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. THE COINCIDENCE OF ACCENT AND ICTUS IN THE ROMAN DACTYLIC POETS By E. H. Stuktevant In Classical Philology, XIV, 234-44, 1 have applied a new criterion to the old problem whether the prevailing harmony of accent and ictus in Plautus and Terence was accidental or intentional. I determined statistically the tendency of the verse-structure to produce harmony in words of the several rhythmic types, and with this I compared the actual usage of the poets. The comparison showed that there was an effort so to place words in the line that harmony would result. There is a similar problem as to the Roman dactylic poets. 1 It is clear that there is usually harmony of accent and ictus in the last two feet of the dactylic hexameter and that clash of accent and ictus is more common than harmony in the first four feet, but scholars are still disagreed as to whether this state of affairs is altogether a result of the Latin system of accentuation and the structure of dactylic verse, or whether the poets paid attention to accent in con- structing their lines. In case we accept the second alternative there is the further problem whether their efforts were devoted solely to the securing of harmony in the last two feet of the line or whether they desired clash of accent and ictus in the first four feet. The dactylic poets had far less choice in the matter than the dramatic poets; in iambic and trochaic verse the ictus can be placed on any syllable of almost any word (firunt or ferunt, niscio or nescio), but most words can be got into dactylic verse with just one incidence of the ictus (ferunt, nescio). In fact the dactylic poets had free choice only in the case of words composed wholly of long syllables. For the rest their only recourse was to avoid such words as ferunt in case they desired harmony of accent and ictus, and to choose them in case they wanted accent and ictus to rest upon different 1 For a brief history of the discussion see Lucian Miiller, De Be Metrica Poetarum Latinorum?, pp. 233-37. Miiller fails to mention a suggestive paper by Humphreys in Trans. Am. Phil. Assoc, IX, 39-58. [Classical Philology XIV, October, 1919] 373 374 E. H. Sttjrtevant syllables. It should be easy to determine whether they did as a matter of fact choose words which require harmony in preference to those which require clash, or whether their choice varied between the different parts of the verse. To secure a standard of comparison I have counted the words of the several rhythmic types containing two, three, four, or five syllables in several prose passages, namely, Caesar B.G. i. 7, Cicero Cat. i. 1-7, Manil. 47-49, Cluent. 51, Caelius ap. Cic. Fam. viii. 1, Livy xxi. 23, 35. I have similarly recorded the rhythmic character of from 2300 -to 2500 words of each of four dactylic poets, Ennius, Catullus, Lucretius, and Vergil. Tables I, II, III, and IV give the detailed results in the form of percentages. Here and throughout this paper words are reckoned with that part of the verse in which their accented syllable falls, except that a dissyllable whose ultima receives the fifth ictus is counted as belonging to the fifth foot. Words with the rhythm ~ ~ and with the rhythms - - and ~ ~ « with elision of the ultima are all included under the heading " Require clash of accent and ictus," although no ictus at all rests upon them. To have taken into account this variation within the types - * and - - a would have made the tables too complicated for con- venience, and so I have preferred to make the necessary correction below (p. 379) in the supplementary discussion. The desire to keep the tables reasonably simple has also led me to include trochaic words with the spondees as permitting either clash or harmony. Trochaic words ending in a consonant had to be included because they are spondaic when an initial consonant follows, and it seemed unsatisfactory to separate forma and mitte from formam and mittit. I assume that such a word as armaque was regularly accented on the antepenult. 1 For convenience of comparison the figures in the several blocks have been added together, and they are presented in compact form in Table IV. The percentage of monosyllables has been appended; but even so the figures in each line of the table aggregate only 99 or a little more. The remainder is accounted for by the words of more than five syllables and by the decimals beyond the fourth place, which have been neglected in computation. i See Shipley, Class. Phil, VIII, 23 ff. Accent and Ictus in Roman Dactylic Poets 375 )l CO >l IN )l OS o >l © )l o >l ■a >l o )l CO )l 5 )l oo a. X c u p< * o o a a W a * o 3 « w OS-* HO «5 COIN O"* 00 CO CO CM. i-l i-H i-l -H OS CN ^ OS Tjl ooo-* o6r-i -HTiiooeqos (h K 03 PhWO - TO 3 3 •♦J , osr3, »h bo s 8 )l OS IN oo c )l -HOO >i o» o (N^Ji rH© )l C0 00-*b- COOO-h )l WCOCD»*lN HrtUJON WIN CO CON >l ) ) cocnkonco OOStOOi'* lo^icooco >l > CO ■»*< T* CO CN INO-*00— i >-l!-lt~.00© r4 r4 i—l a, IT c 1 1 ! c 3 1 c 5 hi 1 > ) >l OS © )l 00-* 00 t» CO COO-si coco T-H )l •>*IT)< s W© >l OS00IN CM NNrtiOH )l ©■* ■* OS -T hH tH iH cN IN COM >l cseocNOstn • 376 E. H. Sturtevant •-H _« i-h y < ft. o S" is o s « w PS o ■< O K s W H a PS s Ph COCNt^CO B cj 3 e p o t— < a Is l toco 00 )l 00 co >l CM CO IO )l OCNCN CO NO-*-* )l ) ) 6.96 7.92 ■ 13.04 9.45 >l ) 13.93 10.77 12.36 14.61 'S := 1 a; 1 1 a > ) a is "A H SS W o u o !» is O s PS < W w PS Of w )l )l CO o> CO CO "* ^ ■* CO )l )l I-H CO •* )l in-* co-* CM CM OS OS ^H CM ^H ^H >l ost-eocM ooo-*rt r-l >l • CO CD O )l .-H-#ascM CO-( l-H >l OS CM CON oocoo^h 10 10 US'* a 'E cr 1 1-; 1 a > 1 Accent and Ictus in Roman Dactylic Poets 377 o z o a < « o < O K H s Cii-t W tO ■<* H Q o H u o < o 3 o )l 1 ) ) )l ) ) 1 )l ) ) 1 ) >l 1 ) l-H )l ) ) co lO )l ) ) CO i-l CO lO i-H )l ) (Mi-no-* i-l •* to to lO " T— l" ' m a •a a a H O a a 6 3 o is O S « W H E B a )l 1 ) >l 1 ■-i-* to ) ) N^.-i ) )l OOlMt^TtH 1 00t~>O"# ) -* (MlOIN ) )l "OlM-^00 ) ■-HTtH-^CO 1 •* t>.io>o )l ) COIMCO t^i-H tO C35 ) )l hOihod ■*lNNO oo to to oo ) i-H t— 1 1-H i-H )l *o o tot~ ) oooioa OCQ^OO 1 T-H (Ni-H-H rr 02 3 nius tullu creti rgil. H O h-1 > 378 E. H. Stubtevant TABLE IV a* ' S 0°S P.sai p§ , .3 As** 8 Prose . (Ennius . . Catullus . Lucretius. Vergil. . . . I Ennius. . . ScS, Vergil I Ennius. . . Catullus. . Lucretius. Vergil. . . . 5.9 25.44 19.12 14.06 19.71 19.01 23.27 20.83 25.09 25.21 10.49 1.6 8.73 5.09 27.87 41.6 45.92 37.33 41.67 42.02 45.05 34.66 39.46 40.74 47.9 42.79 46.54 19.57 18.24 16.64 21 19.3 24.11 24.38 30 28.06 6.24 .41 2.67 .77 18.29 20.64 22.16 20.9 19.49 10.24 9.68 9.3 6.66 41.83 49.36 44.68 46.54 The avoidance of monosyllables in the last two feet of the hex- ameter by Ennius' successors has long been a commonplace; our table shows that Ennius himself used in these feet only about 29 per cent as many monosyllables as might have been expected (10.49 instead of 25.44 per cent). An examination of the hexameter cadences listed by Humphreys, op. tit., p. 43, shows that the use of monosyllables tends to produce clash of accent and ictus, at least in this part of the verse; it is very likely indeed that the reason why the later poets avoided such line-ends as Ennius' Diana Venus Mars, cum magnls dis, and restitult rem was their desire not to do violence to the accent. But that some other influence was at work is indicated by the small number of monosyllables in the earlier part of the line. Since, as we shall see, other rhythmic types which are avoided in the last two feet are as a rule considerably more common in the first four feet than in prose, we should expect to find the percentage of monosyllables in the first four feet higher than the 25.44 per cent which represents prose usage. On the contrary all the poets examined show a lower percentage than this, and the lowest of all appears in Catullus, who most rigidly avoids monosyllables in the Accent and Ictus in Roman Dactylic Poets 379 fifth and sixth feet. What this further motive was which led to the avoidance of monosyllables I am not now prepared to say; and it would be unsafe to base an argument upon the statistics in the second column of Table IV until that factor has been identified. The third column of Table IV shows that polysyllables whose penult and all preceding syllables are long were greatly favored by the dactylic poets, and that after Ennius they were particularly favored in the last two feet. As already noted, these are the words which admit either harmony or clash of accent and ictus. It is therefore likely that poets sometimes preferred, let us say, portant to ferunt in order to avoid the clash which the latter word would involve. Nevertheless there was another motive which led to the use of these words; they would fit into any position in the line, save only the second half of the fifth foot, provided they had the number of syllables required. The use of words composed of long syllables facilitated composition whatever attitude the poets took toward accent, and therefore we cannot base an argument upon the great frequency with which such words occur. We shall presently have to consider the actual incidence of the ictus upon them. The fourth column shows that in the last two feet of the line the poets avoided the use of polysyllables which require clash of accent and ictus. As noted above (p. 374) our figures require a cor- rection at this point; a large proportion of the words of the types ~ * and ~ - x employed in the last two feet carry no ictus, and con- sequently involve no real clash at all; for example, isque pium ex se, U neque posse, aid etiam Ipse haec. Consequently the avoidance of - i and ~ ~ i is even more rigid than the tables indicate; the dactylic poets, especially after Ennius, generally avoided placing in the last two feet words whose rhythm required clash of accent and ictus. The fifth column of Table IV shows that the poets correspondingly favored words requiring harmony of accent and ictus in the last two feet of the hexameter. The difference between the 18 per cent of such words employed in prose and the 42 to 49 per cent employed in these two feet is even greater than the excess just observed in the poets' use of words composed of long syllables. Here again the tendency is clearly marked in Ennius' usage and is carried farther by his successors. 380 E. H. Sturtevant If a word of a particular rhythm was to be avoided in the last part of the line the poet might either use a synonym of different rhythm or recast his sentence so as to place the objectionable word earlier in the line. Our tables indicate that both methods were employed. For, while the third block of Table IV, which represents the usage in the first four feet of the line, shows much larger figures in the fourth column than in the fifth, the difference is not nearly so great as in the fourth block, which represents the usage in the last two feet. Consequently in hexameter verse as a whole the words which require clash are a little less common and those which require harmony are a little more common than they are in prose. It remains to determine whether or not the dactylic poets tried to place the ictus on the penult of words whose penult and all pre- ceding syllables are long, that is, of those words which alone permit either clash or harmony of accent and ictus. Our standard of com- parison here must be the tendency of hexameter verse to place the ictus on the penult rather than the antepenult of such words. To determine this tendency statistically I have "scanned" 200 or more verses of each of four dactylic poets, as follows: Aen. i. 1 Then I have picked out all possible combinations of a long syllable and a following syllable of either quantity, and have recorded the incidence of the ictus. In the above line syllable-groups with the rhythm - * occur with ictus on the penult six times and with ictus on the ultima twice. In 320 lines of the Aeneid they occur with ictus on the penult 1920 times and with ictus on the ultima 233 times; that is, the ictus falls on the penult in 73 per cent of the occur- rences; the structure of Vergil's hexameter tends to put the ictus on the penult of syllable-groups (or of words) with the rhythm - * 73 per cent of the time. For the second term of the comparison I have counted the incidence of the ictus upon words whose rhythm is - ^ in about 400 lines each of the same dactylic poets. In this computation I have ignored the words with elided ultima which carry no ictus. In the first 400 lines of the Aeneid words of the type - * have the ictus on the ultima 258 times and on the penult 466 times or 64 per Accent and Ictits in Roman Dactylic Poets 381 cent of all occurrences. While the structure of the verse tended to place the ictus on the penult of these words 73 per cent of the time, Vergil allowed it to stand there only 64 per cent of the time. Our problem, however, is chiefly a matter of the last two feet. I have therefore divided the line as above (see p. 374) ; each word has been assigned to that part of the verse in which its accented syllable stands, except that a dissyllable upon whose ultima the fifth ictus rests has been reckoned as belonging to the last part of the verse. Counting syllable-groups and words as before, I find that the struc- ture of the fifth and sixth feet tends to place the ictus on the penult of words of the rhythm - » 73 per cent of the time, and that Vergil actually placed it there 100 per cent of the time. A similar calcula- tion for the first four feet indicates a tendency for the ictus to rest upon the penult 73 per cent of the time, whereas Vergil actually placed it there only 39 per cent of the time. Table V gives the results of similar calculations for words of the three types -«, — », », and for the four poets, Ennius, Catullus, Lucretius, and Vergil. At the head of each block is an indication of the rhythmic group to be treated. The Roman numerals indicate the incidence of the ictus; I stands for the ultima, II for the penult, and III for the antepenult. In case a word has ictus on two syllables, only the ictus on penult or antepenult is recorded in our count. In each block the column which corresponds to the accent is printed in heavy-faced type. The third, sixth, ninth, and twelfth number in each heavy-faced column is the difference between the unhampered tendency of the verse-structure and the poet's actual usage. The table shows that the dactylic poets tried to put the ictus on the penult of spondaic words in the last two feet of the verse, and that the tendency became stronger after Ennius. There was even greater regularity in placing the ictus on the penult of words of the rhythms — * and a , but this appears to be chiefly due to the structure of the verse. If, however, we had reckoned to the fifth foot syllable-groups and words of the rhythms and whose ultima has the fifth ictus it would appear that the poets took great pains to avoid clash of accent and ictus in these words. Accord- ing to Humphreys, op. tit., p. 52, the fifth ictus rests upon the- 382 E. H. Stuktevant m m < )1 1 ■O rH M ISO IS a>o T-l §§ OO o>o CBO OO T-l tH T-l T-l T-l T-l M IO t >l 1 M m OIN 1— 1 IO IO 8 1 t-l ISO IO HI. to o»o tH OO o>o OOJ OIO oo En T-l T-l T-l T-l t- t- to to 1-4 IOt- t- S2g t- t>o> IN toco CO t-o in l T-l CO ■* w wco tH to oo tH toco TH M 00 IO OS OOO c# )l t-l w to toco SO-* com O) io C- on CO OtM in o to CO fe l-H t*CO ! tH tO o> ■*lO T-l tUtJI OtH to T-IOJ 00 CO cn •>* coco tH )l M t-Tjl 1 t-co CO 1 t-co CO 1 t-co CO 1 OZO O00 r-1-l r»TH ma INtO (NO IN to )l HH tHOO !NO or- 00 r- M coco CO NH NH a>e> t- 00 T-l CO tHCO (N «eo t-00 TH to to coca in t-CO tH tH s a © i 1 KH ©HO oco oco OS )l KH O-* wco •oco IOCO OIO IO ot- t- ot- t- oco co tH IOIO ■OtO tH to to T-l IO to tH O l M t-co 1 t-to 1 t-to TM 1 t-to 1 00 t- ooo t-r- t-co IN CO coco Accent and Ictus in Roman Dactylic Poets 383 ultima of a dissyllable or polysyllable in only 5.2 per cent of Ennius' verses, and only four-tenths of one per cent of the time in Lucretius and Vergil. My count gives 1.73 per cent for Catullus. This increasing avoidance of line-ends such as ignis mare ferrum, mortales ■perhibebant, amatorim quod amid, Tuditano collega must have been due to the effort to make accent and ictus rest upon the same syllable. Table V shows further a very striking tendency to favor ictus of the ultima of spondaic words in the first four feet. This tendency is no doubt partly due to the restriction of words of the rhythms ~ * and -vsto this part of the verse; for a spondaic word standing next to one of them must ordinarily have the ictus on the ultima; for example, Tyrids olim, currus fuit, mihi causds memord. But words beginning with one or two short syllables are too few to account for so great an excess of spondees with accented ultima as we actually find. Besides many a line could easily be shifted so as to put the ictus on the penult instead of the ultima of a spondaic word. The following lines of the first book of the Aeneid might have run: 1 Anna virtimque can6 qui Tr6iae primus ab 6ris, 50 Talia s£cum flammato dea c6rde vohitans instead of tdlia flammato secum, 104 R6mi franguntiir; turn pr6ra avSrtit et tindis instead of frdnguntur remi, 159 fist locus In sec^ssu 16ngo: insula p6rtum instead of ist in s6cessu longo locus, 167 Diilces fntus aquae viv6que sedflia saxo instead of intus aqude dulces. Similar easy shifts would avoid a clash of accent and ictus in a spondaic word in lines 8, 32, 35, 53, 64, 68, 69, 72. There may be objections to these specific changes; but if such alterations are so easy, Vergil could undoubtedly have found many which were free from objection if he had desired. It seems necessary to conclude that the dactylic poets preferred frequent clash of accent and ictus in the first four feet, at least in the case of spondaic words. 384 E. H. Sturtevant Some will criticize my discussion of the first four feet (particularly some of the suggested rearrangements of Vergilian lines) because no account has been taken of the caesura. I have two lines of defense against such an attack. It is likely that the position and character of the caesura was dependent on considerations of accent (compare Lucian Muller, op. tit., p. 200). For example, the Roman prefer- 1 ence for the masculine caesura was probably due to a desire for con- flict between accent and ictus at that point in the verse. Even if there should prove to have been some other cause for this deviation from Greek usage, that other cause, not the position of the caesura, would have to be taken into account in our investigation. In the second place, if I had taken account of the caesura it would not materially have changed the complexion of the figures. The alterations would have been confined to Table VI, and there would have affected only the first six columns. I have made the calculations for the first four feet in Vergil, with these results : — X _ . . s _ Si I II II III II HI Syllable-groups Words 32 61 68 39 -29 39 46 7 61 54 42 81 31 58 19 All of the above remarks apply to these figures as well as to those given in Table V. It is quite clear, then, that the dactylic poets tried to secure harmony of accent and ictus in the last two feet of the verse; and this effort was stronger, or at least more successful, with the later poets than with Ennius. It seems likely that Ennius, like Plautus and Terence, would have preferred harmony of accent and ictus throughout the verse; but since that was impossible in the hexam- eter he gained much the same effect by preserving fairly regular harmony in the last two feet. Ennius seems thereupon to have made a virtue of necessity and to have employed clash of accent and ictus in the first four feet to help give his language that air of aloofness from ordinary speech which he regarded as essential to heroic verse. Accent and Ictus in Roman Dactylic Poets 385 The later dactylic poets seem to have confined this device to spondaic words, but in this limited field they employed it even more than Ennius had done. 1 If the dactylic poets cared to secure harmony of accent and ictus in the last two feet of the line, or if they aimed at any particular relation of accent and ictus in any part of the line, it follows that both accent and ictus involved stress. The Latin accent, in classical times as well as in the early period, was in part a stress accent. Columbia University 1 Possibly the reason why Latin poets could employ familiar prosaic phrases was that a clash between accent and ictus removed them from their commonplace asso- ciations. At any rate Vergil has such a clash in the phrase (Aen. i. 207) : rebus servdte sec&ndis, and Horace in (Carm. i. 14. 2) (5 quid agis ? (Carm. iii. 5. 42) capitis mindr, etc.