Joyce's contacts with the Berlitz School were frequent and often decisive. Ellmann identifies the first as occurring in December of 1902, coinciding with the young writer's stay in Paris while he was trying to get his studies at the École de Médicine underway. Overwhelmed by financial difficulties, he found in English teaching something which was already in great demand at the time, a means of subsistence, and the Berlitz School in Paris was an initial opportunity. The required full-time dedication to the School, however, led the writer to opt for the alternative of private classes (Ellmann 113).
In the summer of 1904, having already made Nora's acquaintance, Joyce plans elopement with her and so starts looking for employment, precisely at the Berlitz School. He was not to give his first class in Zurich, where he had been promised a post, but rather in Pola, then a part of the Austro-Hungarian Empire, where he remains until 1905, the moment when, for unknown reasons, he moves or is transferred to Trieste. Hardly a year goes by before he accepts a position in Rome which involves being in charge of the correspondence in English for the Kolb Nast and Schumacher Bank. After less than nine months he returns to Trieste with the hope of resuming his teaching. The Head of the School opposes the move, however, raising a number of objections to his chaotic method of work. Despite this, for a few months Joyce combines private classes with occasional teaching at the School (Ellmann 216-18).
This brief incursion into Joyce's biography underscores a well-known fact, namely that he taught English at the then most reputable school of languages for approximately two years, and must have met, in addition, the teachers of other languages, while having at his disposal the proven Berlitz Methods in the field of modern languages teaching.
Joyce, a presumed polyglot, had studied Latin under the Jesuits and probably little or no classical Greek. He had a Modern Languages Degree in French, English and Italian from University College. It is said that he learnt Norwegian in order to read Ibsen, although no clear evidence exists of the skill he acquired in the language, while he also undertook the study of German in praise of Hauptmann. Like all young Irish university students at the time, it can be assumed that he had a rudimentary knowledge of Gaelic, proof of which can be traced in his works. So it would seem that, before reaching the age of twenty, he was acquainted with seven languages, yet, how much did he know of Spanish?
I shall broach the appropriateness of this question later on, while for the moment I would like to emphasize we are completely in the dark about how and when he might have acquired the knowledge of Spanish that his writings reveal. It is true that, as Ellmann indicates, in the Autumn of 1928, discouraged by the unfavourable reviews of the recently published "Anna Livia Plurabelle" and newly suffering from eyesight problems, he took to bed, interrupting the writing of Finnegans Wake, and made the most of his convalescence receiving Spanish lessons from a private tutor (Ellmann 607). Unfortunately, Ellmann does not indicate his sources, so the information is of little use to us.
In any case, the tuition in Spanish that he might have had in 1928 would have been of little consequence, since by then Joyce had already written part of Finnegans Wake. In setting out to answer the question posed, internal evidence must be brought to bear, and cataloguing the Spanish lexis in Joyce's work one sees an ascending curve of incidence, beginning in Dubliners and reaching a highpoint in Finnegans Wake. In Dubliners, A Portrait of the Artist as a Young Man, Stephen Hero (116), and Exiles there are occasional allusions to Spain, to historical figures who are Spanish, or to historical events linked with Spanish history and culture, although only a smattering, it has to be said. However, strictly speaking, there are no samples of Spanish vocabulary, although at times the abundance of quotations in Latin and lexemes scattered about from the same tongue could lead us into error. One such example is the word "sitio" in Stephen Hero, which undoubtedly is not Spanish, but rather the first person singular, present indicative of the Latin verb "sitio, -is, -ire, -ivi, -itum", meaning "to be thirsty". Neither are there motives that would justify the inclusion of Spanish lexis in the earlier writings, given that the themes developed and their aesthetic treatment do not require it, and Joyce's knowledge of Spanish culture at the time was scant or almost non-existent.
It is true that Joyce received a Jesuit education and that, the Society of Jesus being of Spanish origin, the teaching and religious instruction at Clongowes Wood and Belvedere would have held allusions to Spain, at least to sixteenth-century Spain, if little more. However, Ulysses puts a different face on matters, at least to a certain extent, since the female protagonist is of Gibraltarian-Spanish origin, and though the circumstances surrounding Molly's birth have little to do with Spain, but rather with the prehistory of Ireland, to the migrations of the Spanish Milessians and to the novel's mythic backdrop. Joyce had to learn about the geography, customs and history of the Rock and, perforce, about the language to which Bloom refers in "Forgotten any little Spanish she knew" (U 4.60-61). On the other hand, the writer drew information from books by Richard Ford (505) and Henry M. Field (126), as witnessed in Bloom's subsequent: "My wife, so to speak, Spanish, half that is. Point of fact she could actually claim Spanish nationality if she wanted, having been born in (technically) Spain, i. e. Gibraltar" (U 16.876-79) – taken almost literally from the authors mentioned. Meanwhile Molly sets to thinking: "could I get my tongue around any the Spanish como esta usted bien gracias y usted see I havent forgotten it all" (U 18.1471-72), even suggesting that she could teach Spanish to Stephen and he Italian to her in return: "I can tell him the Spanish and he tell me the Italian" (U 18.1476). To my eye, Molly's Spanish approximates that of Joyce, which would not pass today an elementary exam in the language.
In Ulysses, the references to Spanish culture, history and literature are more numerous, while with the exception of episode 18 Spanish vocabulary continues to be scarce, far more so than French, Italian or German. The cultural and historical allusions are fully justified, particularly in episode 12, where the citizen, so as to glorify the place of Ireland in Europe and diminish that of England, exaggerates the importance of Spain. Within the Spanish vocabulary that appears in Ulysses, two types of entries can be traced: those that correspond to the basic use of the language, as in "como esta usted" (U 18.1471-72); "señorita" (U 13.1209, 15.216, 15.1057, 18.1405, and elsewhere); "calle" (U 15.216, 18.763, 18.1465); "torero" (U 15.1069, 17.1810), and so on; and those entries of more restricted use, knowledge of which would require a perfect command of the language or native-speaker status. Examples would include "meadero" (18.557), and "perragordas" (U 18.779). The meaning of the latter term escapes the majority of young speakers of Spanish today.
Leaving aside for the moment the hypothetical origin of the basic vocabulary, let us surmise about the source of the second type. In the biographies of Joyce we are given to understand that he had no contact whatsoever with Spanish people, excepting the artist César Abín, to whom the Jolases commissioned a portrait of Joyce to commemorate his fiftieth anniversary, which eventually became the well-known caricature in the form of a question mark (Ellmann 645). We also know of the letter that Joyce wrote to Dámaso Alonso, an accomplished scholar and poet, in answer to the latter's queries regarding his translation of A Portrait of the Artist as a Young Man (Selected Letters, 311-13). Yet this is all we know. Dámaso Alonso's letter to Joyce is presently among the documents that the scholar's heirs ceded to the Spanish Royal Academy. Regarding César Abín I have only been able to verify that he was born in 1882 in Cabezón de la Sal, a small town in the province of Santander, that in 1924 he moved to Paris, to return in 1939 to Santander, where he no doubt carried on working as a caricaturist and painter, especially since some of his paintings can still be found in certain churches there, as well as in private collections. César Abín died in 1974, the place of his death being in dispute; some sources indicate Paris, and others speak of his having died penniless and alone in Santander (Pereda de la Reguera 12).
This contact with Dámaso Alonso and César Abín aside, Joyce must have met and had some sort of relation with the teachers of Spanish at the Berlitz School in Pola and Trieste. Moreover, during the approximately twenty years that he lived in Paris, Joyce must have had some connection with the large community of Spanish and Latin American artists who populated the streets and quarters of Paris during the twenties and thirties. At least one of the latter would have been the well-known critic Juan Ramón Masoliver, one-time friend of Valery Larbaud, whom he probably got to know during the three-year period Larbaud lived in Alicante, between 1917 and 1920. In an article published in the journal Destino in 1942, Masoliver recalled having heard Joyce read from "Anna Livia Plurabelle" in the backroom of "Shakespeare and Company" (Santa Cecilia 31-2; 111-12). Who might have been another member of the Joyce côterie is the renowned, influential journalist Augusto Assía who, late in life, told me that he had met Joyce in Paris, although in the light of the circumstances and details given, I suspect that the acquaintance was more than likely the product of the imaginings of an elderly man's memory than a real fact. There can be no doubt that the most reliable informant , as well as the one closest to Joyce, was not a Spaniard, but a Frenchman, Valery Larbaud, who not only spent three years in Spain, but also translated into French the work of Spanish authors as reputable as Pío Baroja, Gabriel Miró and Ramón Gómez de la Serna. The Spanish vocabulary of more restricted use, both in Ulysses and Finnegans Wake, then, would have come from one or several of these sources, given that not all dictionaries include the entries in question.
Finnegans Wake entails a marked change in scholarly tack, for the Spanish element is far greater than in Ulysses. This shift corresponds to the aesthetic project of the work: if Finnegans Wake enfolds in its interstices all histories and all languages, Spanish would naturally occupy its corresponding place, one more extensive than the unsuspecting reader might imagine on the surface. After painstaking and repeated readings over many years, the inventory I sum up below divides the evidence drawn from the definitive text into seven categories:
1- Spanish vocabulary without alteration: 291 entries. Examples: "malo" (23.16); "el mundo" (34-32); "blanco" (49.08); "moscas" (84.01); "tugurios" (155.05); "guardacosta" (172.22-23); "tinto" (208.13); etc.
2- Spanish vocabulary with alteration or possible distortion: 244 entries. Examples: "es demasiada gruarso" (54.17); "ohoh" (54.28); "sastra" (61.20); "brisha" (73.12); "espos of" (144.05); "sacer esto" (168.13); "muchas grassyass" (174.15); "chonchambre! Cinsero!" (182.09); etc.
3- Spanish vocabulary interwoven in English: 174 entries. Examples: "youstead" (10.19); "malmarriedad" (20.31); ""muertification" (58.08); "man so" (203.34); "inchamisas" (233.30); "Towntoquest" (279.08); "I dalgo" (281.n.3); "bogalboyo" (329.17); etc.
4- Allusions to Spanish culture: 142 entries. Examples: "Salamangra" (09.13); Mammon Lujius" (13.20); "Blancovide" (43.24); "celestine" (288.21); "Izodella the Calottica" (349.21-22); "Loper de Figas" (440.17); "Crestofer Carambas" (512.07). etc.
5- Homonymous vocabulary in Spanish, Latin and Italian: 21 entries. Examples: "bella" (178.17); "sol" (235.09, 331.14); "luna" (27.15, 340.32); "virvir" ("vivir", vir+vir), (357.34); "hosties" (566.01), "fonte in monte" (202.09); etc.
6- Spanish interwoven in Italian: 9 entries (wide range of error). Examples: "pippa" (55.16); "cara" (327.02, 358.31); "cabronne" (352.21); "pronto" (353.19); etc.
7- Spanish interwoven in Portuguese: 8 entries (wide range of error). Examples: "dia dose Finnandos" (178.26); "peepestrella" (178.27); "no espellor" (179.30); "fiho miho" (287.24); "volomundo" (416.16); etc.
In the latter two categories, entailing a wide margin of error, I suspect the number of entries would vary considerably on comparing the definitive text with the manuscripts. The total number of entries above is 889, yielding an average of almost two per page. The Spanish element in Finnegans Wake is thus of manifest significance.
In viewing more closely the Spanish vocabulary in Finnegans Wake, one sees a twofold division of entries similar to the types established for Ulysses: on the one hand, entries corresponding to the basic use of the language; and on the other, entries of more restricted use, often idiomatic or slang, the knowledge of which would require fluency in the language. I ventured above in the discussion of Ulysses hypotheses regarding the origins of the second type, and in the pages to follow I will detail the process that has led me to identify the genesis of the first.
The lexical entries proper to the first group would reasonably appear in any introductory manual for the teaching of Spanish or in any phrasebook of common words and useful expressions for the traveller. My first step, then, was to go through one by one the titles of the books and journals that belonged to Joyce and that, as catalogued by Thomas E. Connolly, are presently at the University of Buffalo. Joyce's personal library, or what remains of it at Buffalo, is of little help for the research in question, yet it does reveal Joyce's scarce knowledge of Spanish. Of the 426 volumes and journal issues, those in English, French, Italian and German abound, their relative proportion being reflected by the order, while just four books in Spanish appear: the first by Xavier Abril, published in Lima in 1937 and uncut; a Spanish translation of Eliot's The Waste Land; the complete works of San Juan de la Cruz; and a translation into French, surely by Valery Larbaud, of Ramón Gómez de la Serna, of which only the first eight pages are cut. There is also one number of the journal Hélix, the issue corresponding to February-April of 1930, in which an introductory article on Ulysses and several extracts of the novel translated by Manuel Trens y Ribas into Catalan appear.
What can be deduced from this lengthy digression is that Joyce's reading in Spanish was almost non-existent and, as far as the issue which most interests us in this discussion is concerned, no Spanish glossary of any kind formed part of his private library, or, at least, has come down to us. Strictly speaking this is only a minor setback, since we know that Joyce was not a collector of books: he tended to make use of those he had near at hand, giving them away later or simply letting them get mislaid, and often borrowed others from libraries or individuals. Many of the volumes in Joyce's personal library, in addition, disappeared during the German occupation of Paris. When all is said and done, therefore, the search requires a different plan of action, the only viable one, as far as I can see, being the scrutiny of the methods for teaching Spanish at the time. In this field the study carried out by Aquilino Sánchez Pérez is fundamental. According to Sánchez Pérez what prevailed at the end of the nineteenth century and the beginning of the twentieth were those methods which he calls "natural," derived from the pedagogy of Pestalozzi, included amongst which is that of Maximiliam Berlitz (1852-1921). A German immigrant to the United States established the first school of languages bearing his name in Rhode Island, and afterwards created the famous Berlitz manuals in numerous languages, all of them taking the English method – the first to be published – as their basis.
It is clear, then, that since the Berlitz Method was the best known and the most widely used in North America and Europe, and since Joyce taught at the Berlitz Schools in Pola and Trieste, this Method is the place to find the origin of the Spanish used by Joyce, a conclusion I had long suspected. When viewed as a group, the so-called "natural" methods have many things in common, since the teaching starts off from real situations, beginning with objects and circumstances within the classroom itself, moving on to forms of greetings, vocabulary of travel, expressions used in the street and in restaurants, and so on, always with the most basic vocabulary. Thus, it is not enough to identify words such as "book," "window," "mesa," or set phrases such as "good morning," "how are you?," etc., as included in the Berlitz Method, since these can also be found in other similar methods. The dilemma could only be resolved by falling back on the Notebooks kept at Buffalo, given that Joyce gathered in them indiscriminate information, without apparent order and without alteration, thus permitting the identification of sources.
In the Buffalo Notebooks published to date, edited, transcribed and commented by Vincent Deane, Daniel Ferrer and Geert Lernout, I have found scattered words in Spanish and references to Spanish culture, yet not in sufficient number as to allow the identification of a source. However, the same does not occur in the case of the facsimile edition, prefaced and arranged by David Hayman, where the text appears without notes, commentary or transcription, for which reason it is necessary to decipher, at the very least, the Spanish lexis in it. I will limit my remarks to Notebook VI.B.23, composed between November and December of 1928, according to Spielberg, with whom Hayman seems to agree. These dates coincide with the year in which Joyce received private classes in Spanish, this being a further argument in favour of the dates proposed.
On the 84 pages corresponding to Notebook VI.B.23 I have identified 151 instances of Spanish vocabulary, these being largely coincident with the number Hayman points out in his introduction to the manuscript. In proportional terms, nearly a third of the manuscript is in Spanish. The 151 instances, then, had to be compared with an edition of the Berlitz Method for Spanish prior to 1938, when Joyce concluded Finnegans Wake. The copy on which I have based the coincidence of terms in Spanish with the items found in Notebook VI.B.23, belongs to the edition of 1913. Nevertheless, having consulted others that span the period from 1904 to 1939, I have been able to verify that no variations exist among the different editions and, what is more, all of them confine themselves to translating the original English version as set down in 1878 by Maximiliam Berlitz himself, with the exception of the reading passages at the end of each, which vary in terms of the particular language concerned. This means that the Spanish version would have remained unaltered until at least 1939, the edition of that same year being the last I have consulted.
The collation of Notebook VI.B.23 with the 1913 edition of the Berlitz Libro Español demonstrates without the shadow of a doubt that Joyce had this source before him when jotting down the Spanish vocabulary appearing in the manuscript. This conviction only grows as one carries on the comparison. The collation was conducted in the following manner.
In the first place the transcription was undertaken. To begin with the identification of five terms raised doubts, given that this notebook was written by Joyce himself, while the difficulty in reading his handwriting is well known and, as if to complicate matters further, no copy was made of it by the famous (at least amongst geneticists) Madame France Raphael.
The second stage involved comparing the items transcribed with each of the pages in the Berlitz Libro Español. Already in the early pages of the manuscript I observed that the sequence in which the lexis in Spanish arose followed the progression of the Berlitz Method. "Pardo," "libro," and "ventana" on page 17 of the manuscript, for instance, all appear on page 2 of the Libro Español , while "largo," "pequeño," and "estrecho" on page 17 of manuscript appear on page 14 of the Method. On pages 80 to 81 of Notebook VI.B.23 one finds, among others, the following instances of Spanish vocabulary: "cubierto," "llover," "caen gotas," "el piso," "nos ensuciamos," "volvamos a casa," "ropa mojada," "copos de nieve," "lluvia," "frío," and "junto." Each of these, and in the exact order in which they appear, are to be found on page 44 of the Libro Español. My collation of the Spanish vocabulary proceded in an identical fashion and with identical parallel results through to the end of the manuscript. A skeptical voice would object that, despite the evidence, other manuals or methods might well sequence elementary lexical items in similar order. This doubt is dispelled, however, by the appearance of the proper name "Cristobal Colón" on page 55 of the Libro Español, the proper name to be noted, though without acute accents, on page 91 of the manuscript. Exercise II in the Berlitz Method asks: "¿En qué siglo vivió Cristóbal Colón?". In Notebook VI.B.23 only one proper name in Spanish is to be found, that of Cristóbal Colón. Also, by coincidence, on the previous page of the book, page 54, the lexical item "impresos" appears. In the manuscript the same word may be found on page 86, while the proper name occurs on page 91; but between both terms there exists no word in Spanish. Thus it becomes evident that Joyce had the Libro Español before him, when he used the word "impresos" (page 54 of the Libro Español). Meanwhile, for an unknown length of time he continued to gather vocabulary and ideas from other sources for the next five pages. Then, when going back to Spanish, since he had pages 54 and 55 of the book open in front of him, he would have been struck by the name of Cristóbal Colón, which appears in the first line of Exercise II, and likewise by the word "impresos" which can be found within the first line on page 54.
To conclude, of the 151 instances of Spanish lexis in the manuscript, 139 are found in the Berlitz Libro Español. The parallel sequence of entries is identical in both texts, this being, to my eye, irrefutable proof of the source. Joyce, as I had suspected, used the Berlitz Method as his almost exclusive source of the Spanish he would include in Ulysses and Finnegans Wake, at least as regards basic vocabulary. This conclusion accords with what one would expect in the light of the external evidence above: first, the Berlitz School was the most reputable and widespread centre for the teaching of modern languages in Europe and America in the first half of the twentieth century, while the Libro Español constituted the only existing Method for the teaching of Spanish in the Berlitz School; and secondly, having taught at Berlitz Schools, Joyce must have known its teaching methods.
Connolly, Thomas E. The Personal Library of James Joyce. Buffalo: The University of Buffalo Studies, vol. 22, 1955.
Ellmann, Richard. James Joyce. 2nd ed. Oxford: Oxford UP, 1982. 113.
Ford, Richard. A Hand-Book for Travellers in Spain and Readers at Home. London: Centaur Press, 1845.
Field, Henry M. Gibraltar. New York: Charles Scribner's Sons, 1888.
Joyce, James. Selected Letters of James Joyce. Ed. Richard Ellmann. London: Faber and Faber, 1975.
Joyce, James. Stephen Hero. London: Granada, 1977.
Pereda de la Reguera, Manuel. "Prólogo." Cesar Abin, "Leurs Figures" 56 Portraits d'Artistes, Critiques et Marchands d'aujourd'hui avec un commentaire de Maurice Raynal. Paris: Muller, nd. Spanish translation: César Abín, Retratos de Artistas, Críticos de Arte y Marchands. Santander, 2003.
Santa Cecilia, Carlos G. La recepción de James Joyce en la prensa española (1921-1976). Sevilla: Secretariado de Publicaciones de la Universidad de Sevilla, 1997.