Dialectal Variation and SLA: Perception of Aspirated‐/s/ by L2 Learners of Spanish Lauren Schmidt Hispanic Linguistics Symposium 2010 Bloomington, Indiana
Objective of current study • The objective of the current study is to examine the second language acquisition (in terms of perception) of weakening of syllable‐final /s/, a regionally and socially variable feature found across several varieties of Spanish.
Weakening of /s/ in Spanish • /s/ in syllable‐final and word‐final position may be weakened, resulting in aspiration or deletion (Hualde 2005) (1) sibilance [s]: (2) aspiration [h] (glottal frication): (3) deletion:
[kos.ta] [koh.ta] [ko.ta]
• Other variants of syllable‐final /s/ have been reported: • postaspiration of following stop in Western Andalusia (Torreira 2006) • vowel laxing in Eastern Andalusia (e.g., Navarro Tomas 1939) • word‐internal vowel lengthening, in Cuba (Hammond 1978) and in Puerto Rico (Figueroa 2000)
1. [s], male Puerto Rico 2. [h], female Venezuela
3. Ø, male Puerto Rico
Weakening of /s/ in Spanish • Dialectal variation of s‐weakening (Hammond 2001) • s‐maintaining dialects: • northern and central Spain, • highland regions of America
• s‐weakening dialects: • southern Spain and Canary Islands, • lowland, costal, and insular regions of America • up to 50% of the varieties of Spanish spoken in the world are s‐weakening (Hammond 2001)
Factors relating to s‐weakening • Geographic – dialectal variation • Variation in degree of s‐weakening (Silva‐Corvalan 2001, Terrell 1980) and sociolinguistic attitudes relating to s‐weakening
• Sociolinguistic • Gender (e.g., Terrell 1981, Dohotaru 2004), Age (e.g., Cedergren 1973, Samper Padilla 1990), Socioeconomic level (e.g., Calles & Bentivoglio 1986)
• Linguistic • Position within the word (e.g., Samper Padilla 1990, Terrell 1978), Following segment (C, V, pause)
• Stylistic ‐ greater retention in more formal styles • Speech style (e.g., Lafford 1986)
• Lexical frequency effects (e.g., File‐Muriel 2009, Earl Brown 2007) • Functional (Terrell 1978, Poplack 1980, Ruiz‐Sánchez 2005)
Speech perception • Types of variation found in the speech signal: • • • •
Contextual variation (e.g., assimilation processes) Sociolinguistic and regional variation Stylistic variation Individual variation (e.g., different speech rates)
• Spoken word activation – dependent upon legitimacy of variation, not degree of variation (Sumner & Samuel 2005): • Native and second language (L2) listeners – must attend to acoustic variants that create meaning contrast while ‘accepting’ legitimate variants of a single phoneme
Speech perception • In the case of variants of Spanish syllable‐final /s/: • Legitimate variants of /s/ in s‐weakening dialects: [s], [h], Ø • Thus, [fi̯es.ta], [fi̯eh.ta], and [fi̯e.ta] should all activate the construct party (fiesta) (although the different variants may relate sociolinguistic information)
• Illegitimate variations of syllable‐final /s/ in Spanish, such as [fi̯ef.ta], would not be deemed as a possible form and should not be as productive in activating the lexical item fiesta
Speech perception: L2 • Native Language Magnet Model (Kuhl 1992, 1993, 1994): • Language‐specific perception • Distortion (“warping”) of acoustic space early in life from exposure to a particular language
• Perceptual Assimilation Model (PAM, Best 1994): • Perception of non‐native sounds shaped by the listener’s NL • Unfamiliar sounds are assimilated into native categories
• Speech Learning Model (SLM, Flege 1995): • Initially L2 learners may identify L2 sounds in terms of L1 allophones • With increased experience, possibility of enabling the formation of a new category for the L2 sound
English [h] • English glottal fricative /h/ • English /h/ is contrastive in syllable‐initial position, but does not occur in syllable‐finally (McMahon 2002) “hat” vs. “at” • English‐speaking learners of Spanish may not attend to Spanish syllable‐final [h]
• [h] does not occur as a legitimate variant of English /s/ as it does in s‐weakening dialects of Spanish “dust” *[dɅht] • L2 Spanish learners (L1 English) must learn to associate the aspirated variants with Spanish “s”
Spanish /s/‐weakening and SLA • Effect of dialect on L2 comprehension of Spanish (Schmidt 2009)
• Lower comprehension of speech characterized by Dominican dialectal features • Greater effect of some dialectal features (/s/‐deletion and /ɾ/‐lambdacism) on comprehension of Dominican Spanish than others
• Late acquisition of dialectal variants (Geeslin & Gudmestad 2008)
• Of 130 learners, only 9 produced Castilian /θ/ and only 5 demonstrated /s/‐weakening • Production of these dialectal variants occurred only at highest proficiency levels (Levels 4 & graduate students)
In sum • English‐speaking learners of Spanish may not map syllable‐final aspirated variants to Spanish /s/, at least not at initial stages of development • Influence of native system in perception of target language sounds (NLMM, PAM, SLM) • [h] does not occur in syllable‐final position in English • [h] is not a ‘legitimate’ variant of English /s/
• /s/‐weakening is a variable feature in Spanish • Subject to regional, social, linguistic variation • L2 learners may not be exposed to the feature • Variable structures acquired later
Research questions 1. What are the patterns of perception of Spanish aspirated‐/s/ variants by English‐speaking learners of Spanish? 2. Does perception of the aspirated variants change over time (proficiency level) and increased exposure to Spanish dialectal variation?
Methodology • Participants • English‐speaking second language learners of Spanish • 5 proficiency levels: beginning, intermediate, high intermediate, advanced, high advanced (graduate students)
• Native speakers of Spanish • /s/‐aspirating dialect, La Rioja, Argentina
• Instrument • • • •
Written grammatical proficiency test Language background questionnaire Identification task Lexical item recognition task
• Analysis
Participants (N=235) Group
N
Description
Native speakers, Aspirating region
20
La Rioja, Argentina, university students (12 female & 8 male; Mean age = 21.5 years)
Level 1 learners
57
Elementary Spanish (M proficiency score = 4.8)
Level 2 learners
52
Intermediate Spanish II (M proficiency score = 6.5)
Level 3 learners
59
300‐level Spanish literature and culture courses (M proficiency score = 8.7)
Level 4 learners
26
400‐level Spanish literature and linguistics courses (M proficiency score = 10.3)
Level 5 learners
21
University instructors of Spanish courses / graduate students in Spanish literature
Instrument • Language Background Questionnaire (3 versions): • Motivation: to measure proficiency level and exposure to s‐weakening dialects • Linguistic background – L1s, experiences learning L2s • Dialect exposure – places lived, location of study abroad, exposure to Spanish (TV, movies, music, conversation with NSs), origin of Spanish instructors and of other NS contacts
• Identification Task (Praat): • Motivation: to determine how listeners categorize the word‐internal syllable‐final aspirated variants • Participants hear non‐word stimuli (up to 2 times) and choose one of seven options to reflect what heard • Stimuli randomized across participants
Tasks – Identification Task • Demonstration of the Identification Task
Identification Task Stimuli • Disyllabic nonce words [CVC.CV] and [CV.CV] (N=158) • Target stimuli (N=14) • Nonce items with word‐internal s‐aspirated coda [CVh.CV] (14) [bah.pe], [dih.to]
• Control stimuli (N=64) • Nonce items with word‐internal full‐s coda [CVs.CV] (14) • Nonce items with other word‐internal coda contexts: lateral, rhotic, nasal, labiodental fricative, no coda [mis.po], [gal.ka], [sof.te]
• Distracter stimuli (N=80) • Target different positions and sounds, nonce items [fin.ko] [ne.ro], [ka.ti]
Identification Task Stimuli • Speakers (2): 1. Male from Caracas, Venezuela 2. Female from Buenos Aires, Argentina
• Two versions of task • Opposing aspirated‐/s/ – full‐/s/ pairs • Speakers read stimuli (at least) twice, two conditions
Full‐/s/ and aspirated‐/s/ stimuli pair
Analysis 1. Calculation of identification rates of the variants of /s/ • (1) full‐/s/ variants and (2) aspirated‐/s/ variants % = N variants identified as “s” total N stimuli (14) 2. Cross‐tabulation of patterns of identification of aspirated‐ /s/ according to group: • Argentine group • L2 learner group • According to proficiency level (Levels 1‐5)
3. Regression model to determine which dialectal exposure types reported in the LBQ best predict identification of Spanish aspirated‐/s/
Results – Identification of coda [s] and coda [h]
**
Graph 1. Argentina group (N=20) t(19) = 1.996, p = .06
Graph 2. L2 learner group (N=135) t(214) = 39.870, p = .000**
Mapping of aspirated‐/s/, Argentina
• Majority of aspirated variants identified as “s” (82.5%) for Argentine group • When not identified as “s”, mapped to “f” (12%) or “none” (5%)
Mapping of aspirated‐/s/, L2 learners
• Most frequent mapping of [h] for learner group is no coda, [V], (40%) • Majority of remaining data identified as voiceless fricative: “f” (25%) or as “s” (25%)
Results – L2 proficiency level **
** **
• Significant difference in identification of aspirated‐/s/ according to proficiency level, One‐way ANOVA (F(5, 229) = 53.082, p = .000) • All comparisons found significant except between Levels 1 & 2, between Levels 4 & 5, and between Level 5 & Argentine Group (Games‐Howell post‐hoc test)
Distribution of Response Types for Aspirated‐ /s/ According to Proficiency Level • Decrease in no coda [V] response • Emergence of “s” identification at Level 3 • Approaching native aspirating dialect pattern in Levels 4 & 5
Reported Dialect Experience (LBQ) Reported dialect exposure types: 1. Study or living abroad in a /s/‐weakening variety of Spanish 2. Previous and/or current instructors from /s/‐ weakening dialects 3. Frequent contact with /s/‐weakening Spanish speakers (reported talking with NS from /s/‐ weakening dialect frequently) 4. Frequent exposure to /s/‐weakening speakers in the Media (reported frequent exposure to TV, movies, and/or music from /s/‐weakening varieties)
Analysis – Effect of dialect exposure • Linear Regression (enter method) • Variables entered: • Dependent variable = Identification of aspirated‐/s/ (%)
• Independent variables = 1. 2. 3. 4. 5.
Proficiency level Study abroad in /s/‐weakening region Previous instructors from /s/‐weakening region Exposure to media with /s/‐weakening speakers Frequent contact with native speakers from /s/‐ weakening region
Model accounting for variance in L2 identification scores of aspirated‐/s/ • Significant model emerged, F(5, 209)=29.717, p = .000 • Adjusted R2 = .402 • Significant variables: Predicator variable
Beta
t
p
Proficiency Level
.434
6.391
.000
Study Abroad in /s/‐weakening region
.183
3.002
.003
Contact with /s/‐weakening NSs
.155
2.334
.021
• Exposure to /s/‐weakening varieties in Media & Exposure to Instructors from /s/‐weakening varieties were not significant predictors in this model
Identification of aspirated‐/s/ according to study abroad experience type, Mean (N, SD) Study Abroad none
3 4 5
22.8% (42, 29.48) 47.6% (9, 34.44) ‐
s‐conserving s‐weakening both areas area area 17.6% (13, 14.47) 46.2% (13, 32.29) 50.7% (10, 36.34)
50.0% (1) ‐ 54.5% (8, 36.81)
78.6% (3, 12.37) 58.9% (4, 23.60) 71.4% (3, 32.73)
• Higher rates of identification of syllable‐final [h] as “s” for L2 learners who had studied abroad in an /s/‐weakening region than in an /s/‐conserving region or no study abroad at all
Identification of aspirated‐/s/ according to contact with /s/‐weakening speakers, Mean (N, SD) Contact No frequent contact
Frequent contact
3
20.4% (55, 24.26)
87.5% (4, 3.58)
4
48.7% (22, 30.99)
48.2% (4, 36.60)
5
45.2% (3, 46.48)
56.7% (18, 34.17)
• Greater identification of aspirated variants as “s” for those with frequent contact with native speakers from /s/‐ weakening varieties at Levels 3 and 5 • However – unequal (and small) sample sizes, large SDs • Collinearity of contact and study abroad type?
Summary of Results RQ1. What are the patterns of perception of Spanish aspirated‐ /s/ variants by English‐speaking learners of Spanish? • The English‐speaking learners are highly accurate in identification of the full‐/s/ variants as “s” (99%) but do not show a similar tendency for identification of the aspirated variants (21%) • The most frequent L2 identification pattern of Spanish coda [h] is no coda [V] (40% of the responses) • E.g., identification of [leh.po] “lepo” • Followed by the mapping of coda [h] to one of the voiceless fricatives, “f” (25% of responses) or “s” (25% of responses) • E.g., identification of [leh.po] “lepfo” and “lespo”
Summary of Results RQ2. Does perception of the aspirated variants change over time (proficiency level) and increased exposure to Spanish dialectal variation? • Low‐level L2 learners of Spanish (beginner, intermediate) show tendency to “ignore” coda aspiration and/or identify aspirated‐ /s/ as “f” • At Level 3 (high intermediate / low advanced), the same pattern is observed but learners also begin to identify aspirated‐/s/ as “s” (25%) • Levels 4 (advanced) and 5 (graduate students) show tendency for identification of aspirated‐/s/ as “s”, followed by “f” and no coda • *Level 5 (graduate students) are not significantly different from Argentine listeners in identification of aspirated‐/s/
Summary of Results RQ2. Does perception of the aspirated variants change over time (proficiency level) and increased exposure to Spanish dialectal variation? • Depends on the type (and quantity?) of exposure to /s/‐ weakening speech: • Greater identification of aspirated variants as “s” for those learners who had studied abroad in an s‐weakening region and for those who reported frequent contact with native speakers from /s/‐weakening varieties • Having had native instructors from /s/‐weakening regions and exposure to media from these varieties not found to predict perception of aspirated‐/s/
Conclusions 1. The most frequent identification type by the English‐speaking learners of Spanish of the syllable‐ final aspirated variants, [h], is ‘no coda’. • Suggests influence from the native language in perception of the second language • English speakers “ignore” acoustic cues of coda [h] • [h] is not contrastive in syllable‐final position in English (but it is contrastive syllable‐initially!)
• Support for L2 perception models (NLMM, PAM, SLM) • Need for: consideration of effects of position in the syllable of the target sound in cross‐linguistic and L2 perception models
Conclusions 2. Change in patterns of identification of syllable‐final [h] over time (with increased proficiency level): • From a preference at low levels of proficiency for identification of coda [h] as ‘no coda’, to a tendency at advanced levels to identify the aspirated variants as ‘s’ • Evidence of a dynamic, changing interlanguage system – evolving towards native patterns in perception (*for native /s/‐aspirating dialects) • Identification of /s/ does not emerge until Level 3 (300‐ level Spanish literature and culture courses), with tendency for /s/ identification at Levels 4 & 5 • Production of s‐weakening only at Levels 4 and graduate students (Geeslin & Gudmestad 2008)
Conclusions 3. Some (but not all) dialect exposure types found to predict aspirated‐/s/ identification scores: • Dialect exposure • Need for a prolonged and intense contact with /s/‐ weakening, such as through study or living abroad? • Possibility that instructors not producing /s/‐ weakened variants in the classroom • Methodological limitations: • Inaccurate reporting by participants? • Small sample sizes in some cells – need for study of population with more control over different dialect exposure types
¡Muchah graciah!
Lauren Schmidt
[email protected]
References Best, C. T. (1994). The emergence of native‐language phonological influences in infants: A perceptual assimilation model. In C. Goodman & H. Nusbaum (Eds.), The development of speech perception (167‐224). Cambridge: The MIT Press. Best, C. T. (1995). A direct realist perspective on cross‐language speech perception. In W. Strange & J. Jenkins (Eds.), Speech perception and linguistic experience: Issues in cross‐language research (171‐204). Timonium, MD: York Press. Brown, Earl (November 2007). A usage‐based account of final /s/ reduction in Spanish. Conference Presentation. Hispanic Linguistic Symposium, San Antonio, 1‐4 November 2007. Calles, R. & P. Bentivoglio (1986). Hacia un perfil sociolingüístico del habla de Caracas. In J. Moreno de Alba (ed.), Actas del II Congreso Internacional sobre el Español de América, pp. 111‐114. México: Universidad Nacional Autónoma de México. Cedergren, H. (1973). The interplay of social and linguistic factores in Panama.”Unpublished PhD dissertation, Cornell University. Cunningham‐Andersson, U. (1996). Learning to interpret sociodialectal cues. Quarterly Progress and Status Report, Royal Institute of Technology, Department of Speech, Music and Hearing, 2, 155‐158. Dalton‐Puffer, C., G. Kaltenboeck, & U. Smit (1997). Learner attitudes and L2 pronunciation in Austria. World Englishes, 16, 115‐128. Díaz‐Campos, M. & T. Morgan (2002). On the production and perception of Spanish palatal obstruents: An acoustic phonetic study with implications for phonology, dialectology, and pedagogy. In J. Lee, K. Geeslin, & C. Clements (eds.), Structure, Meaning, and Acquisition in Spanish: Papers form the 4th Hispanic Linguistics Symposium. Somerville, MA: Cascadilla, 244‐268. Dohotaru, P. (2004). Variación de ‐/s/ en la habla de habaneros universitarios: Condicionamiento lingüístico y social. In S. Valdés Bernal (comp.), Pensamiento lingüístico sobre el Caribe insular hispánico, 68‐110. Academia de Ciencias de la República Dominicana. Eisenstein, M. & D. Berkowitz (1981) The effect of phonological variation on adult learner comprehension. Studies in Second Language Acquisition, 4, 75‐80. Eisenstein, M. & G. Verdi (1985). The intelligibility of social dialects for working‐class adult learners of English. Language Learning, 35, 2, 287‐298. Figueroa, N. (2000). An acoustic and perceptual study of vowels preceding deleted post‐nuclear /s/ in Puerto Rican Spanish. In E. Herburger, H. Campos, A. Morales‐Front, & T. Walsh (eds.), Hispanic linguistics at the turn of the millennium. Papers from the 3rd Hispanic linguistic symposium. Somerville, MA: Cascadilla.
File‐Muriel, R. (2009). The role of lexical frequency in the weakening of syllable‐final lexical /s/ in the Spanish of Barranquilla, Colombia. Hispania, 92, 2, 348‐360. Flege, J. E. (1995). Second language speech learning. Theory, findings, and problems. In W. Strange & J. Jenkins (Eds.), Speech perception and linguistic experience: Issues in cross‐language research (233‐277). Timonium, MD: York Press. Fox, R. A. & J. T. McGory (2007). Second language acquisition of a regional dialect of American English by native Japanese speakers. Language Experience in Second Language Speech Learning: In Honor of James Emil Flege. John Benjamins. Geeslin, K. & A. Gudmestad (February 2008). The Acquisition of a Second Dialect in Second Language Spanish: Where Shall We Begin? Conference presentation. Current Approaches to Spanish and Portuguese Second Language Phonology, University of Minnesota. February 23, 2008. Goldstein, L. (1987). Standard English: The only target for nonnative speakers of English? TESOL Quarterly, 21, 3, 417‐ 436. Hammond, R. (1978). An experimental verification of the phonemic status of open and closed vowels in Caribbean Spanish. In H. Lopez Morales (ed.), Corrientes Actuales en la Dialectologia del Caribe Hispanico. Hato Rey: Editorial Universitaria, Universidad de Puerto Rico. Hammond, R (2001). The Sounds of Spanish: Analysis and Application (with special reference to American English). Cascadilla Press. Hualde, J. I. (2005). The Sounds of Spanish. New York: Cambridge U Press. Kuhl, P. (1991). Human adults and human infants show a “perceptual magnet effect” for the prototypes of speech categories, monkeys do not. Psychophys, 50, 93‐107. Kuhl, P. (1992). Psychoacoustics and speech perception: Internal standards, perceptual anchors, and prototypes. In L. Werner & E. Rubel (Eds.), Developmental psychoacoustics (293‐332). Washington, D.C.: American Psychological Association. Lafford, B.(1986). Valor diagnóstico‐social del uso de ciertas variantes de /s/ en el español de Cartagena, Colombia. In R. Núñez Cedeño, I. Páez, & J. Guitart (eds.), Estudios sobre la fonología del español del Caribe, 53‐75. Caracas: La Casa de Bello. Major, R., S. Fitzmaurice, F. Bunta, & C. Balasubramanian (2002). The effects of nonnative accents on listening comprehension: Implications for ESL assessment. TESOL Quarterly, 36, 2, 173‐190.
Major, R., S. Fitzmaurice, F. Bunta, & C. Balasubramanian (2005). Testing the Effects of Regional, Ethnic, and International Dialects of English on Listening Comprehension. Language Learning, 55, 1, 37‐69. McMahon, A. (2002). An Introduction to English Phonology. Oxford U Press. Navarro Tomás, T. (1939). Desdoblamiento de fonemas vocálicos. Revista de Filología Hispánica, 1, 165‐167. Samper Padilla, J. A. (1990). Estudio sociolingüístico del español de Las Palmas de Gran Canaria. Las Palmas de Gran Canaria: La Caja de Canarias. Scales, J., A. Wennerstrom, D. Richard, & S. H. Wu (2006). Language learners’ perceptions of accent. TESOL Quarterly, 40, 4, 715‐738. Schmidt, L. (October 2008). Acquisition of sibilant voice assimilation in Spanish as a second language. Conference presentation. The Hispanic Linguistics Symposium, Quebec City, Canada, October 25, 2008. Schmidt, L. (2009). The effect of dialect familiarity via a study abroad experience on L2 comprehension of Spanish. In Joseph Collentine et al., Selected Proceedings of the 11th Hispanic Linguistics Symposium, 143‐154. Somerville, MA: Cascadilla Proceedings Project. Silva‐Corvalán, C. (2001). Sociolingüística y pragmática. Washington D.C.: Georgetown University Press. Smith, L. E. & J. A. Bisazza (1982). The comprehensibility of three varieties of English for college students in seven countries. Language Learning, 32, 259‐269. Tauroza, S. & J. Luk (1997). Accent and second language listening comprehension. RELC Journal, A Journal of Language Teaching and Research in Southeast Asia, 28, 54‐71. Terrell, T. (1977). Universal constraints on variably deleted final consonants: Evidence from Spanish. The Canadian Journal of Linguistics, 22, 2, 156‐168. Terrell, T. (1978). Aspiration and deletion of syllable and word final /s/ in Puerto Rican Spanish. Nueva Revista de Filología Hispánica, 27, 1: 24‐38. Terrell, T. (1981). Diachronic Reconstructions by Dialect Comparison of Variable Constraints: S‐Aspiration and Deletion in Spanish. In D. Sankoff (ed.), Variation Ombibus. Torreira, F. (2006). Coarticulation between aspirated‐s and voiceless stops in Spanish: An interdialectal comparison. N. Sagarra & A. J. Toribio (eds), Selected Proceedings of the 9th Hispanic Linguistics Symposium, 113‐120. Somerville, MA: Cascadilla Proceedings Project. Wilcox, G. K. (1978) The effect of accent on listening comprehension: a Singapore study. English Language Teaching Journal, 32, 118‐127.