This chapter discusses the Ontogeny Phylogeny Model (OPM), which focuses on the formation and development of second language phonological systems. It proposes an interrelationship between L2 native-like productions, L1 transfer, and universal factors. The model argues that chronologically, and as style becomes increasingly formal, L2 native-like processes increase, L1 transfer processes decrease, and universal processes increase and then decrease. It further claims that the roles of universals and L1 transfer are mediated by markedness and similarity, both of which slow L2 acquisition. Specifically, in similar phenomena L1 transfer processes persist, while in marked phenomena universal processes persist. The OPM also argues that these same principles obtain for learners acquiring more than one L2, monolingual and bilingual acquisition, and L1 attrition. In addition to the chronological stages and variation of the individual learner, the model claims that these relationships hold true for language variation and change, including pidgins and creoles.

This chapter brings together the findings from the three studies, which confirmed that Differential Object Marking (DOM) is a vulnerable grammatical area not only in Spanish, but also in Hindi and Romanian as heritage languages, subject to erosion under pressure from, English in this case. This chapter goes deeper into these overall trends, by comparing the three heritage speaker groups, on the one hand, and the three first-generation immigrant groups, on the other, on several background variables related to patterns of language use. A follow-up replication study with Spanish heritage speakers and immigrants from other countries in Latin America is reported, which confirm the attrition effects in the two generations of Mexican immigrants. It is claimed that his finding is strong evidence that while DOM omission may have started as a developmental outcome of heritage language acquisition, it may be on its way to becoming a stable dialectal feature of Spanish in the United States, suggesting language change with respect to DOM in Spanish. The roles of language internal and language external factors are discussed.

Compared to second language speakers who acquired the language later in life, heritage speakers often show native patterns in many areas of their grammar. Nevertheless, heritage speakers’ ultimate attainment in early adulthood is characterized by significant variability in their overall linguistic proficiency of the heritage language, which often falls outside the range of variation seen in monolingually-raised native speakers. Their native language abilities are dissociated by receptive and productive linguistic skills and within the language modules themselves. This chapter provide examples of the structural changes and differences commonly found in heritage language grammars, compared to the grammars of first-generation immigrants, who speak the same variety as the heritage speakers and who are the most comparable baseline. Heritage grammars are characterized by structural simplification, influence of the dominant language and slow and labored language processing during comprehension and production. Quantity and quality of input may drive processes of language change across generations. Although it has been suggested that changes observed in first-generation immigrants due to attrition are directly transmitted to heritage speakers it is argued that direct transmission from first to second generation is unlikely when considering the interaction of the intensity of the exposure and age of acquisition.

This chapter asks whether DOM—which is a vulnerable grammatical area in Spanish in the United States—is also vulnerable in Hindi as a heritage language. The results of the study presented in this chapter show that some Hindi heritage speakers also display omission of DOM in all tasks. But unlike what was found for the Spanish-speaking Mexican immigrants in the Spanish study discussed in the previous chapter, there is no indication of ongoing language change in the Hindi spoken in the homeland nor apparent signs of attrition of DOM in the Hindi-speaking adult immigrant group. The sociolinguistic characteristics of the Hindi/Urdu-speaking population in the United States is discussed. The results of the linguistic background questionnaire and the linguistic tasks (oral narrative task, elicited production task, written task, bimodal acceptability judgment task, auditory/written comprehension task) are presented and discussed.

The acquisition of DOM in Spanish monolingual children and Spanish-speaking bilinguals has been intensely investigated in the last two decades. One of the main objectives of the in-depth study of DOM conducted for this project was to confirm and evaluate the strength of previous findings of DOM omission in Spanish-English bilingual children and in young adult heritage speakers. If erosion of DOM in Spanish as a heritage language is indeed a robust phenomenon, it should be documented in other groups of heritage speakers. Although hypotheses about the potential role of the parental generation in contributing to heritage language grammars abound, at the time the Spanish study reported in this book was conceived, there were no studies of potential attrition of DOM in adult immigrants. Therefore, the first-generation immigrant group was included to put the language transmission hypothesis to the test. This chapter presents a brief overview of Spanish in the United States before summarizing the main results of the language background questionnaire, and of the production, comprehension and judgment tasks described in Chapter 5.