Friday, April 23, 2010

More on Final -d

On pg. 155 VI b, my treatment of the loss of final -d in Latin is much too brief. I wrote there:

-V:d# > -V: in the third century BCE: abl. sg. -ād, -ōd, SENTENTIAD (ILLRP 511 +), POPLICOD (ILLRP 511) >> pūblicō, 36 MED (ILLRP 1197 +) > .

And fn. 37 thereto adds:

Archaizing spellings with final -d continue to turn up until the last quarter of the second century BCE especially in official texts.

First, a cross-reference to pg. 222 E. 2 should be added where the earliest epigraphical evidence for loss of -d (241 BCE) is mentioned.  The very consistent use of final -d in the SCdB is obviously an archaizing orthography.  As many have noted, whereas the text of the SC consistently uses final -d, the concluding paragraph about the placement of the actual document omits final -d (IN AGRO TEVRANO).

Second, it should be noted that forms with final -d after long vowels in polysyllabic ablatives are also found in the literary transmission. For example, Naevius wrote (Blänsdorf 5): amborum uxores / noctu Troiad exibant capitibus opertis where Vossius was the first to recognize Troiad in the transmitted Troiade. In Plautus, final -d was not possible even as an archaism in polysyllabic nominal ablatives—there is no trace of the Troiad type in Plautus. 

And yet the monosyllabic accusative and ablative mēd and tēd— but interestingly not sēd—do occur. For example, at Cas. 143 (Hic quidem pol certo nil ages sine med arbitro (ia6)) the manuscripts transmit and the meter requires mēd. Similarly tēd, here the accusative, is read and required at Asin. 299: Le. quot pondo ted esse censes nudum? Li. non edepol scio (tr7). The first line of the Curculio, Quo ted hoc noctis dicam proficisci foras was quoted by Charisius (Barwick 1964²:143.24) and Diomedes (Keil 1.441.18) precisely on account of its d-final form. The forms mēd and tēd do not occur in Terence and the only generally accepted non-Plautine literary mēd seems to be Enn. frg. var. Epicharm. 45 nam videbar somniare med (ms. me et) ego esse mortuum, but others suggested memet with excision of ego.

On the other hand, Plautus also clearly had vowel final allomorphs of the 1st and 2nd sg. pronouns since these form undergo elision, e.g. Pseud. 375 si id non adfert, posse opinor facere me officium meum (tr 7) and Asin. 44 Dono te ob istuc dictum, ut expers sis metu (ia6). This suggests that the loss of d was sensitive to some prosodic factor, presumably the presence or absence of stress.  Since d was always lost in polysyllabic ablatives which never bore the stress on the immediately preceding vowel it must have been after the unstressed variant that -d was lost.

In fact, it is not really clear that vowel length played a significant role in the loss of -d. A recent article by Martin Kümmel has reopened this question. He notes that there are many epigraphical cases of the omission of final -d in perfect 3rd sg. (IOVSI, ILLRP 129, 3rd cent. BCE, Lacus Albanus, etc.) and 3rd pl. forms and argues that the true condition for loss of final -d was after unstressed vowels. Forms like illud, istud, aliud can be explained as analogical to potentially tonic id and quid. The forms sed, ad, apud, and haud might all be explained as proclitics and hence not true instance of word final -d. This account does seem the best way to handle the dental-less 3rd sg. forms.  For the plural forms in -e:ro(n) it is still possible in my opinion that d was not lost but assimilated and then simplified.

See Kümmel, Martin J. 2007. The third person endings of the Old Latin perfect and the fate of final -d in Latin. In K. Jones-Bley et al. eds., Proceedings of the 18th Annual UCLA Indo-European Conference. Washington D.C.: Institute for the Study of Man, 89–100.

No comments:

Post a Comment