|Abstract:||In acquiring a second language (L2), it has been noted that the phonetics and phonology of an L2 can be difficult for learners to acquire and many sound structures from a learner’s first language (L1) are transferred into the L2, resulting in pronunciation patterns that diverge from the native norms. Pervasive transfer can at times be evident even after the speaker has reached high levels of proficiency. However, not all sound structures pose learners an equal amount of difficulty or are equally subject to persistent transfer effects. Various frameworks have been proposed to predict which sounds will be difficult for a learner based on a given L1-L2 combination (e.g. Best, 1994, 1995; Escudero, 2005, 2009; Flege, 1987, 1995). These traditional models generally account for the acquisition of segments only. However, in more recent years, there has been increased attention given to L2 acquisition of prosody in the field. This body of research has been increasing drastically and includes various aspects of the L2 acquisition of prosody, including the acquisition of prosodic structures from the word level (e.g. Face, 2005; Guion et al., 2004; Nguyen et al., 2008; Zubizarreta et al., 2013) up until the phrase level (e.g. Rasier & Hiligsmann, 2009; van Maastricht et al., 2016; Zubizarreta & Nava, 2011) and considers both the acquisition of the phonological (e.g. van Maastricht et al., 2016; Zubizarreta, 2014) and phonetic (e.g. He et al., 2011) aspects of prosody. Today, thanks to the expanding number of studies available on the L2 acquisition of prosody, a better understanding of this process is beginning to emerge. While frameworks have been proposed in order to predict the relative difficulty of prosodic structures (Albin, 2015; Mennen, 2015; So & Best, 2014; Zubizarreta & Nava, 2011), a broader understanding of the acquisition of L2 prosody still must be achieved before we have a complete picture on the development of L2 prosody. This is no easy task, as prosody, compared to segments, is more intricately intertwined with the meaning of words or utterances. Additionally, prosodic events are often more difficult to measure. Because prosody is so multi-faceted, there are still many research questions needing to be addressed in this field.
The studies in this thesis examine the phonological acquisition of prosody at the level of the phrase. Specifically, L2 acquisition of nuclear accent (NA) placement is investigated in order to better understand which contexts of NA placement and shift in an L2 are more difficult to acquire and why. Two experiments are presented, which examine both L2 English and L2 Spanish. These two languages are interesting to compare as the grammatical rules behind NA placement vary predictably in the two languages. English is categorized as a prosodically plastic language (Vallduví, 1991) as it exhibits flexible NA placement. Generally, NA in English occurs on the rightmost content word, but can move leftwards under certain conditions. Spanish, on the other hand, is often referred to as a non-plastic language. NA in Spanish tends to occur more invariably on the rightmost content word. Traditional theories on Spanish grammar posit that the nuclear accent is not as flexible as in English and can only move leftwards to mark contrastive focus (e.g. Zubizarreta, 1998). Instead of employing prosodic means, Spanish relies on word order inversions to mark some of the distinctions English marks prosodically. By including these two languages, these two studies are able to examine how L1 speakers of a non-plastic language acquire a plastic grammar, how speakers of a plastic language acquire a non-plastic language, and how the acquisition of non-plastic prosody develops in relation to the acquisition of word order inversions.
In order to address these questions, two semi-parallel experiments were conducted. The first, referred to as the English Experiment, tests L1 speakers of Spanish who are acquiring English as an L2. The second, known as the Spanish Experiment, examines L1 speakers of English learning L2 Spanish. The English Experiment includes an Oral Production Task, in which nuclear accent shift is elicited in various contexts, including in compounds, utterances with a final indefinite pronoun, utterances with contrastive focus on a non-final word, and in broad focus intransitives. These various contexts of NA shift, many of which intersect with other domains of grammar, are included in order to examine which factors make acquisition of NA shift more difficult or which areas may be more susceptible to transfer. The Spanish Experiment includes both a Word Order Task and an Oral Production Task. The Word Order Task is designed to test learners' preference of VS word order in context of narrow focus and broad focus intransitives. These same contexts were included in the Oral Production Task in order to examine whether learners are transferring any of the predicted L1 prosodic patterns into their L2.
Results indicate that it is more difficult for the L1-Spanish/L2-English speakers to acquire a plastic prosodic system than it is for the L1-English/L2-Spanish speakers to acquire a non-plastic prosodic system. In L2 English, learners were most accurate at producing NA shift in compounds. NA shift to mark contrastive focus was produced with the second highest level of accuracy, with contexts of NA shift for utterance-final indefinite pronouns and in intransitives as seemingly the most difficult to acquire. Both transfer and difficulty of acquiring phenomena at the interfaces are explored as potential explanations for these patterns. The L2 speakers of Spanish, in comparison, produced prosodic patterns similar to the L1 speaker group in both narrow focus and broad focus constructions. There were no effects of proficiency, suggesting learners were able to learn Spanish prosody relatively quickly. The syntactic acquisition of word order inversions, on the other hand, were less target-like with learners showing an overwhelming preference for SV word orders, regardless of the context.
Findings from these studies have theoretical implications for the trajectory of acquisition of prosody at the phrase level and for the role of transfer versus universals of acquisition. Additionally, the findings from the Spanish Experiment present implications for the relationship between the acquisition of syntax and phonology. Such theoretical implications are discussed in relation to the findings of these two studies and the existing frameworks are evaluated in light of this new data.