Speaking Spanish in the US. Janet M. Fuller
ignore the structural constraints on parental agency. As Zentella (1997a) demonstrated in her landmark ethnographic study of language shift among Puerto Ricans in New York, racism, linguistic subordination and economic precarity are key factors constraining parental language decisions and impacting language shift. Similarly, a person’s desire to maintain Spanish may not be sufficient to make that happen and, despite common feelings of guilt regarding language loss, individual agency is constrained by structural factors. Thus, in addition to looking at social networks, individual factors and personal decisions, we shouldn’t lose sight of the ways in which societal language ideologies and policies impact groups as well as the people within them.
Spanish Language Maintenance and Shift in the US
Historically, immigrants to the US have largely followed a three-generation pattern of language shift to English, with the immigrant generation being mostly dominant in the non-English language, the second generation being bilingual, and the third (and subsequent) generations being monolingual in English (Alba, 2004; Rivera-Mills, 2012; Veltman, 2000). However, some observers have expressed doubts that these patterns would continue to hold for Spanish, given the large number of speakers of Spanish and their relative high density in certain parts of the country – demographic factors that the ethnolinguistic vitality framework posits should contribute to maintenance. In some cases, these doubts are rooted in racially inflected anxiety and entangled with the unfounded and inaccurate claims that Spanish-speakers do not want to learn English, such as we saw in the Tom Brokaw quote presented earlier in this chapter, as well as a broader discourse about Latinxs’ supposed unwillingness and inability to assimilate (Chavez, 2013). Further, some people have suggested that greater ‘tolerance’ of multilingualism and multiculturalism makes it less likely that recent immigrants would assimilate linguistically. We examine patterns of Spanish maintenance and loss in the next sections.
Shift to English
In our earlier discussion of ACS data, we compared the English ability of immigrants and US-born Latinxs and Spanish-speakers, but we did not take into account how many generations someone’s family had been in the US. In order to get a better sense of the intergenerational patterns of language transmission and use, Alba (2004) did just that. Specifically, he compared the ACS language data for first-, second- and third-generation immigrants (i.e. those who immigrated themselves, the children of immigrants and the grandchildren of immigrants). He found that virtually all US-born Latinxs speak English well, but he noted an important difference between the second and third generations in terms of their Spanish use: 85% of the second generation spoke Spanish at home but only 18% of the third generation did. Alba interpreted this as evidence of the endurance of the three-generation pattern of shift to English.
Based on the high percentage of people in the second generation reported to speak Spanish at home, Alba also concluded that second-generation bilingualism is widespread. However, while at first glance this conclusion seems reasonable, it must be taken with a pinch of salt. As we discussed above, the Census Bureau doesn’t ask how well individuals speak Spanish (see Figure 2.2), and thus the count of ‘Spanish-speaking’ persons includes people with only limited Spanish-speaking ability as well as those who are completely proficient (or Spanish dominant), and there is no way of knowing the relative proportion of each. Because even individuals with very limited Spanish ability are included, Census Bureau statistics may give an inflated sense of Spanish maintenance and Spanish–English bilingualism (Leeman, 2018c). This is also the case for research that relies on the Census Bureau’s statistics, such as the Instituto Cervantes’ reports on the number of Spanish-speakers in different countries around the world (e.g. Instituto Cervantes, 2018). For this reason, in order to get a more accurate picture of patterns of language knowledge among the second and subsequent generations, we turn to data from the large-scale surveys of Latinxs carried out by the Pew Research Center, a non-governmental organization that conducts surveys and public opinion polls on a wide array of political and social issues.
In contrast with the Census Bureau’s ACS, the Pew Center surveys ask how well respondents speak both Spanish and English. By inquiring about speaking ability in both languages, the Pew Center is able to compare respondents’ ability in Spanish and English, and thus get a better sense of where they fall on the bilingual continuum. Regarding English, the Pew Center’s findings mirror those of the Census Bureau in that they show that nearly all US-born Latinxs are proficient in English (Taylor et al., 2012). But of greater interest for the question of Spanish maintenance are the data for the responses from the Spanish-speaking ability question. The results show that Spanish knowledge diminishes between generations; 82% of the so-called second generation (i.e. the US-born children of immigrants) speak Spanish ‘very well’ or ‘pretty well,’ but only 47% of the third generation do (Taylor et al., 2012).
Lopez et al.’s (2017) analysis of a subsequent Pew Center survey shows a similar trend. In that analysis, researchers compared individuals’ speaking ability in English and Spanish: respondents who reported speaking both languages ‘very well’ or ‘pretty well’ were considered bilingual, and the remainder were classified as either ‘English dominant’ or ‘Spanish dominant’ (so people classified as ‘dominant’ in a language also included monolinguals). The foreign-born were most likely to be Spanish dominant, although almost one-third of them were bilingual. In the second generation, almost everyone was bilingual or English dominant, and people in the third generation or higher were most often English dominant, with only a quarter reporting speaking both languages well (see Table 2.3).
| Table 2.3 English- and Spanish-speaking ability among Latinxs | |||
| English dominant or monolingual | Bilingual | Spanish dominant or monolingual | |
| Foreign-born | 7% | 32% | 61% |
| Second generation | 43% | 51% | 6% |
| Third or higher generation | 75% | 24% | – |
| All Latinxs | 28% | 36% | 36% |
| Source: Lopez et al. (2017). |
Taking into account these data as well as the ACS statistics on English knowledge and home use that we presented earlier, we have seen that almost all Latinxs born in the US are proficient in English, and the vast majority are either English dominant or bilingual, while the grandchildren of immigrants are either monolingual in English or English dominant. This pattern is consistent not only in studies based on Census Bureau and Pew Center data (e.g. Alba, 2004; Lopez et al., 2017; Taylor et al., 2012; Veltman, 2000), but also ethnographic research (e.g. Schecter & Bayley, 2002; Zentella, 1997a) and studies using a combination of surveys and interviews (e.g. Bills et al., 1999; Porcel, 2006; Portes & Rumbaut, 2001; Rivera-Mills, 2001). In other words, among the children and grandchildren of Spanish-speaking immigrants, the tendency for language shift to English parallels the linguistic trajectories of previous immigrant groups. This pattern holds even in Miami, with its high density of Spanish-speakers and their relatively higher socio-economic status (Carter & Lynch, 2015).
The well-documented pattern of language shift among the children and grandchildren of Spanish-speaking immigrants demonstrates that the apparent vitality and continued presence of Spanish has been the result of continued immigration, rather than intergenerational language transmission and maintenance (Jenkins, 2018; Rumbaut, 2009; Silva-Corvalán, 2004; Veltman, 2000). Further, in addition to disproving the enduring myth that Spanish-speaking immigrants and their offspring don’t learn English, the pattern of intergenerational shift makes it clear that it is Spanish, not English, that is at risk. Nonetheless, despite the universality of English