1 Spanish-English bilinguals’ reactions to felicitous and infelicitous intra- sentential code-switching Tyler K. Anderson Mesa State College [email protected].

1

Spanish-English bilinguals’ reactions to felicitous and infelicitous intra-

sentential code-switching

Tyler K. Anderson

Mesa State College

[email protected]

2

Spanish and English in Contact

• Various Manifestations– Lexical Borrowing

• Lonche, migrante, boila, etc.

– Structural adaptations• Possible increase in overt subject pronouns in

Spanish

• As contact and bilingual proficiency increases, code-switching is manifest.

3

Code-switching

• Code-switching can be defined as– The rapid, uninterrupted alternation between

two languages by bilinguals in an unchanged speech situation

• Two types of CS– Inter-sentential (between sentence

boundaries)– Intra-sentential (between clauses)

4

Grammar of CS

• Beginning in the 1970s it became apparent that CS was constrained by a grammar, much like monolingual utterances.

• Some of the switch sites that have been distinguished as acceptable include

i. between a noun and its relative clause– Había siete enanitos who had worked all day in

the mines.ii. between a subject and the predicate

– The squirrel no tenía nada más que frijoles.

iii. between a verb and its object – Un día la reina miró her magic mirror.

5

Grammar (cont.)

• Other sites have been shown to be incompatible with CS, including

i. between (NEG) and its verb phrase• Graciela ya no had the comforts of a palace.

ii. between auxiliary verb and its verb phrase• Her father had vuelto a casarse con una viuda

que tenía dos hijas.

iii. between an object or subject pronoun and the verb phrase

• I want you to return with me to the city where I vivo y te enseñaré cómo es mi vida en la ciudad.

6

Popular views of CS• Code-switching is commonly disparaged in popular view,

as depicted in the following statements:– El cambio de códigos o code-switching, sobre todo entre

inglés y español, se interpreta como una deficiencia lingüística que revela la falta de proficiencia del hablante en ambas lenguas, la cual le obliga a recurrir a la segunda lengua cuando agota su repertorio en la primera. (Fernández 1990:52)

– … in [sic] a linguistic point of view [codeswitching] is an abomination. Is [sic] a great lack of elementary edutacation [sic] to mix 2 differents [sic] languages, thus [sic] is not the way ordinary poeple [sic] choose to speak but the way choose by people without a trace of education. Don't missunderstand [sic] me I'm no [sic] against the "PEOPLE", I'm against the "LACK OF EDUCATION" that favors such things like the "spanglish” (Aldebarán 2005).

7

• Similarly, even the vernacular nomenclature for the amalgamation carries pejorative connotations that reflect on the perceived intellectual and linguistic limitations of speakers of contact Spanish. – Spanglish, casteyanqui, Tex-Mex, mocho, espalés,

Gringoñol, Englañol and ingleñol• However, as can be assumed, not all references to CS

are negative, as can be witnessed in the following excerpt from a recent editorial:– “Antes, cuando recién acabada de llegar aquí, como

nuevo inmigrante, me horrorizaba al escuchar a la gente hablar… Poco a poco me di cuenta que el spanglish no es un complot…para lavarnos el cerebro a los hispanos. No, el spanglish es una herramienta, que los inmigrantes usamos para vivir diariamente en Estados Unidos” (Zapata, 2004)

• Therefore, from these views we note that the perceptions of code-switching are varied, though generally negative.

8

Research Question

• While linguists may be aware of the grammatical limitations placed on acceptable switch sites, little focus has centered on whether the non-linguist is aware of these restrictions. In order to answer this inquiry,– the present study seeks to evaluate whether bilinguals

react dissimilarly towards grammatical and ungrammatical language alternations.

• Likewise, little has been established on how bilingual proficiency affects these evaluations. In order to answer this inquiry, – the present study seeks to determine if proficiency affects

the judges’ reactions.

9

The study

• 274 Spanish-English bilinguals of differing proficiencies listened to storytellers (2 male, 2 female) retell bilingual versions of four fairytales– El ratón de cuidad y la ardilla de campo / The city mouse and

the country squirrel – El príncipe pordiosero / The beggar prince– La cenicienta / Cinderella – Blancanieves / Snow White

• Each participant listened to four fairytales– 2 containing ungrammatical CS– 2 containing grammatical CS

• After listening to each fairytale, participants were asked to give their first impressions of the storytellers.

10

Data• To encode the responses to

this open-ended question, each comment was evaluated in connection with each storyteller and categorized according to the most common responses. Master categories were then created, based on the most common responses for all speakers. The following scales were devised:– Pronunciation– Neg Personal Char– Pos Personal Char– Grammar– Easy to understand– Hard to understand

– Confusing b/c of CS– CS as aid– Odd switches– Good CS– CS as deficit– General CS ref– Bilingual– Pos English Abilities– Pos Spanish Abilities– Neg English Ability– Neg Spanish Ability– Origin– Pos Voice Quality– Neg Voice Qualitys– Educated – Uneducated

11

Sample comments

• This person seems really nice and educated but it is VERY hard to say what the dominant language of this person is. She seems to be a rare case of a balanced bilingual. It is hard to perceive any obvious "accentedness" in either Spanish or English when she uses them respectively. Also the language switches seem pretty natural. (l0703)

• Chicana probably second or third generation Mexican American (Texan?). (609)

• She sounds very young and her pronunciation of both Spanish and English are very good so it makes me think that she has been learning (or studying) both languages for a long time. (m511)

12

• My first impression of this person is that he is fluent in Spanish. However his English is not very good. He seems friendly and enjoyable to listen to. (m1121)

• He's jumpin all over the place pick one language buddy. Mexican sounding. (n0227)

• She speaks very softly but sounds like a native. I think the switching back and forth between English and Spanish is confusing and I'm unsure as to why she's doing it. The parts that she speaks in either one of the languages sounds fluent enough that she could choose just one of the languages to speak in and she would be fine. (n1225)

13

ResultsAggregate

• Overall we note several trends that indicate that these bilingual judges are distinguishing between Grammatical and Ungrammatical CS

• For example, judges made greater reference to CS proper when listening to ungrammatical switches than in connection with grammatical switches.

14

References to CS

6.0%

19.0%

0.0%

5.0%

10.0%

15.0%

20.0%

References to CS

Grammatical

Ungrammatical

15

Other References

• When listening to the Grammatical versions, judges tended to focus more often on pronunciation, positive characteristics of speaker, origin, and voice quality.

16

Aggregate Results—Other References

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

Pronunciation + Personal Char Origin Voice Quality

Grammatical

Ungrammatical

17

Aggregate results—Linguistic abilities

• An interesting trend concerns the types of reactions that the judges had to the storytellers’ linguistic skills.

• The speakers of the grammatical versions were deemed to have greater English skills, and comparatively weaker Spanish skills.

• The speakers of the ungrammatical versions were deemed to have greater Spanish skills and were more likely to be bilingual.

18

Aggregate results—Linguistic abilities

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

35.0%

40.0%

English Abilities Spanish Abilities Bilingual abilities Limited EnglishAbilities

GrammaticalUngrammatical

19

Results by Proficiency

• Recall that one of the research questions for this study was how proficiency influenced the ways in which judges’ reacted to the grammatical vs. ungrammatical switches.

• All subjects were studying at a university in the Northeast US, and thus were required to be highly proficient in English.

• Thus the variable that was investigate was levels of Spanish proficiency

20

Results by proficiency

• Spanish proficiency was determined by a variation of the DELE (Diploma de Español como Lengua Extranjera)

• 132 (g) and 132 (u) judges tested as low

• 291(g) and 288 (u) tested as intermediate

• 122 (g) and 123 (u) tested as high

21

Results by ProficiencyTrends Grammatical Version

• In the grammatical version the following trends were seen– As proficiency increased there was a general

decrease in attention to:• Pronunciation• Reference to Good CS• Positive English abilities

– And as proficiency increased there was also an increase to general CS references

22

Results by proficiencyGrammatical Versions

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

35.0%

Pronunciation Good CS + Eng Abilities General CS

Trends--References to Grammatical versions

Low

Inter

High

23

Results by ProficiencyUngrammatical Version

• In the ungrammatical version the following trends were seen– As proficiency increased there was a general

decrease in attention to:• Pronunciation• Bilingualism • Good CS• Grammar• Levels of Education

– And a general increase in attention to• Origin of speaker• Voice quality

24

Results by ProficiencyUngrammatical Version

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

Pronunciation Bilingual Good CS Grammar Uneducated

Trends--References to Ungrammatical versions

Low

Inter

High

25

Results by ProficiencyUngrammatical Version

0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

14.0%

16.0%

Origin Voice Quality

Trends--References to Ungrammatical versions

Low

Inter

High

26

Results by proficiencyOther trends

• Another interesting trend is noted in the attributing of higher levels of education as proficiency of the judges increased, as well as an increase in distinction between grammatical and ungrammatical versions as proficiency increases.

0.0%

5.0%

10.0%

Low Inter High

Gram

Ungram

27

Results by proficiencyOther trends

• On the aggregate, the speakers of ungrammatical versions were more likely to be considered bilingual. However, as proficiency increased this distinction diminished.

0.0%

10.0%

20.0%

Low Inter High

Gram

Ungram

28

DiscussionResearch question 1

• Regarding the first research question, ample evidence has been provided to establish that the non-linguist recognizes, at some level, the differences between what linguists have determined to be felicitous and infelicitous CS.

• This is noted in an heightened reference to CS proper when listening to ungrammatical CS, which points to an increase in saliency of the ungrammatical switches, as would be expected.

29

• Because subjects were asked to react in some fashion to the storytellers, it is interesting that judges focused on other linguistic features (i.e., pronunciation, origin, or voice quality) much more in the context of felicitous switches than when infelicitous switches were present. This again points to the lack of saliency of grammatically correct switches.

30

• Finally, an interesting trend is the ascription of Spanish and English linguistic abilities to the storytellers. When the speakers were using grammatical CS, it was more likely to that they were considered English dominant.

• Conversely, when the speakers used ungrammatical switches, they were more likely to be considered bilingual, Spanish dominant or having limited English proficiency.

• These trends point to an perception of limited linguistic abilities and ungrammatical CS.

31

DiscussionResearch question 2

• With regards to the influence of judges’ Spanish proficiency, it was noted that as proficiency increased certain factors became more or less important when looking at grammatical vs. ungrammatical CS

32

• For example, there was a general decrease in reference to pronunciation as abilities increased.

• At first glance this finding can be counterintuitive, considering the ample research that deals with increased abilities to perceive non-target phonemes as proficiency increases (cf. Flege and MacKay 2004).

• However, it seems to be that as proficiency increases so too does tolerance to non-target pronunciations, which finding is supported by prior research (cf. Thompson 1991).

33

• Subjects also tended to decrease their references to ‘good CS’ as proficiency increased, regardless of grammaticality of the switches.

• Again, this can be attributed to an increased tolerance to ‘non-standard’ norms as bilingual abilities increase.

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

7.0%

Good CS (Gram) Good CS(Ungram)

Low

Inter

High

34

Limitations/Future studies

• The first limitation of this study has to do with the lack of ‘rigor’ in the research design. Although the information is promising, other methodologies (i.e. matched-guise technique) will allow a more quantitative view of non-linguists’ abilities to distinguish between felicitous and infelicitous CS. Future studies can implement Likert-scales based on the information from the current study, as suggested by Garrett et al ( 2003) .

35

• Likewise, the use of Fairytales for the speech samples may be considered ‘unauthentic’ material for CS. The use of more realistic material is required.

• Similarly, with regards to bilingual proficiencies, a more balanced design is needed.

• Finally, perspectives from other regions of the United States (i.e. Southwest), and different locales (i.e. urban vs. rural) will also shed light on how non-linguists react to this manifestation of language contact.

36

References• Aldebarán (2005, September 11). What new about Spanglish? [Forum

entry.] antimooon. (http://www.antimoon.com/forum/t493-0.htm). Retrieved May 23, 2006.

• Fernández, Rosa (1990). Actitudes hacia los cambios de código en Nuevo México Reacciones de un sujeto a ejemplos de su habla. In John J. Bergen (Ed.), Spanish in the United Status: Sociolinguistic issues (pp. 49-58). Washington D.C.: Georgetown University Press.

• Flege, James Emil and Ian R. A. MacKay (2004). Perceiving vowels in a second language. SSLA, 26, 1-34.

• Garret, Peter, Nikolas Coupland, and Angie Williams. (2003). Investigating Language Attitudes. Cardiff, UK: University of Wales Press.

• Thompson, Irene (1991). Foreign accents revisited: The English pronunciation of Russian immigrants. Language Learning, 41, 2, 177-204.

• Zapata, Emilio (2004, July 29). Sorry, yo no spekeo español, ni english. Spanglish, for plis. (http://www.elhispanonews.com/editorial_det.php?nid=132). Retrieved February 27, 2007.