Tag Archives: Code-Switching

FAQ: Is mixing languages bad?


Mixing languages is usually a sign of high competence in both languages. Bilinguals often do it even within a sentence (known as code-switching). This does not mean they cannot speak in just one language, or that they are not sufficiently fluent in both languages. It simply means that these speakers use their whole potential in communication, i.e. using the expression that best fits what they want to say. Some studies have found that speakers who are very fluent in both languages are the ones that mix the most.

UIC TiL: Fall 2012 Schedule

Mark your calendars and save these dates because the fall line-up for UIC Talks in Linguistics has been announced. All talks are scheduled on Fridays at 3 PM and will take place in University Hall 1750, located at 601 S. Morgan Street here in Chicago. We look forward to seeing you there for some interesting talks on a wide array of linguistic topics.

  • September 21: Masaya Yoshida, Northwestern (Psycholinguistics)
  • October 19: Kay González-Vilbazo, UIC (Code-switching)
  • November 2: Bernie Issa, UIC (SLA)
  • November 16: Craig Sailor, UCLA (Syntax)
  • November 30: Nicholas Henriksen, Michigan (Phonology)

Titles of the talks as well as abstracts will be announced closer to the dates listed for each.

UIC TiL: Marcel den-Dikken

Greetings everyone!

Next Monday April 23th Marcel den-Dikken from CUNY will be giving a talk at UIC Talks in Linguistics (UIC TiL) entitled ‘Of orphans and twins: Accounting for some peculiar patterns in code-switching’ (see abstract below).

The talk will take place in University Hall 1501 at 3 pm, and as always, light snackswill be provided.

Of orphans and twins: Accounting for some peculiar patterns in code-switching

González-Vilbazo & López (2012) note that in Spanish/German code-switching (CS) a switch at v from a Spanish light verb (inserted in v) to a VP lexified by German vocabulary items leads VO order thanks to the fact that the Spanish v dictates the syntax of the vP (see (1a), analysed as in (1b)). But they also point out that there are speakers for whom the Spanish v=hacer can be followed by a German VP with (German) OV order (as in (2a)). They argue that constructions of the type in (2a) involve what they call an ‘orphan’: a chunk of structure that is not integrated with the rest of the structure of the clause in the regular fashion. The analysis for (2a) that they propose is schematised in (2b), a structure in which there are two vPs, one from Spanish (spelled out as hizo) and the other from German (which is silent); it is the German v that takes the overt VP as its complement, and which causes the VP to be spelled out with (German) OV order.

(1) a.  Juan hizo verkaufen die Bücher Juan did sell the books
‘Juan sold the books’
b. [vP vSp=hizo [VP verkaufen die Bücher]]

(2) a.  Juan hizo die Bücher verkaufen Juan did the books sell
‘Juan sold the books’

       b. [vP vSp=hizo (…)] [vP [VP verkaufen die Bücher] vGe=i]]

González-Vilbazo & López point out that CS constructions with ‘orphans’ such as (2a) behave differently from run-of-the-mill Sp/Ge CS constructions such as (1a) with respect to a number of syntactic properties, including extraction (Worüber has hecho {Tlesen ein Buch/*ein Buch lesen}? ‘about what have you read a book’) and anaphoric dependencies (Juan se ha hecho {Tsehen sich selbst im Spiegel/*sich selbst im Spiegel sehen} ‘Juan saw himself in the mirror’). The ‘orphan’ status of the second vP in (2b) straightfor- wardly explains the fact that nothing can be extracted out of it. But as it stands, the proposal in (2b) leaves two things unexplained: (a) why the object of the OV-VP cannot be anaphorically bound to the subject of the clause, and (b) how the ‘orphan’ relates to the ‘(…)’ in the complement of the non-orphaned v, and what the nature of ‘(…)’ might be.

The problem with (a) is that, since the ‘orphaned’ constituent must be a vP in order for the German v to be able to dictate OV order inside it, and since vP is the locus of base-generation of the external argu- ment, it ought to be possible for the ‘orphaned’ vP in (2b) to have a (null) subject of its own, coreferential with the subject of the clause; an anaphor inside VP should then be able to be locally bound by the null external argument of the ‘orphaned’ vP, and binding should succeed. We will want to make sure that, even though the ‘orphaned’ constituent must indeed be a vP (for word-order purposes), it cannot house an instance of the external argument.

I propose to ensure this by analysing the ‘orphan’ as an asyndetic specifier of the first vP, in a covert coordination structure of the type proposed in work by Koster (2000 et passim). This is a case of asyndetic coordination at the level of vP minus the external argument, which is merged outside the coordinate structure, and introduced by a relator that has the shared external argument as its specifier and the asyndetic coordinat- ion (annotated as ‘:P’, following Koster) as its complement, as in (3).

(3) [RP Juan [RELATOR [:P [vP1 vSp=hizo [VP ec]] [: [vP2 [VP die Bücher verkaufen] vGe=i]]]]]

Coordination/asyndetic specification must be at the level of vP because coordination of acategorial constituents (‘VP’) is impossible; moreover, vP must be present in the second conjunct in order to case- license the object of the German verb. But there is no need to introduce the external argument inside the individual conjuncts: having it be introduced by a relator outside the coordination is derivationally simpler because it involves fewer instances of External Merge (the external argument is merged just once, not twice) and it avoids Across-the-Board extraction of the external argument from the two conjuncts in parallel. (3) is thus the most economical structure for the ‘orphan’ construction — and moreover, it gives the ‘orphan’ an important function: it serves to specify the contents of the empty VP in the first conjunct.

An elliptical VP in the first conjunct thus fills in the ‘(…)’ in the structure in (2b). It must be an empty category because a multi-dominance approach to ‘orphan’ constructions is impossible: having the VP dominating die Bücher verkaufen attached simultaneously to the empty German v and to the Spanish v spelled out as hizo would impose irresolvably conflicting word-order demands upon the VP: since VP would serve simultaneously as the complement of a German and a Spanish v, the object inside VP would have to simul- taneously precede and follow the verb, which is of course impossible. For this kind of Right Node Raising, therefore, an analysis in terms of multi-dominance is out of the question.

Represented this way, the ‘orphan’ construction becomes an instance of a much broader pattern observed in CS constructions: so-called doubling. (2a) is in effectively a doubling construction: the v+VP part of the structure occurs twice, and v is spelled out in both conjuncts — albeit as a null morpheme in the second conjunct, which makes (2a) hard to recognise as a doubling construction. More readily recognisable doubling constructions are utterances such as those in (4), from English/Tamil CS (taken from Sankoff et al. 1990:93).

(4) a.  verb doubling
they gave me a research grant ko`utaa
they gave me a research grant gave.3.PL.PAST ‘they gave me a research grant’

       b.  auxiliary+verb doubling
            I was talking to oru orutanoo`a peesin`u iruntein
            I was talking to one person talk.CONT be.1.SG.PAST ‘I was talking to a person’

       c.  complementiser doubling
just because avaa innoru colour and race engindratunaale just because they different colour and race of-because
‘just because they are of a different colour and race’

I analyse all cases of doubling in CS as involving asyndetic specification, with the ‘shared’ constituent sandwiched between the doublets being the asyndetic specifier of an elliptical constituent in the first conjunct (as in (3)).

Doubling is by no means confined to CS — hence is not a ‘CS-specific’ phenomenon that would justify a separate ‘grammar of CS’. Doubling is found in utterances of monolingual speakers as well. An example occasionally discussed in the literature is complementiser doubling in sentences of the type in (5a) (from Spanish) and (5b) (from Dutch). These constructions are sometimes treated as cases of CP recursion or CP–TopP structures (the latter with a Top-head spelled out as a complementiser) in a strictly right- branching structure. I will present evidence, however, for the conclusion that these are, just like the construc- tions reviewed above, instances of asyndetic specification, with an elliptical TP in the first conjunct. Spanish and Dutch complementiser doubling in addition makes a case for the conclusion that the elliptical TP is a base-generated empty category (ec in (3)), not a PF-deleted full-fledged TP: the sandwiched topic must be base-generated in situ, and can never be followed (in a non-doubling construction) by a non-elliptical TP.

 (5) a.  dice que dinero que no tenía
            says that money that not had
            ‘(s)he says that (s)he didn’t have money’
        b. ik denk dat van brood alleen dat je daarvan niet kunt leven I think that of bread alone that you thereof not can live ‘I think that one can’t live on bread alone’
For more information, contact: Daniel Vergara, Sergio Ramos, or Bernie Issa. Feel free to keep up with past and present talks via the UIC TiL website.

UIC TiL: Ji Young Shim

Join us at our next Talks in Linguistics (TiL)!

This Friday March 30th Ji Young Shim from the CUNY graduate center will be giving a talk entitled ‘A Minimalist Account of Word Order Variation in Code-switching’.

When: Friday March 30th at 3pm

Where: University Hall (601 S. Morgan Street, Chicago, IL 60607) 1750

We hope to see you there!

For more information, contact: Daniel Vergara (dverga3@uic.edu), Sergio
Ramos (sramos@uic.edu) or Bernie Issa (issa2@uic.edu).

Here is an abstract from his work.

A Minimalist Account of Word Order Variation in Code-switching
Ji Young Shim; CUNY Graduate Center

Under the assumption that monolingual and bilingual grammars are subject
to the same principles, the present study aims to provide a principled
account of word order variation in code-switching (CS). Cross-linguistic
CS data show that not only can a switch occur between languages with
different canonical word orders, such as an OV language (e.g., Japanese,
Korean) and a VO language (e.g., English), but the internal order of a
code-switched constituent may also vary, exhibiting either order of the
two languages involved in CS. One immediate question arises as to how
these different word orders are distributed and derived. The present study
employs three different experimental tasks, which are tested against
Korean-English and Japanese-English bilingual speakers’ introspective
judgments of the CS patterns that are presented to them in the form of a
The statistical results from 34 Korean-English bilinguals show that both
the distinction between light and heavy verbs within a code-switched
constituent and the difference between literal and non-literal/idiomatic
meaning of the phrase play a role to derive different word orders in CS,
which reveals that syntax alone cannot account for the various word order
patterns in CS, but both syntax (particularly, the syntax of light verbs,
which differs from language to language) and meaning (the semantic
compositionality of a phrase) contribute to OV-VO variation in CS. We also
found that there is a correlation between the preferred word order and the
syntactic flexibility of a code-switched constituent.
Based on the findings of the tests, the study proposes a syntactic account
of OV and VO derivation in Korean-English and Japanese-English CS in the
framework of Minimalism. The present talk focuses on Korean-English CS
data, and the findings from a small set of Japanese-English CS tests will
be also discussed for the purpose of comparison.

Conference: UIC Bilingualism Forum

We are proud to announce the 2011 UIC Bilingualism Forum, to be held here at the University of Illinois at Chicago, April 14-15.

The UIC Bilingualism Forum is dedicated to research in any area related to bilingualism: theoretical linguistics, codeswitching, SLA, psycholinguistics, sociolinguistics, neurolinguistics, cognitive sciences, heritage acquisition, bilingual acquisition, etc. Presentations will be 20 minutes each with 10 minutes for discussion.

Keynote Speakers

  • Marcel den Dikken, City University of New York
  • Michael Ullman, Georgetown University

Call For Papers

  • Deadline for submission of abstract: 12/1/2010
  • Acceptance response by: 1/15/2011
  • 2 page anonymous abstract including examples and references
  • 1 separate page with name, title, and affiliation
  • Abstract should be submitted via Linguist List

Linguistic Links: Extra, extra…

Some recent bilingualism-related news links to pique your interest:

UIC TiL: Bryan Koronkiewicz & Tara Toscano

This Friday, March 19th, we’ll be having our semesterly student session of UIC Talks in Linguistics. Bryan Koronkiewicz will present a talk entitled, “Exceptional Hiatuses in Spanish: An Extension of Cabré & Prieto (2006)” and Tara Toscano will be presenting a talk entitled, “The Strandability of Prepositions in Spanish-English Code-switching”

Join us at 3 PM in 1650 University Hall (601 S. Morgan St. Chicago, IL 60607) for some talks and refreshments.

Bryan Koronkiewicz (University of Illinois at Chicago)
Exceptional Hiatuses in Spanish: An Extension of Cabré & Prieto (2006)

In Spanish phonology, the syllabification of rising sonority vocoids predicts diphthongization. Yet in various dialects, with a variety of words, speakers favor the creation of a hiatus in such a context. For example, the word *piano* can be realized by Spanish-speakers either as the expected form [‘pja.no] or alternatively as [pi.’a.no]. The work of Hualde (1999, 2002) and Colina (1999) attests that word initiality, distance to stress, and morpheme boundaries all have strong effects on the realization of these so-called exceptional hiatuses. However, more recently these effects have been reexamined by Cabré & Prieto (2006) with dissimilar results, arguing that they are not as strong.

In this talk I will continue to explore the potential explanations for exceptional hiatuses. This current study recreates and expands the work of Cabré & Prieto (2006). Continuing their approach, Peninsular Spanish-speakers are tested on their tendency toward exceptional hiatuses, examining the specific parameters that may be influential. Furthermore, Mexican-Spanish-speakers are also tested to see if the effects are similar cross-dialectally.

Tara Toscano (University of Illinois at Chicago)
The Strandability of Prepositions in Spanish-English Code-switching

I have tested the acceptability of Preposition stranding in English-Spanish code- switching contexts by having sequential bilinguals perform a sentence judgment task. The term Preposition stranding (P-stranding) refers to an instance where the object of the preposition is extracted and the preposition is not pied-piped as shown in (1):

(1) Who did John talk [PP to[ t]]?

While P-stranding is found in some languages, it does not appear in others. P-stranding is quite common in English as shown in example (1). But in Spanish there is a lack of P-stranding:

(2) *Quién habló Juan  [PP con [ t]]?
Who   spoke John        with

Code-switching allows insights into the two grammars that are otherwise opaque in monolingual utterances. I hypothesized that the language of the preposition in code-switching would determine the acceptability of P-stranding regardless of the language of the DP or NP. I explored 2 hypotheses:

1. Spanish prepositions will not allow P-stranding in a code-switching context, as in (3c) and (3d), and English prepositions will, as in (3a) and (3b).

(3) a. Quién salió John with?
b. Quién did John leave with?
c. Who did John leave con?
d. Who salió John con?

2. The language of Tense (T), or more specifically little v, will determine the acceptability of P-stranding. English T will trigger P-stranding (see (3b) and (3c)) while Spanish T will prohibit it (see (3a) and (3d)).

No conclusions can be made regarding the element or layer of the structure that sanctions P-stranding because this phenomenon occurs in the Spanish dialect of the participants.

E-Interview: Natascha Müller, Part 3 – Testable Assumptions

In the final segment of our e-interview with Natascha Müller she presents some testable assumptions, which she has corroborated in her own research.

“The two assumptions for delay effects are testable, since they make different predictions. In my research I have argued against Serratrice, Sorace & Paoli (2004) and I have assumed that the presence and direction of the influence is not a question of MORE-or-LESS constraints, but a question of whether pragmatics decides on syntactic options or not. The invasive function of pragmatics is complex (Italian, Spanish), non-invasiveness is derivationally neutral (English, German, French). If delay is not due to a default strategy (processing load in one language) but motivated by cross-linguistic influence where the linguistically more complex analysis of language A is avoided in favor of the less complex analysis of language B, we can make the following predictions which have been corroborated in the child data we have analyzed:

a)  The effect should only be observable in bilingual children with particular language combinations, i.e. it is not due to the fact that the children acquire two languages, one with more, the other with less constraints, generally speaking. Only an approach to delay effects which takes into account the structure of the two languages involved will be able to account for Continue reading

E-Interview: Natascha Müller, Part 2 – Delay Effects in Bilingualism

Continuing with our e-interview, Natascha Müller talks about delay effects in childhood bilingualism:

“Meanwhile, there is ample evidence for delay effects of early child bilingualism. Bilingualism can slow down the acquisition process with respect to age of acquisition and MLU; in other words, for some grammatical properties, bilingual children reach the adult norm later (age and/or MLU) than monolingual peers. There is also evidence which shows that delay effects are observable in balanced as well as in unbalanced children (Müller & Patuto 2009), which means that an uneven development of the two languages is not a prerequisite for delay. Furthermore, although unbalanced language development can slow down acquisition with respect to age, it does not necessarily lead to differences between bilinguals and monolinguals with respect to MLU in the weakly developed language (Müller & Pillunat 2008). It looks as if delay is related to complexity in the following sense: Language A and B exhibit different degrees of complexity for a particular grammatical property. In Hulk & Müller (2000) and Müller & Hulk (2001), complexity is defined as the coordination of information from different modules, pragmatics and syntax for example. Delay is indicative of target-deviant grammatical representations which, during the course of acquisition, have to be “corrected”. The child will use the less complex analysis of language A in relation to grammatical property X when using language A and language B. Müller & Patuto (2009) further refine the scenario for delay effects of cross-linguistic influence and conclude that in addition to complexity defined as the coordination of information from different modules, the surface strings of the two languages A and B have to be analyzable in terms of the syntactic derivation of language A (which is less complex). This prerequisite looks trivial at first sight, but it excludes the possibility that children come up with analyses for the more complex language B which are not also “supported” by the evidence from language B. Also, it makes the interesting prediction that if the more complex language B is acquired together with a language which encodes the respective grammatical property in such a radically different way than in language A (the less complex language), the derivation of language A would not be “supported” by the evidence of language B when used by the child while speaking language B. Continue reading

E-Interview: Natascha Müller, Part 1 – Two Approaches to Bilingualism

Along with our series on blogs around the world we are also featuring e-interviews with researchers in the field of bilingualism.  We have the privledge this week to present the thoughts of Natascha Müller of the Bergische Universität Wuppertal.  We asked her why the study of bilingualism in children is relevant:


The Wuppertal research group on bilingual first language acquisition, funded by the DFG (Deutsche Forschungsgemeinschaft, German Science Foundation)

Row 1: Veronika Jansen, Natascha Müller, Elisa Turano, Mayte Jiménez Lopez, Laia Arnaus Gil
Row 2: Dunja Stachelhaus, Annette Pötzsch, Nadine Eichler, Vanessa Colado Miguel, Alban Beysson, Tobias Stallknecht, Marisa Patuto.

“Research in bilingual first language acquisition has been guided by two main approaches: Either it has been argued that bilingual children are not able to separate their two languages from early on since the two languages influence each other (Volterra & Taeschner 1978) or it has been shown that separation is possible from early on and that there is no evidence for cross-linguistic influence (Meisel 1989, Genesee 1989, Genesee, Nicoladis & Paradis 1995). Put differently, separation and cross-linguistic influence have been considered as being mutually exclusive in describing early child bilingualism. The main reason for the assumption of mutual exclusiveness is that most research has conceptualized separation and influence as involving whole language systems (or languages). Among the first to question this view were Gawlitzek-Maiwald & Tracy (1996). The main observation is that some grammatical domains develop separately in early child bilingualism while the bilingual child uses language A to bootstrap aspects of the syntactic system of language B for others.

 What does ‘bootstrap’ in this context mean? Gawlitzek-Maiwald & Tracy analyze the monolingual (containing only elements of the context language) and the mixed utterances of Hannah, a bilingual English-German child. On the basis of the respective monolingual utterances, they find that German is much more advanced than English with respect to lexical and syntactic aspects of temporal and modal auxiliary verbs. In order to ‘help herself out’ when speaking English, Hannah produces mixed utterances of the following type:

 (1) Kannst du move a bit                (Hannah, 2;4-2;9, Gawlitzek-Maiwald     
      Can you move a bit                                     &   Tracy 1996:915)

 In example (1), the left periphery comes from German, while the lexical verb and the adverb are in English. Until the English system of modal and temporal auxiliaries has been fully acquired by the child, she will fill in lexical material from German, a strategy which may also help the child to instantiate the English system: “Something that has been acquired in language A fulfils a booster function for language B.” (Gawlitzek-Maiwald & Tracy 1996: 903) The situation can be reversed for other grammatical phenomena

(Click to Read More and Stay Tuned for parts 2 and 3 with Nascha Müller)

Continue reading