Phonological spell-out of Spanish/English word internal code-switching
This paper examines phonological spell-out of Spanish/English word-internal code-switching. Specifically, we empirically testthe claim that a code-switched word (i.e., a word comprised of morphological elements from two different languages) cannot contain phonological elements from two languages (MacSwanColina, 2014; Rao & DenDikken, 2014). For example, consider the code-switched word in (1) comprised of the Spanish verbal root utiliz from utilizar ‘to use’ and the German derivational and inflectional affixes –ier- and –en, respectively. According to a German/Spanish bilingual consultant, the word in (1) is unacceptable if it evidences phonological material from both Spanish and German (2) as well as if it evidences phonological material from just Spanish (3). However, the word in (1) is considered acceptable if it evidences only phonological material from German (4). That is, the code-switched word evinces the phonology of the language of the affixes (González-Vilbazo & López, 2011). We examine these claims viabilingual production of code-switched verbs that have an English root containing a segment that is phonemic in English but not in Spanish (in this case /z/), and Spanish affixes.
Ten early Spanish/English bilinguals that are self-reported code-switchers participated in three experimental sessions on separate days: A bilingual CS session, and monolingual English and Spanish sessions. A picture decision perception task was administered in the English session to confirm perception of the English/s/ and /z/ contrast, and each session comprised an elicited production task. Data from the monolingual production tasks were used to establish a baseline to compare against the CS data. In the production tasks, speakers were auditorilypresented with nonce verbs in the infinitive and prompted to produce them with progressive morphology. The Spanish task stimuli included nonce verbs with /s/ in onset position (e.g. sintear, n = 10) and /s/ in a voicing assimilation context (/s/ followed by a voiced consonant, e.g. mesgear, n = 10). The English task stimuli were nonce verbs with /s/ and /z/ in onset position (e.g. ‘sart’, n = 10, ‘zab’, n = 10). The CS task stimuli were English nonce verbs with /z/ in onset position (e.g. ‘zert’, n = 10). The CS task was administered in code-switching and was designed to elicit morphologically switched verbs consisting ofthe English verbal root combined with Spanish progressive verbal inflection (e.g. estázerteando, see Example 5 for a sample trial). Following González-Vilbazo and López (2011), Spanish/English bilinguals were expected to produce /z/ in the code-switched words as [s] since it is not a phoneme in Spanish (e.g. [s]erteando).
To distinguish between [s] and [z] production in the three tasks, critical segments were analyzed for percentage of voicing (e.g., Schmidt & Willis, 2011). Preliminary results indicate thatthe majority of participants produced /z/in the CS task with (at most) initial voicing, similar to productions in the monolingual Spanish task. This finding points toapplication of Spanish phonology and thus provides preliminary support for the ban on word-internal phonological switches. However,for the remaining participants, CS /z/ was produced with majority voicing (>66%). While the production of [z] in the CS condition could a priori be taken as evidence against a ban on word-internal phonological switches, it must be noted that the participants who produced /z/ with majority voicing in CS also produced segmentswith majority voicing in the voicing assimilation context in monolingual Spanish. The question is, does [z] come from English phonology, or is Spanish phonology being applied, in which case these bilinguals redeploy Spanish [z] from coda position to onset position? Impressionistic analysis of the code-switched words with [z] suggests that English phonology is not applied elsewhere in the verbal root, which suggests that the source of [z] is Spanish and upholds the proposed ban. Further examination of the production data, which crucially includes English phonemes with no corresponding positional allophone in Spanish(/θ/ and /ɪ/), is expected to provide additional evidence to uphold a theory in which words comprised of morphological elements from two different languages evidence only the phonology of the language of its affixes, suggesting that the word is the minimal unit for codeswitching.
Examples:
(1)Utilisieren
(2)*[ut̪ilisiːʁn̩]
(3)*[ut̪ilisi̯eɾen]
(4)[ʡʊtʰɪliːziːʁn̩]
(5)Slide 1: “Repite por favor‘please repeat’. To zert”
Slide 2: “To zert es cuando bailas‘is when you dance’ to your favorite song in an emptyroom. Angela lives in a studio apartment and she zerts every night. Qué está haciendo en la foto?‘What is she doing in the photo?’
References:
Bandi-Rao, S., & den Dikken, M. (2014). Light switches: On V as a pivot in codeswitching, and the nature of ban on word-internal switches.In J. MacSwan (Ed.), Grammatical theory and bilingual codeswitching(pp. 161-183). Cambridge, MA: MIT Press.
González-Vilbazo, K., & López, L. (2011). Some properties of light verbs in code-switching. Lingua, 121(5), 832-850.
MacSwan, J., & Colina, S. (2014). Some consequences of language design: Codeswitching and the PF Interface.In J. MacSwan (Ed.), Grammatical theory and bilingual codeswitching(pp. 185-210). Cambridge, MA: MIT Press.
Schmidt, L. B., & Willis, E. W. (2011). Systematic investigation of voicing assimilation of Spanish/s/in Mexico City. InSelected proceedings of the 5th Conference on Laboratory Approaches to Romance Phonology(pp. 1-20). Somerville, MA: Cascadilla Proceedings Project.