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 PLAUTUS AND TERENCE By E. H. Sturtevant That the Roman poets, and especially the early dramatists, tried to construct their lines so that accent and ictus would fall upon the same syllable has been the opinion of a long line of distinguished scholars. 1 Others, however, have maintained that such harmony as we observe in Latin poetry results solely from the nature of Latin accentuation and the rhythm of ancient verse. In the first place the Latin accent usually rests on a long syllable or, less often, on a short syllable followed by another short, and ancient verse forms place the ictus on a long syllable or, in some cases, on a short syllable before another short; hence the two tend to coincide. In the second place it is argued that the kinds of caesura preferred by the Roman poets favor harmony of accent and ictus. To consider the second argument first, there is undoubtedly a connection between the position of word-ends in the verse and harmony or clash of accent and ictus. Shall we then assume that the Roman poets for an unknown reason departed from their Greek models in the employment of caesura and consequently produced harmony of accent and ictus, or shall we suppose that they tried to secure harmony of accent and ictus and consequently departed from the traditional treatment of caesura ? To me the second alternative seems the more satisfactory. At any rate we are under no obligation 1 For a history of the discussion up to 1870, see Corssen, Aussprache Vokalismus und Betonung d. lat. Sprache, IF, 949 ff. Since that date the doctrine has been sup- ported by Humphreys (although unwillingly), Trans. Amer. Phil. Assoc, VII, 107-45; IX, 39-58; Lindsay, Jour, of Phil, XX, 135-58; The Captivi of Plaulus, pp. 257-374; Skutsch, Plautinisches und Romanisches, pp. 136, 153-56; Satura Viadrina, pp. 122- ii=Kleine Schriften, pp. 92-113; Ahlberg, Sludia de Accentu Latino, De Proceleus- maticis Iarnborum Trochaeorumque Antiquae Scaenicae Poesis Lalinae, De Correplione Iambica Plautina. The other side of the debate has been taken by W. Meyer, Abhand- lungen der k. bayerischen Akad., phil.-hist. Klasse, XVII, 1-120, and Lucian Muller, De Re Metrica Poetarum Latinorum 1 , pp. 233-40. For the theory that Latin accent was a matter of pitch, see especially Vendryes, Recherches sur I'histoire et les effets de VintensiU initiate en latin, p. 74 and references. Bennett's theory of the ictus was propounded in the Amer. Jour. Phil., XIX, 361-83. [Classical Philology XIV, July, 1919] 234 Accent and Ictus in Plauttjs and Tebence 235 to accept the first hypothesis. I shall therefore in this paper dis- regard all questions of caesura on the ground that the reasons why certain types of caesura were preferred are quite unknown, and that caesura is at least as likely to have resulted from as to have caused the position of the accented syllable in the verse. It is my present purpose to test, as far as Plautus and Terence are concerned, the dogma that the harmony of accent and ictus in Latin poetry results from the tendency of both the Latin accent and the ictus of classical verse to stand on long syllables or, less often, on the first of two short syllables. I have undertaken to measure statistically the tendency of iambic and trochaic verse to place the ictus upon the penult when long and upon the antepenult when the penult is short, and I have compared with this the actual usage of Plautus and Terence. The procedure has been as follows. A number of iambic and trochaic fines (about 275 verses of Plautus and as many of Terence) were "scanned" thus: Persa 5: Then all possible combinations of a long syllable and a following syllable of any quantity were picked out and the incidence of the ictus was recorded. In the foregoing line the group _ s occurs with ictus on the ultima six times and with ictus on the penult five times. In the lines of Plautus which were studied the syllable group _ s occurs with the ictus on the ultima 922 times and with the ictus on the penult 1,249 times; that is, the ictus falls upon the penult in 58 per cent of the occurrences. The structure of Plautine iambics and trochaics therefore tends to place the ictus upon the penult of the syllable group _ s 58 per cent of the time. For the second term of the comparison I have recorded the incidence of the ictus upon all the dissyllabic words with long penult in Plautus' Persa. In a number of occurrences such words carry no ictus at all (that is, in case of iambic shortening of the first syllable or elision of the second syllable). We need not decide at present whether any significance is to be attached to the use of a word without an ictus; it is clear that even if such words were accented in ordinary speech Plautine verse cannot teach us anything about their accent. We have therefore disregarded all words which do not have an ictus. 236 E. H. Stubtevant In the iambic and trochaic lines of the Persa dissyllables with long penult have the ictus on the ultima 148 times and on the penult 877 times, or in 83 per cent of the occurrences. Whereas the struc- ture of the verse tended to place the ictus on the penult of words whose rhythm is _ * only 58 per cent of the time, Plautus managed to place it there 83 per cent of the time. At least in words of this type there is evidently some powerful factor at work aside from the struc- ture of the verse; and since an effort to make accent and ictus coincide would work in precisely this way it is natural to infer the existence of such a desire. If we find that words of other rhythmical structure behave similarly the inference will become irresistible. The statistics in Tables I and II cover the dissyllables and trisyllables of the Persa of Plautus and the Andria of Terence, and the quadrisyllables of the Persa, Poenulus, Andria, and Heauton Timoru- menos. I have counted the following numbers of words of the several rhythms: 798 869 877 1244 381 499 309 573 173 235 333 401 113 99 w w w 209 296 55 44 168 187 142 127 160 154 23 31 167 145 Such groups as noctisque, lepidumst, itanest, iubedum, which contain universally recognized enclitics or are customarily printed as single words, have been treated as single words. 1 Our statistics would be more significant if we had gone farther in recognizing sentence accent; but to do so seemed scarcely possible in the present state of knowledge. For a similar reason no account has been taken of the exceptions to the penultima law recorded by the later gram- marians; we do not know the precise limits of such accentuations as illic, adhuc, nostras, and conslli, and we do not know that any such accentuations are as early as Plautus and Terence. Consequently 1 It is assumed that the accent of such combinations followed the penultima law; see Shipley, Class. Phil, VIII, 23 ff. Accent and Ictus in Plautus and Terence 237 such words have been grouped with others of the same quantitative rhythm. Words which have been shortened by the iambic law have been counted according to their original, unshortened form, in the belief that iambic shortening is always a function of the historic accent. Words with elided ultima have not been separated from those whose ultima has syllabic value; in Andria 21, potiiis quam ist6ram obsciiram diligentiam, istorum and obsciiram have been put in the same category. It is altogether likely that the accent of words with elided ultima some- times receded, especially in the case of the enclitics -ce, -ne, and -que; 1 but until the details of the matter are worked out it is impossible to take account of it in a statistical study. It is quite clear at any rate that we cannot disregard elided syllables in the present study; we find such ictuses as repudi(o) and benefici(um) , although trisyllables and quadrisyllables consisting entirely of short syllables never have the ictus on the penult. As might be expected, however, the ictus disagrees with the penultima law somewhat more frequently in elided words than in others; in words of the rhythm _ _ a the ictus stands on the antepenult in Plautus' Persa 35 per cent of the time, while 38 per cent of the elided words show that ictus. Our figures therefore would have been slightly more favorable to the theory of intentional coincidence of accent and ictus if the elided words had been eliminated. Tables I and II exhibit the results of our count in the form of percentage. At the head of each block is an indication of the rhyth- mic group to be treated. The Roman numerals in the second line indicate the incidence of the ictus: I stands for the ultima, II for the penult, III for the antepenult, and IV for the fourth syllable from the end. In case a syllable group or a word has two ictus syllables, one of which is the penult (e.g., libertdtis), it is counted only in column II. If there are two ictus syllables, one of which is the ultima (e.g., ddmigrdnt, fdcilius), the word is counted in column III or IV, as the case may be, and not in column I. In each block the column which 1 It should be possible to fill this gap in our knowledge by the method here employed, and a student of mine is at work upon the problem. 238 E. H. Sturtevant 9 § (Li 00 w b-iO ON )1 M <#CO . ^CO t* c- w tH 1 ■OIO tH iau> » CO d (— 1 X ■•* )l M . H4 oco 00 Off* M 1— 1 fc-00 *H fr-O* « ) - ©-* OCO CO CO w fr- fr- >-t et- fr- oo en «H >l t-at et t-ct» M > ON 1 *tr* 04 WOO ff> > I— | r-< ■ HM 1-4 ) o t-H Hl M NN -*^ T*^ •*H MN o toio o» M ff»*" *H •■* •o w ON WH ©10 ot> om «h t-m co *«*« «*-i *-* M 1 1 oo OG oo OO *-*tH r(rt M >; COM* Hr . -t c iO-tf - «US CO 0»ffl O vats e W ,_, «H too * t> r-t^« oox )l iOCF -*x «H fr-e* to M » « «rt M ) ►H 1 1 ) ) M CO i0 M ■^4 "^ hi »-* ©ft ©a . ,2 t- R- SfcCC S aoc £ & CO P GO I © o s 5 H ^2 !U on, M IS O (8 •o- w o go) fl o o ^ h« ■a a o" es A- *g as Accent and Ictus in Platjtus and Terence 239 corresponds to the accentuation of classical Latin is printed in bold- faced type. The third and sixth number of each bold-faced column is the difference between the two preceding numbers; it represents the difference between the unhampered tendency of the verse struc- ture and the poet's actual usage — in other words, the strength of the effort to produce harmony between accent and ictus. With four exceptions the figures in all blocks of both tables seem quite conclusive. The case appears in fact to be definitely proved by these statistics; but it will be worth while to append a few remarks. The figures in the first block of Table I show that while an effort was made to put the ictus upon the penult rather than upon the ultima of pyrrhic and iambic words, Plautus nevertheless succeeded in doing so only 42 per cent of the time and Terence only 35 per cent of the time. These low figures are largely due to the difficulty of constructing the iambic close, which requires an ictus on the ultima. If we eliminate the iambic close from both sides of our comparison, we find that Plautus puts the ictus on the penult of syllable groups in 38 per cent of the occurrences and of words in 68 per cent, while in Terence the ictus rests on the penult of 34 per cent of the syllable groups and of 66 per cent of the words. On the basis of these figures, the poets' effort to secure harmony between accent and ictus appears to operate in 30 per cent of the words in Plautus and in 32 per cent of them in Terence. It has long been known that words of the type of fatilius usually have the ictus on the first syllable. Lindsay {Latin Language, pp. 173 f.) brings this fact into connection with the post-Plautine syncope of balineae into balneae and the like, which compels us to assume an early accentuation of these words upon the fourth syllable from the end. The recognition of such accentuation brings the figures in the first block of Table II into harmony with the others. Syllable groups and words of the rhythm _ „ _ s can be got into the verse in just one way, and so we cannot expect any evidence from them. The figures in the next to the last block in Table II seem to indicate slightly less harmony between accent and ictus in the actual usage of the poets than the structure of the verse tended to produce. 240 E. H. Sturtevant Since most of the instances of ictus on the long penult show iambic shortening of the second syllable, and since we find two ictus positions aside from the one in harmony with the accent (i x _ £ and *, s. _ ±), it is probable that there was some metrical difficulty in getting words of this type into the verse. The matter evidently has some connec- tion with the tendency to avoid dividing the two short syllables of a dactyl or anapest between two words one of which is a polysyllable; but the limitations of this tendency itself and the reasons for it are far from certain. 1 It is noteworthy that words of the rhythm ~ — a are comparatively rare; our figures are based upon twenty-three words in Plautus' Persa and Poenulus and thirty-one in the Andria and Heauton. This slight discrepancy, however it is to be explained, cannot weigh against the decisive figures in the other blocks. Very significant are the columns which indicate in the syllable groups an ictus which is regularly avoided in the words. The structure of the verse calls for such an ictus as ^ i a in 11 per cent of all occurrences, but it does not occur in any word in the iambic and trochaic parts of the Persa, and in the Andria it occurs only once in the exclamatory hiclne. The ictus _ i. 2 might be expected in from 22 per cent to 32 per cent of all words; but it occurs only in the compound indefinite nesclo quis, where it was probably quite regular, in hocclne, siclne, etc., and in a very few other words. Equally remarkable is the avoidance ofvis and _ i a at the end of words of four or five syllables. This peculiarity of early dramatic verse has long been familiar to scholars; it can scarcely be due to any other cause than a desire to harmonize accent and ictus, especially in view of our demonstration that the structure of the verse tends to place an occasional ictus on these short syllables. 2 Words of more than four syllables are so much less frequent than those we have been discussing that they might safely be neglected. Nevertheless I include a table on the five-syllable words, basing it upon the iambics and trochaics of the Amphitruo, Asinaria, Aulularia, Bacchides, Persa, Poenulus, Pseudolus, Rudens, Stichus, Trinummus, 'See especially Klotz, Grundzilge der altromische Metrik, pp. 254 ft., 307; Leo, Plautinische Forschungen, pp. 236 ff.; Maurenbrecher, Hiatus und Verschleifung im alten Latein, pp. 26 ff.; Lindsay, The Captivi of Plautus, pp. 68 f. 2 See especially Ahlberg, Studia de Accentu Latino, pp. 35 f . Accent and Ictus in Plautus and Terence 241 and Truculentus of Plautus, and the six plays of Terence. I have counted the following numbers of words of the several quantitative rhythms: Plautus . . . Terence... 13 13 15 18 28 12 29 25 124 38 148 66 15 2 201 83 Plautus. . . Terence. .. 76 26 226 99 77 21 108 67 53 32 23 16 10 1 37 10 The totals are in some cases so small that percentages based upon them would be misleading; for example, I have found ten words in Plautus of the rhythm „ a, all with the ictus on the penult, while Terence has only one such word (venustatisque, Hec. 848), and it has the ictus on the antepenult. Under these circumstances it seemed preferable to combine the statistics for Terence with those for Plautus, in order to secure larger totals and correspondingly more reliable percentages (Table III). There are several five-syllable rhythms which can be got into the verse in only one way, and consequently four blocks show 100 per cent for syllable groups and words. In three of these cases the ictus falls upon the accented syllable; but the rhythm _ ~ _ _ x requires an ictus on the antepenult, although the accent falls on the penult. None of these blocks can contribute any evidence. In nearly all the other blocks there is evidence, in most cases very strong evidence, that the poets tried to secure harmony between accent and ictus. Especially noteworthy are words of the rhythm 2 and „ s:, which always have the ictus on the ac- cented antepenult, although the structure of the verse would favor ictus on the penult in 43 per cent and 23 per cent of the occur- rences, respectively. While words of the rhythm _ w w ^ ~ (e.g., adsimiliter) show ictus on the antepenult a little more often than the similar syllable groups, there is a much stronger tendency to put the ictus on the fourth syllable from the end. It seems probable that these words were 242 E. H. Sttjrtevant B °° m & Eh O )l M M oo oo )l 5 ■ » tocn eo tocn i— i CD CO )l > HH B GO >l > •"tcD B 1 t— 1 o )l > HO (-( we •* (-( CO rH )l !> 5 i-i OS CO p. 3 O & CO )i H •*o TttTtt (DO "# m >i B GO OS «»H o t»o» *-l >i 1— 1 1— 1 oc oc >i B OCT **,- eet M ^ (0 00 M >i oo oc r*** >i oe oc lHi-i >i > 1-1 !0^ B enw « CO 00 •* >i > •^ CO > > HH we- ^t > tDO) m >i „ « " c-o «H w s CO >i H n " *•< «* s CO •* A p bo 3 a 0?p Accent and Ictus in Plautus and Terence 243 accented on the fourth-last syllable in the early period, precisely as was the case with facilius, etc. (cf. Lindsay, The Latin Language, pp. 173 f.). That our figures lend no support to the accentuation pukritiae, which is indicated by Horace's puertiae, is probably due to the fact that twenty-two of our twenty-six words of this type are compounds whose second member contains three syllables (bene- ficium, maleficium, fidicinium). At any rate all of these compounds have the ictus on III or V in every occurrence. 1 The figures in our fourth block seem to indicate an early accen- tuation amlcitia; but the frequent iambic shortening of such, words as minlsterium and calefieri (cf. Lindsay, The Captivi of Plautus, p. 34) and the occasional shortening of amlcitia and pudicitia (cf. Skutsch, Satura Viadrina, p. 130 = Kleine Schriften, p. 100) prove that the accent of these words rested on the antepenult. There has been a strange unwillingness to admit the operation of the iambic law in the two words last mentioned, and consequently all editors prefer to scan them with the second syllable long wherever that is possible. In compiling my statistics I have followed the marking of these words in the standard editions; but I am convinced that we should always prefer the scansion amlcitia to amicitia, where either one is possible. This is the case in Amph. 899 (pudicitiae), Poen. 1215, and Stichus 414. If we transfer these three words from one column to the other, our percentages are III, 47.5 and IV, 52.5, and the —3 becomes +4.5. It is noteworthy that of the remaining twenty-one words with ictus on the fourth-last syllable seventeen stand immediately before the iambic close; for example, Rud. 1239: declpitur fn trans&ma av&ritid, su£. Most of them, like this one, are followed by dissyllabic enclitics. Our study, as stated above, has necessarily been confined to word accent and the universally recognized enclitics. On the basis of 1 1 have listed benefacere with words of the rhythm „ _ ^ w a, since bene was origi- nally an iambus. The weakening of a to i in the third syllable of beneficium and male- ficium, however, indicates that these compounds are older than the corresponding verbs, and were originally *beni-ficium and *mali-ficium; wherefore the nouns have been listed here. It is possible that at some time before Plautus the verb benefacere altered the noun to beneficium, in which case I am wrong in placing the nouns here. It is equally possible that after the shortening of the second syllable of the verb it affected the vocalism of the noun, and at the same time the accent of the noun was assimilated to that of the verb. 244 E. H. Sttjktevant this material we find a strong and nearly consistent effort to place the verse ictus upon accented syllables. But it has been demon- strated that the ictus fell upon the syllable which had the sentence accent in many places where it seems to conflict with the traditional word accent. It follows that if we could have based our study upon the sentence accent our percentages would have indicated a still stronger tendency toward harmony of accent and ictus. It is scarcely possible any longer to doubt that accent was an important feature of early dramatic verse; the quantitative nature of the measures was carefully preserved, but at the same time accent was constantly taken into account. Two corollaries follow from this demonstration. It has been held by Bennett and others that the ictus of classical poetry involved little or no stress, that it was merely the quantitative predominance of the long part of the foot. If that were the nature of the ictus it is impossible to explain an effort on the part of the poets to make it fall upon the accented syllables. The verse ictus, at least in the early period, must have involved more or less increase of stress. In the same way it becomes impossible to deny an element of stress to the Latin accent. 1 As I have shown elsewhere (Trans. Amer. Phil. Assoc, XLII, 50), this does not compel us to doubt the explicit statements of the Romans that their accent was one of pitch. It was no doubt expiratory and musical at the same time; which element was the stronger we have no way of determining. Columbia University 1 So Skutsch, Glolta, IV, 188 =KUine Schriften, p. 491.