Commonalities between bilingual and monolingual development
Paradis’s summary of the evidence on sources of individual differences in bilingual development identifies factors in children’s language learning abilities and in the quantity and quality of their language exposure that create individual differences in the rates at which bilingually developing children acquire each of their languages. Individual differences in many of these abilities (e.g., phonological memory and nonverbal analytic abilities) and individual differences in the quantity and quality of single language exposure similarly have documented effects on the rate of monolingual development (Hoff, Reference Hoff2006; Rowe & Weisleder, Reference Rowe and Weisleder2020; Rowe & Snow, Reference Rowe and Snow2020). Although some environmental factors may be more likely to apply in bilingual environments than in monolingual environments, many of the environmental factors and all of the ability factors apply in both (Lauro et al., Reference Lauro, Core and Hoff2020; Unsworth, Reference Unsworth, Nicoladis and Montanari2016).
Differences between monolingual and bilingual development
A good case can be made that the environmental factors that create individual differences in the language growth of monolingual and bilingual children also create differences between monolingual and bilingual development. Children who hear two languages must experience a smaller quantity of exposure to each language, unless their parents talk twice as much or they sleep less than children who hear only one. Also, in the case of immigrant children and children of immigrant parents, the quality of children’s exposure to the host language may be affected by limited proficiency in some of their sources of input and in the limited functions and range of topics to which a home (i.e., heritage) language is applied. It has been well documented that children in immigrant families who are exposed to two languages from birth acquire each of those languages at a slower rate than monolingual children of native parents acquire their single language (Hoff et al., Reference Hoff, Core, Place, Rumiche, Señor and Parra2012; Hoff & Ribot, Reference Hoff and Ribot2017). It is further apparent that these differences are significantly explained by differences in the quantity and quality of the children’s exposure to the language or languages they learn (Hoff, Reference Hoff and Lockman2021; Place & Hoff, Reference Place and Hoff2011, Reference Place and Hoff2016). It has also been documented that language proficiency in the heritage language tends to be weaker in the long run than proficiency in the host language (Giguere & Hoff, Reference Giguere and Hoff2020). On this latter finding, Paradis comments that it would be naïve to expect the heritage language skills of children in immigrant families to be like the skills of children raised in a country where that is the national language. For similar reasons, I would add, it would also be naïve to expect the host language skills of children immigrant families to be like the skills of children who hear only that language and who hear it primarily from native speakers. Because the environmental supports for monolingual and bilingual language acquisition differ while the environmental requirements stay the same, the course of bilingual development must differ in some way from the course of monolingual development.
Future directions
Paradis argues that the field of bilingual development should move away from comparing bilinguals to monolinguals and should instead focus on the more nuanced question of what causes some bilinguals to progress differently from others. Paradis’s argument for moving away from between-group comparisons is twofold: (1) that the results of two-group comparisons can be misleading, and (2) that comparison of bilinguals to monolinguals gives rise to a deficit ideology when bilinguals’ skills are found to be weaker than monolinguals’ skill. Paradis finds this deficit ideology to be particularly prevalent in descriptions of bilinguals’ heritage language proficiency, but I would argue it is prevalent in discussions of the host language skills of bilingual children as well because bilingual children of foreign-born parents, on average, reach school age with weaker host language skills than do monolingual children of native parents.
As a counterargument to Paradis’s suggestion to abandon the between-group comparison of bilinguals to monolinguals, I would argue: (1) that between-group comparisons are not misleading if interpreted correctly, and (2) that sometimes comparisons of bilingual to monolingual development are the appropriate approach to answering theoretical and applied questions. To illustrate her assertion that bilingual-monolingual comparisons can be misleading, Paradis uses the finding in Hoff et al. (Reference Hoff, Core, Place, Rumiche, Señor and Parra2012) that bilingual children had significantly smaller English vocabularies than age and SES-matched monolingual children. She argues that this finding would mislead clinicians to expect that bilinguals would show delays in both their languages. She argues that this is a wrong inference because, in that same sample, the subgroup of bilinguals who heard 70% or more of their input in English did not significantly differ from the monolingual children.
The clinician’s inference that Paradis is concerned about is indeed mistaken–not because an English-dominant subgroup outperformed the average for all bilinguals but because of the statistical fact that average differences between two groups never imply nonoverlapping distributions. Many bilinguals will have single language skills well within the the normal range of variation for monolinguals. Paradis further, and mistakenly, extrapolates from the strong English performance of the subgroup with English-dominant input that all bilingual children will be similar to monolingual children in one of their languages. However, some children have more nearly balanced input. The same analysis in Hoff et al. (Reference Hoff, Core, Place, Rumiche, Señor and Parra2012), which found the English-dominant bilinguals to be similar to the English monolinguals, also found that the subgroup of bilinguals who heard 40% to 60% of their input in English were significantly different from the English monolinguals in the size of their English vocabularies. There was no Spanish monolingual group in this study, but the findings suggest that some bilingual children whose language exposure is relatively equally distributed across two languages may differ from monolingual children in both their languages.
Paradis’s second argument against making comparisons of bilinguals to monolinguals is that such comparisons encourage a deficit view of bilingualism. But whether the language skills of bilingual children are called differences or deficits relative to the language skills of monolingual children does not change the children’s language skills, nor does it change the importance of those skills to academic success. A labeling difference will not remedy the documented academic difficulties that bilingual children experience when their host language skills do not meet the requirements of schooling (e.g., Kieffer, Reference Kieffer2012). Further, I would suggest that the deficit view of bilingualism does not really arise solely from the results of two-group comparisons, nor will it go away if we focus on individual differences instead. Rather, the root problems are prejudice against immigrant groups which results in the racialization of their bilingualism (Hoff, Reference Hoff and Lockman2021; Flores & Rosa, Reference Flores and Rosa2019) and the fact that the clear benefit of bilingualism – the ability to speak and understand another language – is not particularly valued by large segments of the English-speaking population.
Identifying differences that matter
I would suggest that rather than abandoning the between-group approach because it reveals average differences between the single language skills of monolingual and bilingual children, research should focus on investigating the long-term outcomes of the early differences. It would be of interest to know whether, when, and under what circumstances differences in oral language skill diminish. It would also be of interest to know what the consequences of those differences are for longer-term educational attainment, occupational success, and well-being. There are multiple findings in the literature from the comparison of bilinguals to monolinguals which suggest that for practical purposes heritage bilinguals catch up to monolinguals in the host language (if not in the heritage language) (Giguere & Hoff, Reference Giguere and Hoff2020), even as sensitive measures still reveal differences (Giguere & Hoff, Reference Giguere and Hoff2020; Li & Hartshorne, Reference Li and Hartshorne2022). And there is evidence that where differences persist, as in vocabulary size, they do not have the same consequences for reading or academic achievement as they would for monolinguals (Mancilla-Martinez & Lesaux, Reference Mancilla‐Martinez and Lesaux2011; Pearson, Reference Pearson1993).
I agree with Paradis that more could be done to describe – and then explain–the varied outcomes of dual language exposure. The extant literature tends to treat each of a bilingual’s languages as a separate outcome, but bilinguals vary in their profiles of dual language skill. For example, bilinguals vary in the degree to which their dual skills are balanced or one language is dominant. Bilingual children also vary (as do monolingual children) in the total amount of language knowledge they have acquired. (This possibility of knowing more in total removes the need for the trade-off between languages in skill attainment that is implied by the balance terminology.) In a sample of Spanish–English bilingual 5-year-olds in the U.S., we found children who were nearly balanced bilinguals and children who were strongly English dominant, but the English dominant children did not have stronger English skills than the balanced children. Because this study treated profiles of dual language skill as the outcome, it was able to identify factors that support children’s development of strong skills in two languages (Hoff et al., Reference Hoff, Tulloch and Core2021).
Another example of skill constellations in bilingualism is the receptive-expressive gap. In bilingual development, receptive and expressive abilities are often not linked in the same way in both languages. A common pattern in heritage-host language bilingualism is for receptive abilities to be much stronger than expressive abilities in the heritage language (Gibson et al., Reference Gibson, Peña and Bedore2014; Giguere & Hoff, Reference Giguere and Hoff2022). It would be of practical and scientific interest to understand the factors that contribute to this outcome, and an individual difference approach could be key.
But to return to the argument for bilingual-monolingual comparison, I would suggest that there are both theory-driven and practical questions about the effects of knowing one language on knowledge of another that are best addressed by comparing the single language skills and usage of children who know only one language to children who know that language and another in addition. Chondrogianni (this volume) provides examples of theory-driven questions that comparison of bilinguals to monolinguals can address. There are also many practical questions such comparisons answer. A comparison of bilinguals to monolinguals is relevant to answering educators’ legitimate question as to whether bilingual children can be expected to have the same single language skills as monolingual children when they enter school. Finally, and importantly, when parents ask those of us who study bilingual development for our expert opinion on whether exposure to two languages will affect their child’s development of each, to tell them that we prefer not to compare bilinguals to monolinguals because it feeds a deficit approach to bilingualism would be an infelicitous reply.
Commonalities between bilingual and monolingual development
Paradis’s summary of the evidence on sources of individual differences in bilingual development identifies factors in children’s language learning abilities and in the quantity and quality of their language exposure that create individual differences in the rates at which bilingually developing children acquire each of their languages. Individual differences in many of these abilities (e.g., phonological memory and nonverbal analytic abilities) and individual differences in the quantity and quality of single language exposure similarly have documented effects on the rate of monolingual development (Hoff, Reference Hoff2006; Rowe & Weisleder, Reference Rowe and Weisleder2020; Rowe & Snow, Reference Rowe and Snow2020). Although some environmental factors may be more likely to apply in bilingual environments than in monolingual environments, many of the environmental factors and all of the ability factors apply in both (Lauro et al., Reference Lauro, Core and Hoff2020; Unsworth, Reference Unsworth, Nicoladis and Montanari2016).
Differences between monolingual and bilingual development
A good case can be made that the environmental factors that create individual differences in the language growth of monolingual and bilingual children also create differences between monolingual and bilingual development. Children who hear two languages must experience a smaller quantity of exposure to each language, unless their parents talk twice as much or they sleep less than children who hear only one. Also, in the case of immigrant children and children of immigrant parents, the quality of children’s exposure to the host language may be affected by limited proficiency in some of their sources of input and in the limited functions and range of topics to which a home (i.e., heritage) language is applied. It has been well documented that children in immigrant families who are exposed to two languages from birth acquire each of those languages at a slower rate than monolingual children of native parents acquire their single language (Hoff et al., Reference Hoff, Core, Place, Rumiche, Señor and Parra2012; Hoff & Ribot, Reference Hoff and Ribot2017). It is further apparent that these differences are significantly explained by differences in the quantity and quality of the children’s exposure to the language or languages they learn (Hoff, Reference Hoff and Lockman2021; Place & Hoff, Reference Place and Hoff2011, Reference Place and Hoff2016). It has also been documented that language proficiency in the heritage language tends to be weaker in the long run than proficiency in the host language (Giguere & Hoff, Reference Giguere and Hoff2020). On this latter finding, Paradis comments that it would be naïve to expect the heritage language skills of children in immigrant families to be like the skills of children raised in a country where that is the national language. For similar reasons, I would add, it would also be naïve to expect the host language skills of children immigrant families to be like the skills of children who hear only that language and who hear it primarily from native speakers. Because the environmental supports for monolingual and bilingual language acquisition differ while the environmental requirements stay the same, the course of bilingual development must differ in some way from the course of monolingual development.
Future directions
Paradis argues that the field of bilingual development should move away from comparing bilinguals to monolinguals and should instead focus on the more nuanced question of what causes some bilinguals to progress differently from others. Paradis’s argument for moving away from between-group comparisons is twofold: (1) that the results of two-group comparisons can be misleading, and (2) that comparison of bilinguals to monolinguals gives rise to a deficit ideology when bilinguals’ skills are found to be weaker than monolinguals’ skill. Paradis finds this deficit ideology to be particularly prevalent in descriptions of bilinguals’ heritage language proficiency, but I would argue it is prevalent in discussions of the host language skills of bilingual children as well because bilingual children of foreign-born parents, on average, reach school age with weaker host language skills than do monolingual children of native parents.
As a counterargument to Paradis’s suggestion to abandon the between-group comparison of bilinguals to monolinguals, I would argue: (1) that between-group comparisons are not misleading if interpreted correctly, and (2) that sometimes comparisons of bilingual to monolingual development are the appropriate approach to answering theoretical and applied questions. To illustrate her assertion that bilingual-monolingual comparisons can be misleading, Paradis uses the finding in Hoff et al. (Reference Hoff, Core, Place, Rumiche, Señor and Parra2012) that bilingual children had significantly smaller English vocabularies than age and SES-matched monolingual children. She argues that this finding would mislead clinicians to expect that bilinguals would show delays in both their languages. She argues that this is a wrong inference because, in that same sample, the subgroup of bilinguals who heard 70% or more of their input in English did not significantly differ from the monolingual children.
The clinician’s inference that Paradis is concerned about is indeed mistaken–not because an English-dominant subgroup outperformed the average for all bilinguals but because of the statistical fact that average differences between two groups never imply nonoverlapping distributions. Many bilinguals will have single language skills well within the the normal range of variation for monolinguals. Paradis further, and mistakenly, extrapolates from the strong English performance of the subgroup with English-dominant input that all bilingual children will be similar to monolingual children in one of their languages. However, some children have more nearly balanced input. The same analysis in Hoff et al. (Reference Hoff, Core, Place, Rumiche, Señor and Parra2012), which found the English-dominant bilinguals to be similar to the English monolinguals, also found that the subgroup of bilinguals who heard 40% to 60% of their input in English were significantly different from the English monolinguals in the size of their English vocabularies. There was no Spanish monolingual group in this study, but the findings suggest that some bilingual children whose language exposure is relatively equally distributed across two languages may differ from monolingual children in both their languages.
Paradis’s second argument against making comparisons of bilinguals to monolinguals is that such comparisons encourage a deficit view of bilingualism. But whether the language skills of bilingual children are called differences or deficits relative to the language skills of monolingual children does not change the children’s language skills, nor does it change the importance of those skills to academic success. A labeling difference will not remedy the documented academic difficulties that bilingual children experience when their host language skills do not meet the requirements of schooling (e.g., Kieffer, Reference Kieffer2012). Further, I would suggest that the deficit view of bilingualism does not really arise solely from the results of two-group comparisons, nor will it go away if we focus on individual differences instead. Rather, the root problems are prejudice against immigrant groups which results in the racialization of their bilingualism (Hoff, Reference Hoff and Lockman2021; Flores & Rosa, Reference Flores and Rosa2019) and the fact that the clear benefit of bilingualism – the ability to speak and understand another language – is not particularly valued by large segments of the English-speaking population.
Identifying differences that matter
I would suggest that rather than abandoning the between-group approach because it reveals average differences between the single language skills of monolingual and bilingual children, research should focus on investigating the long-term outcomes of the early differences. It would be of interest to know whether, when, and under what circumstances differences in oral language skill diminish. It would also be of interest to know what the consequences of those differences are for longer-term educational attainment, occupational success, and well-being. There are multiple findings in the literature from the comparison of bilinguals to monolinguals which suggest that for practical purposes heritage bilinguals catch up to monolinguals in the host language (if not in the heritage language) (Giguere & Hoff, Reference Giguere and Hoff2020), even as sensitive measures still reveal differences (Giguere & Hoff, Reference Giguere and Hoff2020; Li & Hartshorne, Reference Li and Hartshorne2022). And there is evidence that where differences persist, as in vocabulary size, they do not have the same consequences for reading or academic achievement as they would for monolinguals (Mancilla-Martinez & Lesaux, Reference Mancilla‐Martinez and Lesaux2011; Pearson, Reference Pearson1993).
I agree with Paradis that more could be done to describe – and then explain–the varied outcomes of dual language exposure. The extant literature tends to treat each of a bilingual’s languages as a separate outcome, but bilinguals vary in their profiles of dual language skill. For example, bilinguals vary in the degree to which their dual skills are balanced or one language is dominant. Bilingual children also vary (as do monolingual children) in the total amount of language knowledge they have acquired. (This possibility of knowing more in total removes the need for the trade-off between languages in skill attainment that is implied by the balance terminology.) In a sample of Spanish–English bilingual 5-year-olds in the U.S., we found children who were nearly balanced bilinguals and children who were strongly English dominant, but the English dominant children did not have stronger English skills than the balanced children. Because this study treated profiles of dual language skill as the outcome, it was able to identify factors that support children’s development of strong skills in two languages (Hoff et al., Reference Hoff, Tulloch and Core2021).
Another example of skill constellations in bilingualism is the receptive-expressive gap. In bilingual development, receptive and expressive abilities are often not linked in the same way in both languages. A common pattern in heritage-host language bilingualism is for receptive abilities to be much stronger than expressive abilities in the heritage language (Gibson et al., Reference Gibson, Peña and Bedore2014; Giguere & Hoff, Reference Giguere and Hoff2022). It would be of practical and scientific interest to understand the factors that contribute to this outcome, and an individual difference approach could be key.
But to return to the argument for bilingual-monolingual comparison, I would suggest that there are both theory-driven and practical questions about the effects of knowing one language on knowledge of another that are best addressed by comparing the single language skills and usage of children who know only one language to children who know that language and another in addition. Chondrogianni (this volume) provides examples of theory-driven questions that comparison of bilinguals to monolinguals can address. There are also many practical questions such comparisons answer. A comparison of bilinguals to monolinguals is relevant to answering educators’ legitimate question as to whether bilingual children can be expected to have the same single language skills as monolingual children when they enter school. Finally, and importantly, when parents ask those of us who study bilingual development for our expert opinion on whether exposure to two languages will affect their child’s development of each, to tell them that we prefer not to compare bilinguals to monolinguals because it feeds a deficit approach to bilingualism would be an infelicitous reply.
Acknowledgements
Preparation of this commentary was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development Grant HD068421 to Erika Hoff