For many years the study of language impairment in children focused almost exclusively on children learning English. The past decade has seen a decided and salutary broadening of this exclusive focus as researchers from many nations have become interested in understanding the impaired language acquisition of children speaking diverse languages. The concern for understanding the cross-linguistic nature of developmental language impairment has followed in the path of investigations into the cross-linguistic nature of typical language development that were initiated by Slobin and his colleagues (e.g., Slobin, 1985).

Cross-linguistic studies of language impairment in children have both a theoretical and a practical impetus. Theoretically, the study of typical and impaired language development across a number of structurally different languages should reveal commonalities as well as differences in how children learn languages. Such studies should help researchers determine what is universal and what is variable in the ways that children learn languages. By sorting the variable from the universal, such research enhances knowledge about the properties of language development that are general to the process of language learning and those that are determined by the structure of the language that the child is exposed to. Practically, understanding what form impairment takes in various languages improves the possibilities for assessment and treatment in those languages. Understanding the contribution of input to the nature and timing of acquisition in a specific language has important consequences for developing intervention approaches. On the other hand, understanding the common properties of language learning helps to clarify in what ways impairment results from possible disruptions of basic human biological and cognitive mechanisms.

Most cross-linguistic studies of children with language impairment have been concerned with the morphological deficits experienced by children with specific language impairment (SLI) (see specific language impairment in children). Some such grammatical deficits are found to occur in almost all languages that have been studied so far, regardless of their structure. Other deficits are particular to individual language families. One deficit that appears to exist in many languages has to do with the acquisition of verbal inflection. Finite verbal inflections and auxiliary verbs produced by children with SLI are optionally missing or substituted for in the following languages: German (Clahsen, Bartke, and Gollner, 1997; Rice, Ruff Noll, and Grimm, 1997), Dutch (de Jong, 1999), Swedish (Hansson, 1997), Norwegian (Meyer Bjerkan, 1999), English (Rice and Wexler, 1996), French (Jakubowicz and Nash, 2001; Paradis and Crago, 2001), Italian (Bottari, Cipriani, and Chilosi, 1996; Bortolini, Caselli, and Leonard, 1997; Bottari et al., 2001), Japanese (Fukuda and Fukuda, 2001), Greek (Clahsen and Dalalakis, 1999), Inuktitut (Crago and Allen, 2001), and Arabic (Abdalla, 2002).

Various language families show interesting patterns that relate to the structure or typology of those particular groups of languages. For example, studies of Germanic languages such as Dutch, Swedish, and German have shown that there are word order consequences related to the omission of finite verb inflections and auxiliaries in the speech of children with SLI.

Studies of Romance languages such as Italian and French have demonstrated that children with SLI have greater trouble acquiring the past tense than the present tense, while English-speaking children with SLI have difficulty with both tenses. Impaired speakers of French and Italian also have difficulty with the production of object clitics. Surprisingly, however, Italian and French children with SLI differ in their acquisition of determiners. Italian-speaking children have more difficulty with this aspect of their grammar than the French-speaking children. It is unfortunate that in the family of Romance languages, there are no comparable acquisition studies of children with SLI learning Spanish, either in the Americas or in Europe.

Speakers of non-Indo-European languages such as Inuktitut and Arabic have different patterns of impairment than other groups. For instance, Arabic speakers replace incorrect verbal inflections with a default form that is typically tense-bearing in the adult language. This is different from the Germanic and Romance languages, where tense-bearing morphemes, such as verbal inflections or auxiliary verbs, tend to be dropped, resulting in a nonfinite verb form such as an infinitive, a participle, or the verb stem appearing as the main verb in the sentence. Specific language impairment in Inuktitut has been shown to present yet another dimension. Here, the trouble a child with impairment had with verbal inflection did not resemble younger normally developing children. This is a different pattern than has been found in many other languages where children with SLI show a pattern of optional verbal inflection that resembles that of younger normally developing children.

Cross-linguistic studies can also play a particularly useful role in verifying hypotheses about the nature of the deficits experienced by children with SLI. They allow researchers to check out explanations based on particular languages with results from typologically different languages. For instance, Leonard, one of the pioneers of cross-linguistic studies of child language impairment, hypothesized from his study of English-speaking children that a plausible explanation for the grammatical deficit in SLI was that these children had difficulty establishing a learning paradigm for morphology in languages where the morphemes had low phonological saliency. He and his colleagues sought systematic cross-linguistic verification for this explanation by studying children with SLI who were learning languages with more salient morphology, such as Italian (Leonard et al., 1992) or Hebrew (Dromi, Leonard, and Shteiman, 1993; Dromi et al., 1999). Indeed, such children appeared to have less difficulty with the acquisition of their verbal morphology. However, succeeding studies of other languages with very phonologically salient verbal inflections, such as Inuktitut and Japanese, have shown that children with SLI did, in fact, have difficulty with the acquisition of verbal inflections. This demonstrates how a series of cross-linguistic studies can be useful in establishing whether a certain theoretical explanation of SLI is meaningful across languages.

Even though the number of languages that are being studied is expanding, there are only a few studies that can be considered truly cross-linguistic in design. These few genuine cross-linguistic studies involve either a format that compares in one study two groups of subjects who speak two different languages or a specific language in which the grammatical variables tested are as identical as possible to a previously studied language. In addition, studies in which investigators have examined the grammatical properties of children speaking a single specific language also display some methodological shortcomings impeding well-founded conclusions. They have varied in the age and criteria of selection of the children being studied. They have also varied in the particular properties of the children's grammar that were assessed. Regrettably, such design issues have made the kinds of comparisons that are essential to establishing the universal and variable properties of acquisition and impairment often inconclusive.

In summary, despite certain limitations, the expansion of studies of language impairment in children to include a wider variety of languages has enlivened theoretical debate, led to new, linguistically based understandings, and enriched perspectives on this communicative disorder. It is important that studies of language impairment in childhood continue to encompass more and different languages. In fact, languages spoken by the vast majority of children in the world's population remain virtually unexplored. It is equally important that more studies be designed to be truly cross-linguistic in nature, with variables and criteria for impairment established as uniformly as possible. The study of bilingual children with impairment provides a unique opportunity for the cross-linguistic observation of language impairment. These children are perfectly matched to themselves as the learners of two languages (see bilingualism and language impairment). Finally, just as important as understanding how children with language impairment learn different languages is recognizing that they are also members of different cultures. Language and culture are inextricably linked and as such they influence each other as well as the manifestations of and beliefs about childhood language impairment.