Vulnerability in Heritage Speakers of Spanish in the Netherlands: An Interplay between Language-Internal and Language-External Factors

Brechje van Osch

University of Amsterdam.

E-MAIL: brechje.a.osch@uit.no

Defense date: October 1, 2019.

Supervisors:

Prof. dr. Aafke Hulk,
University of Amsterdam

Dr. Petra Sleeman,
University of Amsterdam

Dr. Suzanne Aalberse,
University of Amsterdam

Defense Committee:

Prof. dr. Jeannette Schaeffer,
University of Amsterdam

Prof. dr. Judith Rispens,
University of Amsterdam

Prof. dr. Enoch Aboh,
University of Amsterdam

Dr. Maria del Carmen Parafita Couto,
University of Leiden

Prof. dr. Jason Rothman,
University of Tromsø

Dr. Cristina Flores,
University of Minho

1. Introduction

This dissertation investigates Spanish heritage speakers (HSs) in the Netherlands, a relatively understudied population. The main aim is to investigate whether linguistic interfaces, especially the external interface between syntax and discourse-pragmatics, are particularly vulnerable in this population, as predicted by the Interface Hypothesis (IH) (Sorace, 2011; Sorace & Filiaci, 2006). While much previous research has tested the IH by comparing two completely different phenomena (e.g., Montrul, 2008; Sorace & Serratrice, 2009), this dissertation compares interfaces within a single phenomenon, to keep possibly confounding variables (e.g., frequency or salience) constant.

An additional objective of this dissertation is to explore several variables external to the heritage language itself, such as age of onset and manner of acquisition, age at testing, task effects, the socio-linguistic circumstances of the host country, and cross-linguistic influence from the dominant language.]

The target phenomena are mood and subject position, both of which can have different functions, pertaining to different (interface and non-interface) domains, as will be explained in the following sections.

2. Mood

For mood, three different functions are distinguished, pertaining to three different linguistic domains: syntax, the syntax-semantics interface, and the syntax-pragmatics interface respectively. The syntactic function concerns contexts in which the main verb obligatorily selects mood in the embedded clause. For instance, episodic predicates in the main clause select indicative mood in the embedded clause (1), while volitional predicates select subjunctive mood (2).

(1) Sé que vas conmigo.

(I) know that (you) go.2SG.PRES.IND / with me

‘I know that you go with me.’

(2) Quiero que vayas conmigo.

(I) want that (you) go.2SG.PRES.SUBJ with me

‘I want you to go with me.’

The syntax-semantics interface function concerns relative clauses specifying indefinite antecedents. In these sentences, both moods are grammatical, but acceptability depends on the specificity of the antecedent.

(3) Buscamos un hotel que tiene/tenga un gimnasio.

(we) look. for a hotel that have.3SG.PRES.IND / have.3SG.PRES.SUBJ a gym

‘We are looking for a hotel that has a gym.’

In (3), the indicative is preferred if the speaker knows that such a hotel exists. Otherwise, subjunctive mood is more acceptable.

The syntax-pragmatic function regards the selection of mood in embedded clauses with certain main predicates in negation. In this case, the preferred mood is determined by the speaker’s commitment to the truth of the proposition expressed in the embedded clause

(4) Pedro no dice que es/sea su culpa.

Pedro NEG say that (it) be.3SG.PRES.IND / be.3SG.PRES.SUBJ his fault

‘Pedro does not say that it is his fault.’

In (4), subjunctive mood is preferred if the speaker does not think that it is Pedro’s fault. Otherwise, both moods are possible.

Of these three functions, the latter is predicted to be most vulnerable by the IH, given that it involves the external interface between syntax and pragmatics.

3. Subject Position with intransitive verbs

In Spanish, the position of the subject respective to the intransitive verb is flexible. However, various linguistic variables determine whether a preverbal or a postverbal subject is more likely to occur. Three of these are taken into consideration in this dissertation.

The first concerns the semantics of the verb. While subjects of unergative verbs tend to be placed preverbally (5), subjects of unaccusative verbs are usually placed following the verb (6).

(5) Juan gritó. (unergative - SV)

Juan shouted

John

(6) Llegó Juan. (unaccusative - VS)

Arrived Juan

‘John arrived.’

The second factor concerns focus. If the subject is in presentational focus, it tends to follow the verb (7), while this is not the case in broad focus, or out-of-the-blue sentences.

(7) ¿Quién gritó? Gritó el niño. (narrow focus - VS)
who shouted (‘Shouted’) the boy

‘Who shouted? The boy shouted.’

Finally, the definiteness of the subject plays a role in determining its position. Definite subjects tend to precede the verb, while indefinite subject usually follow it.

Since both focus and definiteness pertain to the external interface between syntax and discourse-pragmatics, sensitivity to these factors is expected to be more vulnerable than sensitivity to verb type, which is located at the internal interface between syntax and semantics.

4. Methodology

4.1 Participants

Several groups participated in the studies reported in this dissertation:

1. Adult HSs of Spanish in the Netherlands
2. Adult HSs of Spanish in the US
3. Child HSs of Spanish in the Netherlands (ages 5, 9, and 13)
4. A control group of monolingual speakers of Spanish (ages 9, 13, and adults)
5. Adult L2 speakers of Spanish in the Netherlands

4.2 Tasks

The phenomena were tested using a contextualized acceptability judgement task (AJT), which was designed to target explicit knowledge, and an elicited production tasks (EPT), testing implicit knowledge. Both tasks were presented simultaneously in written and in auditory mode. Moreover, general proficiency in Spanish was measured by the DELE (Diploma de Español como Lengua Extranjera), a standardized written proficiency task, as well as by a timed lexical decision task.

5. Results

For mood, the combined results from all studies show that mood is most robust in the obligatory contexts (the syntax condition) and most vulnerable in the negated sentences condition (the external interface). The relative clause condition (the internal interface) was found to be somewhat in the middle. As for word order, verb type, pertaining to internal syntax-semantics interface, was found to be the least problematic for all groups of bilinguals. Sensitivity to focus and definiteness, the two external interface factors, was more problematic. Overall, these results suggest that interfaces—in particular the external interface—are more vulnerable in bilinguals than purely syntactic phenomena.

However, there were important modulating effects of several extra-linguistic factors. One of the variables found to influence the results was the type of task. For mood, HSs of Spanish in the Netherlands performed more in line with the monolingual baseline in the explicit AJT than in the implicit EPT, contrasting with most research from the US (Bowles, 2011). This is attributed to the different socio-linguistic circumstances in the two countries: in the Netherlands, Hispanic communities are smaller and more dispersed, which may lead to fewer possibilities for interaction with other speakers, and thus a smaller advantage on the oral production task. Moreover, the multilingual nature of the Dutch education system and the media may result in a higher degree of meta-linguistic awareness which could lead to an advantage on the explicit task.

To investigate whether age and manner of onset of acquisition plays a role, HSs were compared to L2 speakers of similar general proficiency. While both groups did better on the AJT than on the EPT, the HSs outperformed the L2 speakers in the EPT and vice versa in the AJT, suggesting that early and naturalistic acquisition of a language results in more implicit knowledge than late acquisition, as by the Fundamental Difference Hypothesis (Bley-Vroman, 1990).

To explore the role of the dominant language, HSs of Spanish in the Netherlands were compared to a matched group of HSs of Spanish in the US regarding their knowledge of the constraints on word order in Spanish. In Dutch, the relation between definiteness and word order is stronger than in English. This difference between the two dominant languages was reflected in the HSs’ linguistic behavior in Spanish: The Dutch-dominant speakers were more sensitive to the definiteness effect in Spanish than their English-dominant peers, suggesting an important role for cross-linguistic influence.

Finally, to test how knowledge of the constraints on word order develops in heritage Spanish, a cross-sectional study was conducted comparing HSs in the Netherlands from ages 5, 9, 13 and adults. Knowledge of the effect of verb type on word order was present in all age groups, implying that this factor is acquired early and remains present in the heritage grammar. Sensitivity to the effect of focus emerged much later, after age 13. As for knowledge of the definiteness effect on word order, this was attested in the 9-year-old and the 13-year-olds, but not in the adult group, suggesting that it gets lost between the ages of 13 and adulthood, possibly due to a drop in input in the heritage language after leaving the parental home.

6. Conclusion

In conclusion, this dissertation provides an important contribution to the field by investigating a relatively understudied population of HSs outside the US. The unique design of the studies provides support for the claims made by the IH. In addition, the results emphasize the importance of several language-external variables in explaining HS divergence. It is concluded that both linguistic and non-linguistic factors should be taken into consideration to arrive at an exhaustive and sophisticated theory about heritage language development and outcomes.

References

Bley-Vroman, R. (1990). The logical problem of foreign language learning. Linguistic Analysis, 20, 3–49.

Bowles, M. A. (2011). Measuring implicit and explicit linguistic knowledge: What can heritage language learners contribute? Studies in Second Language Acquisition, 33(2), 247–271.

Montrul, S. (2008). Incomplete acquisition in Spanish heritage speakers: Chronological age or interfaces vulnerability. In Proceedings of the 32nd annual Boston University Conference on language development (Vol. 1). Cascadilla Press.

Sorace, A. (2011). Pinning down the concept of “interface” in bilingualism. Linguistic Approaches to Bilingualism, 1, 1–33.

Sorace, A., & Filiaci, F. (2006). Anaphora resolution in near-native speakers of Italian. Second Language Research, 22(3), 339–368.

Sorace, A., & Serratrice, L. (2009). Internal and external interfaces in bilingual language development: Beyond structural overlap. International Journal of Bilingualism, 13(2), 195–210.