The Latino population encompasses a diverse group of people who self-identify as descendants of individuals who came to the United States from a predominantly Spanish-speaking country. Over the past decade, the Latino population in the United States has increased four times faster than the general population (Guzmán, 2001). It is estimated that the size of the Latino population will represent one-quarter of the U.S. population, or approximately 81 million Latinos, by the year 2050. The large growth of the Latino population is largely attributable to its high fertility rate (National Center for Health Statistics, 1999); a large proportion of the population is in the childbearing years, and families tend to be larger. Children under the age of 15 account for 30.5% of the Latino population.

The Latino population is linguistically diverse with respect to dialects and languages spoken. The dialects spoken in the country of origin and subsequently brought to the United States evolved from the different regional dialects spoken by the original settlers, the languages spoken by the native peoples of the Americas, and the languages spoken by later immigrants. There are two major groups of Spanish dialects, radical and conservative (Guitart, 1996). The radical dialects are spoken primarily in the coastal areas of Spanish-speaking countries and the Caribbean, while the conservative dialects are spoken in the interior parts of the countries. The dialects vary in phonology, morphosyntax, semantics, and pragmatics, with the most drastic qualitative differences seen in phonology and lexicon. The differences in morphosyntax are more quantitative than qualitative. The specific dialects spoken by Latino children will be influenced by the dialects spoken in their community. Other factors influencing the dialect spoken include the degree of contact with Spanish and English speakers, whether the speaker is learning both languages simultaneously or sequentially, and the prestige attached to the various dialects with which the individual comes in contact (Poplack, 1978; Wolfram, Adger, and Christian, 1999).

Speaking Spanish is one of the major ties that bind the Latino population, and approximately 80% of the population reportedly speaks it. The vast majority of the Latino population consider themselves bilingual; a small percentage is monolingual in either English or Spanish. Twenty-eight percent of the Latino population report that they “do not speak English well” or speak it “not at all” (U.S. Census Bureau, 2000). The number of monolingual Spanish speakers and bilingual English-Spanish speakers reflects the fact that 35% of the Latino population is foreign born and that the majority of foreign-born Latinos entered the United States in the last three decades. Continuous immigration and growth of the Latino population, coupled with greater acceptability of linguistic diversity in the United States, might reverse the previous trend, in which immigrants lost their native language by the third generation (Veltman, 1988). The more likely trend is for a continuous growth of a Latino population that is bilingual.

Bilingualism is not a one-dimensional concept. Rather, bilingualism may be viewed as existing on several continua representing different language competencies in form, content, and use of the language (Valdes and Figueroa, 1994). Collectively, these individual competencies will dictate the child's linguistic proficiency in a language. The degree to which proficiency is exhibited in any one language at a particular point in time is influenced by the situation, the topic, individuals, and context (Romaine, 1995; Zentella, 1997). A shift in topic or a shift in participants may result in a switching of the code. This type of code switching is a verbal skill that requires a large degree of linguistic competence. Code switches are also made by less proficient speakers as a way to compensate for insufficient knowledge of one language.

Latino children exhibit varying degrees of proficiency in both English and Spanish. Given the pervasiveness of English language media, the use of Spanish by the majority of Latino families, and the communities in which Latinos are raised, it is doubtful that many Latino children reach school age as true monolingual Spanish speakers. Some of the children may be considered to be sequential bilinguals because their major exposure prior to entering school was to Spanish and their linguistic skills in English are minimal. These children's main exposure to English will come when they enter school. Impressionistically, many of these children are indistinguishable from monolingual Spanish speakers (e.g., Spanish-speaking children in Mexico). However, differences become apparent when their Spanish is compared with the language spoken by true monolingual Spanish speakers (Merino, 1992). Children who have been exposed to both languages at home and who tend to communicate in both languages, the so-called simultaneous bilinguals, show a wide range of linguistic skills in English and Spanish by the time they reach the school-age years. However, their exposure to and use of Spanish makes even the most English-fluent members of this group different from their monolingual peers. In environments that do not foster the development of the child's first language, language attrition occurs. Some language patterns attributable to language attrition are similar to patterns seen in children with language disorders (e.g., gender errors) (Restrepo, 1998; Anderson, 1999).

Given the large linguistic variability in the population, differentiating between a language difference (expected community variation) and a language disorder (communication that deviates significantly from the norms of the community; Taylor and Payne, 1994) is not simple. Our existing literature on language development in Latino children focuses primarily on a limited number of grammatical structures used by monolingual Spanish-speaking children (Gutierrez-Clellen, 1998; Goldstein, 2000; Bedore and Leonard, 2001). However, most Latino children are either bilingual or in the process of becoming bilingual, and therefore existing normative data on language acquisition by monolingual children (e.g., Miller and Leadholm, 1992; Sebastían and Slobin, 1994) do not accurately represent the language development of the majority of Latino children.

Assessments are further complicated by the fact that most of the available assessment protocols assume a high degree of homogeneity of exposure to the content of test items and to the sociolinguistic aspects of the testing situation. The cross-cultural child socialization literature suggests that Latino children's home routines are not always compatible with the content or the routines typically required in a language assessment (Iglesias and Quinn, 1997). Thus, poor performance on a particular assessment may reflect lack of experience rather than a child's inability to learn (Peña and Quinn, 1997).

The growing number of Latinos and their overrepresentation in statistical categories that place children at higher risk for disabilities or developmental delays (Arcia et al., 1993; Annie E. Casey Foundation, 2000; Iglesias, 2002) make it imperative that our assessment protocols not only accommodate linguistic differences across groups but also takes into account children's experiences. A variety of assessment protocols that take into consideration the languages and dialects spoken and the child's experiences have been suggested (Erickson and Iglesias, 1986; Peña, Iglesias, and Lidz, 2001; Wyatt, 2002). These protocols suggest the judicious use of standardized tests, taking into consideration norming samples and possible situational test biases, and consideration of alternative nonstandardized assessment procedures such as ethnographic analyses, criterion-referenced assessments, and dynamic assessments. Further, consistent with IDEA regulations (Individuals with Disabilities Act Amendments of 1997) and ASHA's position statement on the assessment of cultural-linguistic minority populations (American Speech-Language-Hearing Association, 1985), the assessments need to be provided and administered in the child's native language. In many cases this will require the examiner to be bilingual in Spanish and English or will require the use of qualified interpreters (Kayser, 1998).

The interpretation of the assessment results must take into consideration the growing literature on language development in Latino populations (Goldstein, 2000); with recognition that performance may differ from the expected norm because of a language difference rather than a language disorder. Although most children will show normal development in one or both languages, some will demonstrate weaker than expected performance in one or both languages. The data obtained must be carefully examined in the context of the languages and the dialects to which the child has been exposed and the experiences the child brings to the testing situation.

Intervention, if warranted, will require a culturally competent approach to services delivery in which the families' belief systems, including views on disability, are respected and intervention approaches are culturally and linguistically congruent with those of the children's families (Lynch and Hanson, 1992; Maestas and Erickson, 1992; van Kleeck, 1994; Iglesias and Quinn, 1997). The language of intervention should be based on the children's linguistic competencies, parents' preference, and functionality, not on the clinician's lack of proficiency in speaking the child's language. The bilingual literature on typical and atypical language learners strongly supports the notion that intervention should be conducted in the child's strongest and most environmentally functional language (Gutierrez-Clellen, 1999). In many cases, this will mean using Spanish as the language of intervention. The skills gained in the acquisition of the first language will facilitate acquisition of the second.