On meronomy and dimensionality of the models for multilingual language acquisition


UNC Spring Colloquium 2022

Since Flynn et al. (2004), the number of studies in the field of third (and multilingual) language acquisition has soared and over 10 models have been proposed to account for the acquisition process (or parts of it) in trilinguals (broadly conceived). A number of studies attempted to adjudicate between those models and yielded highly divisive results (cf. Slabakova 2017). However, most research to date has treated these models as situated equally within the generative tradition. Models have been interpreted as aiming to describe the same process and identical requirements have been posed to them. Our aim is to show that such an assumption is unwarranted. We attempt to separate strictly generative models from models which actually are only broadly generative, or even functionalist, usage-based. We also aim to show that, to the extent that models have different purposes, sets of requirements for models are different. We review the existing models of multilingual language acquisition and propose a respective taxonomy framework (Table 1 below) based on the approaches models adhere to. Below, the classes of this taxonomy are delineated and justified.

Firstly, we distinguish strictly generative models of acquisition. The key property of these models is adherence to the narrowly Chomskian view of language, and thus reliance on the notion of UG and formal syntax (rarely semantics or phonology). Situated within the Chomskian framework, these models view social factors or artificial notion of typology as holding no explanatory power. The second class is composed of models which rely on extra-linguistic research in the sense that “neurocognitive and psychological foundation of the[se] model[s] is the emerging functional picture of the multilingual brain” (Slabakova 2017:653). They do not rely strictly on rationalistic generative tradition and accept evidence from other fields, notably neurobiology – and equally much, aim to make predictions within those fields.

The third class are usage-based and functionalist models. These models rely on social factors (dominant language of community) or metalinguistic knowledge to characterise the acquisition process. Trivially, these models cannot be tested within the generative framework. At the same time, they stand separate from extralinguistic models per being partially behavioristic/empiricist.

The biggest class we distinguish is broadly “generative-compliant” models. These models test their predictions within the broad generative conception of language, yet often use extra-generative notions at their core. The acquisition process within these models can be characterised by deficiency (negative transfer), typology, “inaccessibility” to UG, etc. – which is impossible in strict generativity. The key component of these models is that they do not ground their theorising in rationalistic generative tradition, but instead merely use formal generative (often parametric) accounts of cross-linguistic variation as one of the means to test their predictions. They should not be held to the same requirements (of keeping with Chomskian economy considerations, etc.) as the models of the first class. The last class is domain-specific models which do not make claims about acquisition process generally, but instead hypothesise about one particular domain of it, in our case phonology.

Concluding, it is important to acknowledge that the holistic account of multilingual acquisitions must inevitably come from a meronomic dimension: it is a cumulative achievement of a wide variety of models adhering to different traditions. Naturally, findings across fields ought to be used to complement each other, and models would benefit from explaining a wide variety of phenomena. However, we hope to have shown that the existing models differ in a significant variety of ways and ought not be held against the same set of requirements. We hope that the proposed frameworks will offer a better vantage point for future literature overviews as well as draw finer boundaries between models which could serve to make sure any given experimentation is well-suited to assess a particular class of models.

Strictly generative models of acquisition

Cumulative-Enhancement model: Flynn et al (2004), 

Micro-cue model: Westergaard (2021),

Grammatical Mapping for L3: Fendandez-Berkes and Flynn (2021)

Extralinguistic models

Scalpel Model: Slabakova (2016)

Usage-based and functionalist models

L2 Status model: Bardel and Falk (2007)

L1 Privilege model: Hermas (2014)

Language of Community: Fallah et al. (2016)

Broadly “generative-compliant” models

Typological Proximity Model: Rothman (2010) et seq.

Linguistic Proximity Model: Mykhaylyk et al. (2016) 

Interlanguage Transfer Hypothesis: Leung (2007)

Full Transfer Full Access to L3: Schwartz and Sprouse (2021)

Domain-specific hypotheses

Phonological Permeability Hypothesis Amaro and Rothman (2010)

Similarity Convergence Hypothesis: Brown and Chang (2022)


  1. Amaro, Jennifer Cabrelli, and Jason Rothman. 2010. “On L3 Acquisition and Phonological Permeability: A New Test Case for Debates on the Mental Representation of Non-Native Phonological Systems.” International Review of Applied Linguistics in Language Teaching 48 (2–3): 275–96.
  2. Bardel, Camilla, and Ylva Falk. 2007. “The Role of the Second Language in Third Language Acquisition: The Case of Germanic Syntax.” Second Language Research 23 (4): 459–84.
  3. Brown, Megan M., Chang, Charles B. 2022. “Regressive Cross-Linguistic Influence in Multilingual Speech Rhythm: The Primacy of Typological Similarity.” 96th Annual Meeting of the Linguistic Society of America.
  4. Fallah, Nader, Ali Akbar Jabbari, and Ali Mohammad Fazilatfar. 2016. “Source(s) of Syntactic Cross-Linguistic Influence (CLI): The Case of L3 Acquisition of English Possessives by Mazandarani–Persian Bilinguals.” Second Language Research 32 (2): 225–45.
  5. Flynn, Suzanne, Claire Foley, and Inna Vinnitskaya. 2004. “The Cumulative-Enhancement Model for Language Acquisition: Comparing Adults’ and Children’s Patterns of Development in First, Second and Third Language Acquisition of Relative Clauses.” International Journal of Multilingualism 1 (1): 3–16.
  6. Flynn, Suzanne, Fernandez-Berkez, Eva. 2021. “Grammatical Mapping in L3 Acquisition: A Theory of Development”. L3 After the Initial State Workshop, October 2021.
  7. Hermas, Abdelkader. 2010. “Language Acquisition as Computational Resetting: Verb Movement in L3 Initial State.” International Journal of Multilingualism 7 (4): 343–62.
  8. Leung, Yan-kit Ingrid. 2007. “Third Language Acquisition: Why It Is Interesting to Generative Linguists.” Second Language Research 23 (1): 95–114.
  9. Mykhaylyk, Roksolana, Natalia Mitrofanova, Yulia Rodina, and Marit Westergaard. 2015. “The Linguistic Proximity Model: The Case of Verb-Second Revisited.” In Proceedings of the 39st Annual Boston University Conference on Language Development, 337–49. Boston, MA: Cascadilla Proceedings Project.
  10. Rothman, Jason. 2011. “L3 Syntactic Transfer Selectivity and Typological Determinacy: The Typological Primacy Model.” Second Language Research 27 (1): 107–27.
  11. Schwartz, Bonnie D., and Rex A. Sprouse. 2021. “The Full Transfer/Full Access Model and L3 Cognitive States.” Linguistic Approaches to Bilingualism 11 (1): 1–29.
  12. Slabakova, Roumyana. 2017. “The Scalpel Model of Third Language Acquisition.” International Journal of Bilingualism 21 (6): 651–65.
  13. Westergaard, Marit. 2021. “Microvariation in Multilingual Situations: The Importance of Property-by-Property Acquisition.” Second Language Research 37 (3): 379–407.