Volume 26, Issue 1

This article investigates how morphological borrowing can take place in situations of language contact, and which factors play a role in this process. The author proposes that synthetic languages can use three strategies to form the plural of borrowed nouns and adjectives: oikomorphological, xenomorphological, and allomorphological. Data collected for varieties of Friulian in contact with different languages show that borrowed nouns can undergo all three types of morphological integration. The use of the xenomorphological strategy, which consists of forming the plural as in the donor language, is becoming widespread through generations of speakers of Friulian living in Italy. Basing on a comparison with the inflection of nouns and adjectives that Friulian has borrowed directly and indirectly from Italian and other languages, the author suggests that, in bilingual communities, the insertional codeswitching of inflected nouns and adjectives could be the “Trojan horse” that makes possible inflectional morphological borrowing.

One of the challenges of child language research is to identify the relevant factors that play a role in the acquisition course of a particular linguistic feature. This article analyzes the role of stress, word position, and word length in the acquisition of coda consonants by Catalan and Spanish children. The fact that the two languages differ substantially in their coda distribution (e.g., stressed word-final codas are more frequent in Catalan than in Spanish) will allow us to test the potential effects of coda distribution in the target language on early coda production. Sixteen Catalan-dominant and Spanish-dominant two-year-olds from the Barcelona area participated in two elicitation tasks with both novel and familiar words of different phonological shapes. Coda productions were assessed in stressed vs. unstressed syllables, in word-medial vs. word-final syllables, and in monosyllables vs. polysyllables. Results showed that the distributional difference between coda consonants in the two languages has crucial effects on the children's coda production. That is, Catalan-dominant children produce significantly more stressed word-final codas than Spanish-dominant children. This result lends support to the idea that when prominence is held equal, as in a controlled experiment, there are still crosslinguistic differences in coda production that are consistent with the frequency distribution of coda consonants in the respective languages, which means that children are very closely attuned to the frequency patterns of prosodic structure in the input language and are aware of their specific distributions across the lexicon.

This paper investigates the phonetics and phonology of initial peaks and final falls in wh-questions produced by speakers of the variety of Spanish spoken in the Castile-La Mancha (Manchego) region of Spain. The acoustic analysis is based on speech data for nine speakers, and the goal is to identify how utterance-initial and utterance-final F 0 gestures relate to broader issues in intonational phonology and the prosodic signaling of wh-questions. The findings for left periphery constituents provide evidence for a H tone at the utterance boundary for all speakers, although the exact autosegmental representation cannot be provided due to variability in peak alignment patterns. The findings for right periphery constituents indicate two distinct speaker groups based on nuclear syllable and posttonic gestures. Specifically, the continuum of final falls is motivated by contrasting bitonal nuclear pitch accent configurations: H + L* vs. ¡L + H*. The boundary L% specification is argued for all speakers in spite of seemingly divergent posttonic gestures. The experimental findings speak to cross-linguistic issues such as prominence marking in wh-question intonation, the syntax-prosody interface in wh-questions, and the internal structure of pitch accent configurations.

The debate over whether preverbal subjects of transitive verbs are canonical or non-canonical elements in Iberian Romance and other null-subject languages over the past twenty-plus years has arguably left the field of Generative syntax at an impasse with respect to the issue. Researchers have argued quite convincingly that preverbal subjects are non-canonical, non-argument (A′) constituents (e.g. Barbosa 1996, 2000; Ordóñez and Treviño 1999; Uribe-Etxebarria 1992, 1995), while others have presented equally compelling evidence supporting the analysis of preverbal subjects as canonical argument (A) constituents (e.g. Costa 2004; Goodall 2001). The results of this debate are of great theoretical importance, as they been claimed to have implications for the universality of the EPP and its associated feature(s). In this paper, I present evidence for Galician in the spirit of previous research and show that, with respect to the standard syntactic tests of A-versus A′-status for preverbal subjects in the literature, Galician is also inconclusive. I discuss how an experimental methodology incorporating a structured discourse context may provide greater insight on the issue and, assuming López's (2009) analysis of CLLD elements, present evidence suggesting that preverbal subjects are not left-peripheral, A′-elements in the discourse contexts examined in Galician, but rather A-elements.