BIODATA

Salvatore Mustois a researcher in Spanish Language and Linguistics at the University of Naples "Federico II". He works on the temporal relationships and their acquisition in Spanish L2 and L3. His research interests are in linguistic acquisition and the syntax and historical grammar of Spanish.

INTRODUCTION

Multilingualism is widely increasing throughout the world. Today, to learn more than one language is not strange, conversely, to learn three or four languages seems to be normal, and in some societies, very common.

Many people learn three or more languages becoming, to all effects, multilingual speakers. In Europe, for example, multilingualism is common among people living in bilingual countries such as Belgium, where two languages – that for historical and/or political reasons imposed themselves over the years – are spoken. The same phenomenon is very common in regions such as Basque Countries, Catalonia, or Galicia, where since the nineteen-eighties, or even before, such is the case of Catalan, they promoted the preservation of the local languages. Along with these examples it is necessary to consider different kinds of dynamics that take part to the spread of multilingualism, first of all immigration, with reference not just to the immigrants who are already bilingual but also to the ones “forced” to learn one or more languages once in the host country.

Nowadays, English is to be considered as the language of communication all over the world, being the third language in bilingual countries and the L2 in the monolingual ones; moreover, the globalization process has enormously increased international contacts, inspiring the need to learn a new foreign language. Simply put, those who can call themselves monolingual become fewer and fewer every day. Different scholars point out that potentially all human beings are multilingual and multilingualism itself is only an intermediate step in the developing of language competence.

“Human beings are remarkable language learners who can easily learn and master several languages throughout their lives. (...) Multilingualism is, no doubt, a common achievement for many people around the world”. (De Angelis 2007)

It is obvious that we all possess the capacities to learn different languages, that is why, even those who speak four, five, or more languages do not have to be considered an exception. We know, indeed, that human beings can learn and manage a certain number of languages but the human brain has a much higher potential that must nevertheless be exploited within the first decade of life (cf. Paradis 1994).

Based on the assumption that all human beings are potentially multilingual, in the last few years studies about language acquisition have grown considerably. Different studies have already proved that the acquisition of a non-native language (L2) is completely different from the acquisition of the mother tongue (L1): this kind of investigation provide us with important conclusions that the study of the L1 acquisition alone could not have provided. Despite that, the study of L2 acquisition is not sufficient to completely understand the human capacity for learning languages; for this purpose, it has been necessary to run investigations about the acquisition of a language subsequent to L2, a language that we are going to call L3.

As well as learning an L2 is different from learning an L1, the acquisitional process of an L3 differs from that of an L2, because the latter can greatly influence the former.

The question here is essentially: which role do the already acquired linguistics skills of the speaker have during the acquisition of an L3?

In order to answer this question, it is mandatory to recognize and assess the skills of the speaker in every different language that he or she already knows, considering that performance and ability are not given parameters and may differ between individuals even in the mother tongue. Building corpora related to the analysis of the acquisitional processes can give a huge contribution to both linguistics theory and language teaching in classroom, in particular for learners focused on the acquisition of a foreign language after an L2. So, through the direct comparison of two or more languages, these corpora can become an important means to develop more effective didactics for the long-lasting acquisition of a different linguistics system.

THE COLLECTION OF THE CORPUS

Despite the significant growth of corpus language studies, building a corpus focused on the analysis of a specific phenomenon is still a challenging task because of the many factors that one needs to consider in order to make it as reliable as possible. It is even harder if the field of study is relatively new or lies behind a much more consolidated one, such as is the case of the study of L3.

The high number of conferences, articles, and volumes published in the last few years has made the acquisition of an L3 one of the most studied areas, and currently it represents one of the fastest growing sectors in the field of language acquisition (cf.Aronin and Singleton 2008; Cenoz 2000, 2001, 2003, Dewaele 2001; Herdina and Jessner 2000, Lindqvist 2015; Lindqvist and Bardel 2010; Murphy 2003; Ringbon 2007).

In the first place, the study of the acquisition of an L3 has been considered just as a branch of second language acquisition (SLA), since L2 was intended to be any language different from mother tongue acquired after L1. In other words, no difference was made between monolingual and bilingual / multilingual individuals engaging the study of a foreign language.

But if before the knowledge of an L2 was considered not relevant for the acquisition of an L3, now all the languages previously learned, including L1, are seen as possibly having a dramatic role in the acquisition of a new language system. As defined by Hammarberg (2001), a L3 is the language one is learning at the moment an investigation takes place, and it can effectively match the third language being studied by the individual or the fourth, fifth etc., if the speaker already knows more than three languages.

These new frontiers not only furnish new horizons to better understand the possibilities in the acquisition of a foreign language, but also start a revolution which complicates the collection of data for new corpora: the higher the number of languages known by a speaker, the more complex it will be to build a framework including all the different systems.

In case of L2 acquisition, according to Cenoz (2001), the learner can count on two different language systems which can mutually influence themselves: it is therefore a bidirectional transfer; instead in the case of the so called ‘cross linguistic influence’ (CLI), the number of transfers increases and they can take place even between languages not related to the same linguistic family.

As an example, during the acquisition of Spanish as L3 by an Italian speaker, it is not certain that the L1, namely Italian, will always support the acquisition of the L3, since a series of factors could intervene: the typological and structural similarities, the learner’s subjective interpretation of linguistic phenomena, the degree of competence achieved in other languages; but also strategies such as avoiding structures too similar to those of the L1 or having recourse to those of the most used language.

All these factors are crucial in cross-linguistic influence and must be considered when building a reliable corpus which aims at accounting for the influence of an L2 over an L3 even though they are genetically distant from one another.

To sum up, to collect data for a corpus finalized for the analysis of L3 acquisition is by all means complicated, because of the high number of variables and strategies involved in this kind of acquisitional process, which depend a lot on the conditions of acquisition, a factor hard to control.

THE CROSS-LINGUISTIC INFLUENCE

Generally, when studying something new, you try to link all the new elements with any previous knowledge you already possess, as Ausubel 1968 points out: “The most important single factor influencing learning is what the learner already knows” and as it is confirmed by the research of Neuner (1992: 158), who affirms that “It is a general and basic law of any kind of learning that we associate new elements, items and structures with elements, items and structures already stored in our memory”.

Someone learning a new language is forced to deal with different kinds of categories, which are precipitously put in relation with the ones of the other known languages. So, the person who learns a new language tries to track down elements from his/her own linguistics knowledge and transfer them into the language that he/she is learning. It is clear that the process of transfer is an important part of the acquisition of a language and of language acquisition research.

Historically, cross-linguistic influence has focused on the acquisition of an L2 and the possible interaction between L1 and L2. The fellow phenomenon increases if we look at the acquisition of an L3, which is why in L2 acquisition CLI is basically different from the one concerned with L3 acquisition.

A multilingual speaker learning a further language has a different approach from an L2 learner, so cross-linguistic influence is different when three or more languages interact. De Angelis (2005) unfolds the matter in these terms:

“Viewing transfer as a one-to-one type of association is a logical and viable option for speakers who are familiar with two languages, but the same type of association ceases to be the only possibility when more than two languages are in the mind. In this case, there are at least two types of CLI that are technically possible. The first is the influence between the source language and the target language, which is the one-to-one type of association already mentioned; the second is the simultaneous influence of more than one language upon a target language, i.e. many-to-one type of association. This second kind of CLI occurs when two or more languages interact with one another and concur in influencing the target language, or when one language influences another, and the already influenced language in turn influences another language in the process of being acquired”.

In these cases, the speaker has already learned different language systems, that could potentially lead to a bigger number of transfers, that can be triggered by any of the languages known. Technically, in the acquisition of L3, the speaker could mix up all the languages he knows.

“Second language learners have two systems that can potentially influence each other (L1 ↔ L2) [...]. Two other bi-directional relationships can take place in third language acquisition: the L3 can influence the L1 and be influenced by the L1 (L1 ↔ L3) and cross-linguistic influence can also take place between the L2 and the L3 (L2 ↔ L3)”. (Cenoz et al. 2001)

The interactions that can take place when the speakers are exposed to the influence of more than a language are probably the most interesting area in the field of L3 acquisition.

We claim that research on cross-linguistic influence should point out the difference between L2 and L3 acquisition and, most of all, seek out the reasons and the circumstances leading to transfer, as well as those bringing the learner to choose one language over another. These reasons will also be our research purposes.

In the last few years, different hypothesis have been formulated concerning the process behind transfer and its activation. It is nevertheless hard to link transfer to universal processes and to explain them according to general and objective rules. Otherwise, the multilingual speaker can know more than one L2, so the processes involved can be even more complex. Quoting Clyne (1997) “the additional language complicates the operations of the processes.” In general, we can assess that in L3 acquisition it is possible to use one’s own linguistic background, whether it is the L1, the L2, or further L2s.

Without considering the number of languages involved, it is reasonable to wonder why in some cases the L1 is the source of transfer while in some other cases transfers come from one of the L2s; what influences the choice of a language over the other known?

According to the results of the research carried until now, transfers from the languages previously acquired are determined by a series of factors. The multilingual speaker can decide to rely or not on his/her linguistic skills and to open or limit the types and number of transfer to the L3. Some of these factors depend on his/her individual knowledge or on the perception of the L3 in analysis, others may vary depending on the characteristics of the L3 itself. Following this point, it is useful to introduce some distinctions proposed by Murphy (2003), who made a difference between learner-based-variables and language-based-variables. Among all the most determining factors are: proficiency, regency, foreignness, context, L2 status, typology, relatedness and psychotypology. All of them are important and hard to separate.

In short, the research on CLI in the acquisition of L3 is influenced not only by the knowledge of other languages but also by how the previous languages have been acquired and by the strategies exploited in the process. Considering the great complexity and diversity in L3 acquisition, the study of CLI has a fundamental role in the analysis of the latter, which is why it is gaining more and more attention by scholars. Moreover, CLI allows us to analyze transfer between languages at different linguistic levels of analysis (lexicon, syntax etc.), which can be important to identify the process of transfer itself and the most influential language.

THE CROSS-LINGUISTIC INFLUENCE APPLIED TO SYNTAX

Until now most of the studies have considered the lexical aspect of L3 acquisition, yet in the last few years the interest in syntax has grown a lot.

Clearly, not all the studies agree that L3 acquisition is something different from L2 acquisition, as it is the case for Na Ranong and Leung (2009). These scholars claim that the L1 is the only source of transfer during all the acquisition process of non-native languages. The investigation they ran is focused on speakers with Thai as L1, English as L2, and Chinese as L3. The results of the research show a total preference for the L1 with no clear example of transfer from the L2. Therefore, the model they proposed assigns to the L1 a preferential, if not exclusive, role in L3 acquisition. Na Ranong and Leung place themselves on the same line as Schwartz & Sprouse (1996), who proposed the Full Transfer/Full Access (FT/FA) model, which claims that all the syntactic properties of L1 will form the core for the development of future foreign language syntax.

Bardel & Falk (2007), on the other hand, agree with the model previously proposed by Williams and Hammarberg (1998) and Hammarberg (2001), in which they show that, among all the languages known by a speaker, the L2 is the one that has the biggest impact on the acquisition process of the L3. The focus of the investigation was to realize the position of negation in Dutch or Swedish native speakers learning Dutch or Swedish as L3 and having English as L2. The results show that, in Swedish and Dutch as L3s, the acquisitional process of negation is the same – in both languages negation is placed after the finite lexical verb; speakers prefer the English pattern placing negation before the lexical verb. In this case they observe a total preference for the L2, proved by the fact that all transfers came from the latter and not from the L1, with the speakers ignoring the resemblance between L1 and L3. Bardel & Falk explained this result by affirming that in some cases the L2 can also stop transfer from L1, and they confirmed it in a later study of 2010.

A different position is proposed by Flynn et al. (2004) who developed the Cumulative Enhancement Model (CEM), in which transfer arises from a previous learned language, be it L1 or L2, if the characteristics of one or the other can match the L3. According to this model, all the languages previously learned can be a possible source of transfer, but the L2 only comes into play when it comes to some structure that does not exist in the L1, and that in order to facilitate a positive transfer.

“Language learning is cumulative. All languages known can potentially influence the development of subsequent learning. The learner’s L1 does not play a privileged role in subsequent acquisition. Where appropriate, other languages known can enhance subsequent acquisition”. (Flynn et al. 2004)

Rothman & Cabrelli Amaro (2010) demonstrated that the CEM model proposed by Flynn et al. works but just partially, since transfer can take place even if the syntactical structures do not match the ones present in L1 or L2, producing by that a negative transfer. That means that transfer from L1 will occur both when the structure in question corresponds with a structure in L1 and when it does not; similarly, transfer from L2 will take place if the structures under focus match some in L3 and also when they are perceived in an incorrect way.

Research by Flynn et al. (2004), concerning the acquisition of pronominal forms, has been performed onto two different groups of learners: 1) English L1, Spanish L2, French L3; 2) English L1, Spanish L2, Italian L3, and it shows that transfers to L3 from L1 or L2 do not always take place in a way which facilitates the acquisition of the correct L3 structures. The study demonstrates that the typological factor prevails on the others, despite the fact that, in the case of French L3, pronouns are much more similar to English than to Spanish.

Rothman’s study (2011) is on groups having: Italian as L1, English as L2, Spanish as L3; and English as L1, Spanish as L2, and Portuguese as L3. This work has led to the creation of a fourth model: the Typological Primacy Model (TPM), which suggests that during the acquisition process the typological and psycho-typological factors are crucial.

Initial state transfer for multilingualism occurs selectively, depending on the comparative perceived typology of the language pairings involved, or psychotypological proximity. Syntactic properties of the closest (psycho)typological language, either the L1 or L2, constitute the initial state hypotheses in multilingualism, whether or not such transfer constitutes the most economical option (Rothman 2011).

All the studies about L3 realized in the last few years can be essentially linked to the four models described above.

DATA COLLECTION AND RESULTS

For the corpus used to run the experiment under focus, we interviewed students of Spanish language at the University of Naples “Federico II” with a B2 level in English and who had never studied Spanish before.

As first, they were given two questionnaires: one to identify their previous linguistic skills and another to know the conditions of their learning (natural or institutional context). Some questions aimed at investigating whether the students had developed linguistic prejudices about the L2 (English) during their acquisitional process and whether this also affected the new language they were learning.

The study could count on 41 students enrolled in the first year with an absolutely homogeneous knowledge of Spanish L3: all of them had Italian as L1 and English as L2 as linguistic background and the same knowledge in Spanish L3. The first test was conducted in November 2016 when the students had a beginner level in Spanish, in order to collect data about the first step of acquisition of the L3.

The questionnaire focused on the different uses of definite and indefinite articles in all three languages, so that for every question the participants had to choose the right answer between two different options. In other words, we gave two options for Spanish L3: one followed the syntax of the L1 and the other one the L2 but just one could be correct in Spanish L3. In this way, we wanted to explore the language that students looked in when acquiring Spanish L3. Not knowing the right structure in Spanish, the participants were supposed to use one of the two languages available.

ANALYSIS OF THE RESULTS

Given the substantial resemblance between L1 and L3, we expected students to exclusively have recourse to Italian L1 to answer, but surprisingly they got the right answer in 90% of contexts, even though the chosen syntactic structure matched English L2 and not Italian L1. This means that students chose Italian when the right answer in Spanish L3 matched Italian L1 as well as they chose English L2 when the right answer in Spanish matched the L2.

We list some examples in order to clarify the kind of operations that the learners did.

In examples 1 and 2 the correct answer for Spanish L3 matched Italian syntax:

(1)

Sólo el 10% aprueba el examen. spanish = italian

Sólo 10% aprueba el examen. spanish ≠ english

(2)

La gramática es más interesante de lo que parece. spanish = italian

Gramática es más interesante de lo que parece. spanish ≠ english

In examples 3 and 4, instead, the correct answer in Spanish L3 matched English syntax namely the L2:

(3)

Conozco muy bien España. spanish = english

Conozco muy bien la España. spanish ≠ italian

(4)

El mi libro es interesante. spanish ≠ italian

Mi libro es interesante. spanish = english

A very low rate, 10%, was observed in mismatched cases, where the wrong option was chosen, such is the case of examples 5 and 6:

(5)

Juan lleva las gafas. spanish ≠ italian

Juan lleva gafas. spanish = english

(6)

José se lava sus manos. spanish ≠ english

José se lava las manos. spanish = italian

The participants were supposed to choose only the L1, since both the mother tongue and the target language belong to the same linguistic family. Yet the results show that the learners use both L1 and L2 and that none used only the L1, as well as no student using his/her knowledge in L2, which obliges us to discard the model proposed by Na Ranong & Leung (2009), Bardel & Falk (2007) and Flynn et al. (2004). So, we have a transfer from the L2 even when the L1 perfectly matches the L3 (negative transfer). In the same way, either the L1 or the L2 is used when the one or the other matches the L3. As a result, these findings are consistent with Rothman (2011).

What has been achieved shows that in the acquisition of an L3 typologically close to the L1, the typological prejudice plays an important role: the learner can persist in this conviction, which leads to always choosing the L1 during the acquisition process of the L3; or he/she rejects this idea, which will trigger the opposite process, in which the L2 is chosen in the L3 acquisition.

In the following example, most of the participants prefer the syntactic form of the L2 although the correct one was the one expressed by the L1 syntax:

(6)

José se lava sus manos. spanish ≠ english

José se lava las manos. spanish = italian

Giving the answer, the students rejected the sentence José se lava las manos because too similar to the structure of their L1 José si lava le mani. The resemblance is perceived to be excessive, which leads the students to ban it as deceptive, and so wrong. Judging this resemblance erroneous, the students have recourse to the L2, in which the possessive is used instead of the article, resulting in the sentence *José se lava sus manos.

CONCLUSIONS

According to the number of languages known by a speaker, the possible elements of influence in the process of acquisition of a new language automatically increase. So, it cannot be assumed that cross-linguistic interference comes from just one of the known languages. When more languages are available, they can all be used. As our test points out, one can make a transfer using both the L1 and the L2, which means, in the case of multilingual speakers, that the mother tongue does not play any preferential role. Nevertheless, it is still not possible to certainly identify the factors which are most relevant in transfer.

The cross-linguistic influence gives and will continue to give answers. If elements like relatedness or typology have proven to be crucial in the learning process of L2, for what concerns L3, they are clearly not the only ones to be taken into account. Factors like psychotypology, proficiency, and L2 status have proven to be even more important due to the results of several studies about L3.

As far as the acquisition of Spanish as L3 by Italian speakers with English L2 is concerned, what seems to play the most important role in the acquisitional process is the relatedness and the typological prejudices, which triggers the phenomenon of hyper-correction. If we consider that other possible factors have not been detected yet and that some others are not properly studied, we can conclude that the available corpora do not yet yield exhaustive results.

If the acquisition of L3 is to be considered just as a cumulative process and not as a bunch of separate systems, further studies crossing L1s, L2s, and L3s different from the ones studied to now will be necessary. L3 acquisition is a complex and huge field, which is why the study of this phenomenon needs new methods of research for what concerns both data collection and the analysis of corpora. For example, it would be helpful to count on studies which focus on the acquisition of more linguistic levels at the same time and to explore the phonological level in order to identify the “bridge-language”.

The distinctive feature of L3 acquisition process, due to the great number of possible combinations and interactions between languages, allows us to face unique situations that make it different from the acquisition of L1 and L2. If on the one hand we can see some similarities with L2 acquisition, such as the research methods, on the other hand we have to admit that there are different characteristics and operations in L3 acquisition, which make it absolutely singular. At the same time, L3 acquisition also allows – through the consideration of the impact of bilingualism on the acquisition of L3 – to deepen the study about the relationship between bilingualism and the way knowledge is acquired.

To acquire a new linguistic system is not just to enrich oneself but also to change forever one’s life and one’s own capacity to learn.

BIBLIOGRAPHY

Aronin, L. and Singleton, D., 2008. "Multilingualism as a new linguistic dispensation". International Journal of Multilingualism, 5 (1), 1-16.

Bardel, C. and Falk, Y., 2007. "The role of the second language in third language acquisition. The case of Germanic syntax". Second Language Research, 23 (4), 459-484.

Bardel, C. and Falk, Y., 2010. "The Study of the Role of the Background Languages in Third Language Acquisition. The State of the Art". IRAL, International Review of Applied Linguistics in Language Teaching, 48 (2), 185-219.

Cenoz, J., 2000. "Research on Multilingual Acquisition". In: Cenoz, J. and Jessner, U., English in Europe: The Acquisition of a Third Language. Clevedon: Multilingual Matters, 39-53.

Cenoz, J., 2001. The Effect of Linguistic Distance, L2 Status and Age on Crosslinguistic Influence in Third Language Acquisition. In: Cenoz, J., Hufeisen, B. and Jessner, U., Cross-linguistic Influence in Third Language Acquisition, Psycholinguistic Perspectives. Clevedon: Multilingual Matters, 8-20.

Cenoz, J., 2003. "The additive effect of bilingualism on third language acquisition. A review". International Journal of Bilingualism, 7, 71-87.

Clyne, M., 1997. "Some of the things trilinguals do", International Journal of Bilingualism, 1, 95-116.

De Angelis, G. and Selinker, L., 2001. "Interlanguage transfer and competing linguistic systems in the multilingual mind". In: Cenoz, J., Hufeisen, B. and Jessner, U., Cross-linguistic Influence in Third Language Acquisition, Psycholinguistic Perspectives. Clevedon: Multilingual Matters, 42-58.

De Angelis, G., 2007. Third or Additional Language Acquisition. Clevedon, Multilingual Matters, 1-3; 20-57; 110-138.

Dewaele, J.-M., 2001. "Activation or inhibition? The interaction of L1, L2 and L3 on the language mode continuum". In: Cenoz, J., Hufeisen, B. and Jessner, U., Cross-linguistic Influence in Third Language Acquisition, Psycholinguistic Perspectives. Clevedon: Multilingual Matters, 69-89.

Hammarberg, B., 2001. "Roles of L1 and L2 in L3 production and acquisition". In: Cenoz, J., Hufeisen, B. and Jessner, U., Cross-linguistic Influence in Third Language Acquisition, Psycholinguistic Perspectives. Clevedon: Multilingual Matters, 21-41.

Hammarberg, B. and Williams, S., 2005. "A study of third language acquisition". In Hammarberg, B., Processes in Third Language Acquisition, Edinburgh, Edinburgh University Press: 17-27.

Herdina, P. and Jessner, U., 2000. "The Dynamics of Third Language Acquisition". In Cenoz, J., Jessner, U., English in Europe: The Acquisition of a Third Language, Clevedon, Multilingual Matters: 84-98.

Lemhöfer, K., Dijkstra, A. and Michel, M., 2004. "Three languages, one ECHO: Cognate effects in trilingual word recognition". Language and Cognitive Processes, 19, 585-611.

Lindqvist, C., 2015. "Do learners transfer from the language they perceive as most closely related to the L3? The Role of Psychotypology for Lexical and Grammatical Crosslinguistic Influence in French L3". In: De Angelis, G., Jessner, U. and Kresic, M., Crosslinguistic Influence and Cross Linguistic Interaction in Multilingual Language Learning. London: Bloomsbury Publishing, 231-252.

Lindqvist, C. and Bardel, C., 2010. "Approaches to third language acquisition: Introduction". IRAL, International Review of Applied Linguistics in Language Teaching, 48 (2-3), 87-90.

Murphy, S. (2003). "Second Language Transfer during Third Language Acquisition". TESOL & Applied Linguistics, 3 (2), 1-21

Na Ranong, S. and Leung, Y.-k. I., 2009. "Null objects in L1 Thai-L2 English-L3 Chinese: An empiricist take on a theoretical problem". In: Leung, Y.-k. I., Third Language Acquisition and Universal Grammar. Bristol: Multilingual Matters, 162-191.

Paradis, M., 1994. "Neurolinguistics aspects of implicit and explicit memory: implications for bilingualism and second language acquisition". In: Ellis, N., Implicit and Explicit Language Learning. London: Academic Press, 393-419.

Ringbom, H., 2007. Cross-linguistic similarity in foreign language learning. Clevedon: Multilingual Matters, 1-9; 54-97.

Rothman, J. and Cabrelli Amaro, J., 2010. "What variables condition syntactic transfer? A look at the L3 initial state". Second Language Research, 25 (4),1-30.

Rothman, J., 2011. "L3 syntactic transfer selectivity and typological determinacy: The typological primacy model". Second Language Research, 27, 107-128.

Schwartz, B.D. and Sprouse, R.A., 1996."L2 cognitive states and the Full Transfer/Full Access model". Second Language Research, 12, 40-72.

Williams, S. and Hammarberg, B., 1998. "Language Switches in L3 Production: Implications for a Polyglot Speaking Model". Applied Linguistics, 19 (3), 295-333.