Brief articleRevisiting the bilingual lexical deficit: The impact of age of acquisition
Introduction
Decades of inquiry on bilingualism within psychology, linguistics, and cognitive neuroscience have established that the frequent use of two (or more) languages exerts an influence on both cognitive and linguistic abilities (for recent overviews, see Bialystok, 2017, Kroll et al., 2014). On the one hand, bilingualism seems to afford a cognitive advantage: compared to individuals who only use one language, bilinguals exhibit enhanced levels of executive control, conflict resolution, and protection against early age-related cognitive decline (but see Lehtonen et al., 2018). However, on the other hand, bilingualism produces a so-called lexical “deficit”: the same studies that document the cognitive advantage also routinely report that adult bilinguals exhibit smaller vocabularies in each language compared to monolingual speakers, as seen, for instance, in picture naming tasks. Moreover, bilingual speakers take somewhat longer than monolinguals to name objects in the same tasks, and to recognize words in lexical decision tasks. It has been suggested that these two effects are epiphenomena of the joint activation of the two language systems of the bilingual, which creates a conflict of selecting the context-appropriate linguistic forms and inhibiting the non-target language (for recent treatments, see De Baene et al., 2015, Luk et al., 2011). This process thus functions as a booster of executive control (giving rise to the cognitive advantage), while at the same time compromising lexical representation and processing speed (giving rise to the lexical deficit).
Recent years have seen an increasing number of discussions on the importance of differentiating between different types of bilinguals in order to better understand the effects of bilingualism on language and cognition (Bialystok, 2016, Bialystok, 2017, Kroll et al., 2014, Luk and Bialystok, 2013). So far, however, this debate has mainly concerned the effects on cognition, largely leaving aside the possibility that different types of bilingualism may also yield different effects on lexical behaviour. The current study addresses this gap. Here, we ask whether the lexical deficit is really an effect of bilingualism alone: a potentially serious problem with interpreting the findings to date on the lexical deficit is that existing studies do not always provide sufficient information on the language acquisition trajectories of their participants, or do not differentiate between individuals who acquired two languages from birth (i.e., simultaneous bilinguals) from individuals who acquired one language from birth and a second language (L2) after that (i.e., sequential bilinguals).1 Evidence from studies on L2 acquisition suggests that such practice is problematic because, first, even speakers with early ages of L2 acquisition (AoA) do not necessarily obtain nativelike L2 proficiency (e.g., Abrahamsson and Hyltenstam, 2009, Nishikawa, 2014, Norrman and Bylund, 2016, Sebastián-Gallés et al., 2005), and moreover, simultaneous bilinguals do not necessarily differ from monolingual native speakers in terms of their proficiency with the dominant language (e.g., Kupisch, 2012, Kupisch et al., 2013). This raises the question as to whether the by now well-known bilingual lexical deficit is indeed an effect of bilingualism alone, or whether there is a potential confound of age of acquisition. Should age of acquisition indeed turn out to play a role here, it could have far-reaching consequences for our understanding of the linguistic costs of bilingualism.
The present study implements a unique methodological design to investigate this question. We test vocabulary knowledge and lexical processing in an unprecedented constellation of speakers where the variables of monolingualism vs. bilingualism and L1 vs. L2 have been fully crossed (Table 1). In this design, simultaneous and sequential bilinguals are thoroughly differentiated, and, crucially, L2 status is dissociated from bilingualism. This eliminates the limitations of previous research where L2 status has been an ever-present confound, and robustly assesses to what extent the lexical deficit is an effect of bilingualism alone.
If bilingualism alone is responsible for producing the lexical deficit, this should be manifested in the current design as a standalone main effect of the factor of bilingualism, whereas the opposite would hold should AoA be the driving factor behind said deficit. However, more nuanced outcomes may also be attested, manifested as an interaction between, or a confluence of, bilingualism and AoA.
In keeping with standard psycholinguistic practice we use the notions L1 and L2 strictly to refer to order of acquisition, regardless of language dominance. The term ‘L1 bilingual’ will be used to describe a person who acquired two languages from birth and uses them on a regular basis (i.e. simultaneous bilingual); ‘L2 bilingual’ refers to a person who learnt a L2 after the onset of L1 acquisition (even if that L2 is learnt in early childhood), and uses both languages on a regular basis (i.e. sequential bilingual); ‘L1 monolingual’ is a person who acquired one language from birth, possibly has some foreign language skills, but uses only the L1 for communication; ‘L2 monolingual’ refers to an individual who at one point in life stopped using his/her L1, lost proficiency in it, acquired an L2, and uses only the L2 for communication (while possibly possessing some foreign language skills). International adoptees are often L2 monolinguals.
Finally, following previous studies on the lexical deficit, the current study assesses lexical knowledge and processing in the societally dominant language of the participants’ residential context (in this case, Swedish in Sweden).
Section snippets
Participants
Eighty adult speakers living in Sweden participated in the study, distributed equally as per the following groups:
L1 monolinguals: These participants (Mage = 29.8) had grown up in Sweden, and had acquired Swedish from birth as only language by native-speaking Swedish parents.
L1 bilinguals: The participants in this group (Mage = 32.2) had one Spanish-speaking parent and one Swedish-speaking parent, and had acquired both these languages from birth. They were fluent in both Swedish and Spanish and
Picture naming
For picture naming accuracy, adding AoA to an intercept model increased the fit significantly, χ2 = 8.663, p = .003, as did adding AoA to a model including bilingualism, χ2 = 9.165, p = .002. Adding bilingualism, however, did not increase the fit, for neither an intercept model nor an AoA model (ps > .10). In no instance did adding an interaction increase the fit (ps > .10). Consequently, only AoA was included in the final model, yielding a significant effect on naming accuracy, b = 0.424, SE
Discussion
The current findings further the results from previous research on the lexical deficit in undifferentiated L1/L2 bilinguals, showing that bilingualism may exert an effect on lexical accuracy and processing even when AoA is taken into account. Crucially, however, the findings reveal a hitherto overlooked effect of AoA in the lexical deficit literature. For lexical knowledge (accuracy on the naming test and lexical decision task), AoA was found to heavily influence performance (more consistently
Conclusion
Taken together, the findings reported in the present study show that recent arguments about bilingualism not being a categorical variable must be taken seriously not only when interpreting the cognitive changes brought about by the use of multiple languages, but also in linguistic behaviour. Progress in the area of bilingual lexical development will require a firmer marriage between age of acquisition studies and bilingualism studies (Li, 2009), in order to understand the influence of
Acknowledgements
We are thankful to the three anonymous reviewers for providing insightful comments on a previous version of the manuscript. This research was funded by Riksbankens Jubileumsfond (The Bank of Sweden Tercentenary Foundation), grant no. M2005-0459 (to K.H.) and grant no. SAB16-0051:1 (to N.A.).
References (53)
- et al.
Mixed-effects modeling with crossed random effects for subjects and items
Journal of Memory and Language
(2008) - et al.
Random effects structure for confirmatory hypothesis testing: Keep it maximal
Journal of Memory and Language
(2013) - et al.
Lexical access in bilinguals: Effects of vocabulary size and executive control
Journal of Neurolinguistics
(2008) - et al.
The emergence of competing modules in bilingualism
Trends in Cognitive Sciences
(2005) - et al.
Critical period effects in second language learning: The influence of maturational state on the acquisition of English as a second language
Cognitive Psychology
(1989) - et al.
Identifying the causal link: Two approaches toward understanding the relationship between bilingualism and cognitive control
Cortex
(2015) - et al.
Vocabulary and verbal fluency of bilingual and monolingual college students
Archives of Clinical Neuropsychology
(2007) - et al.
The influence of initial exposure on lexical representation: Comparing early and simultaneous bilinguals
Journal of Memory and Language
(2005) - et al.
Becoming a balanced, proficient bilingual: Predictions from age of acquisition & genetic background
Journal of Neurolinguistics
(2018) - et al.
The loss of first language phonetic perception in adopted Koreans
Journal of Neurolinguistics
(2004)
The role of age of acquisition on past tense generation in Spanish-English bilinguals: An fMRI study
Brain and Language
Age of onset and nativelikeness in a second language: Listener perception versus linguistic scrutiny
Language Learning
Analyzing reaction times
International Journal of Psychological Research
Fitting linear mixed-effects models using lme4
Journal of Statistical Software
The signal and the noise: Finding the pattern in human behavior
Linguistic Approaches to Bilingualism
The bilingual adaptation: How minds accommodate experience
Psychological Bulletin
Preserved implicit knowledge of a forgotten childhood language
Psychological Science
Cascading and multisensory influences on speech perception development
Mind, Brain, and Education
Brain circuit for cognitive control is shared by task and language switching
Journal of Cognitive Neuroscience
The bilingual brain
Age of acquisition: Its neural and computational mechanisms
Psychological Bulletin
Dominant-language replacement: The case of international adoptees
Bilingualism: Language and Cognition
Boston naming test
Two languages in mind: Bilingualism as a tool to investigate language, cognition, and the brain
Current Directions in Psychological Science
Cited by (29)
The CPH is dead. Long live the critical period hypothesis
2023, Brain and LanguageClinical Assessment and Intervention With Persons Living With Dementia
2022, Comprehensive Clinical Psychology, Second EditionL1 activation during L2 processing is modulated by both age of acquisition and proficiency
2021, Journal of NeurolinguisticsCitation Excerpt :The existing evidence on L2 learning indeed suggests that the lexicon is not exempt from AoA effects.1 In a recent study, Bylund, Abrahamsson, Hyltenstam, and Norrman (2019) investigated lexical retrieval in adult speakers with varying bilingual backgrounds. Instead of assessing L2 learners of varying AoA ranges, this study implemented a design in which the variables AoA and bilingualism/monolingualism were fully crossed, yielding four participant groups: monolingual L1 speakers, simultaneous bilinguals, monolingual L2 speakers (international adoptees) and successive childhood bilinguals.
How words ripple through bilingual hands: Motor-language coupling during L1 and L2 writing
2020, NeuropsychologiaCitation Excerpt :These variables are well-known modulators of various language processes in bilinguals, including action-related sensorimotor resonance (Birba et al., 2020). Typically, lower language-learning ages involve more parallel co-activation of both languages (Canseco-Gonzalez et al., 2010), increased inhibition of the non-target language (Bylund et al., 2019), and differential reliance on lexical and conceptual routes (Sabourin et al., 2014), among other effects. Likewise, higher L2 competence is associated with more efficient lexical access (Kastenbaum et al., 2018) and reduced amplitudes of the N400 and the late positive complex during text processing (Yang et al., 2018).
Motor-system dynamics during naturalistic reading of action narratives in first and second language
2020, NeuroImageCitation Excerpt :For instance, higher proficiency levels are associated with greater reliance on direct (as opposed to L1-mediated) word-concept mappings in L2 (Guasch et al., 2008; Sunderman and Kroll, 2006), less asymmetric processing when comparing L1 and L2 tasks (García, 2015a; García et al., 2014), and more convergent recruitment of neural resources in both languages (Abutalebi, 2008). Similarly, a lower age of L2 learning has been linked to more parallel co-activation of both languages (Canseco-Gonzalez et al., 2010), increased inhibition of the non-target language (Bylund et al., 2019), greater neural sensitivity to fine-grained semantic distinctions in L2 (Vilas et al., 2019), and less marked differences in brain activation between L2 and L1 (Liu and Cao, 2016). More particularly, neurocognitive embodied effects during L2 action-language processing have been shown to correlate with L2 proficiency (Bergen et al., 2010) or even to be present only in high- as opposed to low-proficiency subjects (Ibáñez et al., 2010; Vukovic, 2013).