Facebook Twitter
Cover > Thoughts on translation > Language Models and Catalan Translation

Language Models and Catalan Translation

by Joaquim Mallafrè
The process of how a translator acquires a language appropriate for translation is interesting because it transcends the individual, taking place, as it does in a shared cultural framework, whether in an intra-linguistic or an inter-linguistic community (in both cases the language models with any real influence are limited). In Catalonia, where Catalan has long struggled to survive under the shadow of Spanish, there has been some discussion about language models for translations into Catalan. Under Franco’s dictatorial regime, Catalan was banned in schools and the Catalan cultural tradition was eclipsed by the dominant culture. Thus, the codification and the style of the language have been transmitted by an intelligentsia that was very different from those found in more established societies. For example, the prestigious Catalan poet and translator, Josep Carner, was a symbol of a particular cultural movement at the beginning of the twentieth century, the Noucentisme. This movement belonged to a historical period that could not evolve normally and had no natural continuity. Due to this rupture caused by the Civil War and its aftermath, a present-day reader may find Carner’s style strange. The same is true of the style of Carles Riba, whose translation of The Odyssey is “a model for what some would have whished to be a standard model of cultured Catalan that was brutally frustrated”, according to Parcerisas (1997: 471). Parcerisas includes it amongst “those works that have marked and conditioned how translation has been understood” (485).


Literary canons and linguistic norms

This understanding, the fruit of a literary school and certain political conditions, has been combined with linguistic norms. Together, they have certainly achieved undeniably efficient results. However, there are critics that blame both the literary tradition and the strictly linguistic regulations for creating serious limitations or imposing a single model, that act as a dead weight round the necks of creative writers and translators.

This is the position of Pericay and Toutain (1996), who provide arguments and data, with reference to prose in particular. Their criticism is both skilled and useful and it makes sense to me as a translator; which is justification enough. Nevertheless, they do not so much criticise the model itself as models they consider to be poor imitations. So, despite their contribution, I would question some over-simplifications, their choice of certain examples and partial criticisms of some translators.

In my opinion, there is a tendency in Pericay and Toutain — more exaggerated in other less well documented critics — to group together both formal creations and expressions widely used by Catalan speakers as artificial creations belonging to the Noucentisme, and as such to be excluded. This is hardly justified, any more than would be the exclusion of Joyce’s use of the Dubliners’ language, or his formal expressions based on Tomas de Aquinas. Formal words, such as, tocom, llur, ensems, de bon antuvi, —perfectly legitimate of their kind — should not be classed together with colloquial expressions, such as, abellir, virosta or enguany, that I, and others, have learned in everyday use, not from noucentist texts.


TT norms, the individual translator and the ST author

In any case, I think too much weight has been attributed to a single tradition, even though it has become the principal, official norm. All language models, authors, or translators transcend the official norm, which is no more than a basic guide, some unified rules in the varied and unpredictable game of literary creation. We should not forget that the first obligation of the responsible translator is to the model imposed by the author they are translating, his thought, language and style.

Therefore, I think it may be of interest to consider my own model of language, which was, of course, nurtured by a linguistic community. But it was first structured by conversations in the family and the society, and only later, by more formal levels of language and a variety of cultural components, some of which transcended Catalonia. These later influences, with some additions and reservations, provide literary and stylistic models and more or less shared norms that make up a common, coherent language, and, as such, can be transmitted and exported. These norms and models are superimposed on and enrich the individual linguistic and literary biography. From this acquired foundation we develop our own creativity, or we dedicate ourselves to the service of an alien creation when we decide to translate.

Through experience, we achieve a basic style in accord with our personality and our epoch. The mechanism is essentially the same in any language where there are writers with whom we find points of contact that bring them closer to us. They become translatable if we have the necessary capacities: reading, literary criticism and the adequate expression to reproduce them in our language (Bush: 1997). Only by reading and recreating a work can translator and readers make it theirs, with the admiration of one who sees that the alien environment, its style, context and musicality can be expressed in their own language. Perhaps, this is why the advantages of reading a book in a language one understands compensates and justifies the inevitable disadvantages of translation, as has been affirmed by writers from Goethe to Joan Sales and García Márquez.

If I try to outline how I have acquired a language that could be developed into a literary language, perhaps this will reveal elements that explain both my model, the choice of certain solutions, and analogies with my contemporaries, from which has emerged a common historical-linguistic base. Obviously, the knowledge of some languages and cultures and not others, and personal tastes or affinities explain certain preferences. Beyond personal differences, it is useful to contrast the experiences of an epoch that may possibly be systematised for teaching purposes.

Different genres permit different translation approaches: the translation of a novel can keep very close to the original; the translation of a play will depend on whether it is to be read as a text or performed on the stage; the translation of poetry tends to adaptation, or the personal poetic projection of the translator. Even in poetry, however difficult the task, there are some translators who are determined to re-express the content of the poems, submitting themselves to the form used by the author. Nevertheless, despite inevitable failures, it is usually the source text that lays down the rules of the translation: the long sentences of 18th century prose, the telegraphic density of certain 20th century texts, the registers, technicalities or allusions of others. I may doubt whether to translate the English “you” by vos, vostè or tu, or an exact lexical equivalent may not be possible because it is unknown by the general public. I may be led astray by false friends due to insufficient knowledge of the source language, or by interference due to insecurity in the target language. But the identity of the work will not be lost if we are able to find literal or equivalent coincidences with the source text.


Catalan at home and Spanish at school

My experience of language began with Catalan, with my family, was widened through contacts with neighbours, country folk, priests, jokes told in cafés, the colourful language of the markets, summer holidays. This childish substratum, was almost erased by schooling in another language, Spanish, with alien models, that were useful in so far as they represented universals. My Catalan was reactivated thanks to translation. When translating Fielding and Sterne, Joyce and Beckett, solutions emerged that I had heard from my elder brother, in the market, at snooker games and in the slang used by gypsies. This is where I found readymade solutions that coincided with the author’s expressions: “Shut your eyes and open your mouth” — Obre la boca i tanca els ulls; “Touch the spot” — Tocar el voraviu; “As uncertain as a child’s bottom” — Insegur com el cul del jaumet; “Both ends meet” — Els extrems es toquen; “Collar the leather, youngun” — Enxoma la bimba, baranda.

Apart from textbooks in Spanish, I learnt to read from old Catalan children’s magazines, where I met the humour of Guillem d’Oloró, or of Josep M. Folch i Torres in the Pàgines viscudes and in the Biblioteca Patufet. Folch i Torres was influenced by Dickens, but he was also influenced by a literary apprenticeship of books and plays that was shared by many other children. Folch i Torres adopted the norms established by Fabra (the first modern Catalan grammarian and lexicographer) and this helped to spread a generally accepted external model. Even before I started to learn Catalan grammar, by reading Patufets, I knew enough to find a name for Mr. Allfours — Mr. Quatrepotes; Lord Walkup of Walkup on Eggs — Lord Trepig de Trepitjous; or Kissarcius — Besaculis.

Curiously enough the catechism classes of my childhood were in Catalan and in the Church I was taught a more abstract language. The prayers of the Catholic liturgy gave me the exact translation of some of the references in Ulysses, that were more immediately familiar to me than to an English speaking Protestant reader, for example.

I did not have to go to school to learn, when, outside the classroom, I could hear popular rhymes, fragments of popular poetry by Catalan writers, such as Pitarra or Verdaguer. I got used to more sophisticated language through the plays performed on Sundays at our local theatre, l’Orfeó Reusenc. My literary education continued with my reading, so when I had to translate “Dear dirty Dublin”, I chose to echo the Catalan poet Salvador Espriu, “Pobre, bruta, estimada Dublín”. My translation of “Return of Bloom” echoed another Catalan poet, Àngel Guimerà, “Bloom que torna”.

In some cases the Catalan translation tradition provided solutions for the frequent quotations found in English authors. For quotations of Shakespeare, I could choose the most appropriate version from the translations of Morera i Galícia, Sagarra and Oliva. When translating Tom Jones, I used Miquel Dolç’s version of the Eneas when Fielding had used Dryden’s; and I used Josep M. Llovera’s version of Horace when Fielding used Francis’s. In Tristam Shandy, when Sterne quoted Rabelais, I used the translation by Miguel Ángel Sánchez Férriz.

Other problems with Tristam Shandy were solved by a horse breeder, who confirmed that a horse that “was spavined or greazed”, “tenia esparavanys o aigüerols”. Documentation in philosophy, medicine, building or military fortifications confirmed my translations of some of the archaic expressions that abound in Sterne, such as “homunculus — homuncle”, “radical moisture — humit radical”, “ravelins — revellins” and “fause-brays — falsabragues”.

My knowledge of the rules meant that I had at my disposal an efficient, homogeneous and organised language model, but I rarely felt it was a restriction. I had no need to reject it in order to feel free to use words that are not in the dictionary, such as “cavallbatallívol” for Tristam Shandy’s “hobby-horsical”, or “marcavent” when searching for an equivalent register for Beckett’s “windgauge”.

It is questionable to what extent younger Catalan readers can assimilate my linguistic experience, and this is a serious problem. However, I console myself by thinking that English readers today are equally distant from the linguistic experience of many of their own authors. It is often forgotten that the original text also ages. Of course, the classics resist the ageing process in so far as they remain familiar for their public, through the force of tradition they are read again and again. When I translate a classic I travel to the land of the original text and I describe it in Catalan to a reader of my language. The success of my translation will depend on the degree to which a Catalan reader can speak to an English reader about the same work, the same characters, about a comparable style that acts as a vehicle for the work, and each one can draw closer, sharing a common world from the perspectives of their respective languages.

Of course, cultural differences often mean that exact translations are not viable, expressions that are clear to the reader of the original text are obscure to the reader of the translation. But what about passages that are obscure in the original text, a text that has aged for readers who have lost touch with their own tradition? The role of the language authorities is especially delicate in these circumstances. Perhaps the translator should concentrate on the difficult task of developing good taste as Allén (1997) suggested. In fact, this is what Carles Riba suggested in his preface to the second edition of Fabra’s dictionary (1954): “Good taste: this is a sentiment that is not hereditary; but it is possible to transmit a set of criteria that will improve the education of each and every one”.

If the author can be creative, the translator can only be creative to reflect what the author says. This reflection has to be as natural or artificial as the original. Pym (1997: 51) was quite right to suggest we should encourage the idea that “one should translate as one would whished to be translated”. This seems to me to be a fundamental requirement.

If language is to be authoritative it is not enough for it to be real, but it has to appear real. The artifice that is constructed using the language of the community and the individual should be plausible, natural for the reader. With time, it may influence the reader’s taste, or, he may even use the language of certain authors. This was what happened with a famous author of crime novels. On being told that he seemed very familiar with the underworld that he was able to reproduce with such realism in his novels, he replied that, in fact, it was the cops and robbers who where familiar with his novels. The translator also has to combine the reality and the seeming reality of the original language and reproduce it with the tools of the target language, starting from a landscape that is alien in space, time and culture.


Catalan literary translation in the eighties: A study of two collections

After the Spanish Civil War, translation into Catalan practically disappeared, due to the political repression. The sixties marked the first major recovery, with the successes and failures described by Broch (1991: 189) and Vallverdú (1987: 103). In the eighties there was a second major attempt to publish translations in Catalan. This was a bigger movement and included initiatives to incorporate different genres, such as the collections of the MOLU (Les Millors Obres de la Literatura Universal), MOLU XX (Les Millors Obres de la Literatura Universal del segle XX), Poesia del segle XX, Textos filosòfics, Clàssics del pensament modern. This was possible, in part, due to institutional support and aid from savings banks. In this period the most important works of Joyce, Proust, Kafka, Maiakovsky and Henry Miller were also translated and rigorous criteria were used to translate Boccaccio, Shakespeare, La Fontaine, Sterne and Melville.

In this article, two of these collections will be described: MOLU and MOLU XX, directed by Joaquim Molas and assessed by J.M. Castellet and Pere Gimferrer. The first collection included 50 books published between 1981 and 1986. The first 50 books of the second collection were published between 1986 and 1990. The collections are interesting, not only from the point of view of the translated texts themselves, but also because of the sociological context in which they were produced. They provide us not only with the translations, but also with a whole series of other criteria that can throw light on translation studies. As Parcerisas (1997) suggested, these criteria are commercial, political, institutional and historical, and they may all influence the language model, or models, used. For example, these two collections cannot be explained without the institutional aid from the Servei del Llibre of the Catalan Ministry of Culture and the combined effort of the publishers, Edicions 62, and “la Caixa” savings bank.

It is very difficult to speak of a single language model for these collections, given the variety of genres, source text languages, authors and translators, as the following description shows.

The predominant genre was the novel, taking the place of poetry, which had been very important in Catalan translation in the past. Of the 100 books studied in the two collections, 73 were fiction (34 and 39 in MOLU and MOLU XX, respectively), 14 were poetry (9 and 5) and 13 were drama (7 and 6).

The choice of source text languages shows that our vision of universal literature remained linked to the Western world. There are only 12 languages: English 32 (13 and 19), French 21 (13 and 8), Italian 15 (6 and 9), German 15 (7 and 8), Russian 9 (7 and 2), Galician and Portuguese 3 (1 and 2). In the first collection, MOLU, there are single examples of translations from Latin, Provençal, Swedish and Norwegian, and in the second collection, MOLU XX, there are single examples of translations from Polish and Modern Greek.

The fact that 25% of the translators were also writers proves the dedication of many Catalan authors to translation. Leaving aside the translators of anthologies that included the work of several translators, many of the translators/authors are best known for their own work as writers. 13 of the books were translations by Josep Carner, Manuel de Pedrolo, Josep M. de Sagarra, etc., which had been published previously, but revised for these collections. 12 were translations by 9 modern authors, Pere Gimferrer, Miquel Martí i Pol, Quim Monzó, etc., prepared especially for these collections. However, it is interesting to note the growing number of translators who are not creative writers in their own right, some of them are academics who have specialised in the subject or the author of the book they have translated. It is significant that 48 of the translators also wrote the introductions to the books they translated.


Translation criteria expressed in introductions and footnotes

On the whole, the translation criteria used in these collections show that the books are aimed at the general reader rather than the academic. As Josep Murgades warned when presenting the translation of Goethe, “A translation is not supposed to be a critical edition” (MOLU 41: 11). J.M. Güell did not attempt to represent Gogol’s linguistic variations because he considered them to be “more suitable for an edition addressed at specialists and academics than for an edition addressed at the general public”. Despite this non-academic approach, the collections are full of translator’s notes, some of which illustrate interesting theoretical considerations.

There are different opinions about translator’s notes. Some claim they have no place in an artistic translation, as the translation has to be solved in the text alone, others use them generously, or keep them down to an absolute minimum, using them when they are absolutely essential. It would be useful to study the criteria used and to try to systematise this area where anarchy often reigns, not only in Catalan translations, but elsewhere as well. The MOLU collections allow a great deal of freedom for individual initiatives, although there are some attempts to systematise. In the second collection the translator’s notes are always explicit and regularly marked with an asterisk. However, in the first collection there are notable irregularities.

In the first place, the translator’s notes in the footnotes are not always really translator’s notes. Although it is true that the limits are hard to establish, I believe there are some that are clear. These are: explanations of word play, borrowing (by the author or the translator), the translation or not of quotations, verses included in the original language in the text and translated in a footnote or vice versa, the recognition of earlier translations used by the translator, specific allusions or references that would not be directly understood by the translation reader and require translation decisions. There are other notes that do not affect the translation and are really editorial notes, for example, notes that clarify who a person is, or what an institution is, or encyclopaedic notes. Secondly, it is not always clear if the notes are from the original edition or by the translator. It would be useful to make the distinction. Thirdly, notes that seem to correspond to specific translation problems are not always marked as being by the translator.

Discussion of translation criteria can also be found in the introduction or in a preliminary translator’s note. Many of the theoretical considerations are concerned with faithfulness, in one or other of the senses of the notion, although none would defend the use of archaic language to reflect the original. “As far as the language is concerned, we have tried to maintain, as far as was possible, the unpolished and sometimes confusing style of the original. We have not altered the minimal change in style between the narrative and the dialogues, a change that is hardly noticeable in the original” (Desclot, MOLU 11: 7).

Faithfulness to the source text may make the translator consider a literal translation. “I have preferred to run the risk that my translation may sound rigid because it is so literal, rather than risk the debilitating loss of an overly free translation”, wrote Murgades about his translation of Goethe (MOLU 41: 10). Joan Casas tended towards a literal respect for the original in his translations, despite the difficulties this involved, “so that the deformation is minimal in a process that is in itself extremely traumatic, ‘a betrayal’, the process of passing from the mould of one language to another”. “No concessions should be permitted under the excuse of bringing the original ‘up to date’, a concept which is, after all, no more than a subjective impression” (MOLU 42: 11). The resulting syntax could be complicated, as Joan Casas himself showed with an example of his translation of a verbal structure by Saint-Simon, “es va deixar arrossegar a gosar fer saber que desitjava” (MOLU 37: 10). Casas also translated Bassani, whose precise language was “a sort of spirit that Bassani’s own prose recommended to the translator, who soon found himself submerged in the same feeling, in the certainty that to alter the sense of an adjective, the rhythm of a sentence, could open an irreparable breach in the soundness of the structure” (MOLU XX 33: 10). But when Casas had to translate Céline, he opted for a ‘possible’ translation, given the poverty of non-standard urban registers in Catalan. The lexical limitations were compensated by the use of oral syntax, rhythmical sentences and set expressions that were not necessarily literal translations (MOLU XX 19: 11-12).

Thus, when a literal translation would be in conflict with faithfulness to the original text, because the literal translation would be incomprehensible, the translators opt for a certain degree of freedom, for example the translations of Svevo and Beckett. The same approach is obviously inevitable when translating poetry. As Miralles remarked, translated poetry has to be read “not only as a text that tries to reflect the original correctly, but also as a text with ambitions to achieve a poetic result in the target language” (MOLU XX 25: 16). The translations of the anthologies vary according to the translator’s sensitivity and skill and do not always reflect the original poets’ sensitivity and skill. According to Alain Verjat, this approximation is perfectly valid because “the perfect translation does not exist” (MOLU 44: 18).

The comments made by the translator of Moby Dick, M.A. Oliver, are particularly interesting in this context. She cites the informants who helped her to find Catalan words used in navigation and how to adapt the vocabulary related to whales, animals and plants, the set expressions and registers, to identify the Biblical references or help her to understand certain difficult constructions in English (MOLU 30: 12).

The question of language models and Catalan translation was also sometimes raised in the introduction or in the translator’s preliminary note. Differences on this point can be seen between the new translations and the 13 books that were re-editions of earlier translations within the tradition of Noucentisme. In the earlier translations there are references to the influence of translation on Catalan literature and language. For example, in 1921 Carle Riba talked of “patriotic egoism” when he saluted the translation of Molière into Catalan by Josep Carner. Riba’s appraisal was reproduced as a foreword to the MOLU edition, “This is our hope, that the Catalan word, following the path laid out by the admirable translator, may reach its final and most difficult triumph: to reign supreme in comedy. We need a language that is independent of Rambouillet Palace, a language both fit for the ear of the Great King, and suited for the good bourgeois, whether he believes himself to be a gentleman or not” (MOLU 2: 6). However, even a contemporary writer, like Gimferrer, translator of part of L’Éducation Sentimentale, acknowledged the influence of the Catalan style forged by both translators and writers, and the impact of translation on his own writing. He stressed his obsession with “the prose that was the result of Carner’s translations of Dickens, the prose of Josep Vicenç Foix and Josep Pla”. “I gained another verbal obsession from submerging myself in translating Flaubert and this prevented me from writing myself for a year and a half” (MOLU 20: 7).

However, in the books translated after the 1960’s, few translators discussed language models explicitly. It seems as if there was more confidence in the normative and it was not questioned. These rules in the target language and a more or less shared cultural and linguistic experience, such as the one described in the first part of this paper, provided a language model for translation. Nevertheless, we should not forget the importance of each source text in determining the translator’s use of grammar, syntax, register and the personal rhythm of the work of art.



The analysis of a collection of translations like the one we have described provides significant data about translation criteria in our country at the end of an era. At this time, empirical research can incorporate contributions from the philological and the linguistic stages, as well as the most recent tendencies in translation studies within the framework of communication theory. The present panorama with regard to translators has been summed up by Narcís Comadira. Translation is possible, even though there is no such thing as a polyvalent translator, that is to say, “not everyone can translate everything”. “There are translators who are traitors and translators who are incompetent. The former falsify because of their pride, and the latter because of their ineptitude. However, there are good translators: these are the ones with the adequate technical knowledge who are able to silence their own voices so that, in symbiosis with the other, the voice of the poet they are translating can be heard” (MOLU 40: 11-12).



Beeby, Allison, D. Ensinger & M. Presas (eds.): Investigating Translation. Amsterdam; Philadelphia: John Benjamins, 2000.
Translated by Allison Beeby

Isabel Banal: Llapis trobats, sèrie iniciada el 1999.

Search for articles
A-B-C-D - E-F-G - H - I
J - K - L - M - N - O - P - Q - R
With the support of: