Prosodic Transfer in L2 Speech: Evidence from Phrasal Prominence and Rhythm

The current study examines focus and Nuclear Pitch Accent in the second language (L2) speech of Spanish speakers of English. We begin with a discussion of the distinct algorithms said to account for Nuclear Stress (NS) at work in Spanish and English. We then match the theoretical claims with experimental data on the placement of Nuclear Accent for various focus constructions from native speakers of English, followed by a contrasting set of data from L2 speakers. However, the L2 speakers are not a homogenous group, and indeed the production behavior across proficiency groups gives rise to the hypothesis that the acquisition of nuclear pitch accent is hierarchical in nature: learners acquire anaphoric deaccenting before nuclear stress. We conclude with a discussion of further data that supports the notion that phrasal prosodic patterns might be related to some fundamental rhythmic difference between the two languages. 1. Cross-linguistic differences in NS This paper intends to contribute to our understanding of certain aspects of prosodic transfer in SLA, an area that has only recently begun to be a focus of serious research (cf. Goad &White 2006; Tremblay 2007). More specifically, we look at native Spanish speakers as they progress towards native-like production of Nuclear Stress (NS) in English (their L2), and how this progression toward prosodic representation might be informed by underlying rhythmic/metrical patterns at work in each language. Based on evidence from second language (L2) speech, we argue that the computation of nuclear pitch accent in English arises from the complex interplay of several distinct algorithms. At the heart of this computation is the Nuclear Stress Rule. Due to the pervasive phenomenon of “transfer” of the grammatical properties of the first language (L1) into L2 speech, the latter can constitute a window into the mental representation of the two systems in contact. 1.1. Nuclear Stress: Algorithms at work In Spanish (or more generally in the Romance languages), main phrasal prominence (or NS) aligns with the last word of the Intonational Phrase (IntP) in wide-focus contexts (Zubizarreta 1998, Sosa 1999). As is well known, this is not always the case in English (or more generally in the Germanic languages). The Germanic NS algorithm in (1) applies to a metrical structure, which is a proper set of the syntactic structure. 1 This is not true of emphatic or contrastive stress, which can even target a subpart of a word. (1) Given two sister nodes A and B in a binary metrical structure, if B is selected by A, then assign NS to B. Otherwise, assign NS to the rightmost node. 2 The application of the NS algorithm gives rise to a metrical grid, which can then be further manipulated by other rhythmic related-rules (as in Halle & Vergnaud 1987) We can appreciate the difference between the Germanic NSR and the Romance NSR in the unaccusative and transitive compound data sets. The rule given in (1) correctly captures the patterns given below. In transitive structures, NS falls on the object, whether or not the object has been incorporated into the verb, i.e. whether the order of elements is VO or OV. (The parentheses indicate the metrical structure, italics the metrically invisible constituents, and the underline the word that bears NS.) (2) a. Every summer, (Mary (watches birds)). (SVO) b. Every summer, (Mary (goes (bird-watching)). (SOV) Unaccusative structures provide further support for the Germanic – Romance distinction. In such structures, the subject and the verb are metrical sisters and the subject is the argument in the selectional relation, and therefore the most prominent: (3) a. The magician disappeared. b. The mail arrived. It is to be noted that in the unaccusatives structures, the presence of an adverb to the left or to the right of the verb pulls the NS to the right. This follows from the fact that the adverb and the verb are sister nodes in the metrical tree but they are not selectionally related. (4) a. (A dog (mysteriously disappeared)). b. (A dog (disappeared mysteriously)).