Page images

that it would remain highly probable for ille, inde, immo, nēquis, etc.

It may perhaps be said that, in this discussion of the value of nempe, we are taking refuge in mere verbal quibbles, and that all syllables have, in practice, the value either of one or of two morae. On the contrary, the assumption that early Latin prosody had definitely accepted a sharp division of syllables into those of one and of two morae, is not only, in my judgment, wholly unwarranted, since so rigid a scheme of the syllables confessedly involves many artificial and adventitious elements (cf., for example, Victor Henry, Comp. Gramm., Engl. transl., 85), but is also distinctly negatived by the extensive phenomena of the Iambic Law. As is well. known, this question of syllable values was discussed by the ancient theorists, and gave rise to a division into two schools. One of the two schools, it is true, the metrici, recognized only long and short syllables, that is, syllables of one and two times, but the more scientific school, founded by Aristoxenus and called the rhythmici (or musici), held that many both of the long and of the short syllables differed from each other in quantity, and they expressly recognized in speech syllabae longis longiores, syllables longer than the long, and syllabae brevibus breviores, syllables shorter than the short. Cf. Marius Victorinus, i, 8: nam musici non omnes inter se longas aut breves pari mensura consistere, siquidem et brevi breviorem et longa longiorem dicant posse syllabam fieri. metrici autem, . . . neque breviorem aut longiorem, quam natura in syllabarum enuntiatione protulerit, posse aliquam reperiri. No doubt it is true that, apart from the phenomena of elision2 and from the occasional cases of natural synizesis

1 Cf. Quint. ix, 4, 84, and for a collection of numerous other references to the doctrine of the rhythmici, v. Goodell, Chapters in Greek Metric, 6 f., and Christ, Metr.2 77 f.

2 The elided syllables (so called) were in nearly all cases too short to be definitely measured or to be taken into account metrically, but no one supposes that they were always completely expelled, and were always left entirely unpronounced; at least, this was not the case with Roman elision. Hence we may safely assert that, even if it were necessary to explain the final syllable of nempe -the one word—as neglected in the metre, it would not certainly follow that

in Greek (v. p. 167), this doctrine of the rhythmici is not of such manifest practical importance in Greek poetry, but it is of the greatest practical importance in early Latin versification; for the early dramatic poets, as is now generally recognized, composed primarily by their ear, and according to their general rhythmical feeling, rather than in obedience to a body of precise metrical rules. Their language, it is true, was quantitative, and the weak expiratory accent which it possessed was a wholly insufficient basis for verse, yet this language had not yet fully adapted itself to the somewhat conventional measurements of quantity prescribed by the metrici. Hence, whatever favorite phrase or formula we may adopt, whether we choose to call it "Law of Iambic Shortening," or employ some other name, it is always in reality and in the last analysis the doctrine of the rhythmici which we invoke. This fact has been made fairly plain by W. Christ in his extended article, Die Gesetze der plautinischen Prosodie, Rhein. Mus. XXIII (1868), 559 ff.; and, although many forms of statement employed by Christ in 1868 are now inadmissible, and many of his suppositions are quite untenable in the light of more recent study, yet the rhythmical doctrine to which he appeals still remains highly instructive and substantially correct.

We may then justly claim that the spoken language of Plautus's time possessed syllables which cannot be properly assigned the exact value either of one mora or of two, and without attempting to be over-precise, we may. for the purpose of convenient classification, distinguish the following four classes of syllables: I. 'Heavy' longs (syllabae longis longiores), i.e. those long syllables which (except in the first foot of a hemistich) are rarely shortened in dialogue metres through the agency of the Iambic Law, in other words, those syllables of substantives, verbs, and many adverbs which bear the primary tone; cf. also syllables other than final, which contain a diphthong or naturally long vowel, etc.

it was wholly suppressed in pronunciation, and in this way suffered absolute syncope. According to metrical theory, such an assumption would not be necessary; see below, p. 164 f.

II. Long syllables which are lighter in various degrees than the foregoing, and which are often shortened by the Iambic Law, in other words, unaccented long syllables, including the long penultimate syllables of conjunctions, prepositions, and pronouns. (For the frequent shortening of the latter, see Ahlberg, De corrept. iamb. 66 ff.) III. Short syllables of normal value. IV. Exceptionally short and rapid syllables, which have less than the value of one mora, including those which are diminished in value almost to the point of vanishing (syllabae brevibus breviores). This last class is of special importance for the purposes of the present study, and hence requires more detailed discussion.

Syllables of Diminishing Value. In general, there are very many diminishing syllables in Plautus, some initial, some medial, and some final, which are at least as rapid and as fugitive as the final syllable of nempe. Thus Leo has shown (Forsch. 267 ff.) that the final syllables of enim, nimis, satis, magis, potis are exceptionally weak and very rarely placed under the metrical accent, cf. the monosyllabic forms sat, pot(e), mag(e); still earlier Schrader had shown in his wellknown study (De partic. -ne prosodia, 30 f.) that the final syllable of potin is never placed under the ictus, that of satin very rarely, and so on. Again, we may illustrate this weakness chiefly from medial syllables. The Plautine language possesses the doublets purigo and purgo, iurigo and iurgo, creduas and credas, mavolo and malo, spreverit and sprerit, audiverat and audierat, cognoverit and cognorit, face and fac, siet and sit, laudarier and laudari, dextera and dextra, periculum and periclum.1 We may freely admit that in Plautus these are, for the most part, actual doublets; yet it is evident that,

1 This list includes doublets of two kinds, first, those like iurigo and iurgo, cognoverit and cognorit, in which the shorter forms are actually derived from the longer; and secondly, those like periculum and periclum, creduas and credas, siet and sit, laudarier and laudari, in which the two forms are independent of each other and possess a different origin. Only the first class of cases is strictly pertinent here, but, for convenience, I have enumerated as doublets all the related word-forms which were at the disposal of the poet. For an example of voverat actually written for vårat (Saturnian verse-close), see CIL. I, 541, 7, and cf. Havet, De Sat. 236.

shortly before this time and even during the whole Plautine period, such shortened forms as cognorim or purgo still retained very largely in actual speech the diminishing syllable (e.g. cognoverim), -a syllable having the value perhaps of one half or one fourth of a mora. Moreover, the usage of Terence shows that many of these weak syllables (especially when a long vowel preceded, cf. Spengel on Ad. 304; Schrader, .., 13 ff.) had diminished still further by his time; for Terence is here somewhat more strict than Plautus, and, in general, knows only the forms cognorim, malo, fac, laudari, etc.; in only one position in the verse does he admit the use of the long forms sprēverit, amăverat, face, siet, laudarier, dextera, periculum, viz. in the verse-close,1 and it is evident that in this position he is strongly influenced by metrical convenience or necessity.2 In other words, so great is the demand in the final foot for iambic words and for words ending in an iambus that, in this place, the poet assigns, and the reader expects him to assign, the value of a full mora to syllables which are too short to be often counted at all elsewhere (for additional examples, see below, p. 179). This license is justified, however, by the fact that it rests upon a metrical convention which is thoroughly well known to the reader, and it is possibly also somewhat relieved by the further fact that all quantities, both long and short, tend to be heard more distinctly in the verse-close than elsewhere (cf. in part, Cic. de Or. iii, 50, 192).3

The ancient rhythmical theory of extraordinary short and fugitive syllables applies of course equally well to the verse of other than the classical nations. Thus, in our English accentual poetry we often find the light second vowel sup

1 Cf. Engelbrecht, Stud. Terent. 76 ff.; Stange, De archaismis Terent. 33 f.; Lindsay, Class. Rev. VI, 87 ff.; Brock, .., 75 ff.

2 Cf. Lindsay, Class. Rev. VI, 89: "In other words, Plautus tends to treat the use of the expanded forms (periculum, poculum, etc.) as a license, only to be resorted to in cases of metrical necessity."

3 Cf. Ramain, Études sur les Groupes de Mots, Paris, 1904, p. 202: "Les doubles formes, lexicologiques ou prosodiques, ne sont pas usitées indifféremment: c'est ainsi que evenat, siet, füerit sont reservées exclusivement pour le dernier pied." Cf. also Exon, Hermathena, XIII, 568.

pressed in such dissyllables as never (written also nev'r), seven (sev'n), heaven, power, etc., and in Shakespere also in jewel, being, seeing, playing, cf. also trav(e)lled, rememb(e)red, threat(e)ned (Abbott, Shakespearian Gramm., § 470); conversely, Shakespere sometimes pronounced as dissyllables some of our present monosyllabic words, such as fire (written also fier), dear, fear, hour (ib., § 480). A close parallel to the Terentian usage of měō, amavěrat, periculum is afforded by the termination -tion, which is rarely dissyllabic in the middle of a line, but is frequently so treated at the end of a line (ib., § 479), e.g.:

That shall make answer to such questions.

[blocks in formation]

Compare also the negligible syllables in the following lines from Hamlet:

Ay, thou poor ghost, while mem(o)ry holds a seat, i, 5, 96; And leads the will to desp(e)rate undertakings, ii, 1, 104; Remorseless, treach(e)rous, lech(e)rous, kindless villain, ii, 2, 609; The undiscov(e)red country from whose bourn, iii, 1, 79; And with the incorp(o)ral air do hold discourse, iii, 4, 118.

Synizesis. - Apart from the phenomena of elision, the most perfect illustration of the occurrence of negligible syllables in O. Lat. is afforded by those cases in which two contiguous vowels, belonging to different syllables, are pronounced in immediate succession either in the same word, or in case of elision, in two different words, e.g. meo, meam uxorem.1 Quite similar are the cases where the two vowels belong to differ

1 The second case is of course usually termed synaloepha (so-called elision), but it would be a faulty and pedantic analysis that would admit in the present case any difference of principle between the two examples cited above; see below, pp. 178, 202 f. Besides, the best authorities, both ancient and modern, do not attempt to distinguish sharply between synizesis and synaloepha; cf. Hephaestion, p. 10 W; Scholia A, p. 120 W.; L. Müller, R.M.2 279 ff.

« PreviousContinue »