Lexical access in bilinguals: Effects of vocabulary size and executive control
Introduction
Adult bilinguals who are highly proficient in both languages often reveal a deficit relative to monolinguals in tasks assessing aspects of linguistic processing (review in Gollan & Kroll, 2001; Michael & Gollan, 2005). These processing disadvantages have been found in a variety of psycholinguistic tasks and across groups of bilinguals who vary in their languages, age, and bilingual status, but are especially apparent in tasks based on rapid lexical retrieval. Producing words in a second language takes longer than producing the comparable word in a first or stronger language (Chen & Leung, 1989), but even well-balanced bilinguals responding in their stronger language often perform more poorly than monolinguals in that language (e.g., Gollan, Fennema-Notestine, Montoya, & Jernigan, 2007). There are various explanations for this deficiency in lexical retrieval, such as the inability to inhibit interference from the other language (Hermans, Bongaerts, de Bot, & Schreuder, 1998), and the possibility that each language system in a bilingual is supported by “weaker links” connecting the concept to the word than is the single language of a monolingual (Gollan & Acenas, 2004). The problem of how words are accessed by bilinguals with an evaluation of various explanations for these processes is discussed in detail by Costa (2005).
In addition to less efficient lexical retrieval, research with children has consistently found that bilinguals have smaller vocabularies in each language than comparable monolinguals (reviews in Bialystok, 2001; Oller & Eilers, 2002). In a summary analysis of 963 children (421 monolingual, 542 bilingual) between the ages of 5 and 9 years old, the monolinguals obtained higher scores than the bilinguals on the Peabody picture vocabulary test (PPVT) at every age (Bialystok, 2006). Studies comparing adult monolinguals and bilinguals rarely include formal assessments of vocabulary, but those that do often report a bilingual deficit (Craik & Bialystok, 2006; Fernandes, Craik, Bialystok, & Kreuger, 2007), although an equivalent vocabulary level has also been found (Bialystok, Craik, Klein, & Viswanathan, 2004).
The two disadvantages observed in bilinguals, namely, weaker lexical retrieval and lower vocabulary scores, may be related. It seems likely, that is, that initial differences in vocabulary level will affect performance on lexical-based tasks. However, none of the explanations for the bilingual deficit in lexical tasks takes account of the possibility that vocabulary size for the two groups may be different and considers its role in performance.
The bilingual deficit in lexical retrieval is well documented and usually signaled by longer response times than monolinguals. Randsell and Fischler (1987), for example, showed that monolinguals and bilinguals were equally accurate on a lexical decision task, but the bilinguals were slower to respond, although Soares and Grosjean (1984) reported no difference in either accuracy or reaction time for monolinguals and bilinguals in a similar task. Compromised verbal processing has also been reported in reading: Highly proficient bilinguals read more slowly in the second language than in the first language, in spite of comparable oral facility in the two languages (Favreau & Segalowitz (1983), Favreau & Segalowitz (1982)). Segalowitz and colleagues (Segalowitz, 1986; Segalowitz & Hulstijn, 2005) have identified a number of reasons for these differences, including less automaticity of word recognition and weaker skills in orthographic processing in the second language. The consequence is that these bilinguals find reading more effortful, in a language for which they have comparable oral proficiency, to monolinguals.
Object naming and word retrieval are also slower in bilinguals. In research with children who emigrated from Germany to settle in Sweden and developed Swedish proficiency in school, Mägiste (1992), Mägiste (1979) reported slower response times for these bilingual children than for their monolingual peers in naming objects, words, and numbers. Research with adults has replicated the bilingual disadvantage in naming, mostly by using picture naming tasks in which bilinguals commit more errors than monolinguals (Gollan, Montoya, Fennema-Notestine, & Morris, 2005; Roberts, Garcia, Desrochers, & Hernandez, 2002). A study by Costa, Santesteban, and Ivanova (2006) paradoxically showed that bilinguals were faster to name pictured objects in their second language than in their dominant first language, a result they attributed to setting different threshold levels of activation for responding in each of the two languages (Costa & Santesteban, 2004). This somewhat counterintuitive result is nonetheless consistent with proposals by Green (1998), for the role of inhibition and attention in language selection for bilinguals, and with empirical evidence by Meuter and Allport (1999) demonstrating more difficulty for bilinguals to switch into their dominant language than into their weaker one. The overall picture is one in which knowledge of two languages compromises performance in a simple verbal task in one of those languages.
Picture naming is another method of eliciting lexical responses. A standardized measure of this ability is the Boston Naming Task (Kaplan, Goodglass, & Weintraub, 1983), and a study by Kohnert, Hernandez, and Bates (1998) established norms for a population of English–Spanish bilinguals. These researchers found overall higher performance in English and greater variability in Spanish, with subsets of participants displaying different patterns of language dominance. Subsequent studies comparing monolinguals and bilinguals on this task have corroborated this pattern and reported reduced naming ability by bilinguals (Roberts et al., 2002). In addition to differences in accuracy there are also differences in latency, showing that bilinguals take longer than monolinguals to name the pictures. Gollan et al. (2005) attributed these slower times to the additional time required to retrieve the name of the target item rather than to extra time needed to construct a semantic interpretation of the picture. In their study, monolinguals were faster than bilinguals in naming a set of pictures, replicating the earlier findings, but both groups responded at the same rate to a classification task in which the pictures were categorized as natural or human made.
Finally, some studies have compared monolinguals and bilinguals on time-limited verbal fluency tasks in which participants must generate as many words as possible that conform to a constraint given by an initial letter or category membership (e.g., Borkowski, Benton, & Spreen, 1967). These fluency tasks are normally used as neuropsychological measures of cognitive functioning, and performance can be interpreted diagnostically to indicate changes in cognitive function with age as well as with abnormal patterns of cognitive behavior. Participants are given 60 s to generate as many examples as possible of words that either begin with a particular letter (usually F, A, S) or are members of a particular semantic category (e.g., animals, fruits). Because of the importance of these fluency measures in neuropsychological assessment, it would be important to document persistent differences that might follow from language experience. Bilingual deficits have been reported using both the letter and category tests (Gollan, Montoya, & Werner, 2002; Rosselli et al. (2000), Rosselli et al. (2002)), although again the deficits are not consistent: In the study by Gollan et al. (2002), deficits were found for Spanish–English bilinguals relative to English monolinguals in about three-quarters of the letter and category trials. Many of the bilingual participants reported that their English proficiency was the same as that of monolinguals, although proficiency level was not formally tested. Proficiency, however, has been found to be positively related to performance in verbal fluency tasks in both younger and older (monolingual) adults (Hedden, Lautenschlager & Park, 2005; Salthouse, 1993), indicating the importance of assuring that proficiency in the language of testing is adequate.
Some of the behavioral differences between monolinguals and bilinguals in lexical access have been substantiated by neuroimaging data. Rodriguez-Fornells et al. (2005) combined evidence from behavioral, electrophysiological, and neuroimaging methods in a tacit picture-naming task, in which participants decided if the name of a pictured object began with a vowel or a consonant. Their results showed that bilinguals were slower than monolinguals overall, and were especially slow when the appropriate response was different for the two languages. More importantly, bilinguals but not monolinguals, engaged two frontal areas while performing the naming task, specifically, the left dorsolateral prefrontal cortex and the anterior cingulate cortex. These areas were also reported in a study by Hernandez and Meschyan (2006) in which bilinguals named pictures in their weaker language, although a different pattern was found in the stronger language. Rodriguez-Fornells et al. (2005) concluded that simple naming in one language for a bilingual involves conflict between the two language systems; the conflict carries costs in both time and accuracy, and the frontal cortex responsible for executive functioning is recruited to deal with this conflict.
Task differences can also be reliably identified using neuroimaging methods. Paulesu et al. (1997) used fMRI and found that phonemic (letter) fluency tasks activated a more extensive area in the left inferior frontal gyrus (LIFG) than did category fluency. More recently, Baldo, Schwartz, Wilkins, and Dronkers (2006) examined a group of left-hemisphere stroke patients and confirmed that letter fluency relied on frontal regions, whereas category fluency relied on temporal areas. The authors attributed these results to the increased demands of strategic search in generating words in a phonemic category relative to a semantic category. Furthermore, the activation of the LIFG was not limited to linguistic tasks, but was found as well for perceptual decision-making tasks that did not involve linguistic processing, indicating the involvement of general cognitive mechanisms in linguistic processing where two languages are involved (Heekeren, Marrett, Ruff, Bandettini, & Ungerleider, 2006). Such evidence is essential for interpreting behavioral differences between monolinguals and bilinguals performing these tasks.
These neuroimaging studies demonstrate that some types of simple word retrieval and naming by bilinguals involve frontal areas; it is therefore reasonable to assume that such tasks recruit aspects of executive processing for their performance. Yet, studies consistently report bilingual advantages in nonverbal executive control tasks, both for children (Bialystok, 2001; Carlson & Meltzoff, in press; Mezzacappa, 2004), and adults (Bialystok et al., 2004; Bialystok, Craik, & Ryan, 2006; Costa, Hernandez, & Sebastián-Gallés, in press). This advantage has been attributed to the enhancement of executive processes through their constant involvement in ordinary language use; bilingual language production necessarily involves the resolution of conflict between the two competing language systems, a process that involves frontal executive processes.
The evidence for bilingual advantages in executive processing suggests that lexical retrieval tasks that involve executive control, such as letter fluency, should actually be solved better by bilinguals than by monolinguals, in spite of numerous studies indicating the opposite pattern. The problem is that the research on lexical retrieval has not assessed vocabulary, even though it has been shown to correlate with performance on fluency tasks (Hedden et al., 2005). Therefore, performance in verbal fluency tasks by bilinguals is controlled by two countervailing trends—the lower overall vocabulary and reduced automaticity of word retrieval, predicting bilingual deficits on the one hand, and the higher level of executive control predicting bilingual advantages on the other. Therefore, to assess the effect of language background on lexical retrieval it is necessary to determine the equivalence of the monolingual and bilingual participants on measures of verbal proficiency. An understanding of the relative abilities of monolinguals and bilinguals on lexical retrieval tasks must consider both the executive demands of the task and the vocabulary level of participants in each group. The present studies investigated lexical retrieval in monolingual and bilingual adults in the context of vocabulary size and executive control in order to understand more precisely the differences in lexical retrieval between language groups.
Section snippets
Study 1
Previous research comparing monolinguals and bilinguals on verbal fluency tasks has offered descriptive statements about bilingual processing in terms of overall proficiency, but those claims have rarely been supported by documented performance on a vocabulary test. Therefore, the first study compared monolinguals and bilinguals on picture naming and verbal fluency tasks and included a formal measure of receptive vocabulary. The purpose was to establish whether differences in word retrieval
Study 2
The purpose of this study was to pursue the results of Study 1 using more detailed assessments of language proficiency and a more demanding letter fluency task. In addition, a larger group of bilinguals was tested to verify the effect of vocabulary size on the subgroups of bilinguals.
General discussion
In two experiments, monolinguals and bilinguals performed differently on tasks requiring rapid lexical access. The initial results of the first study appeared to support the findings generally reported in the literature in which bilinguals perform more poorly than monolinguals on these tasks, but more detailed analyses revealed a more complex pattern. First, when initial levels of vocabulary were taken into account, either by dividing bilinguals into subgroups or statistically controlling for
Acknowledgment
The research reported in this paper was funded by a grant from the Canadian Institutes of Health Research (CIHR) to Ellen Bialystok and Fergus Craik.
References (50)
- et al.
Word fluency and brain damage
Neuropsychologia
(1967) - et al.
Lexical access in bilingual speech production: Evidence from language switching in highly-proficient bilinguals and L2 learners
Journal of Memory and Language
(2004) - et al.
Bilingual performance on the Boston naming test: Preliminary norms in Spanish and English
Brain and Language
(1998) The competing language systems of the multilingual: A developmental study of decoding and encoding processes
Journal of Verbal Learning and Verbal Behavior
(1979)- et al.
Bilingual language switching in naming: Asymmetrical costs of language selection
Journal of Memory and Language
(1999) - et al.
Processing semantic anomalies in two languages: an electrophysiological exploration in both languages of Spanish-English bilinguals
Cognitive Brain Research
(2005) - et al.
Role of frontal versus temporal cortex in verbal fluency as revealed by voxel-based lesion symptom mapping
Journal of the International Neuropsychological Society
(2006) Bilingualism in development: Language, literacy, and cognition
(2001)- Bialystok, E. (2006). Bilingualism and cognitive development: Evidence for changes in attentional control.In Paper...
- et al.
Bilingualism, aging, and cognitive control: Evidence from the Simon task
Psychology and Aging
(2004)
Executive control in a modified anti-saccade task: Effects of aging and bilingualism
Journal of Experimental Psychology: Learning, Memory, and Cognition
Patterns of lexical processing in a nonnative language
Journal of Experimental Psychology: Learning, Memory, and Cognition
Lexical access in bilingual production
How do highly proficient bilinguals control their lexicalization process? Inhibitory and language-specific selection mechanisms are both functional
Journal of Experimental Psychology: Learning, Memory, and Cognition
Planning and task management in older adults: Cooking breakfast
Memory and Cognition
Verbal fluency subtest of the Delis–Kaplan executive function system
Peabody picture vocabulary test—III
Automatic and controlled processes in the first- and second-language reading of fluent bilinguals
Memory and Cognition
Second language reading in fluent bilinguals
Applied Psycholinguistics
Effects of bilingualism, aging, and semantic relatedness on memory under divided attention
Canadian Journal of Experimental Psychology
What is a TOT? Cognate and translation effects on tip-of-the-tongue states in SpanishEnglish and Tagalog-English bilinguals
Journal of Experimental Psychology: Learning, Memory, and Cognition
The bilingual effect on Boston naming test performance
Journal of the International Neurological Society
Bilingualism affects picture naming but not picture classification
Memory and Cognition
Cited by (177)
Predicting reading comprehension performance based on student characteristics and item properties
2023, Studies in Educational EvaluationSpanish-English dual language learners' bilingual profiles: Executive function and developmental outcomes
2023, Journal of Applied Developmental PsychologyGeographic variability in limited English proficiency: A cross-cultural study of cognitive profiles
2023, Journal of the International Neuropsychological SocietyA longitudinal investigation of the effects of language instruction versus immersion on cognitive functions in young adult Chinese speakers learning English
2023, Studies in Second Language AcquisitionProcessing to learn noncanonical word orders: Exploring linguistic and cognitive predictors of reanalysis in early L2 sentence comprehension
2023, Studies in Second Language Acquisition