<<

Guglielmo Orlandi

The affinity between

Friulian and Romanian

in an early publication of

The affinity between Friulian and Romanian in an early publication of G. I. Ascoli by Guglielmo Orlandi

All rights reserved. No part of this book may be used or reproduced in any manner without written permission of the publisher, except for the purpose of reviews.

© 2014 - Guglielmo Orlandi - Treviso ()

2

Premise

I decided to publish my thesis in the form of printed book to get her out from the inevitable yel- lowing to which it would be destined remaining buried in the dark meanders and laboriously ac- cessible of the university libraries. It is the thesis for the degree course in Humanities old system, achieved in the classrooms of Pa- dova University under the multi-year guidance of Prof. Lorenzo Renzi, who has been a precise and patient supervisor. All analyzes of were submitted to the vigilant attention of Prof. Laura Vanelli, which I wish to thank again for her courteous and professional support. I can- not overlook the fond memory of the late Prof. Alberto Zamboni, who accepted the task of assis- tant supervisor despite his busy schedule. This book is in full conformity with my original thesis, except the general modification to the for- matting of text and the add a final part which in those days was excluded from my thesis, but al- ways closely related to the topics covered by G. I. Ascoli in his juvenile linguistic essay.

Guglielmo Orlandi

3 I. Introduction

I focused my research on the years preceding 1846, date of publication of the work Sull’idioma friulano e sulla sua affinità colla lingua valaca - Schizzo storico filologico of young Goritian Graziadio Isaia Ascoli, asking myself first of all what had pushed a boy to publish a historical phi- lological essay in the proximity of his seventeenth birthday. He was in fact born in on July 16, 1829. Considering the time required for printing (typesetting did not exist) and for a reasonable period for the research and writing of the work, we can be traced back the beginning of his work at least one year before. It is not integral part of my research determining whether the game of football already existed in in the first half of the nineteenth century, but I would feel more relaxed knowing that guy also involved in some sports, maybe in some square or cobbled alleyway of Gorizia, but well out- doors. And then his course of private studies had already concluded? to enable him to get distracted so long in a matter that could be very challenging even for the few specialists of the time? And the business of the family? He was or was not Graziadio involved in the first person in the daily af- fairs, together with his mother Elena? I’ll try later to find an answer to these questions. That of the young Ascoli is the obvious case of a boy with interests and personality out of the town, who had the ability to quickly neutralize all the normal distractions characteristics of his age, adverse to a strict application to studies. It is only thanks to an iron discipline (which certainly was not of much weight for him) that Graziadio managed to carry out his feat and with so well of print publication. I would add only that the characters of the people can be outlined very clearly from the earliest stages. Beyond the conditioning and the incentives of environment, that we cannot deny, it is cer- tain that the young Graziadio had to contain in his unconscious the germ to the intuition and lin- guistic research, like Mozart had for music. But I would also add that if Leon Flaminio Ascoli, father of Graziadio Isaia, had never moved from Livorno - instead of emigrating to Gorizia - perhaps the fate of future father of Italian lan- guage would have presented quite different implications. While leaving unchanged his deep curi- osity about culture in general and the natural talent for the historical study of languages, the young Ascoli would grow up in a environment less stimulating of multilingual Gorizia, because limited by linguistic monolithism of the Tuscan area.1 But perhaps he could devote himself to study of Portuguese, which was the language used in the ritual of Jewish community in the city. Or at most he could focus his passion on Hebrew, the an- cient language of the fathers, brilliantly retracing the footsteps of his predecessors, supported and encouraged by a climate of tolerance that in Gorizia it was instead denied to him.

1 For a citizen of Jewish origin, however, as you will see in the next lines, the monolithism of the Tuscan language was neutralized by the presence of Portuguese, Hebrew and even the ‘bagitto’, variety of Judeo-Spanish spoken in the city of Livorno by a part of the Jewish community (until the second World War), after the expulsion of the fifteenth century by Catherine of Aragon.

4 But it was really the Gorizia’s environment that saw him grow, and it was this which fueled the many interests of Ascoli for culture in general and linguistics in particular. It was the linguistic complexity of the town of Gorizia, where they intertwined languages as the literary Italian with the German of Habsburg and the Friulian of neighbors with the Triestino of his mother Elena, the Slovenian of near mountains with the Venetian of merchants and the Hebrew of his childhood reli- gious, to shape day after day the ingenious intuition and the pleasure of research. It was further- more the excellent Jewish and Catholic cultural environment of Gorizia that facilitated access to the deepening of his historical and philological studies, offering to him the availability of talented teachers and well-stocked libraries. I will try, in course of my research, examine the environmental and cultural conditions of the time of Ascoli to understand what was the motivation that led him to realize his booklet, and to which model he was inspired. I’ll try to figure out why he wanted to deal with the hassle of a comparative study instead of a dictionary or grammar rules. And then because he wanted to put in their comparison just the Romanian, instead of another Romance language, with the Friulian. And this? Who informed the young Graziadio that the Friulian «idiom» was really a language to consider at the same level of a national language, as the Romanian, instead of a ‘rough’ deformation of Ital- ian? Also to these questions I will devote a specific comment within my work. Even knowing that in the years of his maturity Ascoli will qualify this pamphlet as “un lavoruccio insig- nificantissimo”,2 I will proceed in the second part of my job to a thorough linguistic analysis of his comparative research by comparing the descriptions and the analysis of the results described by Ascoli with those of later linguistic research, to understand the extent to which came his intuition, despite his young age and the feeble ‘technical’ resources put available by the precocity of time.

2 G. I. Ascoli, Studj critics, vol. I, extracted from “Studj orientali e linguistici”, fasc. III, Gorizia, Paternolli 1861, p. 57, n. 3.

5 II. Graziadio Isaia Ascoli

II.1 Premise

The younger days of Graziadio Isaia Ascoli are essentially linked to the territory of Gorizia, where they developed the corporate commitments of Leon Flaminio Ascoli, dad of Graziadio. As you will see in detail later, the phase of Gorizia will represent for Ascoli a enough short period of his life, which goes inserted between the Tuscan roots of his family and the Milanese period of his long academic career. For this reason, I need to expose a summary of the historical events that in- volved closely the family of Ascoli, starting from the events of Jewish community of Livorno, then move on to historical description of Gorizia and its Jewish community.

II.2 Jewish community of Livorno

Among the various nationalities who resided in Livorno, the Jewish community was without a doubt the most numerous and economically and culturally the most important. The Jewish pres- ence grew slowly in the late sixteenth and early nineteenth century to represent approximately 10% of the total population of the city. The Jews will make in Livorno the trade brokerage be- tween the ports of Levant and the squares of Italy and Northern , taking advantage of the privileges granted by the Medici government to the city. The very baggy privileges granted by Grand Duke Ferdinand I (1587-1609) with letters patent of June 10, 1593, known as the “Constituzione Livornina”, aimed to summon at Livorno the Spanish- Portuguese Jews who had been expelled from the Iberian Peninsula at the end of the century XV, they had enabled these last to reach in Livorno an economic prosperity and cultural rarely equaled in other communities of Mediterranean: there was guarantee for the marranos1 to freely practice Judaism without being prosecuted by Inquisition, there was freedom to study and to achieve aca- demic qualifications, to own property, to reside in the district open (in Livorno there was never ghetto), to settle in the city and depart freely with their goods, to print Jewish books, to independ- ently administer justice in lawsuits between Jews. The climate of relative tolerance and freedom established with such privileges favoured the flow- ering of Jewish studies for which Livorno became famous for at least three centuries: rabbis and scholars flocked to Livorno, where they found a favorable environment, patrons willing to help and finance studies and publications, education institutes and Talmudic academies provided of no- table libraries. The decline of Livorno as square of international trade, accelerated by the Napoleonic storm2 and by the absence - following the abolition of free port - of traffic of warehousing and brokerage, in- volved the Jewish community, which saw decrease steadily in the nineteenth century its importan-

1 Jews converted to Christianity against their will, but suspected to preserve the old faith in secret. Exiled from Spain and emigrated to Portugal, were also persecuted and driven from the country. 2 Occupied by the French in 1795, Livorno remained united to Napoleonic empire until 1814.

6 ce and its numerical consistency, and therefore the functionality of the institutions. The Jews of Livorno, abandoned the big trade, gave themselves to the retail sales, to liberal professions and gave their valuable contribution to the local and national culture.

II.3 Gorizia

Where the plain of Friuli goes to meet the last offshoots of Julian , in a valley surrounded by low hills of Collio and karstic plateau, rises Gorizia (Slovenian Gorica, German Görz), a small border town collecting on the slopes of hill where owes its name (gorica in Slovenian it means ‘lit- tle mount’). This place is first documented in 1001, when Gorizia appears in the diploma with which Emperor Otto III donated a house for half to Patriarch of and the other half to Count of Friuli. The part of the latter then passed to Dukes of , from which was passed to the house of Bavar- ian origin who assumed the name and possession until 1500. Maximum expansion of the small state of Gorizia occurred between the mid-thirteenth century and beginning of the next, when it was easy to Counts to annex the Tyrol and ensure the control of Treviso March, even for short pe- riods. With the death of Henry II began the decline of County that, in the grip between the Repub- lic of Venice and the Empire ruled by the Habsburgs, gradually it lost importance. In 1500, the Counts of Gorizia died out and their estates were inherited by Habsburgs, who incorporated Go- rizia in the Empire of Austria. In new political and territorial arrangement the extended from the high valley of Isonzo to Aquileia, from the hills of Cormons to the rocky outcrop of Duino and also included the fortress of Gradisca, built by Venice between 1478 and 1498 to defend frequent incursions that Turks were making in Friuli. Gorizia reached its maximum degree of political stability and territorial during the eighteenth cen- tury with the government of Empress Maria Theresa, and in the second decade of her kingdom there was introduced the silk industry. This century was a period of moderate prosperity for the County of Gorizia, which developed an intense cultural and social life. It was also the century of reforms and the loss of old autonomy: II at the end of the century subjected administra- tively the county to , which became the main port of empire. In the nineteenth century Gorizia continued to be in Austrian territory and was dominated by the , which accompanied the Friuli from many centuries. It kept intact, however, its origins and its traditions, emotionally binding itself to Italy, to whose culture it was push- ing the orientation of Italian bourgeoisie, and which was involving also the popular classes of Friulians and Slovenians (the latter, however, will be geared later toward Jugoslavia). It had to been surely widespread in the educated classes the knowledge of Italian literary language and as in other parts of Empire where there was an Italian component (the nearby Trieste, the whole , different centers of , the ) it was widespread hope of a future unifi- cation with young . The Italians of Austria in fact considered themselves “very Italian” and “unredeemed”.

7 II.3.1 Jewish community in Gorizia

The presence of first Jewish communities in the Isonzo valley dates from the late fifteenth century. In 1500, at death of Leonardo, the last Count of Gorizia, the county passed to the Hapsburgs and under Emperor Maximilian I the Jews got a series of privileges and the support of local authorities. One of the Jewish families that stood out was that of Morpurgo. The activity to which devoted most of Jews was banking. During the seventeenth century, the Jewish community was enriched by new members. Many families obtained the title of Hofjuden (Jewish of court) and this allowed them to have more free- dom in business, not to carry the hallmark and being considered stable citizens of Empire, as well as owning real estate. At the end of century, March 24, 1696, was imposed, however, the transfer of Jews in Contrada del Ghetto, the current via Ascoli. It was in fact issued by Emperor Leopold I (1658-1705) an edict which enjoined the Jews to Gorizia locking themselves into a separate dis- trict by Christians. Two gates were closing the Gorizia ghetto in which it lived an industrious community devoted to silk spinning and production of wax. Here was built in 1756 also the Syna- gogue,1 even now preserved and can be visited also as Museum of Jewish community. During the 1700s the Jews enjoyed a special economic development and a population growth, but spite of the privileges renewed also by Charles VI, they remained confined in ghetto, so to many restrictions, including a ban of using servants or Christian nurturers, of open their shops dur- ing Christian holidays, get out of the ghetto during the processions nor during the last three days of Holy Week, to own homes or funds outside the ghetto and the obligation during Lent to be present at least twice at Christian sermon. The gates of ghetto were also constantly monitored by guards. Many Jews continued to live in the ghetto after the “License of tolerance” of Joseph II (13 October 1781) which established religious freedom for all subjects. The same Emperor gave off January 21, 1790 a “Special Regulations for the Jews of Gorizia and Gradisca” (Judenordnung), according to which the Jews had equal rights and duties to other subjects, including permission to practise any trade or profession, it being their prohibited only the management of taverns, considered pro- moters of idleness, and was abolished badge to be worn on dress.2 For the first time the Jews were able to choose their profession and access to each school. Joseph II, in fact, was intended to facilitate the process of assimilation of the Jewish community in Go- rizia. It was granted the freedom of religion and German schools were established within the Jew- ish community with obligation of study of German language. The certificate of knowledge of German language was required to begin the study of Talmud and before they can marry. It became mandatory also military service and they were forced to submit to the authority of State and not more to that rabbinic one. Banking activity was suppressed after the notification of first Archbishop of Gorizia, Carlo Michele Attems, for excessive wear applied by banks.

1 Where now you can see ritual objects, books, documents and photographs. To learn more about the history of Jews of Gorizia you can visit the old cemetery Valdirose (Rozna Dolina) now in Slovenia, but not far from the state border (crossing of Red House). Many famous people are buried here, including Rabbi Isaac Reggio, who is responsible for the flowering of Jewish culture in the nineteenth and twentieth centuries.

8 Gorizia became a center of Jewish culture, as to be called “Little Jerusalem on the Isonzo”, but also a center of the Enlightenment movement. Jews were allowed access to every type of studies. Trade and industry flourished, especially after the expulsion of the Jews (in 1777) from the territo- ries of Venetian State free of ghetto, when the community of Gorizia became more numerous. After the death of Joseph II (1790) was recorded a decline of liberal policy towards the Jewish subjects. This situation had, however, a limited duration because in 1797 began the French occupa- tion that lasted until 1814 with ups and downs. With arrival of French, the situation of Jews improved. In 1812 the gates of ghetto were demol- ished, the compulsory residence was abolished, and for the first time the Jews were able to hold public office. With regard to the facts of Gorizia, Carlo Dionisotti informs us that “negli ultimi tre anni dell’impero napoleonico (1811-1814), essendo stato istituito a Gorizia un liceo, l’ebreo Reg- gio3 era stato preferito a ogni altro come insegnante di materie letterarie nelle classi superiori. Il successivo governo austriaco aveva posto fine allo scandalo.” 4 Gorizia was the only one of Jewish communities of Habsburg Italy that, from the Restoration, saw clear all the rights gained by the Napoleonic legislation.5 Only after January 15, 1818 were ac- knowledged on a provisional basis to the Jews all the rights previously acquired, except admission to public office. During the nineteenth century the population in ghetto declined much, because the Jews of Gorizia moved to more central areas and residential of city. The number of Jews in the community was stabilized thanks to arrival of teams from central Europe. At the times of Ascoli, the Jewish com- munity of Gorizia was composed of 300 people in a town of about 13,500 abitanti.6

2 PC Ioly Zorattini (ed.), Gli ebrei a Gorizia e Trieste, tra «Ancien régime» ed emancipazione, Atti del Convegno di Gorizia, Udi- ne, Del Bianco, 1984, pages 125-126. 3 This is Isacco Samuele Reggio (1784-1855). 4 C. Dionisotti, Appunti su Ascoli, in “Le tradizioni del testo. Studi di letteratura italiana offerti a D. De Robertis.” edited by F. Ga- vazzeni e G. Gorni, Milano-Napoli, Ricciardi, 1993, p. 422. 5 M. Berengo, Gli Ebrei dell’Italia asburgica nell’età della Restaurazione, in “Italia”, n. 1-1, 1987, pages 62-103. 6 C. Dionisotti, Appunti su Ascoli, cit., p. 421.

9 II.4 Graziadio Isaia Ascoli in Gorizia: his younger days

Because this work is devoted to an early essay of Ascoli, I consider it appropriate to present a de- tailed biographical account of his youth, to which I will follow a brief reference to maturity. Graziadio Isaia Ascoli was born in Gorizia July 16, 1829, second son of a family of wealthy Jew- ish entrepreneurs, in the paternal elegant building on Via San Giovanni al Municipio, in Contrada del Ghetto, now via Ascoli. The ancestors of his family came from Ascoli Piceno - then Papal ter- ritory - and they were emigrated first to Massa and then to Livorno, in the Grand Duchy of Tus- cany. The Graziadio father, Leon Flaminio, from Livorno had moved to Gorizia (as we know from a let- ter of 1878 of the same Ascoli to De Gubernatis1), where in January 1799 he had married Bersabea [sic] Luzzatto.2 In Hapsburg Friuli, Leon Flaminio Ascoli had built a substantial fortune by start- ing his industrial activity with two paper mills, one in Podgora3, the other Passariano4, and a textile mill near Strassig5 by buying November 25, 1807 the house on which Graziadio Isaia will live un- til 1861. Widower, November 2, 1827 he married Helen Norza by Trieste, with whom he had Ber- sabea, called Betty, born on August 8, 1828 and in July of following year Graziadio Isaia. When Graziadio was just eight months Flaminio Leon dad died prematurely on March 10, 1830. Mama Elena Norza took on the entire administration of the company inherited from the husband and the onerous task of raising and educating children Betty and Graziadio. It is certain that the Goritian young never performed regular studies, due to a ban that struck then the Jews, which excluded them from public schools.6 The same Ascoli confirms the fact, in a letter dated 4 May 1872 to De Gubernatis: “Non sono mai stato a scuola, ma sono autodidatto nella più schietta significazi- one del vocabolo; ed ho, degli autodidatti, tutti i vizj e qualche virtù”.7

1 Ascoli wrote about his name: “ma la mia firma ufficiale ... altro non è se non Graziadio Ascoli, combinazione onomatologica avutasi più volte fra gli ebrei di To- scana, ai quali mio padre apparteneva” (BNCF, Carteggio De Gubernatis, lettere Ascoli, ins. 22, 3, 1). Angelo De Gubernatis (Turin 1840 - Rome 1913) represented a unique and eclectic figure of scholar. He was a pioneer of the italian orientalistic, polygraph publicist, pacifist, participating of the positivist debate and open to all problems of his time, in particular to the improvement of education and women’s emancipation. His degree in Literature was the first of Kingdom of Italy. He perfected himself abroad, in Berlin, in phi- lological studies, especially in Sanskrit. At just 23 years Angelo had the chair of Sanskrit at the Institute of Higher Studies in Flor- ence. Collaborator and founder of magazines and newspapers, in 1891 he passed from to the University of Rome as ordi- nary professor of literature also performing the teaching of Sanskrit. He wrote a universal history of literature in 18 volumes, a mythical history of the animals and a number of dictionaries. His production has not always generally been assessed positively; but was instead evaluated his tireless work, even abroad, to divulge an image of Italian secular scientific world and open to more differ- ent orientations. 2 O. Altieri, La comunità ebraica di Gorizia: caratteristiche demografiche, economiche e sociali (1778-1900), , Del Bianco, 1985, p. 130. 3 Slovenian version of Piedimonte del Calvario, today western outskirts of Gorizia, on the right of river Isonzo. 4 Little town at just four kilometers to the east of . 5 Now Straccis, between Gorizia and Podgora, on the east bank of the Isonzo. 6 F. D’Ovidio, Commemorazione dei Soci G. Ascoli e G. Carducci, « RAL», 5, XVI, 1907. 7 G. Breschi, L’«Autobiografia» di G. I. Ascoli, in “Archivio glottologico Italiano”, LVIII, fasc. I, 1973, lettera XI, MIlano, 4 may 1872, p. 77.

10 The well-off conditions of his family, however, allowed him a private education with good teach- ers and permitted him to get all those books that were needed to deepen his studies. Among his masters, there was also a certain Solomon Lolli, Israelite, agent and employee of his mother in the family firm, especially well versed in Hebrew, which at the time of his death (in 1865) will be re- membered by Ascoli with painful affection. But after the education received by his tutors, the young Graziadio followed his vocation to study alone. This natural inclination was certainly fa- vored by the particular ethnic and linguistic conditions of Gorizia.8 This city indeed offered to him with ease the printed languages of literary Italian, of German and Hebrew, in addition to intermin- gling of languages such as Slovenian and Friulian, spoken by the common people, the Venetian of the bourgeoisie and the Triestin9 of Mama Elena. What language of colloquial use he spoke the young Ascoli is difficult to establish, given the com- plex linguistic reality in which he lived, but one cannot exclude use of , which was lan- guage of wealthy bourgeoisie. As for use of Friulian is interesting an confidence written by Graziadio to Marco Luzzatto in a letter dated September 10, 1846: “A San Daniele ho parlato friulano a lungo credo la prima volta in vita mia, e v’appresi qualche voce nuova”.10 We deduce that Ascoli had a predomi- nantly passive knowledge, and not perfect, of the language to which he would give a contribution so formidable in the future. The small Graziadio had a private library of which he himself informs us. At the age of ten, July 12, 1839, he drew up a catalog of 49 volumes among which there were four German grammars, a dictionary German-Italian and a Hebrew grammar. In March 1842, the number of volumes had risen to 171, but to the books of linguistic nature he had not added that a German grammar. On 11 August 1845 the volumes had risen to 380 and there were increased especially the historical and literary works. Works of out-and-out linguistic we find only in the catalog of 1849, when volumes reach the large number of 1318.11 But the factor that contributed perhaps more than any of cultural formation of young Ascoli was certainly the environment that he was lucky enough to find around his house. Coming out of his house, in fact, he could take a few steps to left and he was finding the Jewish world of Synagogue, very active center of studies since 1756. Here he met Rabbi Abraham Vita Reggio and his son Isaac Samuel Reggio,12 linguist and scholar of great stature.

8 See the previous paragraph. II.3. 9 In Trieste they were spoke until about 1830 a dialect of Friulian type, which is now called Tergestino or old Trieste, to distinguish it from the present Tergestino dialect, which is formed instead on basis of Veneto. It is certainly this latter which they were speak- ing in Ascoli home in the forties of nineteenth century. See M. Doria, Storia del dialetto triestino, Trieste, 1978. 10 G. I. Ascoli, Composizioncelle, Archivio Ascoli, 3/163, c, p. 40. 11 G. I. Ascoli, Cataloghi di libri, Archivio Ascoli, 5/153. 12 Isaac Samuel Reggio, rabbi, mathematician, Hebrew scholar, teacher (Gorizia 1784 - 1855 therein). He was rabbi of Jewish community in Gorizia, where he also distinguished himself as an educator. He published, among other things: Piccola enciclopedia scritta per uso dei giovanetti (Gorizia 1812); Dissertazione sulla divinità della legge ebraica (Vienna 1818); Die Religion und die Philosophie (Wien 1827); Il libro di Isaia: versione poetica (Udine, 1831). He also compiled, from 1852 to ‘55, a “Jewish strenna” output in Gorizia, exemplary document of his vast learning and his moral qualities. Read more: G. Hughes, Di alcuni illustri semiti- sti e orientalisti della Venezia Giulia, SG XXIV, 1958; Enciclopedia Motta, 1970; Enciclopedia universale Rizzoli-Larousse, Milan 1971; M. Grusovin, Abram Vita e I. S. R., in La speranza. Attraverso l’ebraismo goriziano, 1991; Id, I. S. Reg- gio rabbino e filosofo, in “Quaderni Giuliani Storia”, 1996, 17, 2, p. 7-29. [from Dizionario biografico friulano, 2002]

11 A few steps to right and he could cross the Central Library of the Theological Seminary, where he met the rector Don Stephen Kocjančič,13 distinguished multilingual, which deserves to be counted among the leading friends of Ascoli. Since you still retain the records of the volumes of library of the Seminary loaned to Ascoli, it is reasonable to infer that Don Kocjančič was the only Christian to have a direct impact on his personality, given that the letters of their correspondence, which date back to years 1855-59, testified that their contacts were maintained for a long time.14 He had a great reverence for the famous semitist Samuel David Luzzatto,15 a professor at the Rab- binical College of Padua. Luzzatto was one of the renovators of poetry and Hebrew philology in Italian, translator of the Pentateuch and author, among other things, of a Grammatica ebraica. To him Ascoli sent a copy of the «Sull’idioma friulano» to receive a judgment. Luzzatto replied him with a long letter on November 8, 1846 expressing some criticisms which I will discuss in more detail in Sec. V.1.1.16 Graziadio Ascoli was bound by a deep friendship with the son of Samuel Luzzatto, Filosseno (Tri- este 1829 - Padua 1854), he also big fan of oriental studies and linguistic. As we know from the biography drawn by his own father, in a letter to Immanuel Wolf,17 Filosseno, after a regular course of high school and one year of college courses, he studied at the rabbinic Institute of Padua, without receiving a degree. Very precocious Epigraphist and linguist, died at only twenty-five years, having written a few articles published in Milanese magazines and on «Journal Asiatique». Umberto Saba said of him, in the notes to the story Un letterato ebreo: “A Parigi, dove aveva avuto il permesso di recarsi per ragioni di studi, si prese invece subito una malattia venerea. Mancandogli il coraggio di confessare a suo pa- dre il delitto e il castigo, la malattia fece rapidi progressi ed ebbe, per il povero giovine, le conseguenze che si possono immaginare”.18

13 Stephen Kocjančič, priest, teacher, linguist, writer of history (Vipacco/Vipava 1818 - Gorizia 1883). He was rector of the Theo- logical Seminary of Gorizia, librarian of the central and metropolitan Seminary canon. It was perhaps the greatest connoisseur of languages (at least thirty between living and dead, between ancient and modern) lived in Gorizia. He had sure knowledge, in par- ticular, of the , especially of Hebrew and Arabic, of Old Church Slavonic, of Sanskrit, Persian and Coptic. He left unpublished a large dictionary Slovenian - German and one of Ethiopian language. He encouraged the young G. I. Ascoli in the years of its preparation. He published, inter alia, the Historia Archidioecesis Goritiensis (Gorizia 1875). He translated also in Friulian the first three chapters of De imitatione Christi. Read more: A. Cronia, Un inedito e grande dizionario sloveno-tedesco di S. K., SG XVI, 1954; B. Marušič, Il Friuli goriziano nelle opere di S. K., Atti del congresso internazionale di linguistica e tradizio- ni popolari, Udine 1969; P.C. Ioly Zorattini, S. Kocjančič (1818-1883). Un ecclesiastico al servizio della cultura fra Sloveni e Friu- lani, Gorizia, 1984. [from Dizionario biografico friulano, 2002] 14 G. Hugues, Di alcuni illustri semitisti e orientalisti della Venezia Giulia, «Studi Goriziani», XXIV, 1958/2, p. 75 and foll. 15 For the biographical synthesis of S. D. Luzzatto see par. III.2.2. 16 Epistolario italiano francese latino di S. D. Luzzatto di Trieste pubblicato da’ suoi figli, Padova, Tipografia della Minerva dei fratelli Salmin, 1890, parte I, n. CCXCVII, pages 485-91. 17 Epistolario Luzzatto, cit., parte II, n. DLXXI, pages 883-86. 18 U. Saba, Prose. Ricordi-Racconti, editrd by L. Saba, Milano, 1964, p. 23.

12 A significant role was certainly the abbot Jacopo Pirona19, “l’illustre filologo” by which Ascoli was al- ready in correspondence from the 40s, author of first Vocabolario friulano, to whom he dedicated his first historical-linguistic Sull’idioma friulano....Schizzo storico-filologico.20 How it took place the life of young Graziadio and what were his interests, we can see it in the let- ters he wrote to many friends and relatives. “Cerco poi nelle mie letture” he wrote to his cousin Cesare 21 Norza of Trieste, May 9, 1846 “cose per lo più gravi, evito la lettura di romanzi, preferendo quella della storia, e di quella d’Italia particolarmente m’occupo. Sorto molto di rado, eccettuandosene una mezz’ora la sera e formo la mia distrazione alterna- tivamente o col canto, o colla pittura, o colla poesia … Un’ora o due al giorno sono dedicate al mio commercio, il quale fortunata- mente non è di tal natura da impedire le mie occupazioni letterarie ... I divertimenti che offre la nostra città e i suoi dintorni sono una buon’aria, una ridente campagna, un cielo puro purissimo ed una temperatura mitissima”. In another letter written in to his friend Moses Gentilli of Padua in the same period (“VI ante Idus Majas”) he was re- peating “Raro egredior et tantum ad vesperum ut scis”.22 What were the patterns of comparative linguistics to which the sixteen Graziadio gleaned to design and then make his first linguistic essay «Sull’idioma friulano» we do not know, because he never quoted his linguistic patterns, but only works as the Storia Universale of Cesare Cantù and the Deutsch-Walachischen Grammatik of Jon Molnar. As is well known Ascoli was preceded by Carlo Cattaneo on this path, although it seems certain that he was unaware of his work, entitled Del nesso fra la lingua valaca e l’italiana, appear in May 1837 in Vol. LII of «Annali universali di Statistica, Economia politica, Storia, Viaggi e Commercio», that young Goritian would have read only in November 1849.23 We only know he sent a copy of his paper to Samuel David Luzzatto to receive his authoritative judgment. And he did not delay to express to him that “la lingua della Crusca non può essere illustrata senza il confronto degli altri dialetti italiani, e questi non possono illustrarsi senza il confronto delle lingue dei barbari, germani e slavi, che diedero origine alle nuove lingue latine.” 24

19 Jacopo Pirona, abbot, teacher, writer ( al 22/11/1789 - Udine 01/04/1870). Teacher and then director of the school of Udine, the first curator of the Museum of which he was the promoter and president of Academy of Udine. It was pol- ished writer in Italian and Latin, had corresponded with many scholars of that time. G. I. Ascoli dedicated him the early work Sull’idioma friulano e sulla sua affinità alla lingua valaca. His name is linked particularly to Vocabolario friulano that began to emerge periodically since 1868. The work was printed in its entirety in Venice in 1871. In 1848, in collaboration with C. A. Car- nier, directed the political journal “Lo spettatore friulano”. [from Dizionario biografico friulano, 2002] 20 G. D’Aronco, G. I. Ascoli, in “Studi Goriziani”, XXIII, 1958. I. Bonomi, Note sulla formazione e gli interessi linguistici del gio- vane Ascoli, in “Parallela 5. Atti del VI Convegno italo-austriaco dei linguisti” edited by M. Dardano, W. Dressler, C. Di Meola, Milano, Bulzoni, 1995, p. 30. 21 I. Bonomi, Note sulla formazione e gli interessi linguistici del giovane Ascoli, cit., p. 41. 22 From letters that are found in Ascoli Archives; see M. Radoni Zucco, Profilo biografico di G. I. Ascoli, in G. I. Ascoli e l’Archivio Glottologico Italiano, edited by M. Cortelazzo, Udine, Doretti, 1973, pages 23-24. 23 Cfr. B. Terracini, G. I. Ascoli, direttore dell’«Archivio» (dal carteggio Ascoli-Salvioni), in «AGI», LII, 1967, pages 1-54. 24 S. D. Luzzatto, Epistolario italiano, latino e francese, cit., parte I, n. CCXCVII, pages 485-491. See par. V.1.1.

13 Later, in the enthusiasm aroused by the revolution of ‘48, he wanted to launch a proclamation to his countrymen, by publishing the pamphlet Gorizia italiana, tollerante, concorde. Verità e sper- anze dell’Austria del 1848, at the Printing Paternolli, to illustrate the reasons of the separatist ten- dencies of the Lombardo-Venetians, talking about the issue of nationality, of italian education and by discussing of language and the problem of religious tolerance. On 10 October 1848, the «Gazzetta di Venezia» - the official gazette of the Venice Republic of Manin - published a paper titled Due parole di un crociato goriziano ai fratelli italiani e alla po- tenza d’Europa which said “la lingua e non solo la lingua delle persone colte, ricche, ecc., ma che pure la lingua del volgo, dell’artigiano, dell’infima plebaglia delle vie è l’Italiano” 25 referring to Gorizia. Now we know that - which referred to independence of Italy - was of Ascoli. In 1850 we have news of his engagement with Fanny Beatrice Cohen, which took place in Trieste, March 23, as per contract, while his sister Betty will marry Isaac Pesaro on September 18, taking residence in Venice. Despite the new affective commitments with his girlfriend Fanny, the growing tasks in his paternal firm26 and the occupation to his difficult studies, Ascoli had also the time to devote himself to the duties to Jewish community, that he headed since November 1850 until 1853. In 1851 he devoted himself to a curious work, the Pasitelegrafia. System of universal language to apply to electric telegraph, in which he proposed a system of universal language to be applied to electric telegraph.27 From a travel diary28 of 1852, you know that Ascoli conceived the idea of founding a magazine and for this purpose he went on the road in with his young wife - who had married January 4 of that year - and his mother Elena, to make contact with renowned Orientalists and Ital- ian linguists. During this trip he talked with Paolo Marzolo, Samuel David Luzzatto, Gabriele Rosa and then with Carlo Tenca, Cesare Cantu, Giovanni Flechia, Amedeo Peyron and others. On his return to Gorizia he founded a linguistic magazine entitled Studj orientali e linguistici, writ- ten almost completely from Ascoli himself (one article was by G. Rosa) and printed by typography Paternolli in three booklets (1854, 1855, 1861). The magazine contained a large collection of translations of Indian texts, of remarks on topics Semitic and Italian , in addition to in- formation of Indo-European linguistic, taken from the most recent studies of the German linguists. Following extensive experience accomplished on their studies, together with the originality of his research and his methodical rigor, the fame of Ascoli soon spread throughout Italy, so much so that in 1860 he was offered the chair of Semitic languages at the University of Bologna.

25 The original autograph is located at the , Fondo Ascoli, pacco 163. Cited in M. E. Loricchio, Graziadio Isaia Ascoli, biografia di un intellettuale, Mariano del Friuli, Edizioni della Laguna, 1999, p. 21, note 12. 26 There is a letter from his mother, kept at the Archivio di Stato di Gorizia. Cited in M. E. Loricchio, Graziadio Isaia Ascoli, bio- grafia di un intellettuale, cit., p. 21, note 14. 27 Read more, see I. BONOMI, Idee per un progetto di lingua universale in uno scritto inedito del giovane Ascoli, in Milano e l’Accademia Scientifico-Letteraria. Studi in onore di M. Vitale (edited by G. Barbarisi et alii), Milano, Istituto Editoriale Cisalpino, 2001, pages 385-402. 28 Note letterario-artistiche minori durante il viaggio nella Venezia, nella Lombardia, nel Piemonte, nella Liguria, nel Parmigiano, Modenese e Pontificio. Maggio-giugno 1852, published in the “Annali della Scuola Superiore di Pisa”, XXVIII, 1959, pages 151- 191, by S. Timpanaro.

14 It was therefore time, for the enterprising Goritian linguist, to leave Austria and his beloved Go- rizia. The designation at the Bologna University was recommended to the Minister Terenzio Ma- miani from Gabriele Rosa29, that Ascoli had met during his travels in northern Italy in 1852. The Goritian scholar, however, asked they changed destination. On the unexplained reasons which led to rejection of Ascoli, we are aided by this hypothesis of Carlo Dionisotti “È probabile che Ascoli pru- dentemente diffidasse di una sede universitaria prestigiosa, ma presieduta fino all’anno prima dall’autorità ecclesiastica, e in tale sede di una cattedra che lo avrebbe fatto successore e rivale dell’ancora famoso poliglotta cardinal Mezzofanti”,30 but it is also pos- sible that Ascoli wanted go to teach in Florence at the Istituto di Studi Superiori, as is evident by a letter to Michele Amari June 186031. From Turin it came January 2 the following year this letter in his own hand of the Minister of Public Education:

Mio Signore.

Veggo dalla cortesissima Sua, ch’Ella preferisce Milano a Bologna e l’insegnamento della Grammatica e San- scrito a quello delle Lingue semitiche. O’ per fortuna segnalata il poterLa contentare appunto. Faccia conto insino da ora di essere professore nella nuova facoltà milanese di Filosofia e Lettere la cui apertura avrà luogo li 14 del presente mese. Si affretti, per tanto di trasmutare la Sua dimora e mi avvisi dove e quando Le debbo far tenere il decreto di nomina il quale domani stesso verrà sottoposto alla firma di S. Maestà. Mi creda con alta stima Suo devotissimo Terenzio Mamiani

The letter was sent by way of Vienna and bore the address:

Al Chiarissimo G. I. Ascoli GORIZIA.32

Not without some other correspondence with the Minister Mamiani, where Ascoli dictated other conditions and details, the scholar of Gorizia finally accepted the chair of Comparative Grammar and Oriental languages at the Academy Scientific-Literary Milan, thus inaugurating the first chair linguistics in Italy.

29 To find the correspondence between Rosa and Ascoli on his appointment university, see D. Santamaria, G. I. Ascoli e la lingui- stica italiana del primo Ottocento, in G. I. Ascoli, attualità del suo pensiero a 150 anni dalla nascita, Firenze, 1986, p. 237. 30 C. Dionisotti, Appunti su Ascoli, cit., p. 425. 31 Letter published in M. Ziino - G. Gabrieli, Lettere di Michele Amari a Graziadio Ascoli, in “Archivio storico Siciliano”, LIII, 1933, pages 225-258. 32 The correspondence between Ascoli and the Minister Mamiani is reported by Carolina Coen Luzzatto in Graziadio Ascoli. 1860- 1861, Gorizia, Paternolli, 1907, p. 18 and foll.

15 Before moving to Milan with the whole family, June 2 1861 Ascoli concluded the contract of sale of the paper mill and the spinning-mill in Piedimonte with Baron Hector Ritter de Zahony33 and gave in rent his paternal house to Gorizia town. After the post of lecturer at the University of Mi- lan Ascoli abandon forever, it seems, every business activity.

II.5 Graziadio Isaia Ascoli in Milan: maturity

Since the activity of Ascoli which corresponds to period of his maturity is more known than of his youth (which I discussed in the previous section), I will limit myself to exhibit in this section of the ascoliana biography only a few brief notes about the fundamental moments of his work done in Milan.1 In 1861, though working in Milan, he wanted to publishing in Gorizia the first volume of the Studi critici, in which he examined the Bernardino Biondelli’s work of 1856, the «Saggio sui dialetti galloitalici» (Milan, 1853-56). From 1867 to 1870 his research has looked more and more on issues of and especially of phonetics Indo-European. Part of these studies was published in 1870 under the title Corsi di glottologia and with subtitle Lezioni di fonologia comparata del sanscrito, del greco e del latino. The key date of ascoliana career is represented by 1873, the year of birth of famous Archivio glot- tologico italiano (AGI), magazine that saw him founder, curator and director until to the XV vol- ume (1899-1901). The first volume of AGI (curious acronym of his initials), written entirely by Ascoli, includes the Saggi Ladini, landmark study that was directing in Italy to the systematic and scientific analysis of Italian , but which is considered a milestone of dialectological studies in general.2 On 26 January 1889 he was appointed a senator of the kingdom. He continued scientifically to work until 25 June 1902. He died in Milan January 21, 1907.

33 Carolina Coen Luzzatto, Graziadio Ascoli. 1860-1861, cit., p. 28. 1 About the academic activity of Ascoli see B. Terracini, La paleontologia linguistica: Ascoli, in Guida allo studio della linguistica storica, I, Profilo storico-critico, Roma, 1949; T. Bolelli, the entry “Ascoli, Graziadio Isaia” in Dizionario Biografico degli italiani, Roma, Istituto della Enciclopedia Italiana fondata da G. Treccani, 1962; S. Timpanaro, Classicismo e Illuminismo nell’Ottocento italiano. Carlo Cattaneo e G. I. Ascoli, Seconda edizione, Pisa, 1969; S. Timpanaro, Graziadio Isaia Ascoli, in “Belfagor”, XXVII, 1972; D. Santamaria, G. I. Ascoli e la linguistica italiana del primo Ottocento, cit.; D. Santamaria, Orientamenti della linguistica italiana del primo Ottocento, in The history of linguistics in Italy, edited by P. Ramat et alii, 1986; C. Dionisotti, Appunti su Ascoli, cit.; A. Brambilla, Appunti su G. I. Ascoli. Materiali per la storia di un intellettuale, Gorizia, 1996; G. Lucchini, Graziadio Ascoli: per una biografia di un intellettuale, in A. Casella, Graziadio e Moisè Ascoli: scienza, cultura e politica nell’Italia liberale, Pavia, Università di Pavia, 2002. 2 See S. Pop, La dialectologie. I. Dialectologie romane, Louvain, Duculot, 1950 ; B. E. Vidos, Manuale di linguistica romanza, Firenze, Olschki, 1959; A. Varvaro, Storia, problemi e metodi della linguistica romanza, Napoli, Liguori, 1968; C. Tagliavini, Ori- gini delle lingue neolatine, Bologna, Patron, VI ed. 1972; I. Iordan - J. Orr, Introduzione alla linguistica romanza, Torino, Einaudi, 1973; P. Benincà, Piccola storia ragionata della dialettologia italiana, Padova, Unipress, 1988.

16 III. Linguistics of early nineteenth century

III.1 General considerations

The framework of language studies, which was emerging in Italy (and in fact even in other coun- tries) in the early nineteenth century, shows a decidedly hectic laboriousness. It was, however, ini- tially oriented so poorly coordinated, toward goals that had little or nothing in common with the linguistics that it will impose itself over the course of century. The problems that caught the atten- tion of Europeans ‘protolinguisti’, and the curiosity of many, they were still represented by phi- losophical research on the origin of language, theme typically eighteenth century, and by the rela- tionships among the languages and peoples. They were then interests ranging from prehistory to the links between language and what it can be called ethnography (but not in the disciplinary sense that it will assume later and it has today). Those that I have defined as protolinguisti, in fact they were scholars of different cultures, such as writers, philologists, orientalists or more, who charmed by the new line of studies, devoted them- selves with great fervor in search of new solutions. These early decades of the century can be considered as the transition from an initial classification of historical-ethnographic nature to the specialization of linguistic discipline in the sense said ‘sci- entific’ in the nineteenth century, and that will depend decisively by the studies coming from Ger- many, to which it will make a crucial contribution just the work of the scholar Graziadio Isaia As- coli in the second half of the century. The questions related to the origin of language were debated between the divine and human origin. Among the ‘monogenesis’, that is the derivation of all languages from a single language, which perhaps - according to some - was no stranger the divine intervention, and the ‘polygenesis’, which assigned to the languages different origins. How the polygenism of the German Friedrich Schlegel who considered the ‘inflectional’ languages Indo-European like perfectly structured bodies from birth. Fact, this, that also had to imply a divine intervention. Against him had placed himself the Lombard Carlo Cattaneo with his thesis ‘agglutinant’, who wanted to remain into the sphere of the rational explanation, which did not admit they existed languages perfect from the start, attributing to all languages a purely human origin.1 What the two language versions were trying to mask in a certain way (and justifying), was nothing but the very origin of the mankind. Schlegel wanted to derive humanity from an upper phase, al- tered then gradually over time. Cattaneo instead wanted to prove that mankind was progressing from a primitive stage, which progressively evolved towards more complex civilizations. It was the example of the conflict existing between German Romanticism and the Enlightenment positiv- ist of Cattaneo. The study on the relationship between languages and nations (or races) was bound to raise contro- versy between the Enlightenment and Romantic.2 The claim of scientific progress of the first was

1 See S. Timpanaro, Classicismo e illuminismo, cit., p. 275 and foll. 2 See S. Timpanaro, Classicismo e illuminismo, cit., p. 241 and foll.

17 in clear conflict with metaphysics of the seconds. The fundamental contrast was outlined on the is- sues of so-called theories of superstrate and substrate. The German Romantics were attributing importance to Germanic superstrate of the , wanting to defend the theory of in- fluence barbaric on the greek-latin element (and typically medieval culture), to support their equal- ity toward the ‘Latin’ nations. The substrate, defended by classicists (especially French), was meaning to reevaluate the presence of the pre-Latin element, which was later incorporated by cen- turies of Roman rule. In the beginning of this century, the study of new linguistic discipline was so torn between materi- alism and spirituality. They wanted to take advantage of linguistic research to support different po- sitions which concerned the rivalry and superiority among the peoples of Europe and not just in Europe. The view was already world: the theory of polygenesis for example was able to represent a convenient springboard to support the «innate» superiority of the Indo-Europeans on blacks in the of America or on the Jews. It may be added that Ascoli also, as a boy, will have personal interest to defend the theory of secular monogenesis to validate the unity of origin of Indo-Europeans and the Semites, theory to which he will be faithful until the end. Up to half of the century, only Carlo Cattaneo will propose insistently in Italy a clear distinction between linguistic affinity and racial affinity. As for the interest in studies of the romance area, the dialect survey was developed initially as a means to access backwards to the linguistic cultural heritage of the ancient Italian peoples. In this perspective, the dialects were no longer regarded as the product of a process of corruption of Latin, according to what had been the first humanistic conception, but recovered with full dignity and scientific value. The promoter of this research in Italy was just Cattaneo, who with his brilliant insights formed the new and large group of scholars who helped to refine and disseminate the fledgling dialectological discipline, to consign later the rein to Ascoli, meanwhile well informed about new European per- spectives. The other controversy of nineteenth century, destined to dim a bit at a time, was the one that con- cerned the common origins between Latin and Germanic.3 This commonality had been lately dis- closed by German discoveries of Indo-European linguistics. This theory aroused the hostility of Italian classicists, who were seen suddenly deprived of the ancient supremacy of a noble origin sinking its ties to the and Roman civilization, considered a model unapproachable and never exceeded. This new linguistic theory was born in Germany at the beginning of the nine- teenth century and proved conclusively the kinship of Sanskrit with Greek and Latin, followed by gradually adding the Germanic, as well as the Persian, Slavic, Celtic and all the other languages that will be included in the Indo-European family. Artificers of these new discoveries were largely German scholars, such as Friedrich Schlegel, who in 1808 opened the way for a new linguistic discipline with the essay Über die Sprache und Wei- sheit der Indier (“On the language and the wisdom of the Indians”), who first had the merit of hav- ing established the concept of linguistic relationship, which included morphological traits com- mon, as well as lexical.

3 See S. Timpanaro, Classicismo e illuminismo..., cit., p. 269 and foll.

18 In 1816 Franz Bopp published the first systematic work Über das Conjugations system der San- skritsprache in Vergleichung mit jenem der griechischen, lateinischen, persischen und germani- schen Sprachen (“Comparative Grammar of the Indo-European languages”), which was intended to provide new solutions about the genealogy of languages. In 1818 (but it was written in Danish in 1814 with the title Undersögelse om det eller gamle Nord- iske Islandske sprogs Oprindelse “On the origin of the ancient Irish and Norwegian”) came out an essay of Rasmus Rask on the origins of ancient Icelandic, in which, in addition to anticipate some basic Bopp’s lines, he expressed for the first time clearly the concept of constancy in phonetic rela- tionships. The founder of the Germanic linguistics was Jacob Grimm, who in 1819 published his Deutsche Grammatik, which was the comparison of the Germanic languages and their historical develop- ment.4 Grimm later offered a model to the comparative study of Friedrich Diez on the Romance languages, already comparatively investigated by François Raynouard. The contribution of these great pioneers of linguistics of early nineteenth century was, therefore, the birth of a new scientific discipline, called Sprachwissenschaft ‘science of language’ in German and ‘linguistics’ in Italian (or ‘glottologia’, a word coined by Ascoli), which favored in the course of a very few decades, the acquisition of a number of important results ‘technical’ in the domain of linguistic evolution. These include the perfection of a method of inquiry, known as the historical- comparative, which enabled the systematic and structural comparison of languages, and which led, among other things - as already mentioned - at the identification of a linguistics family (by the la- pels sensational and - for some - disconcerting) that has an undocumented ancestor in the Indo- European (mostly in German said indogermanisch). In the next section I shall list in chronological order a detailed biography of the great pioneers of European linguistic nineteenth century, to which be followed by a brief biographical sketch of Ital- ian scholars.

4 See G. Mounin, Storia della linguistica, Dalle origini al XX secolo, Milano, Feltrinelli, 1968; V. Pisani, Le lingue indeuropee, Brescia, Paideia, 1971; L. Renzi, Nuova introduzione alla filologia romanza, Bologna, Il Mulino, 1987; A. Morpurgo Davies, La linguistica dell’Ottocento, Bologna, Il Mulino, 1996.

19 III.2 Biographical profile of earlier scholars to Ascoli

In order to consider more clearly and organic the cultural and linguistic environment that were configuring at the time of the first studies of young Ascoli, I consider it useful to expose this brief biographical sketch of foreign scholars1 and Italian that may have contributed to his historical edu- cation and linguistics.

III.2.1 Foreign scholars

François Juste Marie Raynouard (1761-1836), French, is interested in and is the first scholar, still previous the model of grammars of Bopp and Grimm, to deal with the historical grammar of Romance languages by publishing in Paris a series of six volumes of texts, titled Choix de poésies originales des troubadours (1816-21). The first volume of the series is entirely occupied by Recherches sur l’origine et la formation de la langue romane and by a Grammaire de la langue romane (1816), which is the first grammar of an ancient Romance language based on texts. The sixth volume instead is entitled Grammaire comparée des langues de l’Europe latine (1821), in which he enlarges the comparisons also to Friulian and Romanian, besides the as ones of Ferrara, Mantua, Milan and others. Raynouard bases his comparative method on a long series of characters that relate to elements of lexicon, the morphological facts and of syntax, but he gets wrong to consider the Provencal as the direct heir of Latin and the founder of all Ro- mance languages, from which he excludes however the Romanian, considered too isolated to be descended from the Roman language. His best work, however, remains the Lexique roman ou dictionnaire de la langue de l’Europe La- tine (Paris, 1838-44). Raynouard is considered the precursor of Romance Philology. His work comes out simultaneously with that of Franz Bopp, who in 1816 publishes in Frankfurt a comparative grammar of the Indo- European languages (see more under the heading Franz Bopp). Friedrich Diez and the other nine- teenth-century scholars of inspired themselves at François Raynouard.

Friedrich Schlegel (1772-1829), German, Greek literary scholar and man of letters in general, he arrives to Paris in 1802 for a period of study until 1805. Here he meets the Scot Alexander Hamil- ton, who introduces him to the study of Sanskrit and in part inspires him in the writing of the essay Über die Sprache und Weisheit der Indier of 1808, in which he argues the close linguistic affinity of Sanskrit (ancient Indian) with the Greek and Latin, Persian and German. He says indeed that these languages are derived really from Sanskrit. In addition to criteria of phonetics and historical morphology he resorts to the comparison of morphological characters, founding what will be called the modern “typology”.

1 See C. Tagliavini, Panorama di storia della linguistica, Bologna, Pàtron, 1963; Id., Origini delle lingue neolatine, cit.; G. Mou- nin, Storia della linguistica. cit.; L. Renzi, Nuova introduzione alla filologia romanza, cit.; A. Morpurgo Davies, La linguistica dell’Ottocento, cit.

20 Schlegel lays the foundation of the emerging historical-comparative method and at the same time he founded the linguistic typology.

Wilhelm von Humboldt (1767-1835), German, after a brief sojourn in Paris in 1797 and a few years later in the Basque country to learn the language, begins a long career as a diplomat in Rome, Vienna and London. He studies the classical languages and many modern languages including Chinese, Burmese, Tartar and Sanskrit in order to write a comparative anthropology, who will never realize. In 1810 he founds the University of Berlin and until 1819 he holds posi- tions as high-level politicians. While not a linguist, the language philosophy of Humboldt will ex- ert an enormous influence on his contemporaries, despite the haziness of his writings, in part common to many idealists and thinkers of the German Romantics.

Rasmus Rask (1787-1832), Danish, born in Brøndeskilde, he studies at the University of Copen- hagen without completing his studies. In 1811 he published a grammar of the ancient Icelandic and in 1817 one of the ancient English. With the wise Undersögelse om det eller gamle Nordiske Is- landske sprogs Oprindelse (“On the origin of ancient Norse and Icelandic”)2 he obtains in 1814 the award of Denmark Danish Academy of Sciences, which then forces him to embark on a long jour- ney to the Orient until to India and Ceylon (1816-1823).3 From 1823 he writes many descriptive grammars of various classical and modern languages. Also important was his contribution occa- sional (but crucial) to decipher the characters cuneiformi.4 Shortly before his death gets a chair of Oriental languages in Copenhagen. The work of Rask, probably because it was written in a language few read, remains limited to the environments in Denmark. Outside Denmark only Jacob Grimm will immediately praise him un- conditionally, but the real Rask will be known and enabled only in the late nineteenth century.

Jakob Ludwig Karl Grimm (1785-1863), German, leads for many years the task of librarian in Kassel, interspersed with a few trips to Paris, and then in Göttingen. In 1819 he issues his first vol- ume of the Deutsche Grammatik, a historical and comparative grammar of Germanic languages, which renews with various editions up to 1837, using the results of Rasmus Rask and Franz Bopp. In 1840 he is invited along with his brother Wilhelm in Berlin, in charge of academic and of uni- versity professor. In 1848 he publishes the Geschichte der deutschen Sprache, devoting the first volume to the morphology and reserving the phonetics to second volume. At Grimm is credited the first scientific treatment of the historical phonetic and the first phonetic law called Lautverschiebung, ie the consonantic shift that separates the Germanic languages from all other Indo-European ones, although in reality the concept of this law was already exposed in

2 The work, however, was published until four years later, in 1818. Given that it came out after the Bopp’s book of 1816. (Über das Conjugationssystem der Sanskritsprache in Vergleichung mit jenem der griechischen, lateinischen, persischen und germanischen Sprachen) and was written in Danish, his influence was not as wide as he could been. See A. Morpurgo Davies, La linguistica dell’Ottocento, cit., p. 182 anf foll. 3 G. Mounin, Storia della linguistica. Dalle origini al XX secolo, cit., p. 150. 4 A. Morpurgo Davies, La linguistica dell’Ottocento, cit., footnote 3 on p. 211.

21 the work of Rask so already clear enough.5 In 1854, Grimm publishes the first part of the Deutsches Wörterbuch.Within the German culture the interests of Jacob Grimm are addressed, in addition to linguistics, even to literature, to law and religion. He is considered the founder of the Germanistics.

Franz Bopp (1791-1867), German, born in Mainz, initially is interested in Oriental languages and from 1812 to 1816 he studies in Paris Persian, Arabic, Hebrew and Sanskrit, which at the time was considered to be the progenitor of some European languages. The Parisian setting contributes to the realization of his major work Über das Conjugationssystem der Sanskritsprache, in Vergleichung mit jenem der griechischen, lateinischen, persischen und germanischen Sprachen (Frankfurt 1816), which establishes him as the founder of comparative grammar. But already in 1820, he is sure that Sanskrit is one of the many related languages, and not the ancestor from which they derive. In his initial work Bopp focuses exclusively on morphology. Only Rasmus Rask, and then Jacob Grimm in 1822, will show interest also in the study of phonetics. In 1821 he was offered a chair of extraordinary professor at the University of Berlin. From here, after an extensive Sanskrit grammar, Ausfürliche Lehergebäude der Sanscrit Sprache (1827) and a glossary of Sanskrit-Latin, Glossarium sanscritum (1830), he begins to publish in 1833 a series of editions of the Vergleichende Grammatik des Sanskrit, Zend, Griechischen, Lateinischen, Lithuaischen, Gothischen und Deutschen (1833-52), in which time by time he recognises and adds Slavic in 1835, Celtic in 1838, Albanian in 1854 and Armenian in 1857. Bopp defines the forefa- ther of these languages with the Indo-European term, according to the use of English Indo- European, rather than Indo-Germanic, which was instead the use of German.6

Friederich Diez (1794-1876), German, born in Giessen (Hesse), begins as a lecturer of Romance languages and literatures at the University of Bonn in 1821, where he continues as a scholar of medieval Romance literature. Since 1830 he hold in Bonn the first chair of Germanic philology, but inspired by the model of comparative grammars of Bopp and historical of Grimm, edits a com- parative-historical grammar of the Romance languages, called Grammatik der romanischen Sprachen (1836-44), which is parallel to that of the Germanic languages of Jakob Grimm. In 1854 publishes the Etymologisches Wörterbuch der romanischen Sprachen. Diez is considered the founder of Romance philology. As much as his work depends from the in- tuition of Raynouard and from method of Schlegel and Bopp, Diez takes first, in the Romance domain, the historical-comparative method that was used in indo-European languages and the reading of which would have sent the method of working also to Ascoli.

6 See G. Mounin, Storia della linguistica, cit., p.157; V. Pisani, Le lingue indeuropee, cit., p. 27.

22 III.2.2 Italian scholars

In the chronological list of Italian scholars1 who have influenced largely on cultural education of young Ascoli I also include the historian Cesare Cantu, who despite not being a linguist, was quoted in his early essay, and the semitist Samuel David Luzzatto, who had direct contact with the young Goritian.

Carlo Ottaviano Castiglioni2 (1785-1849), born in Milan, but spent part of her childhood in Vi- enna, where he gained a good knowledge of the German language. He returned to Milan continues the study of classical and oriental languages and many modern languages such as Arabic, Persian and Turkish. In subsequent years, uses these language skills to deal with numismatic studies, publishing numer- ous essays and catalogs. Later called by Angelo Mai to the deciphering of some Ambrosian pal- impsest, primitively written in Gothic, in 1819 he published his first linguistic work with the long title Ulphilae partium ineditarum in Ambrosianis palimpsestis ab Angelo Maio repertarum speci- men coniunctis curis eiusdem Maii et Caroli Octavii Castilionaei editum, followed by the publica- tion in 1829 of Gothic text of second letter of St. Paul to the Corinthians, with Latin version, notes and glossary. In 1834 he will publish the letter of Saint Paul to the Romans and Ephesians. Until 1839 he continues his publications of the Gothic text of the Pauline letters, without however to re- ceive great praise. He died April 10, 1849 at Cornigliano near Genoa. Bernardino Biondelli will compile a list of published and unpublished writings by Castiglioni in his book «Scritti linguistici» in 1856. Domenico Santamaria will tell of him3: “Al Castiglioni che è stato quasi ignorato da studiosi e storici della linguistica spetta un posto di primo piano nell’ambito della linguistica preascoliana ... La procedura del Castiglioni risulta segnatamente moderna rispetto a quella adottata dagli al- tri studiosi italiani del tempo. Egli infatti è pienamente consapevole che la comparazione tra le lingue ... va condotta non solo a livello lessicale ma specialmente a livello morfologico. È pure consapevole che voci di lingue imparentate sono etimologicamente affini anche se si presentano diverse per forme e per significato, poiché la ricerca etimologica deve basarsi sulla applicazione delle leggi fonetiche. Di tale procedura non v’erano validi modelli in autori italiani, bensì nella linguistica europea del tempo... In definitiva [Castiglioni] si distingue marcatamente da Biondelli, Cattaneo e da tutti gli altri preascoliani, in quanto si rivela uno studioso più orientato in senso tecnico della comparazione.”

1 See S. Timpanaro, Classicismo e Illuminismo. cit.; D. Santamaria, G. I. Ascoli e la linguistica italiana del primo Ottocento, cit.; Id., Orientamenti della linguistica italiana del primo Ottocento, in The history of linguistics in Italy, edited by P. Ramat et alii, 1986; P. Benincà, Piccola storia ragionata della dialettologia italiana, cit. 2 See A. Paredi, s. v., in Dizionario Biografico degli italiani, Roma, Istituto della Enciclopedia italiana fondata da G. Treccani, 1962. 3 D. Santamaria, G. I. Ascoli e la linguistica italiana del primo Ottocento, cit., pages 244-45.

23 Francesco Cherubini4 (1789-1851), Milan, enters the age of six in the Barnabite high school of St. Alexander in Milan. In 1805 he becomes corrector at the Reale Stamperia and from 1808 to 1815 contributes to the Giornale Italiano and to other periodicals with articles almost always anonymous. Subsequently he looks after translations from English and German, thanks to which he gets various government positions and the direction of a state school. The patient collection and study of a very extensive dialectological material they fruit to him the implementation in 1814 of the Vocabolario milanese-italiano, based more on literary language that on language alive (according also to A. Manzoni, who will use it initially to draft of the Promessi Sposi) and many other dialectal vocabularies, remained manuscripts, including one on the Friulian in three volumes, also reported by Ascoli in the Saggi ladini. Also his Dialettologia italiana re- mained handwritten5, who wanted to be a general description of all Italian dialects, including Friulian, and a collection of evangelical sermons in local dialect. He died in Misson (Como), June 4, 1851.

Pietro Monti6 (1794-1856), abbot, he devotes himself to works that reveal an approach that re- flects the influence of Cattaneo. In 1844 he publishes on the «Politecnico» the article Florilegio di voci comasche, estratto da un Dizionario inedito della Diocesi Comasca and the following year publishes the Vocabolario dei dialetti della città e diocesi di Como con esempi e riscontri di lingue antiche e moderne. We know, by the premise of Cattaneo to the Florilegio, that Monti was parish priest of Brunate, an alpine village of 240 souls, and in another review on the «Politecnico» always Cattaneo reveals us that Monti had edited the Italian version of the Spanish poem El Cid. In 1856 he publishes the Saggio di Vocabolario della Gallia Cisalpina e Celtica.

Samuel David Luzzatto7 (Trieste 1800 - Padua 1865), from 1829 to his death he was professor at the Rabbinical College of Padua. He devoted himself to the study of Hebrew language and litera- ture, which he renewed through the use of modern criticism, that is scientific. Among his major works of linguistic in Italian there are Prolegomeni ad una grammatica ragionata della lingua ebraica (Padova, 1836), the Grammatica della lingua ebraica (Padova, 1853-69), the Elementi grammaticali del caldeo biblico (1865) and comments on various books of the , as well as a study of Jewish poetry in Italy during the . In the Prolegomeni, which Ascoli, boy of just seven years, appears have bought almost certainly on the advice of his teacher of Hebrew Samuel Life Lolli8, Luzzatto shows to know the comparative method, but rejects its consequences, regarded as dangerous for the principles of Israel. In fact he still believed the Hebrew as holy lan- guage, derived directly from the first language of mankind.9

4 See F. Vittori, s. v., in Dizionario Biografico degli italiani, cit. 5 G. I. Ascoli, Saggi ladini, “AGI” I, 1873, p. 477. 6 See D. Santamaria, G. I. Ascoli e la linguistica italiana del primo Ottocento, cit., p. 221. 7 See G. Hugues, Di alcuni illustri semitisti e orientalisti della Venezia Giulia, in “Studi goriziani” XXIV, 1958, p. 51 and foll.; G. Lucchini, Graziadio Ascoli: per una biografia di un intellettuale, cit., p. 5 and foll. 8 G. Lucchini, Graziadio Ascoli: per una biografia di un intellettuale, cit., p. 8. 9 G. Lucchini, Graziadio Ascoli: per una biografia di un intellettuale, cit., p. 6 and foll.

24 His translation of the Pentateuch made text for over a century. He left a rich collection of letters on Hebrew, Italian, French and Latin of considerable interest philological and religious, which testi- fies to his stature at European level in the «science of Judaism». Of Enlightenment mentality, even if believer, he put himself in contrast with the mystical tendencies of E. Benamozegh.

Carlo Cattaneo10 (1801-1869) Milanese, begins his studies at the seminary in Lecco, going then to that of Monza, but continues them in high school S. Alessandro (now high school Beccaria) in Milan. On August 19, 1824 he earns a degree in law at Pavia. Despite the professional title, he continues to be teacher in the grammar school of S. Marta in Milan, where he worked for four years. In 1835 he leaves teaching and starts an activity as free journalist, collaborating with the magazine «Annali universali di statistica» of publisher Lampato with numerous articles on various topics such as economics, history, law, the technical and natural sciences. In 1837 he publishes an essay on the Nesso della nazione e della lingua valaca coll’italiana, of which I’ll report in more detail in section IV.2.1. Twisting the activity as a scholar and publicist with that of the businessman, Cattaneo designs and manufactures at its own expense the publica- tion of the «Politecnico», magazine of which he takes the helm, continuing the themes already dealt with in the «Annali universali di statistica» and adding new ones, including linguistics. In this domain he publishes in 1841 the essay Sul principio istorico delle lingue europee, on which emphasizes the importance of study of dialects, inviting the Italian scholars to carry out adequate research on Italian dialects which should be compared with the national language, for noting down similarities and differences and get to formulate hypotheses on the language of substrate. In 1844 he publishes the essay Sulla lingua e le leggi dei Celti. Completed in 1844 the adventure of «Politecnico», Cattaneo collects in the following years the best of his production in three volumes, entitled Alcuni scritti. You can enter in recent years perhaps even the preparation of an unpublished manuscript,11 which contains interesting lexical notes on the Veneto and the Friulian and a historical overview of the two regions, from which one can infer a tacit understanding on the distribution of roles with Ber- nardino Biondelli (see biography below), formerly collaborator of the «Politecnico», who will look after the study of another family of Italian dialects with the Saggio sui dialetti Gallo-italici of 1853. Involved, at the beginning against his will, in the political events of 1848 in Milan, Cattaneo gets to a certain point the true entertainer of the armed uprising against the Austrian government. But a few days later he resignes from his role for misunderstandings with the filosabaudo team and re- sumes activity journalism. After a few months of other political commitments he retreats into vol- untary exile in Castagnola near Lugano, continuing from Canton Ticino his journalism activity and entrepreneurial.

10 See E. Sestan, s. v., in Dizionario Biografico degli italiani, cit. 11 Now edited by Francesca Agnoletti, Appunti linguistici sul friulano in un manoscritto inedito di Carlo Cattaneo, degree thesis, supervisor Laura Vanelli, Università di Padova, 2001.

25 In the elections of March 1860 he is elected deputy in the fifth college in Milan, but to Parliament of Turin he will never go, believing to be more useful to his cause with the action of journalism than with the parliamentary one. On January 21, 1861 refusing the chair offered by the scientific and literary Academy Milan, be- cause - he says - is busy to the rebirth of magazine «Politecnico», in which resumes frantically to write about everything from philosophy to economics, from transport policy to military matters, and even linguistics, history, geography and many other topics. In 1865 he also resignes from the post of professor in the school of Lugano. He dies in Lugano, February 5, 1869. The figure of Cattaneo will be located by E. Sestan with this judgment: “Due pensieri sembrano dominare la mente del Cattaneo: il rifiuto di ogni metafisica; l’idea e la fede nel pogresso”.12 In the political perspective of Cattaneo, and the United States of America are the mod- els, and he comes to hope “il giorno che l’Europa potesse, per consenso repentino, farsi tutta simile alla Svizzera, tutta simile all’America, quel giorno ch’ella si iscrivesse in fronte: Stati Uniti d’Europa”.13 For him, the federal system not only descended from the entire history of Italy, but it is the only one that can guarantee the exercise of freedom, against the prevailing trends centralist and bureaucratic. And in fact, at the end of 1859, suggests “Arrischiate, accettate, rifate a voto universale le quattro assemblee; poi fate un Congresso federale in Roma - Stati Uniti d’Italia; e avete un modello bello e grande e tutte le questioni già sciolte dall’esempio e dalla pratica di ottant’anni”14 con allusione al mo- dello statunitense. According to him a central Parliament, whom he calls Congress, only had to look af- ter to general problems, but otherwise “ogni fratello padrone in casa sua. Quando ogni fratello ha casa sua, le cognate non fanno liti”.15 Of Cattaneo linguist is essential to remember his insistence on importance of study of dialects, that he defines as the only relic of the primitive times, in order to discover the ancient races that inhab- ited the Italian peninsula. Important are his main insights in the field of languages - often later also accepted by the specialists of the subject - such as the distinction between linguistic affinity and racial affinity, the opposition to the theory of monogenism linguistic, and the theory of the sub- strate, at first fighted, but then accepted even by Ascoli mature, once become the most important Italian linguist.16

Bernardino Biondelli17 (1804-1886), from Verona, teaches mathematics, history and geography in schools Venetian. Having moved to Milan in 1839 to collaborate «Politecnico» of Carlo Catta- neo with the essay Sullo studio comparativo delle lingue e Sull’origine e lo sviluppo della lingua italiana, besides various reviews, including the Grammatica di tutte le lingue germaniche del Dott. Grimm, of 1840.

12 E. Sestan, s. v., in Dizionario biografico degli italiani, cit., p. 434. 13 C. Cattaneo, Scritti storici e geografici (edited by G. Salvemini-E. Sestan), I-IV, Firenze, 1957, II, p. 178 and foll. 14 C. Cattaneo, Epistolario (edited by R. Caddeo), I-IV, Firenze, 1949-56, III, p. 231. 15 C. Cattaneo, Epistolario, cit., III, p. 373. 16 See S. Timpanaro, Classicismo e Illuminismo nell’Ottocento italiano, cit. 17 See T. De Mauro, s. v., in Dizionario biografico degli italiani, cit.; Benincà, Piccola storia ragionata della dialettologia italiana, cit.

26 He collects various articles of Indo-European linguistics and Germanic languages, appeared on the “Polytechnic” between 1839 and 1845, under the heading Studii linguistici, that he reprints without revisions, drawing for this the rebuke of Ascoli, who accuses him of poor bibliographic update. In 1841 he publishes Sullo svolgimento delle lingue indoeuropee, which he intended as first step in a planned Atlante linguistico d’Europa, intended to remain however an unfinished project. In addition to Indo-Europeistic interest, Biondelli also deals with Italian dialectology, publishing in 1845 the essay Della linguistica applicata alla ricerca delle origini italiche and in 1846 the Studii sulle lingue furbesche. Still in 1846, probably on the wake of the invitation launched by Cat- taneo in 1841 with his «Principio istorico delle lingue europee», writes for the Encyclopedia Pomba Lingue e dialetti d’Italia, in which he divides the dialects into eight families: Carnian, Ve- netian, Gallo-Italic, Ligurian, Tuscan-Latin, Samnite-Japigic, Lucan-Siculian, Sardinian. Then analyzes one of these families, the Gallo-Italic, which includes the dialects of Lombardy, Pied- mont and Emilia, collecting numerous dialectal versions of the parable of the prodigal son, and be- gins to publish in 1853 under the title Saggio sui dialetti Gallo-italici, which can be considered his most important work. Other interests of Biondelli are archeology and linguistics of the pre-Columbian cultures of Amer- ica and the numismatics, to which devotes several publications, including the Evangelarium, epis- tolarium et lectionarium Aztecum (Milan 1858), and the Glossarium Azteco-Latinum et Latino- Aztecum (Milan, 1869). In 1859 he was appointed professor of archeology at the scientific and literary Academy Milan, being the year following fellow of G. I. Ascoli and conserving his chair until 1884. He dies in Milan on 11 July 1886.

Cesare Cantù18 (1804-1895), from Como, professor in the gymnasium of Sondrio, then Como, and finally Milan, is removed and imprisoned by the Austrian government in 1833-34 under the accusation of complicity with liberal conspirators. Exile in Turin since 1848, adheres to Neoguel- phism, and, MP in the House from 1861 to 1867, supports their thesis. His literary and histo- riographical production is enormous: more than 500 writings of varying size and varied in nature, beginning with the novel in verse Algiso (1828) with which he reaffirms his adherence to the Ro- manticism; they follow a Storia della città e della diocesi di Como (1829-31) and Lombardia nel sec. XVII (1832), which illustrates and comments on the historical facts narrated in «Promessi Sposi». In prison he begins to compose the historical novel Margherita Pusterla (1838) which meets a great popularity. In 1838 begins also the publication of his most monumental work, the Storia universale (35 volumes, 1838-1846), of which he will cure subsequently nine others edi- tions and that will have a significant influence on the young man Ascoli, who will find described and commented many linguists of eighteenth and nineteenth century.19 Of the countless other historical works we still remember only his Gli eretici d’Italia, the Storia degli Italiani (1854-56, in 6 vol.), the Storia di cento anni (1750-1850) in 1851. Between 1857 and

18 See F. Della Peruta, C. Marcora, E. Travi (edited by), Cesare Cantù nella vita italiana dell’Ottocento, Milano, 1985. 19 See I. Bonomi, Note sulla formazione e gli interessi linguistici del giovane Ascoli, in “Parallela 5, Atti del VI convegno italo- austriaco dei linguisti” edited by M. Dardano et alii, Milano, Bulzoni, 1995, p. 35.

27 1861 directs the publication of the Grande illustrazione del Lombardo-Veneto. In 1865, they issue his Storie della letteratura greca, latina e italiana.

Gabriele Rosa20 (1812-1897), Lake Iseo, he attendes the private gymnasium of Don Archetti, but after the death of his mother he abandons his studies for need to help his father in his work as a baker. He continues to study alone, however, in the few hours per day granted by the family activ- ity. Soon he begins to work in the most important magazines of the time, from the «Politecnico» to the «Rivista europea» with the article Della geografia e della storia dell’antica Italia sino alla dominazione romana (1844), in «Giornale Euganeo» of Padua with La pianura lombarda conqui- stata dai Galli (1845). In 1847 he publishes on the «Rivista europea» an article on the Vicende delle lingue in relazione alla storia dei popoli where, referring explicitly to the linguistic research of Cattaneo, contends the idea that the development of language cannot be explained on the basis of ethnic factors and racial. He also believes that the assimilation of Latin among the conquered peoples into the Roman Em- pire did not take place uniformly, but in proportion to the progress of individual cultures, which retained linguistic elements found today in the dialects, precious relics of the Roman languages. The differences between the phonetic elements, morphological and lexical of the different dialects offer, according to Rose, extremely interesting insights for the study of ancient Italian peoples. On the contrary, the similarities are not that proof of a common influence exerted later from Latin. Rosa preferably deals of historical themes, with a marked predilection for the works of the great German linguists: from Jacob Grimm to Wilhelm Humboldt. Especially they impassion him the studies on European populations in prior times to Roman rule. He is convinced that the historical events of the peoples are the result of successive layers, generated by continuous contacts with other particular reality. He focuses in particular on the analysis of elements characterizing the ori- gin and development of populations such as the Umbrians, the Celts, the Gauls. Supports the con- cept of language as a free human production, denying any innatism. In 1850 he publishes the essay Documenti storici posti nei dialetti, nei costumi, nelle tradizioni e nelle denominazioni de’ paesi intorno al lago d’Iseo. The predilection for linguistic and philological research are the basis of the friendship with the young Ascoli, met in 1852 on the occasion of his journey in the various Italian provinces in search of consensus and financial aid to support the publication of Studi orientali e linguistici. Rosa is the only one among the scholars met from Ascoli offering himself in the compilation of Studi orien- tali. It is of 1855 the work Dialetti, costumi e tradizioni delle provincie di Bergamo e di Brescia.

20 G. Rosa, Autobiografia, Brescia, 1912.

28 IV. Sources and models of Ascoli’s essay

IV.1 Historical grammars before Ascoli

Although it was chronologically possible, we will see in this section that Ascoli, at the time of his Schizzo storico filologico, did not know or did not use the comparative grammars that had begun to circulate to European universities in those years. The first studies on the Sanskrit and its compari- son with the classical languages (and then with the other European languages), were largely works of German scholars, and is due to their work if this new discipline developed quickly, meeting a broad interest, especially in Germany. The curious paradox is represented by the silence of French science in terms of comparative grammar, which lasted for more than fifty years, despite the fact that François Raynouard had an- ticipated the German Romanists in his Grammaire comparée des langues latines (1821). It is known that the French were not ignorant of comparative grammar processed by the Germans, but decidedly they refused it, at least until the Franco-Prussian War (1870), when the young Gaston Paris and Paul Meyer, followers of the ‘German science’, will be appointed to section IV of the Ecole des hautes études. This singular contradiction is explained by the fact that historical linguistics was developed thanks to the vigorous influence of Wilhelm von Humboldt and the climate of German Romanticism, which was entirely turned round to the discovery and the exaltation of the national past, against the inertia of French classicism. It is in this environment that emerges the role of the German language in its confrontation with the Sanskrit and it was from this bond is coined the indogermanic term, that wants to stress indirectly (but firmly) the unmistakable clarity of the German origins. This orientation, which sometimes ominously will characterize the folds of certain German philosophy until the twentieth century, was opposed with strict tenacity to classicist opinions (not only of the French cultural environ- ments) that supported the indisputable separation of Germanic element - considered of ‘barbaric’ origin - from the one of Greek-Latin origin. Working in this place a thorough research on the linguistic theories of the early nineteenth century that may have influenced the young Ascoli, is a difficult task, that while presenting interesting, de- cisively prevails the goals of my work. Nevertheless, it did not dispense me from a synthetic look, and at the same time complete, to the hard facts. As we shall see later, Ascoli shows really to possess some basis of Indo-European linguistic, but also to ignore the comparative method based on the comparison of morphological elements. This was stated (albeit in synthetic mode) since 1808 by the work of Friedrich Schlegel, Sprache und Weisheit der Inder, and confirmed shortly afterwards by Rasmus Rask and Franz Bopp in their co- lossal, systematic comparative grammars of Indo-European languages. The skills of indoeuropeis- tic of Ascoli could not derive from the reading of the two German scholars, much less of Rask, who had written in Danish a work that Ascoli probably would not have been able to read. The fact is that the ascolian essay remains essentially pegged to the method of lexical comparison.

29 If we restrict the field to the Romanistic, he could have access to the Grammaire comparée des langues latines (1821) by François Raynouard, which exposed the comparative method on the “main elements” of the lexicon, ie long lists of words, and on the morphological facts and of syn- tax. From this reading, however, Ascoli would also inherited the strange conception of Provencal, which Raynouard considered the ancestor of the Romance languages,1 but Ascoli never makes mention of this language nor to its singular role. You just have to remember the work of Friedrich Diez, who had begun to publish in 1836 his Grammatik der romanischen Sprachen (1836-1844), carrying on the work of Raynouard and the indications of Schlegel, but tempered by the strict les- son of Grimm (which in 1819 published the first volume of the Deutsche Grammatik, a compara- tive grammar of Germanic languages), the reading of which would have transmitted the work method also to Ascoli. However, he certainly did not know these sources, and this confidence comes from list of his library, where these works are not inserted before the 1849,2 and from a let- ter he wrote the same Ascoli July 15, 1847 to his friend Filosseno Luzzatto “my studies on philosophy of language and so on. date from 18 June last year”.3 There is every indication that the young scholar of Go- rizia has approached the design of its short essay without a clear guide ‘technique’, but in the wake of an echo of voices bouncing through the various cultural centers of Europe aimlessly and without a clear scientific construction. In addition to the comparative grammars, we must consider the possibility that Ascoli knew the first study of Cattaneo Del nesso fra la lingua valaca e l’italiana (1837), which probably he has only read after 1846, the year of the publication of his Idioma. Certainly he did not know the Paralelism între dialectele român şi italian (Bucharest 1841) of Ion Heliade Rădulescu, who claimed that Romanian and Italian, rather than two languages very similar, must be regarded as two dialects of same language,4 theses that had also circulated previously. In view of these observations, the only positive award to be granted to young Ascoli is that its con- tribution is the first comparative study between the Friulian - which was still far from possessing the dignity of language - and the Romanian.

1 Note the strong opposition of the French theories based on the Sanskrit language. 2 See par. II. 4 about the youth of Ascoli. 3 G. I. Ascoli, Raccolta delle composizioncelle fatte nelle ore di distrazione da G I.Ascoli dalla sua più tenera fanciullezza in poi, Fondo Ascoli, tomo II, p. 106. 4 C. Tagliavini, Un frammento di storia della lingua rumena nel secolo XIX. L’italianismo di Ion Heliade Rădulescu, Roma, 1926, pages 30-31.

30 IV.2 Romanian

In the prologue of his essay, Ascoli says he was impressed by the great affinity between the Friulian and the Romanian. This reason seems sufficient to his curiosity to devote himself to analysis of relations between the two languages, “attingendo alle fonti inesauribili dell’istoria”. And sincerely pleased and gratified by his research, thinks of making partecipant even the audience. With this decision we are now in a position to examine his work. But on who really he took the spark to ven- ture himself into a challenge if nothing else so unusual, especially for his time? Ascoli had the good fortune to be in close contact with eminent personalities of Gorizia, which certainly affected his interests of history and linguistic, as Isaac Samuel Reggio and Don Stefano Kocjančič (as you can see in both the Jewish and Catholic area), which have facilitated the reading of the publications from Milan - which in those years was still under Austrian rule - or from Vi- enna. The first decades of the nineteenth century were in fact the years of a rich and explosive flowering of dialectal vocabularies in the whole Italian territory, by distinguished scholars passionate about language issues (edited essentially for educational purposes), and we cannot exclude that Ascoli has never heard of. We cannot forget the first work of historical-comparative linguistics, Del nesso tra la lingua valaca e l’italiana, made by Carlo Cattaneo and published in Milan in 1837, while in the European background had appeared for few years a number of studies - mostly published in Germany - of the great pioneers of European linguistics (see chap. III), of which the young Ascoli has perhaps received at that time, some prior information.

IV.2.1 Cattaneo and Romanian

Among his earliest essays, Carlo Cattaneo wrote one, in 1830, of comparative linguistics, entitled Nesso della nazione e della lingua valaca coll’italiana, which came out in 1837 in Milan in the «Annali universali di statistica». It was republished in 1846 by the same Cattaneo in the first vol- ume of Alcuni scritti under the title Del nesso fra la lingua valaca e l’italiana.1 The text is divided into eight short chapters dealing with a discursive style (but so little organic) of history and the language of the Romanians. “Tra le nazioni che ad onta dei favori d’una prodiga natura, ad onta d’un bel cielo, di fertili campagne e di larghi fiumi giaciono da secoli sepolte in oscura e misera esistenza, è l’antichissima gente valaca”. So Cattaneo begins the first chapter and goes on to say “Eppure ella non era venuta da stato barbaro ad artificiale e comandata civiltà. Dal cuore dell’imperio romano furono condutti i coloni e apportarono seco agricoltura, arti, lettere e leggi civili tra i barbari della Dacia. La loro stirpe si perpetuò; la loro lingua assorbì le favelle indigene, e si conserva ancora; e di tutte le figlie della latina nessuna è da molti lati più simile alla madre”. He adds a little later that “invece di splendide ed eleganti opere d’ingegno i Valachi altro non produssero finora che lavori grammaticali ... Tre alfabeti e forse quattordici diversi sistemi d’ortografia dividono fra loro i pochi letterati e i pochissimi libri.” He then goes on a series of reflections on topics ranging from the history of to the Romance philology. As when he says that “Uno scrittore, osservando che il valaco per molti lati ritrae più dal latino

1 Collected in 1948 in Scritti letterari, artistici, linguistici e vari by Bertani, Firenze, Le Monnier.

31 che non il moderno italiano, disse «stimar egli che quando Dante, Boccaccio e Petrarca dai barbarismi longobardici e gallici e dalle re- liquie del latino familiare non avevano ancora composto questa nuova lingua italica tutta elegante e quasi divina, il linguaggio valaco doveva essere per tutto simile all’italiano» ... Egli è indubitato però che se la lingua dacoromana o valaca non s’accosta al latino più dell’italiano, ella s’accosta ed al latino ed all’italiano commune più di molti dialetti d’Italia, quali si parlano dagli uomini culti di molte cultissime città.” Towards the end of the chapter he switches to the story of Romanian population, saying that it de- rives from the ancient “Daci, Mesi e Geti commisti ai coloni romani” and continues with a brief description of the conquest of Trajan. But he immediately inserts an interesting etymological observation and historical on the name Valachi, which I will report in more detail in par. IV.2.3 (The term “valaco”). All these observations of Cattaneo are nowadays still accurate, and can be considered as an anticipation of the rich framework that provides us Carlo Tagliavini in his essay Origini delle lingue neolatine (Bologna 1962). In the second chapter he goes on to examine more directly the relationship between the Italian and Romanian, anticipating some thoughts on the linguistic relationship in general. From the similarity that resides only in the dictionary (we would say the lexicon), such as that between the English and the Romance languages, goes to description of the grammatical similarity, that which “sopravive anche quando tutto il dizionario della lingua si muta”, as in case of English with German. But there is - still - the affin- ity that includes both the dictionary and grammar, which is the case of the Romanian and Italian. To confirm his statement first lists a long series of lexical affinity, noting also any phonetic dis- crepancies in relation to Latin word from which derive the Romanian and Italian voices, noting perfectly for example the outcome of Latin nexus CT of “nocte, octo, pectine, luctare” in Italian “notte, otto, pettine, luttare”and in Romanian “nopte, opto, peptine, luptare”. In the third very short chapter he considers the subject of foreign loans that justifies with the en- tanglement of the overlap of peoples throughout history. In particular, the transition from the Latin domination of Rome to the one Greek of Constantinople led to the entry into Romanian numerous Graecisms as sografu ‘painter’, dascalu ‘master’, cusutoriu ‘pond’ icona ‘image’, calogeru ‘brother’, etc. In the fourth chapter he gets to the heart of the matter that concerns the possible linguistic influ- ence of the ‘barbarian’ dominations that affected Romanian territory. After a prelude of historical events that saw the rotation of Germanic peoples, without to omit sometimes rich and interesting details, he concludes that the Romanians were found to coexist in the same regions with Hungari- ans, , Turks, Gypsies and Saxons. And he immediately adds that such coexistence was based on the mutual hatred, on the suspicion and contempt, besides on diversity of languages and relig- ions. Cattaneo ends the chapter by illustrating a short list of foreign words entered in the Romanian lexicon, such as Gothic, Slavic, Turkish and Hungarian. “Anzi talvolta un solo oggetto è significato con due vocabo- li, l’uno nazionale, l’altro straniero; ... un nobile ora nobilu, ora nemisciu; invece di repa’usare (riposare), si dice anche hodignire; invece di pericolosu si dice primes’diosu; invece di patientu si dice bolnavu.” In preparing the sixth chapter finally he takes into consideration the outright morphological com- parison between the Romanian and Italian. Definitely original and acute, for the times, is the ob- servation of Cattaneo on the location of the article: “È singolare ed esclusiva proprietà del valaco che mentre l’italiano, il francese e lo spagnolo prepongono l’articolo, il valaco lo pospone ai nomi, anzi lo inserisce tra il nome e l’aggettivo.” Not only, se- eking the explanation he proves to know that “L’uso di posporre gli articoli è commune a qualche altro linguaggio, al basco, per esempio e all’islandese; ed eziandio ai linguaggi di due popoli confinanti col valaco, cioè l’albanese e il bulgarico”, proving

32 himself with this last observation as a forerunner of Balkan linguistics. He continues his reflection on the factors that may have led to the phenomenon of article put behind, to insert at the end of his reasoning the controversy on the relations of connection between the linguistic affinity and racial affinities. Cattaneo is firmly determined to prove that such a connection does not exist, saying among other things that “questa maniera di classificar le nazioni su la sfumata simiglianza d’una sola forma grammaticale è troppo ardita”, just referring to position of the article. He then expounds the rest of chapter examining gradually the unique characteristics of the Romanian personal , the confluence of the cases, the formation of plurals, the construction of the future with the “voiire” (actually voi + , from VOLO) and more. He devotes the seventh chapter to the phraseology of the Romanian, illustrating a long series of “tratti di colloquio famigliare”, accompanied by the Italian version. Finally, the eighth chapter exposes a sequence of Romanian grammars, commenting on the histori- cal path and expressing his opinion on issues he considers most relevant. He is very well informed about the grammar of physician Johann Molnar, of which he says it was printed in Vienna in 1788 and whose author “giudicò bene di scrivere ogni voce valaca con ambo gli alfabeti latino e cirillico.” He had in fact antici- pated at the beginning of chapter that the first publications about the Romanian language appeared in Venice in 1770, and were “il lessico quadrilingue di Daniele Moscopolita, e il glossario trilingue di Caballiotti; ma erano am- bidue in greca scrittura.” He added that “nel 1780 Giorgio Scinkai2 publicò in Vienna gli Elementa linguae daco-romanae di Samuele Klein, adattando al valaco le lettere latine per zelo di catolicismo, giacchè i partigiani della religione greca e riformata si fanno dovere d’adoperar l’alfabeto ci- rillico. Questo alfabeto fu inventato dai missionari Cirillo e Metodo pei cristiani Slavi delle rive del Danubio, verso la fine del secolo IX; e set- te secoli dopo fu dai Valachi riformati applicato alla loro lingua nei libri sacri.”3 Cattaneo continues by saying that Cyrillic alphabet had no less than 44 letters, to which the Roma- nians added two other, because “hanno il vantaggio di poter esprimere senza composizioni e contorsioni molte preferenze che il nostro alfabeto non raggiunge. Per esempio distinguono i due z, il ronzante e il tagliente, come già voleva il nostro Trissino; e parimente distinguono le due s; distinguono il k dal c aperto; il gh chiuso da g parimente aperto; ed indicano molti altri suoni cioè il th greco, il ch aspirato, il sci, il j francese, l’n nasale, il gni, il gli; e il suono composto sct, simile al tedesco scht. Fu in quell’alfabeto che nel 1787 il boiaro Vacarescu stampò a Rimnico le sue Osservazioni sulla lingua Valaca.” In the long list of grammars and Romanians dictionaries, Cattaneo also refers to the reprinting of grammar of Molnar published at Hermannstadt (now Sibiu) in 1810, and as we shall see it is just this edition that will be used, and cited in note, from Ascoli. Very interesting the consideration that Cattaneo expresses toward the end of chapter on the question of handwriting: “Quanto alla scrittura dei tre alfabeti, il greco è affatto insufficiente, il cirillico è troppo esclusivo e tende troppo a isolare la nazione; il latino offre l’uso più facile, più elegante, men dispendioso perchè applicabile ad altre lingue, più commodo agli stranieri, e utile ai nazionali; e basterebbe aggiungervi qualche segno in più, come gli Europei occidentali han già fatto introducendo le lettere J, K, W, ed U, perchè bastasse alla certezza dei suoni e per facile comprensione del vulgo.”

2 That is Gheorghe Şincai, representative of school of Transylvania, as well as Samuel Micu-Klein.. 3 In fact it is the Glagolitic alphabet which must be considered the true original creation of Cyril, who in some cases seems to be inspired by the Greek lower-case writing (developed in later age, before in the papyri, and then, only from the ninth century after Christ, in the manuscripts on parchment) and by other alphabets. Today we know that Cyrillic alphabet is a post-processing of Glagolitic and was based on the Greek uncial writing (greek capital letters used both in the inscriptions and in the papyri). See. N. Radovich, Introduzione allo slavo ecclesiastico antico, Istituto di Filologia Slava, Università di Padova, 1982, p. 16.

33 Considering that in effect the use of Latin alphabet was introduced in Romania in 1840, the obser- vation of Cattaneo sounds like a prophecy that would be realized soon.

IV.2.2 The Romanian Grammar of Ion Molnar

As we have already mentioned, Ascoli, in a footnote on page 15 of his essay, says to use, for the realization of his «Idioma», the Deutsch-Walachische Sprachlehre of Molnar, printed in Hermann- stadt (today Sibiu, Transylvania, now in Romania, at the time instead in Hungary) in 1810, but did not add anything else. But already from reading the chapter VIII of work of C. Cattaneo Del Nesso fra la lingua valaca e l’italiana you can find some more information on the grammar used by Ascoli. This is the work of Johann Molnar printed the first time in Vienna in 1788 and reprinted in 1810 in Hermannstadt. The real name of Molnar was Ion Piuariu, and he was a educated ophthalmologist Romanian of Cluj, naturalized hungarian. Come soon fruitless the searches of this grammar in the Italian libraries, Professor Lorenzo Renzi has found a specimen of the press of Vienna (1788) at the University Library in Cluj, Romania, giving me a photocopy of it. Although it is not the reprint of Sibiu (1810), I believe this difference should not be severe for our purposes.

The grammar of Ion Molnar Piuariu, which is very voluminous, is called Deutsch-Walachische Sprachlehre and is written in German with Gothic characters. After the first chapter, devoted, as we shall see in more detail in the next section, to phonetic aspect, continues as in a regular norma- tive grammar, developing first the morphological part, then followed by a rich lexical part the- matically organized, from conversational exercises and short stories. The Romanian text is always written in the Cyrillic alphabet1 and then in Latin one.

1 The Cyrillic alphabet was based on the Greek uncial writing (greek capital letters used both in the inscriptions and in the papyri). N. Radovich, Introduzione allo slavo ecclesiastico antico, cit., p. 16.

34 IV.2.2.1 The writing Romanian of Molnar

In the first chapter Molnar describes with great precision all the elements relevant for the proper understanding of the Cyrillic alphabet. First he lists the Cyrillic letters, flanked by their name with . Then to each letter he associates its sound, presenting a Romanian word that contains it, written in Cyrillic and Latin, followed by a German translation in Gothic characters. In the case of phonemes that do not correspond to German (eg. /Z/ e /dZ/), Molnar refers to the equivalent French or Hungarian or Latin, while also specifying the relevant examples with the words of cor- responding language written in Latin characters. In the third chapter illustrates the accents and punctuation, specifying the use, always with accom- panying examples. All Romanian examples are always presented in Cyrillic and Latin writing, fol- lowed by the German translation. Molnar describes with accurate precision the phonetic of Romanian starting by the graphic render- ing of vocalism, in which presents firstly the є, that can count [ýe] when it has got the ac- cent, or [e] if it does not have got the accent. With the term ‘accent’, I indicate a special graphic mark of Molnar, with double prime, which he places above the Cyrillic letter. In this case, it is ê with the double accent [unfortunately my digital graphics software will not let me write ê with the double prime]. With the sign ъ Molnar indicates the vowel e centralized, today corresponding to ă [.], whose ex- amples are moaрtъ/moarte (current moartă, ‘dead’). The sign ъ is transliterated with a ė, that is a e superimposed by a short vertical line. With the sign shows the current central Romanian not labialized î, â, namely [4], trans- literated with the sign æ. The example is м нa/mæna (current mîna, ‘the hand’). Compared to the Cyrillic alphabet applied to Slavic, Molnar also uses the sign ↑, used in the Ro- manian Cyrillic spelling (until 1840), to indicate the vowel î followed by nasal n. The value of this sound is therefore [4n] and the example associated is ↑дъpъπt/ėndėrėpt (current îndărăt, ‘back’). For the consonants, things get complicated a bit more. We find the sign Cyrillic Ж, that he ex- plains “wie ein gelindes [ie soothed] Sch”, followed by the example жŏдek/schudek (current judec, ‘I judge’). The absence of this phoneme in German put in difficulty Molnar, who may use more luck with the French and Hungarian. Soon after he adds really “oder wie im Französischen j, juger, jaloux, und im Ungarischen zs, mázsa”. In Latin transliteration Molnar therefore makes the sign ж with sch, to indicate the palatal spirant voiced [Z], which in modern Romanian is written j. But of course Sch indicates also the transliteration of the voiceless palatal spirant [S] which is rep- resented by the Cyrillic Щ. The example is щєZo/schesu (current şes ‘plain’), “und das Ungarische s, als sereg”. The corresponding modern spelling is ş. The Cyrillic letter ц corresponds at voiceless dental z [ts]. The example is цïe/zie (current ţie, ‘to you’) “und wie im Ungarischen das cz, als czegér”. Even here the Romanian transliteration of Molnar of the sign ц with z may give rise to easy misunderstandings, because the sign Latin z is polyvalent, since it includes both the voiceless version and the voiced. Wisely, the choice of Molnar to pro- ceed with the use of both spellings, Cyrillic and Latin, contributes to dispel all doubt. The current transliteration of Cyrillic ц is the Romanian ţ.

35 The Cyrillic letter Z is explained “wie das deutsche S” (that is voiced s) and from the example Zнk/sik (current zic, ‘I say’) “und im Ungarischen das z, zöld”. This is the voiced sibilant [z] that in modern Roma- nian is z. The corresponding voiceless of Z is represented by Cyrillic C “wie in deutschen und lateinischen Ss” (that is voiceless s, for example - we might add - in Ger. Straβe and groβ, where β is ss), with the example ctapћ/sstarea (current starea, ‘the position’), “und wie im Ungarischen sz”. The handwriting of the - less sibilant it is in Romanian the modern s. The Cyrillic letter ч is followed by the German phonetic correspondence Tsch, with the example чapa/tschara (current ceara, ‘the wax’). It is the voiceless palatal affricate [tS], that in Romanian present, as in Italian, it is c, ce. The Cyrillic letter џ indicates the voiced palatal affricate [dZ] that Molnar makes with Dsch, as in џєm/dschem (current gem, ‘I moan’), “und im Wälischen g bor e und i, z. V. gemere, ächzen, giacere, liegen”. For the modern Romanian is used, as in Italian, the sign latin g. I add below my summary schematic of the transliteration of the Cyrillic graphic system of Molnar applied to Romanian. Between round brackets, I indicate the Latin transliteration of Molnar, fol- lowed, without parentheses, by writing of current Romanian. The letters between square brackets represent the phonetic signs international IPA (International Phonetic Alphabet):

А = [a] Y,Oy,o = [u] Б = [b] Ф = [f] В = [v] Х = [χ] = (ch) Г = [g] Ц = [ts] = (z) = ţ Д = [d] Ч = [tS] = (tsch) = c Є = [e], [ýe] Щ = [S] = (sch) = ş Ж = [Z] = (sch) = j Ш = [St] = (scht) = şt Z = [z] = (s) Ъ = [.] = (ė) = ă И,I = [i] = [4] = (æ) = î, â К = [k] Ћ = [ea] = (ea) Л = [l] = [ýa] = (ia) М = [m] ю = [ýu] N = [n] Θ = [T] W,О = [o] Ψ = [ps] П = [p] ξ = [ks] Ρ = [r] ↑ = [4n] = (ėn) = în С = [s] = (Ss) = s Џ = [dZ] = (dsch) = g Т = [t]

36 IV.2.3 The term «valaco»

The title of the Ascoli’s essay refers to affinity between the Friulian idiom and the “valaca” lan- guage. Also Cattaneo a few years earlier had published an essay on the relationship between the language “valaca” and the Italian.1 Both use the term “valaco”, while we now use the term “Ro- manian”, to mean always the language spoken in Romania. The first attempt at an explanation of the term ‘valaco’ we find it right in the first chapter of the Cattaneo’s essay quoted above. The name Valachi “vien loro dato dagli stranieri; e sembra il nome generale con cui le nazioni slave dinotavano gli abitanti dell’imperio [romano], simile a quello ch’era dato loro dalle genti gotiche, che ignare della conquista romana chiamarono Valli, Valloni, Velsci o Velschi, dapprima i Galli o Celti, poi anche li Italiani che li avevano conquistati.” Then he says that the name Velsci still survives in “Britannia, dove vien dato ai montanari Cambri, che noi diciamo Gallesi ... e dai Germani in generale vien dato agli Italiani e Francesi. Si dice che il nome Vlah presso i Dalmatini significhi ad un tempo pastore e valaco; e quindi alcuni eruditi vogliono che quel nome sia loro venuto dalla vita pastorale ed errante che conducono in Illiria; se non che potrebbe darsi viceversa che alla professione stessa fosse venuto il nome del popolo che la esercita. Secondo alcuni, anche gli Albanesi chiamano i Valachi ciubani, che suonerebbe pastori.” A modern contribution to the explanation of term ‘valaco’ we can read in the Origini delle lingue neolatine of Carlo Tagliavini,2 which so completes the informations of Cattaneo. He says that the Germanic peoples used the ethnic wal(a)hisc to denote the descendants of latinofone populations (or, as he himself points out, the ‘Romanici’ of and Italy) and Walcho-lant to indicate Italy. In modern German welsch has the meaning of ‘Roman, Italian’, but used with a de- rogatory sense. The Dutch waals has kept the same principle to distinguish the Latin ethnic from Germanic and namely the Walloons of by the Flemish. And he then continues “ma la storia di questo nome non è finita. Dal Germanico, Walha passa nello slavo *valh; nello Slavo ecclesiastico il termine vlah vale «Vlachus, generatim homo romanae originis», ma con particolare riguardo si riferisce ai Romani dell’Oriente, coi quali gli Slavi ebbero maggiori contatti, e cioè i Rumeni (Valacchi; non si dimentichi che una parte della Rumania, e cioè uno dei due principati che riunendosi formarono la Rumania, si chiama Valacchia nella tradizione storica e geografica, anche se il nome indigeno è Ţară Rumânească). Nel Russo voloh’ vale specialmente «valacco», ma nella letteratura antica (per es. presso il cronista Nestore) è usato nel senso di «italiano»; nel Polacco Włoch è fino ad oggi l’unica denominazione usata per «italiano» (e quindi Włochy «Italia»), come nello Sloveno lah «italiano». Gli Ungheresi prendono la parola dagli Slavi e dal singolare vlah fanno il loro olàh «rumeno», dal plurale vlasi traggono olasz che designa «italiano» (e di qui: Olaszorszàg «Italia»).” Today, the term ‘Valacco’ is no longer used either in Romanian or in other languages, replaced by native term ‘Romanian’, the use of which was imposed only in eighteenth century. We see, how- ever, that in the nineteenth century ‘valac(c)o’ was very still used, and this is an indication of the fact that knowledge of Romanians in the West was generally mediated by Hungary.

1 C. Cattaneo, Del nesso tra la lingua valaca e l’italiana, cit. See par. IV. 2.1: Cattaneo e il rumeno. 2 C. Tagliavini, Origini delle lingue neolatine, cit., page 124.

37 IV.3 Friulian

The Schizzo storico filologico of Ascoli was preceded in 1837 in Milan by the work of Cattaneo Del nesso fra la lingua valaca e l’italiana, which dealt with a historical-linguistic comparison be- tween Italian and Romanian.1 The young linguist Goritian would only read it in November 1849.2 as is clear from his perusals of Cattaneo’s works. Despite this fact and despite the fact that Ascoli has never quoted the essay of Cattaneo in his Schizzo storico filologico, we observe many points in common between the two works: Cattaneo cites the two editions of the Grammar German- Romanian of Johann Molnar, ie to that one Vienna in 1788 and the reissue of Sibiu in 1810, the latter also used by Ascoli for his comparative work. Secondly both Cattaneo and Ascoli note the Cyrillic alphabet best meets the needs of Romanian phonetics. Both of them finally compare not only single terms but also whole sentences. As a result of these observations remains lawful the suspect that Cattaneo has been for the young Ascoli a valid model,3 from which, however, the Goritian will operate a customized version, com- paring the Romanian with the Friulian, in place of the Italian. We can ask ourselves now, given that young Graziadio was not yet a formed linguist, why he chose to realize just such a essay of comparative linguistics, rather than a dialect vocabulary or anything else. The abbot Jacopo Pirona, to whom Ascoli dedicated his short work, had he already begun writing his Friulian vocabulary? It is likely yes. Why then compare Friulian with Romanian and not with French or Spanish? Probably, I think, to emulate Cattaneo.4 But in this case also would have been able to conserve the Italian as a term of comparison (to stay in the model of Cat- taneo) and place it in analysis with French or Spanish or Portuguese, or (why not) with the same Friulian.5 Ultimately Cattaneo had called into question the Italian and not the Milanese. He had de- liberately chosen the Tuscan, language universally recognized literary, nominated by many schol- ars and Italian patriots to represent the common language of a long-awaited united Italy, free and independent (republican, monarchical or federal it was). Before proceeding further with the examination of essay of Ascoli, I believe interesting to known more strongly even the manuscript comparative of Cattaneo between Italian and Friulian, recently unearthed and analyzed with competence by Francesca Agnoletti.

1 According to the custom of the time said “valaco”, see par. IV.2.3. 2 See B. Terracini, G. I. Ascoli, direttore dell’«Archivio» (dal carteggio Ascoli-Salvioni), in “AGI”, LII, 1967, p. 10; G. Lucchini, G. I. Ascoli, per una biografia di un intellettuale, cit., p. 9 and foll. 3 But not all scholars agree on this point. 4 See note n. 3. 5 Ascoli could not have known the work of Cattaneo which concerned the comparison between the Italian and Friulian, for the sim- ple reason that it was never published. See next par.

38 IV.3.1 Cattaneo and Friulian

When in 1873 Ascoli will publish in Milan the Saggi ladini, he shall appoint, in the chapter on the Friulian territories, two scholars that before him dealt with Friuli, with these words: “In Lombardia, il Cherubini e il Cattaneo si fermarono con predilezione a questo dialetto, e in ispecie il primo, che sen era pur formato un vocabolario abbas- tanza copioso”.1 In a footnote, then we find the cataloging, at the Biblioteca Ambrosiana in Milan, of the Cherubini’s manuscript to whom he referenced. He gave no reference instead for contribution of Cattaneo about the Friulian. Today we know the work of Cattaneo was an unpublished manu- script and his papers had not yet been collected and collated, because the Milanese scholar was re- cently passed away (in 1869). Ascoli says only later, in 1873, that Cattaneo had dealt with Friulian, but we cannot know to what extent and in which year he came to know the author’s work in Milan. It is to be excluded in any case that he had read it before 1846, when it went out to Udine his essay «Sull’idioma friulano», just because the manuscript was unpublished. It will be discussed and analyzed for the first time only in 2001, by Francesca Agnoletti.2 The Cattaneo’s manuscript is not dated, but can be placed in that time that sees the large produc- tion of works about dialectology from the circle of scholars who were collaborating to «Politec- nico». And then between 1830 and 1850. Together with his previous work, Del nesso fra la lingua valaca e l’italiana (1837), the manuscript on Friulian is the clear intention of the Milanese author to carry out the research on Italian dialects and on the Romance languages. Indeed he had insis- tently urged to this research even his colleagues and other researchers, in a perspective that com- bined historical research and dialectological with that on the ancient people who inhabited the Ital- ian peninsula. As stated Francesca Agnoletti in his thesis, Cattaneo had based his research on the poems in Friulian of conte Ermes di Colloredo.3 It was a collection of poems (Poesie scelte), edited by Pietro Zorutti,4 published in Udine in 1828. It was the text of literary Friulian of the seventeenth century, therefore, the highest level of Friulian spoken three centuries before, which it certainly could not adequately represent aspects of popular language. Maybe Cattaneo had not at hand the most suitable material, but on this text, however, he did all his analytical work, by organizing it into phonetic and morphological sections and placing in direct comparison the Friulian with the Italian, with only occasional references to Latin.

1 G. I. Ascoli, Saggi ladini, “AGI” I, 1873, p. 477. 2 F. Agnoletti, Appunti linguistici sul friulano in un manoscritto inedito di Carlo Cattaneo, cit. 3 Ermes di Colloredo, man of weapons, poet, literatus (Colloredo di Montalbano 03/28/1622 - Gorizzo, Codroipo 09/21/1692). Brother of Giovanni Battista. He lived at the courts of Florence and Vienna, he fought in Germany and in Dalmatia, and finally re- tired to Gorizzo where he composed verses among the most famous Friulian literature, of which is considered to be the initiator. [from Dizionario biografico friulano, 2002] 4 Pietro Zorutti, officer, poet (Lonzano, Dolegna del Collio 27/12/1792 - 23/02/1867 Udine). He experienced great popularity by publishing for many years the “strolics”, popular poetic almanacs. He was praised by Tommaseo and Carducci. For the theater he wrote Il trovatore Antonio Tamburo. Among his compositions: Il Bon Pari, written for a collection of poems in all languages of the Austro-Hungarian empire, in honor of Francis Joseph. In 1990, edited by R. Pellegrini it was reprinted in Udine the complete work, already published by B. Chiurlo in 1911. It has been partially translated in Italian, English and Slovenian. [from Dizionario biogra- fico friulano, 2002]

39 Cattaneo identified primly the phonetic features that separate Friulian from other northern dialects, such as sigmatic plurals (“le inflessioni plurali in s come pastorelis”), the palatalization of CA and GA (“il can- giamento dell’iniziale ca e ga in cia e gia, come ciaval, gialina”), the preservation of Latin au and of the nexus Consonant + L (“la conservazione della forma latina in flor e clar, aur etc.”), the outcome of the initial voiced velar [g ] in [w] (“la soppressione del g iniziale in guerra e guadagno, uera e uadagn”) and the diphton- gment of O in [wa], [w] [wi] (“la distorsione dell’o nei dittonghi ua, ue ed ui come muart, limuesine, puint per morte, limo- sina, ponte”). If, therefore, we evaluate the approach of Cattaneo towards the ‘dialect’ of Friuli, watching it in relation to the times and to the limited linguistic tools that the of his time gave him - as you would expect from a period of pioneering studies - and if we consider that the Milan- ese student probably did not know directly the Friulian, we cannot but admire the excellent results which the brilliant Cattaneo had got in the linguistic field as in other domains.

40 V. Analysis of «Schizzo storico filologico» of Graziadio Isaia Ascoli

V. 1. Previous reviwes

The essay of the young Ascoli had already been the subject of attention from scholars soon after its appearance. Just off the Vendrame typography of Udine, the short pamphlet was sent by young Graziadio to his master Samuel David Luzzatto,1 professor of Hebrew at the rabbinical college of Padua. He re- plied with a long letter on 8 November 1846,2 in which did not spare some criticism to boy, that was able to mitigate, however, with a benevolent sense of understanding, certainly for not tear apart in the bud the momentum of youth. As I think there are good reasons to know more about the contents of this letter, I intend to trace further detailed examination. A short but interesting consideration of this first work of Ascoli was also advanced by Carlo Salvioni, important dialectologist of Ticino,3 in an article published in the journal «Memorie storiche forogiuliesi» in 1907 with the title G. I. Ascoli e il dialetto friulano.4

V.1.1. The letter of Samuel David Luzzatto

In response to young Graziadio Ascoli, who had sent him a copy of his first essay of linguistics in order to obtain judgment, the Hebraist of Trieste Samuel David Luzzatto addresses him calling “carissimo” and giving him always the more polite ‘voi’ form. After a brief introduction laudatory, he transmits to him the first important advice: “uopo non è arrestar- si al primo aspetto delle cose; non contentarsi, non appagarsi d’un’apparenza, che colpisce; e, sopra tutto, uopo è non chiudere un oc- chio e passare avanti sulle difficoltà, sulle contraddizioni, che si presentano opposte alle nostre opinioni, le quali, sinchè non siano ben bene eliminate, non potremo esser sicuri d’avere scoperto alcuna parte di vero. Voi, mio caro, avete raccolte le parti omogenee del friulano e del valaco, ed avete chiusi gli occhi sulle parti eterogenee, che non sono nè meno numerose, nè meno gravi ed importanti.” The latter statement is followed by a Luzzatto’s list of morphological phenomena and phonetic of Romanian which are not reflected in Friulian, such as the postponement of article, that he supplies with examples, and the future of , “composto (alla tedesca) da un verbo ausiliario; non è esso nel valaco una dif- ferenza essenzialissima?” to which follows an extensive list of examples. He then returns to comment the common phenomena to the Romanian and the Friulian “egualmente che al latino ed al greco, a cui si accostano anche i tedeschi ein, zwei, drei...” referring to similarity of the numbers.

1 For the biographical synthesis of S. D. Luzzatto I send back to par. III.2.2. 2 S. D. Luzzatto, Epistolario italiano francese latino, cit. 3 Carlo Salvioni (Bellinzona 1858 - Milano 1920), founder of the Vocabolario dei dialetti della Svizzera italiana, author of a fun- damental Fonetica del dialetto moderno della città di Milano (Torino, 1884), he is one of the largest Italian linguists. See P. Benin- cà, Piccola storia ragionata della dialettologia italiana, cit. 4 C. Salvioni, G. I. Ascoli e il dialetto friulano, in “Memorie storiche forogiuliesi”, III, 1907, p. 116 and foll.

41 In a few lines Luzzatto proves to Ascoli to know not only Latin, Greek and German, which cer- tainly were part of his professional knowledge, but also effortlessly master the morphological structure of Romanian, that he knew from the Grammar of Şinkai (see more below) and of the Friulian. If we consider that these two languages should not be a normal object of study of a com- mon Hebraist and that the same European scholars of his time had opened themselves only by few years to the study of Romance language, we must agree that Ascoli had got the great luck to stand in contact with a scholar really extraordinary.1 Always from the few lines above we can also deduce that Luzzatto was condensing in short stretches the recent achievements of Indoeuropean linguistics, by bringing back to a common fund Latin, Greek and German both the Romanian and Friulian, to which they were referring. Many other acute observations go to credit of Luzzatto, among which I would notice that about numbers “11, 12, 13 ecc. non già in una parola composta, come fa il latino co’ suoi derivati, ma con una preposizione tra l’unità a la decina: un spre zece, doi spre zece, trei spre zece ecc. Questa è una stranissima originalità del valaco.” 2 If Ascoli had listed at the end of his work a very long list of Romanian words and Friulian, in his view, similar to each other, Luzzatto instead lists a series of 72 Romanian words unknown to Friu- lian, including haina ‘dress’, mare ‘great’, noroc ‘luck’, spuza ‘spark’, barbat ‘husband’ and so on, gotten, he says, “da una piccola grammatichetta di Giorgio Sinkai (Buda 1805)” e continua “Voi ne addu- cete molti di più, i quali però sono per la maggior parte comuni al valaco ed al friulano col latino e suoi dialetti, i quali quindi null’altro provano se non che il valaco è un dialetto romano, come tanti altri.” The concept of Luzzatto here is very clear: he leads the Romanian directly to Latin. He goes on to reprimand Graziadio, saying “a qual pro adunare gli Om, gli an, i vin, i kapon, gli urs e le scroafe, e tanti altri simili vocaboli, che non sono niente più valachi e friulani, che latini, italiani e francesi? Arrestiamoci dunque ai pochi, che sono, o sembrano a noi essere, esclusivi a quei due dialetti.” It is evident here the reference to the bottom Romance that Ascoli instead had ignored. Luzzatto then moves a critique toward Ascoli about sorèli, strongly denying the postposition of ar- ticle in this form, finding “più semplice il credere che da sole siasi fatto soleli... come in francese soleil, e che la delicatezza delle orecchie e delle bocche friulane abbia voluto evitare la cacofonia di quel leli, e l’abbia trasformato in reli.” We will see in my commentary on the Ascoli’s work [XV] (whose I reference), that hypothesis of Luzzatto is very close to reality. The letter of the scholar of Trieste continued with a long series of observations of linguistic nature, which I intend to resume from time to time in the course of paragraph V.5.3: Analysis of the lin- guistic part.

1 For the biographical synthesis of S. D. Luzzatto see par. III.2.2. 2 The formation of numbers from 11 to 19 with the preposition ‘on’ (in Romanian spre

42 Luzzatto starts at the end of his letter by sending to Graziadio other helpful hints, of linguistic na- ture and of psychological level, encouraging him to continue his linguistic studies, but also he in- serts a historical-linguistic observation that clashes sharply with the description of theories he had anticipated before and that I must mention: “La lingua italiana fu sinora studiata per lo più pedantescamente. La lingua della Crusca non può essere debitamente illustrata senza il confronto degli altri dialetti italiani, e questi non possono illustrarsi senza il confronto delle lingue dei barbari, germani e slavi, che diedero origine alle nuove lingue latine.” If by ‘Latin lan- guages’ he means the Romance languages, the concept of Luzzatto is unequivocal: the Romance languages are generated from Latin with the help of languages of the Germanic and Slavic super- strate. And this is in stark contrast to what he had said previously, when he was affirming that Romanian and Friulian were nothing more than Romans dialects, like so many others.

V.1.2 The article of Carlo Salvioni

Under the title G. I. Ascoli e il dialetto friulano the linguist Carlo Salvioni was publishing in 1907 (the year of death of Ascoli) a short article inside the magazine «Memorie storiche forogiuliesi» on the contributions of scholar by Gorizia to Friulian language. In his brief overview of Ascoli’s ca- reer, Salvioni examines the first work of young Ascoli Sull’idioma friulano e sulla sua affinità colla lingua valaca. Schizzo storico-filologico (1846), premising immediately that the little work was repudiated by his author himself on numerous occasions. Salvioni instead follows a discursive line very lenient towards the youth work of Graziadio, noting that “le osservazioni dell’Ascoli sono prudenti e non isconfinano dal buonsenso: merito non lieve, se si tien conto della età dell’autore, che ancora non s’era accostato nè al Diez nè agli altri luminari della linguistica tedesca”. Salvioni proves to have carefully read the “scritterello” of Graziadio, because he reveals with preci- sion some phenomena treated by Ascoli, such as presence of in the Friulian muart ‘dead’ and in Romanian moartă ‘death’, of the z in the Romanian plural talenzi and of the Friulian plural talenz, however attributing them to “coincidenze fortuite”, explaining, for example, with regard to this fact of z, that “nel primo dei quali il z è dovuto all’intacco da parte dell’i, nel secondo alla stretta aderenza di -t-s [talent-s].” He defines “curiosa e strana assai la ricerca dell’articolo posposto nel -li del friulano soréli sole” and “un’illazione d’altra natura, ma non meno strana” what makes him recognize the Germanic influence in -t of avut and in -d of dur- mind. The article of Salvioni then examines the activity of the scientific maturity of Ascoli, with no re- turn on our work.

43 V.2 Description of the book

Of the printed edition of Ascoli’s essay I found a copy in the Biblioteca Civica «Vincenzo Joppi» in Udine, marked with placement Misc. SFF 51.6.1 It consists of 35 pages including the cover, which is repeated on page 3, but without frame and with addition of extremes of the press: Udine, Tipogr. Vendrame 1846. Continued on page 5 the dedication “ALL’ILLUSTRE FILOLOGO E DELLA PATRIA STORIA BENEMERITO AB. JACOPO PIRONA R. CENSORE PROVINC. PROFESSORE E BIBLIOTECARIO NEL R. LICEO SOCIO ORD. E VICE-PRES. DELL’ACCAD. DI UDINE CORRISPONDENTE DELL’I. R. SOCIETÀ AGRONOMICA DI GORIZIA. DELL’ATENEO DI BASSANO. DEI CONCORDI DI BOVOLENTA. DELLA PONTIFICIA ROMANA ACCADEMIA DI ARCHE- OLOGIA. E DELL’ALTRA DI RELIGIONE CATTOLICA. DELL’I. R. SOCIETÀ ARETINA DI SCIENZE LETTERE ED ARTI. DELL’I. R. ATENEO ITALIANO. EC. IN SEGNO DI VENERAZIONE QUESTO SUO PICCOLO SAGGIO DEDICARE OSAVA G. I. A.” On page 7 a brief prologue precedes the text titled IDEA. which I reproduce in full in the next sec- tion (along with the historical part). The text is divided into two parts, which do not have the title: the first, carried out in 6 pages, deals with the history of Romanian people, while the second part, which expatiates for 18 pages, is about the affinity of the Friulian idiom with the Romanian lan- guage, central theme of the work. It closes the publication an epilogue of 3 pages. The inside has an elegant frieze declaring the price of the book: an Austrian lira.

1 I cannot help but thank prof. Piera Rizzolatti, University of Udine, for her valuable help in the search for the printed volume of G. I. Ascoli in the Biblioteca Civica “Joppi” of Udine.

44 V.3 Historical premise of Ascoli

I reproduce below the full text of essay by Ascoli.

IDEA.

Non è già che io voglia imprendere a far pompa di nuove fonti istoriche per vie meglio dilucidare le emigrazioni dei popoli, nè ch’io cerchi trarre dalla mia scopertuccia (credo così poter nomare il mio saggio), nuovi lumi alla bella ed utile scienza della filologia, mentre qui alla soglia di questa mia pic- cola fabbrica, m’affretto a dichiarare essere stato il mio, un lavoro di piacere e di ricreazione a cui mi trassero i miei, finora, piccoli studj filologici. La grande affinità fra l’idioma friulano ed il valaco mi colpì; volli spiegarmene il motivo, attingendo alle fonti inesauribili dell’istoria, m’appagai, e decisi partecipare al pubblico questo mio ritrovato. Un piccolo sorriso di compiacenza di qualche dotto filo- logo, se mai vi getterà uno sguardo, un piccolo incitamento agli abitanti dell’ameno Friuli, a coltivare e ad illustrare il loro bello e dolce idioma, la soddisfazione della curiosità di molti che cercarono l’origine o le parentele del friulano con altri idiomi, ed oltre a tutto il compatimento del benigno let- tore, ecco ciò che, ottenuto col suo piccolo parto, appagherà il poco più che trilustre

AUTORE

Gorizia, Luglio 1846

45 Domiziano, le di cui scelleratezze, confrontate alle pie azioni del predecessore Tito, più orrenda mo- stra di sè far doveano, aggiungendo al cuore il più crudo, una sfrontatezza senza pari, entrò trionfan- te in Roma, (a. 90 d. C.) qual debellatore dei Daci, da cui all’opposto vergognosa pace coll’oro aveva ottenuto. Eletto però nell’anno 98 Trajano a salire sul trono imperiale, siccome colui che accoppiava all’amore dell’equità e della giustizia, sommo valore, nutriva ognora il pensiero di vendicare collo sconfiggere i Daci, la poco onorevol tregua che Domiziano da loro comperata aveva. Contro a questo popolo bellicoso e sprezzatore della vita mosse dunque l’imperatore Trajano l’anno 102 di Cristo. Offrì battaglia a Decebalo re dei Daci, e totalmente lo sconfisse (105). Ma Decebalo riavuto dal colpo terribi- le, risorse più forte l’anno appresso: Trajano fabbrica il celebre ponte, passa il Danubio, fa prodigi di valore, e ridotta la capitale a rendersi, riduce la Dacia a provincia romana, vi pone numerosa colonia di soldati, che v’apportano lingua e costumi romani, ed uniti agli indigeni rimasti, formano il popolo Valaco, che tuttora con orgoglio ti vanta la sua origine. Qual lingua parlavano questi Daci? Difficile oltremodo e quasi impossibile sarebbe il diffinirlo, mentre a fronte degli ostinati studj di moderni filologi, non si può penetrare nella misteriosa complicazione, nè linguistica, nè politica dei popoli che, o dall’oriente*), o dal settentrione onde traessero origine, calarono a distruzione del colosso romano. E così per esempio trovasi

*) Molti sommi moderni sono di questa prima opinione, che io pure debolmente ammetterei, e ciò dalla parte filologica particolarmente attestar si dovrebbe come di fatto par che succeda. Un bell’esempio si potrebbe darne nella voce tanto comune ape, acqua, valaca, in friulano aghe, che forse più s’avvicina all’ape valaco che all’aqua latino; or chi dubite- rebbe che nelle irruzioni dei popoli l’ape valaco non sia portato da chi conosceva l’ap sanscrito e l’ab persiano che pa- rimenti significava acqua? Altri molti esempj io potrei recarne, ma non è qui il luogo d’allungarsi su questo argomen- to. in istorici antichi il nome di Daci applicato ad una gente speciale, mentre lo si ha pure per nome di tutta l’immensa nazione, che probabilmente dalla Persia e dal resto dell’Oriente, minacciosa sorten- do, innondò tutt’Europa, e diede origine al generico nome deutsch**). Però

**) Cantù, storia universale. probabilmente la lingua di questi Daci sarà stata quella che diede origine alle germaniche, ed in ogni modo, attaccati restando alla storia, pare che l’imperatore Trajano abbia col suo sistema di coloniz- zazione estirpato quasi del tutto e lingua e costumi dei precedenti abitatori dei paesi corrispondenti all’odierna Valachia. Fu appena nel 250 che i Goti invasero questi paesi ed abbenchè i Romani, con va- rio successo (270) contro a loro si volgessero, pure alla fine Aureliano imperadore costretto si trovò di lasciarli in libero possesso di quella provincia, che Trajano all’impero aveva conquistata. A seconda poi dell’energia o della debolezza degli imperadori più tardi, o restavano nella provincia a loro assegnata, o più arditi, scorrerie faceano sul territorio romano. Furono finalmente costretti ad indietreggiare ai tempi di Costantino il grande. Ma durante l’imperio di Valente (376) dai fieri Unni re-

46 spinti dalle loro contrade, furono costretti a ricadere sui romani dominj, che già d’ogni parte vacilla- vano. Ed ecco appunto che questi Goti comparvero di nuovo sul danubio nella posizione precisa ove oggi giace la Valachia. Ecco appunto che questi Goti stanziati per qualche tempo in quei paesi furon quegli stessi, che, tentato di volgersi verso Costantinopoli e respinti, s’addirizzarono verso l’Adriatico ed agognarono alla bella Italia. Teodosio se li fece amici, ma alla morte di quel grande, per la fatale divisione dell’impero, toccato l’Occidente al debole Onorio (396), i Goti tanto si fecero arditi, che sotto la condotta d’Alarico, benchè battuti più volte, finalmente la spuntarono, e diedero al mondo lo spet- tacolo d’un re Goto che fece tremare il superbo Campidoglio. Ed Attila (449) quel tremendo gigante di guerra, prima s’arrestò nei paesi che corrispondono all’odierna Valachia, e qualche anno più tardi flagellò i paesi vicini all’Adriatico. Tuttavia passeggiero fu questo turbine, e l’irruzione degli Unni non potè lasciare traccia nei costumi e nella lingua dei popoli. Furono poi al certo quegli stessi Goti prima nominati, che da ogni parte (e dai paesi al sud del Danubio particolarmente) concor- sero all’invito di Teodorico, che a Zenone prometteva riscattare Italia (489), e che batterono, disfece- ro totalmente Odoacre alle sponde dell’Isonzo presso Aquileja. Nelle terre vicine, con Teodorico che poscia divenne signore di tutta Italia, questi stanziarono. Erano insomma i paesi al sud del Danubio, che servivano da punto di riposo alle barbare torme che sconquassarono particolarmente Italia. Per- ché precisamente nella Valachia, dal nord della Germania provegnenti, s’arrestarono (540) i Longo- bardi per prender fiato all’irruzione nell’Italia, giardino del mondo, di cui volle la benigna providen- za, che tanti popoli da lei creati godessero! Ed il primo fondamento della grandezza longobarda in Ita- lia si fu il ducato del Friuli, sottoposto ad Alboino, provegnente direttamente dai paesi valachi, al ni- pote Gisulfo. Per non andare insomma troppo a lungo con queste piccole riflessioni storiche, diremo, che se prima e dopo di questi tempi fino ai nostri giorni, le terre ove si parla la lingua valaca, ebbero a soffrire invasioni di Slavi, d’Ungheri, di Tedeschi, spesso pure il Friuli vide calpestar le sue belle contrade da Slavi, Ungheri e Tedeschi, che vi potrebbero nella lingua aver lasciato le stesse traccie che nella Valachia. Se adunque i paesi dove tuttora si parla la lingua valaca, furono ai tempi di Trajano ridotti a coltura romana, come contemporaneamente quelli formanti l’odierno Friuli lo erano; se le stesse barbare ir- ruzioni, infestarono queste due contrade, chiaramente scorger potrassi che le lingue della Valachia e del Friuli, nella loro più importante parte, esser dovranno una composizione della romana, mista agli idiomi delle stesse barbare famiglie*) cui queste orde appartenevano, nè più stupore recare ci dovrà se immensa somiglianza fra questa e quella troveremo

*) Avanzi di lingue germaniche potrebbero scorgersi in ambo gl’idiomi, oltrecchè in molte voci, nel terminare per e- sempio i participj in t come avut, valaco avuto, friulano vut, gehabt dei tedeschi, ed i gerundij in una d tronca in fine: durmind dei friulani, dormind dei valachi, schlafend dei tedeschi.

Che se taluno argomentare volesse, che allora questa corruzione di lingua romana avrebbe dovuto passare in altre provincie d’Italia e della Lombardia particolarmente, soddisfacente risposta a tale obbiezione sarebbe l’osservazione che il Friuli, spesso staccato dal rimanente dell’Italia, fu quella

47 provincia che meno prendendo parte alle rivoluzioni delle altre d’Italia, ed aggrevata essendo pure in molti tempi alla Germania, conservò più intatte le impressioni cagionatele particolarmente dalle germaniche irruzioni, e forma, al dire del celebre Leo, siccome il Piemonte, passaggio dalla Francia all’Italia, egli, passaggio dalla Germania all’Italia. Che se poi in quanto ai costumi, coll’accomunarsi questa provincia alle sorti del Veneto e del Lombardo, questa idea germanica quasi intieramente spa- rì, e l’ameno Friuli rappresenta oggi vera parte d’Italia: non della lingua necessariamente lo stesso doveva succedere, come da questa comparazione col valaco vedrassi non essere succeduto. Non sarà già però che io nieghi, che del francese, del provenzale, e molto dello spagnolo, esista nella lingua friulana, e ch’io la voglia precisamente sorella della valaca; io non volli che mostrare che le affinità, però molto grandi, che colla valaca tiene, confermandole storicamente; nè credo sarebbe impossibile, per chi volesse spiegare le somiglianze che con altre lingue porta, parimenti dal vario giro delle emigrazioni dei popoli, provarle.

48 V.4 Comment on the historical premise of Ascoli

Ascoli opens his essay with a historical introduction, stating the facts that led to the Roman con- quest of Dacia. Domitian, who succeeded Titus, entered “trionfante in Roma (a. 90 d.C.) qual debellatore dei Daci, da cui all’opposto [di Tito] vergognosa pace coll’oro aveva ottenuto”. But it was Trajan, the successor of the “buon Nerva”, which in 105 finally defeated Decebalus, king of Dacians, in battle. He so wanted avenge, according to Ascoli, “la poco onorevol tregua che Domiziano da loro comperata aveva”. Trajan therefore reduced the Dacia as a Roman province, there he installs his legions “che v’apportano lingua e costumi romani, ed uniti agli indigeni rimasti, formano il popolo valaco, che tuttora con orgoglio ti vanta la sua romana origine”. At this point Ascoli wonders what language they spoke these Dacians, but without specifying when: before or after the romanization? while nothing is known, despite the “ostinati studj di moderni filologi” (we do not know to which he refers), of those peoples of East, or of North - he says - who contributed to the fall of Roman Empire. In the footnote Ascoli adds to share the opinion of those scholars (which?) that support the oriental origin of these peoples. And proposes the example of the Romanian term “ape” (ie, apă ‘water’) that may be linked to Sanskrit ap and Persian ab, which always mean ‘water’ rather than Latin aqua.1 The observation of Ascoli shows he was already familiar with concept of Indo-European and he had some knowledge about it, but also that he was insecure on method: he was trying far away what he should first look nearby, since in reality the Romanian apă is the regular continua- tion of Latin AQUA, according to a process of develarization of labiovelar nexus QU, which, however, retains the condition of labialization, becoming the occlusive p of apă. The same phe- nomenon also occurred in Sardinian, where the p, in intervocalic position, was later voiced in b, as in abba.2 Ascoli continues exposing an example this time of ethnic character, referring in a footnote to the “storia universale” of Cantù3, in which they say that “il nome di Daci […] lo si ha pure per nome di tutta l’immensa nazione, che probabilmente dalla Persia e dal resto dell’Oriente [...] innondò tutt’Europa, e diede origine al generico nome deutsch”, as if to confirm the link between the Dacians of the Roman province and those coming from Persia, who, having in common the name deutsch, have been without a doubt - according to him - the ancestry of Germanic peoples that “calarono a distruzione del colosso romano”. It is a hypothesis that is now com- pletely abandoned, as the other (that was circulating at the time) that was correlating the Daco- Getae with the Goths. And once proposed the ethnic bond, it results to him easier to conclude that “la lingua di questi Daci sarà stata quella che diede origine alle germaniche”, adding immediately after “ed in ogni modo, attaccati restando alla storia, pare che l’imperatore Trajano abbia col suo sistema di colonizzazione estirpato quasi del tutto e lingua e costumi dei precedenti abitatori dei pa- esi corrispondenti all’odierna Valachia” thus confirming earlier: “Trajano [...] riduce la Dacia a provincia romana, vi pone nume- rosa colonia di soldati, che v’apportano lingua e costumi romani, ed uniti agli indigeni rimasti, formano il popolo Valaco, che tuttora con orgoglio ti vanta la sua origine.”

1 Italic is also used by Ascoli. 2 F. Dimitrescu, Întroducere în fonetica istorică a limbii române, Bucarest, 1967, pages 129-130. 3 For the biographical synthesis of C. Cantù see par. III.2.2.

49 He continued his exposition of historical character saying that in 250 the Goths invaded the terri- tory, forcing the Emperor Aurelian to give what Trajan had conquered. They were forced, then, to step back at the time of Constantine, but anew were driven back in the Roman area (“dove oggi giace la Valachia”) “dai fieri Unni”, during empire of Valens (376). Rejected, after an attempt to go toward Con- stantinople, the Goths were brought to Rome by Alaric. Even Attila some time after he went to It- aly, but his presence was so fast as to leave no trace “nei costumi e nella lingua dei popoli”. The Goths, later, they returned to Italy under the command of Theodoric, “che poscia divenne signore di tutta Italia”. Even the Lombards, continues Ascoli, they passed through the land of Wallachians “per prender fiato” before to break “nell’Italia, giardino del mondo”. And he remembers that Alboino put in Friuli “il primo fondamento della grandezza longobarda in Italia […..] provegnente direttamente dai paesi valachi”. Either Wallachia or Friuli, formerly belonging to Roman Empire, knew the same invasions, be- cause they were overrun by Slavs, and German populations that could have left traces in the same language. From these premises Ascoli deduces “che le lingue della Valachia e del Friuli, nella loro più importante parte, esser dovranno una composizione della romana, mista agli idiomi delle stesse barbare famiglie cui queste orde appartene- vano”. No wonder then - still - if among the two languages you will find many similarities. In the footnote he adds that could be identified in both languages ”scrap of Germanic languages “avanzi di lingue germaniche”. And he brings as examples the participles in t, as avut of Wallachians, vut of Friulians and gehabt of Germans and the gerunds in d as the Friulian durmind, the Wallachian dormind and German schlafend. We know instead that the voiceless dental of Romanian and Friulian participles is common to all Romance languages and comes from the Latin participle, whereas the voiced dental is what remains of Latin gerund, not to mention that in Friulian the voiced consonants word-finally undergo a process of devoicing, for which the graphic forms as durmind presuppose the outcome phonetic durmint. If the German forms gehabt and schlafend show some elements in common with what Romances this is due to the fact the Germanic family (from which is derived the German) is closely related to kinship with the Latin (and all other Indo- European languages) because both refer to a common Indo-European root. The linguistic affinities involving Romanian and Friulian did not affect the other Italian regions, according to Ascoli, such as Lombardy, because the historical developments and political of Friuli were different enough to allow an easier preservation of its idiom compared to that of neighboring areas. However, this is a hypothesis that already Luzzatto, as we have seen, will show unfounded.

50 V.5 Linguistic section

V.5.1 Introduction

From the title of his essay it clearly emerges that Ascoli had set the objective of verifying the exis- tence of common elements between the Friulian and the Romanian. He believed these elements needed to justify his theories of ethnological, according to a perspective similar to that of Cattaneo and, as we have seen, even of Luzzatto. He excluded instead the examination of any difference be- tween the two languages: we have already mentioned the limitations of this opinion, already de- tected by Luzzatto. I set the task of verifying the correctness of affinities considered in the essay of Ascoli, in the light of recent research in the field of Romanian and Friulian linguistics. However, I believe that it is not out of place to expose, albeit in summary form, the framework of the most distinctive differ- ences, also using the help of essay Del nesso fra la lingua valaca e l’italiana of Carlo Cattaneo1, which has, as we have already seen in Sec. IV.2.1, a series of observations more ‘linguistic’ of As- coli, in the sense he lists already departing the peculiar characteristics of a language than the other, as for example the postposition of article in Romanian (even noting that the phenomenon is also common to the Albanian, thus proving to be a forerunner of Balkan linguistic), which instead As- coli considers only in passing. For the realization of the second part of his historical-linguistic essay Ascoli uses the Romanian grammar of Ion Molnar, as he himself declares in a footnote on page 15 of his essay. The path of his work is the result of a careful reading of the text of Molnar, from which he points out step by step all the elements that according to him meet the character of affinity between Friulian and Ro- manian. Following this path too methodically the young Ascoli ends up neglecting the peculiar elements that stand out instead the Friulian from Romanian. And this is also the reproach that gave him Luzzatto, defining these differences “né meno numerose, né meno gravi ed importanti”,2 reminding him a series of phonetic and morphological phenomena of Romanian that are not reflected in Friulian, such as the postposition of article, the declination bi-casual, the periphrastic future, the outcome of nexus Latin CT giving rise to Romanian pt, the rhotacism and other yet. Do not forget that many of these phenomena had been observed also by Cattaneo,3 who was leading them to action of substrate, while the position of Luzzatto still remains to be clarified, having regard to final part of his re- sponse to Ascoli.4 With these exact words Luzzatto would have sealed his opinion about the theory of superstrate, that perhaps he may have had the opportunity to convey to young Ascoli on other previous occasions, either directly or through his son Philoxenus, who was his esteemed friend, but that it is as I repeat of an opinion to be verified.

1 Although there is no real evidence that Ascoli has read the Cattaneo’s essay. 2 For a more detailed analysis of letter of S. D. Luzzatto see par. V.1.1. 3 C. Cattaneo, Del nesso tra la lingua valaca e l’italiana, cit., see par. IV.2.1. 4 See par. V.1.1.

51 As you proceed in reading the linguistic part becomes tangible the general flaw of the setting of Ascoli, noted by Luzzatto, which is to have performed a dry comparison between Friulian and Romanian, totally detached from all ties with Latin and with the Romance overall framework.

V.5.2 Writing of Romanian in Ascoli

In his essay Ascoli uses for the Romanian a writing in Latin characters that deviates so much from the Molnar’s Grammar as the writing they will impose later. If Molnar associated with the Cyrillic sign ъ the Latin ė, to indicate a vowel more closed than a, which corresponds to current script of Romanian ă [.], Ascoli uses a simple e without accents or other distinguishing marks, as in the pair of examples ascoliani “vizel”, with e (current viţel ‘calf’) and “vizei”, with ă (current viţăi ‘calves’), where instead Molnar was marking more clearly the vowel variation with two different spellings “e” and “ė”, as we can see in “вицєл/vizel” and “вицъи/vizėi”.1 Ascoli therefore would have to distinguish between two vowels Romanian using two different spellings, just as he had done Molnar, because in Romanian the different degree of openness of ă and a is phonologically distinctive. In nominal morphology in fact the final vowel -ă indicates a morpheme without article, while -a indicates the presence of article. For the sign ê, which Molnar indifferently transliterated with either ie or e, Ascoli uses ë, which is very close to the Cyrillic spelling of ê, but differs from the current e [ýe], as when he writes ël in- stead of el ‘he’. The letter , transliterated with æ, is taken by Ascoli with the sign ä [4], which in the present Romanian is marked with â or î, depending on the position in the word. The letter ч, transliterated with tsch (according to the German standard), is taken up with ćh, as in taćhi instead of present taci, to indicating the phoneme of palatal affricate /tS/. The letter ц, transliterated with z (according to the German spelling), which indicates the voiceless dental affricate, whose current writing is expressed by sign ţ [ts], is recouped with z also from As- coli, coinciding with the norm. In order that the correspondence between the Romanian spelling of Molnar and the transliteration of Ascoli is the clearest possible I will report as a rule to both spellings throughout the course of my linguistic analysis.

1 See page 46 of his Grammatica.

52 V.5.3 Writing of Friulian in Ascoli

For the representation of Friulian phonemes Ascoli uses essentially the graphic characters of Ital- ian. To make phonemes not present in Italian, as the phoneme /c/, characteristic of Friulian, of voice- less prepalatal occlusive, Ascoli uses the spelling chi, as in “chiapiel” ‘hat’, “vachie” ‘cow’, “chiase” ‘home’, “pechiat” ‘sin’, “chiarnos” ‘fleshy’. He uses once incoherently the mark g in “ding” to indicate the plural of dint ‘tooth’, where t is palatalized in c’ [c]: dinc’ [dinc] ‘teeth’. For the corresponding voiced phoneme [F], Ascoli shows no example. Indicates the voiceless palatal affricate [tS] with the marking chi, as in “chizze” ‘bitch’, “chiness” ‘Chinese’. This graph usage, however, could lead to confusion the reader, because it overlaps with chi derived from the palatalization of nexus CA (see above), which does not indicate a affricate, but a occlusive. Ascoli always indicates the final consonants with handwriting etymological, as in lung, instead of lunk [luŋk], which represents the actual pronunciation.

53 V.5.4 Analysis of the linguistic part

In this paragraph I will examine in more detail the text of Ascoli dedicated to comparison between Friulian and Romanian, reproducing a piece at a time* and commenting soon after to verify the correctness of linguistic phenomena considered. Since he makes constant reference to Italian trans- lation after introduction of the Romanian examples, I cannot help but put the Italian language in analytic correlation with the Romanian and Friulian.

[I]. Venghiamo al confronto propostoci. Vedrassi dalle voci valache ch’io raccolsi senza stento, aver queste una immensa somiglianza con friulane dello stesso significato; nè punto dubito che chi volesse darsi la pena di gir più oltre, con un vocabolario alla mano, con tutta facilità ne accrescerebbe di molto il numero. Ma quelle che io riporterò, vedrassi essere parole essenziali alla lingua familiare, alle arti più indispensabili, parte delle favelle umane, ove il loro vero tipo ad evidenza incontrasi. Pe- rò non ardiva mostrare al pubblico queste somoglianze di voci, che, quantunque avrebber abbastanza soddisfatto, tuttavia sempre avrebbero lasciato al lettore l’idea d’una fortuita combinazione, se rego- le importanti della Grammatica io non avessi pure avuto in appoggio della mia comparazione.

Ascoli founds the principle of his comparison on a number of voices Romanian and Friulian simi- lar in shape and with same meaning. He also emphasizes that the lexical corpus will concern essentially terms the colloquial language, discussed with reference to a set of morphological rules as comparative ‘support’. Unfortunately for us he does not cite the source of these important gram- matical rules.1

* For the sake of convenience I find it useful identifying the sections of Ascoli with Roman numerals in square brackets. 1 On this point, I refer to my reflection on the historical grammars of par. IV.1.

54 [II]. E queste per le prime esporrò: Nei verbi per i primi: I pronomi possono essere preposti tanto nel valaco che nel friulano, e posposti in casi pure non interrogativi. Esempj: in valaco *) ël are, ël poarte, kalćhi-tu (le lettere ćhi con questo accento, avranno un suono aspro di tsc come se in questo esempio fosse scritto kalsci) in italiano egli ara, egli porta, tu calchi. Ed eccone la traduzione friulana al are, al puarte, chialchistu.

Speaking of the position of pronouns with respect to verb, Ascoli refers to the personal pronouns subject, distinguishing the Romanian “el” from Friulian “al”. He does not realize that in Friulian as in many linguistic systems (in particular in the northern Italian, but also in French), there are two pronominal series with subject function. One is represented by free pronouns, which are phonologically tonic and are not tied syntactically to a particular location. The other set consists of atonic and clitic pronouns. As atonic pronouns cannot enjoy life of their own, but they are necessarily linked to other tonic elements. As clitic they are stably bound to the verb to which they refer, except the event in which there are present other clitic elements, such as other oblique pronouns, that may cause the separation from the verb.2 The Romanian rather not have this feature. In today’s terminology Romanian and Italian are defined languages ‘at null subject’, since the presence of personal subject is not compulsory neither before nor after the verb to which it relates, and also these languages have only one series of subject pronouns. Instead the Friulian possesses a double series pronominal, of which only the clitic series is mandatory.still 2 Below I present the full list of subject personal pronouns related to Romanian, Italian and Friulian.

SUBJECT PERSONAL PRONOUNS Romanian Italian Friulian FREE CLITIC FREE CLITIC FREE CLITIC eu - io - jò o tu - tu - tu tu el - egli - lui al ea - ella - je e noi - noi - no o voi - voi - vo o ei - essi - lo:r a ele - esse - - -

2 L. VANELLI, Il sistema dei pronomi soggetto nelle parlate ladine, in I dialetti italiani settentrionali nel panorama romanzo, Stu- di di sintassi e morfologia, Roma, Bulzoni, 1998, p. 105 and foll.

55 Throughout the section, I will link to each fragment of Ascoli’s essay a graphical depiction where I summarize the examples of the Goritian scholar, preceded from those presented by Molnar’s Grammar, either in the graphical version in Cyrillic and Latin, to check for any discrepancies be- tween the two authors, as in the chart below:

Romanian of Molnar Romanian Italian Friulian

êл apъ/el arė (p.158) ël are egli ara al are

êл поартъ/el poartė (p.9) ël poarte egli porta al puarte

kaлчи tŏ/kaltschi tu (p.165) kalćhi-tu tu calchi chialchistu

In these examples, the association of pronouns is not consistent. While in fact the Romanian and Italian subject pronouns are free, those of Friulian belong to the series of clitic pronouns. From the writing of Molnar we notice that tonic a of “kaлчи tŏ/kaltschi tu” had not yet centralized in ă, as in current Romanian călci. All three verbal examples reported by Ascoli are the direct continuation of Latin ARĀRE, POR- TĀRE and CALCĀRE. I will report at the end of the commentary of each fragment, if possible, a table that reproduces the correct examples as they should have to be, compared to the initial one:

Latin Romanian Italian Friulian ARAT el ară egli ara lui al are PORTAT el poartă egli porta lui al pwarte CALCAS tu călci tu calchi tu tu calchis

In the examples of Romanian poartă and Friulian pwarte Ascoli should have found the affinity of diphthongization oà and wà as phonetic outcome of a common O tonic Latin.

56 [III]. *) Quelle e valache, a cui verrà sovrapposto questo segno (¨) avranno un suono quasi di je p. es. qui pronuncia jel are, dolcemente. Notisi ch’io mi servo della Grammatica valaca-tedesca di Molnar. Hermannstadt. 1810.

This phonetic note of graphic nature is suggested to him by Molnar, who uses the same artifice for the Cyrillic letter ê, superimposing a special accent mark when it corresponds to sound [ýe], writ- ing it ê, but in the Latin transliteration he uses the double spelling e and ie. On the Grammar of Molnar and other Romanian grammars you can find news very detailed in the eighth chapter of Carlo Cattaneo Del Nesso fra la lingua valaca e l’italiana output in Milan in 1837.3 Here he presents a long list of grammars and Romanian dictionaries, commenting on the historical journey and also expressing his opinion on the aspects he considers most important. He is very well informed about the grammar of the physician Johann Molnar, of which he says that was printed in Vienna in 1788 and whose author “giudicò bene di scrivere ogni voce valaca con ambo gli alfabeti latino e cirillico.” Later he also refers to reprint of this grammar, output at Hermannstadt (now Sibiu) in 1810, which is precisely the edition cited in note by Ascoli. The young scholar by Gorizia says almost nothing in theme of Romanian vocalism, apart from the brief mention in footnote to the Romanian vowel e- initial of word. It is the ascending diphthong [ýe] that represents the actual pronunciation of e- of the subject pronouns el, ela, ei, ele and of the forms of verb to be, eşti, era, and este. The apparent incongruity arises most likely from the transliteration of the Cyrillic4 spellings є and іє, which in turn correspond to a single Glagolitic5 sign З. The conventional reading of Glagolitic was З [ýe] at the beginning of word and after vowels, [e] in the other cases. This phenomenon was neutralized only in loans from Greek and in Slave words eda, ei, ese, eterъ,where З- had the initial sound [e]. Not even Cattaneo, in his essay of 1837, notes that Romanian has a very unique vowel system compared to all other Romance languages. In fact they do not possess the two phonemes /./ and /4/, graphically ă and î, which instead are common within the Balkan linguistic area. The central- ized voice /./ is represented by ă in Romanian and by ë in Albanian and it is not very different from the Bulgarian ъ, while the not labialized /4/ is equivalent to Romanian writing î, â and is similar to Russian Ы.6

3 See Cattaneo e il rumeno par. IV.2.1. About the Grammatica rumena di Molnar see par. IV.2.2. 4 The Cyrillic alphabet was based on the uncial Greek writing (Greek capital letters used in the inscriptions and in the papyri). N. Radovich, Introduzione allo slavo ecclesiastico antico, cit., p. 16. 5 The Glagolitic alphabet can be considered an original work of Cyril and in some cases seems to be inspired by the lower-case Greek writing (developed in age back, before in the papyri and then, only from the ninth century after Christ, in the manuscripts on parchment) and to other alphabets. N. Radovich, Introduzione allo slavo eccles. antico, cit. 6 G.B. Pellegrini, Introduzione alla linguistica balcanica, cit..

57

SYNCHRONIC CHART OF ROMANIAN TONIC VOCALISM Front unrounded Central Back rounded

i î,â u

é ă o è

a

Of Friulian vocalism Ascoli says nothing. Escapes him that the vowels in Friulian express a quan- titative dual system, phonologically relevant because distinctive. It is a innovative phenomenon of Friulian within the Romance, also present in other Italian varieties northern, eg. in Lombard and regards the presence of tonic vowels phonologically long.7 It is true, however, that this distinction does not regards the Goritian area, where the vowel system is more simplified. But even here particular vowel elements hint at the possibility that the vowel length has interested the same Goritian in ancient times.8 This clarification could, however, justify the omission of Ascoli, which moreover must not have known to perfection the Friulian, as he wrote in his letter to Marco Luzzatto on September 10, 9 1846: “A S.Daniele ho parlato friulano a lungo credo la prima volta in vita mia, e v’appresi qualche voce nuova”. Before illustrating the graphical depiction of Friulian vocalism I consider appropriate to display a quick diachronic overview to focus more clearly the singular phenomenon of Friulian tonic vocal- ism. As it is known, the vowel system of Latin was based on the principle of quantity, so every voice could have two durations: short or long. This quantitative opposition was representing a distinctive value, which however over time has weakened, until it was supplanted by a new mechanism based on different quality of vowels, which could be opened or closed.10

7 L. Vanelli, Le vocali lunghe del friulano, in “Quaderni della grammatica friulana di riferimento”, I, Udine, Forum, pages 69-107. 8 P. Rizzolatti, Elementi di linguistica friulana, Udine, Società Filologica Friulana, 1981, p. 23; L. Vanelli, Le vocali lunghe del friulano, cit., p. 70. 9 G. I. Ascoli, Composizioncelle, Archivio Ascoli, 3/163, c, p. 40; M. Radoni Zucco, Profilo biografico di Ascoli, 1973, p. 20. 10 V. Väänänen, Introduzione al latino volgare, Bologna, Pàtron, 1982.

58 It is well to specify that the vowel quantitative system should not necessarily be associated to an- cient languages, because it is also used by modern languages as the Hungarian11 and Gheg12, as well as by the Friulian. In the framework of Romance languages only the Friulian maintains a dual vowel system, returning to path of quantitative vocalism, without however lose that qualitative in- herited from . The following depiction represents the tonic vocalism of Friulian according to concept of strong position13, accompanied by that of weak position.

SYNCHRONIC CHART OF FRIULIAN TONIC VOCALISM strong position weak position

i: u: i u è, (é) ò, (ó)

e: o: jè, (jà) wè, (wà)

a: a

This dual system of quantitative tonic vocalism can be examined more clearly in its diachronic process which I shall explain in the next diagram. For now it is easy to see the two systems are equivalent in sense of the quantitative opposition (long ~ short), except for inclusion of diphthongs jè e wè. From this chart you can see that in strong position it appears no trace of tonal modulation, which instead it reveals itself clearly into the elements of weak position (distinction of tone open and closed of average vowels), which become from 5 to 7 (the vocal elements in parentheses are to be considered as alternative and not additional). In other words it can be said that the Friulian outcome of the Latin tonic vowels in strong position is identical to the vowel system of Sardinian14, but in a longer version. And that the outcome of vowels in weak position is very similar to the Italian one in open , as shown in the follow- ing diagram:

11 P. Fabiàn, Manuale della lingua ungherese, Budapest, 1970. 12 Albanian variety that is spoken in northern Albania; G. B. Pellegrini, Introduzione alla linguistica balcanica, cit., p. 10. 13 A tonic vowel in open Latin syllable, in a word that does not end in -a and that into Friulian result is in closed syllable, word- finally, is defined in strong position; see G. Francescato, Dialettologia friulana, Udine, Società Filologica Friulana, 1966. 14 Into Sardinian the short and long vowels of Latin merge into one vowel, giving rise to a system of five vowels, neutralizing the quantitative Latin opposition and excluding the pan-romance qualitative system; V. Väänänen, Introduzione al latino volgare, cit.

59

DIACHRONIC CHART OF TONIC VOCALISM Latin Vulgar Romanian Italian Friulian Latin closed open strong weak position syllable syllable position

Ī ī i i i i: i Ĭ - Ē é é é é e: è, (é) Ĕ è jè è jè i: jè, (jà) Ā - Ă a a a a a: a Ŏ ó ò wò u: wè, (wà) o Ō ó ó ó o: ò, (ó) Ŭ u Ū u u u u: u

The Romanian continues the pan-romance system only into the higher-mid vowels, while into the lower-mid vowels follows the Sardinian pattern. Singular the result of Friulian i: and u: - in strong position - for the short-mid Latin Ĕ and Ŏ, while the higher-mid Ĭ, Ē, Ō, Ŭ - in weak position - when they are followed by final vowel they always expand in è and ò open, such as in FĒTA > fède ‘sheep’, SĒRA > sère ‘evening’, GŬLA > gòle ‘throat’, LŬPA > lòve ‘wolf’.15 Only Italian keeps - in closed syllable - the character of Vulgar Latin.

15 P. Rizzolatti, Elementi di linguistica friulana, cit., p. 25.

60 [IV]. Le desinenze dei verbi valachi regolari sono queste: Della I. Congiugazione a lungo. Esempio luminà, illuminare, leudà, lodare; la di cui traduzione friula- na sarà appunto luminà, laudà. Della II. Congiugazione ea lungo. “ III. “ e breve. “ IV. “ i lungo. Ognuno s’accorgerà che le desinenze della I. II. e IV. Congiugazione valaca, sono le identiche della maggior parte dei verbi friulani: La Grammatica valaca ch’io m’ho innanzi gli occhi, ha per esempio della III. Congiugazione i due verbi bàte, vìnde, - battere, vendere; osservisi la somiglianza con bati, vendi, dei friulani: sicchè la desinenza della III. Congiugazione valaca (e breve) che pare non esista in friulano, v’è rimpiazzata dall’i breve. Per esempio d’i lungo trovo dai valachi dormì, dormire; eccoci al friulano durmì. We can summarize in this manner the Ascoli’s reasoning of the preceding lines:

Molnar’s Romanian Romanian Italian Friulian I A lungo luminà illuminare luminà leudà lodare laudà II EA lungo - - - III E breve батє/bate (p.257) bàte battere bati vìnde vendere vendi IV I lungo a дорми/a dormi (p.215) dormì dormire durmì

We can grant to Ascoli ea ‘long’, as diphthong. But there is no reason to define “breve” the e, of III (unstressed), and long the i, of IV (tonic). The true meaning that in reality Ascoli intends to assign to vowels, using the terms ‘short’ and ‘long’, is undoubtedly what of ‘unstressed’ and ‘tonic’. So we have to read the accent of his examples exclusively in its value of tonic sign. The correct exposure will be:

coniug. Latin Romanian Italian Friulian I -ĀRE ILLUMINĀRE -a a lumina -are illuminare -a: lumina: LAUDĀRE a lăuda lodare lauda: II -ĒRE - -ea - -ere - -e: - III -ĔRE BATTUĔRE -e a bate -ere battere -i bati VENDĔRE a vinde vendere vendi IV -ĪRE DORMĪRE -i a dormi -ire dormire -i: durmi: [I marked in bold type all the tonic vowels]

61 The four conjugations of Latin are preserved intact both in Romanian and in Friulian, but com- pared to Italian the desinences of the Romanian and Friulian show the apocope of final syllable -RE, without however causing any modification of tonic accent, which remains anchored to primary vowel. Therefore, the common element among Romanian and Friulian is not the tonic final vowel of the infinitive, but the fall of final syllable of the Latin infinitives. Today we know this phenomenon is widespread in many Italian and Romance varieties (eg. central-southern Italian dialects, in those of the north-west area and in the same southern Tuscany).16 The elements instead that distinguish the infinitives of the two languages are the preposition of the vowel a ( -i. It is possible indeed that after the apocope of final-E also the vowel -Ĕ- is fallen, leaving a consonant nexus in word-finally in intolerable execu- tion, as in VENDĔRE > *vendr. At this point it intervenes the support vowel -i, typical of Friulian, which closes the word (*vendri) and finally it falls also the -r- realizing the current outcome vendi.17

[V]. Il participio passato, formano i valachi aggiungendo agl’infiniti una semplice t; dunque avut, arat, dormit, leudat, - avuto, arato, dormito, lodato; saranno participj valachi, abbastanza compren- sibili mi sembra ad ogni friulano.

Molnar’s Romanian aвoт/avut (p.151) apaт/arat (p.165) дopmит/dormit (p.215) Romanian avut arat dormit leudat Italian avuto arato dormito lodato Friulian - - - -

If we consider the topic in historical perspective, the t is not “aggiunta” in Romanian to form the past participle, but is derived, as well as in Italian and Friulian, from the Latin form of the partici- ple, as we see in the table just below. It should be emphasized, however, that the Romanian infini- tive linked to participle avut is a avea. Consequently he has no basis the Ascoli’s theory that just add to the infinitives a -t to get the corresponding participle, because avea + t would give *aveat and not avut.

16 G. Rohlfs, Grammatica storica della lingua italiana e dei suoi dialetti, Morfologia, Torino, Einaudi, 1968, p. 339. 17 R. Maschi, Morfologia storica del friulano: l’evoluzione del sistema verbale, in “Ce fastu?”, LXXVI (2000), 2, Udine, p. 202.

62 Ascoli does not explicit the Friulian form corresponding to that of the Romanian verbs, only say- ing they would be easily understandable in Friulian, thus suggesting that are similar. Instead I add it in the following table, pointing out its specific characteristics, including the common ancestor Vulgar Latin which they belong.

Latin Vulgar Latin Romanian Italian Friulian I ARĀTUM ARĀTUM arat arato ara:t LAUDĀTUM LAUDĀTUM lăudat lodato lauda:t II HABĬTUM HABŪTUM avut avuto vu:t IV DORMĪTUM DORMĪTUM dormit dormito durmi:t

The ones shown above are past participles ‘weak’, derived from verbs with weak perfect -ĀVI, - ĪVI, which have long vowel in penultimate Latin position, and so tonic. In Friulian the past parti- ciples are characterized by long desinential vowel -a: t, u: t, and -i: t, with the lengthening closely associated with a final voiced consonant devoiced: originally this (tonic) consonant was voiced be- cause it was in intervocalic position and after the fall of final vowel is devoiced, according to evo- lution -ATU(M) > ado > ad > a:t.18 Ascoli does not consider any form of past participle ‘strong’, which in Friulian would not present any vowel lengthening, since in these verbs the final dental -t is not a -d devoiced, but comes from the simplification of a consonant nexus or from the shortening of a geminate, whereby also in the 19 corresponding feminine form there is -t-, like the past participle fat ‘made’, from the Latin FACTU(M), which has the corresponding feminine fate ‘made’.

18 L. Vanelli, Le vocali lunghe del friulano, in “Quaderni della grammatica friulana di riferimento”, cit., p. 89 and foll. 19 L. Vanelli, Le vocali lunghe del friulano, cit., p. 89 and foll.

63 [VI]. I gerundj valachi sono pretti pretti i friulani nella loro forma: di dormì, avremo in valaco dormind; di arà, arare, aränd *); d’un ausiliare troveremo fiind, essendo, friulano sind, (jessind). [Note Ascoli] *) Gli a con questa linea – sovrapposta devono essere nasali e profondi, come gli a ottusi degli ingle- si.

Molnar’s Romanian Romanian Italian Friulian dormì дopmинд/dormind (p.215) dormind - - arà ap нд/araend (p.165) aränd - - - фїинд/fiind (p.157) fiind essendo sind (jessind)

In this scheme, Ascoli does not represent the infinitive form of Romanian verb to be a fi and indi- cates with script ä the phoneme /4/, which is one of the peculiar characteristics of Romanian vowel system, namely the central vowel not labialized î. Furthermore he does not explicit the forms of Friulian gerunds, but only says that Romanian gerunds are “pretti pretti i friulani”. We should there- fore interpret the forms dormind and aränd as Friulians, but we will see later that this association is not entirely correct. In the note ‘*)’ Ascoli proposes the phonetic implementation of a marked with the line ‘–’ (but ac- tually the typography writing is a ä) as a vowel “nasale e profonda, come gli a ottusi degli inglesi”, per- haps that one reduced of but [bVt], while in reality we know that it is a high vowel and centralized, as in children [‘tS4ld0.n], which is neither nasal nor deep. In Romanian the suffixes of Latin gerunds -ANDO and -ENDO have the outcome -înd into verbs of conjugation in -a, -ea, -e and -î with consonantic root. Instead, they have the outcome -ind into verbs of conjugation in -i, -a, and -e with vocalic root. In the Friulian the tonic vowel -A of desinence -ANDO of Latin gerunds remains unchanged, while the -E of -ENDO becomes -i, taking the vowel character typical of IV conjugation. All Friulian vowels different from -A fall, and the word-finally voiced dental becomes devoiced by going to -t, while the examples of Ascoli end up in voiced dental, rather than voiceless: *dormind for durmint, *arand for arant, *jessind for jessint. In the example also he presents the Italian translation of the gerund auxiliary only. For complete- ness, I add (in gray) also the forms that are missing:

Latin Romanian Italian Friulian ARANDO arînd arando arant DORMIENDO dormind dormendo durmint - fiind essendo jessint (sint)

The Friulian form of the gerund of auxiliary ‘to be’ jessint may also be assumed on the basis of Latin infinitive ĔSSE, where the initial vowel Ĕ- undergoes the regular outcome je- in the context of weak position and loss of final vowel. To form jess- is added the ending -int, analogic with

64 other verbs of third conjugation, as in bati > batint ‘beating’, for the formation of gerund of auxil- iary, which did not exist in Latin.

[VII]. Un tempo composto del verbo avere valaco sarà ël au avut, come il friulano el a avut.

Molnar’s Romanian Romanian Friulian

єл aŏ aвŏt/el au avut (p.146) ël au avut el a avut

In ancient times, the ‘reduced’ auxiliary of singular third-person was au for the compound tenses, then in Romanian modern is simplified in a. It should be remembered, however, that auxiliaries of Romanian past composed are not exactly the same as the present tense of verb a avea, as we see on the table that follows:

simple present am ai are avem aveţi au auxiliaries in compound tenses am ai a am aţi au

In the example of Friulian compound tense “el a avut” the personal pronoun el is decidedly out of place because the two options are the free pronoun lui or the clitic al, as I has already explained in the commentary on the passage [II], and el does not exist. The past participle of the Friulian infini- tive ve: ‘have’ is not *avut, but the apheretic vu:t. So it would have been more accurate to write:

Romanian Friulian el a avut al a vu:t

Even in this form, however, both expressions confirm an excellent executive affinity.

65 [VIII]. Notisi pure che le varie maniere, in cui il friulano può unire al verbo il pronome (forse più facilmente che l’italiano) trovansi pure tutte nel valaco. Esempj: dice il valaco: demi, dei, daune, dauve, - dammi, dàgli, dannoci, dannovi, confrontar li potremo coi friulani: dami, dai, danus, daus (al sing.). Il valaco ha ancora più complicate queste regole, perchè può dire tutto in una voce miau - mi hanno, fr. mi an.

Molnar’s Romanian дъми/dėmi дъи/dėi даонє/daune даовє/dauve (p.338) Romanian demi dei daune dauve miau Italian dammi dagli dannoci dannovi mi hanno Friulian dami dai danus daus mi an

Graphically the unstressed personal Romanian dative pronouns postponed to the verb are not united, but separated by a hyphen. The Italian forms dannoci and dannovi are now no longer used, but were common in the nineteenth century, at least in written literary language. These examples concern the relationship between the verb and the unstressed personal pronoun and we can see that in Romanian only the first two forms dă-mi, “demi”, and dă-i, “dei”, are perfectly regular. In most cases, in fact, the unstressed forms precede the verb and are postponed only when they go with the affirmative imperative (as is in fact the case of dă-mi and of dă-i) or the gerund. And so the post- position of pronoun in dau-ne a in dau-vă would indicate that the verb is imperative. But the VI person does not exist, in the imperative mood. We must therefore consider them, as well as mi-au, present indicatives, and as such usually have the pronoun placed before: ne-dau, vă-dau. The same holds true for the Friulian. So It would have been better write:

UNSTRESSED DATIVE PERSONAL PRONOUNS Romanian Italian Friulian imperative dă-mi dammi dami

dă-i dagli dai mi-au mi hanno mi an present ne-dau ci danno nus dan indicative vă-dau vi danno us dan

In these examples Ascoli presents the development of the personal pronouns unstressed dative ap- plied to the verb ‘give’ and to just one expression of verb ‘to have’. The Romanian forms refer to

66 verbs a da and a avea, while the Friulian ones to verbs da: and ve: and in both cases they are de- rived from Latin DARE and HABERE. Both a da and da continue properly the Latin first conjugation, while a avea and ve: continue the second:

Latin Romanian Italian Friulian I DARE a da dare da: II HABĒRE a avea avere ve:

To complete the picture of the personal pronouns dative I think it is useful to illustrate the com- plete development of them in all three languages in question, by coupling to the clitic series also the one free:

DATIVE PERSONAL PRONOUNS Romanian Italian Friulian FREE CLITICS FREE CLITICS FREE CLITICS mie (î)mi a me mi a mi mi ţie (î)ţi a te ti a ti ti lui (î)i a lui gli a lui j ei - a lei le a je - nouă ne, ni a noi ci a noaltris nus vouă v(ă), vi a voi vi a vualtris us lor le a loro loro a lo:r ur - li - - - -

67 [IX]. Osservisi questi modi di dire friulania, con l impersonali: nus l’a dat, us l’ai dat – ce l’ha dato, ve l’ho dato: ed io per tradurli in valaco dovrò dire precisamente nilau-dat, vilam-dat. (In alcuni luoghi del Friuli dicesi pure vi l’ai dat a grande somiglianza del valaco).

Molnar’s Romanian (p.339) Romanian Italian Friulian нилаŏ-дат/nilau-dat nilau-dat ce l’ha dato nus l’a dat - vilam-dat ve l’ho dato us l’ai dat (vi l’ai dat)

Even in this case it is necessary to immediately correct the discrepancy existing in the first Roma- nian example “nilau-dat”, because in the Romanian of today is no longer used the auxiliary ‘re- duced’ au, but a. The dative clitic pronouns of plural that are used in front of another pronominal form are precisely ni and vi. For the rest, it should be only changed the look and feel in ni-l a dat and vi-l am dat. This paragraph, which would like to introduce the curious concept of “l impersonale”, actually de- scribes adequately the relationship of the double personal pronoun with the verb, within which emerges a new aspect of the series of the personal pronouns that had not yet been called in cause. What Ascoli curiously defines “l impersonale”, on the other hand, is nothing more than the accusa- tive element of the pronominal clitic series, as can be seen in the following table (third row):

ACCUSATIVE PERSONAL PRONOUNS Romanian Italian Friulian FREE CLITICS FREE CLITICS FREE CLITICS mine mă, (m-) me mi me mi tine (-)te(-) te ti te ti el -l, l-, -l-, îl lui lo, l+voc. lui lu, l+voc. ea (-)o lei la je le noi (-)ne(-) noi ci noaltris nus voi vă, (v-) voi vi vualtris us ei (-)i(-), îi loro li lo:r ju ele (-)le(-) - - - lis

The Friulians examples of Cattaneo20 and Marchetti21 already showed, however, that when an ac- cusative pronoun follows another pronoun (usually a dative) it is used another series of clitics:

20 F. Agnoletti, Appunti linguistici sul friulano in un manoscritto inedito di Carlo Cattaneo, cit., p.161. 21 G. Marchetti, Lineamenti di grammatica friulana, Udine, Società Filologica Friulana, 1985, p. 212.

68

singular plural masc. al ai femin. e es

Therefore, the Friulian examples “nus l’a dat” and “us l’ai dat” should be corrected in nus al a dat and us al ai dat. Unless, Marchetti always specifies, especially in areas exposed to extraneous influ- ences, does not react a ‘recent’ analogy factor with Italian. For ‘recent’ he means that the phe- nomenon may have developed not before the nineteenth century, to which follows, however, a du- bious question mark. The precise framework should therefore have been:

Romanian Old Italian Italian Friulian ni-l a dat ne l’ha dato ce l’ha dato nus al a dat vi-l am dat - ve l’ho dato us al ai dat

[X]. Indi noteremo altre regole: Coll’allungarsi delle voci valache, sempre più scorgonsi avvicinarsi alle friulane: es. mort, - morto, per dire in femminino avrassi moarte, ed eccoci al friulano muart; port, - io porto, in terza persona diventa ël poarte, quasi l’identico puarte dei friulani; di dormì, avremo ël doarme, eccoci all’al duar dei friulani.

Molnar’s Romanian Romanian Italian Friulian masculine мopt/mort (p.9) mort morto muart

feminine моартъ/moartė (p.9) moarte - -

Ia person порт/port (p.9) port io porto -

IIIa person поaptъ/poartė (p.9) ël poarte - puarte

infinitive а дорми/a dormi (p.215) dormì - -

IIIa person єл доармє/el doarme (p.208) ël doarme - al duar

All versions represented by Ascoli can be traced back to a single Latin form from which they de- rive: MORTŬU(M), PORTĀRE, DORMĪRE.

69 The real affinity regards the Latin tonic vowel which diphthongizes in Romanian and in Friulian (but not in Italian). In Romanian the tonic vowels diphthongize in rising eà, oà when they are in Romanian penultimate syllable, as in fact MORTŬA(M) > moartă ‘(she) dead’, but MORTŬU(M) > mort ‘(he) dead’, DORMĬT > doarme ‘he or she sleeps’ but DORMIO > dorm ‘I sleep’, POR- TAT > poartă ‘he or she carries’ against PORTO > port ‘I carry’. Ascoli brings only the examples concerning the evolution of O Latin and none for vowels me- dium-high not labialized, but you might remember chances like NĬGRA(M) > neagră ‘(she) black’, CĒRA(M) > ceară ‘wax’, GĔMĬTU(M) > geamăt ‘moan’. As you can see from these ex- amples the Latin tonic vowels affected by the phenomenon can be both long and short, in stark contrast to the diphthongization of Friulian, which develops exclusively from the short Latin me- dium vowels in the context of weak position, as I report below. The phenomenon that makes possible the diphthongization of Romanian tonic vowel is the meta- phony, which is caused by quality of the vowel (ă and e) of final syllable.22 We have just seen in the chance of MORTŬU(M) > mort ‘dead’ and DORMIO > dorm ‘I sleep’, that the metaphony do not occur in absence of final vowels and even in presence of final vowels other than -ă and -e, as in chance of the -i of plural form ploi ‘rains’. It also seems that the Romanian diphthongs eà and oà wish to assume the function of distinguishing in more marked mode the forms singular femi- nine by ones masculine. It should be noted also that the diphthongization occurs regardless of the nature of open or closed syllable, as one can easily see in the above examples and also in this it distances a lot from the diphthongization of Friulian, which strictly requires closed syllable Latin. Contrarily to Romanian the Friulian diphthongization is also developed in the monosyllabic forms, as we can observe in the forms that makes us notice the same Ascoli: DORMĬT > al dwar ‘he or she sleeps’, with total disappearance of the final syllable Latin -MĬT and MORTŬU(M) > mwart ‘dead’. Together even with the example PORTAT > (al) pwarte ‘he or she carries’, we are in the Friulian context of weak position (closed syllable Latin), where the outcome of tonic Ŏ is the as- cending diphthong wà. When moreover the tonic vowel is followed by a final -r it also triggers the process of elongation, for which we have DORMĬT > (al) dwa:r. For the first time Ascoli uses the clitic subject pronoun al in front of verb dwa:r, but not in front of pwarte and once again he omits any mention of the long vowel. Following these considerations I can write:

Latin Romanian Italian Friulian masculine MORTŬUM mort morto mwart feminine MORTŬAM moartă morta - Ia person PORTO port io porto - IIIa person PORTĂT el poartă egli porta al pwarte infinitive DORMĪRE a dormì dormire - IIIa person DORMĬT el doarme egli dorme al dwa:r

70 The other phenomenon common to both languages, that Ascoli should have noticed in the exam- ples reported, is the fall of the Latin final vowels other than -A as in MORTŬU(M) > Rom. mort, Friul. mwart, but MORTŬA(M) > Rom. moartă (Friul. mwarte). We must emphasize however that in Romanian this phenomenon involves only the singular , while, as regards the Friulian, the appearance of the plural is slightly more complex, as will be seen below in the discussion of the plural.

[XI]. I nomi femminili valachi, nella di cui penultima sillaba evvi oa, e che terminano in e cangiamo l’e in i es. ploae pioggia, ha in plurale ploi. Ciò scorgiamo anche in friulano che da ploë abbiamo plois, da puarte, puartis. (Però tanto per questa regola quanto per due altre che in seguito verranno, bisogna dire, che queste semplici i sono strane un po’ per i plurali valachi, che così rassomigliano ai friulani, mentre per i friulani, l’i plurale è comunissimo, avendo tanti plurali in i, is ecc.)

Molnar’s Romanian Romanian Friulian sing. плоає/ploae (p.64) ploae ploë puarte plur. плои/ploi ploi plois puartis

Ascoli notes that the -e final, of some Romanian feminine, changes at the plural in -i. Then he adds that the same happens in Friulian, as in ploe ‘rain’, plois ‘rains’, where in fact the same vowel mu- tation occurs, but does not consider at all the presence of -s final. In his reflection brackets he notes also the strangeness of Romanian plurals in -i, while believes that in Friulian the exit in -i is very common. Now we say that the plural in -i is a really peculiar feature of Romanian (and Italian) within the romance domain, where instead the common rule for forming the plural is the use of -s. Cattaneo23 was come out very simply about the Romanian plural noting the close relationship with the Italian one: “Nella formazione dei plurali la simiglianza tra la lingua valaca e la nostra è quasi perfetta. Lupi, ursi, tauri, vermi nel mascolino; e nei feminili capre, persone, camiscie, funi, legi, ec., con poche eccezioni. Lo stesso dicasi degli aggettivi negru e negra, negri e negre; verde e verdi. Alcune voci rammentano quelle antiche forme italiane tèmpora, càmpora, ràmora imitate dai neutri latini; poiché hanno i plurali campuri, tempuri, ec. Alcune, come da noi, mutano il c chiuso coll’aperto, e come presso di noi amico e amici, fanno sacu e saci.”

22 It is up to A. Mussafia the merit of having first recognized the nature of Romanian diphthongs ea and oa, developed by the metaphonic influence of final vowel. See. L. Renzi, Gli studi rumeni di Adolfo Mussafia, in Omagiu lui Alexandru Rosetti la 70 de ani, Bucharest, 1965. For a comment of same Romanian diphthongs see also T. Ferro, Latino e lingue balcaniche nella formazione del rumeno, Catania, CUECM, 1992, p. 244 and the foll. 23 C. Cattaneo, Del nesso tra la lingua valaca e l’italiana, cit., pages 228-29. Cattaneo, however, makes no mention of the Roma- nian diphthongization.

71 Friulian is placed in an intermediate position, adopting both the vowel plural and sigmatic one. And really this is its peculiarity. This dual mechanism used by the Friulian for plural construction can be traced back to an old bicasual declension, in which the language had at its disposal two dif- ferent morphemes to indicate the plural: the -i of nominative of second declination and the -s of accusative -AS, -OS and -ES of their Latin declensions. In the transformation of Latin in the an- cient Friulian, the bicasual declension could be productive as long as the words derived from the II declination, which could choose between a quantitatively marked plural (with the addition of -s) and a qualitatively marked plural (palatalization of the final consonant due to the action of -i, which would later fall) then opposing a to a sibilant. Not all the words of II declension could, however, choose between the sigmatic plural and the palatal one, because some already ended with -s at singular. The only path indeed that could guar- antee the morphological distinction of number was what linked to palatalization of final sibilant (as nas/naS), because in the case of sigmatization the plural remained identical to singular (nas/nas). If the palatal way was the most productive for this limited group of words ending in -s, it could therefore also be followed by those forms which, though ending in a palatalizable conso- nant, they could freely choose the sigmatic plural. The words deriving from third declension instead had no other choice but to use the morpheme -s, since it involved both forms derived from nominatives and the accusatives.24 The desinence -is of plois and pwartis affects the feminine nouns ending in vowel, in which the addition of the morpheme -s causes an elevation of the voice than in the singular, for a general phonological rule that raises an post-tonic underlying /a/ in front of an consonant, thus generating the evolutionary sequence PLŬVIAS > ploias > ploiis > plois and PŎRTAS > pwartas > pwartis.

Romanian Friulian sing. ploaie ploe pwarte plur. ploi plois pwartis

Both Romanian and Friulian examples derive from Latin PLŬVIA and PŎRTA, but the outcomes of the tonic vowel follow different paths. When tonic vowels of Romanian singular forms are con- ditioned by presence of an open or mid vowel in final syllable, they give rise to the formation of the diphthong oà, as in PLUVIA > ploaie ‘rain’, while the Latin final -A which should switch to - ã, as in neagră < NĬGRA(M) and moartă < MORTŬA(M), is reduced instead to -e for influence of semivowel j that precedes it. In ploi the ending -i (closed vocal) of plural neutralizes the phe- nomenon, recreating the conditions of the original syllable plo- < PLŬ-.25 In Friulian the diphthong wà of pwarte ‘door’ rises, only in context of weak position, from mid short Latin tonic vowels, and this condition remains unchanged even for the plural.

24 L. Vanelli, La formazione del plurale in friulano e la ricostruzione diacronica: l’ipotesi della declinazione bicasuale, in “I dialetti italiani settentrionali nel panorama romanzo. Studi di sintassi e morfologia”, Roma, Bulzoni, 1998, p.153 and foll. 25 T. Ferro, Latino e lingue balcaniche nella formazione del rumeno, cit., p. 244 and foll.

72 But these examples also show another correlation that Ascoli would not have to ignore, and it is the preservation of link consonant PL, which we note either in Romanian ploaie or in Friulian ploe, both from PLŬVIA, but you might remember other Romanian cases as plus ‘more’ < PLUS, plumb ‘lead’ < PLUMBU(M) and the Friulian plume ‘feather’ < PLŪMA(M). From this passage to the one [XXII] (except [XV] and [XVI], in which he speaks respectively of the article and gender) Ascoli talks exclusively about the formation of plural.

[XII]. I nomi valachi che terminano in g duro lo cambiano in plurale in g dolce. Es. lung – lungo – lungi, come il friulano lung, nel plurale femminile lungis.

Molnar’s Romanian Romanian Italian Friulian sing. лŏнг/lung (p.11) lung lungo lung plur. лŏнџи/lungi lungi - lungis

This exposition of Ascoli on palatalization of voiced velar [g] comes back in the morphological analysis of plural of the previous examples. The adjective of first class LŎNGU(M) expounds differently the final syllable -GU(M) at singular and the plural alike. In the plural indeed it is easy to notice the presence of palatal ending -i for the Romanian and sigmatic ending -is for the Friulian. In the plural the velar [g] is conditioned by -i in Romanian and -is (< -AS) in Friulian, becoming from voiced velar to palatal affricate [dZ], while in Italian (although Ascoli omits the plural version lunghi) it remains occlusive, but it becomes palatal for the presence of final -i. There would have been also to be observed the outcome of the velar, which in end position re- mains sonorous in Romanian, but not in Friulian. Indeed the apocope of final vowel -U causes in Friulian the desonorization of the voiced consonants, which deprived of their syllabic nucleus (the vowel fallen), are forced to emigrate, in the condition of final consonants, in the pre- ceding syllable. In this step from the final syllable to the previous one they undergo a process of transformation from voiced to voiceless. In the final syllable -GU(M) of LŎNGU(M) therefore, after the fall of final vowel -U, you have the desonorization of the velar [g] in [k], with Friulian outcome [lunk]. To complete the analysis of these forms is well to remember that the actual Romanian pronuncia- tion of -i of lungi is that of a vowel more short and less voiced than that which is used as mor- pheme of the article: [lung‘]. This results in the following chart:

73 Latin Romanian Italian Friulian masculine singular LŎNGU(M) lung lungo lunk masculine LŎNGI lungi lunghi - plural feminine LŎNGAS - lunghe lungis

[XIII]. Alcuni dei nomi valachi che hanno per ultima sillaba le preceduto da vocale, perdano la l. Es. moale, - molle, - moi; piiale, - pelle, - piei, appunto come il friulano chiapiel, - cappello, - chiapiei; biel, - bello, - biei ecc. ecc.

Romanian Italian Friulian sing. moale piiale molle pelle cappello bello chiapiel biel plur. moi piei - - - - chiapiei biei

Molnar’s Romanian (p.13) sing. moaлє/moale пї лє/piiale plur. moн/moi пїєи/piei

The following forms proposed by Ascoli are the evolution of Latin BĔLLU(M) ‘lovely’, a diminutive of BŎNU(M) ‘good’,26 and MŎLLE(M) ‘soft’, of PĔLLE(M) ‘skin’ and of me- dieval Latin CAPPĔLLU(M) ‘hat’, a diminutive of CĂPPA(M) ‘hood’.27 These examples are characterized by the etymological presence of the geminate liquid, which tends to simplification in almost all the Romance languages. It is in fact preserved only in Tuscan, in south- and in Sardinian.28 Even Cattaneo had noticed “quel medesimo aborrimento delle consonanti doppie che distingue la lingua spagnola e molti de’ nostri dialetti, massime il veneto; e ne sian esempio stele, vale, curere, ec.” 29

26 Väänänen, Introduzione al latino volgare, cit., par. 158, p.146. 27 M. Cortelazzo - P. Zolli, Dizionario etimologico della lingua italiana, Bologna, Zanichelli, 1979, I, p. 129. 28 Väänänen, Introduzione al latino volgare, cit., par. 110, p.115. 29 C. Cattaneo, Del nesso tra la lingua valaca e l’italiana, cit., p. 218.

74 The Romanian forms at the singular, moale and piiale, when they pass to plural lose the l follow- ing the palatalization caused by final -i, generating the outcome moi and piei. This is determined from the palatal plural of MOLLE(M) and PELLE(M) which conditions the geminate liquid LL, palatalizing it first and then pushing it to the zero degree, as happens starting from vulgar Latin CABALLI > cabal’i > cai.30 In Friulian, the use of the palatal plurals primarily interested the words belonging to II Latin de- clension, which were ending at singular in coronal consonant and it originates from plural of nominative in -I, while in all other cases the sigmatic plural comes from the suffix of accusative plural in -S. These coronal consonants are characterized by their easy adaptability to palatalization when followed by /j/, such as /g/ + /j/ > /dZ/, /n/ + /j/ > / /Y/. But since in the phonemic inventory of modern Friulian there is no the liquid palatal /Y/ we are faced with the (singular) absence of the palatalized consonant, because it was replaced by the pho- neme /j/. The presence of final -i in the forms chjapjei and bjei therefore does not represent an ex- ception to the rule that sees the fall of all vowels other than -A, but the semi-vowel /j/ as reflex of transformation of phoneme of the palatalized coronal /Y/.

Romanian Italian Friulian sing. moale piele molle pelle cappello bello chjapjel bjel

plur. moi piei molli pelli cappelli belli chjapjei bjei

As you can see by the outcome of palatal plurals of these examples, there is a real affinity between Romanian and Friulian. Ascoli, however, would have to also point out the typical peculiarities of Friulian represented by the palatalization of nexuses CA and GA, that he proposed in his exam- ples, albeit unintentionally.

30 F. Dimitrescu, Întroducere în fonetica istorică a limbii române, Bucarest, 1967, p. 98.

75 [XIV]. I nomi sostantivi valachi terminanti in t cangiano t in z. Esempj: moarte, - morte, -moarzi, - le morti, nel friulano vediamo pure muart, muarz; dei nomi valachi coll’ultima sillaba ent, unt, at, ot, molti pure cangiano questa t in zi. Es. talent, - talento, - talenzi; kumnat, - cognato, - kumnazi; funt, - funti, funzi; ecc. confrontiamoli un po’ col friulano e ne avremo talent, talenz, cugnat, cugnaz, ecc. ecc.

Romanian Italian Friulian kumnat cognato cugnat singular funt - - moarte morte muart talent talento talent kumnazi - cugnaz plural funzi funti - moarzi le morti muarz talenzi - talenz

Molnar’s Romanian singular plural кŏмнат/kumnat (p.53) -ци/-zi фŏнт/funt (das Pfund) (p.53) фонци/funzi moaptъ/moartė (p.80) -ци/-zi талєнт/talent (p.53) -ци/-zi

Some Romanian consonants, when followed, in contact or at a distance, by a i both tonic and atonic (and therefore not only with morphological function) undergo a form of palatalization, con- ditioned by regular semivocal jod. They are part of this group, not only the voiceless dental occlu- sive [t], which becomes coronal affricate [ts], as in cumnat > cumnazi [ku’mnatsi] ‘cognates’, but also the relative voiced [d] > [z], as in rotund > rotunzi [ro’tunzi] ‘round’, the voiceless velar [k] > [tS], in bunic > bunici [bu’nitSi] ‘grandparents’, the voiced velar [g] > [dZ], in plug > plugi [‘pludZi] ‘plows’ and the dental [s] > [S], in nas > naşi [‘naSi] ‘noses’. The notch of [s] > [S] also acts at a distance, as in the case of artist > artişti [a’rtiSti] ‘artists’. The semivocal element is not only represented by the morphological final -i, unstressed, but also from internal -i-, tonic, creating new shapes without i, but which presuppose its presence in ancient times, as in ţară according to the possible evolution TERRA > *tjeară > ţară.31 The occlusive t + j gives rise to the voiceless dental affricate [ts] and this is what we find in the examples of Ascoli, which all refer to notch of the final morphologic -i of plural. Graphically, it must be indicated by ţ,

76 instead of z, because the latter can be a use phonic multi-purpose: the voiceless dental affricate [ts], the sonorous one [dz] and also the voiced sibilant [z]. It is interesting to note that even the Friulian maintains a parallel case in the form tjare (and tjere) < TERRA(M), with regular diph- thongization of E tonic in weak position. This phenomenon, however, is conditioned geographi- cally, because in some areas of Friuli it also proceeds to palatalization of tj > c’ [c], with the result cere [‘cEre], without affrication and coronalization. The reference of Ascoli to typology of the last syllable, “ent, unt, at, ot”, as a causative element of transformation phenomenon of t to z is completely out of place, because all comes down to the re- action of t followed by jod. The Friulian plural forms of cugnaz, mwarz, talenz, while resembling very to related Romanian forms, they are actually the application of Friulian sigmatic plural. The voiceless dental affricate z [ts], which closes the word, is nothing indeed that the result of fusion of t of singular with s of plu- ral, t + s > z [ts], as in cugnat > cugnats > cugnaz [ku’

Romanian Italian Friulian cumnat cognato cugnat singular funt libbra lire moarte morte mwart talent talento talent cumnaţi cognati cugnaz plural funţi libbre liris morţi le morti mwarz talenţi talenti talenz

All Romanian examples fit easily to respective Friulians, because they are based on the same Latin root, but follow morphological different paths to realize the plural. The article le in Italian expres- sion “le morti”, as translation of Romanian morţi is out of place, because the Romanian form is not articulated, as we will see later.

31 T. Ferro, Latino e lingue balcaniche nella formazione del rumeno, cit., p. 260 and foll.

77 [XV]. Quanto agli articoli che sembrerebbero ostacolo insormontabile a chi vuol comparare questi due idiomi, perchè l’articolo valaco viene posposto ai nomi e nel friulano li precede, diremo, che appunto nel friulano si scorgono ancora delle traccie di questo articolo posposto e ne daremo degli esempi: Ssoare dice il valaco per dire Sole; coll’articolo Ssoarele, ecco spiegato il Soreli friulano. Credo pure in friulano possa dirsi egualmente ploë e ploje, ed ecco che il valaco per dir pioggia coll’articolo fa ploa-ja ed anche ploja come pure da voe, - voglia fa vo-ja; il friulano pure può dire voe e voje; e così di molti altri. Sstëä dice il valaco alla Stella, e non è che coll’articolo genitivo che fa Sstëä–lei, donde potrebbe dedursi il friulano Stele. Riesce al certo Stele più affine a Stella italiano e latino, ma fo’ per ispiegare come meglio vedrassi in seguito, queste e pronunziate friulane, nel singolare, che se anche esistono nel francese sono però mute.

Romanian Italian Friulian without article with article without article with article Ssoare Ssoarele Sole - Soreli - ploa-ja, ploja pioggia ploë ploje voe vo-ja voglia voe voje Sstëä Sstëä-lei Stella - Stele

Molnar’s Romanian without article with article

- плоа- /ploa-ia (p.19) воїє/voie (p.19) во- /vo-ia стћ/sstea (p.18) стћ-лєи/sstea-lei (p.23)

Cattaneo had been more explicit about the phenomenon of postponed article: “È singolare ed esclusiva proprietà del valaco che mentre l’italiano, il francese e lo spagnolo prepongono l’articolo, il valaco lo pospone ai nomi, anzi lo inserisce tra il nome e l’aggettivo.” Not only, seeking the explanation, he proves to know that “L’uso di posporre gli articoli è commune a qualche altro linguaggio, al basco, per esempio e all’islandese; ed eziandio ai linguaggi di due popoli confinanti col valaco, cioè l’albanese e il bulgarico.” 32 Romanian soarele is the articulated form of soare, regular outcome of SOLE(M). Friulian sore:li ‘sun’ derives instead from *SOLĬCULU(M), with rhotacism of liquid L > r, tonic Ĭ with romance result e, sonorization of intervocalic voiceless velar C > G (romance phenomenon which however does not involve the Romanian and Central-Southern Italian)33 and apocope of final -U.

32 C. Cattaneo, Del nesso tra la lingua valaca e l’italiana, cit., p. 226. 33 V. Väänänen, Introduzione al latino volgare, cit., p.105, par. 104.

78 The syncope of interconsonantal -U-, ie *SOLĬCULU(M) > *SOLĬGLU(M), occurs only after the voicing of intervocalic consonant C > G, resulting in the form *soregl, which must be supported by the support vowel -i, due to impractical pronunciation of the consonantal nexus -gl remained free at the end of word. At this point, it triggers the phenomenon of increased of tonic vowel with its lenition of postonic voiced consonant g, which is led to grade zero. The final outcome therefore will have got the form sore:li ‘sun’, where it is easily seen the last syllable -li represents no article, but the Friulian regular result of the last syllable of *SOLĬCULU(M), as in GENŬCULU(M) > *genogli > geno:li ‘knee’. As in all Romance languages, except Romanian, the Friulian article is always put before the noun and has the form il. In ancient times it was lu < ĬLLU(M), of which it continues the second sylla- ble, for the singular masculine. It has the form ju for the plural, formerly gli < ILLI, with palatali- zation of L conditioned by final -I. The old singular article lu, corresponding to basic form of Tus- can lo, has passed, through the stage post-vocalic apocopate l and with subsequent vocalic prosthe- sis in initial position, to il, alternating with l in postvocalic context and finally also occupying the contexts before reserved only to lu. The ancient plural glu [Yu], who coexisted with the variant gli [Yi] as early as the fourteenth cen- tury, passes to ju when the Friulian loses the use of the phoneme /Y/ of palatal lateral (which is re- placed with /j/), alternating with the variant i in the contexts with apocope. The most common form of singular feminine article is la < ILLA(M), of which it always consti- tutes the second syllable, alternating with the less common form le. For the feminine plural we find only the form lis < las < ILLAS in central Friulian (li in Goritian), on analogy of the plural ending -lis of feminine nouns.34 With this brief explanation of Friulian article35 we can finally exclude that feminine plural article li (without final -s) of Goritian, though much resembling the final syllable -li of sore:li, is the post- poned article of it, if only for the contrast of gender and number: sore:li is masculine singular against article li, which is plural feminine. The real articulated form of sore:li remains therefore il sore:li ‘the sun’. On page 19 of his grammar Molnar indicates the articulated forms ploa-ia and vo-ia internally separated by a hyphen, like to highlight that the postponed article is -ia. Instead, we know that in Romanian the article -a of feminine replaces the final vowel -e when it is preceded by vowel or semivowel, then ploaie plus -a gives ploaia ‘the rain’ and voie plus -a gives voia ‘the will’. The Friulian form ploie, cited by Ascoli, must be the earliest phase of present ploe, which loses the etymological jod according to the evolution PLŬVIA(M) > ploie > ploe. None of these forms (and even voie and voe) includes a trace of postponed article. In the Romanian forms ending in stressed -a, as stea, the article -a of feminine is added through an epenthetic semivocalic -u, from which we have the articulated form steaua ‘the star’, that Ascoli however ignores for considering instead the genitive case stelei fully in harmony, he was thinking, with the Friulian stele, which instead derives regularly from Latin STĒLLA(M). He, however, cor- rects himself at the end of the passage saying that stele is more akin to Italian and Latin stella.

34 G. Francescato, Dialettologia friulana, cit., p. 71. 35 All information relating to Friulian article are taken from L. Vanelli, I dialetti italiani settentrionali nel panorama romanzo, cit.

79 He mentions immediately afterwards a timid comparative analysis about quality of the final vowel Friulian -e, saying it is the same as the French, where, however, we have a more centralized execution (“muta”). This consideration does not help to suggest to him that both vowels are two different outcomes of the same vowel final Latin -A. Luzzatto’s response on this point is clear and firm: “In quanto all’e finale femminino, esso è proprio del friulano insieme col francese. Ma il valaco conserva spessissimo l’a latino”.36 In the case of the Romanian genitive stelei Ascoli reaches to touch a phenomenon of great impor- tance, such as the conservation of genitive, but does not confer to it the due linguistics importance. I realize that in the long run is pesky be pitiless on the poor Graziadio citing the continuing linguistic statements of Cattaneo, but even in this case he had clearly identified the singular phenomenon expressing a very precise analysis: “In italiano sono rimasti pochissimi genitivi; diciamo per esempio la cui casa, l’altrui vita, la costui, la costei. Ma presso i Valachi questa inflessione è universale, e vi si distinguono sempre i generi: acestui, acestei, acestor (di costui, di costei, di costoro); unui, unei, unor; altui, altei, altor, ec. Ne proviene pertanto una specie di declinazione, per mezzo dell’articolo posposto ai nomi. Il nominativo e l’accusativo fan esempio campu-l; il genitivo e il dativo campu-lui; il nominativo e l’accusativo plurale fa campi-i; gli altri casi campi-lor. E nei femminili il nome capra’ fa coll’articolo capra; nel genitivo e dativo capr-ei; nel plurale capre-le e nel genitivo e dativo capre-lor. Forse anche le declinazioni delle lingue latina, greca, gotica in origine non furono altro che semplici nomi con un articolo affisso. Onde questa, ch’è la più notabile proprietà della lingua valaca, è cosa di molto momento nell’istoria generale delle lingue.” 37 The table below brings together all the correct forms:

Romanian Italian Friulian soare sole sore:li without ploaie pioggia ploe article voie voglia voe stea stella stele soarele il sole il sore:li with article ploaia la pioggia la ploe voia la voglia la voe stelei (gen.) della stella da la stele (gen.)

We can conclude the comment to passage [XV] of Ascoli criticizing him hotly. In fact, we find his reflection on the presence of Friulian traces of an postponed article totally without foundation. May have contributed to this result, in addition to an apparent lack of adequate linguistic tools, even an analysis certainly rushed and superficial.

36 S. D. Luzzatto, Epistolario, cit., p. 490. 37 C. Cattaneo, Del nesso tra la lingua valaca e l’italiana, cit., p. 228.

80 [XVI]. Il valaco come il friulano ammette per regole generali che i nomi che terminano in a ed e sono femminini, in una consonante maschili; abbenchè in ambo gl’idiomi vi sieno molte eccezioni.

Romanian Friulian masc. -consonante -consonante femin. -a, -e -a, -e

In his observation Ascoli does not mention at all the neuter gender, an other element of Latin con- servation that distinguishes the Romanian within the Romance field. The Romanian nouns of neu- ter gender call the things and inanimate objects, such as cap ‘head’ (pl. capete), braţ ‘arm’ (pl. braţe), fel ‘sort (pl. feluri) and so on. The characteristic of these neuters is that they behave in the singular as the masculine nouns and in the plural as the feminine, as in Italian (il) ciglio, (le) ciglia. The Romanian masculine nouns (and adjectives) are characterized by endings in consonant, as cumnat ‘brother in law’, in vowel -u, as socru ‘father in law’, in semi-vowel -w, as leu ‘lion’, in semi-vowel -j, as pui ‘chick’ and in vowel -e, as arbore ‘tree’. The same desinences are also valid for the neuter, as in calendar ‘calendar’, muzeu ‘museum’, stilou ‘fountain pen’, cui ‘nail’ and nume ‘name’. The Romanian feminine nouns are characterized by the vowel -a, as moartă ‘dead’, by the stressed vowel -a, as stea ‘star’, by vowel -e, as moarte ‘death’ and exceptionally by the vowel -i, as zi ‘day’. As for the endings of Friulian masculine nouns we can add to those consonant, as cugnat ‘brother in law’, even those vowel, such as sore:li ‘sun’, zeno:li ‘knee’ and many other words ending in -i as mestri ‘teacher’, salvadi ‘wild’, ne:ri ‘black’, where the final -i may be an etymological i, as in salvadi < SĬLVATICU(M), but more simply a support vowel -i, to close the complex consonant nexuses, such as -tr of mestri < MAGISTRU(M) and -gr of ne:ri < NĬGRU(M). The Friulian feminine is also characterized by nouns ending in consonant as su:r ‘sister’ and bru:t ‘daughter in law’ and by vowel different from -e, like in ma:ri ‘mother’, where we find the support vowel -i, typical of Friulian. The more complete table would have been:

DESINENCES OF THE NOUNS (C = consonant) Romanian Friulian masculine -C, -u, -w, -j, -e -C, -i feminine -ă, -a, -e, -i -C, -e, -i neuter -C, -u, -w, -j, -e -

81 [XVII]. Anche gli aggettivi con una t in fine la cambiano al plurale in s ovvero z come in friulano. Esempio valaco: sskurt, sskurzi, - corto, corti, a grande somiglianza del friulano curt, curz. Alcuni nomi valachi, con d finale, cambiano d in si. Esempio. Leopard, - Leopardo, - Leoparsi. Pure il friulano farebbe Leopars.

Romanian Italian Friulian sing. scurt leopard corto leopardo curt - plur. scurzi leoparsi corti - curz leopars

Molnar’s Romanian скŏрт/sskurt (p.36) лєопард/leopard (p.42) скŏрци/sskurzi -Zи/-si

The subject of nouns that change in the plural the consonant t in z has already been discussed by Ascoli in the passage [XIV], where he had better treat even the adjectives which are involved by the same phenomenon. Or at least he could write a note for reference, artifice that the young Go- ritian scholar uses only twice in his essay. For a full examination of this topic therefore I refer to the comment of passage [XIV]. So the Romanian form leopard cannot do “leoparsi” in the plural, but leoparzi, unless you interpret the s of “leoparsi” as a graphic version of the voiced sibilant [z]. Although not specifically mentioned by Ascoli, the Friulian form *leopard cannot exist, because we have already said that the voiced consonants at the end of word undergo the effect of devoicing, by switching to the form leopart. As we have seen, the Friulian plural is a bit more complex than that Romanian and this determines a larger series of consonant desinences in general, because you have to take into account the out- come of consonants followed by sibilant (for the sigmatic plural) and the result of consonants fol- lowed by the i of palatal plural. The construction of sigmatic plural generates almost no difficulty, given that is added the regular sibilant /s/ to singular nouns ending in voiceless velar [k], as in bosc > boscs ‘forests’, in labial na- sal [m], fum > fums ‘smokes’, in velar nasal [ŋ], furlan > furlans ‘Friulians’, in palatal nasal [<], pugn > pugns ‘punches’, in voiceless labial [p], intop > intops ‘unexpected’, in vibrant [r], pùar > pùars ‘poor’. For the forms ending in voiceless labiodental [f], u:f > u:s ‘eggs’, it can occur the fall of fricative f, while for those ending already in sibilant [s] there is no change, as in vo:s, that so remains also in the plural (but can alternate with [vo:S]). The only real change occurs for the forms ending in voiceless dental [t], which summed to sibilant [s] of plural passes to affricate [ts], as in plat > plaz ‘dishes’. The Friulian nouns that instead have the palatal plural they highlight a number of specific out- comes that relate to the final consonant -l, -n, -t, and the endings -st and -li. The lateral [l] + [j] in ancient times had as result the phoneme /Y/, then replaced by /j/, as in metàl > metài ‘metals’, the nasal [n] + [j] > [<], an > agn [a<] ‘years’, the nexus [t] + [j] > [c], insec’ [insEc] ‘bugs’, the nexus [st] + [j] > [sc], artist > artisc’ [artisc] ‘artists’ and the ending li + j > i (as well as lis), as in sore:li

82 > sorèi (and sore:lis) and zeno:li > zenòi (and zeno:lis). In sorèi, we note that the jodization of ending -li has neutralized the lengthening of tonic e, phenomenon that does not occur instead in the sigmatized form sore:lis. So:

Romanian Italian Friulian sing. scurt leopard corto leopardo curt leopart plur. scurzi leoparzi corti leopardi curz leoparz

[XVIII]. Oltre dei nomi terminanti in le (vedi pag. ant.) altri della desinenza el fanno il plurale in ei precisamente come il friulano. Esempio: vizel, vizei, - vitello, vitelli; il friulano vigiel, vigiei, (vidiel, vidiei).

Molnar’s Romanian Romanian Italian Friulian sing. вицєл/vizel (p.46) vizel vitello vigiel (vidiel) plur. вицъи/vizėi vizei vitelli vigiei (vidiei)

In this part Ascoli has noticed the similar results, which are due in the same phonetic contexts to action of the palatalization, both in Romanian and in Friulian. All these forms derived from Latin VITĔLLU(M) (see Plautus), diminutive of VITULU(M), con- tinue what has already been said about moale and piele. In the path from VITĔLLU(M) to Roma- nian viţel there has been the shortening of geminate liquid LL > l, the apocope of -U, the diph- thongization of Ĕ > iè, reaching the outcome *vitiel, followed by notch of dental t > ţ, conditioned by jod of diphthong iè, with outcome *viţiel and subsequent fall of i semi-consonantal. Even the -i of plural palatal contributes to palatalization of degeminated liquid l + j, which has the result i, as in viţei. In addition to the form vigjel Ascoli also adds vidjel to the Friulian correspondents of ‘calf’. And in these two forms is possible seeing the diachrony of VITĔLLU(M) > vidjel > vigjel [viFel], with the voiced dental d that palatalizes in g [F] (as also in gjàul < DIABOLUM), as a result of the j of diphthong jè, normal development of Ĕ in weak position. So:

Romanian Italian Friulian sing. viţel vitello vidjel, vigjel plur. viţăi vitelli vidjei, vigjei

83 [XIX]. Osservisi, che quei molti nomi valachi colle desinenze orju ed arju, sono della stessa specie, e molte volte corrispondenti, a quei friulani in ori, ar ed ari.

Romanian Italian Friulian - orju - - ori - arju - - ar, - ari

All Romanian nouns and Friulian with suffix built on the nexus R + j derive from the Latin nouns ending in -ORIUM and -ARIUM, which generated a large number of Romance outcomes. The suffix -ORIUM gave rise to Romanian -or and -oriu, to Italian -oio38 and Friulian -ori, as we can see in the examples I propose:

SUFFIXES FROM -ORIUM Latin Romanian Italian Friulian LABORATORIU(M) laborator - laboratori SANATORIU(M) sanatoriu - sanatori

The suffix -ARIUM gave rise to Romanian -ar, -arie and -ariu, to Italian -aio, -aro (and the Ve- netian -er) and to Friulian -ari and -a:r, where the apocope of final -i is compensated by the lengthening of tonic vowel, as we can see in the examples that I propose in the following scheme:

SUFFIXES FROM -ARIUM Latin Romanian Italian Friulian ADVERSARIU(M) adversar - riversari NŎTĀRIU(M) notar notaio noda:r IANUARIU(M) ianuarie gennaio gena:r SALARIU(M) salariu - salari

But we must point out that most of the Romanian forms in -ariu are presented in modern dictionar- ies with the suffix -ar simplified. For the Friulian are correct the suffixes -ori, -ar and -ari cited by Ascoli. It follows, therefore, the following chart:

Latin Romanian Italian Friulian -ARIUM -ar -aio, -ario, -aro -ari, -a:r -ORIUM -oriu, -or -orio -ori

38 The Italian siffix -orio (and -ario) is a cultism.

84 [XX]. I nomi valachi che hanno nell’ultima sillaba ae ne fanno un ei: betae, - battaglia, plurale be- tei, ed il friulano batae, batais.

Molnar’s Romanian Romanian Italian Friulian sing. бътає/bėtae (p.64) betae battaglia batae plur. бътъи/ bėtėi betei - batais

Here, too, the write of Molnar of “betae” is bătae (modern bătălie) and in the plural is not “betei”, but bătăi. I have already pointed out in section 4.1 (The handwriting Romanian of Ascoli) that he dis- tinguish never the mid-vowel e by the central-vowel ă. In the Friulian batais it does not occur the palatalization of tonic a, because in this case they use the sigmatic plural. So:

Romanian Italian Friulian sing. bătae battaglia batae plur. bătăi - batais

[XXI]. Tutti i molti nomi valachi in je cangiano l’e finale in i, esempj: sentenzie, - sentenza, - senten- zii; quitanzie, - quitanza, - quitanzii; restanzie, - restanza, - restanzii; corrispondenti al friulano sen- tenzie sentenziis; quitanzie, quitanziis: restanzie, restanziis.

Romanian Italian Friulian singular plural singular plural singular plural sentenzie sentenzii sentenza - sentenzie sentenziis quitanzie quitanzii quitanza - quitanzie quitanziis restanzie restanzii restanza - restanzie restanziis

Molnar’s Romanian singular plural сєнтєнцїє/ssentenzie (p.68) - квитанцїє/kvitanzie (p.65) - рєстанцїє/resstanzie (p.65) -

In Romanian the feminine nouns ending in the singular in -e form the plural by replacing this vowel with an -i. The forms cited from Ascoli - all latinisms entered into Romanian by different

85 ways (especially from French) - have the modern equivalents in sentinţă, chitanţă and restanţă, which in the plural close the vowel -ă in -e, resulting in sentinţe, chitanţe and restanţe. The dental reveal the presence of a j that has notched the dental occlusive (SENTENTIA > sentinţiă > sentinţă) and then has fallen, while the tonic E has been closed in i by action of the nasal.

Romanian Italian Friulian singular plural singular plural singular plural sentinţă sentinţe sentenza sentenze sentence -is chitanţă 39 chitanţe quietanza 39 quietanze restanţă restanţe restanza 40 restanze

[XXII]. I nomi valachi in ate scambiano in plurale l’ate coll’ezi, esempj: ssträmbetate, curva, alle volte ingiustizia ssträmbelezi; come il friulano strambetat strambetaz figurato per goffaggine. *) Troveremo pure il valaco dinte, dinzi, - dente, denti, abbastanza somigliante al friulano dint, ding. [note of Ascoli:] *) La lingua italiana non ha precisamente il sostantivo strambità.

Romanian Italian Friulian sing. ssträmbetate dinte curva-ingiustizia dente strambetat dint plur. ssträmbelezi dinzi - denti strambetaz ding

Molnar’s Romanian sing. стр мбътатє/sstræmbėtate (p.78) динтє/dinte (p.80) plur. -ъци/-ėzi -ци/-zi

Here, too, Ascoli considers again the phenomenon of plural palatal which affects the dental both in Romanian and in Friulian (see above the passages [XIV] and [XVII]): the phenomenon affects both languages, as Ascoli rightly notes. In Friulian the plural of dint cannot be ding, but dinc’ [dinc], because at the end of word it ex- cludes any sonority of obstruent consonants. In the last examples, rather than the development of plural, it is the common outcome of Latin nexus EN > in, for the raising of the vowel in front of nasal, which characterizes both the Roma- nian and Friulian: DENTE(M) > Rom. dinte, Friul. dint ‘dente’.

39 In analogy with the French quittance, from quitter ‘release from an obligation’. The modern quietanza was built on quieto. See M. Cortelazzo - P. Zolli, Dizionario etimologico della lingua italiana, cit., p. 1013. 40 Archaic, with meaning of ‘surplus, rest’.

86

Romanian Italian Friulian sing. strîmbătate dinte curva-ingiustizia dente strambetat dint plur. strîmbătăţi41 dinţi - denti strambetaz dinc’

[XXIII]. Le regole abbastanza importanti in una grammatica, per formare dagli aggettivi e nomi mascolini, aggettivi e nomi femminini, le troveremo identiche nelle due lingue, cioè: I. Aggiungendo al maschile dell’aggettivo una e lo rendesi femminino. Esempj grass, - grasso, - grasse, - grassa, - lung, - lungo, - lunge, - lunga, le quali voci dicendo, nessuno saprà s’io mi parli valaco ovvero friulano.

Molnar’s Romanian Romanian or Friulian? Italian masc. лŏнг/lung (p.104) grass lung lungo fem. лŏнгъ/lungė grasse lunge lunga

That of Ascoli, who perhaps is using the operation of “addendum” in the context of ‘quattuor spe- cies’,42 is a mechanical synchronic description. In reality he should have considered the inverse operation, because it is the formation of the masculine which depends by fall of final -U of GRASSU(M)43 and LONGU(M), while the feminine retains the etymological vowel -A, which is modified according to outcome of Romanian and Friulian, ie - ă and -e respectively.

Romanian Italian Friulian masc. gras lung grasso lungo gras lunk fem. grassă lungă grassa lunga grasse lunge

41 The l of “ssträmbelezi” is certainly a misprint. 42 L. Renzi, Come gli umanisti non scoprirono le leggi fonetiche, in Diaz-Rizzotto (edited by), Hommage a J. Brunet. Annales littéraires de Franche-Comté, Vol. 2, pages 571-584. 43 Graphic variant of CRĂSSU(M) for influence of Grossu(M). See M. Cortelazzo - P. Zolli, DELI, cit., p. 517.

87 [XXIV]. II. I nomi sostantivi qualificanti un grado riduconsi femminini aggiungendo ëässe ed alle volte semplicemente esse: Esempj valachi baron, - barone, - baronesse; - konte, - conte, - kontesse, ed anche qui i due idiomi si fondono quasi in un solo.

Romanian Italian Friulian masc. baron konte barone conte baron konte fem. baronesse kontesse - - baronesse kontesse

Molnar’s Romanian (p.393) бapoн/baron koнtє/konte бapoнєcъ/baronessė koнtєсъ/kontessė

The feminine forms in Italian -essa derive from the medieval Latin in -ISSA, such as COMITĬSSA(M), formed from Latin CŎMITE(M) ‘companion’ (later ‘high imperial official’) with the suffix -ISSA typical of titles, as contessa, baronessa, and so on.44 All forms reported by Ascoli therefore derive from a common source Medieval. In Romanian, however, the feminine in -essă are loans from modern French, while in Friulian the feminine in - esse might be loans from Italian. In the Romanian of today, however, the feminine of baron is baroană.

Romanian Italian Friulian masc. baron conte barone conte baron conte fem. baronessă contessă baronessa contessa baronesse contesse

44 M.Cortelazzo - P.Zolli, DELI, cit., 1, p. 273.

88 [XXV]. E se poi guarderai, quali gli eccettuati sieno, che per esprimere il femminino abbiano una voce propria, più ancora ne resterai stupito, scorgendo principali fra loro: pork, - porco, femminino skroafe; käne, keza; dove trovi evidentemente scrove, e chizze dei friulani. Nessun’idea di questo chizze scorgesi nel francese o nell’italiano.

Molnar’s Romanian (p.105) Romanian Italian Friulian m. порк/pork k нє/kæne pork käne porco - - f. сκроафъ/sskroafė къца/kėza scroafe keza - scrove chizze

Despite the different phonetic changes, you can easily identify that the feminine correspondences of porc and cîine date back to a common source Latin and this determines their affinity, both in Romanian and in Friulian. They are lexicalized forms of distinction between masculine and femi- nine, present to some extent in all languages. From Latin PŎRCU(M), SCRŌFA(M), CANE(M) derive both Romanian forms porc, scroafă and cîine, and Friulian forms purcit (dim.), scrove and chian [caŋ], which correspond to Italian porco, scrofa e cane. The Romanian form “keza” presented by Ascoli, corresponds to modern write căţa, as best shown by Molnar’s Cyrilllic/Latin, “къца/kėza”, where the final -a of căţa is tonic, as in cafea. The modern form căţea conserves (or has reactivated) the semi-consonantal element of diphthong ea [ýa], which led to etching of dental t > ţ (but the form căţa, reported by Molnar, is definitely an ac- tual pronunciation). As I have already explained in paragraph IV.2.2.1 the Cyrillic sign ц is exclusively the voiceless dental affricate [ts], that Molnar explains with Romanian example “цïe zie” followed by the Ger- man translation “dir”, Italian ‘a te’. It can therefore be ruled out that it implies the vowel e. The Romanian căţa is the feminine of căţel ‘doggie’ < Lat. CATELLU(M), perhaps through the form *căţelea. The Friulian form “chizze” is cizze [tSitse] (see Pirona) ‘bitch’ (currently cice [tSitSe]). Noting that “nessun’idea di questo chizze scorgesi nel francese”, for the first time Ascoli calls into question the comparison with French, but without bringing examples. With the enlargement to French lan- guage, in addition to languages already considered so far by the young Goritian scholar, that is, Romanian, Italian and Friulian, he outlines at least a segment of the Romance framework already made in 1821 by François Raynouard with his Grammaire comparée des langues latines, and Friedrich Diez in 1840, with the Grammatik der romanischen Sprachen, all scholars still unknown to Ascoli.45

Romanian Italian Friulian masc. porc cîine porco cane purcit chian fem. scroafă căţea scrofa cagna scrove chice

45 The biographical summaries of F. Raynouard and F. Diez are presented in par. III.2.1.

89 [XXVI]. Avanzati pure, cerca come i valachi formino il diminutivo, e troverai che lo formano aggiungendo al nome uz o zel; cosa che non riuscirà nuova ad orecchio friulano. [see p. 106 of Molnar]

The topic covered by this passage of Ascoli is on page 106 of the Molnar’s Grammar (as I indi- cated and I will indicate in square brackets at the end of each passage), from which Ascoli derives, not without flaws as we shall see, a correspondence between Friulian and Romanian for which he not give examples (but I shall explain the argument). Like all Romance languages except French, Romanian is a language that has morphological proc- esses of diminutivization, not less than Friulian.46 To form the diminutive it makes use of a num- ber of suffixes that can be added to nouns, to adjectives and even to as the suffix -as (Ion ‘John’ >) Ionaş, (băiat ‘boy’ >) băieţaş, as -el of Ionel, (mic ‘small’ >) mitiţel, băieţel, -uţ of Ionuţ, (pat ‘bed’ >) pătuţ, as -şor of (lucru ‘work’ >) lucruşor, (rîu ‘bad’ >) rîuşor, as -ţă of (Ana ‘Anna’ >) Aniţă, Ioniţă, (casă ‘house’ >) căsuţă and others, not to mention the possibility of forming the diminutive of diminutive, as in cîine > căţel > căţeluş. There are, as in Italian, diminutives even of adverbs, as încetişor, ital. ‘pianino’. The Friulian form the diminutive adding to the singular noun the suffixes -ut for the masculine and -ute for the feminine, -uz, and -utis in the plural. As representative examples I can remember the forms omenut ‘little man’, chiasute ‘little house’, and omenuz ‘little men’, chiasutis ‘little houses’. The Friulian form -uz (to which seems alluding Ascoli) results from -ut + s > -uts > -uz and is used for the masculine plural of diminutives. From the Latin diminutive suffix -(C)ELLUS, as in ASELLUS < *ASIN(O)LOS, AGELLUS < *AGR(O)LOS, TABELLA < *TABEL(O)LA and ANELLUS for ANULUS, PORCELLUS for PORCULUS47 you get to the Proto-Romance suffix -ELLU(M), from which the Romanian diminutive -(j)el of băieţel (< băiat), Ionel (< Ion) and Friulian -el, as in pwarte ‘door > pwartel (but, as we shall see, probably through the Italian). The voiceless dental affricate of the Romanian suffix “uz” (see above Ionut, pătuţ) is the result of an etymological dental occlusive followed by a final -i, which then fell, -uti > -uţi > - uţ, while “zel” implies an ancient -t + iel, where the dental is the final consonant of the word you want to diminutivize (and thus not part of suffix) and ie is the spontaneous diphthongization of tonic Ĕ of - ĔLLU(M). It also follows here the etching of dental and the disappearance of jod. The real suffix that is added to the Romanian noun for the diminutive form, is not - ţel, as suggested by Ascoli, but -el. In Friulian does not occur the diphthongization of tonic Ĕ of -ĔLLU(M) in jè, as it should have been, being in the context of weak position, because in all likelihood the use of suffix -ĔLLU(M) in diminutive function took place at a later date, perhaps by analogy with the Italian or the Ve- netian. In any case, the nexus t + j in Friulian could never have given [ts], but [dj] when the conso- nant is placed in intervocalic position and then eventually [F], as already seen in the example of VITELLU(M) > vidjel > vigjel [viFel].

46 L. Renzi, Nuova introduzione alla filologia romanza, cit., par. 12, p. 193 and foll. 47 V. Väänänen, Introduzione al latino volgare, cit., p.162, par. 189.

90 When the dental had been word-initial or preceded by another consonant, the nexus would have remained [tj], as in tjere (and tjare) or it could have gone to palatal [c] of cere [cEre] ‘earth’ and as in the plural of dint > dinc’ [dinc] ‘teeth’ and of dut’ > duc’ [duc] ‘all’.

[XXVII]. I valachi formano Comparativo e Superlativo coll’ajuto d’Avverbj; il friulano fa sempre così nel Comparativo e quanto al Superlativo credo che l’issim dal latino issimus nel friulano non suoni molto bene. Nelle belle poesie friulane del Zorutt, molto di rado parmi aver trovato un superlativo a tal foggia. [see p. 106 of Molnar]

Already Classical Latin had periphrastic comparatives with MAGIS and later (by imitation of Greek πλέον) with PLUS48 next to the organic comparatives in -IOR, -IUS. The Romance languages continue only the analytical forms, which can be divided, for the positive degree, among the elements MAGIS and PLUS. MAGIS established itself as the only comparative form in the peripheral Romania, as in Dacia (mai) and in the Iberian area (sp. màs, port. mais, cat. més). The central Romania opted instead to build with PLUS, as in French and Provencal plus, in the outcome Italian più and Friulian plui.49 The Romanian is forming the comparative by prefixing to adjective the element mai for the positive degree. At page 107 Molnar presents only the examples “mai frumos, mai mare, mai mik” and “mai énvézat” (= învăţat ‘educated’), that, however, Ascoli does not report, and it’s clear he thinks only to first term of majority comparative: I do not report then the second term, that he does not deal. Even in Friulian is expressed the comparative by prefixing the element plui. To express the absolute superlative the Latin used the suffix -ISSIMUS, but also a set of adverbs as FORTITER, MULTUM, BENE, TRANS and others too, which are precisely those inherited from Romance languages, such as in Romanian foarte bun, in Italian molto bello, in French bien peu, très beau, ecc.50 The suffix -ISSIMUS was reintroduced in Italian, Spanish and Portuguese,51 and is therefore an element of learned language, which does not appears instead in most of Romance varieties, which do not have the re-introduction of -ISSIMUS, so that the agreement Friulian and Romanian is only in normal Latin heritage.

48 G. Rohlfs, Grammatica storica della lingua italiana e dei suoi dialetti, Morfologia, cit., p. 83. 49 V. Väänänen, Introduzione al latino volgare, cit., p. 207 and foll. 50 V. Väänänen, Introduzione al latino volgare, cit., p. 208. 51 V. Väänänen, Introduzione al latino volgare, cit., p. 208. W. Meyer-Lübke, Grammatica delle lingue romanze, II, Lipsia, 1899, p. 86.

91 In Romanian the absolute superlative can be expressed with the use of the foarte ‘strong’, but also prea (the only one used by Molnar) ‘much’, ‘too’ and also extraordinar ‘extraordinary’, grozav ‘terribly’ and many others, followed by preposition de and the adjective bun (for the posi- tive form) or puţin (in the negative form), as o prea frumoasă fată ‘a beautiful girl’ (Eminescu, Luceafărul), extraordinar de bun ‘extraordinarily good’, ‘very good’. In Friulian instead the absolute superlative is expressed by use of the augmentative -on, as in grandon ‘huge’. The Latin suffix -ISSIMUS is continued therefore neither by Romanian nor by Friulian, as he rightly observed Ascoli. With the statement that “issimus in friulano non suoni molto bene” Ascoli has certainly meant to say that use of suffix would be in Friulian a extraneous influence, certainly Italian. The poet “Zorutt” cannot be other than Pietro Zorutti, presented in a footnote in section IV.3.1.

[XXVIII]. Nè s’arresteranno ai pronomi i nostri ritrovati che anzi molti dei valachi scorgeremo infrà i friulani. Esempj: il femminino valaco a jei, jei, non sarà nuovo a chi conosce il friulano je; il dativo maschile plurale lor valaco, ed il singolare lui neppur nuovi riusciranno.

Molnar’s Romanian Romanian Friulian

sing. а єи/a ei (p.128) a jei, jei je

лŏи/lui (p.127) lui lui plur. лор/lor (p.127) lor lor

All pronouns here considered from Ascoli belong to series of pronouns ‘free’ and represent the Romanian outcome and Friulian of dative pronouns derived from *ILLAEI,52 ILLUI and IL- LŌRU(M). *ILLAEI, which is the vernacular form instead of usual ILLI, gave the Romanian re- sult ei and Friulian je. ILLUI, which appears in the imperial era,53 passes to Romanian and Friulian lui. ILLŌRU(M) > Rom. lor and Friul. lo:r, because the Friulian Ō toned is located in the context of strong position and therefore becomes longer. In all these cases the Romance languages continue always the second part of Latin pronouns: the etymological basis for the Romance lan- guages are so *(IL)LAEI, (IL)LUI, (IL)LORU(M). In order to justify the Romanian outcome ei we can assume the path *(IL)LAEI > lei > ei, where the initial l of lei was palatalized by one of following vowels and then vocalized, as LEPORE(M) > l’epure > iepure.54 You can also think to alternative outcome *(IL)LAEI > lèi > lièi, with relating

52 CIL VI, 14 and 484; ILLEI, Egitto, II secolo. V. Väänänen, Introduzione al latino volgare, cit., p. 213. 53 A. Ernout, Morphologie historique du latin, Parigi, 1953, p.132. 54 F. Dimitrescu, Întroducere în fonetica istorică a limbii române, cit., p. 96.

92 palatalization of the initial lateral, conditioned by the jod of diphthong or from the final -i, and apheresis following. From the phonetic point of view it cannot be excluded either that the personal pronoun dative Romanian ei derives for direct transmission from the equivalent Latin EI, with a pronunciation Slavicized [ýei], but we know that the pronoun IS, EA, ID is passed away early from Latin and this hypothesis for this reason is to be set aside, given also that the masculine cor- responding is lui < (IL)LUI. The Friulian path of *(IL)LAEI may be continued from *(IL)LĔI, where the Ĕ in weak position diphthongizes in je giving rise to ljei, followed by palatalization of the initial liquid l because con- ditioned by jod, resulting in Yei. When Friulian loses the use of phoneme /Y/ of palatal lateral (which is replaced with /j/) the form Yei passes to jei. At this point the final -i is interpreted as a semi-consonantal element, whereby we have jej, in which -ej emerges as descendant diphthong, which is simplified, giving rise to je. Ascoli doesn’t notice a syntactic important feature: only the Romanian retains the Latin synthetic form, while the Friulian has made the innovation of Romance domain for the oblique cases, devel- oping the analytic form with use of preposition a, that you need to premise therefore to “je, lui, lor”. For the analysis of the initial vowel e of Romanian ei [ýei], which is pronounced as a diphthong [ýe] to Slavic influence,55 I send back to my comment about the passage [III] of Ascoli.

rumeno friulano sing. ei a je lui a lui plur. lor a lo:r

55 See even F. Dimitrescu, Întroducere în fonetica, cit., p. 80.

93 [XXIX]. Ed i possessivi almieu, alteu, alseu, alnostru, alvostru, - il mio, il tuo, il suo, il nostro, il vostro, reggeranno ad un bel confronto coi friulani al miò, al to, al so, al nestri, al uestri; ed i plurali ancor meglio, perchè vedremo nel valaco ai miei, ai tei, ai ssei, - i miei, i tuoi, i suoi, che quasi s’identificano coi friulani dello stesso numero. [see p.128 of Molnar]

POSSESSIVES Romanian Italian Friulian almieu il mio al miò alteu il tuo al to singular alseu il suo al so alnostru il nostro al nestri alvostru il vostro al vestri ai miei i miei ~ ai miei plural ai tei i tuoi ~ ai tei ai ssei i suoi ~ ai ssei

Molnar’s Romanian singular plural алмїєŏ/almieu аи мїєи/ai miei алтъŏ/altėu аи тъи/ai tėi алсъŏ/alsėu аи съи/ai ssėi алнострŏ/alnostru - алвострŏ/alvostru -

The Romanian possessive, if does not follow the noun (eg. fratele meu ‘brother-my’), requires the mandatory presence of so-called genitival articles (or possessive) al (< alu < ILLU), a (< ILLA), ai (< ILLI) e ale (< ILLAE), which precede the possessive adjective, but separate, not only when it is pronoun (as in Italian and Friulian), but even if the noun is not given, as in citesc o carte a ta ‘I’m reading one of your books’ (against the articulate expression citesc cartea ta ‘I’m reading your book’).56 The other case where the possessive may be preceded by the Romanian genitival occurs when the noun, though determined, is followed by an element that intrudes between the noun and the geniti- val article as casa nouă a mea ‘my new house’ (in the absence of nouă you would have had casa mea ‘my house’).

56 B. Cazacu et alii, Cours de langue roumaine. Introduction à l’étude du roumain à l’usage des étudiants étrangers, Bucarest, 1967, p. 150 and foll.

94 In Friulian, as in Italian, the possessive is preceded by the definite articles il, la, i and lis when it is both adjective and pronoun.57 Ascoli however, in his examples (all masculine), uses the articles al and ai, which are variants of il and i. It should be noted, however, that in Romanian the use of arti- cle with the pronoun is larger than in Friulian.

MASCULINE POSSESSIVES Romanian Italian Friulian al meu il mio il mjò, il <ò singular al tău il tuo il tò al său il suo il sò al nostru il nostro il nestri al vostru il vostro il vestri ai mei i miei ~ i miei plural ai tăi i tuoi ~ i tiei ai săi i suoi ~ i siei

[XXX]. Se passeremo ai dimostrativi, troveremo pure di che soddisfarci: i valachi ċhest e ċhel, che bene sono rimpiazzati dai friulani chest e chel, massimamente dall’ultimo che ha nel plurale chei come appunto il ċhel valaco ha ċhei. [see pages 130 and 132 of Molnar]

DEMONSTRATIVES Molnar’s Romanian Romanian Friulian S. чєст/tschesst (p.130) чєл/tschel ćhest ćhel chest chel P. - чєн/tschei (p.132) - ćhei - chei

The Latin basis of demonstratives presented by Ascoli is ECCU-ISTU (ACCU for ECCU, accord- ing to some) or ATQUE ISTE both for Romanian “ċhest” and for Friulian “chest”, and ECCU-ILLU or ATQUE ILLE for the respective “ċhel” e “chel”.58

57 G. Marchetti, Lineamenti di grammatica friulana, cit., p. 228. 58 ATQUE ISTE and ATQUE ILLE in the language of Plautus were very frequent. See G. Rohlfs, Grammatica storica della lingua italiana e dei suoi dialetti, cit., p. 203, note n. 2.

95 The corresponding Italians would be questo and quello, if the latter is followed by a vowel (in front of which it elides), from x, z and nexuses s + consonant, gn, ps, otherwise we use the variant quel, when followed by all other consonants. In Old Italian there was also the form chesto59 and the reduced form esto, from which the modern Italian sto, like sto cretino, stamattina. Ion Molnar on page 130 of his grammar specifies clearly that the Romanian pronouns are “ачєст/acest, oder чєст/cest, oder ъст/ėst” and then “ачєл/acel oder чєл/cel”. Even in Old Friulian there were the forms achest, achel,60 where a- is derived from ĔCCU(M). In current Romanian the form cest was dropped to use only acest. Instead cel, in the constructions as Ion cel bun, and acel, they continue to be used both. The Friulian made instead the contrary choice. The table below compiles all modern forms, accompanied by some correspondence (gray) of Old Italian and Old Friulian.

DEMONSTRATIVES Romanian Old Italian Old Friulian Italian Friulian sing. acest acel chesto questo quel achest chest chel quello plur. - acei - questi quei achel - chei quegli quelli

Ascoli has well grasped the fact that there are not significant discrepancies between Romanian and Friulian. Both in fact possessed both forms with initial a- than those without, even if they made choices, in part, in different direction. A Romance general setting (by enriching the framework) would show that the type with initial a- is present in Spanish, the one without a- in Italian and French.61 So the similarity between Roma- nian and Friulian is merely a special case, of one between all Romance languages, for uniform derivation from Vulgar Latin.

59 G. Rohlfs, Grammatica storica della lingua ialiana. e dei suoi dialetti, cit., p. 203. 60 G. Marchetti, Lineamenti di grammatica friulana, cit., p. 227. 61 W. Meyer-Lübke, Grammatik der romanischen Sprachen, cit., III, p. 95.

96 [XXXI]. Nei pronomi interrogativi troveremo kare, che a prima vista ci sgomenterà, ma andiamo innanzi ed ecco che kare è pieghevole ed in dativo e genitivo ha kui? ed a kui? il pretto friulano, cui? [see p. 135 of Molnar]

INTERROGATIVE PRONOUNS Molnar’s Romanian Romanian Friulian nominative-accusative карє/kare kare - genitive а кŏи/a kui a kui? - dative кŏи/kui kui? cui?

The Romanian pronoun care derives from Latin QUALIS, with L > r for rhotacism. The form cui is the direct continuation of Latin dative, which remains the same in Italian as well as in Romanian and Friulian. The genitive of Romanian care is usually distinct by presence of the possessive da- tive article al, a, ai, ale. The Friulian pronoun cui, however, is also used in the nominative with value of chi. The affinity of the forms considered is indisputable, but as in the previous case (and in many other cases), it’s all due to the common Latin ancestor.

INTERROGATIVE PRONOUNS Romanian Italian Friulian nominative-accusative care quale - genitive a cui di cui - dative cui a cui cui

97 [XXXII]. Dite finalmente ad un valaco che conti, ed ei comincierà un, doi *) come farebbe ogni buon friulano. [note of Ascoli:] *) Per il numero tre i valachi tréi, e trei precisamente nell’alto Friuli. [see p. 141 of Molnar]

The form un, both Romanian and Italian and Friulian, continues the Latin UNUS. The Romanian and Friulian doi (such as the Italian due) continue the masculine DUO, through the phase DUĪ,62 while the Romanian feminine două and Friulian do:s continue the feminine DUAS. For the Romanian feminine, Molnar shows the form “доаw/doao”. Even Luzzatto, in his letter to Ascoli, writes “doao zeci” as number translation 20,63 probably drawing it from the grammar of Gheorge Şincai (Buda 1805).64 From the Latin TRES you have Romanian trei and Friulian (and Italian) tre. Trei, as asserts As- coli, is actually attested in .

[XXXIII]. Se cerchiamo fra altre parti del discorso, troveremo il niċhe un, pronome improprio, il friulano nanchie un; troveremo fra gli avverbj di tempo per il primo akum, - ora il vero acumò dei friulani; ċhe interrogativo, come il friulano tanto comune ce? [see pages 39, 137 and 314 of Molnar]

Molnar’s Romanian Romanian Friulian ничє оyн/nitsche un (p. 39) nićhe un nanchie un акоŏм/akum (p. 314) akum acumò чє/tsche (p. 137) ćhe ce?

The adverb “niċhe un” (now nici un) derives from NEC UNUM o NEQUE UNUM. Both Romanian acum and Friulian cumò (current version apheretic of ascolian acumò) could derive from Vulgar Latin *ECCU MODO. The form ce of the Romanian and Friulian interrogative pronoun is the regular result of Latin QUĬD. The labiovelar QU- was passed to velar occlusive [k] already in the early Middle Ages, when it appears from the sixth century a new element written que (quae, quem), which was cer- tainly pronounced ke and represented the outcome of QUID (as ca of QUIA). These two forms were alternating with QUOD, from which you had the Romanian outcome ca (and că) and Italian

62 Vaguely attested by later grammarians. See V. Väänänen, Introduzione al latino volgare, cit., p. 209. 63 S. D. Luzzatto, Epistolario, cit., p. 487. 64 See H. Tiktin, Rumänisch-Deutscher Wörterbuch, Bucarest, 1903, 3 vol., s. v. Among the forms reported by Tiktin there is also dóao next to doa, doo, doaoa, etc.

98 ko, as in Placito Capuano of 960.65 Beyond these graphic alternations, however, like QUI > it. chi, French qui, it is certain that the labiovelar was pronounced [k], already during the Proto-Romance epoch in many (all?) areas of Romània. The Ĭ of QUĬD in the meantime (or perhaps even before the of labiovelar QU-) was open in e closed, giving rise to Romance common result [k] in Italian, French, Spanish, etc. Con- tinues at this point the palatalization, both in Romanian and in Friulian (but not in Italian), of velar [k], generating the result chie [ce] and then ce [tSe].66 Ascoli noted, rightly, the phenomenon of palatalization of Romanian forms nici, ce and Friulian nanchie, ce, which is only of some Romance varieties.

Romanian Italian Friulian nici un neanche uno nanchie un acum - cumò ce che ce?

[XXXIV]. Quanto alla costruzione dei discorsi, il valaco lungi dall’avere le inversioni germaniche o latine, è piano affatto nella maggior parte delle sue frasi a mo’ del friulano, che forse non potrebbe dirsi dell’italiano, perchè l’italiano pure piegasi ad essere passabilmente inversivo. Qualche piccola frase offriremo in seguito.

Ascoli notes implicitly that Romanian and Friulian observe an unmarked word order, which we call SVO (= Subject, Verb, Object), as we can see in his Romanian example “lui nu jaste voe” and Friulian “lui a nol a voe” ‘he has not will’, while Italian (literary, still in use in his days) had still rever- sals, “ebbe il giovane visto...” (Boccaccio!). In fact, Friulian has the inversion of interrogative clitic pro- nouns, not noticed by Ascoli. And then Old Friulian had the same ‘reversals’ of German and Ro- mance Old languages.67

65 V. Väänänen, Introduzione al latino volgare, cit., p. 274. 66 For a complete discussion of [k] > [tS] in Romanian see. F. Dimitrescu, Introducere în fonetica, cit., p. 106 and foll. 67 L. Renzi, Nuova introduzione alla filologia romanza, cit., cap. 12.

99 [XXXV]. Ora passiamo a confrontare alcune delle voci più ovvie nella vita comune. E per il primo nome della nostra specie VALACO ITALIANO FRIULANO Om Uomo Om Sulla divisione del tempo, del giorno ecc. ssoarele il sole il soreli amiäsi mezzodì miesdì ssare sera sere aċhaste ssare questa sera cheste sere un an un anno un an an nou anno nuovo an nuv timp tempo timp lune luna lune aer aria aer boare vento forte buere sefir zeffiro zefir Nomi dei giorni della settimana mierkuri mercoledì miercui gioi giovedì joibe vineri venerdì vinars I due liquori più comuni vin vino vin ape acqua aghe Nomi d’animali raze anitra razze vizel vitello vigiel (vidiel) taur toro taur keza cagna chizze jepure lepre jeuur vake vacca vachie kapon capone chiapon leopard leopardo leopard urs orso ors scroafe scrofa scrove Nomi di metalli fier ferro fier plumb piombo plomb aur oro aur arċhint argento ariint

100 Osservisi la forma della parola, così tronca in fine; quelle e nel singolare, e vedrassi l’identico spirito delle due lingue.

As you can see Ascoli has a number of comparative lists, partitioned into lexical semantic fields, in which lists “alcune delle voci più ovvie nella vita comune” to show the similarities that are part - without As- coli telling - to a same Latin basis. Here, and in subsequent lists, he does not comment the corre- spondences that occur between the two languages under consideration. For this reason, I will ana- lyze only the cases that I believe more interesting from the point of view linguistics. From the forms of om ‘man’ he goes to the words concerning the “divisione del tempo”, in the days of week, liqueurs and metals. The correspondence noted by Ascoli, certainly on an intuitive basis, in the Rom. “amiäsi” (now amiază) and Friul. “miesdì” ‘noon’ leads the Romanian amiază to Latin AD MEDIAM DIEM > a-mia-ză, while the Friulian form miesdì (also misdi, see Pirona, s. v.) is the result of (AD) MEDIUM DIEM > (a)-mies-di. This is therefore two forms that derive from same Latin source and have the only difference in the initial a-.68 The recurrence of obvious phonetic regularities as in Rom. “vake, kapon” and Friul. “vachie, chiapon” (and up ahead - Romanian/Friulian - “kase/chiase, sporkat/sporchiat, pekat/pechiat, predikator/predichiador”) or as in couples “plumb/plomb” and “aur/aur”, should have attracted his attention on the elements that differen- tiate the two languages in question, represented by the Friulian palatalization of Latin nexuses CA and GA and from conservation, both in Romanian and in Friulian, of nexus -PL and diphthong AU, which instead Ascoli has never detected in his work. With regard to the affinity of the couple “raze/razze”, Luzzatto will write to Graziadio “La razze (ani- tra) è veramente una proprietà esclusiva, almen ch’io sappia, dei valachi e dei friulani. Ma un uccello non fa primavera.” and will also detect that “questo ca, o ka è conservato nel valaco in molti luoghi, dove il friulano lo raddocisce in chià”.69 From the etymological point of view the word razze is in fact a loan common, rare, both of Friulian and Romanian, from Slavic: see Serbian, Croatian, raca [‘ratsa]. In the two final lines Ascoli emphasizes the affinity of truncated words and the final -e, which are common - he say - to the country language and home, of terms of kinship and family, home tools, clothes and body parts, names of people and various other nouns, adjectives, verbs, pronouns and propositions. “L’identico spirito delle due lingue”, as rightly pointed out by Ascoli, is due to apocope of Latin vowels other than -A, that generates the truncated forms of masculine, while the -e of singu- lar are the result of Latin final -A , both in Romanian and in Friulian.

68 Personal contribution of prof. L. Renzi. 69 S. D. Luzzatto, Epistolario, cit., p. 489.

101 [XXXVI]. Indi notinsi queste voci tanto comuni alla lingua di campagna e di casa: sspinoss spinos spinos un fuss un fuso un fus keldarea la caldaja la chialderie lemn legno len ssemn segno segn ssac sacco sac sspin spino spin sskaun scagno scagn baston bastone baston fän fieno fen plan piano plan foae foglia fuee piärseke pesca piersul fum fumo fum jarbe erba jarbe karete carretta carete ozet (azet) aceto aset ulei oglio ueli ai aglio ai lumin lumiccino lumin sspik spiga spic vass vaso vas paket pacco pachett puz pozzo poz fäntäne fonte fontane ssape zappa sape kuseture cucitura cusidure arature aratura aradure aze matassa azze Termini di parentela, di famiglia muër moglie muir tate babbo tate *) *) In alcuni luoghi del Friuli orient. tutor tutore tutor mame mamma mame Attrezzi di casa, vestimenti, parti del corpo, per esempio: furkuzze forchetta furchiuzze *)

102 *) Abbenchè furchiuzze friulano non significhi che forca in diminutivo avendo per forchetta il piron. dinze denti ding kase casa chiase poarta porta poarte almar armadio armar Nomi di popoli e di nazioni: grek greco grec angless inglese ingless obreju ebreo (obreo, goriz.) turk turco turc indian indiano indian chiness (ch guttur.) chinese chiness russ russo russ croat crovato croat kosak cosacco cosac Altrj varj nomi, aggettivi, verbi, proposizioni, pronomi ecc. bune buona bune betran vecchio vedran berboss barbuto barbos amar amaro amar vojoss voglioso vojos grass grasso grass disculz scalzo discolz dojoss doglioso dojos kurat nettato, scelto curat larg largo larg luminoss luminoso luminos lung lungo lung moarte morte muarte neskut nato nassut pierdut perduto pierdut rar raro rar sporkat lordo sporchiat tiner tenero tener frekat fregato frejat ċhernut scelto cernit en askuns di nascosto in scuindon enke ancora anchie

103 nule zero, nulla nule un uno un doi due doi gèmut gemuto zemut avut avuto avut o vut ard del verbo ardere ard ku voe con voglia cun voe lok luogo luc kredinze credenza credinze ploae pioggia ploë piälea pelle piel faur fabbro fari cu bukurìe volentieri cun bon cur ark arco arc acum o amu adesso cumò (mò, acumò!) kascì così cussì fündke essendo che sind che ċhe? che cosa? ce? uss osso uess fok fuoco fuc nass naso nas funt funto (funt) pekat peccato pechiat palosc (sc dolce) scialba paloss quietanzie quitanza quitanzie istorie storia istorie sentenzie sentenza sentenzie Osservinsi questi ie finali in singolari, tanto nel valaco che nel friulano. ssume somma sume forme forma forme mesure misura misure kelkà calcare chialchià ram ramo ram plin pieno plen strämb curvo, strambo stramb unit unito unit fund fondo fond un prinzip un principe un prinzip

104 kurioss curioso curios komediant commediante comedeant prokurator procuratore procurator kontesse contessa contesse predikator predicatore predichiador Notinsi queste desinenze ant, tor, nel valaco e nel friulano. Se anche l’italiano da predicatore può far predicator, il valaco al pari del friulano da predicator non può fare predicatore che è la vera voce dell’italiano dall’ablativo latino; e il latino se ha le voci in or non vi poggia però l’accento;

Many correspondences are correct and based, as in the previous list, on the common Latin origin and on similar treatments, as Rom. plin, Friul. plen ‘full’ < PLĒNU(M), Rom. păcat, Friul. pechia:t ‘sin’ < PECCATU(M), Rom. us, Friul. wess ‘bone’ < ŎSSU(M). But in some cases it is cultured latinisms as curios, comediant. A few remarks, however, is required. The Friulian form “bune” does not correspond to the central Friulian bwine but could belong to the Goritian area. The Friulian “zemut” means ‘how’, not “gemuto”, and “paloss” means ‘saber’, not “scialba”. Put close cu bukurie to cu bon cur is certainly a gamble. Writes Carlo Salvioni in his article Ascoli e il dialetto friulano70 “A vere violenze, per amor della tesi, l’Ascoli veramente non si spinge. Tra queste porrei fiindke (il testo ha per errore fündke) messo con sind che, dove però il ragguaglio s’intende forse solo pegli elementi non iniziali, bukurie paragonato a con bon cur”. The term Bukurie is not Romance, but a word of substrate (see Albanian bukur). The verb “fiindke” is actually fiind că, and is related to a fi (< FIERI), while sind comes from ESSE: it is the same type of training as in French étant que, Ital. essendo che, etc. The i of nexus of the final vowel ie, both in Romanian and in Friulian, is an integral part of the original Latin form, as in HISTORIA, while the -e (final of ie) is the Romance outcome of Latin - A. The suffixes -ant and -tor both derive from the oblique Latin forms in -ANTE(M) and -TORE(M), with apocope of final -E. The confirmation that the Romance words derive from oblique Latin forms is given us just by the vowels -A and -O of -ANTE(M) and -TORE(M), which are tonic only in oblique Latin cases and remain tonic even in the respective Romanian and Friulian results -ant and -tor.

70 C. Salvioni, G. I. Ascoli e il dialetto friulano, cit., p. 124. See par. V.1.2.

105 [XXXVII]. ciò vale anche per le altre di queste desinenze. boü (l’u appena inteso) bove bò sstat stato stat pietenat pettinato petenat (pitin) poliit nettato pulit dormind dormendo durmind ssaltà saltare saltà viu vivo vif kosse cucire cusì ċhel quello chel inel anello anel bunetate bontà bontat liturgje liturgia liturgie protokol protocollo protocoll termin termine tiermin gular collare golar pass passo pass magistrat magistrato magistrat nipot nipote nevod kumnat cugnato cugnat talent talento talento laz laccio laz betae battaglia betae materie materia materie restanzie restanza restanzie skrütorie scrittojo scritori arie aria arie (di musica) fantazzie fantasia fantasie pronunzie pronuncia pronunzie kemeruze cameretta chiameruzze strämbetate goffaggine strambitat sigur sicuro sigur babe donna vecchia babe barbe barba barbe veduve vedova vedue arände ciò che s’ha in affitto arende korone corona corone pulver polvere polvar

106 pumn pugno pugne koaste coste cuestis Si osservino questi oa, cogli ue, ua friulani. Esempj moarte valaco, friulano muarte; poarte valaco, friulano puarte, e qui koaste valaco, cuestis friulano. Per dire scorza, guscio, dicono i valachi scoarze, e pure i friulani adoperano la voce scuarze per dire corteccia. gust gusto gust mut muto mut pipe pipa pipe pompe pompa pompe pulpe polpa polpe krepeture crepatura crepadure note nota note ( di musica) krete creta crete dlhäze dolcezza dolcezze mänuscie guanti manezzins ċhenusce cenere cinise despojat spoglio dispojat deslegat slegato dislejat kernoss carnoso chiarnos monete moneta monede leudat lodato laudat niċhe un nemmeno uno nanchie un Molti pronomi valachi assomigliano a’ friulani, vedi pag. 21;

Among the many observations as possible, we cannote that the word “pulit” has in Friulian (as in Veneto) the adverbial meaning of ‘well’. With the meaning of ‘clean’ instead it would be more Friulian the sense netat. The corresponding Friulian of Romanian “pumn” is not “pugne”, but pugn [pu<].

107 [XXXVIII]. indi per non dilungarsi segneremo ancora questi verbi: arà arare arà arde ardere ardi armà armare armà bate battere bati vinde vendere vendi jeu blestem io bestemmio jo blestemi jeu vajet io grido dal dolore jo vai (io piango) So d’aver veduto in altre Grammatiche valache il pronome di prima persona anche jo precisamente come il friulano. vini venire vigni gustà gustare cibi gustà (pranzare) jeu sbier io grido jo sberli lassà lasciare lassà mutà cambiare mudà numerà numerare numerà ċherni scegliere cerni ecc. ecc.

The phenomenon of apocopated infinity (already dealt in the comment on the passage n. [IV]) is present in many Romance varieties. The Romanian personal pronoun of first person jo [ýO] is ac- tually a phonetic variant of eu [ýeþ].

[XXXIX]. Sicchè un valaco potrebbe pronunciar delle frasi comprensibili del tutto ad un friulano, ed alcuni esempj ne potrei recare, aggiungendo sempre che chi volesse aumentarne il numero potrebbe farlo con tutta facilità. I. val. un om curioss au avut fr. un om curios al a avut it. un uomo curioso ha avuto II. val. demi vin nu ape fr. dami vin no aghe it. dammi vino non acqua III. val. femi un kar bun fr. fami un chiar bon it. fammi un carro buono IV. val. aċheste lam kumperate ku 6 fiorinzi

108 fr. cheste l’ai cumperade cun 6 florinz it. questa l’ho comperata con 6 fiorini V. val. tae cu cuzit fr. tae cul curtis it. taglia col coltello VI. val. känd poate ël veni? fr. quant puedial vigni? it. può egli venire? VII. val. lui nu jaste voe fr. lui a nol a voe it. egli non ha voglia VIII. val. din ċhe ċhetate ëscti ? (sc come nell’it. scelta) fr. di ce citat estu? it. di qual città sei tù? IX. val. aċhasta je grasse adveretate fr. cheste je grasse in veretat it. questa è grassa in verità.

This list of Ascoli represents a series of examples not so much of phraseology, id est expressions common to the two languages, but of correspondences that may extend to portions of text of a cer- tain length (4 to 7 words). We can see in fact that in the example n. VII Ascoli does not notes that the Romanian expression lui nu jaste voe represents a dative of possession, which literally means ‘it is not allowed to him’ (not ‘wish’). The dative of possession is a typical construction of Romanian, as for example mi-e sete to say ‘I am thirsty’ and is inherited as is from Latin, in which the real subject was placed in the dative case, followed by the verb ‘to be’ and then by the object ‘owned’ placed in the nomina- tive case. The indication of this singular mechanism, which would have alerted Ascoli, is the per- sonal pronoun free lui, which acts as a subject only in Friulian, but not in Romanian and even in Italian71 (the Italian spoken however makes extensive use of lui even with function of subject), and especially the use of verb ‘to be’ instead of ‘to have’ (the Rom. jaste is equivalent to modern Ro- manian este [‘ýeste]). In the example n. VI the exact translation of Romanian “känd poate ël veni?” and of Friulian “quant72 puedial vigni?” is ‘when can (he) come?’. Here Ascoli has left out, certainly for distraction, the Italian adverb quando.

71 See in this regard my scheme of free personal pronouns in the passage [II] of Ascoli. 72 Here, for the first time, Ascoli uses in word end the actual pronounce of the final consonant, namely the t, in place of the etymological d of QUANDO.

109 [XL]. Qualche parola più strana prendiamo ad esaminare: mi cuvente dicono i friulani per esprimere m’abbisogna, mi conviene, ed io troverò lo stesso un verbo impersonale valaco kuvine. Clopadiz dicesi d’un vaso rotto, che ha cattivo suono; non mi pare di farlo derivare da sclopà, crepare, ma piuttosto da klopot valaco che significa campana; perchè quest’aggettivo dicesi particolarmente in rapporto al suono. Ed il grazioso interrogativo friulano isal? isel? non sarebbe molto lontano dal valaco jiste ël. E che dirassi del pierdut, del manuscie, del piarseke del jepure che prima abbiamo mentovato?

The Friulian mi cuvente means truly ‘I need’, as says Ascoli, and is voice of verb coventâ73 ‘to need’. The Romanian verb a cuveni ‘be convenient’ (of which cuvine ‘is better’ represents the third person) is derived from Latin CONVENIRE. As for the second approach, Friulian “sclopà”, which does not mean “crepare”, has the exact corre- spondence in Italian scoppiare, from Latin SCLŎPPU(M), from which you have also schioppo ‘musket’.74 The Romanian clopot ‘bell’ is of Slavic origin. The Friulian interrogative expressions “isal?, isel?” contain the personal pronoun subject enclitic -al, - el, while in the Romanian form jeste el? the reversed pronoun el is free. We already seen above, however (see the commentary on the passage [II]), that the clitic subject pronouns Friulian are al in the masculine and e in the feminine. The form with the clitic -el then does not exist in Friulian. The forms proposed by Ascoli should be corrected therefore with isal? ‘is he?’ (mod. Ital. ‘è lui?’) and ise? ‘is she?’ (mod. Ital. ‘è lei?’), which are the Udinese version (and vicinity) of ésal? e ése?75 and that correspond, if you will, to Romanian este el? and este ea? (but not with subject pronouns free). The Romanian forms “manuscie” (mod. Romanian mănuşă) and Friulian “manezzins” (mod. Friulian manece) derive from a single model Latin MANUS, but with different suffixes. The common source of Romanian “piarseke” and Friulian “piersul” is the Latin PERSICA, neuter plu- ral of PERSICU(M), from which even Tuscan and Italian pèsca, properly ‘[the fruit of] Persia’. Even the modern meanings are respectively piersică and piersul. In both cases it occurs the diph- thongization of tonic E in je. The Romanian jepure and Friulian jeur derive from Latin LĔPORE(M), as well as Italian lepre. In jepure the initial lateral was palatalized by a vowel that followed and then is vocalized, according to the path LĔPORE(M) > l’epure > iepure,76 while the path of Friulian was ljevur > ljewr > Yewr > jeur.77 it is possible, however, that even in Romanian has occurred the diphthongization of Ĕ > iè, followed by palatalization of the initial liquid due to the jod of diphthong and relating aphere- sis: LĔPORE(M) > liepore > l’iepure > iepure.

73 See AA.VV., Dizionario etimologico storico friulano [DESF], II vol., Casamassima, 1987, p. 505. 74 M. Cortelazzo - P. Zolli, DELI, cit., p. 1153. 75 G. Marchetti, Lineamenti di Grammatica friulana, cit., p. 240. 76 F. Dimitrescu, Întroducere în fonetica, cit., p. 96. 77 F. Agnoletti, Appunti linguistici sul friulano, cit., p. 145.

110 [XLI]. Conchiuderemo coll’arrischiare di proporre a chi scrive il friulano, di servirsi dell’alfabeto illirico, *) che è pure preferibile per il valaco, al latino, parendoci poterne trarre grande utilità. Scrivendo il friulano con lettere latine, l’ortografia riesce arbitraria, mentre con più precisione il tutto potrebbe farsi adottando l’alfabeto illirico. Suoni, per esprimere i quali, due e tre lettere s’adoperano, potrebbero così in molti luoghi essere rappresentati da una sola. *) Non vadano in collera i friulani, che io non intendo già di parlare dell’illirico che adoperano gli slavi vicini a noi, ma bensì dell’alfabeto illirico propriamente detto, che viene adoperato anche dai Russi (il cirillico).

In his final passage Ascoli proposes to make use of the alphabet “illirico” to write the Friulian. And then he specifies that intends “l’alfabeto illirico propriamente detto, che viene adoperato anche dai Russi (il cirillico)”, not the one “che adoperano gli slavi vicini a noi”. If we understand it, this argument is supported by a series of confusions. First, an alphabet called ‘Illyrian’ seems us to have never existed: with this term has never been coined no alphabet. The term ‘Illyrian’ (from Lat. ILLIRICUM) was given by the Romans to the province inhabited by the Illyrians, formed about 27 av. Christ, which included a part of modern Albania, Montenegro, Croatia and the Istrian coast. The Illyrian territory was then invaded around sixth century by the Slavs, who introduced the use of Slavic language. After long and bloody reli- gious events, in most of this area, which is currently the Croatian coast, was adopted the Latin al- phabet. It does not seem to exist here then a link between the terms ‘Illyrian’ and ‘Cyrillic’, because the former should be associated, if anything, with the use of Latin alphabet, while the second refers to the alphabet which was based on the Greek uncial writing (ie, the greek capital used both in the in- scriptions and the papyri), introduced all over the orthodox world from monk greek Cyril,78 evan- gelizer of the Slavs, as Glagolitic, and then reformed.79 The Russian Cyrillic and the one of neighbors have therefore the same origin and are very similar. It makes no sense, therefore, that Ascoli refers to these alphabets as if they were two different things. With the expression “slavi vicini a noi” he must have thought probably to the Serbs, but they are not so close to the Friulians, since among them there are always the and Croats, who speak Slavic, but as we said they use the Latin alphabet. The observation of Ascoli on the use of Cyrillic alphabet, as far as antihistorical, may have a its justification in the fact that the wealth of consonantal Friulian sounds is made with difficulty by Latin alphabet. This is the case for example of the voiceless and voiced sibilant, which could have been distinguished with the use of Cyrillic letter s and z, or of dental affricate z, whose the voice- less expression is expressed with the sign ц. Of course with the Latin alphabet you could have cre- ated some diagrams, such as was done by Ascoli with writing chi to make sound [c]. However, as I was saying, the proposal of Ascoli was antihistorical because no language changed its alphabet for the technical opportunities that this can offer, but always after very important historical up- heavals (such as when the Romanian abandoned the Cyrillic for Latin or again, when Turkish left the Arabic letters for the Latin ones).

78 But his baptismal name was Constantine. 79 N. Radovich, Introduzione allo slavo ecclesiastico antico, cit., p. 16.

111 [XLII]. Imperfetto è quest’abozzo. In ogni modo l’imperfezione del nostro lavoro consisterà nel non aver ben ispiegate le cause che portarono all’effetto, giacchè l’effetto, crediamo aver dimostrato suf- ficientemente, esistere, e trovarsi un’assoluta affinità fra questi due idiomi. I pochi dati Storici pre- messi al confronto, crediamo tuttavia mostrare alquanto la possibilità d’una stessa origine delle due lingue; e se certo è come abbiamo veduto, che il friulano ha il vero tipo della lingua daco-romana, sarebbe a desiderarsi, che qualche dotto friulano, delle patrie cose amante, s’interessasse di meglio illuminare storicamente e filologicamente questa relazione della lingua di Bukarest con quella d’Udine, ciocchè alla Storia ed alla cultura della lingua patria, per vero un po’ trascurata, gran lume apporterebbe.

The young Ascoli is perhaps succeeded to give an overall impression, that is intuitive, about the affinity among Romanian and Friulian, but without ever invoking the common origin, the Latin, and the fact the similarity between the two languages is only a particular aspect of affinity of all Romance languages between them, due to common origin. The serious “imperfezione” of his work consists precisely in the absence of research into the causes that led to this affinity. Causes, as we have often repeated, that are easily traceable to the language from which they originated both Romanian and Friulian, namely the Latin, especially the Vulgar Latin. Ascoli noted as a whole some special convergences, which are not more common in Friulian and Romanian than in any other possible pair of Romance languages. Later studies have estab- lished that in particular the Friulian - despite its eastern location - typologically belongs to group of Western Gallo-Romance languages, along with the northern Italian (except the Veneto), while Romanian belongs, with the Italian south-central, to eastern Romance area.80

80 See L. Vanelli, La posizione del friulano nel dominio romanzo, in Tullio De Mauro and Shigeaki Sugeta, Lesser-used languages and Romance linguistics, Roma, Bulzoni, 2002. .

112 VI. Conclusions

At this point you can make a quick assessment of work of the young Goritian linguist. Let’s observe first of all the numerous ‘improvisations’ within his historical premise: - The ethnic connection of Dacians with Germans, deducted hurriedly by similarity between the two ethnic, daci and ‘deutsch’, whose ancient form tiudisk (from which Ital. tedesco) should have warned the author.1 From the acceptance of this theory Ascoli obtained, in addition, that language of pre-Romanized Dacians had given rise to Germanic languages. - Romanian language and Friulian appeared to him as the result of the interlacement of Latin with languages of peoples who invaded both Rumania and Friuli, namely the Slavs, Ungars and Ger- manic peoples. Hence the examples of morphological correspondence put forward to justify the presence of Germanic element in Romanian and in Friulian (cfr. Par. V.4). Moving on to his comparative analysis, we see a long line of ‘inaccuracies’, which I summarize, combining them in typological groups:

Formation of article: if the Romanian has the postponed defined article, for Ascoli would exist traces of postponed article also in Friulian (see passage of Ascoli [XV]), which is obviously wrong;

Formation of plural: Ascoli does not notice the use of sigmatic plural in Friulian (see passages [XI], [XXI]), and therefore does not justify properly the phonetic phenomena derived from it (see [XIV] and [XVII]);

Past participle: the past participles in Romanian would be formed for Ascoli by adding a t to the infinitives: the description in itself is possible, but it is not clear what is the perspective that allows it (see [V]).

Personal Pronouns: Ascoli does not distinguish the statute of free pronoun of Romanian from the one of clitic in Friulian (see [II]). He gives to Friulian the pronominal element el, which is Veneto, instead of al (see [VII]). He does not use the clitic in the Friulian examples of compound verbs (see [IX]);

Finally, among the historical arguments, he seems to confuse the term ‘Cyrillic’ with ‘Illyrian’ (see [XlI]).

From the analysis of this brief exposition we note that the inaccuracies of Ascoli refer broadly to the Friulian language, while about the Romanian one you can speak only of the error (relative, as we have seen) in description of the formation of participles. This observation may be sufficient by itself to prove that the young linguist did not have a great knowledge of Friulian, and has not been

1 But this is the connection that was circulating at the time and that was accepted, among others, also by the man who will become the great Romanian national poet Mihai Eminescu, who in his note Germanophily was satisfying himself.

113 able to penetrate in many cases, albeit in an intuitive way, the structure of it, whereas for the Ro- manian was supported with good security from the Molnar’s Grammar. In the above exposition I have confined myself to the list of ‘errors’ present in the essay of Ascoli, respecting his primary setting, that wanted be only a description of lexical affinities between Friulian and Romanian. This approach, which we know limited because lacks entirely its comple- mentary corpus, that is the description of what distinguishes the two languages, as will make noted Luzzatto (see Par. V.1.1), could still have been validly accepted, whether Ascoli had endowed it with a Latin common platform, to which connect the affinities. So there is not even a presentiment of the method of that historical linguistics was already risen in France in the early nineteenth cen- tury to work Raynouard (see Par. III.2.1), and especially in Germany, of which Ascoli would have been soon the largest standard bearer (and recognized exponent) in Italy. We observe that Cattaneo, initially higher than the young Ascoli, although he pioneer, he restrict to build a system of contrastive comparisons between Italian and Romanian (see Par. IV.2.1), and between Italian and Friulian (see par. IV.3.1) without recourse to historical criteria, except on rare occasions, as when he notes the outcome of Romanian nexus pt from Latin CT and in Friulian the palatalization of Latin nexuses CA and GA, the preservation of diphthongs AU and of nexuses CL, FL. Cattaneo, however, even if remaining within the pure comparison, goes further than Ascoli, because he considers, in addition to affinities, even the distinctive aspects of languages under con- sideration. Ascoli then he could have emphasized at least the elements that characterize the specific peculiar- ity of Romanian and that of Friulian, within the Romance area, which had already been just de- scribed exactly by Carlo Cattaneo (see Par. IV.2.1 and IV. 3.1), such as the postposition of the ar- ticle in Romanian and the palatalization of CA and GA in Friulian. We cannot forget, finally, that it was the author himself, Graziadio Isaia Ascoli, to condemn his early work, calling it in his mature years “un lavoruccio insignificantissimo”.2 Our thesis does not failed to overturn the harsh self-condemnation of Ascoli, but had to confirm it. His work is consisted in hasty notes, noticed while reading of the excellent grammar of Molnar, of real or alleged similarities detectable in Friulian. Ascoli did not have the initiative to consider this examination, certainly useful, as a start to be completed with other research on the Friulian, who at the time - it is true - it was still very little studied. It will do so later, and with such great results we know. For the moment the young scholar will stop at what was supposed to be just the beginning, and rushes too soon to look for a press for his little essay. As a whole and in details, therefore, the Schizzo storico filologico of Graziadio Isaia Ascoli on the affinity of Friulian with Romanian remains - after the accurate examination to which we have submitted it - an early work, scarce as a whole and in details, in which it is not still possible to catch a glimpse of the future greatness of most renowned Italian linguist of the nineteenth century and one of the first in Europe.

2 G. I. Ascoli, Studj critici, cit., p. 57, n. 3.

114 VII. Bibliographic references

VII.1 Linguistics and philology

F. ALBANO LEONI - P. MATURI, 1999, Manuale di fonetica. Nuova edizione, Roma, Carocci P. BENINCÀ, 1996, Piccola storia ragionata della dialettologia italiana, Padova, Unipress M. CORTELAZZO - P. ZOLLI, 1979, Dizionario etimologico della lingua italiana, Bologna, Za- nichelli A. ERNOUT, 1953, Morphologie historique du latin, Parigi I. IORDAN - J. ORR, 1973, Introduzione alla linguistica romanza, Torino, Einaudi W. MEYER-LÜBKE, 1890-1902, Grammatik der romanischen Sprachen, 4 voll. (facsimile repro- duction Hildesheim, Olms, 1972), Lipsia A. MORPURGO DAVIES, 1996, La linguistica dell’Ottocento, Bologna, Il Mulino G. MOUNIN, 1968, Storia della linguistica. Dalle origini al XX secolo, Milano, Feltrinelli G. B. PELLEGRINI, 1977, Introduzione alla linguistica balcanica, Università di Padova, Istituto di Glottologia V. PISANI, 1971, Le lingue indeuropee, Brescia, Paideia S. POP, 1950, La dialectologie. I. Dialectologie romane, Louvain, Duculot N. RADOVICH, 1982, Introduzione allo slavo ecclesiastico antico, Università di Padova, Istituto di Filologia Slava L. RENZI, 1987, Nuova introduzione alla filologia romanza, Bologna, Il Mulino L. RENZI, Come gli umanisti non scoprirono le leggi fonetiche, in Diaz-Rizzotto (edited by), Hommage a J. Brunet. Annales littéraires de Franche-Comté, Vol. 2, pages 571-584. G. ROHLFS, 1966, Grammatica storica della lingua italiana e dei suoi dialetti, Fonetica, Torino, Einaudi G. ROHLFS, 1968, Grammatica storica della lingua italiana e dei suoi dialetti, Morfologia, Tori- no, Einaudi C. TAGLIAVINI, 1963, Panorama di storia della linguistica, Bologna, Pàtron C. TAGLIAVINI, 1972, Le origini delle lingue neolatine, VI ed., Bologna, Pàtron V. VÄÄNÄNEN, 1982, Introduzione al latino volgare, Bologna, Pàtron A. VARVARO, 1968, Storia, problemi e metodi della linguistica romanza, Napoli, Liguori B. E. VIDOS, 1959, Manuale di linguistica romanza, Firenze, Olschki

115 VII.1.1 About Friulian

AA.VV., 1987, Dizionario etimologico storico friulano [DESF], II voll., Udine, Casamassima M. DORIA, 1978, Storia del dialetto triestino, Trieste G. FRANCESCATO, 1966, Dialettologia friulana, Udine, Società Filologica Friulana G. MARCHETTI, 1985, Lineamenti di grammatica friulana, Udine, Società Filologica Friulana J. PIRONA, 1868, Vocabolario friulano, Venezia P. RIZZOLATTI, 1981, Elementi di linguistica friulana, Udine, Società Filologica Friulana, L. VANELLI, 1998, I dialetti italiani settentrionali nel panorama romanzo, Studi di sintassi e morfologia, Roma, Bulzoni L. VANELLI, 1998, Le vocali lunghe del friulano, in “Quaderni della grammatica friulana di rife- rimento”, I, Udine, Forum, pages 69-108 L. VANELLI, 2002, La posizione del friulano nel dominio romanzo, in Tullio De Mauro and Shi- geaki Sugeta, Lesser-used languages and romance linguistics, Roma, Bulzoni

VII.1.2 About Romanian

B. CAZACU ET ALII, 1967, Cours de langue roumaine. Introduction à l’étude du roumain à l’usage des étudiants étrangers, Bucarest F. DIMITRESCU, 1967, Introducere în fonetica istorică a limbii române, Bucarest T. FERRO, 1992, Latino e lingue balcaniche nella formazione del rumeno, Catania, CUECM M. LÖRINCZI ANGIONI, 1986, Nouvelles données pour l’étlablissement d’une probable tradi- tion graphique latino-moldave: l’ortographe polonaise chez M. Costin et D. Cantemir in Critique et éditions de textes. Actes du XVII Congrès international de linguistique et philologie romane, Vol. n. 9, Aix-en-Provence, Université de Provence J. MOLNAR, 1788, Deutsch Walachische Sprachlehre, Vienna A. NICULESCU e F. DIMITRESCU (a cura di), 1970, Testi rumeni antichi (secoli XVI-XVIII) con introduzione, studio linguistico e glossario, Padova, Antenore L. RENZI, 1965, Gli studi rumeni di Adolfo Mussafia, in Omagiu lui Alexandru Rosetti la 70 de ani, Bucarest, pages 745-750 H. TIKTIN, 1903, Rumänisch-Deutscher Wörterbuch, 3 voll., s. v., Bucarest

116 VII.2 Studies about Ascoli, Cattaneo and others

F. AGNOLETTI, 2001, Appunti linguistici sul friulano in un manoscritto inedito di Carlo Catta- neo, thesis, rel. L. Vanelli, Università degli Studi di Padova, Facoltà di Lettere T. BOLELLI, 1962, s. v. “Ascoli, Graziadio Isaia”, in Dizionario Biografico degli italiani, Roma, Istituto della Enciclopedia Italiana fondata da G.Treccani I. BONOMI, 1995, Note sulla formazione e gli interessi del giovane Ascoli, in “Parallela 5, Atti del VI Convegno italo-austriaco dei linguisti”, edited by M. Dardano et alii, Roma, Bulzoni, pages 29-55 I. BONOMI, 2001, Idee per un progetto di lingua universale in uno scritto inedito del giovane A- scoli, in Milano e l’Accademia Scientifico-Letteraria. Studi in onore di M. Vitale (edited by G. Barbarisi et alii), Milano, Istituto Editoriale Cisalpino, pages 385-402 A. BRAMBILLA, 1996, Appunti su G. I. Ascoli. Materiali per la storia di un intellettuale, Gori- zia, Istituto Giuliano di storia, cultura e documentazione G. BRESCHI, 1973, L’”Autobiografia” di Graziadio Isaia Ascoli, in “AGI”, LVIII, pages 39-81 C. COEN LUZZATTO, 1907, G. I. Ascoli (1860-1861), Gorizia, Paternolli E. COSERIU, 1986, Gli “antenati” di Ascoli, in “Attualità del suo pensiero a 150 anni dalla nasci- ta”, Istituto per gli incontri culturali mitteleuropei (Gorizia), Firenze, Licosa, pages 21-36 G. D’ARONCO, 1958, G. I. Ascoli, in “Studi goriziani”, XXIII F. DELLA PERUTA, C. MARCORA, E. TRAVI (edited by), 1985, Cesare Cantù nella vita ita- liana dell’Ottocento, Milano T. DE MAURO, 1962, s. v. “Biondelli, Bernardino”, in Dizionario Biografico degli italiani, Ro- ma, Istituto della Enciclopedia Italiana fondata da G.Treccani M. G. D’IMPERIO, 1999, Graziadio Isaia Ascoli: la vita, thesis, rel. C. Griggio, Università degli Studi di Udine, Facoltà di Lettere M. G. D’IMPERIO, 2002, I «Pensieri» di Graziadio Isaia Ascoli, in “Studi goriziani”. XCI/XCII, pages 39-46 C. DIONISOTTI, 1993, Appunti su Ascoli, in Le tradizioni del testo. Studi di letteratura italiana offerti a Domenico De Robertis, edited by F. Gavazzeni and G. Gorni, Milano - Napoli, Ricciardi, pages 419-432 G. FRANCESCATO, 1961, G. I. Ascoli e il friulano del suo tempo, in “Studi Goriziani”, XXIX, pages 27-36 G. HUGUES, 1958, Di alcuni illustri semitisti e orientalisti della Venezia Giulia, in “Studi Gori- ziani”, XXIV, pages 33-80 M. E. LORICCHIO, 1989, G. I. Ascoli tra politica e cultura, in Friûl di Soreli Jevât. Setante ains di storie, di culture, di Filologiche (1919–1989), edited by E. Sgubin and M. Michelutti, Udine, pages 29-44

117 M. E. LORICCHIO (edited by), 1999, G. I. Ascoli. Biografia di un intellettuale, Monfalcone (Go), Edizioni della laguna G. LUCCHINI, 1999, Il giovane Ascoli e la tradizione ebraica, in “Studi di grammatica italiana”, XII, 1, pp. 329-435 G. LUCCHINI, 2002, Graziadio Ascoli: per una biografia intellettuale, in A. Casella, Graziadio e Moisè Ascoli: scienza, cultura e politica nell’Italia liberale, Pavia, Università di Pavia, pages 1- 172 S. D. LUZZATTO, 1890, Epistolario italiano francese latino di Samuel David Luzzatto da Trieste pubblicato da’ suoi figli, Padova, Tipografia della Minerva dei fratelli Salmin, part I, n. CCXCVII G. MANZINI, 1957, Attualità di Ascoli, in “Studi goriziani”, XXI, pages. 60-73 A. PAREDI, s. v. “Castiglioni, Carlo Ottavio”, in Dizionario Biografico degli italiani, Roma, Isti- tuto della Enciclopedia Italiana fondata da G.Treccani M. RADONI ZUCCO, 1973, Profilo biografico di G. I. Ascoli, in G. I. Ascoli e l’Archivio Glotto- logico Italiano, edited by M. Cortelazzo, Udine, Doretti, pages 11-35 G. ROSA, 1912, Autobiografia, Brescia F. SALIMBENI, 1986, Ascoli, intellettuale del Risorgimento, in Attualità del suo pensiero a 150 anni dalla nascita, Istituto per gli incontri culturali mitteleuropei (Gorizia), Firenze, Licosa, p. 39 and foll. F. SALIMBENI, 1986, Graziadio Isaia Ascoli tra cultura e politica, in “Studi goriziani”, n. 64, pages 99-111 C. SALVIONI, 1907, G. I. Ascoli e il dialetto friulano, in “Memorie storiche forogiuliesi”, III, pa- ges 122-126 D. SANTAMARIA, 1986, G. I. Ascoli e la linguistica italiana del primo Ottocento, in G. I. Ascoli, attualità del suo pensiero a 150 anni dalla nascita, Firenze D. SANTAMARIA, 1986, Orientamenti della linguistica italiana del primo Ottocento, in The hi- story of linguistics in Italy, edited by P. Ramat et alii E. SESTAN, 1962, s. v. “Cattaneo, Carlo”, in Dizionario Biografico degli italiani, Roma, Istituto della Enciclopedia Italiana fondata da G.Treccani B. TERRACINI, 1949, La paleontologia linguistica: Ascoli, in Guida allo studio della linguistica storica, I, Profilo storico-critico, Roma, Edizioni dell’Ateneo B. TERRACINI, 1967, G. I. Ascoli, direttore dell’«Archivio» (dal carteggio Ascoli-Salvioni), in “AGI”, LII, pages 1-54 S. TIMPANARO, 1969, Classicismo e Illuminismo nell’Ottocento italiano. Carlo Cattaneo e G. I. Ascoli, Seconda edizione, Pisa, Nistri-Lischi S. TIMPANARO, 1972, Graziadio Isaia Ascoli, in “Belfagor”, XXVII, pages 149-176 D. TROTTA, 1986, «Modernità e storicità» di Graziadio Isaia Ascoli, in “Studi goriziani”, LXIII, pages 83-107

118 L. VALMARIN, 1989, Il contributo di G. I. Ascoli allo studio della linguistica romena in Italia, in “Cultura neolatina”, XLIX, pages 55-67 F. VITTORI, 1962, s. v. “Cherubini, Francesco”, in Dizionario Biografico degli italiani, Roma, Istituto della Enciclopedia Italiana fondata da G.Treccani

VII.3 Works of Ascoli and Cattaneo

G. I. ASCOLI, 1846, Sull’idioma friulano e sulla sua affinità colla lingua valaca. Schizzo storico- filologico, Udine, Tipografia Vendrame G. I. ASCOLI, 1852, Note letterario-artistiche minori durante il viaggio nella Venezia, nella Lombardia, nel Piemonte, nella Liguria, nel Parmigiano, Modenese e Pontificio, maggio–giugno 1852, in “Annali della Scuola Normale”, edited by S. Timpanaro, XXII, 1959 G. I. ASCOLI, 1872, Autobiografia inedita, riportata da A. L. Prosdocimi, Carteggio di Ascoli a Mussafia, in “AGI”, LIV, 1969 G. I. ASCOLI, 1873, Saggi ladini, “AGI”, I G. I. ASCOLI, 1929, [i frammenti del] Diario giovanile, edited by B. Terracini, published in A- A.VV, Silloge di scritti in onore di Graziadio Isaia Ascoli, Torino, pages XXVIII-XXXVII (= “AGI”, XXII-XXIII) C. CATTANEO, 1842, Sul principio istorico delle lingue europee, Milano C. CATTANEO, 1846, Del nesso fra la lingua valaca e l’italiana, reprinted in Scritti letterari, ar- tistici, linguistici e vari, collected by Bertani, Firenze, Le Monnier, 1948 C. CATTANEO, 1949-56, Epistolario (edited by R. Caddeo), I-IV, Firenze C. CATTANEO, 1957, Scritti storici e geografici (edited by G. Salvemini - E. Sestan), I-IV, Fi- renze

VII.4 Studies about Judaism

O. ALTIERI, 1985, La comunità ebraica di Gorizia: caratteristiche demografiche, economiche e sociali (1778-1900), Udine, Del Bianco M. BERENGO, 1987, Gli Ebrei dell’Italia asburgica nell’età della Restaurazione, in “Italia”, n. 1-1 M. DEL BIANCO, 1981, Ebrei e industria della seta nel Gradiscano attraverso gli atti del Magi- strato e del Consesso commerciale, in “Quaderni Giuliani di Storia”, 2, pages 41-66

119 M. DEL BIANCO, 1991, La vita privata degli ebrei nei territori italiani della casa d’Austria e nel Friuli veneto in età moderna, in Gli Ebrei tra Italia nord-orientale e impero asburgico dal medio- evo all’età contemporanea (edited by G. Todeschini and B. C. Ioly Zorattini), , Studio Tesi, pages 181-213 P.C. IOLY ZORATTINI (edited by), 1984, Gli ebrei a Gorizia e Trieste, tra «Ancien régime» ed emancipazione, Atti del Convegno di Gorizia, Udine, Del Bianco, P.C. IOLY ZORATTINI, 1991, Gli Ebrei del Friuli e dell’Istria nelle fonti inquisitoriali, in Gli Ebrei tra Italia nord-orientale e impero asburgico dal medioevo all’età contemporanea (edited by G. Todeschini and B. C. Ioly Zorattini), Pordenone, Studio Tesi, pages 130-154 M. LUZZATI (edited by), 1995, Le tre sinagoghe. Edifici di culto e vita ebraica a Livorno dal Seicento al Novecento, Torino A. MILANO, 1963, Storia degli Ebrei in Italia, Torino, Einaudi B. SEGRE, 1993, Gli Ebrei in Italia, Milano, Fenice 2000

120 The following section has been excluded from my thesis, but it contains an argument closely re- lated to the topics covered by G. I. Ascoli in his youth linguistic essay.

VIII. The polemic about the issue of the influence of barbaric languages on the for- mation of Vulgar Italian: in the fifteenth century, Leonardo Bruni1 believed that Latin and Vulgar had coexisted as two languages quite distinct. Instead Flavio Biondo2 argued that Latin had been unitary and the Vulgar Italian had born later, due to a contamination of Latin first to the large number of foreigners come to Rome, then, a much greater extent, to the barbarian invasions. Bruni was seeing well the close connection between Italian and Vulgar Latin, but was denying any evolution of languages and each derivation connection. Biondo attached the loss of Latin purity to outside infiltrations, those barbarian, and he linked this fact to loss of racial purity of Romans. in the eighteenth century, Scipione Maffei3 stated that Italian was the continuation of Vulgar Latin. He believed that during this transformation, the influence of barbarian languages had been negligible, especially since their number was very small. Ludovico Antonio Muratori4 instead tended to deny the continuity with the Latin and give the highest prominence to the Germanic ele- ment.

1 Leonardo Bruni, born in Arezzo in 1370 ca. and very young moved to Florence, undertook studies of rhetoric and law, soon be- coming a refined humanist, from 1405 embarked on a political career in Rome, in the service of Innocent VII. In 1411 he partici- pated in the Council of Constance in the wake of the antipope John XXIII and in 1415 he settled in Florence, holding the position of Chancellor of Republic from 1427 to his death. Personality of great culture, Bruni was one of the most distinguished Latin writers of his time, and he engaged in fine translations from Greek (Plato and Aristotle). Among his many works stands the Historiae flor- entini populi in 12 books, in which the historical method adopted is based on the comparison of the documents, removing any providential interpretation of history. In the Vulgar he wrote a Vita di Dante and a Vita di Petrarca (both 1434), some poems and short stories. He died in Florence in 1444 and was buried in Holy Cross. [from G. Bongos, Biografia di L. Bruni] 2 Flavio Biondo, historian and archaeologist, born in Forlì in 1388. He was the founder of archeological science and of Christian topography of Middle Ages. He served in numerous popes, from Pope Eugene IV, to Nicholas V, Calixtus III and Pius II. He was the author of three encyclopedias, which represented the basis for all subsequent dictionaries of archeology and antiquities of Rome: Italia Illustrata (1474), accurate description of four Italian regions, Romae Instauratae Libri Tres (1482) study of the an- cient monuments of Rome and Romae Triumphantis Libri Decem (1482), on the manners and customs of the ancient Romans. In his works converge, in way for his time singular, philology, archeology and history. He died in Rome in 1463. [from the Catholic Encyclopedia] 3 Scipione Maffei, born in Verona in 1675, from a noble family, founded in 1710 with Zeno and Vallisnieri the «Giornale de’ let- terati italiani». His multiform activity worked courageously to the renewal of the Italian social structures and cultural. He was, with Muratori, an important step in the transition from epoch of scholars to that of reformers. He gave a contribution to the progress of philology and palaeography with l’Italia diplomatica (1727), which establishes in the inspection of medieval codices an evaluation criterion chronological-historicist. A historical and scholarly masterpiece of Maffei was the monumental Verona illustrata (1732), dedicated to the history, writers and monuments of his city. He died in Verona in 1755. [from the Catholic Encyclopedia] 4 Ludovico Antonio Muratori born at Vignola in 1672. As a modest family, he majored in philosophy and canon law. In 1695 he moved to Milan as Prefect of the Biblioteca Ambrosiana and the same year he was ordained priest. Scholar of extraordinary ability and tenacity, he directed his studies to the Middle Ages that had been hitherto little studied and considered unworthy of attention. He sought and gathered by itself the sources of Italian history starting from 500 to 1500, which published in the monumental collec- tion Scrittori di cose italiche (Rerum italicarum scriptores 1723-1751) in 25 volumes. He died in Modena in 1750. Muratori is con- sidered the founder of modern historiography on scientific and documentary basis. [Lydia Pavan]

121 in the nineteenth century, a staple of the romantic doctrine was that modern civilization had be- gun after the fall of Roman Empire. The dominance of Christianity over paganism and replacing the new national language to Latin were marking for romantics the end of classical period and the beginning of the romantic or modern, of which Middle Ages was an integral part. So also in the question of origins of vernaculars the romantics were induced to minimize the continuity with re- gard to Latin and instead give the highest prominence to Germanic element. It was the theory of Muratori. It was the Medieval birth of a new Europe, unhooked from paganism, from literature and from language of Roman world. Romagnosi5 answered to these theoretical formulations of the romantics by denying, relative to opinion of Maffei, the forced association between the cultural education of a people and its an- thropological composition. Ascoli himself had a direct contact in youngness with a supporter of so-called theory of ‘super- strate’. When he sent a copy of his essay on the affinities between Friulian and Romanian to Sa- muel David Luzzatto to receive his authoritative opinion, he did not delay to express to him that “la lingua della Crusca non può essere illustrata senza il confronto degli altri dialetti italiani, e questi non possono illustrarsi senza il con- fronto delle lingue dei barbari, germani e slavi, che diedero origine alle nuove lingue latine.”6 So we can suppose that thought of his master had come to Graziadio directly, or by means of his friend Filosseno Luzzatto (son of Samuel David), before the preparation of his essay about the Friulian and Romanian.

5 Gian Domenico Romagnosi, born in 1761 in Salsomaggiore, was a distinguished jurist and philosopher. He graduated in law at the University of Parma in 1786, he was appointed Magistrate in Trento in 1791 and in 1804 named lecturer at the Department of Public Law at the same university of Parma. In 1806 he was called to Milan with a mandate to review the draft Code of Criminal Procedure, where he became a consultant to the Ministry of Justice. In 1807 he was given the chair of civil law at Pavia. In 1817, following the return of Austrians in Lombardy, lost his job as a teacher in the special schools of law, which were suppressed. In 1821, for connections with progressive circles, he was imprisoned in Venice. Released shortly afterwards, he remained out of work. He died in Milan in 1835. Follower of Francis Bacon and John Locke, as well as of the French illuminists, he left many legal and philosophical works. 6 S. D. Luzzatto, Epistolario italiano, latino e francese, cit., part I, n. CCXCVII, pages 485-491. For a full commentary on the letter of S. D. Luzzatto see par. V.1.1.

122 VIII.1 The theory of substrate

If up to now has been considered the relationship of Latin with the languages that would have su- perimposed in later times, namely the languages of the ‘superstrate’, the theory of ‘substrate’ con- cerned instead the relationship of Latin with languages of the peoples conquered from the Romans, that is Etruscans, Osco-Umbrians, Celts, Iberians, and so on. In Italy the problem was posed by the differentiation of Latin in the various Italian dialects and Maffei was the first to enunciate it, talking of difference between the dialect of Brescia and Vero- na, “certa cosa essendo che i nostri odierni dialetti non altronde si formarono, che dal diverso modo di pronunciare negli antichi tempi e di parlar popolarmente il Latino”, in «Verona illustrata» of 1825. Pursuant to his precise definition, even Muratori, already a staunch defender of the Germanic superstrate, confirmed the thesis of Maffei, but he placed it side by side to the one of superstrate: it was the thesis of ‘dual influence’. From this moment was born and developed in Italy the fashion of the ethnic and linguistic ‘ori- gins’ of Italian peoples, that it appeared broadly divided into two sections, one advocate of pre- Roman thesis, the other of the Germanic influence. It is curious as both of these factions founded their animosity on a common attitude of anti-Roman patriotism. The theory of substrate was firmly anchored to sympathy to the pre-, which they wanted re-emerged in the Middle Ages, after a long period of submersion to Roman domination. On the contrary, the theory of superstrate favoured the individuality and autonomy of the new European nations - especially those of German origin - from the shackles of Latin culture. At European level, the substrate theory was also applied to Indo-European language. Friedrich Schlegel had made derive Greek, Latin, German, Persian from the interweaving of Sanskrit lan- guage with the languages of primitive peoples of Europe. And on this linguistic relationship is rec- reated the conditions of a vast movement of masses from Asia to Europe. Asia it was considered for a long time as the “cuna del genere umano”, as was expressing Cesare Cantu,7 repeating a concept which then enjoyed great credit.

7 C. Cantù, Storia universale, I, Torino 1838, p.191

123 VIII.2 The linguistic thought of Carlo Cattaneo

The first essay of linguistics published by Cattaneo was the one on the Nesso della nazione e della lingua valaca coll’italiana, published in 1837, with the aim of studying “l’influenza dell’invasione dei barbari sulla favella italica”. It was a work of historical and not purely linguistic, because Cattaneo was allud- ing, as was his habit, even to matters of legal and economic history. In this paper Cattaneo was driven by self-interest for the primitive peoples. And he was following, loosely based, on a con- cept of scientific research more advanced than the style of time, because he corrected the excesses of Enlightenment rationalism and anti-historical, without ever give up the concept of Enlighten- ment progress. In his linguistic essay more meaningful, Sul principio istorico delle lingue indo-europee (1841), Cattaneo insisted on the importance of the study of dialects and the need to collect them with soli- citude, because “i dialetti rimangono unica memoria di quella prisca europa, che non ebbe istoria, e non lasciò monumenti. Giova dunque raccogliere con pietosa cura tutte queste rugginose reliquie; studiare in ogni dialetto la pronuncia e gli accenti; notare quanto il suo dizionario ha di commune colla lingua nazionale, e quanto ha di diviso.” He took, therefore, the commitment to draw the attention of scholars on the need to conduct research on dialects, which he considered the only evidence of pre-Roman languages and cultures. This incitement was soon followed by the work of Cherubini, Biondelli, Monti, Rosa and others, who in turn prepared the ground to work of Ascoli. Cattaneo’s interest was directed even to the relations between Romans and barbarians and the ori- gins of Italian language, supporting many times the argument of the small number of barbarian in- vaders and the limited influence of their languages in the Italian vernacular. Both of these theses of Cattaneo will be confirmed by later scholars. Cattaneo demonstrated interest from the beginning even to the Germanic languages and Romanian, the study of which helped him to know the new comparative linguistics of the Danish and German scholars. It was certainly from them that he learned the principle that the linguistic relationship of two languages is shown by the affinity of the grammatical structure, rather than by lexical similari- ties. He was contrary to the great Asian migrations who wanted to justify the affinity of Indo-European languages. He believed that did not exist a parallel between the theory of great migrations of bar- barian invaders of Roman Empire (to which Cattaneo had never joined) and the one of migration of Asian peoples colonizers of early Europe. He noted that the barbarians who invaded Roman Empire never succeeded to impose their language, while Asians (which he said showed up in the European territories for purposes other than military invasion or mass) imposed their language through the foundation of colonies of emigrates, including the possible presence of political refu- gees or religious in search of freedom, not only of living space or wealths. Though remaining polemic toward the migration of peoples, Cattaneo defended with great deter- mination the concept of clear distinction between linguistic affinity and racial affinity “Communanza di lingue non prova communanza di origini”,8 distinguishing, well ahead of its time, the linguistics from anthro- pology.

8 C. Cattaneo, Politecnico, VII, 1845 = SF, I, p. 220.

124 The other polemic destined to uncertain assuagement was that which concerned the common ori- gins among and Germans. This commonality was recently disclosed by the German discov- eries of the Indo-European linguistics and it was the new theory that aroused the hostility of the Italian classicists, who were seen suddenly deprived of the ancient record of a noble origin, sinking its ties on the ancient Greek and Roman civilizations, considered a unapproachable model and never exceeded. Cattaneo, however, he realized that these linguistic connections between Greek- Roman peoples and ‘barbarians’ Germans were indisputable. However he justified the linguistic colonization in European zone from Asians as an event occurring by sea rather than by land. This migration involved first the Mediterranean peoples and only later continental peoples such as Celts, Germans and Slavs. It may not be a case - says Cattaneo - if the languages extraneous to Indo-European contamination like Basque, Finnish, Samoyed, they are in “estremo contorno dell’Europa”.9 As for his position on the origin of languages, he defended the theory of ‘polygenism’ which had been inspired to him especially by Cesarotti10 and Romagnosi11, but he did not admit that they ex- isted languages perfect from the beginning, because according to him all languages had humble origins. On this point, Cattaneo was contrasted with polygenism of Schlegel, who considered the Indo-European ‘inflected’ languages as bodies perfectly structured from birth. Fact, this, which had to imply a divine intervention, while the thesis ‘agglutinative’ of Cattaneo wanted to be in the limit of the rational explanation, attributing to all languages an origin purely human. What the two linguistic versions were trying in some way to hide (and justify), it was no more than the very origin of mankind. Schlegel wanted to derive humanity from an upper phase, corrupted then gradually over time. Cattaneo instead wanted to prove that the mankind was proceeding from a primitive phase, which gradually evolved towards a civilization more complex. It was the exam- ple of conflict existing among German Romanticism and the Enlightenment positivism of Catta- neo. The theory of substrate (an idea circulating for some time in European circles and Italian)12 was enunciated in the linguistic writings of Cattaneo when he made to date back the difference of dia- lects to the “differenza delle popolazioni primitive, le quali non si sradicarono mai dal terreno nativo, né dopo i Romani né prima, e assumendo dai Romani il linguaggio latino, lo modificarono a seconda del loro anteriore idioma etrusco, o celtico, o veneto, o carnico, e del- la domestica loro abitudine di pronunciarlo”.13

9 C. Cattaneo, Scritti letterari, I, p.170 10 Melchiorre Cesarotti born in Padua of a noble family in 1730. Priest, taught Greek and Hebrew at Padua. Modest supporter of the ideals of Enlightenment, had even the protection of Napoleon. Writer abundant, he ordered the complete edition of his works in 40 volumes (1800-1813). Famous for Cesarotti was the Saggio sulla filosofia delle lingue and the Saggio sulla filosofia del gusto, published in 1785. Cesarotti denies the principle of authority, identifies both in the use and in the reason the key factors for the his- torical evolution of languages, equally distant from the purist rigor and the innovator excesses. Cesarotti was not a great, but was central in the literary development between the eighteenth and nineteenth century, he traced some important guidelines of the pre- Romanticist gusto, though moving within a Enlightenment formation. He died in Selvazzano (Padua) in 1808. 11 See footnote on par. in the nineteenth century. [from Liber liber, Biblioteca informatica] 12 See par. La teoria del sostrato. 13 C. Cattaneo, Scritti letterari, I, p. 408.

125 In the formation of Indo-European languages, he attaches fundamental importance to the role of substrate, but not considered as an element of corruption of languages superimposed later, as in- stead reported by Friedrich Schlegel in the case of Sanskrit, but a good thing, because it was the principle of linguistic variety.

126 IX. Historical summary of linguistics of the early nineteenth century

To understand the early essay of Ascoli it is unthinkable to leave out of consideration the general aspects of time when it was designed and realized. The first half of the nineteenth century repre- sents without doubt the most effervescent period for the history of linguistics, since in a few years they concentrated the most sensational discoveries in the Indoeuropeistic field - and soon after in that Romance - by some European scholars, in especially Germans. The problems which had caught the attention of the first scholars were represented by the philoso- phical search of language and by the relationship between languages and peoples. They were stud- ies that had not yet focalized the specific objectives that will be later of scientific language, but which proceeded on a random line of philosophical and religious kind. It ranged in fact from the ‘monogenesis’, a theory which wanted to derive all languages from a single mother language, to the ‘polygenesis’, which assigned to the languages different origins. This apparent linguistic conflict was masking actually the search for the origin of humankind. Even the study on the relationship between languages and peoples (or races) was bound to raise several controversy between Enlightenment philosophers and Romantic, giving rise to the contrast of so-called theories of the substrate and superstrate. In this case, the linguistic research was ma- nipulated to support different positions which concerned the rivalry and superiority among the peoples of Europe and not only Europe. In this respect you can also place the relentless contro- versy that concerned the common origins of Latins and German, raised by the discoveries of Indo- European linguistics, which showed conclusively the relationship of Sanskrit with the Greek and Latin, and all the other languages that will be included later in the Indo-European family. The framers of such upheavals were largely German scholars, like Friedrich Schlegel, to whom it comes back the first linguistic approach with Sanskrit, Franz Bopp, author of the first comparative grammar of Indo-European languages, Jacob Grimm, who dealt with the comparison of the Ger- manic languages and Friedrich Diez, the first scholar who moved the historical-comparative also to the Romance languages. This method, which allowed the systematic comparison and structural language, was discovered and perfected during this period from these scholars. At that time also the Italian scholars were interested in the new line of linguistic research in the wake of the growing success of German developments without nevertheless reaching significant results. Among the pioneers of linguistics in Italy there were Carlo Ottaviano Castiglioni, Fran- cesco Cherubini and Pietro Monti, but especially Carlo Cattaneo and Bernardino Biondelli, who with their work specifically of dialect investigation, they opened the studies of dialectology in It- aly. As will be seen better from the detailed examination of the linguistic corpus of his essay, Ascoli demonstrates yes to possess some basic of Indo-European linguistics, but also to ignore the com- parative method based on the comparison of the phonetic and morphological elements. This fact leads us to think that when he was writing his short study he did not know yet the linguistic theo- ries perfected a few years earlier by the European scholars quoted above. Ascoli himself confirms this hypothesis in a letter written the 15 July 1847 to his friend Philoxenus Luzzatto: “I miei studj su filosofia delle lingue ecc. datano dal 18 giugno dell’anno scorso”.

127 In view of these observations, the only positive allocation to be granted to young scholar of Go- rizia is that his contribution is the first comparative study between the Friulian - which was still far from possessing dignity of language - and Romanian.

128 Index

Premise 3

I. Introduction 4

II. Graziadio Isaia Ascoli 6 II.1 Premise 6 II.2 Jewish community of Livorno 6 II.3 Gorizia 7 II.3.1 Jewish community in Gorizia 8 II.4 Graziadio Isaia Ascoli in Gorizia: his younger days 10 II.5 Graziadio Isaia Ascoli in Milan: maturity 16

III. Linguistics of early nineteenth century 17 III.1 General considerations 17 III.2 Biographical profile of earlier scholars to Ascoli 20 III.2.1 Foreign scholars 20 III.2.2 Italian scholars 23

IV. Sources and models of Ascoli’s essay 29 IV.1 Historical grammars before Ascoli 29 IV.2 Romanian 31 IV.2.1 Cattaneo and Romanian 31 IV.2.2 The Romanian Grammar of Ion Molnar 34 IV.2.3 The term «valaco» 37 IV.3 Friulian 38 IV.3.1 Cattaneo and Friulian 39

V. Analysis of «Schizzo storico filologico» of Graziadio Isaia Ascoli 41 V. 1. Previous reviwes 41 V.1.1. The letter of Samuel David Luzzatto 41 V.1.2 The article of Carlo Salvioni 43 V.2 Description of the book 44 V.3 Historical premise of Ascoli 45 V.4 Comment on the historical premise of Ascoli 49

129 V.5 Linguistic section 51 V.5.1 Introduction 51 V.5.2 Writing of Romanian in Ascoli 52 V.5.3 Writing of Friulian in Ascoli 53 V.5.4 Analysis of the linguistic part 54

VI. Conclusions 113

VII. Bibliographic references 115 VII.1 Linguistics and philology 115 VII.1.1 About Friulian 116 VII.1.2 About Romanian 116 VII.2 Studies about Ascoli, Cattaneo and others 117 VII.3 Works of Ascoli and Cattaneo 119 VII.4 Studies about Judaism 119

VIII. The polemic about the issue of the influence of barbaric languages on the formation of Vulgar Italian: 121 VIII.1 The theory of substrate 123 VIII.2 The linguistic thought of Carlo Cattaneo 124

IX. Historical summary of linguistics of the early nineteenth century 127

Index 129

130