metric of the cantica rather from necessity than from passionate devotion and with the courage of making mistakes. The great majority are accordingly quite helpless. They are best content if they see that the rhythmical periods coincide with the grammatical. They attach themselves to that which is most familiar from the schoolroom and is obviously also in the intention of the poet the predominant thing, namely, the meaning of the words, and let the rest pass as an accompaniment. In particular, when older poets, like Pindar, often intentionally avoid coincidence of grammati- cal and rhythmical pauses, that is to say, when the enveloping rhythmic garment rather veils the members of the thought than moulds itself to them like a mesh, these scholars feel uncomfortable. That strophes repeat a fixed group of verses with the same melody they accept as the familiar fact, but they never ask whether there may not be possible or demonstrable relations also between rhythmi- cally different strophic parts or members. What a lack of under- standing, nay I will venture to say, of reflection, when Friedrich Leo 2 exclaims, "Was soil denn das Ineinander zweier Responsionsarten bedeuten?" What is meant are the freely varying echoes within smaller and greater periods, and the antistrophic congruity of the repetition of such structures. It was a revelation when, more than twenty-five years ago Fr. Blass had the idea of explaining as dimeters the three-stressed 1 Vorarbeiten 152. ' Gmt. Gel. Am. 1911, 104 Anm. 2. 160 Otto Schroeder dactylic cola of the dactylo-epitrite, following therein certain Greek metricians who have long been objects of contempt. And how timid, even up to the present day, have scholars been in adopting this extraordinarily fruitful idea, although in the interim super- abundant confirmation had presented itself from Bacchylides and the drama. It looks like three dactyls; let us call it dactyls of three theses, or better yet, dactylic trimeters. This metric, which at bottom remained only a metric for the eye, in reality triumphed only in the huge mass of repeated spoken verses of the epic and of the dramatic dialogue. But the spoken verse is unambiguous and is essentially finished at the moment where we meet it first in literature. The sung verses on the contrary, apart from a handful of the strophes of popular song, 1 are in perpetual flux: what seemed like becomes unlike, and vice versa. Nothing will help here but to lose our- selves in loving study of this to us so alien world of verses born in song and dance. We shall never find either meaning or connection even in the simplest rhythms and sentences without a correspond- ing secret vibration of the auditory nerves and those of the motor organs. But now, in fine, what is connection and context in Greek rhythm ? When has a period a beginning and end in itself, in Aristotle's phrase? when has it hand and foot? when does it round itself to a whole ? The penetrating interpretation of some strophes and passages not repeated in strophes will more advance us than any theoretic explanations. In the first rough working up of the material — my analyses of the Cantica have never been offered as anything more — there was no space for a loving exposition of the rhythmic course of thought. Some samples were presented some years ago, but they have effected little more than the fixation of the concepts irepiohos and avafioXr}. The metrical guild that has its headquarters at Gottingen has no use for anything of the kind. 2 However avSp&v hucaloav xpdvos crwTrjp dpicrro<:. Everybody knows the strophe fiev 7rov vaieis aXi-irXaicTos evBaifiwv, iraaiv irepi<$>avro<; aei. 600. eya> S' 6 rXd/jLav "7raXcub<; a(fV ov %povo<; 'ISaia fii/xvcov XeificaviSt iroia firj- Xwv, avrjptdfios alev, evv&fiat,, 605. XP^ V< ? Tpi>x6fievo$, /caiccbv e\7rt'S' ej(a>v en jxe 7tot' avvtreiv rov cvk6- rpoTrov atSrjXov "AiSav. The text is in general that received: 601, Xeipa>v(AI for Xei/juoviAI with Erfurdt; 603, evv&fiai for evvo/xa with Bergk, and in the antistrophe, 616, %epolv fieyio-rw; for xep-)/* 01 /tot, which here before a concluding paroe- miac of the form Irj'ie $>oi-fie, /iai = (ev-)p t|T«i, which at the close of a double trimeter and before the beginning of the last sentence again mark the closing emphasis and at the same time tension of feeling, in this instance surely more especially tension. For it would be impossible for the strophe to end with this after-echo of the first sentence, augmented by the two prefixes or preludes. In the whole of Greek lyric there are no structures composed solely of sentences of unequal compass. There is to be sure the form a a+b, that is to say, unequal sentences in which the excess of the one is distinctly detached as a third sentence (whether before, after, or in the middle makes no difference) but there is no lyric constructed on the formula a b or even abed, Kara Trepiopia-ftov? avurovs, as the "eye metric" believed, still believes, and teaches, without being able to offer the slightest proof. We must then in the present state of our knowledge expect a further sentence of seven or nine measures to round out the strophe. The continuation now presents according to our colometry a glyconic with a choriambic metron added, and then with agitated resolutions, a lecythion closely united to an ithyphallic, with syllabic prefix and also containing resolutions; that is, even as in the first sentence, three dimeters with an interpolated measure, which in this case however is one dimeter farther back, and in its full complement of syllables seems more animated. We have thus three sentences of seven, nine, and again seven measures, and may conjecture that the course of the melody illustrated this correspondence of two sentences of equal compass which we without difficulty read in the metrical structure. We may indeed everywhere infer identity of melody from antistrophic correspondence. Whether and how far this The New Metric 163 structure was also reflected in the evolutions of the chorus and espe- cially in the lyric dance of Pindar or the tragic stasima is a question that will probably always remain obscure. But since even aTpoi], avrioTpoo<; and e7r&>So? (fiei\a trap' ai\ o t ? eVecr' eVecre fjteXe 01? 'AT/aetSat?, though to be sure in the corresponding place of the strophe there is no division. Decisive for me is, first, the fact that the continuation of XP° V( ? Tpvxofievo'i does not begin with a long syllable which would more distinctly detach the sesquimetron, whereas now Kaicdv almost automatically completes the glyconic, and, secondly, the insinuating echo of the paroemiac (599) and the ithyphallic (608), the content and sound of which are especially felt in the strophe: •k a er iv ir e p I avr o ? a el and {rov airo-^r p o tr o v a ihrfkov "AiBav. This impression is strengthened by the fact that the first sentence of the second strophe repeats the two first trimeters of the middle sentence of the first, and rounds them off precisely with our ithyphal- lic beginning with four shorts: (vocrovv-^ra p e v o p, opws aicovari. Yet whichever of the interpretations of the final sentence is preferred the structure of the whole remains essentially the same. The remainder of the second strophe, apart from the purely Aeolic, 1 alXivov alXivov, admits of Aeolic and Ionic interpretation up to the close of 630-1, (navrivav, asclepiadean double dodrans, catalectic, and dodrans, rfkOes e« Trepdroav, that is to say, as two dovetailed asclepiadeans. After this intentionally somewhat uneven composition the seven metra of the concluding sentence flow on in a strong and powerful stream; two pherecrateans and a phalaecean. The most advanced of American metricians, John Williams White, 2 who has just laid down his professorship but will, I hope, not soon bury his hatchet, likewise favors Aeolic interpretation here, but secures a different articulation by treating on principle all three- quarter glyconics as dimeters, when they do not in juxtaposition yield an asclepiadean. But even those who adopt this radical and as such convenient principle would still hardly maintain that this measurement here yields a better connection. It is true from the purely numerical point of view no fault could be found with the equation 33 2 2; 3 2 2: 2 2 3 = 10; 7: 7 metra but in every other respect it would be a change for the worse, notably in the first sentence, where an ithyphallic would be followed by a markedly acephalous glyconic (« ~ -«- ~- ~-). However it must be observed that White seems sometimes to attach no importance to syntactic articulation and the connection of periods. Let us now take a duet. In the first strophe of the parodos of Sophocles' Eledra, the chorus begins with expressions of sympathy which Electra rejects with thanks. Both sing in 17 measures, unless with unpardonable thoughtlessness we disregard the prolonged alal in the strophe and antistrophe. 1 For ithyphallics among glyconics see supra p. 141, n. 2. * Classical Quarterly III (1909), 302. 166 Otto Schroedeb XOPOS [t] cS irot vaX dvirTavordras 'HAEKTPA 3 7eW0Xa yevratuv, [Taripuy] 'HX&rpa /MTp&s, t£»' del 130 f/Ker' 1 ip.wii xandrav wapa/jdeiov Tiieas US' dicSpearov oi/ioyav oldd re rat S-vvtriiu rd$\ oSn pe tpvyydvu, oiS' 4$£kw irpoXnrei» ride, rhv iriXot 4k SoXepas d0eWTdVa[s] M <"& Toy ^y arevaxtiv vartp' 1 125 fiarpbs dXo>r' dVdVeus 'Aya/i4/imva KaKf re X «ol xpoSoro,., is 4 roSe dX ^ f J a ""°i" „, "■ OS (p raXaiv' iya> tcaic&v trim, ap' €? to /eaWo? eKKeKaxjxovrai ^ir) ; trim. rdy^a Tt? 'Kpyeuov evoirXo [t]? opfirjaa 1 ; 8 — 8 1290 tto8\ @OT]8pd(i

KeXevOov ba ba o-Koirevovcr' wrrdinq. ba ba E lcb TLeXao-ybv "Apyos, oWv/xai «a/ew?. trim. X 1 rjicovcraQ'; dv8pe<; %et/>' e^ovaiv ev 6vt\oio-iv. / \ E MeveXae, . Ovrjaicw o~v 8e trapdrv p! ovk a>$eXe«. H oXXvre KaiveTe, ShrTV^a 8taTop,a cbryava - 0e(veT ? - eK %epo? lepevoi, 1 3 05 Tav XnroTraTopa Xnroyafiov, a irXelo-To v? etcavev 'EXXapaw 8dpt trapa troTafMv 6\op,evov<;, 66i 8cucpva SdicpvcrLv erre- &e m8apeoi<;h] 1310 fie\eira Set (col fiiXa k&^/ov, Si ; K aVaircu , 310 pd AP otic ?7(i>7« v$t> old' bvbSev ri Seiirror tarai. II tI pie drJT\ <5 peXta pij- rep, Ktiktcs; K tra pot irpdypara poaxeiv irapixys- II avbvi)- top dp', ev. If it be granted that it is not too violent to double an interjection which occurs but once in the MSS, we shall then have two sentences of three dimeters each, which clearly correspond and are characterized by the anaklomena, and which are followed in each instance by a sentence spoken by the boy, a dimetron with a single metron as a conclusion. Together the speeches of the coryphaeus and the boy form in each instance a well rounded period of 2+2+2/2-fT metra. The remaining metra again make a total of 9 on each side, reckoning in the clausulae. The third and fourth on either side have contractions in the arsis, 1 the only contractions in the whole duet. The periods are both divided into 2+T/2+2 2 metra, but through the diaereses they are brought into an order the reverse of that of the preceding periods: 6/3:3/6. Both parts of the duet have accordingly exactly the same extent, 30 metra each, and are divided in exactly the same way into 12; 9:9 metra, clearly having in many ways mutual internal relations, both in the position of the anaklomena and of the contractions and in the clausulae. All who are not blinded will see here either two • The MSS tradition offers in 314 a metrical monstrosity, ~ - — - ~- . A reference to 5ie5i)er' (281) is of no avail, for there it is not an unretarded Ionic, but a choriambus that follows. But still the tradition, &v6virrov &pa , so as to make an "Ionic" of the form ~ - - ! s 309 without . 'An "Ionic" of the form „ ^ " w - __ _ ! ' Berliner Klassikertexte V. 2, 44. 8 Sitzungsb. d. preuss. Ah., loc. tit. 172 Otto Schkoeder Poor Gottfried Hermann! Never before has such treatment been thy lot! And commiseration is due Aristophanes, that of a sudden he is not allowed even those corrections for decency's sake without which no editio princeps today appears. Nor, finally, are the Caniica placed in an enviable position, for they are equally implicated with the others. Happily all this does no more harm to them than to the two great dead. But perhaps we may learn something from the example which Wilamowitz gives of an analysis of the aria of Philocleon which immediately follows : 317 Ckoi, TrjKonai fiev ba ba 2- 2 irdKai Bia t?)? ovn'S » glycon 2 vfi&v xnraKovav. * pherecr 2- aKXa yap oi>x oU<; t' ei/t' ( ch mol ( 2- iXdelv rv 7ro(i)rjv8\ en-el I glycon I 2 fiov\ofia( ye irdXat fied' v- \ glycon \ 2 fi&v i\0a>v irrl tov? Kahi- ) glycon J 2 aKom kclkov n 7ro(()?)o-a( \ pherecr \ 2- 8- e\0elv is the reading of von Wilamowitz (aiXdtv for atSeiv), and he deviates from our interpretation of the meter only in this, that he would have the first two cola made up of a dochmius and a gly- conic, and the fourth and fifth of a dochmius and a pherecratic. This we cannot accept, for while the dochmiacs, it is true, balance each other, together they do not balance the following period of glyconics. There follows a prayer in anapaests, the first verse of which, with Porson's emendation, runs aXX & Zev fjxya- ftp6vTa. " What reason in the world there is, " says Wilamowitz, "to change the last pherecratic into a paroemiac, does not appear." " The last pherecratic ? " — can we believe our eyes ? Here we have a nonstrophic member, in which the rhythmical and the logical pauses may coincide in a totally different way from that to be expected in strophic recurrence. Are we to say then that the first dimeter of the prayer, which preserves its anapaests throughout, rhythmically belongs still to the Aeolic sentences, which are from the periodological point of view completely rounded; and that the continuation of the The New Metric 173 prayer on the other hand is no longer considered a song? Porson was also certainly an afiovao<{, but a natural sense for the correct and the reasonable preserved him, as it has Wilamowitz a hundred times on other occasions, from making a division so lacking in harmony. The anapaests run on in 8- 3 3 8- metra. "Doesn't catalexis mean anything then ? " objects Wilamowitz, who is content to divide them into 6 and 14 metra. Certainly, the catalexis after the fourth dimeter of the prayer does have a meaning for us, and for us alone: taken together with the two following distinctly detached trimeters, which in turn are followed by a period of 8 metra, it shows that Philocleon's prayer is musically divided and is therefore to be counted as at least half singing. Who shall say whether these examples of the old method of interpretation will win new adherents? The guild will not fail to enter it most dutifully on the books. Still perhaps some who are not yet wholly enslaved may undertake an independent discussion of the new method: to be sure, a man must familiarize himself with that which he wishes to attack. "Metric demands patience," says Wilamowitz (p. 526), and he is right in more than one sense. So much for the present of the nature of period formation. Our attitude of opposition to the worshippers of the irepiopurpol aviaoi has involved the necessity of repeatedly defending afresh the thesis: no strophic or astrophic members, no rhythmical forms of greater or less extent, exist in Greek sung verse without a dominat- ing duality of clearly related, equally extended 1 periods, capable of extensive internal variations. Many questions of rhythmic stylistic it was possible to touch only in passing. My task was to illustrate by a few perspicuous examples the method which I had already found confirmed through the greatest part of Greek lyrics when Timotheos' Persae brought complete certainty. If any critic can test it without prejudice, that is, without inadvertently or unreflect- ingly identifying it with the superstitious faith in numbers of the 'Exceptions, antistrophioa of unequal content: Aristoph. Cantica pp. 15, 46, 59-60, 83, 84 (Vesp. 410 ff„ 468 ff., Nub. 811-13, Thesm. 312 ff.-352 ff.; Eccl., 912-13, 954 ff.-963 ff.) ; nowhere however in these examples is the balance of the periods destroyed. U. von Wilamowitz goes much farther (loc. dt., pp. 526-35). He will have nothing to do with this balance of periods and therefore cannot see that under certain circumstances a strophe can more easily spare two dimeters than one metron or even only a thesis. 174 Otto Schroedeb methods from Lachmann to Oeri, he will be equally welcome to us whether he improves it in detail or substitutes a wholly new method. But to accept the results in the interpretation of the single cola and then vilify the method to which we owe their secure possession is a procedure which recalls too vividly the poem of Goethe that ends — "er ist ein Recensent. " But I must break off here with the addition of only a brief word with regard to the problems of the history of verse. 4. History of Verse It was a promising step in advance when in 1886 in the excursus to his Isyllos mentioned above U. von Wilamowitz undertook to follow out the historical development of a single verse form. A second attempt followed in the Choriambische Dimeter (1902). But a weakness, not overcome to the present day, as we have seen, boded disaster to both works — he extended the term "Ionics" to include measures of bacchic catalexis, especially choriambics, and further to Aeolics like the Telesilleion 1 (6p-6pevop,eva joo «) and its cata- lectic form (rep- 0e2ia, — aWa vaei tivX kcu iv0ov