The Impact of Teaching Adverbial Pronouns in L4 (French) on L1/L2 (Catalan) Writing Scripts

Josep R. Guzman

guzman@uji.es

Universitat Jaume I, Spain

ABSTRACT

The aim of this study was to determine whether learning how to use certain microparametric aspects – adverbial pronouns – in L4 (French) had an effect on students’ L1/L2 (Catalan) writing scripts. The study was conducted on 427 learners who had Catalan as their L1 or L2, English as an L3, and were studying a translation degree at university. The students were divided into two groups according to whether they were studying French or German as their L4. Altogether 1,620 texts produced over four academic years were reviewed. Although both French and Catalan, unlike German, have similar adverbial pronouns, the results of the analysis show that there is no statistically significant difference in the use of adverbial pronouns between the German group and the French group. However, there is some evidence of a more intense use of these pronouns among the Catalan learners in the French group.

Keywords: Writing Scripts; Cross-linguistic Influence; Adverbial Pronouns; Translation.

I. INTRODUCTION

In recent times, the increasing recognition of and respect for the existence of multilingual communities has lead multilingual education to gradually take on a more central role at various educational levels (Cenoz & Görter, 2012; Portolés, 2020), both in the Valencian education system (Baldaquí-Escandell, 2020) and in other communities in Spain (Cenoz, 2009; Slabakova & García Mayo, 2015). With this approach, we move away from the monolingual approach to language teaching and learning that has been predominant for so long.

When we speak of multilingualism, we are talking about a concept that can be considered from a social or an individual point of view, where in this second case it is often referred to as plurilingualism (Council of Europe, 1992; Aronin, 2018). In this paper we will use the term multilingualism in a generic sense as the contact of two or more languages in a learner, or a group of language learners, and we specifically move away from the distinction between bilingualism and plurilingualism. The reason for this decision lies in the fact that in many cases we are talking about bilingual individuals or individuals in bilingual contexts and learners of a third language, as frequently occurs in the field of translator training. In these studies, the overlapping of languages has been present right from the start. In other words, we are referring to the competences, in this case basically of a linguistic nature, which each translation student develops according to their linguistic repertoire.

Research on the role of multilingualism in the acquisition and use of third languages has grown exponentially since the turn of the century. This research has taken the form, for example, of studies on attitudes towards different languages, especially with English as a reference (Lasagabaster, 2005; Lasagabaster & Huguet, 2007; Ibarraran et al., 2008; Portolés, 2015; Madariaga et al., 2016), pragmatic awareness or production (Safont-Jordà, 2005; Cenoz, 2007; Alcón-Soler, 2010; Safont-Jordà, 2012; Safont-Jordà & Alcón-Soler, 2012; Stavans & Webman-Shafran, 2018), sociolinguistic, cultural and translinguistic analyses (Gorter & Cenoz, 2015; Gardner-Chloros & Weston, 2015; Beatty-Martínez et al., 2020) or the transfer of certain linguistic patterns between languages that are related to a greater or lesser extent (Rothman, 2014; Cuza et al., 2018; Ortin & Fernandez-Florez, 2019), to mention, as we said, but a few examples.

The transfer of linguistic patterns between languages, or more specifically the transfer of similar usages between languages, is a factor that is clearly involved in studies where language contact is a determining factor. As pointed out by Cenoz and Gorter (2014, p. 240), “…languages have been considered as separate entities and the transfer of elements from one language into another has traditionally been regarded as negative”, adding that this stance is related to a monolingual ideology that establishes firm barriers between languages. This isolation is clearly manifested in educational contexts and marks a clear separation between languages. Generally, and in the context of multilingualism, this osmosis between languages has been seen, on the one hand, in the framework of code-switching or code-mixing in oral (Musk, 2010) or written contexts (Sebba et al., 2012), to mention just a couple of examples. On the other hand, it has also been observed in the reinforcement of the accomplishment of shared linguistic patterns acquired during the study of previously learnt languages. Research in this last line, and more particularly that referring especially to the third language (L3), has given rise to different theoretical models, which coincide in the fact that multilingualism is conditioned by the L3 learner’s prior knowledge of the first (L1) or second (L2) languages. However, on the one hand, we have the cumulative improvement model that considers that grammatical elements acquired in previously learned languages improve the acquisition of the next language but in a non-redundant way (Flynn et al., 2004; Berkes & Flynn, 2012). This means that the mind avoids repetition and redundancy, and does so regardless of the order in which the languages are acquired. On the other hand, the typological primacy model proposes that it is the similar structures of previously learned languages that are transferred to the L3 (Rothman, 2010, 2014). That is to say, “…the extent to which two or three languages in contact are typologically closer/ more similar leads to either facilitative transfer or non-facilitative transfer due to economy principles…” (Cuza et al., 2018, p. 7). Consequently, structural similarity between the L3 and previously acquired languages would facilitate transfer.

As we can see, in all these cases we are talking about a process of L1/L2 → L3, with a whole variety of languages involved, with or without linguistic proximity, and in certain cases with typological proximity. Nevertheless, we must not overlook the effect of subsequently acquired languages on the L1. That is to say, language transfer must be understood as a bidirectional phenomenon (here we will not discuss the various terms with which the phenomenon is defined) where multilingualism occurs as a dynamic and cumulative process in which the languages interact in both directions. This influence of the L2 on the L1 has been studied extensively, especially as regards the acquisition of L2 (Kecskés & Papp 2000; Kecskés & Cuenca, 2005; Kecskés, 2008), where the need to understand the constant interaction between the channels of the two languages is underlined, thus highlighting the bidirectional influence. Furthermore, the effect of L1 on L2 is clearly differentiated from the inverse process, in the sense that “L2 influence is cognitive and pragmatic rather than syntactic or lexical” (Kecskés, 2008, p.34). This bidirectional relationship has been studied at various educational levels and with different pairs of languages. Thus, Liu and Ni (2016) studied the effects of L2 (English) on L1 (Chinese) at the semantic level on the use of question tags in a group of speakers who have little exposure to the L2. In this paper the authors concluded that there was an influence between knowledge of question tags in English (L2) and “tag question conduciveness judgment” in Chinese (L1), noting that this occurs in a context in which L1 is clearly dominant. In university settings, the use of English (L2) writing by speakers of Farsi as their L1 has been studied by Agheshteh (2015). Although the aim of the paper was to examine the concept of “dual language” rather than “interlanguage”, the results indicated a positive effect of L2 on L1 writing, in that bilingual (Farsi/English) subjects were more proficient than monolingual subjects. Similarly, this bidirectional relationship has been studied between Spanish (L1) and English (L2) and in oral production (Luque-Agulló, 2020). This work points to the existence of involuntary transfer beyond certain very general lexical and syntactic aspects and with a positive nuance among more advanced learners. The study by Brown and Gullberg (2008) examined the transfer between Japanese (L1) and English (L2) in the expression of motion and found that there was a difference between monolingual speakers of Japanese as opposed to bilingual speakers, since the latter made a shift towards the English rhetorical style in the expression of motion. Finally, a last study (Tsang, 2016) analysed the use of plural markers in combination with quantity markers by Chinese (Cantonese) speakers (L1) with English as their L2 and who were learning French as an L3. The conclusion was that there was a certain transfer from L3 to L2 especially in subjects with higher proficiency in L3.

While a large number of studies have analysed the influence of L1 on L2, a smaller number have studied the influence of L2 on L1, or of L3 on preceding languages. All these studies have been carried out taking into account various cultural and linguistic aspects (phonetic, lexical, syntactic, pragmatic, discursive), and with subjects who were bilingual or exposed to the L2 from school up to adults at university.

Nevertheless, there is an area where the overlapping of languages is extremely significant, as we pointed out at the beginning of this work. University translation studies are the natural environment where all the necessary conditions occur for osmosis between languages to take place. The aim of these studies is for students to acquire professional skills in several languages and, in some cases, in territories where there are two official languages, which means that at least four possible languages are involved. In these cases, the effect of some languages on others is an element that has not been studied sufficiently. To simplify the structure of influence of certain languages upon others in this context, we would speak of the following structure: Ln ↔ Ln-x, (L = language; n= order of acquisition of the language; ↔ = bidirectional influence; x= any other language). In other words, any language, or any linguistic, discursive or cultural aspect, could have an effect on similar aspects in a previously or subsequently acquired language regardless of the order of acquisition. This study is carried out with the aim of testing this possible transfer.

II. THE PRONOMINAL CLITICS EN AND HI

Some Romance languages, in particular French, Italian and Catalan, unlike others such as Spanish, Portuguese or Romanian, currently have a series of adverbial pronouns derived from the Latin deictic forms INDE and IBI/HIC. In Catalan, the topic has been addressed in the grammars of Badia i Margarit (1994: 474-478), Todolí (2002, p.1364-1365, 1377-1388, 1402-1418, 1423-1429), Acadèmia Valenciana de la Llengua (AVL) (2006: 181-183), Institut D’Estudis Catalans (IEC) (2018), to cite but a few of the more recent authors. All of them broadly coincide in their uses and functions, although they also present differences and restraints regarding their use in certain circumstances. In fact, these two clitics present more restraints and limitations than other pronouns, due to the fact that they are invariable and, consequently, it is far more complex to establish relations of concordance with the antecedent. These restraints become apparent, on the one hand, in the difficulty in being coreferentials of a complement of the subject (Solà, 1994), as well as in the blockage that occurs in noun subordinate sentences or in the interposition of two prepositional phrases between a pronoun and its antecedent (Todolí, 2002).

II.1. The clitic en

This clitic can be found accompanying unaccusative verbs of motion and also intransitive verbs (Badia i Margarit, 1994, p.475; AVL, 2006, p.181; Todolí, 2002, p.1427). This usage is fully valid in Valencian-speaking areas – although it has clearly disappeared since the early 18th century (Ribera, 2018) – and its meaning is that of movement from a place. They are forms in which the pronoun has been lexicalised, such as anar-se’n, baixar-se’n, eixir-se’n, pujar-se’n, tornar-se’n, vindre-se’n. This lexicalisation is also apparent in a series of idioms, for example: tenir-ne prou/massa and haver-n’hi prou/massa; saber-ne, de la missa, la meitat; dir-ne per a tots els gustos; no cantar-se’n gall ni gallina; fer-ne de verdes i de madures, etc.

This usage has also occurred in French, where in some cases “L’agglutination est complète dans enlever, entrainer, emportar, emmenar et dans les pronominaux s’enfuir, s’envoler” (Grevisse & Goosse, 2008, p.878), and in others it remains exactly as in Catalan with verbs such as s’en aller, s’en retourner, s’en venir “où il fait double emploi avec un complement introduit par de” (Grevisse & Gosse, 2008, p.880).

One of the most characteristic features of the clitic is its partitive sense and, both in French and in Catalan, it can perform several common syntactic functions, including those of subject, direct object, noun complement, predicative (attribute and complement of an attributive adjective) and prepositional complement. In French, we also find it as a substitute for a direct object infinitive introduced by de.

The representation of a noun phrase in a partitive sense, unlike other pronouns, is defined by the fact that the pronoun does not refer to the whole noun phrase and that the reference can be established with only the head (Example 1), or with the head together with the complements (Example 2) of a noun phrase that may or may not be quantified (AVL, 2006, p.181). The same is true in French (Example 3 and Example 4). The noun phrase can act as the direct object or as the undefined subject of certain unaccusative verbs. These latter verbs are those that have a subject that is not semantically the one that performs the action of the verb and is generally in the post-verbal position commonly occupied by the direct object. In this section we would include verbs of motion such as anar, baixar, caure, eixir, tornar, sortir, venir, etc., and verbs that indicate a change of state, including néixer, florir or fondre’s. Todolí (2002) also added sentences containing impersonal or existential verbs (Example 5) and those with periphrastic or pronominal passive forms.

(Example 1) Ven les meues pells, aconseguiràs, si més no, cinc mil, i en tinc unes altres tantes en la companyia. (London, 2000) [Sell my furs; they’ll fetch at least five thousand, and I’ve got as much more with the company]

(Example 2) El capità es va beure un got de vi, després n’afegí un segon i un tercer (London, 1995) [The captain, however, drank a glass of wine, and topped it off with a second and a third]

(Example 3) Mon père, je me suis promis de vous obéir; mais ne puis-je connaître vos raisons? J’hésitais à lui en donner. (Gide, 1974) [“My father, I promised myself to obey you, but can I know your reasons?” I hesitated to give them to him]

(Example 4) Une petite commotion qu’il en avait reçue la veille, lui en avait fait connaître l’effet. (Itard, 2012) [A trifling shock that he received from it diverted him from his state of reverie]

(Example 5) Però jo sé que en queden quatre pots al rebost, perquè els he vists,… (Saki, 1994) [I know there are four jars of it in the store cupboard, because I looked]

Badia i Margarit (1994) considered that these usages are fully valid in Valencian speech.

The pronoun en establishes a close relationship with the preposition de. As pointed out by Todolí (2002), this relationship is quite obvious with thematised NP (Example 6).

(Example 6) …però ara està tan inquiet per la política que mai no n’estic prou segura, d’ell. (Wilde, 2007) [but he is so agitated about politics at present, that I never feel quite certain about him]

Likewise, Badia i Margarit (1994) recalled this relationship when the noun phrase refers to a place name, whether common or proper (Example 7 and Example 8), and stressed the fact that this usage is alien to Valencian speech. He added a couple of examples:

(Example 7) Si vas a la cuina de dalt, puja’n les patates [If you’re going up to the kitchen, take the potatoes up with you]

(Example 8) Vas a casa teva? ― No, jo ara en vinc [Are you going home? – No, I’ve just come from there]

This same partitive sense also appears in French (Example 9), although more so as in (Example 8):

(Example 9) Vas-tu a ta maison? Non, j’en reviens [Are you going to your house? – No, I’ve just come from it]

Even so, as in Valencian speech, its use is becoming increasingly less common (Sabio, 2005).

Moreover, French also has a partitive article, in addition to the two definite or indefinite articles, which can also be replaced by the clitic (Example 10):

(Example 10) Veux-tu de l’eau? No, j’en veux pas [Would you like some water? – No, I don’t want any]

which is used in the same way in Catalan (Example 11):

(Example 11) Vols aigua? No en vull [Would you like some water? – No, I don’t want any]

In Catalan, this partitive use presents a series of restraints related to the syntactic function or the position of the noun phrase, with the specifier or with the complements (Todolí, 2002). As for the restraints regarding the syntactic function, this pronominalisation is feasible if the verb requires the noun phrase and the position of this phrase is post-verbal; in that case, pronominalisation of both the head and the objects would be possible (Example 12).

(Example 12) Quants cartutxos dius que et queden? Va preguntar. “Tres” —va respondre Bill—. “Però m’ agradaria tindre’n tres-cents” (London, 1996) [“How many cartridges did you say you had left?” he asked. “Three,” came the answer. “An’ I wisht ‘twas three hundred.”]

In this case we observe that the noun phrase must play the role of direct object. As we pointed out earlier, the specifier also establishes restraints in such a way that the pronominalisation cannot be accompanied by any determiner like the definite article, demonstrative or possessive, or by tot [all] or by numerals modified by a definite article. Finally, as for the object, Todolí (2002) offered several specific examples to highlight aspects related to the nominalisation of ordinals or in the case of adjectives, depending on whether they are classificatory or relational.

Ruaix (2003) stressed the limit of replacing the adverbial pronoun en in the substitution of the subject block, when the pronoun is to the left of the verb, in both non-copulative and copulative sentences. However, he also acknowledged the possibility of substitution when the subject is modulated by quantifiers or the subject is an unstressed relative pronoun. Similarly, in interrogative sentences, this limit occurs when the pronoun is to the right of the verb, since in interrogatives the order of the elements in the sentence is altered.

We have already indicated that, as far as Catalan is concerned, the clitic en can act as a complement of the noun, as in French (Grevisse & Gosse, 2008, p.870). Todolí (2002) remarked that most grammars accept the substitution of the pronoun en by some object within the noun phrase acting as the subject of unaccusative verbs or the direct object of transitive verbs, when the substitution is carried out by an indefinite phrase or, conversely, if it indicates a fraction. However, the problem arises when it comes to establishing with which phrases this substitution can be performed. In fact, if, as we have said, the noun depends on a quantifier, this possibility is accepted. However, others (Ruaix i Vinyet, 1981; IEC, 2018; GramaTIC) also accept it not only when it is modified by a quantifier. Todolí, in turn, agreed with Jané (1997) that the pronoun en cannot replace just any complement. This latter author admitted the possibility of a theoretical substitution rather than a real one in the case of “grups nominals formats plenament per un nom determinat per un altre” (Jané, 1997: 13) [nominal groups fully formed by a noun determined by another – Our translation]. Todolí (2002) pointed to syntactic restraints as well as others of a semantic and referential nature. The former refer to the fact that the verb needs the substituted phrase, which must not be linked to the verb indirectly, that is to say, it must be a prepositional phrase. Moreover, the substituted complement must immediately follow the head of the noun phrase and must not be placed in front of the verb; likewise, it cannot be an infinitive sentence.

As far as semantic and referential restraints are concerned, it should be mentioned that substitution is easier when it is not carried out with phrases that refer to animate beings, especially with abstract or not very concrete concepts. Hence, if the phrases refer to individualised elements, it will be more difficult. Finally, if the object of the noun designates a single unit with the head, this substitution is also impossible.

In French we also find restraints on the use of the clitic en. However, there is no unanimity on the matter among grammarians. For some, such as Grevisse and Gosse (2008), it can replace a noun phrase that does not indicate place and which often designates things, animals or abstract ideas and, less frequently, people. Others, in turn, consider that in current French, en can only designate people when it has a partitive sense as an object to the verb, noun, numerical adjective, pronoun or adverb (Pinchon, 1972).

Another function of the clitic en in Catalan is predicative. This function has also been much debated. Thus, on the one hand, there are those who recognise that en substitutes quantified phrases with the verbs ser and estar in the spoken language, but that this substitution is not acceptable in the written language (Badia, 1994) (Example 13).

(Example 13) El preu de la casa és molt alt, però no ho és tant com em pensava [The price of the house is very high, but not as much as I thought – Our translation]

In this example, Badia stresses that the neutral pronoun ho has to be used instead of the clitic en.

On the other hand, there are others such as the IEC (2018) that do consider it acceptable in the written language, although some others recognise that the substitution is not performed in the Balearic and Valencian speech (Solà, 1972). Likewise, we have others, such as Fabra (1968), who do admit it and add the verb fer-se (Solà, 1972), as well as incorporating other verbs such as tornar-se, convertir-se and esdevenir. Finally, Todolí (2002, p.1413) included some quasi-copulative verbs such as sentir-se, trobar-se and some designative verbs like dir, posar, nomenar or elegir. To these, she adds certain fixed constructions amongst which we find an exclamative quantifier (Example 14):

(Example 14) ― Que n’és de solitari, aquest lloc! — digué Orli — (Wilson, 1998a) [“It’s lonely in here,” Orli said.]

It is fair to say that, in French, this usage is clear when the noun appears without an article or “précédé d’un article indefinit ou partitif” (Grevisse & Gosse, 2008) (Example 15).

(Example 15) Est-ce de l’or ou n’en est-ce pas? [Is it made of gold or not?]

In this same section we must include the function underlined by the Gramàtica essencial del catalá (2018) (Example 16) as the object of an attribute that we also find in French (Example 17).

(Example 16) Aquell protector era en realitat el propietari de Véronique, feia d’ella el que volia i n’era el senyor. (Aventin, 1988) [That protector was in actual fact the owner of Véronique, he did whatever he wanted with her and he was her master – Our translation]

(Example 17) Ce protecteur était en réalité le propriétaire de Véronique , il en disposait à sa guise et en était le maître . (Aventin, 1988) [That protector was in actual fact the owner of Véronique, he did whatever he wanted with her and he was her master – Our translation]

Finally, we will underline here another function of the clitic, which is that of prepositional complement. In this function, the clitic en is related to prepositional phrases introduced by the preposition de and, despite the fact that they are usually phrases required by a (pronominal or non-pronominal) predicative verb (Example 18), they can also be related to copulative or quasi-copulative verbs (Example 19).

(Example 18) “Hauries d’haver sabut que se n’adonarien.” [You might know they’d find out]

(Example 19) …no hi va haver cap cançó de resposta que se’n fera ressò. (Robertson, 1994) [No answering song echoed]

The same usage is evident in French: “Il replace un groupe de mots ou une proposition complément d’un verbe” (Delatour et al., 2004) [It replaces a group of words or an object proposition of a verb] (Example 20):

(Example 20) Ma mère en parlait souvent comme d’un refuge possible, si la vie lui devenait vraiment insuportable (Aventin, 1988) [My mother often spoke of it as a possible shelter, if life ever became truly unbearable – Our translation]

II.2. The clitic hi

The clitic hi also accompanies pronominal verbs, as occurs with en, which have become lexicalised forms integrated within the verb, in this case with verbs of perception when they are used intransitively, such as caure, mirar, tornar, veure, etc. (Example 1) (Badia i Margarit, 1994). Here we find verbs like valer-s’hi, veure-s’hi, which, as Todolí (2002) remarked, are prefixed in Valencian speech. She also pointed out that this is because in colloquial Valencian the clitic does not perform any syntactic function and this is the reason why in these speeches it has disappeared from the pronominal paradigm.

(Example 1) Llevat de Lip-lip, tots els altres s’hi veien obligats a moure’s en grup per tal de protegir-se els uns a els altres (London, 1996)1 [With the exception of Lip-lip, they were compelled to hunch together for mutual protection]

The lexicalisation of the clitic hi also occurs when the clitic accompanies the verb haver and gives it an impersonal sense with meanings referring to existence or occurrence (Example 2). This same sense appears in French with the use of y (Example 3):

(Example 2) En el prestatge de sota hi havia les obres completes de Wordsworth (Joyce, 1992) [A complete Wordsworth stood at one end of the lowest shelf]

(Example 3) Et il y a une carte du Grand Nord, avec une région peinte en rouge,… (Pennac, 1984) [And there’s a map of the Far North, with an area marked in red,…]

In terms of usage and the functions it performs, we can say that they are those of prepositional, indirect, circumstantial, attributive and predicative complements, which coincide with the uses of the pronoun y in French (Grevisse & Gosse, 2008).

In Catalan, functioning as a complement, the clitic hi appears related to phrases introduced by the prepositions a, en, amb and per – not per a (GramaTIC), although Todolí (2002) does admit it – that is to say, following Badia i Margarit (1994), by any preposition except de or an adverb equivalent to a prepositional complement (Example 4).

(Example 4) Aleshores es retirà a un còmode taüt de plom i s’hi quedà fins a la nit. (Wilde, 2007) [He then retired to a comfortable lead coffin, and stayed there till evening]

It must be said, as we have already mentioned, that in Valencian speech this use is not present and thus some empty form or another anaphoric expression is used instead of hi. In this function, it can also become pronominalised when the complement includes a subordinate sentence of finite or infinitive form where the preposition may not appear (Example 5).

(Example 5) Ja veuràs com s’hi acostumarà de seguida. (London, 1996) [It’s all right. He’ll learn soon enough]

This usage also appears in French (Example 6):

(Example 6) Oui, une étrangère qui me souriait machinalement, sans y penser. (Aventin, 1988) [Yes, a foreign girl who smiled at me mechanically, without thinking about it – Our translation]

As an indirect object, it replaces the weak dative pronouns, i.e., li or els (los), in certain cases. It can also appear with verbs such as lliurar, presentar or recomanar (7) or even with other verbs, depending on the speech community (IEC, 2018). This usage also appears in French (Grevisse & Gosse, 2008) (Example 8).

(Example 7) En arribar en una ampla habitació a la part posterior del quart pis (la porta de el qual calgué forçar, perquè era tancada amb la clau posada per dins) s’hi presentà un espectacle que estremí tots els presents tant d’horror com de sorpresa. (Poe, 1998) [Upon arriving at a large back chamber in the fourth story (the door of which, being found locked, with the key inside, was forced open,) a spectacle presented itself which struck every one present not less with horror than with astonishment]

(Example 8) je vivais avec cette étiquette depuis si longtemps. A quoi bon y attacher, aujourd’hui, de l’importance? (Aventin, 1988) [I had been living with this label for so long. What do you attach importance to today? – Our translation]

The clitic hi can also replace adverbials. In these cases, pronominalisation is possible when the object performs as an object of the verb and not of the whole sentence or of the subject. When we speak of adverbials, we are talking about comitative, instrumental and locative adverbials, amongst others. The last type in this list can be dependent upon static verbs or verbs of directional motion, which can indicate destination (Example 9) or origin.

(Example 9) L’endemà de matí, després de desdejunar, el pare va cridar Joe a la biblioteca i ell hi anà amb un sentiment quasi d’ alegria perquè la inquietud de l’espera havia acabat. (London, 1995) [On the following morning, after breakfast, Joe was summoned to the library by his father, and he went in almost with a feeling of gladness that the suspense of waiting was over.]

When they indicate the origin, the corresponding clitic is en, especially with copulative verbs, always related to the preposition de. In the other cases it is hi. In any case, as Todolí (2002, p.1407) remarked, while there are verbs that obligatorily require the locative, there are others that do so in a more lax manner. The only ones that are really clear are the situational locatives (Example 10).

(Example 10) Quan de temps feia que hi vivien? (Faville, 1999). [How long had they been there?]

Rigau (2020) indicated that the difference between the presence or absence of the clitic hi in locatives is related to the fact that the locative is not identified with the place where any of the participants in the communicative act may be found. Thus, when the place of destination cannot be linked to the participants, the clitic hi is necessary and, conversely, when the locative has a deictic value, the clitic cannot appear. However, “la pronominalització sí que és possible quan l’antecedent es troba exprés (tematitzat o en una oració diferent), encara que siga per mitjà d’un adverbi amb funció díctica” (GramaTIC) [pronominalisation is possible when the antecedent is clearly expressed (thematised or in a different sentence), even if it is by means of an adverb with a deictic function].

In French, this same function (“complément adverbial d’un verbe”) replaces noun phrases introduced by several different prepositions. Delatour et al. (2004: 86) spoke of “un groupe de mots ou une proposition introduite par la préposition à”. Thus, it appears in adverbials of place (Example 11) or manner (Example 12), etc.

(Example 11) … l’industrie consiste à louer dans les villes un magasin vacant et à y débiter de l’horlogerie à des prix fabuleux de bon marché. (Allais, 1993) […the industry consists in renting a vacant shop in cities and selling watches at fabulously cheap prices – Our translation]

(Example 12) …commença de caresser mon front, tandis que j’enfonçais dans les draps mon visage pour lui cacher mes larmes et pour y étouffer mes sanglots (Gide, 1974) […began to caress my forehead, while I sunk my face into the sheets to hide my tears from her and to smother my sobs]

We have already remarked that the clitic hi can have a predicative function. This function can appear in copulative, quasi-copulative or predicative verbs. In the case of the copulative verbs ser and estar (13), it alternates with the neutral ho in the Balearic dialects (Todolí, 2002; IEC, 2018). Moreover, with the quasi-copulative verbs (such as acabar, anar, eixir, parar, restar, presentar-se, romandre, sortir, tornar-se, etc.), that is, verbs that semantically modify the predication expressed by the attribute by adding or attenuating certain values of the proposition (IEC, 2018), the clitic most used in this function is hi (14). At least that is what Fabra (1968, p.80) indicated and, as we have already recalled when talking about en, he added that with fer-se it was necessary to use en. Finally, the clitic appears in predicative verbs whether the secondary predication refers to the subject or a complement of the verb, and whether it is optional or compulsory (IEC, 2018) (Example 15).

(Example 13) No hi va haver descans, perquè quan acabaven de buidar una draga ja hi era la següent (London, 1995) [There was no rest, for by this time the other dredge required emptying.]

(Example 14) …i allà s’hi va quedar, dreta i sola, durant un temps. (London, 1996) [For some time she stood alone.]

(Example 15) Les he qualificades com a “deduccions legítimes”, però el meu parer no hi resta plenament expressat (Poe, 1998) [I said “legitimate deductions”; but my meaning is not thus fully expressed.]

In French, this function as an attribute also appears in Example 16 and Example 17:

(Example 16) “Je vais pleurer, se dit Afrique, ça y est, je vais pleurer!” (Pennac, 1984) [“I’m going to cry,” Africa said to himself. “I just can’t help it, I’m going to cry”]

(Example 17) Il y restait jusqu’à l’heure du déjeuner (Pennac, 2003) [He stayed there until lunchtime]

Among the restraints on the use of the clitic hi in Catalan, it should be borne in mind that:

• it needs a referent which may, or may not, be thematically displaced. However, a preceding phrase cannot be pronominalised when the rheme is introduced in a sentence by the verb ser and subsequently reappears in a relative clause or one that is introduced by the completive conjunction que (GramaTIC).
• it cannot be substituted when it acts as a framing element (time expressions and certain adverbs) (Ruaix, 2003).
• it cannot be substituted when it is pleonastic.

III. METHODOLOGY

The participants in the study (n =427, 14.3% males and 85.7% females) were first and second-year students of a bachelor’s degree in Translation and Interpreting. Participants had Catalan as their L1 or L2, English as an L3, and were divided into two groups according to whether they were taking French or German as a fourth language (L4). The German group (G) consisted of 208 subjects and the French group (F) had 219. In group G, 92% had studied Catalan (Valencian) at secondary school, while in group F the figure rose to 96%. There were 166 bilingual subjects with Spanish as their dominant language in group G and 171 in group F; in contrast, there were 42 bilingual subjects with Catalan as their dominant language in group G and 48 in group F.

The data were collected over four academic years, from 2016–2017 to 2019–2020. To determine whether there were significant differences in language proficiency between one group and the other, at the beginning of each academic year the subjects took a Catalan language proficiency test using one of the C1 exams of the Junta Qualificadora dels Coneixements de Valencià (Valencian Language Proficiency Assessment Board). After applying the Kruskal-Wallis test to the results obtained, the difference in proficiency between the groups was not seen to be statistically significant (p=0.3986).

Each year, subjects from group G and group F were instructed in the use of adverbial pronouns in Catalan in two 2-hour sessions. Group F subjects were instructed in the use of the pronouns en and y in French, in a 2-hour session. In the Catalan language sessions attended by members of group F, reference was always made to the similarities in usage between French and Catalan in relation to the use of the clitics en and hi.

The participants wrote a series of scripts in class on a wide range of topics, mostly on issues related to female discrimination, resulting in a total of 1,620 texts. Each token was analysed by two independent researchers to determine their classification and the agreement scale was set, following Cohen (1960), at 97%.

Table 1. General characteristics of the texts.

Group G		Group F
Words	33,4225	37,2362
Types	30,153	31,729
Sentences	12,453	13,874
Word/ sentence ratio	26.8389	26.8388
Number of texts	791	829
Words/text	422.53	449.17
Mean texts/subject	3.8	3.78

As can be seen, the differences in the number of words used, tokens, sentences, number of words per sentence or words per text are very similar. The small differences in favour of group F with respect to group G are due to the fact that the first had eight more members than the second. Group F used 10.2% more words than group G. However, this difference is not due to the greater number of subjects in group F (eight more) than in group G, since this only represents a 3.65% increase in words in group F, but rather results from the higher number of words per text (6%) in group F compared to group G.

IV. RESULTS AND DISCUSSION

The results obtained by the two groups for the clitic en are shown in Table 2.

Table 2. Distribution of tokens according to the function of the adverbial pronoun en.

Function	F	G
Subject (S)	1	0
Attribute (At)	1	0
Attributive complement adjective (ACA)	3	3
Direct object (DO)	40	37
Prepositional complement (PC)	16	18
Noun complement (NC)	27	4
Lexicalising (L)	19	25
TOTAL	107	87

To determine whether there were any differences between the tokens in the two groups, we used a non-parametric test, i.e. normality was not assumed in the data, so a statistical analysis carried out with the Kruskal-Wallis test gave us a p value of .60928, which indicates that the result is not significant at p< .05. However, we noted that while the difference in the number of words between group G and group F is 10.2% higher in group F, the use of the adverbial pronoun en is 18.7%, that is, group F uses this pronoun 8.5% more than group G, regardless of the number of words used.

Secondly, it must be said that, with regard to the use of the clitic functions, while in F we find all the functions of the clitic, albeit in just a single case, and G does not have the S function (“Comencen a copiar aquestes idees de compartir el seu moment més important amb un miler de persones a les xarxes socials, que molta falta no en fa, encara així se senten feliços…”) [They begin to copy these ideas of sharing their most important moment with a thousand people on social networks, which is not really necessary, even so they feel happy…] or the At function (“I és que vivim en una societat en la qual les desigualtats entre dones i hòmens es veuen a simple vista, i el fet que les dones hàgem de tapar-nos i els hòmens no, ja n’és una») [And we live in a society in which the inequalities between women and men are quite clearly appartent, and the fact that, as women, we have to cover ourselves up and men don’t is already one], two functions that we also find in French.

The use of the adverbial pronoun en to replace a DO is the most common form in both groups. It is fair to say that this is a fully valid function in Valencian. This function represents 37.28% of all the uses of the pronoun in group G and 42.52% in group F.

The three transitive verbs in the two groups (haver-hi, fer, tenir) account for 56.76% of the cases in group G and a slightly lower percentage in group F (52.87%). This indicates that there is a higher (albeit only slightly) concentration in certain verbs in group G than in group F.

In the case of the use of the clitic acting as a PC, we observe a greater use in G than in F, although in the case of group G, this is concentrated in the use of the verb adonar-se’n, with 76.19%. This verb is also the most used in group F, but its use is reduced to 43.75%.

As a noun complement it is clearly used more in group F than in group G: 85%. This use is also quite apparent in French, as we have seen in the previous section. Finally, as far as lexicalised forms are concerned, in both groups the form used is that of the motion verb anar-se’n. However, in the case of group F, we also find three tokens with eixir-se’n.

In short, the use of the most common forms in Valencian speech such as verbs (haver-hi, fer, tenir), lexicalised forms or the use of certain verbs where the pronoun acts as a PC is observed to be higher in group G than in group F. Subjects in group F were instructed in the use of the pronoun en in forms which also coincide with those used in the French language.

There is a final aspect to be considered which, in part, also reveals this use of en – that of the ungrammatical uses of the pronoun. In this sense, two types of uses should be distinguished: pleonasm, that is, the appearance of the pronoun in cases in which the element it is supposed to replace is in fact present, and the other instances in which it appears by analogy in cases that either clash with grammatical logic or uses in which the rules do not recognise them. In these cases there is also a difference between the two groups, as can be seen in Table 3. However, with the use of the Kruskal-Wallis test in com combination with the data on the use of the clitic en gives a p-value (0.47993) that is not significant.

Table 3. Distribution of errors in the use of the clitic en between groups F and G

	F	G
Pleonasm	54	47
Other types of error	8	1
TOTAL	62	48

In the case of pleonasms, both groups coincide in the use of the verb haver-hi, with the pronoun, when the element supposedly substituted is also present (“…n’hi han [sic] moltes coses que els homes tenen permés fer…”) [there are a lot of things that men are allowed to do…]. Even so, there are certain differences: in the case of F this type of pleonasm represents 33.3%, while in G the figure remains at 25%. This fact is rather the result of linguistic interference with Spanish. It should not be forgotten that in the whole linguistic area, these pronouns are in decline. Therefore, the speakers, as a reaction to the awareness of loss, add them when substitution is not necessary or not possible (“…les dones de la tribu Mursi es perforen els llavis i els estiren ficant-ne plats de diferents mides.”) […the women of the Mursi tribe pierce their lips and stretch them by inserting plates of different sizes].

The appearance of many more errors in group F also responds, in part, to the same awareness of the loss of the clitic, even introducing it without any antecedent to which it refers, or even confusing it with the clitic hi (“A les dones sempre ens ha estat prohibit parlar de sexe, a voltes inclús pensar-ne”) [It has always been forbidden for women to talk about sex, sometimes even to think about it].

As for the clitic hi, the tokens are clearly higher in group F (27%) than in group G (Table 4), even though with the clitic en the p-value (0.87278) is not significant with the Kruskal-Wallis test.

Table 4. Distribution of tokens according to the function of the adverbial pronoun hi

Function	F	G
Attribute (At)	3	2
Indirect object (IO)	0	2
Predicate adjunct (PA)	50	41
Prepositional complement (PC)	40	13
Predicative (P)	0	0
Lexicalising (L)	814	604
TOTAL	907	662

In both groups, the function with the highest number of tokens is the lexicalising function, with 89.74% in group F and 91.23% in group G. We have to bear in mind that this function is made up of structures where the clitic accompanies the verb haver (“En general, hi ha moltes cançons com aquestes, plenes de masclisme, que tothom escolta i de les quals no som conscients”) [In general, there are many songs like these, full of male chauvinism, which everyone listens to and which we are not aware of]. However, in group F we have three cases where the clitic hi accompanies pronominal verbs of perception (“M’hi veig donant classes d’anglés a un institut…”) [I can see myself giving classes in a secondary school], something which does not occur in group G.

The next group with the most tokens is PA. Those in group F (“En segon lloc, hi podem veure com les dones han de patir per poder ser considerades belles, sota el principi “para presumir hay que sufrir” que trobem en castellà”) [Secondly, we can see how women have to suffer in order to be considered beautiful, under the principle of “to look good you have to suffer”, as they say in Spanish] are 20% more numerous than those in group G. In this section, the presence of locative circumstantial complements is totally predominant, including situational locatives (“Les meues amigues també hi viuen i ens agrada molt sortir pel carrer, anar al bar, caminar per les muntanyes, banyar-nos en la piscina municipal i milers de coses més”) [My girl friends also live there and we like to go out in the street, go to the bar, walk in the mountains, swim in the municipal swimming pool and thousands of other things], which reach 86% in group G and 83% in group F.

Uses of the pronoun hi acting as a PC are also far more numerous in the case of group F than in group G – 67.5% more. In all cases in both F and G with non-pronominal verbs (“Mentre que l’home en tot moment es comporta de manera superior a la dona i pot expressar i dir lliurement tot allò que hi pensa (“te quiero amarrar”), la dona ha de dissimular…”) [While men behave as though they were superior to women at all times and can freely express and say whatever they think about them (“I want to snag you”), women have to conceal things…].

Also in the primary predication function, At, the use by group F is greater than group G, even though the number of tokens is very small (“Resulta impressionant la quantitat de connotacions que es poden extraure d’una cançó que no s’estén més de tres minuts i que ha estat creada amb l’objectiu de gaudir-la, però que enceta el debat sobre diferents qüestions que hi són, amb força, d’actualitat, tot i que pareixen no incomodar-nos”) [It is impressive how many connotations can be extracted from a song which does not last more than three minutes and which has been created with the aim of entertaining, but which sparks a debate on different issues that are very topical, even though they do not seem to bother us].

The only function where group G has more tokens than group F is that of IO. (“M’he rigut bastant amb la carpeta compartida i la hi he mostrat a la meva família perquè es rigués amb mi”) [I had a good laugh with the shared folder and I showed it to my family so that we could have a laugh together], but both cases respond to the binary pronominal combination of indirect object and direct object, where the pronoun li (IO) becomes hi when it comes into contact with a third person pronoun acting as the direct object. This is a fact that reveals a speaker from the northern part of Valencia or of eastern Catalan.

Finally, it should be noted that the non-normative uses (Table 5), whether due to pleonasm or for any other reason, reflect the users’ need to employ a pronoun that does not currently belong to the Valencian pronominal system. The Kruskal-Wallis test for all the uses of the adverbial pronoun yields a p-value (0.59951) that is not statistically significant.

Table 5. Distribution of the errors in the use of the clitic hi between groups F and G

	F	G
Pleonasm	36	15
Other types of error	17	12
TOTAL	55	27

The frequency of errors is much higher in group F than in group G – almost 50% higher. This evidences a greater awareness of the pronoun hi in group F than in group G. Pleonasms are the most common and occur in all functions, even those with fewer tokens, such as At (“Jo no hi sóc partidària del matrimoni però reconec que existeixen proposicions que són vertaderament romàntiques i emocionants”) [I am not in favour of marriage but I recognise that there are propositions which are truly romantic and exciting]. However, the most frequent are seen in locative complements, since these are also the most frequent type of complement (“La belladona va ser gastada per moltes dones per posar-la-hi en els ulls aconseguir que les pupil·les es dilataren”) [Many women used to put Belladonna in their eyes to dilate their pupils].

Errors not attributed to pleonasm occur when the pronoun hi represents a phrase that cannot be replaced by this pronoun, that is, it is an attempt to replace phrases that cannot be pronominalised or it is simply confused with another weak pronoun. In most cases, this pronoun is the clitic en (“Cal tindre present, però, que les convencions que ens fan belles a un lloc, ens farien indesitjables a un altre, per no parlar-hi, de com canvien amb el pas del temps”) [We must bear in mind, however, that the conventions that make us beautiful in one place would make us undesirable in another, let alone how they change with the passage of time].

Finally, we have to say that the differences in the use of the two pronouns between group G and group F are not statistically significant with the Kruskal-Wallis test, since the p-value is 0.41827.

To sum up our findings partially support previous studies that have indicated a clear influence of L2 or L3 on L1 (Kecskés & Papp, 2000; van Hell & Dijkstra, 2002; Kecskés, 2008). Similarly, our findings support Agheshteh’s (2015) study that observed a positive effect of L2 on L1, and that conducted by (Liu & Ni, 2016), carried out in a context that was clearly L1-dominated (Liu & Ni, 2016), Thus, our work partially support previous research providing further evidence on the use of the L4 clitics in the L1/L2. More specifically, it is observed a higher percentage of usage of the clitic pronouns by those learners who had received instruction on the French adverbial pronouns. In addition, although this difference is not statistically significant, this group also uses a greater variety of adverbial pronouns than the other group, especially as the subject, attribute, complement of the noun, or as prepositional complement. Besides, this group also shows a lower occurrence of empty forms or other anaphoric expressions, such as substitutes of adverbial pronouns.

Different reasons can explain why in our study the impact of teaching adverbial pronouns in L4 (French) is not statistically significant in the use of adverbial pronouns in L1/l2 (Catalan). First, the effect of instruction on adverbial pronouns. We acknowledge that group F received instruction on adverbial pronouns both in French and Catalan, while group G only received instruction on adverbial pronouns in Catalan. Thus, further research should carry out to analyse the use of adverbial pronouns out of a teaching environment. Situations such as writing emails or professional encounters are likely to provide further information on this issue.

Secondly, we acknowledge that the teacher did not follow a multilingual approach to language teaching. It is possible that teaching from a multilingual perspective encourages making references to different languages, or even translanguaging. As a result, this could have an impact on students’ use of adverbial pronouns.

Thirdly, we have only analysed learners’ production in L1/L2 (Catalan), so the study has analysed the influence of L4 on L1/2. Future studies could be conducted to analyse learners’ use of adverbial pronouns in different languages such as L4 (Italian) or L1/L2 (Catalan). This, in turn, may help to explore bidirectional transfer across languages. Moreover, further studies could benefit from discriminating between subjects according to their dominant L1 language, since in line with previous studies (Tsang, 2016; Luque-Agulló, 2020), transfer may happen with greater knowledge of L3 or L2.

V. CONCLUSIONS

The aim of the present study was to determine whether learning the use of certain microparametric aspects – adverbial pronouns – in L4 had an effect on their use in L1/L2, Catalan. For this purpose, an analysis was performed on a sample of 1,620 scripts, collected over four years and produced by 427 participants and divided into two groups. These groups were determined according to whether the L4 instruction was in French or German, given that adverbial pronouns are part of both French and Catalan, while no equivalent form with a similar function exists in German. Both groups had the Catalan language as L1 or L2 and similar language knowledge as a group, as confirmed with the corresponding test. The following results were found:

• There is no statistically significant difference in the usage of group F with respect to group G. However, there is a higher percentage of use in group F, which had received instruction on the use of clitics, than in group G, which had not received instruction on the use of clitics in French.
• A lower occurrence of the most common forms in the Valencian speech in F than in G, and therefore a higher weight of the normative forms in group F.
• There is, however, a greater overall use of empty forms or other anaphoric expressions in the case of group G, which reveals a lower awareness of the use of clitics.

To conclude, in contrast to what was expected, that is, an effect of L4 on L1/L2, no statistically significant difference was detected between the two groups. Therefore, the typological similarity with L4 did not have an effect on L1/L2 language development. Care should be taken to generalize the results from the present study since some limitations are worth pointing out. First, we have not considered all language repertoire of participants. Secondly, we have focused on adverbial pronouns and different target forms have not been analysed. Thirdly, we have considered on type of instruction, but different types of activities or different methodological approaches may result in different learning outcomes.

In spite of these limitations, it is worth pointing out that our result suggest some pedagogical implications. Among them, the importance of training teacher and learner on multilingual education, its advantages and the benefits of learning different languages at school and beyond the school. So, if instead of teaching adverbial pronouns in a particular language, the focus were on the use of adverbial pronouns in different languages, positive transfer may be encouraged, or, at least, language awareness across languages may be activated.

Notes

Acknowledgements

This work was supported by Universitat Jaume I [UJI-B2017-58] and by the Spanish Ministry of Science and Innovation [PID2019-103953GB-I00].

VI. REFERENCES

Agheshteh, H. (2015). On the Effects of L2 on Iranian Bilinguals’ L1 Writing Ability. Advances in Language and Literary Studies, 6(4), 48-52. https://doi.org/10.7575/aiac.alls.v.6n.4p.48

Alcón-Soler, E. (2010) Dealing with the acquisition of pragmatics in English as a third Language. Quaderns de filologia. Estudis de comunicació, 73, 29-51.

Allais, A. (1993 [1891]) Les petits cochons. Éditions fata morgana

Aronin, L. (2018). Lecture 1: What is Multilingualism? In D. Singleton & L. Aronin (Eds.), Twelve Lectures on Multilingualism (pp. 3-34). Multilingual Matters. https://doi.org/10.21832/9781788922074-003

Aventin, C. (1988). Le coeur en poche. Hachette Livre.

Aventin, C. (1992). En la foscor de la nit. Bromera. Translated by Francesc Machirant.

AVL (2006). Gramàtica Normativa Valenciana. Acadèmia Valenciana de la Llengua.

Badia i Margarit, A. M. (1994). Gramàtica de la llengua catalana: Descriptiva, normativa, diatòpica, diastràtica (1. ed). Enciclopèdia Catalana.

Baldaquí-Escandell, J.-M. (2020). La “Llei d’ús i ensenyament del valencià» i la configuració del sistema educatiu plurilingüe valencià. Caplletra. Revista Internacional de Filologia, 68, 149-176. https://doi.org/10.7203/caplletra.68.16475.

Beatty-Martínez, A., Navarro-Torres, C., & Dussias, P. (2020). Codeswitching: A Bilingual Toolkit for Opportunistic Speech Planning. Frontiers in Psychology, 11, 1699. https://doi.org/10.3389/fpsyg.2020.01699

Berkes, É., & Flynn, S. (2012). Further evidence in support of the Cumulative-Enhancement Model: CP structure development. In J. Cabrelli Amaro, S. Flynn, & J. Rothman (Eds.), Studies in Bilingualism (pp. 143-164). John Benjamins Publishing Company. https://doi.org/10.1075/sibil.46.11ber

Brown, A., & Gullberg, M. (2008). Bidirectional Crosslinguistic Influence in L1-L2 Encoding of Manner in Speech and Gesture: A Study of Japanese Speakers of English. Studies in Second Language Acquisition, 30(2), 225-251. https://doi.org/10.1017/S0272263108080327

Cenoz, J. (2007). The Acquisition of Pragmatic Competence and Multilingualism in Foreign Language Contexts. In E. Alcón-Soler & M. P. S. Jordà (Eds.), Intercultural Language Use and Language Learning (pp. 123-140). Springer Netherlands. https://doi.org/10.1007/978-1-4020-5639-0_7

Cenoz, J. (2009). Towards multilingual education: Basque educational research from an international perspective. Multilingual Matters.

Cenoz, J., & Görter, D. (2012). Una perspectiva multilingües en la enseñanza de lenguas. En E.Alcón & F. Michavila (Eds), La Universidad Multilingüe (pp. 131-149). Tecnos.

Cenoz, J., & Gorter, D. (2014). Focus on Multilingualism as an Approach in Educational Contexts. In A. Blackledge & A. Creese (Eds.), Heteroglossia as Practice and Pedagogy (pp. 239-254). Springer Netherlands. https://doi.org/10.1007/978-94-007-7856-6_13

Cohen, J. A. (1960). A coeficient of agreement for nominal scales. Educational and Psychological Measurement, 20, 37-46. https://doi.org/10.1177/001316446002000104

Consell d’Europa (1992) European Charter for Regional or Minority Languages. [https://www.coe.int/t/dg4/education/minlang/textcharter/Charter/Charter_es.pdf]

Cuza, A., Jiao, J., & Lopez Otero, J. C. (2018). Does Typological Proximity Really Matter? Evidence from Mandarin and Brazilian Portuguese-Speaking Learners of Spanish. Languages, 3, 13. https://doi.org/10.3390/languages3020013

Delatour, Y., Jennepin, D., Léon-Dufour, M., & Teyssier, B. (204d. C.). Nouvelle grammaire du Français: Cours de civilisation française de la Sorbonne. Hachette Livre.

Deuchar, M. (2020). Code-Switching in Linguistics: A Position Paper. Languages, 5(2), 22. https://doi:10.3390/languages5020022

Fabra, P. (1968). Gramàtica catalana. Editorial Teide. –

Faville, B. (1999). El supervivent. Bromera. Translated by Víctor Oroval Martí.

Flynn, S., Foley, C., & Vinnitskaya, I. (2004). The Cumulative-Enhancement Model for Language Acquisition: Comparing Adults’ and Children’s Patterns of Development in First, Second and Third Language Acquisition of Relative Clauses. International Journal of Multilingualism, 1(1), 3-16. https://doi.org/10.1080/14790710408668175

Gardner-Chloros, P., & Weston, D. (2015). Code-switching and multilingualism in literature. Language and Literature: International Journal of Stylistics, 24(3), 182-193. https://doi.org/10.1177/0963947015585065

Gide, A. (1974 [1919]). La simphonie pastorale. French & European Pubris.

Görter, D., & Cenoz, J. (2015). Translanguaging and linguistic landscapes. Linguistic Landscape, 1, 54-74. https://doi.org/10.1075/ll.1.1-2.04gor

GramaTIC [http://www.salt.gva.es/auto/ajuda-salt/puntuacio/index.html]

Grevisse, M., & Goosse, A. (2008). Le bon usage. Gramaire française (14.a ed.). De Boeck & Larcier. https://books.google.es/books?id=lU3crQEACAAJ

Ibarraran, A., Lasagabaster, D., & Sierra, J. (2008). Multilingualism and Language Attitudes: Local versus Immigrant Students’ Perceptions. Language Awareness - LANG AWARE, 17, 326-341. https://doi.org/10.1080/09658410802147311

IEC. (2018). Gramàtica essencial de la llengua catalana (1.a ed.). Institut d’Estudis Catalans. https://doi.org/10.2436/10.2500.04.1

Itard, J. (2012 [1801]). De l’éducation d’un homme sauvage ou des premiers développemens physiques et moraux du jeune sauvage de l’Aveyron, Tredition Classics.

Jané, A. (1997). “El pronom “en” fent de subjecte”. Llengua Nacional. 7(20), 12-13. https://llenguanacional.cat/pdf/20.pdf.

Joyce, J. (1992). Els Morts. 3i4. Translated by Joan Talens.

Kayne, R. S. (2005). Some Notes on Comparative Syntax, with Special Reference to English and French. Oxford University Press. https://doi.org/10.1093/oxfordhb/9780195136517.013.0001

Kecskés, I. (2008). The effect of the second language on the first language (The dual language approach). Journal Babylonia, 2(3), 31-34.

Kecskés, I., & Cuenca, I. M. (2005). Lexical choice as a reflection of conceptual fluency. International Journal of Bilingualism, 9(1), 49-67. https://doi.org/10.1177/13670069050090010401

Kecskés, I., & Papp, T. (2000). Foreign language and mother tongue. Erlbaum.

Lasagabaster, D. (2005). Attitudes Towards Basque, Spanish and English: An Analysis of the Most Influential Variables. Journal of Multilingual and Multicultural Development, 26(4), 296-316. https://doi.org/10.1080/01434630508669084

Lasagabaster, D., & Huguet, A. (Eds.). (2007). Multilingualism in European bilingual contexts: Language use and attitudes. Multilingual Matters.

Lawrence, D.H. (1992). La verge i el gitano. Amós Belinchon. Translated by Ines Costa i Granell.

Liu, P., & Ni, C. (2016). Effects of L2 on the L1 at Semantic Level: An Empirical Study. Journal of Language Teaching and Research, 7(2), 425-431. https://doi.org/10.17507/jltr.0702.26

London, J. (1995). El creuer del Dazzler, Bromera. Translated by Remei Bataller i Gerard Bataller.

London, J. (1996). Claublanc. Bromera. Translated by Josep Franco.

London, J. (2000). Els aventurers de l’àrtic. 3i4. Translated by Blai Beneyto.

Luque-Agulló, G. (2020). Unintentional Reverse Transfer from L2 (English) to L1 (Spanish) in Tertiary Levels. International Journal of English Studies. 20(3), 57-76.  https://doi.org/10.6018/ijes.406901

Madariaga, J.-M., Huguet, Á., & Janés, J. (2016). Language attitudes in Catalan multilingual classrooms: Educational implications. Language and Intercultural Communication, 16(2), 216-234. https://doi.org/10.1080/14708477.2015.1118111

Musk, N. (2010). Code-Switching and Code-Mixing in Welsh Bilinguals’ Talk: Confirming or Refuting the Maintenance of Language Boundaries?. Language, Culture and Curriculum, 23(3), 179-197.

Ortin, R., & Fernandez-Florez, C. (2019). Transfer of variable grammars in third language acquisition. International Journal of Multilingualism, 16(4), 442-458. https://doi.org/10.1080/14790718.2018.1550088

Pennac, D (2003). Cabot-Caboche. Nathan.

Pennac, D. (1984). L’oeil de loup. Nathan.

Pinchon, J. (1972). Histoire d’une norme, emploi des pronoms “ lui “, “ eux “, “ elle(s)”, “ en “, “ y “. Langue française, 16(1), 74-87. https://doi.org/10.3406/lfr.1972.5705

Poe, E: A. (1998). Els misteris de Paris. Bromera. Translated by Ramon Cohen.

Portolés, L. (2015). Multilingualism and Very Young Learners. An Analysis of Pragmatic Awareness and Language Attitudes. De Gruyter Mounton https://doi.org/10.1515/9781501500022

Portolés, L. (2020). El multilingüisme a l’escola: Tendències educatives i nous reptes. Caplletra. Revista Internacional de Filologia, 68, 125-147. https://doi.org/10.7203/caplletra.68.16474

Ribera, J. E. (2018). Açò és veritat, y·n fas memòria per a l’esdevenidor. Una introducció a la sintaxi del pronom adverbial en en valencià modern (1601-1832). eHumanista IVITRA, 14, 621-645.

Rigau, G. (2020). Presència i absència del clític locatiu “hi”. Caplletra. Revista Internacional de Filologia, 68, 99-116.

Robertson, C. (1994). Cartes d’una jove enamorada. Bromera, Translated by Víctor Oroval Martí.

Rothman, J. (2010). On the typological economy of syntactic transfer: Word order and relative clause high/low attachment preference in L3 Brazilian Portuguese. IRAL - International Review of Applied Linguistics in Language Teaching, 48(2-3), 245-273. https://doi.org/10.1515/iral.2010.011

Rothman, J. (2014). Linguistic and cognitive motivations for the Typological Primacy Model (TPM) of third language (L3) transfer: Timing of acquisition and proficiency considered. Bilingualism: Language and Cognition, 18(2), 179-190. https://doi.org/10.1017/S136672891300059X

Ruaix i Vinyet, J. (1981). El català en fitxes. 2: Morfologia i sintaxi. L’Arc de Berà.

Ruaix, i Vinyet, J. (2003). Límits en l’ús dels pronoms febles «en», «hi», Llengua Nacional, 44, 15-34. https://llenguanacional.cat/pdf/LN44_08.pdf

Sabio, F. (2005). Les pronoms clitiques et l’expression du lieu: l’usage de y et en dans le français contemporain. Le français parlé au XXIe siècle, 87-105. hal- 00576847

Safont-Jordà, M. P. (2005). Third Language Learners: Pragmatic Production and Awareness. Multilingual Matters. https://doi.org/10.21832/9781853598043

Safont-Jordà, M. P. (2012). Pragmatic Competence in Multilingual Contexts. In C. A. Chapelle (Ed.), The Encyclopedia of Applied Linguistics. Blackwell Publishing Ltd. https://doi.org/10.1002/9781405198431.wbeal0927

Safont-Jordà, M. P. (2013). Early stages of trilingual pragmatic development. A longitudinal study of requests in Catalan, Spanish and English. Journal of Pragmatics, 59, 68-80.

Safont-Jordà, P., & Alcón-Soler, E. (2012). Teachability of request act peripheral modification devices in third language learning contexts (pp. 275-314). https://doi.org/10.1075/pbns.217.09saf

Saki (1994). Tobermory, Bromera. Translated by Doménec Ardit.

Sebba, M., Mahootian, S., & Jonsson, C. (Eds.). (2012). Language mixing and code-switching in writing: Approaches to mixed-language written discourse. Routledge.

Slabakova, R., & García Mayo, M. D. P. (2015). The L3 syntax-discourse interface. Bilingualism: Language and Cognition, 18(2), 208-226. https://doi.org/10.1017/S1366728913000369

Solà, J. (1972). Estudis de sintaxi catalana. Edicions 62.

Solà, J. (1994). Sintaxi normativa: Estat de la qüestió. Editorial Empúries.

Stavans, A., & Webman Shafran, R. (2018). The pragmatics of requests and refusals in multilingual settings. International Journal of Multilingualism, 15(2), 149-168. https://doi.org/10.1080/14790718.2017.1338708

Todolí, J. (2002) “Els pronoms”. A Solà, Joan; Lloret, Maria-Rosa; Mascaró, Joan; Pérez Saldanya, Manuel (dirs.) In J. Solà, M.R. Lloret, J. Mascaró and M. Pérez-Saldanya (Dirs.) Gramàtica del català contemporani (pp.1337-1433). Empúries.

Tsang, W. L. (2016). Acquisition of English number agreement: L1 Cantonese–L2 English–L3 French speakers versus L1 Cantonese–L2 English speakers. International Journal of Bilingualism, 20(5), 611-635. https://doi.org/10.1177/1367006915576398

Wilde, O. (2007). El crim de Lord Arthur Savile i el fantasma dels Canterville. Bromera.

Wilson, E. (1998a). Els vampirs d’Ottawa. Bromera. Translated by Àlan Greus Bosch.

Wilson, E. (1998b). Pànic a Vancouver. Bromera. Translated by Àlan Greus Bosch.

Received: 10 March 2021

Accepted: 5 July 2021