See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/320394588

Third

Chapter · January 2018

CITATIONS READS 0 2,836

2 authors:

Jennifer Cabrelli Mike Iverson University of Illinois at Chicago Indiana University Bloomington

36 PUBLICATIONS 388 CITATIONS 36 PUBLICATIONS 315 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Third language acquisition View project

Second language acquisition View project

All content following this page was uploaded by Jennifer Cabrelli on 14 October 2017.

The user has requested enhancement of the downloaded file.

Cabrelli Amaro, J., & Iverson, M. (in press). Third language acquisition. In K. Geeslin (Ed.), Handbook of Spanish Linguistics. Cambridge: Cambridge University Press.

1 Introduction Language acquisition research has long made distinctions between first and second language acquirers, recognizing the potentially different paths and outcomes and carrying out systematic investigation of these acquisition scenarios. However, only recently—within the past 20 years—has the distinction between second and third language acquirers been treated with the same fervor. Acknowledging that acquiring a second and third language are different processes is a crucial step for both second language (L2) and third language (L3) research. For L2 research, it means abandoning the tacit assumption that all acquisition beyond the native language is comparable and identifying the possible confound of including multilinguals in an L2 study. For L3 research, it has meant the birth of a new subfield. In the following chapter, we give a brief history of L3 research, detail the current state of the field with a focus on the impact of Spanish on the study of L3 acquisition, and provide possible directions for future lines of inquiry.

2 Early studies of third language acquisition To understand the origin of the lines of inquiry that currently drive L3 acquisition, it is necessary to consider a set of studies which arguably constructed the field’s foundation at the end of the twentieth century. The research we report on here ranges from early impressionistic diary studies to large-scale quantitative studies, and dates to Vildomec (1963), which to our knowledge was the first systematic treatment of . Larger-scale experimental work began later in the 1960s (e.g., Rabinovitch & Parver [1966] for phonology) and 1970s (e.g., Stedje [1977] for lexical transfer), but as our review will reflect, appearance of these studies was sporadic and an appreciable increase in output began in the 1980s and 1990s. From this body of work, we have pulled out the variables with the most substantial presence, and detail general findings.1 We then segue into a brief discussion of the limitations of this research before moving onto research from the last fifteen years. We note that, while most of the studies mentioned in this section do not involve Spanish, the incorporation of these studies is necessary to establish the early trajectory of this area of investigation. When considering the L3 research that came out of the last decades of the twentieth century, a central theme can be identified, which is that of transfer/cross- linguistic influence (CLI). Herein, we focus on this theme, while acknowledging that this is one of several lines of inquiry in the field of multilingualism. The fundamental questions related to CLI come out of the notion that, unlike second language acquisition, there are two background languages to contend with. So, what CLI patterns are present at

1We recognize that this overview of L3 research in the twentieth century is abbreviated and (necessarily) simplified. For exhaustive reviews of earlier L3 research, we direct the reader to De Angelis (2007) for a general overview, Cabrelli Amaro (2012), Cabrelli Amaro and Wrembel (2016), and Wrembel (2015) for phonology, García-Mayo (2012) for cognitive approaches, and García-Mayo and Rothman (2012) for morphosyntax. different stages in L3 acquisition, and what are the variables that yield these patterns? One primary variable proposed to drive L3 acquisition has been language status. Is the source of transfer more likely to be the L1, the L2, or both, and why? Early research suggested that reliance on the L1 as opposed to the L2 depends on the domain of grammar, with research from this period limited primarily to lexical transfer and, to a lesser degree, phonetic/phonological transfer. Several studies showed that lexical transfer comes primarily from the L2 (e.g., Rivers, 1979; Stedje, 1977; Vogel, 1992), although semantic transfer has been posited to originate with the L1 (Ringbom, 1987). Regarding phonological transfer, Llisterri and Poch-Olivé (1987) and Ringbom (2001) report evidence of L1 transfer (with Ringbom [2001] showing long-term effects of L1 transfer, particularly for intonation), while Bentahila (1975) and Chumbow (1981) report L2 phonological influence in learner speech. How might age of acquisition of a language relate to source of transfer? There was a common assumption that L1 transfer occurs due to the entrenched nature of the and the cumulative experience that a learner has with the language compared to the L2, referred to as the ‘mother tongue effect’ (Chumbow, 1981). Some researchers proposed that L2 transfer might be driven by a conscious desire to avoid the L1 to sound ‘non-foreign’ (see e.g., Hammarberg, 2001, for a discussion of this phenomenon). Other proposals for a ‘foreign language effect’ (e.g., Meisel, 1983) focused on similarities in L2 and L3 acquisition processes (cf. Stedje [1977] and Ringbom [1987], who claim that L2 transfer is more likely if acquired naturalistically), which results in the blocking of L1 transfer (e.g., Dewaele, 1998). Still others noted that L2 transfer could be the product of recency, whereby the L2 is transferred because it has most recently been activated (Hammarberg, 2001; Vildomec, 1963), or that L2 transfer will be more likely in an L2 context (Stedje, 1977). How can we be certain of an L2 effect and explain contradictory findings such as those reported earlier for L3 phonological transfer? Unfortunately, we cannot isolate the role of language status without considering other potential variables in the research design. For example, Stedje’s (1977) investigation of L1 Finnish/L2 Swedish L3 German learners’ lexical transfer reported L2 transfer, but the learners’ L2 also happens to be more similar than Finnish to German. Therefore, it was not possible to pinpoint whether there was an L2 effect or whether the typological relationship between Swedish and German drove transfer. This confound was present in most early research, which primarily consisted a single learner or group of learners with similar linguistic backgrounds. Several studies reported (perceived) typological similarity as a catalyst for transfer (e.g., Bild & Swain, 1989; Singh & Carroll, 1979; Singleton, 1987) but none of these could rule out the role of other factors in their design. The few exceptions are studies by Ringbom (1987), Cenoz and Valencia (1994), and Llisterri and Poch-Olivé (1987), all of which employed a design using mirror-image groups: they observed two groups of sequential bilinguals that had each acquired the same language pair, but in reverse order; this design made it possible to tease apart language status and similarity. For example, Ringbom (1987) compared two groups of L3 English learners, L1 Finnish/L2 Swedish and L1 Swedish/L2 Finnish, and found that both groups transferred Swedish lexical items independently of whether Swedish was the L1 or L2. These early studies informed methodological issues, and prompted foundational questions of current research, which we review in the following section.

3 Research questions that drive the study of L3A and a review of the literature The research discussed in Section 2 made clear the value of examining third language acquisition not just as another instance of L2 acquisition and established several research questions that have been further refined over the last two decades. Researchers have begun to fill in some of the gaps pointed out by Fouser (1995) in his synthesis of L3 acquisition research, including the lack of investigation of both process and product, production and comprehension, and facilitative and non-facilitative transfer. Within this section, we address the progress that has been made towards what we see as the primary research questions that drive this field: understanding a) what catalyzes initial transfer to the L3 and b) developmental processes and their effects on the L3 as well as the L1/L2. Throughout this discussion, the substantial role of Spanish in the investigation of L3 acquisition will become clear.

3.1 Cross-linguistic influence during the L3 initial stages While L3 acquisition is a distinct area of inquiry, it borrows from insights in the second language acquisition (SLA) literature. L2 research shows that a learner’s prior linguistic experience (the L1) can influence L2 development, and that the entire L1 grammar may form the initial state of L2 acquisition (e.g. Schwartz & Sprouse, 1996). This is readily acknowledged in L3 research, as is the additional layer of complexity of two potential sources of cross-linguistic influence, the L1 and the L2. The first explicit models of L3 acquisition aimed to predict patterns of linguistic transfer for beginning L3 learners (specifically, morphosyntactic transfer, which was not a focus of the research reported in Section 2). However, they differ in their conception of a) the consistency of transfer—is the source of transfer fixed (always the L1 or L2) or variable (possibly the L1 or L2)?—and b) the selectivity of transfer—is a full grammatical system transferred, or only a partial system?—as well as the underlying catalysts for transfer. One logical possibility in L3 acquisition is that the L1 serves as the L3 initial state and influences behavior in the L3 at the initial stages. This is consistent with models of L2 acquisition in which the L1 serves as the initial state of the L2 (e.g. Full Transfer/Full Access, Schwartz & Sprouse, 1996). In such a model, the L1 holds a privileged role as the sole source of transfer, to the exclusion of other factors. This should give rise to two observable consequences. First, behavior at the initial stages of L3 acquisition of (for example) French by a speaker of (for example) L1 Spanish and L2 English should be similar to the behavior of a native speaker of Spanish learning French as an L2, as L2 English won’t factor into the initial state of L3 French.. Second, acquisition beyond an L3—L4, L5, etc.—should yield similar results, at least at the outset. In cases where a linguistic property is shared in the L1 and the L3, acquisition will be facilitated (i.e., positive or facilitative transfer); however, in cases where the L1 and L3 do not align, acquisition will be hindered (i.e., negative or non-facilitative transfer). L1 transfer has been demonstrated. In L1 Moroccan Arabic/L2 French learners of L3 English, L1 influence has been evidenced by placement of negation and adverbs (Hermas, 2010), subject-verb inversion in declarative sentences (Hermas, 2014), and knowledge of restrictive relative clauses (Hermas, 2015), despite the closer typological relation between L2 French and L3 English. Sanz, Park, and Lado (2015) examined reliance on L1 (English) or L2 (Japanese or Spanish) morphosyntactic cues in the processing and interpretation of transitive sentences in L3 Latin at the initial stages. In English, thematic roles in a transitive sentence are established primarily via rigid SVO word order; both case and subject-verb agreement are not consistently marked. In contrast, word order in Japanese and Spanish is flexible, and is therefore not a reliable indicator of an entity’s role. The relationship among entities is established through morphology: Case in Japanese and subject-verb agreement in Spanish are (sufficiently) unambiguously marked. Latin, like Japanese and Spanish, exhibits free word order, and also marks both Case and agreement. Despite the possibility of using a morphological cue as in L2 Japanese or Spanish, L3 learners of Latin relied on word order for interpretation—consistent with transfer from L1 English. Another logical possibility is that the L2 always serves as the L3 initial state. This has been formalized as the L2 Status Factor (e.g., Bardel & Falk, 2007; Falk & Bardel, 2011). In this model, the L2 serves as the source of transfer because of its status as a non- native language, a characteristic which it shares with the L3 and all subsequently acquired languages. This distinction has roots in Paradis’ (2004) Procedural/Declarative model, in which native and non-native languages are stored in fundamentally different ways, and therefore rely on different types of knowledge. Native languages are stored in procedural memory, while non-native languages (learned in adulthood) are learned skills that utilize declarative memory. One consequence of such a model is that the initial state of L3 acquisition should be the same for learners who share an L2, regardless of L1. Seminal work by Bardel and Falk (2007) tested the placement of negation in L3 Swedish or Dutch by two groups of speakers of Germanic L1s and L2s. Both Swedish and Dutch are classified as verb-second (V2) languages, in which the verb must be the second element in main clauses. Placement of negation is considered to be a property related to V2. In this study, one group had an L1 without V2, but spoke an L2 with V2, and the other group had an L1 with V2, but spoke an L2 without V2. The group who spoke an L2 with V2 showed no difficulty with the placement of negation, outperforming the non-V2 L2 group. Given these results, the nature of the L2 seemed to be the decisive factor in L3 performance. This study was followed by an examination of L1 French/L2 English and L1 English/L2 French learners of L3 German. In German, object pronouns are post-verbal in matrix clauses, patterning with English, and preverbal in subordinate clauses, patterning with French. Both groups showed non-facilitative influence from their respective L2s, failing to reject ungrammatical sentences in German if the word order was grammatical in the L2. The L2 status factor has also been posited to explain L3 Spanish phonological transfer of voiceless stops by mirror-image groups of English/French bilinguals (Llama, Cardoso, & Collins, 2010); both groups were shown to produce L2-like voice onset time (VOT) in Spanish. To our knowledge, this is the only mirror-image study involving Spanish to show an L2 effect for phonology. The scenarios mentioned above assume that transfer is non-selective—an entire grammatical system is transferred—and that transfer is fixed—it is always the L1 or always the L2 that is transferred. However, the empirical evidence supporting both L1 and L2 transfer calls this assumption into question, and suggests that transfer into an L3 may be variable. Unlike models in which the source of transfer is fixed, models in which the L1 or the L2 may be transferred require some sort of trigger for transfer, tipping the proverbial scale to one grammar or the other. One such model is Rothman’s Typological Primacy Model (TPM) (2011, 2013, 2015). In this model, structural similarity serves as the trigger for the transfer of the L1 or L2 into the L3. Early conceptualizations of this model claimed that it was the learner’s perception of overall typological similarity, in the vein of Kellerman’s (1983, 1986) psychotypology, which drives transfer. If a learner (unconsciously) deemed the L3 more globally similar to their native language, the L1 would be transferred. In cases where the L2 was determined to be more similar, the L2 would be transferred. Because this is a model which envisions the transfer of an entire grammatical system, transfer can be both facilitative or non-facilitative. Later instantiations of this model (Rothman, 2013, 2015) detailed how selection of the source of transfer might happen. As learners are confronted with L3 input, they make use of the most salient cues across linguistic domains. Rothman (2015, p. 185) states that the following factors impact the source of transfer, with decreasing influence: lexicon, phonological/phonotactic cues, functional morphology, syntactic structure. The lexicon is the first point of comparison. In the case that the learner is able to parse individual lexical items in the L3 and a sufficient number are similar to those in a previously acquired language, that language will serve as the source of transfer. If, on the other hand, both the L1 and the L2 lexicons are comparably similar to the L3 lexicon, or neither the L1 nor the L2 lexicon aligns with the L3 lexicon, or individual lexical items in the L3 are not able to be parsed (and cues from the L3 lexicon are therefore unavailable), the learner will make use of the next set of cues—phonological and phonetic cues. If these cues are unusable, or result in a tie of sorts, the learner moves on to the next set of cues, and so on. In this model, typological/structural similarity must be assessed in a global sense, and independently of the linguistic property under investigation. Furthermore, because the source of transfer must be determined early, the learner makes this assessment with extremely limited L3 data, and any claims as to whether the L1 or L2 grammar will be transferred should take this into consideration. Although the expected source of transfer is not always easy to determine, much of the work on the TPM, including the foundational research, has avoided potential confounds in gauging the more similar language pair by exploiting the close historical, typological, and linguistic ties between Spanish and other Romance languages. Rothman and Cabrelli Amaro (2010) tested L1 English/L2 Spanish learners of L3 Italian or L3 French on properties related to the availability of null subjects. Because Italian and French are both Romance languages, L3 learners are expected to transfer Spanish. However, Spanish and Italian are null subject languages, while French and English are not, so transfer of Spanish would facilitate acquisition of only Italian, while transfer of English would facilitate acquisition of only French. L3 learners showed behavior consistent with transfer of Spanish, treating both Italian and French like null-subject languages. This is also consistent with the L2 Status Factor, leading Rothman to incorporate mirror-image groups in subsequent studies. Many studies evaluating the TPM have examined the acquisition of L3 (Brazilian) Portuguese (BP), often using mirror-image groups of L1 English/L2 Spanish and L1 Spanish/L2 English speakers. The grammars of Spanish and Portuguese are similar in many respects, and an L1 or L2 speaker of Spanish is expected to use the Spanish grammar at the initial stages of L3 Portuguese. Knowledge of Spanish as an L1 or an L2 has been shown to facilitate acquisition of a range of properties in L3 Portuguese, including adjective placement and interpretation of pre-nominal and post-nominal adjectives (Rothman, 2010), and gender/number marking and noun-drop (Iverson, 2009). For properties in which Spanish and Portuguese diverge, non-facilitative transfer is predicted. This has obtained for a range of properties, including relative clause attachment and word order (Rothman, 2011), differential object marking (Giancaspro, Halloran, & Iverson, 2015), mood distinctions (Child, 2017), verb raising across dative experiencers (Cabrelli Amaro, Amaro, & Rothman, 2015), object expression (Montrul, Dias, & Santos, 2011), as well as in intervocalic stop realization and mid-vowel contrasts (Cabrelli Amaro & Pichan [unpublished data]). Considering the entire body of work on English/Spanish/L3 Portuguese, it seems evident that transfer may be non-facilitative. It is also notable that Spanish, whether it is the L1 or L2, is transferred for each property that has been tested—consistent with the transfer of an entire grammar. The models mentioned so far all assume transfer of an entire grammatical system. However, this is not the only logical possibility, and there are models in which transfer is assumed to happen on a property-by-property basis. The Cumulative Enhancement Model (CEM) (Flynn, Foley, & Vinnitskaya, 2004) claims that third language acquisition is both a cumulative and non-redundant process. Transfer is not wholesale, but rather structures or properties from any previous language are equally available to an individual L3 learner. Additionally, transfer into the L3 will only happen in cases in which the transfer is facilitative, which distinguishes this model from the rest. The basis of this model is data from L1 Kazakh/L2 Russian speakers of L3 English. While Kazakh is a head-final language (i.e. complements precede heads), both Russian and English are head-initial languages. These learners’ production of relative clauses L3 English was head-initial, suggesting influence from L2 Russian and contradicting a claim of any privileged status for the L1. These results where corroborated by data from L1 German (head-final in embedded clauses) and L1 German/L2 Hungarian (head-initial) learning L2 or L3 English (Berkes & Flynn, 2012). L3 learners showed influence from L2 Hungarian, facilitating the acquisition of relative clause structure as compared to L2 English learners. More recent models of L3 acquisition are similar to the CEM in that they question wholesale transfer. The Linguistic Proximity Model (LPM) (Mykhaylyk, Mitrofanova, Rodina, & Westergaard, 2015; Westergaard, Mitrofanova, Mykhaylyk, & Rodina, 2016) claims that transfer into the L3 happens on a property-by-property basis, and can be sourced from all previously acquired languages. Transfer is conditioned by abstract linguistic structure. In cases where a target L3 property is structurally similar to a property from a previously acquired language, transfer into the L3 will be facilitative. If, however, the L3 input is misanalyzed, resulting transfer may be non-facilitative. This process is active at any stage of L3 acquisition, not just the beginning stages. Development is not just comparing initial hypotheses about the L3 grammar to the L3 input and making subsequent adjustments to the L3 linguistic system. Learners entertain their initial hypotheses, the L3 input, but also continue to access any previously acquired linguistic structures (and perhaps recruit them based on the L3 input), even at later stages of the acquisition process. Slabakova’s (2016) Scalpel Model also stipulates transfer on a property-by-property basis, and allows for facilitative and non-facilitative transfer. In contrast with the LPM, however, transfer is conditioned by other factors in addition to structural similarity, such as frequency and misleading input. Low-frequency constructions and those that would require negative evidence for acquisition (with respect to the competing property in the L1/L2) are predicted to be susceptible to non-facilitative transfer.

3.2 L3 development Transfer at the initial stages is (literally) just the beginning in terms of the questions we can explore at different stages of L3 acquisition and how these questions can inform linguistic theory more generally. In this section, we focus on two questions related to the nature of cross-linguistic influence throughout acquisition: What are the roles of the L1 and L2 after initial transfer, and what effects does the development of an L3 have on the L1 and L2? A small but growing body of work indicates that the roles of the L1 and L2 may change substantially. Ringbom (2000) put forth the logical notion that the L1 and L2 exert the heaviest amount of influence on the L3 at the earlier stages of L3 acquisition, and that as L3 proficiency increases, L1 and L2 influence decreases. Early research suggested that the influence of the L1 and L2 might change throughout development. Specifically, analysis of production data from case studies by Vogel (1992) and Williams and Hammarberg (1998) suggests that the learners’ reliance on the L2 diminishes earlier than the L1. Without controlling for variables such as order of acquisition (via inclusion of a mirror- image learner group) and similarity, however, it was not possible to definitively conclude that these learners overcame non-facilitative transfer from the L2 earlier than the L1 because of language status. As shown in Section 3.1, most studies involving Spanish as an L1 or L2 and a Romance L3 find that Spanish transfers regardless of other factors. To tease apart the predictions of the CEM and TPM, the property selected should be a mismatch in its realization between the two similar languages, resulting in non-facilitative transfer of Spanish. The subsequent L3 learning task is for the learners to retreat from initial non-facilitative transfer. Cabrelli Amaro and Rothman (2010) first noted that this task might pose more of a challenge for learners that transfer their L1 than for learners that transfer their L2, citing the entrenched nature of the L1 as a potential source of difficulty. Recently, researchers have begun to examine differences in rate of L3 morphosyntactic development in cases of initial non-facilitative transfer. Following up on the initial stages DOM data from Giancaspro et al. (2015), Cabrelli Amaro, Iverson, Giancaspro, and Halloran (unpublished data) administered the same acceptability judgment task to a pair of mirror image English/Spanish groups of advanced L3 BP learners. A cross-sectional analysis reveals that the L2 Spanish advanced group patterns with BP native controls and appears to have overcome Spanish transfer of DOM, while the L1 Spanish advanced group continues to pattern with the initial stages groups. Cabrelli Amaro (2015) has replicated this finding with raising across a dative experiencer (RExp) in L3 BP, a structure that is ungrammatical in Spanish (*Pedro me parece estar triste ‘Pedro seems to me to be sad’ vs. Me parece que Pedro está triste ‘It seems to me that Pedro is sad’; Cabrelli Amaro et al., 2015), but not in English or BP. She took the initial stages data from Cabrelli Amaro et al.’s (2015) study of RExp in L3 BP and compared it with data from the same groups of advanced L3 BP speakers from Cabrelli Amaro et al. (unpublished data). The L1 Spanish advanced group continued to rate grammatical RExp in BP lower than the BP control, although the ratings were significantly higher than the L1 Spanish initial stages group. The L2 Spanish advanced group, on the other hand, rated RExp higher than the L2 Spanish initial stages group and the advanced ratings were not different from those of the BP controls. Therefore, the L2 Spanish group appears to have converged on the L3 target while the L1 Spanish group is still in the process of L3 reconfiguration. The question that follows from this apparent consequence of L1 transfer is whether these learners can eventually converge on the L3 target. Based on his study of L1 Greek/L2 English/L3 Spanish learners, Lozano (2002) states that fossilization will occur when the relevant L1 features do not match the L2 or L3 features, a finding supported by Jin (2009) for L1 Mandarin/L2 English/L3 Norwegian learners. However, there is evidence from other language triads that L1 transfer can be overcome when the relevant L3 features are available in the background language(s) (Hermas, 2014). This finding is supported by Slabakova and García-Mayo (2015), which examines the acquisition of L3 English syntax-discourse interface properties by Spanish-Basque bilinguals that exhibit Spanish transfer. The results show non-facilitative Spanish transfer of topicalization to be persistent even at advanced L3 English proficiency for Spanish- dominant and Basque-dominant speakers alike. However, a subset of learners in both groups exhibit L3 English convergence, and the authors conclude that “trilingual learners are not permanently constrained by the native grammar in this domain” (p. 14). Even if access to linguistic universals is limited in L3 acquisition, the logical prediction is that the L1 Spanish group in this case will eventually converge on the L3 target. Cabrelli Amaro et al. (unpublished data) propose that the asymmetry between the L1 Spanish and L2 Spanish groups is due to probabilistic processes in L3 acquisition, and that feature reconfiguration will eventually obtain for both groups. Specifically, L3 input must reach a certain threshold at which point the mismatch between the L3 and the Spanish hypothesis will drive reanalysis. They assume that the difference between the L1 and L2 Spanish speakers comes down to cumulative Spanish experience, whereby a greater amount of L3 input is necessary for the L1 Spanish speakers to reach the same threshold. Our discussion up to this point has revolved around progressive transfer, that is, transfer from the L1 and/or L2 to the L3. However, a large body of work has established that cross-linguistic influence is bidirectional and an L2 can affect an L1 (see Schmid, 2016, for an annotated bibliography). A logical extension of this is that an L3 can affect previously acquired systems, and researchers have begun to examine a couple of questions via observation of L3 effects on the L1 and/or L2. The first is a new take on the long-standing question of whether L1 and L2 systems are fundamentally different (see e.g., Rothman & Slabakova [2017] for discussion), and the second is whether L3 acquisition can result in facilitative effects on the L2. To understand potential differences in stability between native versus non-native systems when an L3 is acquired, Cabrelli Amaro and Rothman (2010) and Cabrelli Amaro (2013a, 2016) compare the Spanish perception and production by L1 English/L2 Spanish and L1 Spanish/L2 English speakers. Cabrelli Amaro (2013a) did not find any significant differences in effects of L3 BP word-final vowel reduction on L1 versus L2 Spanish perception or production. As a follow up, Cabrelli Amaro (2016) reduced the data set to include only L3 BP speakers that patterned with a group of BP controls. She found that, while perception appeared to remain stable for both groups, L2 Spanish vowel production was more BP-like than L1 Spanish vowel production, although productions in both groups showed L3 effects. Sypiańska (2016) found L3 English effects on L1 Polish but not L2 Danish vowel production, although proficiency was not controlled for in this study. Cabrelli Amaro (2017) found similar results for morphosyntactic competence in a comparison of initial stages L3 BP data (from Cabrelli et al., 2015) with advanced L3 BP data. As in Cabrelli Amaro (2016), only the advanced learners that rated RExp in BP within the confidence interval of a BP control group were included. While both learner groups rate RExp higher than a group of Spanish controls, the L1 Spanish speakers rated RExp lower in Spanish than the L2 Spanish speakers and the BP controls. These initial findings from phonology and morphosyntax suggest that there might indeed be a critical period for stability of a linguistic system, but further (longitudinal) research is necessary to confirm these data. The L3 can also have beneficial effects on a background language(s). Recent research by Hui (2010), Matthews, Cheung, & Tsang (2014), and Tsang (2015) compares an L1 Cantonese/L2 English control group with an L1 Cantonese/L2 English/L3 French or German experimental group. In each case, the French or German property patterns with English and differently than Cantonese. Results show that the L2 English of the experimental groups exhibits acquisition of properties that the control groups’ does not; this is the case for subject relative clauses (Hui, 2010), tense/aspect (Matthews et al., 2014), and number agreement (Tsang, 2015). A similar result comes from Llinàs-Grau and Puig Mayenco (2016), who examined that-deletion in L3 English/L4 German learners (2L1 Catalan/Spanish); they report a higher rate of that-deletion in L3 English by the L4 German learners than by a control group of 2L1 Catalan/Spanish speakers of L3 English. This finding is attributed to (facilitative) regressive influence from the L3/L4. Although cross-sectional studies illustrate the potential impact of developmental data on our understanding of how multiple systems interact across the lifespan, researchers cannot go back in time and confirm how (advanced) L3 learners’ L2 looked like prior to L3 acquisition. Subsequent calls have been made for longitudinal research given the ability to control for variation (e.g., Cabrelli Amaro & Wrembel, 2016; Cenoz, 2003). Currently we know of only a few longitudinal studies involving Spanish (see e.g., Kopeckova [2016], for acquisition of L3 Spanish rhotics; García-Mayo & Villareal Olaizola [2011] and Ruíz de Zarobe [2005, 2008] for L3 English acquisition by Spanish/Basque bilinguals; Sánchez [2015], for L3 German lexical acquisition by Spanish/Catalan bilinguals). By observing learners over time, it is possible to first establish the L3 initial state for each learner, that is, the state of the L1 and L2 prior to L3 exposure. Learners can then be tested in all three languages over time to observe a) initial stages transfer, b) the roles of the learner’s background languages during acquisition, and c) how the background languages are affected as L3 proficiency increases. Without this research, we cannot confirm whether the existing developmental findings hold up. For example, consider the studies that show regressive transfer: It is not outside the realm of possibility that the L3 groups had already acquired those properties in their L2 (in the case of facilitative transfer) or that their L2 was not native-like at the onset of L3 acquisition (in the case of non-facilitative transfer). If we are unable to test those learners prior to L3 acquisition, however, we simply cannot know.

4 Future directions As with any line of inquiry, progress in the field of L3 acquisition has given way to new questions (and extensions of existing questions) as well as new ways of addressing these questions. In this section we outline a set of research questions yielded by the existing literature, followed by a brief discussion of some of the novel methods that are being explored.

4.1 New and understudied research questions As with any young field of scientific inquiry, progress is incremental. To date, the primary concerns of the study of multilingual acquisition have overwhelmingly addressed the initial stages of acquisition and the nature of linguistic transfer, and to a lesser extent, L3 development. While questions in these areas are still being answered, new questions are being asked, and additional lines of investigation are ripe for research. Herein, we consider some (but certainly not all) areas that merit further pursuit.

4.1.1 What happens with three unrelated languages? Practical limitations often restrict the scope of research, and studies on L3 acquisition are no exception to this. Researchers use the populations to which they have access, and L3 learners are more difficult to find than L2 learners. For many, language learning in adulthood is a costly endeavor in both time and effort, especially to reach the highest levels of proficiency, and learning a third language compounds this. This immediately restricts the number of L3 learners. The combinations of languages are also generally restricted. A person learning a third language does so for many reasons, but advancing linguistic research is not one of these. The combinations that are found are often the result of geographic and/or educational practicalities, as well as the status of English as the current lingua franca. Consequently, many L3 studies involve English and a major European language, and what has been very rarely seen are studies involving languages that are truly (typologically) unrelated. We only know of a handful of examples of these to date: L1 Tuvan/L2 Russian/L3 English (Kulundary & Gabriele, 2012), L1 Polish/L2 French/L3 English (Wrembel, 2014), L1 English/L2 Spanish/L3 Arabic (Goodenkauf & Herschensohn, 2015), L1 Arabic/L2 French/L3 English (Hermas, 2010, 2014, 2015), Basque/Spanish/L3 English (Slabakova & García Mayo, 2015). The purposeful inclusion of unrelated languages might offer insights that using related languages cannot. For example, some models such as the TPM relies on the L3 learner’s sensitivity to typological similarity. These relationships might be obscured in unrelated languages, perhaps to the point where the learner is either severely delayed in identifying similarities, or perhaps completely unable to do so. What happens in such a scenario? It could be the case that the strategy for selecting the L1 or L2 for transfer is variable. If the learner is unable to recognize any similarities between the L3 and L1/L2, perhaps they resort to a default secondary strategy, such as using the L2 as the initial hypothesis about the L3 grammar. Such an insight is only available if we examine unrelated languages.

4.1.2 Is L4 acquisition (and beyond) another instance of L3 acquisition? To our knowledge, there has not to date been a systematic investigation comparing L3 acquisition and L4 acquisition. At first glance, it may seem like an uninteresting comparison. On the other hand, it may be an entirely new scenario with results that differ from L3 acquisition. Consider the learning task as a speaker progresses from a monolingual to a multilingual. In the case of L2 acquisition, there is one possible source of transfer, and it is the first experience learning an additional language. L3 acquisition is slightly different: The speaker has experience acquiring an additional language, but this is the first experience in which there are multiple possible sources of transfer. For any language beyond the L3, the speaker has already had experience acquiring an additional language and dealing with multiple possible sources of transfer. Perhaps this could result in further streamlining of the acquisition process.

4.1.3 Do different types of bilinguals show similar patterns in L3 acquisition and development? A central finding from research on second language acquisition is that there are general differences in path and ultimate attainment among cases of simultaneous bilingualism, childhood second language acquisition, and adult second language acquisition (see Montrul [2004, 2008] for an overview). It is not unreasonable to expect that any differences in the linguistic systems and language experience among these types of bilinguals could influence the acquisition of a third language, warranting a systematic investigation of their L3 acquisition patterns. This line of research would be complementary to the increasingly prominent body of research on heritage speakers, which in general seeks to go beyond treating age-of-acquisition as a macro-variable, instead seeking a more nuanced look at variables relating to the relationship between the L1 and the L2 (or the two L1s). These variables include effects of age-of-acquisition, language dominance, metalinguistic knowledge, and relative influence of a native vs. non-native language. Although the current L3 models do not make explicit predictions for different scenarios of bilingualism, a handful of studies have examined L3 acquisition by heritage speakers. Results from Giancaspro et al. (2015) and Cabrelli Amaro et al. (unpublished data) showed that both English-speaking heritage speakers of Spanish) and sequential bilinguals of English and Spanish incorrectly used Spanish-like DOM at the initial stages of acquisition of L3 Brazilian Portuguese (BP). At later stages of L3 proficiency, this initial non-facilitative transfer was overcome by L1 English/L2 Spanish speakers. However, errors persisted for native speakers of Spanish, including the English-dominant heritage speakers of Spanish, suggesting dominance was not a deterministic factor in rate of acquisition. It is unclear if this result applies to all areas of the grammar, however, and language dominance may be an explanatory variable for L3 phonology. In fact, English- dominant heritage speakers of French have been found to produce longer English-like VOTs in L3 Spanish, despite the more similar, shorter VOTs found among Romance languages (Llama & López-Morelos, 2016). Language dominance has also been shown to influence an L3 accent more globally, as German-dominant heritage speakers of Turkish learning L3 English are perceived to exhibit German-like accents (Lloyd-Smith, Gyllstad, & Kupisch, 2017).

4.2 New methodological considerations As the questions at the core of third language acquisition research are further developed, a need arises for new methodologies. L3 methodological considerations have been addressed previously for syntax (e.g., García-Mayo & Rothman, 2012) and phonology (e.g., Cabrelli Amaro, 2013b), and recent applications of methods that have been gaining traction in the field of second language acquisition have the potential to provide new insights into the cognitive organization of multiple linguistic systems. Although selective transfer has been examined exclusively via behavioral methodology, ranging from grammatical judgments to story retelling to spontaneous production, none have captured online linguistic processing. Rothman, Alemán-Bañón, and González Alonso (2015) suggest using neurological data, such as event-related potentials (ERPs) measure by electroencephalography (EEG), for testing initial stages models and argue that online methods can “add new insights to and strengthen the descriptive and explanatory power of these models” (p. 5). If learners show evidence of native-like ERP signatures in their L2, the L2 should be a viable source of transfer and these learners can be tested in their L3. Rothman et al. (2015) propose a study that examines two ERP signatures associated with gender agreement violations and explain how the initial stages models could be tested in L3 learners of artificial languages (ALs) that are English/Spanish bilinguals. One of the artificial languages is lexically similar to Spanish (mini-Spanish) while the other is similar to English (mini-English), and both languages have number (like English and Spanish) and gender agreement (like Spanish only). The ALs consist of Spanish or English lexical items and word order but instantiate novel number and gender morphology. Since English does not have gender, transfer of English to mini-English would not be facilitative. Transfer of Spanish, however, would be facilitative for both ALs. To carry out the study, a pair of mirror-image groups would be trained on mini-English or mini-Spanish, for a total of four groups. After training, brain activity would be recorded with EEG while the learners completed a grammaticality judgment task in the relevant mini-language. The authors predict that sensitivity to gender violations will reflect Spanish transfer while absence of sensitivity will point to English as the source of transfer. The methodology the authors propose is promising in its ability to tap unconscious processes, and can be extended to other domains. Rothman et al.’s (2015) use of artificial language learning (ALL) warrants a brief discussion of the benefits of using artificial languages in L3 acquisition research. As alluded to in the previous paragraph, an artificial language is constructed “according to rules and properties (deterministic or probabilistic) that generate language-like sentences using pseudowords” (Onnis, 2012, pp. 35-36). While these systems are not natural languages, ALL stimuli have been found to elicit similar ERP signatures to those elicited by natural language stimuli (e.g., Friederici, Steinhauer, & Pfeifer, 2002, but cf. Christiansen, Conway, & Onnis, 2012, who found signatures suggestive of domain- general processing). The use of the ALL paradigm gives the researcher control over the input the learner receives, which is especially important in L3 initial stages research, where one of the greatest challenges is finding a sufficient sample of learners that have comparable outside exposure and prior knowledge of the target language. Another benefit is the control that the researcher has over the experimental design. As outlined in Section 3.1, testing all the initial stages models requires the selection of a linguistic phenomenon that patterns a specific way across the three languages under investigation. In practice, when working with the natural language triplets, the number of properties or phenomena that meet the criteria for initial stages testing is limited. However, with ALL, it is possible to construct a language with the configuration needed to specifically examine the role of the L1 and L2 we are investigating, while controlling for variables that are otherwise uncontrollable when the L3 is a natural language.

5 Conclusion As the field of third language acquisition continues to grow, its contribution to our understanding of language acquisition and linguistic theory more generally becomes more apparent. As seen in this chapter, there are a number of questions related to cross- linguistic influence that currently drive the field. As we get deeper into their examination, new questions arise that have the potential to make their own mark on how we look at the multilingual mind, and the ways in which we approach these questions evolve.

References

Bardel, C. & Falk, Y. (2007) The role of the second language in third language acquisition: the case of Germanic syntax. Second Language Research, 23, 459−484. Benrabah, M. (1991). Learning English segments with two languages. In Actes du XIIème Congrès International des Sciences Phonétiques. Aix-en-Provence: Université de Provence, Service des Publications, 334−337. Bentahila, A. (1975) The influence of the L2 on the learning of the L3. MA Bangor University. Berkes, É. & Flynn, S. (2012) Further evidence in support of the Cumulative- Enhancement Model: CP structure development, in Cabrelli Amaro, J., Flynn, S. & Rothman, J. (eds), Third Language Acquisition in Adulthood. Amsterdam: Benjamins, 143-164. Bild, E. R. & Swain, M. (1989) Minority language students in a French immersion programme: Their French proficiency. Journal of Multilingual & Multicultural Development, 10(3), 255-274. Cabrelli Amaro, J. (2012) L3 phonology: An understudied domain, in Cabrelli Amaro, J., Flynn, S. & Rothman, J. (eds), Third Language Acquisition in Adulthood. Amsterdam: Benjamins, 33-60. Cabrelli Amaro, J. (2013a) The Phonological Permeability Hypothesis: Measuring regressive L3 influence to test L1 and L2 phonological representations. PhD University of Florida. Cabrelli Amaro, J. (2013b) Methodological issues in L3 phonological acquisition research. Studies in Hispanic and Lusophone Linguistics, 6(1). Cabrelli Amaro, J. (2015). Does the source of transfer affect the rate of L3 morphosyntactic development? Poster presented at the Boston University Conference on Language Development, Boston, MA. Cabrelli Amaro, J. (2016) Testing the Phonological Permeability Hypothesis: L3 phonological effects on L1 versus L2 systems. International Journal of Bilingualism, 1367006916637287. Cabrelli Amaro, J. (2017). Regressive transfer in L3 syntactic development, in Hahn, A. & Angelovska, T. (eds), L3 Syntactic Transfer: Models, New Developments and Implications. Amsterdam: John Benjamins. Cabrelli Amaro, J., Amaro, J. F. & Rothman, J. (2015) The relationship between L3 transfer and structural similarity across development: Raising across an experiencer in Brazilian Portuguese, in Peukert, H. (ed), Transfer Effects in Multilingual Language Development. Amsterdam: Benjamins, 21-52. Cabrelli Amaro, J. & Rothman, J. (2010) On L3 acquisition and phonological permeability: A new test case for debates on the mental representation of non- native phonological systems. IRAL, International Review of Applied Linguistics in Language Teaching, 48, 275-296. Cabrelli Amaro, J. & Wrembel, M. (2016) Investigating the acquisition of phonology in a third language–a state of the science and an outlook for the future. International Journal of Multilingualism, 13(4), 395-409. Cenoz, J. (2003) The additive effect of bilingualism on third language acquisition: A review. International Journal of Bilingualism, 7(1), 71-87. Cenoz, J. (2013) The influence of bilingualism on third language acquisition: Focus on multilingualism. Language Teaching, 46(01), 71-86. Cenoz, J. & Valencia, J. F. (1994) Additive trilingualism: Evidence from the Basque country. Applied Psycholinguistics, 15(2), 195-207. Cheung, A. S., Matthews, S. & Tsang, W. L. (2011) Transfer from L3 German to L2 English in the domain of tense/aspect, in De Angelis, G. & Dewaele, J. M. (eds), New Trends in Crosslinguistic Influence and Multilingualism Research. Clevedon, UK: Multilingual Matters, 53-73. Child, M. (2017) The Typological Primacy Model and bilingual types, in Bellamy, K., Child, M., González, P., Muntendam, A. & Parafita Couto, M. C. (eds), Multidisciplinary Approaches to Bilingualism in the Hispanic and Lusophone World. Amsterdam: Benjamins, 41-64. Christiansen, M. H., Conway, C. M. & Onnis, L. (2012) Similar neural correlates for language and sequential learning: evidence from event-related brain potentials. Language and Cognitive Processes, 27(2), 231-256. Chumbow, B. S. (1981) The mother tongue hypothesis in a multilingual setting, Proceedings of the 5th Congress of the International Association of Applied Linguistics. De Angelis, G. (2007) Third or Additional Language Acquisition. Clevedon, UK: Multilingual Matters. Dewaele, J.-M. (1998) Lexical inventions: French as L2 versus L3. Applied Linguistics, 19(4), 471-490. Falk, Y. & Bardel, C. (2011) Object pronouns in German L3 syntax: Evidence for the L2 status factor. Second Language Research, 27, 59-82. Friederici, A. D., Steinhauer, K. & Pfeifer, E. (2002) Brain signatures of artificial language processing: Evidence challenging the critical period hypothesis. Proceedings of the National Academy of Sciences, 99(1), 529-534. 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. The International Journal of Multilingualism, 1, 3-16. Fouser, R. (1995) Problems and prospects in third language acquisition research. Language Research, 31(2), 387-414. García-Mayo, M. d. P. (2012) Cognitive approaches to L3 acquisition. International Journal of English Studies, 12(1), 129-146. García-Mayo, M. d. P. & Rothman, J. (2012) L3 morphosyntax in the generative tradition: The initial stages and beyond, in Cabrelli Amaro, J., Flynn, S. & Rothman, J. (eds), Third Language Acquisition in Adulthood. Amsterdam: Benjamins, 9-32. García Mayo, M. D. P. & Villarreal Olaizola, I. (2011) The development of suppletive and affixal tense and agreement morphemes in the L3 English of Basque-Spanish bilinguals. Second Language Research, 27(1), 129-149. Giancaspro, D., Halloran, B. & Iverson, M. (2015) Transfer at the initial stages of L3 Brazilian Portuguese: A look at three groups of English/Spanish bilinguals. Bilingualism: Language and Cognition, 19, 191-207. González Alonso, J. & Rothman, J. (2016) Coming of age in L3 initial stages transfer models: Deriving developmental predictions and looking towards the future. International Journal of Bilingualism, 1367006916649265. Goodenkauf, J., & Herschensohn, J. (2015). Gender feature transfer from L2 Spanish to L3 Arabic. Paper presented at the International Symposium on Bilingualism, New Brunswick, NJ. Hammarberg, B. (2001) Roles of 1 and L2 in L3 production and acquisition, in Cenoz, J., Hufeisen, B. & Jessner, U. (eds) Cross-linguistic Influence in Third Language Acquisition: Psycholinguistic Perspectives. Clevedon: Multilingual Matters, 21- 41. Hermas, A. (2010) Language acquisition as computational resetting: verb movement in L3 initial state. International Journal of Multilingualism, 7, 343-362. Hermas, A. (2014) Multilingual transfer: L1 morphosyntax in L3 English. International Journal of Language Studies, 8. Hermas, A. (2015) The categorization of the relative complementizer phrase in third- language English: A feature re-assembly account. International Journal of Bilingualism, 19(5), 587-607. Hui, B. (2010) Backward transfer from L3 French to L2 English production of relative clauses by L1 Cantonese speakers in Hong Kong. Hong Kong Journal of Applied Linguistics, 12(2), 45-60. Iverson, M. (2009) N-drop at the initial state of L3 Portuguese: Comparing simultaneous and additive bilinguals of English/Spanish, in Pires, A. & Rothman, J. (eds), Minimalist Inquiries into Child and Adult Language Acquisition: Case Studies Across Portuguese. Berlin: Mouton de Gruyter, 221-244. Jin, F. (2009) Third language acquisition of Norwegian objects: Interlanguage transfer or L1 influence?, in Leung, Y.-k. I. (ed), Third Language Acquisition and Universal Grammar. Bristol, UK: Multilingual Matters, 144-161. Kellerman, E. (1983) Now you see it, now you don’t, in Gass, S. & Selinker, L. (eds), in Language Learning. Rowley, MA: Newbury House, 112- 134. Kellerman, E. (1986) An eye for an eye: crosslinguistic constraints on the development of the L2 lexicon, in Kellerman, E. & Smith, M. S. (eds), Crosslinguistic Influence in Second Language Acquisition. New York: Pergamon, 35-48. Kopečková, R., Marecka, M., Wrembel, M. & Gut, U. (2016) Interactions between three phonological subsystems of young multilinguals: the influence of language status. International Journal of Multilingualism, 13(4), 426-443. Kulundary, V. & Gabriele, A. (2012) Examining the role of L2 syntactic development in L3 acquisition: A look at relative clauses, in Cabrelli Amaro, J., Flynn, S. & Rothman, J. (eds), Third Language Acquisition in Adulthood, 195-222. Llama, R., Cardoso, W. & L., C. (2010) The influence of language distance and language status on the acquisition of L3 phonology. International Journal of Multilingualism, 7, 39-57. Llama, R. & López-Morelos, L. P. (2016) VOT production by Spanish heritage speakers in a trilingual context. International Journal of Multilingualism, 13(4), 444-458. Llinàs-Grau, M. & Mayenco, E. P. (2016) Regressive Transfer from L4 German to L3 English: the Case of that-deletion, in Ibarrola-Armendariz, A. & Ortiz de Urbina Arruabarrena, J. (eds), Glancing Backwards To Build a Future in English Studies. Bilbao, Spain: University of Deusto, 281-288. Llisterri, J. & Poch-Olivé, D. (1987) Phonetic interference in bilingual’s learning of a third language, Proceedings of the XIth International Congress of Phonetic Sciences. Tallinn, Estonia: Academy of Sciences of the Estonian SSR, 137-147. Lloyd-Smith, A., Gyllstad, H., & Kupisch, T. (2017) Transfer into L3 English: Global accent in German-dominant heritage speakers of Turkish. Linguistic Approaches to Bilingualism, 7(2), 131-162. Lozano, C. (2002) The interpretation of overt and null pronouns in non-native Spanish, in Marsden, H., Pourcel, S. & Whong-Bharr, M. (eds), Durham Working Papers in Linguistics. Somerville, MA: Cascadilla Press, 53-66. Matthews, S. J., Cheung, S. C., & Tsang, W. L. (2014). Anti-transfer effects in third language acquisition. Paper presented at the International Conference on Third Language Acquisition and Multilingualism, Uppsala, Sweden. Meisel, J. (1983) Transfer as a second language strategy. Language and Communication, 3, 11-46. Mykhaylyk, R., Mitrofanova, N., Rodina, Y. & Westergaard, M. (2015) The Linguistic Proximity Model: The case of verb-second revisited, in Grillo, E. & Jepson, K. (eds) Proceedings of BUCLD 39. Somerville, MA: Cascadilla Press, 337-349. Montrul, S. (2004) The Acquisition of Spanish: Morphosyntactic Development in Monolingual and Bilingual L1 Acquisition and Adult L2 Acquisition. Amsterdam: Benjamins. Montrul, S. A. (2008). Incomplete Acquisition in Bilingualism: Re-examining the Age Factor. Amsterdam: Benjamins. Montrul, S., Dias, R. & Santos, H. (2011) Clitics and object expression in the L3 acquisition of Brazilian Portuguese: Structural similarity matters for transfer. Second Language Research, 27, 21-58. Onnis, L. (2012) Artificial language learning, in Robinson, P. J. (ed) Routledge Encycopedia of Second Language Acquisition. New York: Routledge, 35-41. Paradis, J. (2004) The relevance of specific language impairment in understanding the role of transfer in second language acquisition. Applied Psycholinguistics, 25, 67- 82. Rabinovitch, M.S. & Parver, L.M. (1966) Auditory discrimination in monolinguals and poliglots. Paper presented at the Meeting of the Canadian Psychological Association, Montreal, Canada. Ringbom, H. (1987) The Role of First Language in Foreign Language Acquisition. Clevedon, UK: Multilingual Matters. Ringbom, H. (2001) Lexical transfer in L3 production, in Cenoz, J., Hufeisen, B. & Jessner, U. (eds) Cross-linguistic Influence in Third Language Acquisition: Psycholinguistic Perspectives. Clevedon: Multilingual Matters, 59-68. Rivers, W. M. (1979) Learning a sixth language: An adult learner's daily diary. Canadian Modern Language Review. La Revue Canadienne des Langues Vivantes Welland, Ontario, 36(1), 67-82. Rothman, J. (2010) On the typological economy of syntactic transfer: Word order and relative clause high/low attachment preference in L3 Brazilian Portuguese. International Review of Applied Linguistics in Teaching (IRAL), 48(2/3), 245-273. Rothman, J. (2011) L3 Syntactic Transfer Selectivity and Typological Determinacy: The Typological Primacy Model. Second Language Research, 27, 107-127. Rothman, J. (2013) Cognitive Economy, Non-Redundancy and Typological Primacy in L3 Acquisition: Evidence from initial stages of L3 Romance, in Baauw, S., Dirjkoningen, F. & Pinto, M. (eds), Romance Languages and Linguistic Theory 2011. Amsterdam: John Benjamins, 217-248. Rothman, J. (2015) 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, 179-190. Rothman, J., Bañón, J. A. & Alonso, J. G. (2015) Neurolinguistic measures of typological effects in multilingual transfer: introducing an ERP methodology. Frontiers in Psychology, 6. Rothman, J. & Cabrelli Amaro, J. (2010) What variables condition syntactic transfer?: a Look at the L3 initial state. Second Language Research, 26, 198-218. Rothman, J., & Slabakova, R. (in press, 2017). The state of the science in generative SLA and its place in modern second language studies. Studies in Second Language Acquisition. Ruiz de Zarobe, Y. (2005) Age and third language production: A longitudinal study. International Journal of Multilingualism, 2(2), 105-112. Ruiz de Zarobe, Y. (2008) CLIL and foreign language learning: A longitudinal study in the Basque Country. International CLIL Research Journal, 1(1), 60-73. Sanchez, L. (2015) L2 activation and blending in third language acquisition: Evidence of crosslinguistic influence from the L2 in a longitudinal study on the acquisition of L3 English. Bilingualism: Language and Cognition, 18(2), 252-269. Sanz, C., Park, H. I. & Lado, B. (2015) A functional approach to cross-linguistic influence in ab initio L3 acquisition. Bilingualism: Language and Cognition, 18(2), 236-251. Schmid, M. S. (2016) First language attrition. Language Teaching, 49(2), 186-212. Schwartz, B. & Sprouse, R. (1996) L2 cognitive states and the Full Transfer/Full Access model. Second Language Research, 12, 40-72. Singh, R. & Carroll, S. (1979) L1, L2 and L3. Indian Journal of Applied Linguistics, 5, 51–63. Singleton, D. (1987) Mother and other tongue influence on learner French. Studies in Second Language Acquisition, 9(3), 327-345. Slabakova, R. & García Mayo, M. P. (2015) The L3 syntax-discourse interface. Bilingualism: Language and Cognition, 18, 208-226. Slabakova, R. (2016) The Scalpel Model of third language acquisition. International Journal of Bilingualism, 1367006916655413.

Stedje, A. (1977) Tredjespråksinterferens i fritt tal-en jämförande studie, in Palmberg, R. & Ringbom, H. (eds), Papers from the Conference on Contrastive Linguistics and Analysis. Abo: Abo Akademi, 141-158. Sypiańska, J. (2016) Multilingual acquisition of vowels in L1 Polish, L2 Danish and L3 English. International Journal of Multilingualism, 13(4), 476-495. Tsang, W. L. (2015) Acquisition of English number agreement: L1 Cantonese–L2 English–L3 French speakers versus L1 Cantonese–L2 English speakers. International Journal of Bilingualism, 1367006915576398. Vildomec, V. (1963) Multilingualism: General Linguistics and Psychology ofSspeech. Leiden, Netherlands: AW Sythoff. Vogel, T. (1992) Englisch und Deutsch gibt es immer Krieg': Sprachverarbeitungsprozesse beim Erwerb des Deutschen als Drittsprache. Zielsprache Deutsch, 23(2), 95-99. Westergaard, M., Mitrofanova, N., Mykhaylyk, R. & Rodina, Y. (2016) Crosslinguistic influence in the acquisition of a third language: The Linguistic Proximity Model. International Journal of Bilingualism, 1367006916648859. Wrembel, M. (2014) VOT patterns in the acquisition of L3 phonology. Concordia Working Papers in Applied Linguistics, 5, 750-770. Wrembel, M. (2015) In Search of a New Perspective: Cross-linguistic Influence in the Acquisition of Third Language Phonology. Poznan: Wydawnictwo Naukowe UAM.

View publication stats