LSA abstracts (all years)

2021/09/20

LSA 2021 (poster): Linguistic transfer, or there and back again: A chronological study of terminological meandering

Keywords: transfer, second language acquisition, third language acquisition, contrastive analysis, multilingualism

The notion of transfer has remained impervious in second language acquisition ever since it “transferred” from experimental psychology in the late 1940s (cf. references in Lado 1957). Similarly, transfer has been crucial to L3 acquisition, since the field sets out to ascertain the influence of previous languages on the one being learnt (sec. Slabakova 2018). With two major subfields of linguistics working with transfer, one would expect it to be a clear and well-defined term. Yet, even a superficial survey of recent literature exposes various debates and uncertainties which imbower the definition of transfer (cf. Westergaard 2019, Schwartz and Sprouse 2021, and commentaries on both keynotes). In order to understand what transfer is and how it came to be what it is, we set out to trace the origins and development of transfer: from James (1890) to the modern day. Notably, this paper is the first one to undertake such a holistic evaluation of terminology spanning over 130 years, including the research in psychology where transfer came from in the first place. We consider it essential that to investigate phylogeny, the paper should be holistic in this way.

The analysis of literature we performed divided the history of transfer into five main stages: transfer in philosophy (up to 1890), transfer in psychology (1890-1948), R. Lado’s transfer (1948-1957), uncertain transfer (1957-1990), all of which led to the modern-day transfer (1990s-present). During the first stage, transfer came up in philosophical literature as an abstract term (e.g., “transfer of judgement from one proposition to another”). Transfer in psychology refers to the period when transfer emerged as an applied notion: “the influence of prior learning […] upon the learning of, or response to, new material” (McGeoch 1942:394). McGeoch was the first one to actually define transfer, despite dozens of works on transfer between 1910s-1930s. The transfer in McGeoch’s definition was welcomed to linguistics by Robert Lado (1957) and was used pervasively in his work. Our analysis shows that Lado collated under the same term both McGeoch’s (often conscious) psychological transfer (e.g., “reading habits”, Lado 1957:94) and transfer of linguistic structure (case and gender morphology, Lado 1957:58). We establish that this kluge of definitions in an influential framework of contrastive analysis was the primary cause of much of the confusion in the next stage of the uncertain transfer.

During much of the second half of the 20th century, SLA inquiry centered around transfer (notably Gass and Selinker 1983): what constitutes it and what causes it? Various papers tried to draw distinctions between transfer and adjacent notions, such as interference – but none of those distinctions survived to the present. It was as early as in the 1970s that acquisitionists admitted that “transfer can mean anything to anyone” (Kellerman 1977, hence uncertain transfer). Upon our careful dissection of attempts to formalize or define transfer in papers between 1957 and 1990, we ineluctably find Lado’s traces in every one of them. Our analysis leads us to the modern-day studies which are still trying to find out what transfer is and find the line between transfer and e.g., “cross-language influence” (Rothman et al. 2019). Trivially, however, it is impossible to separate notion A from notion B if one of them is not defined at all or ill-defined. We thus conclude our analysis.

As a footnote to the main analysis, we also outline some suggestions for moving forward in transfer studies. The main suggestion is to do away with transfer completely, liberating the field from the endowment of misinterpretations and abstrusity which transfer carries with it. To achieve that, we briefly outline a draft of a proposal to look back to James (1890) and supplant the notion of transfer with that of merge, to avoid treating languages as entirely separate entities. We speculate that such a proposal should be based on theories of language universals and emergentist approaches to parametric theory. The merge approach also resolves modern debates on “copying” being a part of transfer and offers a venue for easy formalizations, essential for generative inquiry.

We sincerely hope that our investigation of 120+ papers will be insightful for both experienced acquisitionists and the newcomers, lending itself to didactical use: both as a warning to handle terminology carefully, and as a unique “transfer” lens to observe the development of the SLA. It is this didactical purpose coupled with the tumultuous personal history of transfer that leads us presently to draw attention to the story of linguistic transfer, or there and back again.

References

  1. Gass, Susan M., and Larry Selinker. “Language Transfer in Language Learning.” Lald.5, John Benjamins Publishing Company, 1983.
  2. James, William. The Principles of Psychology: Volume II. Henry Holt and Company, 1890, pp. vi, 704. APA PsycNET.
  3. Kellerman, Eric. “Towards a Characterisation of the Strategy of Transfer in Second Language Learning.” Interlanguage Studies Bulletin, vol. 2, no. 1, Sage Publications, Ltd., 1977, pp. 58–145.
  4. Lado, Robert. Linguistics across Cultures : Applied Linguistics for Language Teachers. University of Michigan Press, 1957.
  5. McGeoch, J. A. The Psychology of Human Learning: An Introduction. Longmans, 1942, pp. xvii, 633.
  6. Rothman, Jason, et al. Third Language Acquisition and Linguistic Transfer. Cambridge University Press, 2019. CUP.
  7. Schwartz, Bonnie D., and Rex A. Sprouse. “The Full Transfer/Full Access Model and L3 Cognitive States.” Linguistic Approaches to Bilingualism, vol. 11, no. 1, John Benjamins, Mar. 2021, pp. 1–29. www.jbe-platform.com, doi:10.1075/lab.20055.sch.
  8. Slabakova, Roumyana. “The Scalpel Model of Third Language Acquisition.” International Journal of Bilingualism, vol. 21, no. 6, SAGE Publications Sage UK: London, England, 2017, pp. 651–65.
  9. Westergaard, Marit. “Microvariation in Multilingual Situations: The Importance of Property-by-Property Acquisition.” Second Language Research, SAGE Publications Ltd, Nov. 2019. SAGE Journals.

Comments From Reviewers

Reviewer 1:
I’m pretty skeptical about the goals here. Just because something has meant a lot of things over the years (particularly a term that started as and continues to be an innocuous English word) doesn’t mean that it’s problematic. I also suspect that replacing it with an approach called “merge” - itself a widely used and multiply-defined word in linguistics - won’t improve matters. That said, the LSA seems like a good venue for work on the history of the field.

Reviewer 2:
This abstract examines the importance of linguistic transfer, a term that is still not well-defined, according to the author(s). It gives a historical overview of how the term has been used over the last 130 years and examines how the definition of transfer has changed or evolved over time.
Though the historical overview might be interesting, and I find the last part of the abstract (the suggestions moving forward) useful, I believe the contributions that this research makes overall are not spelled out. Additionally, it is unclear how it helps researchers currently working on transfer to get a clearer idea of what the term represents/denotes.

Reviewer 3:
I am always happy to see papers which help the field out of the ‘terminological weeds’. ‘Transfer’ has been one such topic that umbrellas across L1, L2 and L3 and thus i am happy to see the authors take it on. Here is a strong suggestion: i sincerely hope that in their comprehensive lit review, the authors take on not spoken languages but also sign languages, and only unimodal bilingual cases but also bimodal bilingual cases. To the degree that these areas remain untouched, ‘transfer’ is a topic is not covered comprehensively.

LSA 2021 (paper): L1 vs. L2 vs. L3 transfer: grammatical gender and determiner acquisition in sequential quadrilinguals

Keywords: fourth language acquisition; L4; grammatical gender; French; English, Ukrainian, Russian

While research in acquisition of third language (L3A) flourishes, the acquisition of the fourth (L4A) remains to be underinvestigated. It, nonetheless, could alleviate the tension between L3A theories. After a persuasive rejection of the wholetransfer models for L3A by Slabakova (2017), there is still a need for further scrutiny of partial-transfer models: cumulative-enhancement model (Flynn et al., 2004), linguistic proximity model (Westergaard et al., 2017), and the Scalpel Model (Slabakova, 2017). The goal of the current study is to test the prediction that all previous grammars are all equally available for transfer while constructing the grammar of a fourth language. The study is pioneering in looking at three languages as a potential source of transfer; it is also novel in looking at typologically proximate L1 and L2.

The experiment to test the hypothesis engaged three groups of adolescent quadrilinguals (L1 Russian - L2 Ukrainian - L3 English - L4 French), divided by their mastery of the 4th language (7 Beginners, 7 Pre-Intermediates, and 8 Intermediates). With their levels being confirmed by PET, the reported standardized grades, as well as self-reporting, the participants completed a “yes-no” grammaticality judgment task which included (a) sentences with gender mismatch and b) sentences with a missing article which indicated L3 transfer. Distraction stimuli were also included. (a) FRA *La[F] nom[M] est très long[M] vs. ENG The name is very long.
(b) FRA *Lit est un meuble que nous utilisons.
vs. * Bed is an element of furniture which we use.

The offered stimuli tested transfer of grammatical gender from L1, from L2, L1=L2 (matching genders), and transfer of determiners from L3. The data was analyzed using MANOVA, ANOVAs, and MANCOVA to estimate the effects of transfer from L1, from L2, mixed L1-L2 transfer, transfer from L3 and their interaction on the correctness of judgment with respect to students’ level. All tests returned significant within p=.001.

The results confirmed the hypothesis, showing transfer from all L1-L3. Thus, the study counters the claim that any language offers a basis for a full transfer of grammatical categories to Ln (whole-transfer models), and further supports the partial-transfer models which claim that the transfer occurs property by property from any of the previous languages. However, with regard to the latest developments (LPM, CEM, and Scalpel model), the study, extending their claims to L4A, reveals a number of novel questions. Namely, the study scrutinizes the notion of structural/linguistic proximity so widely used in both LPM and Scalpel models, urging further elucidation in measuring this proximity, and asking what are the margins. There is, beyond any doubt, a precise mechanism guiding how any given property is derived from the UG for the L4A even when choosing between typologically proximate languages like Russian and Ukrainian. This mechanism is unknown so far and it is exigent that it be investigated. With Slabakova (2017) suggesting that “the search for the definitive L3 acquisition account continues,” it is time to introduce the search for the definite L4 acquisition account which will help with the former.

References

  1. Flynn, Suzanne, Claire Foley & Inna Vinnitskaya. 2004. The Cumulative-Enhancement Model for language acquisition: Comparing adults' and children’s patterns of development in 342 first, second and third language acquisition of relative clauses. International Journal of Multilingualism 1. 3–16. http://doi.org/10.1080/14790710408668175.
  2. Slabakova, Roumyana. “The scalpel model of third language acquisition.” International Journal of Bilingualism 21.6 (2017): 651-665.
  3. Westergaard, Marit, et al. “Crosslinguistic influence in the acquisition of a third language: The Linguistic Proximity Model.” International Journal of Bilingualism 21.6 (2017): 666-682.

Comments from reviewers:

Reviewer 1:
The abstract clearly states its objective in relation to past works in language acquisition. In addition, it articulates its contribution to the field.

Reviewer 2:
This study is novel in that it focuses on fourth language acquisition and looks at three languages as a source of transfer. The researcher learned that transfer does indeed occur from the L1, L2 and L3. Although much research has been done on L1-L2 transfer, studies rarely, if ever, pay attention to the role that L4 transfer plays.The small sample size (i.e. 22 participants) is, however, a limitation that would hinder generalization of the findings. However, on the whole much can be provided to the field of linguistics based on the work that the researcher has done.

Reviewer 3:

  1. How were the order of non-L1 determined? That is, how do you know French is the L4 for any individual child, as opposed to English?
  2. How are the gender systems in these languages different? I know they are, but how are they different in a way that is relevant to this study?
  3. I have no idea how the data (whatever that looks like) actually show that the authors purport them to show. How do you go from yes-no judgments on these items to showing what has transferred and what has not? I’m sure there is a way, but it is not explained.
  4. How is adding a fourth language any different from looking at 3, except in the number of languages? I don’t see how the same finding could not be found using just 3 languages (the right set of 3). In other words, what is the added benefit to quadrilingualism over trilingualism? That seems unmotivated.
  5. Having said all that, it’s pretty marvelous to be looking at quadrilingual populations.