North Picene reloaded

From: dgkilday57
Message: 70659
Date: 2013-01-09

The angle brackets which I have been using for years in place of italics are being dropped along with their contents. I will now see whether direct posting solves the problem.

Conventional wisdom holds that North Picene is an unclassified language whose corpus consists of the Novilara stele and three short fragments from the vicinity of Pesaro. My opinion is that we also have an important North Picene gloss on the stone of Cafatius, commonly known as the bilingual of Pesaro (CII 69 = CIL 1{2}:2127 = CIL 11:6363 = PID 346 = NRIE 1150 = TLE 697 = ET Um 1.7).

This stone was discovered in the 16th century within the walls of ancient Pisaurum. It contains a Latin epitaph in monumental characters typical of Caesar's time, followed by an inscription in smaller, simpler Etruscan script. The Latin text, with damaged parts restored, reads as follows:

[L.CA]FATIVS.L.F.STE.HARVSPE[X] / FVLGVRIATOR

'Lucius(?) Cafatius, son of Lucius, of the tribe Stellatina, haruspex (i.e. diviner from entrails), fulguriator (i.e. diviner from lightning).'

The Etruscan text reads as follows:

[c]afates.lr.lr.nets'vis.trutnvt.frontac.

The first part is easily understood as 'Cafates Laris Larisal [clan]', i.e. 'Laris Cafate, son of Laris', in abbreviated form, with South Etruscan conventions for sibilants and order of names. Scholars have found it difficult to equate the last three words with the Latin text. Some have supposed that <frontac> indicates origin, like <rumach> 'of Rome, Roman', <velznach> 'of Volsinii, Volsinian', and might mean 'of Ferent(i)um or Ferentinum, Ferentine'. But such a designation would be equivalent to a cognomen, and should follow the personal name, as <Ste(llatinas)> does in the Latin text. Moreover the letter <o> is not used in ordinary Etruscan. To be sure, the Lemnos stele employs <o> throughout in place of <u>, but <trutnvt> shows that no such convention was in use here.

The word <nets'vis> is identical to North Etr. <netsvis'> from the ossuary of Poggio al Moro (CIE 978 = TLE 524):

nae.cicu / pethnal / netsvis'

'Nae Cicu, (son) of Pethnei (mother's gentilicium), nets^vis.'

This word presumably corresponds to Lat. <haruspex> 'entrail-looker'. Various scholars have commented on the similarity to Greek <né:duia> nt. pl. 'entrails', <ne:dús> f. 'belly' without providing a plausible explanation. I believe one can be given, since <ne:dús> is characteristic of Epic and later poetic Greek, like <ple:thús> f. 'throng, crowd' beside the usual prosaic <plêthos> nt. 'id.'. This presupposes an obsolete prosaic /es/os/-stem *nêdos. Now in the compound <pho:s-phóros> 'light-bringing; morning star' the combining stem has the same surface form as the nom./acc. sg., <phôs> from *phawes-, *pháwos. (In classical Greek the combining stem and the oblique cases of <phôs> acquired a -t- which apparently originated in the neuter /n/-stems; cf. A.L. Sihler, New Comp. Gr. of Grk. & Lat. §§ 288-93, 1995.) From *nêdos we can imagine a similar archaic Western Greek compound root-noun *ne:des-wid-, nom. sg. *ne:deswi:s 'belly-looker, entrail-examiner' (cf. Attic <eîdon> aor. 'I saw' from *h1e-wid-o-m). Borrowed into archaic Etruscan, this would yield *ne(:)tes(^)vi(:)s, and with regular syncope of a short middle vowel, recent Etruscan <nets^vis> as found. The same Greek compound was likely calqued to yield Lat. <extispex>.

Late Etr. <trutnvt> is evidently the same word as <trutnuth> from the cippus of Arnth Aprie of Tarquinii (CIE 5487 = TLE 118):

apries.ar.vth / trutnuth

'Arnth Aprie, (son) of Velthur, trutnuth.'

The order (typical recent South Etruscan) is identical to that of our Laris Cafate text, while the latter's spelling <trutnvt>
shows loss of word-final aspiration and vocalic <v> (apparently to distinguish close [u] from open [U]) as late features observed elsewhere (e.g. <Arnt>, <Lart>, <Pvrni>, <Apvcuia>). This word presumably corresponds to Lat. <fulguriator>, and is probably a native compound.

J. Whatmough drew a possible connection between the endings of <frontac> and <-tenag>, which occurs twice on the North Picene stele of Novilara (The Prae-Italic Dialects of Italy 2:557-8, 1933). He also drew one between <trutnvt> and five words on the stele ending in <-út>, but this should be rejected because the evidence points to <trutnvt> being a native Etruscan word. Only <frontac> requires a non-Etruscan explanation. Indeed if it cannot correspond to <Ste(llatinas)>, it creates serious problems for the interpretation of the Cafatius stone as a mere bilingual, since three terms must somehow be equated with two.

Some scholars, including M. Pallottino in his youth (Elementi di lingua etrusca 99, 1936), took <nets'vis trutnvt> as a unit corresponding to <haruspex>, thus equating <frontac> with <fulguriator>. V. Pisani went so far as to regard <nets'vis> as the genitive (with metathesis from *netus-is) of a Mediterranean congener to Grk. <ne:dús>, and <trutnvt> as the mere equivalent of Lat. <-spex> (Le lingue dell'Italia antica oltre il latino 218ff., 1953). Obviously this viewpoint clashes with the attestation of Etr. <netsvis'> and <trutnuth> as self-standing words. S. Ferri took this evidence into account and equated <nets'vis> with <haruspex>, and <trutnvt> with <fulguriator>, which I believe to be correct (Osservazioni alla "bilingue" di Pesaro, Rend. Acc. Lincei (ser. 8) 13:323ff., 1958). But he then explained <frontac> as formed like Etr. <rumach> 'Roman', and based on the town of Ferentium (Tac. Hist. 2:50; now Fèrento near Viterbo), whose inhabitants were counted among the tribus Stellatina (CIL 11:3008). While ingenious, this suffers from major problems. It is true that <trutnvt> shows late word-final deaspiration (or loss of the distinction in that position), so earlier Etr. -ach would be expected as -ac. But <frontac> still contains that glaring <o>, and moreover the accent of <Fèrento> demands word-initial ancient accent, <Férent(i)um>. (This disobeys the classical penultimate rule, as do some other place-names not in Rome's immediate vicinity, e.g. <Fálerii:> as demanded by (S. Maria de) <Fàlleri>.) This accent makes it extremely implausible, even in theory, that the first vowel of <Ferent(i)um> would suffer syncope going into Etruscan. In practice, names borrowed into Etruscan underwent weakening and syncope of vowels in medial, not initial, syllables, regardless of the original accent. This is well illustrated by Greek <Klutaimn:éstre:> appearing in Etruscan as <Cluthumustha>, later <Clutmsta>, and <Aléxandros> as <Alechsantre> etc., <Elachsantre>, <Elachs'ntre>, <Alcsentre>, <Elsntre>, <Elcste>. Another difficulty is Ferri's equation of the Roman political concept of tribal membership with the ordinary notion of provenience from a particular town.

G. Alessio rightly rejected Ferri's explanation of <frontac> (Mediterranei ed Italici nell'Italia centrale, Studi Etruschi 29:191-217, esp. 205ff., 1961). Unfortunately he was so enamored with his chain of etymological reasoning, equating <frontac> with <fulguriator>, that he took a backward step and insisted that the two words <nets'vis trutnvt> must correspond to <haruspex>, not in a syntagm as Pisani assumed, but with the second word delimiting the first. As a Latin parallel he cited CIL 11:5824:

L. VENTVRIVS RVFIO [A]RISPEX EXTISPICVS

'Lucius Venturius Rufio, extispician haruspex.'

One might ask why Rufio was not remembered simply as an extispex. Apparently he had been trained as a haruspex, an interpreter of divine signs, but most of his practice involved examining the entrails of sacrificial animals. (I find the native derivation of <haruspex> from PIE *g^Her- 'gut, entrail' formally difficult, and I prefer to see the Latin word as a half-calque of Western Greek *hiaróskopos or *hiarósko:ps 'sacred-sign-seer'.) Rufio's career probably reflected a trend during the late Republic in which <haruspex> became practically synonymous with <extispex>. Alessio argued that since an extispex is a specialized type of haruspex (extispicus), a <trutnvt> could equally well be a specialized <nets'vis>, and he provided unconvincing reasons (based on wild etymological speculation which I will not repeat) that a <trutnvt> dealt specifically with examining livers. The obvious problem is that if <nets'vis> requires qualification on the Etruscan side, then so does <haruspex> on the Latin side, and we find no *jecinispicus or the like there.

As a result I see no other option than to understand <frontac> as an Etruscan scribe's rendering of the North Picene word for a soothsayer whose practice involved divination from both entrails and lightning. This was probably very close to the original <haruspex>, but as Rufio's epitaph suggests, the role of <extispex> had come to take up most of the duties of the traditional <haruspex>, and the <fulguriator> was now considered a separate specialist in both Roman and Etruscan society. Thus the Pesaro text is not bilingual, but trilingual. Alessio deserves credit for recognizing this, since he saw that <frontac> cannot be a native Etruscan word, but his conclusion that the non-Latin text is in a "mixed language" cannot be taken seriously.

As for the pronunciation of <frontac>, the Etr. <f> probably represents a voiced fricative [B] or [v] in North Picene, which natives would have written <b>. Had this been a voiced stop, one would expect Etr. <p> (as in the cognomen Palpe for Lat. Balbus, etc.). The <o> must represent a vowel-timbre completely foreign to contemporary Etruscan; otherwise <u> would have been used. The <c> presumably represents a voiced sound, parallel to <-tenag> in the native Novilara text. Whether this was a stop or a fricative in word-final position, I cannot say. At any rate I believe natives would have written this word as <brontag>. Details of its etymology, along with my views on North Picene phonology, morphology, and classification, will follow some other week.

DGK