Dialectologia. Special issue, VII (2017), 19-35. ISSN: 2013-2247 Received 28 August 2017. Accepted 28 October 2017. ANTONI M. ALCOVER AND ILDEFONS RUL·LAN: TWO VIEWS ABOUT THEORETICAL LEXICOGRAPHY AT THE BEGINNING OF THE 20TH CENTURY IN MAJORCA1 Eloi BELLÉS Universitat de Barcelona*∗ [email protected] Abstract This paper studies the conception of lexicography in Ildefons Rul·lan’ work (Palma 1856-1911) and compares it with the one that appears in three lexicographic works by Antoni M. Alcover: the Mostra de Diccionari mallorquí (1882), the Lletra de convit (1901) and the Diccionari català-valencià- balear (1930). Several views about lexicography, etymology and the role of proverbs in a dictionary are shown. Finally the thesis about the existence of a Majorcan (linguistic) school around the members of the journal L’Ignorancia, particularly Tomàs Forteza, is evaluated. Rul·lan and Alcover are part of this school. Keywords lexicography, history of linguistics, linguistic ideologies ANTONI M. ALCOVER E ILDEFONS RUL·LAN: DOS VISIONES SOBRE LA CONCEPCIÓN DE LA LEXICOGRAFÍA EN LOS INICIOS DEL SIGLO XX EN MALLORCA Resumen Este trabajo estudia la concepción de la lexicografía en el trabajo de Ildefons Rul·lan (Palma, 1856-1911) y la compara con la que aparece en tres obras lexicográficas o reflexiones sobre esta labor de Antoni M. Alcover: la Mostra de Diccionari mallorquí (1882), la Lletra de convit (1901) y el Diccionari 1 This paper takes its origins in Bellés master’s thesis L’obra lingüística d’Ildefons Rul·lan, presented at the Universitat Autònoma de Barcelona in July 2017, and supervised by Maria-Pilar Perea. * Departament de Filologia Catalana i Lingüística General. Facultat de Filologia. Gran Via de les Corts Catalanes, 585, 08007 Barcelona. 19 ©Universitat de Barcelona Eloi BELLÉS català-valencià-balear (1930). Se muestran las concepciones de la lexicografía, la etimología y el papel que en un diccionario desempeñan locuciones y proverbios. Al final se evalúa la tesis según la cual se podría hablar de escuela (lingüística) mallorquina) en torno a la revista L'Ignorancia y Tomàs Forteza, de la cual formarían parte Rul·lan y Alcover. Palabras clave lexicografía, historia de la linguística, ideologías lingüísticas 1. Introduction Undoubtedly, the history of the Catalan language is linked to two great linguists who lived at the term of the last years of the 19th century and the first years of the 20th century: Pompeu Fabra and Antoni M. Alcover. If Fabra is known as the creator of standardization of modern Catalan, Alcover, in the words of his disciple Francesc de B. Moll (2004: 505), is known as the «[f]undador de la filologia catalana» (‘the founder of Catalan philology’) and, especially, of Catalan dialectology (according to Wilhelm Meyer-Lübke (BDLC, XIV, 1925-1926, 280-281)). The greatest works of both authors are dictionaries: the Diccionari general de la llengua catalana (DGLC) (1932) by Fabra — which was the academic or normative dictionary until 1995 — and the descriptive Diccionari català-valencià-balear (DCVB) (1930-1968) by Antoni M. Alcover and Francesc de B. Moll. The work of these two authors is so great that Moll, in a laudatory speech devoted to his teacher, remembered that before them en aquells primers anys del nostre segle [20th century], la filologia catalana tot just si existia. Un Milà i Fontanals o un Balari, a Catalunya, i un Tomàs Forteza a Mallorca, eren gairebé els únics al nostre país que tenien notícia de les obres dels primers romanistes europeus com Friedrich Diez i Meyer-Lübke (Moll 2004: 507). Nevertheless, Moll in his speech did not talk about the development of linguistic studies in different areas of the Catalan speaking territories during the 19th century. It 20 ©Universitat de Barcelona Dialectologia. Special issue, VII (2017), 19-35. ISSN: 2013-2247 is true that many of the linguistic works published during this period can be considered as traditional ones. They are often grammars, dictionaries or spelling works which aim is either to teach their own language or to facilitate the learning of the Spanish language (see Rico & Solà 1995: 19-20). However, since the second half of the century there has been a certain interest in linguistics per se, and some scholars begin to be in contact with the European developments in Romance philology and have news about the works of Rasmus Rask, Franz Bopp, Wilhelm von Humboldt, Friedrich Diez and Wilhelm Meyer-Lübke. Among these first Catalan linguists and grammarians are traditionally mentioned Manuel Milà i Fontanals, Josep Balari and Tomàs Forteza. But others should be added: in Catalonia, Alfons Par, Josep Calveras, Marià Grandia and (perhaps) Jaume Nonell, together with Pompeu Fabra and the group of L'Avenç; and in Majorca, Antoni M. Alcover, Marià Aguiló and Tomàs Forteza. Aguiló and Forteza greatly influenced Alcover’s linguistic conscience (Massot 1983: 79). In this group of Majorcan scholars, there were other lesser-known linguists, among which Ildefons Rul·lan, born in Palma in 1856, stands out. He led a life always related to linguistics: he studied Arts and Humanities in Barcelona with Josep Balari between 1880 and 1884, was a teacher of rhetoric, wrote several linguistic works and translated El Quijote into Catalan.2 Despite the fact that Rul·lan lived in Barcelona during his studies, he participated in the intellectual and linguistic life of Palma (Majorca) and was part of the core of L’Ignorancia — one of the main groups that implicitly defended the unity of the Catalan language, among which there was Alcover (Massot 1985: 106-114). The most important work by Rul·lan is the Apuntes para un vocabulario etimológico mallorquín (Rul·lan 1899), published between 1899 and 1902, which includes a reflection about lexicography and explains how a dictionary should be. Rul·lan and Alcover were friends since their youth, and both started their linguistic careers at the same time. In 1881 L’Ignorancia organized a contest to award the best sample of a Majorcan dictionary. This contest invited to «presentá ben espinzellades tres paraules, vêus ò térmes de ses més notables en sa nòstra llengua, y 2 See Bacardí & Estany (2006), Riera & Cotoner (2005), Bordoy (1956) and Bellés (2017) for a larger Rul·lan’s biography. 21 ©Universitat de Barcelona Eloi BELLÉS es verbs fér, dur y anar, junt amb tots es modismes en qu’hey sònan» (quoted from Perea 2001: 15). Alcover obtained the first prize and Rul·lan, who participated by the pseudonym Un estodiant doblegat, won the second one.3 Their interest in lexicography lasted for a long time and at the turn of the century both began a new dictionary. Firstly, Rul·lan published the already mentioned Apuntes, an etymological dictionary ordered by etymons. In 1901 Alcover published the Lletra de convit (‘Letter of invitation’) (Alcover 1902), where he invites all Catalan society to collaborate in his large dictionary project. The aim of this paper is to analyse and describe the theoretical conceptions that Alcover and Rul·lan had about lexicography, and debate over the hypothesis about the existence of a group of scholars with common linguistic references in Majorca in the late 19th century. 2. Lexicography understood by Ildefons Rul·lan Rul·lan (1899) includes some reflections about how a dictionary should be. They are analysed in this paper, after describing the conception and the structure of the work and giving some examples. The Apuntes is a project about an etymological Majorcan dictionary, but it only takes into account the letter A, which gives the dictionary the aspect of an unfinished work. However, a conclusion closes the work and a very interesting prologue (Rul·lan 1899: 5-55) includes several ideological, historical and linguistic views. The introduction of the Apuntes consists of a letter as a preface in which the structure of each lexical entry is presented. It begins with a consideration about etymology and philology as sciences, and about the use of the comparative method. The author justifies the use of Spanish rather than Catalan as the metalanguage and explains the process of Spanishization since the 18th century after defending the unity of Catalan and the Majorcan dialect. Rul·lan also compares some historical processes 3 See Massot (1985: 76, No. 28). Bacardí & Estany (2006: 139) erroneously asserted that Rul·lan obtained ad aequo with Alcover the first prize. Riera & Cotoner (2005: XVI) consider him the unique winner of the award. There is no mention of this prize in Perea (2001). This first work by Rul·lan is lost. 22 ©Universitat de Barcelona Dialectologia. Special issue, VII (2017), 19-35. ISSN: 2013-2247 of linguistic change between Catalan and Spanish, which come from Vulgar Latin. Finally, some spelling notes and a list of references close the Apuntes. The second part of the Apuntes, which are the lexicographic part, is a mixture of etymological and cultural, historical and linguistic notes. It includes information about the sources used, which are basically documentation from the dictionaries by Pere Labèrnia (Barcelona, 1839-1840) and Juan José Amengual (Majorca, 1858-1878) and also from some medieval Catalan texts. The ordering of the words in the vocabulary is not alphabetical, as could be expected. Words are ordered by root, and each root includes derivatives and related concepts, in a similar way that Coromines ordered them in his dictionaries. Finally, the conclusion of the Apuntes contains some considerations about how the perfect dictionary should be (see Rul·lan 1899: 138-141). Rul·lan (1899: 139) thinks that “el diccionario de una lengua debe ser al mismo tiempo la gramàtica é historia de la misma” and, if possible, it must contribute to establish the history of the language itself. By history he understands both the internal history, that is, a description of the phonetic and morphological changes that the words have undergone, and the external influences that they have had. Thus, in the Apuntes Rul·lan indicates the words and expressions that have a Castilian origin (for instance, see Rul·lan 1899: 95, where he explains that agulla del sombrero ‘needle of the hat’ comes from the Spanish). The author also reflects about the lexicographic methodology, and adds that “no pueden desconocerse la fonologia y morfología del latín y del griego” (Rul·lan 1899: 139), since it is in those languages where the etymologist and the lexicographer must settle. It is obvious that Rul·lan has taken into account the Romance studies, and he declares: para tener un conocimiento exacto de una lengua neo latina se necesita también hacer un detenido estudio de su fonología y morfología, comparándolas con las de sus hermanas; empezando por el análisis de cada letra, y siguiendo por el de los grupos para formar afijos, por las permutaciones y substituciones, hasta llegar á la formación de las palabras (Rul·lan 1899: 139). 23 ©Universitat de Barcelona Eloi BELLÉS Later on, spelling, which is not the result of the routine, the caprice or the mere authority of the classical authors, can be imposed. From this view, Rul·lan conceives a dictionary as a total work that must contain all the possible information about a language: the lexical explanation of each word, with etymological information, phonetic and morphological aspects — even diachronic data — and the spelling system to write it. Finally, the author also reflects on the system of the dictionary word ordering. He defends the system adopted in the Apuntes: an ordering based on roots and not a conventional alphabetical order, because [n]o cabe dudar de que es más científico, abriendo el camino á investigaciones que el mismo lector se complace en seguir, y convirtiendo el libro en algo más que un mero almacén de palabras (Rul·lan 1899: 139). Hence, he does not want the dictionary to become a mere list of words, similar to the Eduardo Benot’s Diccionario de ideas afines, which is one of the bibliographic references of the Apuntes (Rul·lan 1899: 55). However, a dictionary is not a linear reading tool, but rather a query tool. Therefore, a mere alphabetical order can sometimes facilitate a specific query, a task that is really difficult in works ordered etymologically if they do not include a proper index of words, as in the case of the DECat by Coromines (see Colón 2011: 42). In the Apuntes, the information about methodology is not clear. In the first pages, when Rul·lan honours his teacher Josep Balari, he says: sus consejos [of Dr. Balari] […] son: recoger notas filológicas, siempre y en cualquier sitio demos con ellas, guardarlas cuidadosamente y clasificarlas cuando tengamos oportunidad para ello. […] Hace tiempo que recojo notas filológicas donde quiera las haya encontrado, y éstas son las que me propongo presentar, deseoso de contribuir al estudio de nuestra lengua. […] 24 ©Universitat de Barcelona Dialectologia. Special issue, VII (2017), 19-35. ISSN: 2013-2247 El primer trabajo que debería tomarse es el de clasificarlos, aprovechar los útiles y separar el ripio. Esto es lo que puedo hacer con mis notas, disponiendo las aprovechables en forma de apuntes (Rul·lan 1899: 6). Next, without mentioning which sources are considered, Rul·lan explains the structure of the book. At the end of the introduction, just before the vocabulary, he offers a short reference list: the “Abreviaturas de las obras que se citarán con más frecuencia”, which includes dictionaries and other works. Although he does not state that all the content of the Apuntes is based on this list, it can be deduced that most of them come from it. After examining the list, some conclusions about the sources used by Rul·lan can be drawn. On the one hand, many of the lexicographical materials are not only Spanish, but also Catalan. Thus, it includes the most important Catalan dictionaries of the 19th century: the Labernia and Amengual’s dictionaries (see Colón & Soberanas 1986: 153- 160; 177-178), and also the Diccionari Torra, of the 17th century.4 The Diccionari Amengual is the most important Majorcan dictionary since the publication of the DCVB. It was also one of the main sources used by Alcover in his Mostra de diccionari mallorquí (Perea 2001: 12). On the other hand, the other works are mainly medieval texts, with special importance of the Spanish ones. The author includes literary texts as the poem of El Cid and Calila é Dymna. Concerning Catalan, he includes the works by Ramon Llull, Ramon Muntaner, King Pere IV, etc., and some modern ones; for instance, the journals La Renaixença and L’Ignorancia. Finally, in a third section, he states: los nombres de los autores mallorquines, catalanes y valencianos que se citen, serán fácilmente entendidos cuando se escriban abreviados (Rul·lan 1899: 56). The mention of abbreviations suggests that Rul·lan was not systematic in the compilation of the materials that were to provide the words for his dictionary. In 4 A usual practice of traditional lexicography was to copy and update dictionaries as a main tool to create new works. See Perea (2001: 57) and Colón & Soberanas (1986). 25 ©Universitat de Barcelona Eloi BELLÉS addition to these theoretical considerations, he also explains how the information of each lexical entry will be structured. It can be summarised in the following paragraphs. Firstly, Rul·lan says that the materials must be ordered by etymology from the simplest to the complex form. For example, casa ‘house’, casar ‘to marry’, caserío ‘homestead’, caseta ‘small house’ (Rul·lan 1899: 7). Secondly, the author studies the etymology of the word — or it should be said the family of the word. This becomes a methodological problem because firstly he associates each word with its derivatives, but the knowledge of the etymon is necessary to carry this out. Thus, using the aforementioned example, casar (‘to marry’) is a derivate form of casa (‘house’) (cf. DECat, II: s. v. casa), although the etymology is not always evident. Thirdly, Rul·lan proposes to mark if a word is common or not. Finally, from an ethnological point of view, he tries to collect proverbs and idioms related to each word. He also expects to include in the same lexical entry some related ideas, based on Benot’s dictionary. However, the dictionary is too short to analyse if he succeeded in this task. The analysis of word abat ‘abbot’ in relation to its etymology, derived forms, etymologies of these forms, information sources, etc. will show if Rul·lan was able to accomplish this aspects. Rul·lan gives the Catalan, Spanish, Hebrew, Syrian, Greek and Latin translation of abat and explains its original meaning of ‘father’ or ‘head of the family’ (cf. DECat, I: s. v. abat: “del ll. ABBAS, ABBATIS, i aquest de l’arameu abba ‘pare’, passant pel gr. άββα”; and Ernout i Meillet (1939, s. v. abba): “[m]ots d’Église, introduits seulement dans la basse latinité, et empruntés sans doute au gr. άββα qui lui-même est emprunté de l’araméen”). Rul·lan (1899: 69-71) offers abundant documentation in ancient Castilian, but what is more interesting is his study about the semantic change of abat from ‘clergyman’ to ‘leader of a monastery’ in the 15th and 16th centuries (cf. DCECH, sv. abad). Form abat the word abadia ‘abbey’ is derived, which, apart from designating the buildings or rooms destined to the abbot, is also the base of “los apellidos Abadía, 26 ©Universitat de Barcelona Dialectologia. Special issue, VII (2017), 19-35. ISSN: 2013-2247 Abadías, Abadiano y el mall. Badía” (Rul·lan 1899: 75). This definition coincides with the DCVB (s.v. badia, acc. 4) and Moll (1987: 208). Several samples of etymological discussion are also found: from abad the word abadejo is derived, which can refer to “tres especies zoológicas” (Rul·lan 1899: 79). These three species are a) “avecilla comunmente llamada reyezuelo ó régulo”, reietó, in Catalan; b) “pescado seco, por otro nombre balacao”; and c) “carraleja”, a sort of beetle. Rul·lan explains why this animal is called thus. First, the bird was called the diminutive of ‘abbot’ “porque en lo vistoso de sus plumas semejaba al traje de los abades primitivos»; the fish, «por el color aceitunado y por tener una mancha amarillenta en la cabeza, semejando su aspecto al del traje de los abades primitivos»; and, the insect, «por recordar [its body] una mitra” (Rul·lan 1899: 79). Next, Rul·lan explains abadejo’s ‘cod’ etymology. From the beginning, he notes that the etymology of a priest (‘abad’) is not found in the Scottish bodach, but rather it has Castilian origins because the appearance of the suffix -ejo. This proposal does not appear in either of the two etymological dictionaries mentioned in the introduction: the Diccionario etimológico de la lengua española by Barcia and the Diccionario etimológico de la llengua castellana by Monlau. The Diccionario etimológico de la lengua española by Barcia (1881-1883, s.v. abadejo), considers that abadejo is a “[f]orma diminutiva de abad, aludiendo á su color aceitunado y á tenir una mancha amarilla en la cabeza, semejante al traje de los abades primitivos”, which coincides with the etymon proposed by Rul·lan. The other Castilian etymological dictionary mentioned, the Diccionario etimológico de la llengua castellana by Monlau (1856), does not include the word abadejo. The Diccionario Enciclopédico Hispano-Americano (Hispano-Americano, from now on) proposes a Celtic etymology (Hispano-Americano: I, 27b: “ABADEJO (del escocés bodoch): s. m. Pez correspondiente al orden de los malacopterigios, familia de los gádidos, denominado por Linneo Gadus Morrhua”). In the definition, Rul·lan also mentions two Andalusian farmhouses called from abadejo — information obtained precisely from Hispano-Americano (I, 29b) — and concludes that it is easier to decide which ones are derived from a abada ‘inflated or bulky thing’ because it is necessary to swell the dry cod with water before eating it. 27 ©Universitat de Barcelona Eloi BELLÉS Coromines (DCECH, s.v. abadejo; DECat, s.v. abat) accepts that abadejo derives from abat or, rather, from ABBADAGIUM, which would be a contribution in the form of a cod that would be paid to the abbots. The option chosen by the DCVB (s.v. abadeco) starting from ABBATICULU derives from abat, which was formed in Low Latin. Finally, Colón (2014: 21-22) considers it a derivative from abat in Spanish, which was exported to the other peninsular languages (see also Colón & Soberanas 1992: 305, 27). Colón’s opinion is the simplest and also the most lucid and credible — the one that would be expected from Rul·lan, as he includes abadejo in the lexical entry of abat. This long example has been useful to illustrate some theoretical aspects that Rul·lan describes in the introduction of the Apuntes. As we have seen, it is a mixture of Catalan and Spanish references, and the latter are the majority. This aspect suggests a new analysis based on the discussion if the word mallorquín in the title of his work is appropriate or not. 3. Lexicography understood by Antoni M. Alcover 3.1 The first step: the Mostra de diccionari mallorquí (1882) Fifty years before the publication of the Apuntes para un vocabulario etimológico mallorquín, Rul·lan had participated in the contest organized by the journal L’Ignorancia. It was won by Antoni M. Alcover, and Rul·lan obtained the second prize. Unfortunately, Rul·lan’s contribution is lost, but Alcover’s Mostra de diccionari mallorquí is preserved and was published by Perea in 2001. Some aspects of the Mostra will help to contextualise Rul·lan’s work. Alcover’s youth work is a bit special. In fact, it is not a classic dictionary, despite its title, Mostra de diccionari mallorquí (Mostra, from now on) (‘A sample of Majorcan Dictionary’). Divided into two parts, the first includes an analysis of the verbs anar ‘to go’, dur ‘to carry’ and fer ‘to do/to make’, and a large sample of idioms and collocations using these verbs. In the second part, Alcover studies the historical evolution of the gender of the words amor ‘love’, color ‘colour’ and olor ‘smell’ from 28 ©Universitat de Barcelona