Project

Romance languages; language change; Latin; Italian dialects

Lead
Lay summary
The evolution of Latin GN is a major issue of Romance and Italian historical phonetics. In this project the relationships between the different outcomes of GN will be considered, starting from the pronunciation of GN in Latin and following its evolution in Mediaeval Latin and Early Romance until today. Four different tendencies can be observed in the evolution of GN, i.e. palatalization (ñ in Italian and all Western Romance languages, -in- in Engadinese and in most Southern Italian dialects), velarization (-un- in a few dialects of Southern Basilicata, Northern Calabria and Puglia), labialization (-mn- in Rumanian), and assimilation (-nn- in Sardinian). These tendencies have been variously explained, among others, by Lausberg, who postulates an Early Romance pronunciation of GN resulting either in spirantization (GN > -*yn-) or palatalization (GN > -in-), and by Merlo, who supposes that in Southern Italy an epenthetic vowel would have broken up the GN cluster already in Latin, thus giving GiN > -in- and GuN > -un-. However, both explanations do not take into account the fact that, according to Indo-European linguistics, the first segment of the GN cluster was not pronounced as a voiced velar stop, but as a nasal velar [n]. The aim of the project is to test the different hypotheses that have been given by following the evolution of the GN cluster from Classical Latin to the contemporary Romance languages and dialects. By collecting a large number of data from old and contemporary dialects and analysing them by means of experimental phonetics and philology, a possibly comprehensive explanation of the different outcomes of GN will be proposed. The project intends to contribute not only to Romance dialectology, but also to Indo-European linguistics and generally to historical phonetics. Moreover, by analysing the distribution of the different outcomes of GN in the Romania and in the Latin borrowings of non-Romance languages, it will help to reconstruct the linguistic history of Europe from antiquity until today. Finally, it will test the most recent theories on language change by applying them to a highly problematic issue, which apparently cannot be satisfyingly explained in terms of traditional historical phonetics.

Direct link to Lay Summary

Last update: 21.02.2013

Active participation

Communication	Title	Media	Place	Year

The evolution of the Latin consonant cluster GN is a major issue of comparative Romance historical phonetics. At least six major outcomes are to be found in the Romance languages and dialects: 1) GN > [ñ] in Portuguese (lenha, punho, anho ‘lamb’), Spanish (leña, puño, añejo), Catalan (llenya, puny, anyell), Occitan (lenha, ponh, anhèl), French (O.Fr. leigne, poing ['poñ], M.Fr. agneau) and Italian (legna, pugno, agnello); 2) GN > [jn] in Engadinese (lain, quinó < *coinó < COGNATU), in several dialects of Central and Southern Italy and in some peripheral varieties of Sardinia (áino, -? < AGNU, púinu, -? < PUGNU); 3) GN > [wn] - with frequent epenthesis of g, v and m before the velar glide [w] - in some dialects of Southern Basilicata, Northern Calabria and Puglia (líunu, lígunu e lívunu < LIGNU, áunu e ámunu < AGNU); 4) GN > [mn] in Romanian (lemn, pumn, cumnat) and probably in Dalmatian; 5) GN > [nn] in Sardinian (linna, sinnu < SIGNU, connadu < COGNATU); 6) GN > [n] in some dialects of Central and Southern Italy such as Neapolitan (lena < LIGNA, prena < PRAEGNANS). Thus, referring to the final effect of the phonetic evolution, four different tendencies can be identified, i.e. palatalization (GN > [jn] and GN > [ñ]), velarization (GN > [wn]), labialization (GN > [mn]), and assimilation (GN > [nn]). For what concerns GN > [n], it is highly probable that this outcome represents a secondary evolution of GN > [jn], whose palatal glide has been absorbed by a preceding anterior vowel.Two main hypotheses have been proposed for the evolution of GN. According to Lausberg, whose explanation is mostly based on Diez’ and Meyer-Lübke’s grammars, the first segment of the cluster was subject in Late Latin to a weakening process resulting either in spirantization (gn > *?n) or palatalization ([gn] > [jn]): from *[?n] the outcomes [wn] and [mn] would have originated; from [jn], still to be found in Southern Italy, the assimilated outcome of Sardinian and the palatal nasal of Tuscan, Northern Italian dialects and Western Romance would have arisen. In his classical article "Degli esiti del latino GN nei dialetti dell’Italia centro-meridionale", though, Merlo proposes a wholly different explanation for GN > [jn] and GN > [wn]. According to Merlo, an epenthetic [i] or [u] would have broken up the GN cluster already in Latin, thus giving [gin] and [gun]: [gin] would have evolved into [jn], while [gun] would have developped into [wn]. Merlo’s hypothesis is followed by Rohlfs, who derives also Tuscan and Northern Italian [ñ] from [gin] through [jn], and by Ambrosini, who considers Sardinian assimilation [nn] as a secondary outcome of [wn] (through *[nw]). Both Lausberg’s and Merlo’s explanations derive the Romance outcomes from [gn]. However, most Indo-European linguists believe on the basis of comparative evidence that in Classical Latin the first segment of the GN cluster was not pronounced as a voiced velar stop, but as a nasal velar [?]. Although the majority of Romance linguists have rejected or ignored this hypothesis, some of them (Millardet and Bourciez) have tried to explain the Western Romance palatalized outcome [ñ] from [?n], thus postulating a reciprocal assimilation of the two nasals in the palatal zone. This explanation has been followed by the French linguists Fouché and Zink and more recently by the Spanish linguists Torreblanca and Wireback. Nevertheless, Loporcaro has shown on the basis of the symmetry with the evolution of GR (NIGRU > *néiro > nero and *níuru > ñur?) that the Southern Italian outcomes [jn] and [wn] are much more easily explained from [gn] than from [?n].The aim of my research is to test the different hypotheses that have been proposed by following the evolution of the GN cluster from Classical Latin to the contemporary Romance languages and dialects. Comparative studies have been recently performed on other problematic phenomena such as vowel diphthongization and nasalization. However, a comprehensive study of the evolution of GN in all the Romance varieties is still a desideratum. By collecting a large number of data from old and contemporary dialects and analysing them by means of experimental phonetics and philology, I will try to give a phonetically plausible explanation of the different outcomes of the Latin cluster. My major concern will be to reconcile the achievements of the Indo-European linguistics with the evidence shown by the Romance languages, thus attempting to derive all Romance outcomes from a generalized pronunciation [?n] of the GN cluster in Late Latin and Proto-Romance.The first part of my research will be devoted to tracing a detailed map of the outcomes of GN in the contemporary Romance languages. Between November 2009 and February 2010 I will gather data from the available bibliography, by inspecting all phonetic studies concerning Romance varieties and going systematically through the large collection of linguistic atlases kept in the Jud Library of the University of Zürich. At the end of February 2010 I will have located those dialects of Engadin, Southern Italy and Sardinia in which two or more different outcomes of GN are present. From March until June 2010 I will conduct linguistic inquiries in Central (Latium, Umbria) and Southern Italy (Lucania, Northern Calabria). During the summer (July-August 2010) and in the first months of autumn (September-October 2010) I will conduct inquiries in those peripheral areas of Sardinia (Barbagia, Trexenta) and Grisons (Engadin) where GN > [jn] is to be found. I will analyse all the data gathered during the inquiries in the Phonetic Laboratory of the University of Zürich. Between November 2010 and February 2011 I will compare my data with the information extracted from the existent studies and write down the first part of the work.In the second part of my research I will concentrate on ancient and mediaeval records. From March until June 2011 I will go through Late Latin inscriptions and Medieval Latin diplomatic corpora looking for non-standard orthographies of GN (as for instance , , ). From July until October 2011 I will inspect corpora of medieval Romance languages, especially those available on-line such as TLIO for Italian, DMF for French and CORDE for Spanish. The PHOENIX program, which is coordinated by Professor Gleßgen of the University of Zürich, will also be used for the analysis of Old French documents. Finally, between November 2011 and February 2012 I will go systematically through Romance medieval texts written in non-Latin characters (e.g. the aljamiado texts in Arabic and Hebraic characters from Spain and the documents in Greek and Hebraic characters from Southern Italy), looking for information on the phonetic value of the writing in Old Romance texts. The final part of my research will be dedicated to the outcome of GN in learned words. Since many Romance (especially Western Romance) languages show today learned or semilearned outcomes of the Latin cluster (e.g. Sp. magnífico, Port. assinar < *ADSIGNARE), reflexes of Latin GN in borrowings of non-Romance languages such as Basque, Arabic, Greek, etc. will be analysed. By comparing the data from the borrowings with the information given by medieval and modern grammarians (March-June 2012), I will try to follow the pronunciation of GN in learned and semilearned words from the Middle Ages until the present. Finally, in the summer and autumn 2012 I will attempt to propose a general explanation for the evolution of GN from Latin to the Romance languages on the basis of all the data I will have gathered during the three years of the grant.As a result, at the end of my research in November 2012 I will be able to write a comprehensive essay in the form of a book on the Romance outcomes of the GN cluster. This study aims to be an important contribution not only to Romance dialectology, but also to Indo-European linguistics and generally to historic phonetics.