Migration

Migration has a major impact on multilingualism and multiculturalism throughout the world. In 2005, there were nearly 200 million international migrants (9.2 million of whom were refugees) around the world, a significant increase from 82 million in 1970 (Global Commission on International Migration for the United Nations, 2005, http://www.gcim.org/en/). Sollors (2009) reported that there are 56 million migrants in Europe, 50 million in Asia, 40 million in North America, 16 million in Africa, 6 million in Latin America, and nearly 6 million in Australia (making up 18.7% of the total population of Australia). In Japan, the number of expatriate workers increased from 750,000 to 1.8 million between 1975 and 2001. With China and India gaining economic momentum, it is likely that there will be a large migratory flow to these countries also in the near future, which will create new forms of multilingualism (Sollors, 2009).

In the United States, 75% of population growth between 1995 and 2000 was the result of migration (Sollors, 2009). At the time of the 2007 U.S. Census, more than 55.4 million (20%) people older than 5 years spoke languages other than English (Shin & Kominski, 2010). Although the predominant non-English language was Spanish (62% of those who spoke a language other than English), 19% spoke another Indo-European language (e.g., German, Italian), 15% spoke an Asian and Pacific Island language (e.g., Chinese, Tagalog, Vietnamese, Korean, Japanese), and 4% spoke another language (e.g., Haitian Creole). The majority of people in each of these language groups reported that they spoke the other language “very well” (Shin & Kominski, 2010). The uneven distribution of speakers of languages other than English in the United States reflects the continuous flow of non-native populations, with about 500,000 to 1 million migrants becoming U.S. citizens annually (U.S. Census Bureau, 2000). Sollors (2009) states that in this era of global migration and regional cultural and linguistic diversity, many countries are endorsing educational curricula that encourage the learning of languages other than the respective country’s official one. This is a challenging goal because there is still a dearth of knowledge about multilingual speech and language acquisition, particularly regarding whether multilingual children show similar patterns of acquisition to those of monolingual speakers.

Although immigrants in a given country differ linguistically, culturally, and socioeconomically, many share common characteristics, such as insufficient knowledge of the languages, dialects, and accents spoken by the residents in their new country (Cheng, 2004). Cheng reports that immigrants often have different medical and social experiences as well as different educational backgrounds and home-language literacy skills compared with residents in their new country. In addition, immigrants may experience stress from immigration depending on the circumstance of the reason for migrating. Consequently, Cheng and Butler (1993) report that immigrants may experience tiers of adaptation in the country in which they land:

It seems that a core underlying issue of migrant dissatisfaction is difficulty communicating (Grech & Cheng, 2010). Different nations respond to multilingual issues relating to migration in different ways. European Union (EU) educational policies and strategies have attempted to address diverse cultural and linguistic characteristics by offering support services and expanding resources for optimal learning of all children, including those of ethnic and migrant communities. National curricula are increasingly endorsing two languages, particularly in many European bilingual contexts (Huguet & Lasagabater, 2007). For example, the 1995 White Paper (CEC, 1995) on teaching and learning proposed that EU citizens should be proficient in three European languages (i.e., mother tongue and two other EU community languages). Additionally, the European Commission (EC) is offering incentives to individuals with multilingual proficiency, including jobs (mobility opportunities) and opportunities for educational staff and student exchange. There is also a push toward multilingual educators and professionals. Lassagabaster and Huguet (2007) report the results of a transnational study among trainee or student teachers in nine bilingual areas/states in Europe, indicating that 70% of student teachers claimed to have high proficiency in a minority language; almost all of the student teachers reported high proficiency in the majority language. The multilingual areas/states involved in the study were Basque Country, Catalonia, Galicia, Valencian Community, Belgium, Friesland, Ireland, Malta, and Wales. Some of the student teachers were proficient in at least three languages (but this varied from one area/state to another; in some, more than 50% of the student teachers reported proficiency in at least three languages). The importance of multilingual educators (and SLPs) is reinforced by the fact that the course of language development in monolingual children may be different from that of bilingual children and that language acquisition is influenced by culture, attitudes, and beliefs.

International adoption

Within many Western countries, another significant group of migrants are children who are adopted internationally. For example, in the United States in 2009, 12,753 international adoptions took place, with the majority of children coming from China, Ethiopia, Russia, South Korea, Guatemala, Ukraine, Vietnam, Haiti, Kazakhstan, and India (Child Welfare Information Gateway, 2010). In previous years, even more children were adopted by families in the United States. In 2004, 22,990 children were adopted internationally. In 2008, when 17,229 international adoptions took place, the majority (48%) were of children 1 to 4 years old, an additional 34% were of children younger than 1 year, and the remainder (18%) were of children 5 years and older (Child Welfare Information Gateway, 2008).

Children who are adopted internationally are often referred to as second first language learners because they are neither monolingual nor multilingual (Pollock, 2007). Typically children who are adopted internationally begin learning one language in their home country, then at the time of adoption, exposure to their first language ceases, and they begin to learn another language. A meta-analysis of language outcomes for children who were adopted internationally showed that they were more likely to have poorer language outcomes; however, there was great variation (Scott et al., 2011). One mediating factor was the age of adoption; there was a slight trend in favor of children who were adopted at younger age groups. Gauthier and Genesee (2011) also indicate that children adopted from China tended to have better language outcomes than children adopted from countries in which deprivation was more common.

Phonological development

In the late 1960s, Roman Jakobson offered the following statement regarding children’s speech acquisition across the globe: “Whether it is a question of French or Scandinavian children, of English or Slavic, or Indian or German, or of Estonian, Dutch or Japanese children, every description based on careful observation repeatedly confirms the striking fact that the relative chronological order of phonological acquisitions remains everywhere and at all times the same . . . the speed of this succession is, in contrast, exceedingly variable and individual . . .” (Jakobson, 1968, p. 46, emphasis added). This statement, although appealing in its simplicity, has been proved on numerous occasions not to be true. A more current understanding is that children’s speech acquisition is influenced by a complex interrelationship between phonetic complexity, functional load, and phonetic frequency and this varies by language (see Ingram, 2011, for an overview). Phonetic complexity or ease of production refers to the articulatory difficulty or ease producing a sound. For example, fricatives are said to be more complex than stops because they require finer motoric movements to produce, have lower acoustic saliency, occur less frequently in babbling, and so on. Functional load refers to how often a sound contrasts with other phonemes (Meyerstein, 1970). For example, in English /ð/ (voiced th) has a low functional load (because it is in only a few words); however, in Greek it has a high functional load. Phonetic frequency refers to how often a sound occurs in spoken language. For example, in English /ð/ has high phonetic frequency because words such as the, this, and that are used frequently. Stokes and Surendran (2005) demonstrated the relationship between these three factors for different languages. They found that a child’s age of emergence of sounds in Chinese was predicted by phonetic frequency, but in English it was best predicted by functional load. They found that accuracy of production was best predicted by phonetic complexity in English, but not in Dutch. Late acquisition of complex consonants was predicted by phonetic complexity in Arabic. It appears that, to date, one generalization can be made: “Phonemes with high functional load in a language will be acquired earlier by both typically developing children and by children with phonological deficits” (Ingram, 2011, p. 12). A comprehensive summary of monolingual speech acquisition for more than 20 languages is provided by McLeod (2010). However, little is known about the relative factors that influence multilingual speech acquisition for different languages, so SLPs need to constantly update their knowledge on this topic.

Lexical acquisition

Multilingual and monolingual children typically have similar numbers of words in their lexicon; although the multilingual children’s words are distributed across the languages that they speak (Peña et al., 2002). Some words are unique (termed singlets), whereas others overlap (termed translation equivalents) (e.g., a translation equivalent occurs if a child knows both the English and Spanish words for the same object: pen and pluma) (Hemsley et al., 2010). Perhaps unsurprisingly, 7-year-old French-English bilingual children were found to have more unique words (singlets) in their dominant language than their nondominant language (Paradis et al., 2003). It has been found that multilingual and monolingual children tend to use similar strategies for categorizing words (Peña et al., 2002) but will rely on context for words that they choose. For example, although most Spanish-English bilingual children said cake when asked to name foods at a party, when speaking Spanish children followed this by arroz (rice) and frijoles (beans), whereas when speaking English, they listed hamburger and hotdog.

Morphosyntactical and narrative acquisition

To communicate effectively, multilingual children need to follow the grammatical rules for each language. When language mixing occurs, children typically mix nouns with nouns and verbs with verbs (Paradis et al., 2000). Nicholls and colleagues (2011) conducted a study of the acquisition of English morphological skills by 148 3-year-old children: 74 were multilingual, speaking over 31 different languages in addition to English, and were 74 monolingual children matched for age. The multilingual and monolingual groups showed a similar developmental progression in the acquisition of morphology, with the same grammatical skills being more difficult for both groups; however, the multilingual children had a slower rate of acquisition. In another study, Cleave and colleagues (2010) assessed 26 children with specific language impairment; almost half the children were monolingual speakers, whereas the other children were dual-language learners, with English being the dominant language. They found that the multilingual children achieved higher scores on the production of language from using narrative assessment, rather than using standardized tests of morphosyntax. However, all the children scored below average on measures from narrative samples for productivity, narrative structure, literate language, and language form. Bedore and colleagues (2010) similarly found the usefulness of eliciting narratives to determine 170 bilingual Spanish-English children’s language abilities. They found that English mean length of utterance (MLU), as well as English and Spanish grammaticality, was the best predictors of language use in these bilingual children.