The demographic changes in the Western world in the last two decades have led to rapid growth in the number of children being raised bilingually, and in many locations they represent a majority of the school population. With this increase in the number of bilingual children, researchers as well as educators and practitioners, face a diagnostic dilemma which arises from similarities in the linguistic manifestations of child second language (L2) acquisition and of Specific Language Impairment (SLI). This dilemma has motivated a new field of research, the study of bilingual children with Specific Language Impairment (BISLI), which aims at disentangling the effects of bilingualism from those of SLI, making use of both models of bilingualism and models of language impairment. The majority of the studies are currently focused on morphosyntax as a key direction of research. The present issue, which originated in papers presented at a scientific workshop funded by the Israel Science Foundation and The Hebrew University's Center for Advanced Study, held in February 2009, aims at broadening this area of research to other linguistic domains.

Definitional and methodological controversies in research on both bilingualism and SLI are major obstacles to addressing the dilemma outlined above and to ensuring comparability of across languages and across research sites. Selection of homogeneous groups of bilingual children is complicated by social status (e.g., immigrant, indigenous, privileged minorities), differences in age (early/late age of onset of L2 acquisition), birth order, family size, acquisition order (simultaneous/sequential), degree of exposure, acquisition contexts (e.g., one parent for each language, one language at home vs. another language at kindergarten or school), and relative prestige of languages (Meisel, Reference Meisel2007). These variables influence success in language acquisition of both the first and the second language. Identifying children with SLI on the background of this heterogeneity is a great challenge, since SLI is a primary deficit in linguistic abilities and language development (Bishop, Bright, James, Bishop & van der Lely, Reference Bishop, Bright, James, Bishop and van der Lely2000).

Inclusionary criteria make the population with SLI very heterogeneous, with differing criteria and a controversy about which linguistic aspects are affected and to what extent (Bishop, Reference Bishop, Bishop and Leonard2000; van der Lely, Reference van der Lely1998). SLI is manifested at different linguistic levels (Leonard, Reference Leonard1998), but not all linguistic skills are equally impaired. A deficit in morphosyntax is often viewed as the key symptom (Crago & Gopnik, Reference Crago, Gopnik, Watkins and Rice1994; Rice & Wexler, Reference Rice and Wexler1996), but due to the heterogeneity of the population, some children might show a more severe impairment at a different linguistic level. For both SLI and bilingual children, representation of linguistic knowledge may be incomplete or different. Processing of linguistic stimuli may also differ under the influence of duration, rate and salience, showing up as difficulties in memory, temporal integration or word-finding/lexical retrieval problems. Developing bilinguals, when compared to monolinguals, may also show evidence of different linguistic representations and/or difficulties in fluency related to their lexical knowledge or reduced exposure to each language.

In light of these confounds associated with language impairment and bilingualism, five central issues are addressed by the articles in this issue:

1. 1. Can bilingualism be disentangled from SLI in impaired children?

2. 2. How should one diagnose SLI in bilingual populations?

3. 3. Are bilingualism and SLI “two of a kind” (Crago & Paradis, Reference Crago, Paradis, Levy and Schaeffer2003)?

4. 4. Do bilingual children with SLI show a “double delay” (Paradis, Reference Paradis2007; Paradis, Crago, & Genesee, Reference Paradis, Crago and Genesee2005/2006; Paradis, Crago, Genesee & Rice, Reference Paradis, Crago, Genesee and Rice2003)?

5. 5. Can bilingualism be “instructive” for children with SLI (Roeper, Reference Roeper2009).

To fully answer these questions a two by two design (shown in Figure 1) is necessary to compare four different possible groups of children, bilingual and monolingual with typical language development (TLD) and with SLI, where data collected from the bilingual children is collected in both languages.

Figure 1.

Informative between group comparisons in the two by two design.

Comparisons of bilingual TLD with Monolingual SLI test the “two of a kind” hypothesis. Comparisons of bilingual TLD with bilingual SLI and monolingual TLD with monolingual SLI are informative in terms of the influence of SLI on linguistic knowledge and diagnosis of SLI in bilingual populations. Comparisons of bilingual TLD with monolingual TLD and bilingual SLI with monolingual SLI reveal the effects of bilingualism addressing the “double delay” question. More often than not, however, studies (including those in the present issue) focus on a part of this design. This is done in order to focus the studies on one of the questions. Earlier studies (e.g., Paradis, Reference Paradis1999) compared bilingual TLD with Monolingual SLI, while later studies, in particular when the issue at stake is identifying children with SLI among bilingual children, made comparisons among different groups of bilinguals (Kohnert, Reference Kohnert2007).

This special issue provides initial answers to the above questions. The similarities and differences between monolingual English-speaking children with SLI and bilingual children for whom English is the second language (Crago & Paradis, Reference Crago, Paradis, Levy and Schaeffer2003; Paradis, Reference Paradis1999; as well as Paradis, Genesee & Crago, Reference Paradis, Genesee and Crago2006) is addressed here by Chondrogianni and Marinis. These researchers offer a window into the processing differences between English/L2 children and children with SLI. Comparing production and performance in an online task, they bring evidence for intact underlying syntactic representations in child L2 acquisition despite non-target-like production.

Moving away from tense morphology and away from English, Jacobson's contribution looks into the use of object clitics in the first language (L1) Spanish of Spanish–English school-age bilinguals with and without SLI. Jacobson shows that despite the continued development of object clitics in both populations, a gap remains even into later grades, which is indicative of incomplete acquisition by bilingual children with SLI, a feature distinguishing them from bilingual children with TLD. Verb inflection does not seem to offer such a distinction. The author further argues that the continued development in the L1 of the children with SLI shows that bilingualism does not exacerbate the difficulties in language acquisition these children exhibit.

BISLI research in Europe, unlike the work in the US and Canada, explores a wide variety of language pairs, addressing other domains of syntax and morphosyntax, beyond tense morphology. Rothweiler, Chilla and Clahsen capitalize on previous findings, addressing both empirical and theoretical issues in a group study of monolingual German-speaking children with SLI and bilingual Turkish–German-speaking children with SLI. In earlier studies, Rothweiler, Babur and Kroffke (Reference Rothweiler, Babur and Kroffke2007) found agreement and case errors in both languages of Turkish–German bilingual children with SLI, with more errors in L2 German. Chilla and Babur (Reference Chilla, Babur, Topbas and Yavas2010) reported that SLI in the German of successive bilinguals matches SLI characteristics of monolingual German SLI. Empirically, the present study shows that agreement errors, rather than tense errors or problems at the CP level, can serve as a clinical marker for SLI in both monolingual and bilingual German, with no differences between the two populations. Theoretically, these findings support the Agreement Deficit Hypothesis (Clahsen, Reference Clahsen1989, Reference Clahsen1991) over alternative explanations.

A handful of studies explore bilingual language impairment in narrative and discourse domains. Iluz-Cohen and Walters report on two studies of bilingual English–Hebrew speaking children with and without SLI, one replicating the findings by Pearson (Reference Pearson, Oller and Eilers2002) that narrative abilities are not impaired by SLI in bilingual children while other morphosyntactic measures are, and the second examining frequency and direction of code-switching as a possible marker of SLI in bilingual children.

Only a few studies have looked into other aspects of BISLI. Roeper's contribution builds on initial evidence from studies of the syntax–semantics interface with bilingual children in Germany that show a rapid progress in bilingual children which does not resemble the persistent deficits reported for children with SLI (Schulz & Wenzel, Reference Schulz and Wenzel2007). It presents a theoretically driven, but practically grounded, approach to the role of bilingualism as a bootstrapping mechanism in BISLI, giving examples from impaired and unimpaired monolingual and bilingual children. His line of argumentation gains support from the empirical evidence presented in the above articles.

The above studies focus on spoken language in bilingualism with SLI. In contrast, Woll and Morgan offer an insight into impairments in the morphosyntactic domain in cross-modal bilingualism of British Sign Language and English. Drawing from several case studies and one group study of cross-modal bilingualism in impaired populations, they show that SLI manifests itself in morphosyntactic errors in both modalities, while other impairments affect only one of the modalities reflecting the nature of the impairment. Theoretically, their study lends support to the claim that SLI is not related to auditory processing, but rather to linguistically related processing or linguistic representation.

Addressing the first two questions presented earlier, both Jacobson and Iluz-Cohen and Walters point to features which can be used to disentangle bilingualism from SLI in bilingual populations, object clitics in the L1 of Spanish–English bilinguals and the degree of code-switching on controlled tasks for English–Hebrew bilinguals, respectively. While the former is specific to the studied population, the latter is expected to reoccur in other bilingual populations, and could help in diagnosing SLI across bilingual populations. Another more general observation which has impact on diagnosing SLI across bilingual populations emerges from Woll and Morgan's study that shows that SLI, unlike other impairments, is manifested in both modalities, supporting the call for diagnosing children with BISLI in both languages.

Addressing the three questions which elaborate on the possible relations between bilingualism and SLI, all contributors show that despite the superficial similarity in language production between bilingualism and SLI they are not “two of a kind”, no matter which linguistic system one studies. To this end, Chondrogianni and Marinis's innovative study is a major contribution, offering a window into syntactic representations in child L2, showing that it is intact despite the limited input unlike what was found for children with SLI. Moreover, Rothweiler, Chilla and Clahsen, as well as Jacobson, show that there is no cumulative effect for bilingualism and SLI, that is, there is no “double delay”. Jacobson further argues that the constant development in the L1 of the children with SLI indicates that bilingualism does not necessarily aggravate the difficulties in language acquisition, while Roeper takes this claim one step further, providing theoretical arguments for the “instructive” role of bilingualism as a bootstrapping mechanism in BISLI.

Although the work on BISLI has only begun recently, these studies clearly advances our understanding of the nature of the problem and offer some directions for resolution of the challenge of disentangling bilingualism and SLI and setting guidelines for diagnosing SLI in bilingual population.