A new look at an old problem: on the origin of the pronominal augment -ne.
At the beginning of the last century, the origin of the extension -ne was discussed in several brief notes, mainly in connection with the Italo-Romance forms, on the pages of the Zeitschrift fur Romanische Philologie, and echoes of that discussion have found their way into the major historical grammars of individual Romance languages and Romance as a whole. In general, those who discussed these pronominal forms in the individual languages, especially Romanian and Italo-Romance, tended to favor their polygenesis, while those with a broader Romance perspective tended to view them as related. But even while agreeing on the overall status of these forms, each category of experts had independent views about the etymological source of the augment, which, in the words of Allieres, remains "un probleme ardu" (64). (3) This note offers a new perspective on the origin of this particle, which does not seem to be found in the earlier literature on the problem. Before proceeding, I briefly survey the ideas on the origin of -ne which were put into circulation at the time when this augment formed the topic of a lively scholarly exchange.
The discussion of the forms with -ne was primarily concerned with the Italo-Romance data and took different directions depending on whether the pronominal augment was considered separately from or together with the paragogic -ne ~ -ni of other lexical classes. The former approach was adopted by Subak and Bourciez who, however, proposed different sources for the particle itself. Subak derived mene, tene and sene from stressed forms ultimately going back to the emphatic conglomeration mem[et ipsum] (cf. also Bartoli, Italienische, 112). Bourciez saw the origin of the pronominal -ne in the Latin interrogative particle -ne, deriving the relevant pronouns from their use in questions and exclamations. His examples, cited in support of this view, are Virgil's mene incepto desistere victam and Cicero's tene haec posse ferre (238).
Those who connected mene, tene and sene with the verbs and other lexical classes containing the augment syllable -ne, saw the connection in the fact that all the relevant forms were vowel-final oxytones. The specific forms discussed in this context included verbs like fane 'fa', vane 'va', puone 'puo' and saline 'sali', found already in Dante (Parodi, 116); and central and southern Italo-Romance verbs, adverbs, nouns, numerals and non-personal pronouns like piune 'piu', dine 'di', none 'non', quine 'qui', purcene 'perche', piene 'pie', trene 'tre', cosine 'cosi' and pensone 'penso' (more examples are cited in Rohlfs, Grammatica, 468-469). As previously, the origin of the particle and the starting point for its extension received different treatments. Thus, Pieri viewed its pronominal use as original and derived it from the presence of -ne, derived from inde, in such combinations as mene vado. According to Pieri, the particle -ne could later have become generalized as a mere euphonic addition to vowel-final oxytones (340). A different school of thought originated with Meyer-Lubke, who saw the use of -ne with personal pronouns as just another manifestation of its use with oxytonic words in general. Meyer-Lubke derived -ne from the effects of phrase-level phonetics on the negative non, with none being the stressed and no the proclitic allomorph. According to Meyer-Lubke, a paradigmatic relation could have been established between the stressed none and unstressed no; from there, the syllable -ne could have been extracted and pressed into service as a marker of the stressed allomorphs of other vowel-final oxytones, including mene and tene in opposition to me and te (Italienische, 172-173). An alternative theory was suggested by Grandgent, who started off with the Italian verb forms tiene and viene and their truncated variants tie and vie. These appear to have been modeled on the die for diede which, in its turn, was derived on the model of pie for piede and that of the abstract nouns in -de. (4) According to Grandgent, the syllable -ne was extracted from the pairs tie/tiene and vie/viene and extended, first, to other verb forms (whence fane for fa and other similar creations in Dante and the modern dialects), and later to pronouns and adverbs (Grandgent 47, 135). D'Ovidio and Meyer-Lubke offer essentially the same solution by identifying the pair sono/so as the point of departure (167). Rohlfs does not take a firm stand on the origin of -ne, although he is inclined to favor its derivation from the juxtaposition of ve/vie, te/tie and be, respectively, with vene/viene, tene/tiene and bene in central and southern Italy (Grammatica, 469). Lausberg (vol. 2, 161), writing from the pan-Romance perspective, derives the augmented pronouns from the bases *mene and *tene, which according to him were built analogically on the pattern of *quene < quem. His explanation of this development is pragmatically and discourse-based, and is adopted for Romanian by Rosetti (146), and both Romanian and Sardinian by Wagner (113).
All of the above derivations assume a Latin or at least a Romance source for -ne, whether seen as an originally independent morpheme or an analogically created one. A competing proposal relates it to the Balkan areal phenomena, deriving the Romanian particle and its equivalents in Albanian and New Greek from a substratum source (Rosetti et al., 326; Du Nay, 48, 55; Allieres, 64). The following table from Rosetti et al. (326) demonstrates the parallelism among the Balkan forms. The authors also suggest a substratum or adstratum source for the augmented forms in south Italy, citing Messapic and Greek as the possible source languages.
Free form Clitic Romanian mine ma Albanian mua (< mene) me New Greek (e)mena me
It is interesting to note that despite the breadth, variety and phonetic/semantic/pragmatic plausibility of the proposals put forth in explanation of the etymology of -ne--including Bourciez' derivation from the Latin interrogative particle -ne, Subak's from the emphatic m[et ipsum], Pieri's from the adverbial inde, Lausberg's from quem, and others--the dialectal distribution of the augment remains a mystery. The only explanation to take the geographical factor into consideration, that of Rosetti et al., has the disadvantage of divorcing from one another the facts of the individual Romance languages. Therefore, if the hypothesis that the Romance pronominal forms with -ne are related is to be salvaged, we must look for a different source for this particle, one that would both work phonetically and semantically and explain the augment's dialectal range. From this point of view, it seems that a substratum language or languages from the requisite regions of Italy might provide just such a source. Provided that the phonetic and semantic sides of the development can be shown to be accounted for satisfactorily, such a source would be able to explain the dialectal distribution of -ne both inside and outside Italy.
There appear to be two etymological possibilities for deriving the particle -ne from languages spoken in the center and south of ancient Italy. The first possibility is the accusative of the first person singular pronoun mi in Etruscan, variously spelled as mini, mine, min, mene, men and, once, mi (the last variant explained as an instance of haplography by Rix, La scrittura, 229). It is conceivable that in the process of language shift to Latin, in the wake of Roman colonization, the original nominative vs. accusative contrast between mi and mini could have been reanalyzed as that between the shorter and longer, and subsequently the unstressed and stressed oblique forms. The formal similarity between the pair mi/mini and the oblique forms of the Latin first person pronoun would have favored such a reanalysis at a time when the local population was switching to the suppletive pattern of Latin ego versus the oblique m-initial forms. After the reanalysis would have been complete, the final syllable of mini could have been extended to other vowel-final oxytones as a marker of emphasis, essentially following the scenarios proposed by MeyerLubke, Pieri, Grandgent, and others, for alternative starting forms.
The second etymological possibility is to derive -ne from one of the Sabellian languages, which covered the southern regions of ancient Italy and the central regions not covered by Etruscan. In Sabellian, the attested accusatives of singular personal pronouns include the Volscian miom, Umbrian and South Picene tiom and Oscan siom (Rix, La lingua, 231, n. 8). Among the attested nominative forms, the Oscan tiium deserves a mention (Buck, 139). The concurrently attested Umbrian accusative singular forms tiu, tio, teio seem to indicate that the final -m might originally have been a particle, possibly related to the one that occurs in the Sanskrit nominative singular tvam and dative singular tubhyam. (5) If not the forms themselves, at least the final particle could have been transferred to the similar-sounding accusative singular of the personal pronouns in local Latin (> *me-m, *te-m, *se-m), which would have produced -ne in the relevant Romance dialects by a regular phonetic development (*mem > mene, etc., just as quem > Romanian cine and spem > Italian spene) (Meyer-Lubke, Grammatica, 115; Lausberg, vol. 1, 425). The original impetus for the transfer could have been reinforcement of the relevant accusative pronouns in Latin; later on, the pairs of allomorphs me/mene, te/tene and se/sene could have been reanalyzed as unstressed versus stressed, etc., following the path of development already outlined in the preceding paragraphs.
To recapitulate, it is proposed that the longer pronominal form(s) was/were borrowed into Latin from a substrate source or replicated language-internally under substrate influence. At first, it/they existed in free variation with the shorter native form(s); later, each allomorph was reallocated to a specific function and syntactic environment, producing the coherent systems of present-day languages. While using a different starting point, the proposed development naturally relies on the mechanisms of analogical extension already identified by earlier researchers. In addition, the proposed scenario works not only sociolinguistically, by relying on the common processes of bilingual transfer and functional reanalysis, but also phonetically and semantically; in fact, derivation of a marker of oblique pronouns from the accusative pronouns of one or more substrate languages is able to explain the current semantics of the augmented forms somewhat better than the alternative proposals. (6) The final advantage of the present etymology is to explain the dialectal distribution of the augmented forms. The south of Italy has traditionally been an area of emigration due to overpopulation, and because of this historical fact Romance scholars have long speculated that many of the shared features of southern Italo-Romance, Sardinian, Dalmatian and Romanian may be due to the shared south Italian origin of the colonists (Terracini; Silva Neto Fontes and Historia; Meyer-Lubke, Grammatica, 4; Wartburg, 26ff; Tagliavini, 91-158; Du Nay, 41ff; Rohlfs, Estudios). The substratum influence of Etruscan and especially Sabellian on southern Italo-Romance has been amply discussed in the literature and includes not only lexical and phonological traits but also morphological patterns; to mention one salient example, one of the accepted explanations for the southern Italian tt-Perfect (stetti, detti etc.) is to derive it from a similar formation in Oscan (Pisani, Il sostrato, 159-160 and Le lingue, 65; Buck). Therefore, it is both plausible and likely that the pronominal augment -ne represents yet another instance of the preservation in the Romance languages of a morphological feature of Italic and/or Etruscan origin.
CALIFORNIA STATE UNIVERSITY, FULLERTON
Allieres, Jacques. Manuel de linguistique romane. Paris: Honore Champion, 2001.
Bartoli, Matteo Giulio. Das Dalmatische. 2 vols. Vienna: Alfred Holder, 1906.
--. "Italienische Sprache." Kritischer Jahresbericht uber die Fortschritte der Ro manischen Philologie 10 (1906): 110-13. Blasco Ferrer, Eduardo. La lingua sarda contemporanea. Cagliari: Edizioni Della Torre, 1986.
Bourciez, Edouard. Elements de linguistique romane. Paris: Klincksieck, 1967.
Brancug, Grigore, Adriana lonescu and Manuela Saramandu. Limba romana. 4th ed. Bucharest: Editura Universitacii din Bucuregti, 1999.
Buck, Carl Darling. A Grammar of Oscan and Umbrian. Hildesheim/Zurich/New York: Georg Olms Verlag, 1995 [1928J.
D'Ovidio, Francesco and Wilhelm Meyer-Lubke. Grammatica storica della lingua e dei dialetti italiani. Trans. Eugenio Polcari. Milan: Ulrico Hoepli. 1919.
Du Nay, Andre. The Early History of the Rumanian Language. Lake Bluff, Illinois: Jupiter Press, 1977.
Grandgent, Charles H. From Latin to Italian. Cambridge: Harvard UP, 1927.
Jones, Michael A. "Sardinia." The Dialects of Italy. Ed. Martin Maiden and Mair Parry. London/New York: Routledge, 1997. 376-384.
Kemmer, Suzanne. The Middle Voice. Amsterdam/Philadelphia: John Benjamins, 1993.
Lausberg, Heinrich. Linguistica romanica. 2 vols. Madrid: Gredos, 1966.
Meyer-Lubke, Wilhelm. Grammatica storica della lingua italiana e dei dialetti toscani. Turin: Loescher Editore, 1955.
--. Italienische Grammatik. Leipzig, 1890.
Parodi, E.G. "La rima e i vocaboli in rima nella Divina Commedia." Bullettino della Societa Dantesca Italiana 3 (1896): 80-156.
Pieri, Silvio. "L'it. -ne." Zeitschrift fur Romanische Philologie 30 (1906): 339-340.
Pisani, Vittore. "Le lingue preromane d'Italia: origini e fortune." Popoli e civilta dell'Italia antica. Vol. 6. Rome: Biblioteca di Storia Patria, 1978. 56-77.
--. "Il sostrato oscoumbro." I dialetti dell'Italia mediana con particolare riguardo alla regione umbra. Perugia: Facolta di Lettere e Filosofia dell'Universita degli Studi di Perugia, 1970. 149-169.
Poultney, James Wilson. The Bronze Tables of Iguvium. Oxford: Blackwell, 1959.
Rix, Helmut. "La lingua del Volsci. Testi e parentela." Kleine Schriften. Bremen: Hempen Verlag, 2001. 229-241.
--. "La scrittura e la lingua." Gli etruschi: una nuova immagine. Florence: Giunti Martello, 1984. 210-238.
Rohlfs, Gerhard. Grammatica storica della lingua italiana e dei suoi dialetti. Vol. 1. Fonetica. Trans. Salvatore Persichino. Turin: Giulio Einaudi, 1966.
--. Estudios sobre el lexico romanico. Trans. Manuel Alvar. Madrid: Gredos, 1979.
Rohrsheim, Ludwig. "Die Sprache des Fra Guittone con Arezzo." Beihefte zur Zeitschrift fur Romanische Philologie 15 (1908).
Rosetti, Alexandru. Istoria limbii romane de la origini pina in secolul al XVII-lea. 2nd ed. Bucharest: Editura Ctiinjifica [section]i Enciclopedica, 1978.
--. B. Cazacu and I. Coteanu. Eds. Istoria limbii romane. Vol. 2. Bucharest: Editura Academiei Republicii Socialiste Romania, 1969.
Silva Neto, Serafim da. Fontes do latim vulgar (o Appendix Probi). Rio de Janeiro: Imprensa Nacional, 1946.
--. Historia da lingua portuguesa. 2nd ed. Rio de Janeiro: Livros de Portugal, 1970.
Subak, J. "Epitese." Zeitschrift fur Romanische Philologie 30 (1906): 581-583.
Tagliavini, Carlo. Le origini delle lingue neolatine: Introduzione alla filologia romanza. Bologna: Riccardo Patron, 1969.
Terracini, B.A. "Sostrato." Scritti in onore di Alfredo Trombetti. Milan: Ulrico Hoepli, 1938. 321-64.
Vignoli, C. "Il vernacolo di Castro dei Volsci." Studj Romanzi 7 (1911): 117-296.
Wagner, Max Leopold. "Flessione nominale e verbale del sardo antico e moderno." Italia Dialettale 14 (1938): 93-170.
Wartburg, Walther von. La fragmentation linguistique de la Romania. Paris: Klincksieck, 1967.
(1) I thank Edward Tuttle and the anonymous reviewer for their valuable comments.
(2) The distinction between stressed and unstressed personal pronouns is panRomance, and consists in using formally distinct variants of the pronoun in complementary functions and syntactic positions. In the case of oblique pronouns, the unstressed forms are used in constructions with verbs, while the stressed forms are used with prepositions and to convey emphasis (cf. overview in Kemmer 1993:163ff). The syntactic and semantic differences between the two sets of forms may be illustrated with the following examples from Romanian, in which te is the unstressed and tine the stressed allomorph of the accusative case of the second person singular pronoun. It will be observed that te is used as the ordinary object of the verb, including pleonastically in the second example, whereas tine indicates contrast with another referent.
Te caut. 'I am looking for you.'
Te cautpe tine, nupe ea. 'I am looking for you, not for her.'
(3) Cf. also Rohlfs (Grammatica, 469): "La provenienza di questa sillaba paragogica non e ancora definitivamente chiarita".
(4) Grandgent explains pie for piede and the abstract nouns without the final syllable -de as resulting from such haplological combinations as alli piede de Deo, la mercede de Deo and la fede de Deo; whence pie, merce and fe, respectively (47).
(5) For alternative explanations of the final -m, cf. Buck (139-140) and Poultney (108).
(6) Thus, Rosetti et al. object, on both semantic and geographical grounds, to Bourciez's proposal to derive -ne from the homophonous Latin interrogative particle (325-326).
|Printer friendly Cite/link Email Feedback|
|Date:||May 1, 2012|
|Previous Article:||The failure of language in Ernesto Sabato's El tunel.|
|Next Article:||The (w)hole in the text: a hermeneutic reading of Marguerite Duras's la pluie d'ete.|