This licence is known to every reader of Lucretius, and its extension may be shown by a line from Ennius' Annales (601 Vahlen) tum lateralis dolor, certissimus nuntius mortis. Corssen (1 286. 599) gives lengthy lists of names from inscriptions, many of which are as old as the Punic wars, and in which a final s is entirely omitted, and the same fact occurs again in inscriptions of the decline of the Roman empire'. It would, therefore, be very surprising if no traces of it were found in the prosody of the comic writers. As instances of it will be frequently met with in Plautus, we shall confine ourselves to some examples from the Aulularia. Thus we should pronounce We shall now easily understand such endings of iambics as the following, all of which are taken from Terence's Hecyra: aúctus sit 334. deféssus sum 443. incértús sum 450. expertus sum 489. núllus sum 653. úsus sit 878. Comp. occidistis me Bacch. 313. r was, in many instances, merely a substitute for an earlier s, and we should therefore be prepared to find that by comparing two lines of Ennius and Virgil. Aen. xII 115 we read Solis equi lucemque elatis naribus efflant, on which words Servius has the following note 'Ennianus versus est ordine commutato: funduntque elatis naribu' lucem.' (See Ennius ed. Vahlen p. 85 and the pas- once. 1 See Schuchardt, on Vulgar Latin, II p. 445 p. xxxii p.xxxiii occasionally a final r is dropped. Thus we should pronounce soro (= soror) in a line from the Poenulus (1 2,84) Sátis nunc lepide ornátam credo, sóror, te tibi vidérier. and in two short anapaestic lines from the Stichus (18 and 20): haec rés vitae | me, sóror, | saturant ne lácruma, sóror, | neu túo id | animo Another line in the same play furnishes a fourth example cf the same pronunciation (v. 68): quíd agimús, soror, si óffirmabit páter advorsum nós :: patiIn Terence we have the same, Eun. I 2, 77 soror dictast cupio abdúcere ut reddám suis. This is the reading of the Bembine ms., and the prosody of -soro dict- is rightly explained by Faërnus in his note on the passage1. The word color should be pronounced colo in the following line : color vérus, corpus sólidum et suci plénum. and amor loses its final r in Ter. Andr. 1 5, 26 amor, mísericordia huíus, nuptiárum sollicitátio. pater follows the same analogy, e.g. né tibi aegritúdinem, pater, párerem, parsi sédulo. 1 Liber Bembinus quocum hic consentiunt omnes fere libri recentes nec versus repugnat, si abicias r ex soror, ut primus pes sit anapaestus. FAERNVS. -If we adopt Corssen's views (krit. beitr. zur lat. formenlehre, p. 399 s.), we should have to acknowledge the possible dropping of a final r only in those words where it had supplanted Trin. 316. an original s. The sole exception to this law would be pater, and this instance has been neglected by Corssen.-Comp. also the Italian suora frate moglie. Schuchardt I p. 35 shows that the popular pronunciation dolo instead of dolor gave rise to a confusion between dolor and dolus in the later stages of Latin. quid ego agam? pater iam híc me offendet míserum adve niens ébrium. Most. 378 (according to the mss.) pater vénit. sed quid pértimui autem, bélua, = Ter. Phorm. 601. In these cases Prof. Key adopts a monosyllabic pronunciation Fr. père. The possibility of such a pronunciation is questioned by Ritschl (Proll. Trin. CLV) whose words are as follows: 'In quibus (i.e. monosyllabis) si etiam pater habitum est, eius rei et rationem et documenta desidero. et omnium minime ex eo argumentandum esse quod, ut e soror monosyllabum soeur, ita e pater similiter factum esse père dicunt, vel hinc intelligitur quod, etsi frère quoque et mère e frater et mater contracta sunt, tamen haec latina nec contendit quisquam nec poterit contendere unquam monosyllaba fuisse.' This is, indeed, the best argument' which can be alleged against Prof. Key's way of pronouncing and contracting Latin words according to the analogy of the corresponding French forms; but has it been understood and appreciated by Mr Parry? This 1 I have left this passage exactly as it was written ten years ago. In his work on 'Language,' p. 133, Mr Key alludes to the above as follows-' In parricida for patricida we see already that change which led to the Fr. père from pater; and here again when pater appears in Latin comedy, as it sometimes does, to need a shortened pronunciation, it seems simpler to drop the t than to drop the r, as Dr Wagner proposes. Of course māter and frater, with their long penults, were better able for a time to resist such compression, so that Ritschl's contention has I think little weight.' This is a remarkable instance of perverse argumentation. Mr Key apparently assumes a form somewhat resembling paer, to be W. P. pronounced like the Fr. père. But père is not, as he thinks, descended from pater, but from patrem, comp. the Italian padre, and see e. g. Brachet, Dict. étym. de la langue franç. p. 404. It is evident that père patre (m) cannot represent pater. But what weight shall we attribute to the assertion of a modern writer of the 19th century that he considers this or that pronunciation to be 'simpler,' when this is quite contrary to the very evidence of the inscriptions and earliest mss.? See the instances collected by Schuchardt, On Vulgar Latin II 390 sq., where both pate and soro are quoted from ancient testimony. See also Corssen II 656. 3 scholar accuses Ritschl of 'losing sight of the difference in quantity' between frater mater and pater1. But Ritschl's argument is entirely based on this very same difference. He means that, if we once begin to remodel the old pronunciation of Latin upon that of the French of the nineteenth century, we must be prepared to find a contracted pronunciation of mater and frater just as well as of pater, all these words being treated alike in French as mère frère père. But we never meet in Plautus and Terence with mater or frater as monosyllables, on account of their different quantity, and this fact proves that, as we cannot draw a correct inference from mère and frère as to mater and frater, we cannot p. xxxiv. consequently rely upon the comparison of père and pater. And indeed in Plautus or other poets, we never find mate and frate = mater and frater, though in a Faliscan inscription we actually read MATEHECUPA, i.e. mater hic cubat. But this is of course a low dialectic corruption3. The same theory accounts for the loss of a final t and d. An old form hau (instead of haud) owes its existence to this process (see note on v. 170): it remained in use until the time of Tacitus, if we may trust the authority of the Medicean ms. In the Aulularia we have apu (= aput or apud)* in several instances (v. 83. 340. 736.), in the same we should pronounce = decet dece 136. (See M. Crain, Plaut. Stud. p. 10)". 1 Parry's Introd. to Terence, XLVI. 2 See Ritschl, Corp. Inser. Lat. I 89, or Rh. Mus. xvi 603. 3 We need not add how dangerous, nay how fallacious, it is to draw inferences from French with regard to the pronunciation of Latin. I do not hesitate to accede to Ritschl's assertion' omnino tam esse lubricum hoc genus comparationis arbitror, nihil ut inde proficias.' (Proll., 1. 1.) 4 For ape apud see Schuchardt, On Vulgar Latin 1 p. 123. 5 See also the instances given by Corssen 11 650. Thus we find dedit written as dede in three very inscriptions, C. I. L. 1 62. 169. 180. old The preposition ad is thus often degraded to a simple ă, e. g. séd ăd postrémum. Poen. iv 2, 22. quís ad forés est? Amph. Iv 2, 1. et ăd pórtitóres. Phorm. 1 2, 100. ut ad paúca rédeam. Phorm. Iv 3, 43. But it would be superfluous to accumulate more instances of this fact we shall only add that even nt was entirely dispensed with in the rapid pronunciation of the time of Plautus. Bentley has quoted in his Schediasma (p. xv ed. Lips.) the following instances: student fácere studě fácere. = habent déspicatu = habě déspicatu. To these we might easily add other instances from Plautus, but to prove the existence of such forms as we assume here in the metres of the comic poets, we mention the p. xxxv. form dedro, which in an inscription from Pesaro (C. I. L. I 177) stands as an equivalent to dederunt. This form is an unmistakable precursor of the corresponding Italian form diedero1. But precisely the same kind of form as is assumed exists in emeru = emerunt C. I. L. 11148, in an inscription later than the second Punic war, but earlier than the Lex Julia de civitate sociis danda. This emeru forms the stepping-stone from emerunt to the secondary form emere. The final letters m srtd are more frequently dropped than two others which we have yet to mention. The first is l, which is sometimes cut off in the word semol (simul), e. g. Aul. 617. Mil. gl. 1137. Ter. Eun. 11 2, 10. Haut. tim. Iv 5, 55: the second n, which is dropped in the word 1 See Corssen, I 186 sqq., where further materials are produced from the Inscriptions. 2 These instances are taken from Corssen 111, 96 (11 643). |