We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Published online by Cambridge University Press: 11 February 2009
In several recent studies we have developed precise statistical methodologies which have demonstrated that the cursus mixtus was the dominant rhythmical system for final clausulae in Latin prose from the third century a.d. to the fifth. The cursus mixtus consisted of four standard metrical forms derived from the richer variety of Cicero's Asiatic tradition – cretic-spondee, dicretic, cretic-tribrach and ditrochee –, which were structured according to three accentual patterns – planus, tardus and velox. The latter are differentiated by the number of unstressed syllables intervening between and following two accented syllables. The planus has two unaccented syllables between two word accents and one after the last accent. The tardus has two between two accents and two after the last accent. The velox has four between two accents and one after the last accent. The four metrical forms are contained within the parameters of the accentual cadences. The planus contains either the cretic-spondee (epistulae nostrae) or the ditrochee (manu ferire); the tardus either the dicretic (carcer includeret) or the cretic-tribrach (estis indigenae); the velox either the ditrochee (experientiam singulorum) or the cretic-spondee (nimium videbatur. Authors who used the cursus mixtus, however, preferred to effect an exact coincidence between the accent and ictus, and as a result evinced primarily the following standard rhythmical forms: planus/cretic-spondee; tardus/dicretic or cretic-tribrach; and velox/ditrochee.
So far now we have examined only the rhythmical properties of final clausulae, but it is appropriate also to investigate for rhythm at internal positions within the sentence.
1 Oberhelman, S. M. and Hall, R. G., ‘A new statistical analysis of accentual prose rhythms in imperial Latin authors’, CP 79 (1984), 114ff.Google Scholar; ‘Meter in accentual clausulae of late imperial Latin prose’, CP 80 (1985), 214ff.Google Scholar; Hall, R. G. and Oberhelman, S. M., ‘Rhythmical clausulae in the Codex Theodosianus and the Leges Novellae ad Theodosianum Pertinentes’, CQ 35 (1985), 201ffCrossRefGoogle Scholar.
2 For more detailed discussion of the cursus mixtus, see our paper in CP 80 (1985), 215ff.Google Scholar; for a bibliography on the cursus mixtus, see ibid. n. 11.
3 We use the term ictus merely to denote the first syllable of the metrical unit. The question of whether an ictus existed in metrical prose, not to mention poetry, is too complicated to be dealt with here. See our comments in CQ 35 (1985), n. 5Google Scholar.
4 There is a fourth accentual scheme, the trispondaicus, which is the accentual abstraction of the metrical pattern paeon-spondee. The trispondaicus has three unaccented syllables between two word-accents and one after the second accent, e.g. inveníre valeátis and vérba gradiátur. Among the cursus mixtus authors the paeon-spondee often falls under the accentual scheme, e.g. 
and 
. As we have demonstrated elsewhere (CP 79 [1984], n. 6Google Scholar), this rhythm cannot be admitted as an integral part of the cursus mixtus system. It can provide a useful guide to determining an author's rhythmical tendencies; see the discussion in note 11 below.
5 Arnobius: Hagendahl, H., ‘La Prose métrique d'Arnobe. Contributions à la connaissance de la prose littéraire de l'Empire’, Göteborgs Hogskola Årsskrift 42 (1937), 1ff.Google Scholar; Ammianus: Harmon, A. H., The Clausula in Ammianus Marcellinus (New Haven, Conn., 1910)Google Scholar; Historia Augusta: Zernial, H. L., ‘Über den Satzschluss in der HA’, Deutsche Ak. der Wiss. zu Berlin. Schr. der Sektion für Altertumswissenschaft II (1956)Google Scholar.
6 As a rule we sampled clausulae only if the number of words exceeded four.
7 For example, at Symm. Orat. 2.4 we read ‘quisque arcana scrutari postulat, te sequamur’. Here postulat is needed to form the first three syllables of the velox configuration and the metrical scheme, which is a cretic-ditrochee.
8 We have placed the standard metrical forms in bold face for easy identification.
9 See our tables in CP 79 (1984), 122ffGoogle Scholar. and 80 (1985), 222ff.
10 For ancient opinions on Cicero's practice and the lengths to which his imitators went, see Tac. Dial. 23 and Quint. 10.2.18.
11 The greater number of the trispondaicus óo/ooóo forms in Palladius' internal clausulae, as compared to the number of velox óoo/ooóo forms, reinforces Michael Winterbottom's ‘velox rule’: review of Janson, T.'s Prose Rhythym, Medium Aevum 45 (1976), 298ff.Google Scholar, and ‘Aldhelm's prose style and its origins’, Anglo–Saxon England, 6 (Cambridge, 1977), 71ffGoogle Scholar. Winterbottom has accurately oberved that, because Latin has a natural preference for paroxytones rather than proparoxytones as the penultimate word in a sentence-closing, a favouring of the proparoxytone velox form óoo/ooóo over the paroxytone trispondaicus form óo/ooóo indicates rhythm in an author. Palladius' greater use of trispondaicus clausulae over velox supports our assertion below that he was not as careful in his adherence to accentual rhythms internally.
12 L. Stephens, ‘Some generalizations concerning syllable quantity in late Latin clausulae’, forthcoming in Phoenix.
13 We use the term miscellanei to describe those accentual clausulae that do not fit under any of the other rubrics.
14 Some, e.g. Hagendahl, op. cit. 32–3, add a fifth pattern: óoooóoo, which is a resolution of the tardus or velox. This pattern does appear in Ammianus and in cursus mixtus authors, since the metrical coincidence is a paeon-cretic or paeon-tribrach. The incidence of this pattern, however, is never great enough (about 2%) to justify treating it as a major component of the cursus mixtus.
15 We have sampled all the genuine works of Tertullian and have found them to contain purely metrical rhythms (66·5–80·2%). No accentual tendencies can be determined according to our statistical models.
16 Although the percentages of accentual clausulae containing a standard metrical form irrespective of coincidence of accent and ictus are fairly high, they do not approach Symmachus' practice in final clausulae. For reasons why Symmachus' internal rhythms do not show similar properties, see Specific Observations (8) and (9).
15 The slightly greater diversity of typologies in final clausulae could be due to the very large sample taken of final clausulae (572) vs. internal (248).
18 De mortalitate, accentual 81·1% and metrical 86·0%; De habitu virg., 76·3 and 78·9%; Epistulae, 79·2 and 81·9% ; De cath. eccl. imitate, 72·3 and 74·6%; and De dom. oral., 701 and 79·7%.
19 OCD 2, 772, s.v. Palladius. R. H. Rodgers in his 1975 Teubner edition of Palladius states, xvif., that Palladius was a ‘homo…eruditus et numerorum rationem numquam oblitus’. Rodgers may therefore be presumed to have used prose rhythms in drawing up his edition.
20 CP 79 (1984), 120ffGoogle Scholar.
21 Wilkinson, L. P., Golden Latin Artistry (Cambridge, 1970), 157ffGoogle Scholar.
22 We exclude from consideration the works De mundo, De Platone and Peri Hermeneias which, as we have demonstrated elsewhere (CP 79 [1984], 122)Google Scholar, contain accentual rhythms. The debate over the authenticity of these works is still raging. See our discussion in CP 79 (1984), 127 n. 35Google Scholar.
23 CP 79 (1984), 122fGoogle Scholar.
24 Ibid. 130.
25 Wilkinson, op. cit. 163, following Norden, E., Die antike Kunstprosa 5 (Berlin, 1898), ii. 951Google Scholar.
