Journal of Child Language Acquisition and Development – JCLAD Vol: 5 Issue: 4 193-207, 2017, December ISSN: 2148-1997 193 Subject realization in bilinguals: A comparative study of German-Turkish and Russian-Turkish bilingual children Çiğdem Sağın Şimşek1 Middle East Technical University Elena Antonova-Ünlü2 Hacettepe University Abstract The aim of this study is to examine the vulnerability of subject realization in Turkish as an interface structure at the syntax-pragmatics interface. The study compares subject realization of four Turkish monolingual, three German- Turkish bilingual and two Russian-Turkish bilingual children. The language combinations investigated in the study were determined by the fact that Russian is a partially null-subject language, while German is a non-null-subject one and Turkish is a null-subject language. Thus, focusing on the comparison of two different language combinations, the study aims to provide new insights about bilinguals’ subject realization patterns and their possible relation to cross- linguistic influence. The data for the study were collected by recording the natural language production of the three groups of children. Analysis of the data revealed that both the German-Turkish and the Russian-Turkish bilingual children overused overt subject pronouns in their Turkish more than their monolingual counterparts. Hence, we hypothesized that the inappropriate subject realizations of the bilingual children cannot merely be explained as evidence for cross-linguistic influence but also as a language processing problem. Keywords Subject realization, syntax-pragmatics interface, German- Turkish, Russian- Turkish, Turkish 1. Introduction Sorace and Filiaci (2006) proposed the Interface Hypothesis (IH) according to which structures involving interface between syntax and other cognitive domains are more prone to fossilization and incomplete acquisition in L2 end-states. Initially, this hypothesis was suggested for the very advanced level of ultimate attainment in L2 acquisition, however, later it was expanded to bilingual first language acquisition (BFLA) and to initial stages of L1 attrition. Several studies (Müller & Hulk, 2001; Paradis & Navarro, 2003; Tsimpli, Sorace, Heycock & Filiaci, 2004; Belletti, Bennati & Sorace, 2007; Haznedar, 2010) validated the IH demonstrating that at the syntax-discourse interface, language behavior of L2 learners as well as that of children 1 Bio: Çiğdem Sağın-Şimşek is an associate professor at Middle East Technical University, Department of Foreign Language Education. Her domains of research include aspects of bi-multilingualism, language contact, second and third language acquisition, Turkish linguistics and sociolinguistics. Contact: sagin@metu.edu.tr 2 Bio: Elena Antonova-Ünlü is an associate professor at Hacettepe University, Department of Translation and Interpreting. Her domains of research include bi- and multilingualism, bilingual first-language acquisition, second- and third-language acquisition and sociolinguistics. Received : 17.11.2017 Accepted : 13.12.2017 Published : 26.12.2017 DOI:10.5281/zenodo.7697213 mailto:sagin@metu.edu.tr Subject realization in bilingual Turkish children Sağın-Şimşek, Antonova-Ünlü 194 acquiring languages in BFLA differs from monolingual acquisition. Of all the domains that have been examined in relation to the IH, subject realization in null-subject and non-null-subject languages is probably the most representative one. In the BFLA context, the findings of the studies that examine subject realization were consistent with the L2 research (for instance, see Hacohen & Schaeffer, 2007 for Hebrew–English bilinguals; Müller, Kupisch, Schmitz & Cantone, 2006 for Italian–German bilinguals; Paradis and Navarro, 2003 for Spanish–English bilinguals; Pinto, 2004 for Italian–Dutch bilinguals; Serratrice, Sorace & Paoli, 2004 for Italian–English bilinguals), indicating that bilingual children who acquire a null subject language and a non-null subject language from birth tend to overuse overt pronouns in their null- subject language. These results were interpreted in favor of the IH as evidence that the interface conditions on the use of subject pronouns are susceptible to developmental delays and cross-linguistic influence in BFLA. Regarding the possible sources of vulnerability of interface structures, Sorace (2011) suggested two plausible explanations: (1) differences between bilinguals and monolinguals at the level of knowledge representation that occur due to the interaction of two competing grammatical systems and, (2) differences in processing resources and strategies between monolinguals and bilinguals. While numerous studies demonstrated that the process of subject realization in null-subject languages is affected by other non-null-subject language in bilingual acquisition, which can be attributed to cross-linguistic influence, several other studies provided a piece of evidence that even in cases when both languages of bilinguals are non-null-subject languages, their subject realization might be different from that of monolinguals (Lozano, 2006; Margaza & Bel, 2006; Sorace, Serratrice, Filiaci, & Baldo, 2009). In this respect, Sorace and Serrtrice (2009) and Sorace (2011) suggested that differences between monolinguals and bilinguals at the syntax-pragmatic interface seem to reflect differences in processing rather than cross-linguistic influence. Within this framework, the aim of this study is to contribute to the debate about the possible sources of vulnerability of interface structures by comparing the subject realization of Russian-Turkish and German-Turkish bilingual children, who have been exposed to both languages from birth and to examine the extent to which their dominant Russian or German languages may account for non-monolingual-like subject realization in their non-dominant Turkish language. We believe that the comparison of two groups of bilinguals having different subject realization patterns will allow us to speculate on the role of cross-linguistic influence and potential differences in processing between monolinguals and bilinguals at the domain of syntax- pragmatic interface. The choice of the Russian and German languages as dominant languages of the bilingual participants has been determined by the fact that Russian is a partially null-subject language, while German is a non-null-subject one and Turkish is a null-subject language. Regarding these differences between the languages and assuming potential cross-linguistic influences that might take place between the languages as a possible source of vulnerability of interface structures, it can be hypothesized that the Russian-Turkish bilingual Journal of Child Language Acquisition and Development – JCLAD Vol: 5 Issue: 4 193-207, 2017, December ISSN: 2148-1997 195 children will reveal a better overall performance in the realization of null- subjects in Turkish and their performance is expected to be unmarked or less marked with overuse of overt subjects, if compared with the German- Turkish bilingual children. As for the subject realization of German-Turkish bilingual children, it can be hypothesized that their acquisition would be more prone to overuse of overt subjects in Turkish for two reasons. First, even though overt subjects are realized in both Turkish and German, Turkish as a null-subject language that allows omission of subjects, while German is a non-null subject language that relies only on overt subjects. Second, null-subject realization in Turkish is restricted not only by syntactic but also by pragmatic constraints. The study is structured as follows. First, in order to provide a framework for the study, studies that focused on subject realization among bilinguals will be reviewed. Next, a brief overview of subject realization in the Turkish, Russian and German languages will be described to be able to estimate possible cross-linguistic influences of Russian and German languages on the Turkish language. Then, the present study, the participants and the method will be introduced. Finally, the results of data analysis will be presented and discussed. 1.1. Studies on bilingual subject realization Several studies focusing on subject realization of simultaneous bilingual children who acquire a null-subject and a non-null subject language have suggested that the topic is vulnerable to cross-linguistic influence. The important research in this respect was conducted by Paradis and Navarro (2003) who investigated the subject realization of a Spanish-English bilingual child between the ages 1;9-2;6. The data collected during the natural speech production of the simultaneous bilingual child revealed that the child used more overt subjects in his Spanish than his Spanish monolingual counterparts, which according to the researchers indicated possible cross-linguistic effects from English to Spanish. Similar results were also reported by Serratrice and Sorace (2003) and Serratrice et al. (2004), who collected data from one Italian-English bilingual child, and by Hacohen and Schaeffer (2007), who collected data from one Hebrew-English bilingual child. Based on the inappropriate overuse of overt subjects in Italian of the Italian-English bilingual child and in Hebrew of the Hebrew-English bilingual child, the researchers discussed the vulnerability of the pragmatic constraints resulting from cross-linguistic influence. These studies, in common, suggested that if a null-subject language is acquired together with a non-null subject language, bilingual children would be inclined to use more overt subjects in their null-subject language than their monolingual counterparts. If the development of syntactic and pragmatic knowledge in coordination with each other is a demanding task for young children (Avrutin, 1999), then it would not be wrong to assume that bilinguals who acquire two languages having different subject realization patterns (requiring the acquisition of syntactic parameters and the pragmatic constraints of their languages) will experience more difficulties than monolinguals and that the difficulties may Subject realization in bilingual Turkish children Sağın-Şimşek, Antonova-Ünlü 196 be related to cross-linguistic influence. In fact, the above studies validated this assumption. Nevertheless, if language combination is an essential factor for the existence of cross-linguistic influence, an interesting question would arise how bilinguals who acquire two null-subject languages both of which require syntactic parameters and pragmatic constraints to be developed acquire the appropriate realization of subjects. To our knowledge, very few studies examined the above question. In one of such studies Schmitz, Patuto and Müller (2011) examined three different language combinations German-Italian, German-French and Italian-French, two of these combinations contained both a null-subject (Italian) and a non- null-subject language (French and German), all of which have different pragmatic characteristics. The study highlighted the importance of the language combinations and demonstrated that while German-Italian bilingual children produced too many subject pronouns (which they interpreted as evidence for cross-linguistic influence), such overuse patterns were not observed in Italian-French bilingual children, although French is a non-null-subject language, like English and German. The researchers argued that not all diverse forms can be explained due to cross-linguistic influence. As for the studies examining subject realization among bilinguals whose language combinations include the null-subject Turkish language, not much has been revealed yet. In one of the studies, Haznedar (2010) investigated subject realization in Turkish in spontaneous data collected from one simultaneous Turkish-English bilingual child and one Turkish monolingual child. The researcher reported that the bilingual child made use of overt subjects in Turkish at a rate more than 10 times higher than the monolingual child and the bilingual child’s use of overt subjects was pragmatically inappropriate. The results of the study were interpreted as evidence for cross-linguistic influence from English regarding the realization of overt subjects in the context of Turkish-English bilingual acquisition. Similar results were also reported by Sağın Şimşek (2009), who compared the subject realization of four Turkish monolingual and four Turkish- German bilingual children aged between 5 and 7;3. The study reported high percentage of inappropriate use of overt subjects and subject pronouns by the Turkish-German bilinguals in Turkish in comparison to their Turkish monolingual counterparts who had the tendency to use null-subjects. Accordingly, the study also suggested cross-linguistic influence as the main source of difficulty in acquiring the pragmatic constraints of Turkish with regards to subject realization. This study focusing on the comparison of two different language combinations, one of them containing a null-subject and a non-null-subject language (German-Turkish) and one containing one partially-null-subject and one null-subject languages (Russian-Turkish) might allow us to provide new insights about bilinguals’ subject realization and its relation to cross- linguistic influence. 1.1.1. Subject realization in Turkish Turkish is a null-subject language with subjects that are identified via agreement morphemes on verbs as exemplified in (1a) and (1b) (Kornfilt, Journal of Child Language Acquisition and Development – JCLAD Vol: 5 Issue: 4 193-207, 2017, December ISSN: 2148-1997 197 1997; Enç, 1986; Özsoy, 1987). In Turkish, subjects can be omitted when their interpretations are discourse or context predictable and when there is an overt agreement marker on the predicate. In examples (2a) -(2c), Ali is introduced as the subject of the event (2a) and in the following utterances, (2b) and (2c), it is possible to drop the subject (Ali) since it is discourse predictable (1a) Ben kitab-ım-ı oku-yor-um. I book-POSS-ACC read-PROG-1.sg ‘I’m reading my book.’ (1b) Ø kitab-ım-ı oku-yor-um. book-POSS-ACC read-PROG-1.sg ‘I’m reading my book.’ (2a) Ali ev-e erken gel-di. Ali home-DAT early come-PAST-3.sg ‘Ali came home early.’ (2b) Önce ev-i temiz-le-di. First house-ACC clean-CAUS.-PAST-3.sg ‘First, he cleaned the house.’ (2c) Sonra yemeğ-i piş-ir-di. Then meal-ACC cook-CAUS.-PAST-3.sg ‘Then he cooked the meal.’ The realization of subjects in Turkish is mainly determined by pragmatic considerations of the speaker such as expressing new and/or old information, contrast, subject change or indicating emphasis/focus (Erguvanlı, 1984; Kornfilt, 1984; 1997; Enç, 1986; Özsoy, 1987). As presented in (3a)-(3c) in order to indicate new information, contrasting, changing the subject and emphasizing a constituent, subjects are overtly stated. (3a) Ödev-i ben yap-tı-m Ali yap-ma-dı. Homework-ACC I do-PAST-1.sg Ali do-NEG-PAST-3.sg ‘I did the homework, Ali didn’t do it.’ (3b) Ali ev-de uyu-yor-du. Ali home-LOC sleep-PROG-PAST-3.sg ‘Ali was sleeping at home.’ (3c) O uyu-rken, Ayşe ev-e gel-di. He sleep-CON Ayşe home-DAT come-PAST-3.sg ‘While he was sleeping, Ayşe came home.’ Subject realization in bilingual Turkish children Sağın-Şimşek, Antonova-Ünlü 198 1.1.2. Subject realization in German German is considered to be a non-null-subject language in which the use of subject realization is mainly restricted by syntactic rules. Nevertheless, only rarely, German allows the use of null-subjects. However, the use of null- subjects is not regulated by pragmatic rules rather it only entails informal speech. For instance, both (4a) and (4b) present the same information with a difference in the level of formality. While example (4a) with a null-subject indicates informal speech, example (4b) with an overt subject indicates formal speech style. Thus, subject omission it is not pragmatically but syntactically determined pattern (Müller, 2007). (4a) Ø Hab das schon gemacht. Have that already done. ‘I have already done that.’ (4b) Ich hab das schon gemacht. I have that already done ‘I have already done that.’ 1.1.3. Subject realization in Russian Russian is a partial-null-subject language which means that Russian allows null-subjects but under more restricted conditions than consistent null- subject languages. Subject omission is determined by syntactic and discourse conditions in Russian and is limited to the 1st and 2nd person in finite clauses, and 3rd person pronouns “bound by a higher argument” (a context that Holmberg (2005: 539) stated is “rather poorly understood”). Generic pronouns are also not realized overtly. To illustrate, examples (5a) and (6a) define the context, making the subjects in the following examples of (5b) and (6b) discourse predictable and therefore in example (5b) and (6b) the subjects can be omitted. (5a) Ты что делаешь? you what do-2.sg-PROG ‘What are you doing?’ (5b) Ø Доклад готовлю report-ACC prepare-1.sg-PROG ‘I am preparing a report.’ (6a) Где Олег? where Oleg-NOM ‘Where is Oleg?’ (6b) Ø Работает в библиотеке. work-3.sg-PROG ‘He is working in the library.’ Journal of Child Language Acquisition and Development – JCLAD Vol: 5 Issue: 4 193-207, 2017, December ISSN: 2148-1997 199 1.2. The study This study aims to compare the subject realization of Russian-Turkish and German-Turkish bilingual children in their non-dominant Turkish language. Precisely, the study aims to answer the following research questions: 1. Does the subject realization in Turkish of the German-Turkish bilingual children differ from that of Turkish monolingual children? 2. Does the subject realization in Turkish of the Russian-Turkish bilingual children differ from that of Turkish monolingual children? 3. Does the subject realization in Turkish of the German-Turkish and Russian-Turkish bilingual children differ from each other? 2. Methodology 2.1. The Participants The participants of this study include three groups of children (Table 1). Group 1 is a control group consisting of four normally developing monolingual Turkish children with the age range of 4;2 to 8;0 (Mean=6;5). The monolingual data examined in the study is taken from Sağın Şimşek (2009). Group 2 consisted of three German-Turkish bilingual children with the age range from 6;5 to 8;7 (Mean=7;5). The German-Turkish participants had been raised in the German-dominant environment and had been exposed to both languages from birth. The children were all born in Germany in families where mothers were native speakers of German and fathers were native speakers of Turkish. The German-Turkish participants had been living in Germany but they had regular contacts with their Turkish relatives and had been visiting Turkey twice a year. Due to the German-dominant context of the language acquisition, the German language of the participants developed as dominant in their linguistic repertoire. Group 3 consisted of two Russian-Turkish bilingual children whose ages were 7;0 and 8;6 (Mean=7;8). The Russian-Turkish children had been raised in the Russian-dominant environment and had been exposed to both languages from birth. The children had both been raised in Russia from birth in families where mothers were native speakers of Russian and fathers were native speakers of Turkish. The Russian-Turkish participants had regular contacts with their Turkish relatives and had been visiting Turkey two-three times a year. Due to the Russian-dominant context of the language acquisition, the Russian language developed as dominant in their linguistic repertoire. Subject realization in bilingual Turkish children Sağın-Şimşek, Antonova-Ünlü 200 Table 1 Information about the participants Turkish monolingual controls German-Turkish bilinguals Russian-Turkish bilinguals Age Age Age C1 4;2 6,5 7;0 C2 6;8 7,3 8;6 C3 7;2 8,7 C4 8 Mean 6,5 7;5 7;8 2.2. Data collection Taking into consideration the young ages of the participants, the data were collected recording their natural language production. The participants’ utterances were recorded while they were describing their past experiences such as their summer holidays and school experiences and while they were in interaction with their parents. 3. Results 3.1. Subject realization of the Turkish monolinguals The results regarding the subject realization of Turkish monolinguals were obtained from Sağın Şimşek (2009). Table 2 presents the distribution of null- subjects and overt subjects in the monolingual data referring to their functions and the frequency of use. As can be seen, the Turkish monolingual children used both null-subjects and overt subjects in their language production and most of their usage was considered accurate and appropriate. The monolinguals used null-subjects when the same subject was used within the same discourse and when the subject was discourse predictable and there were very few instances of misuse of null-subjects. The use of overt subjects was preferred in cases when subjects presented new information, when they were contrasted and/or emphasized. As for the inappropriate use of overt subjects, they were few in number. Table 2 Distribution of subject realization in the Turkish monolingual data Distribution of null-subject Distribution of overt subjects Age Utterance Same subject Context recover- able Misuse New info. Contrast Emphasis Overuse C1 4;2 118 46 24 2 15 16 15 0 C2 6;8 173 40 72 0 44 3 13 1 C3 7;2 149 53 54 2 22 3 9 6 C4 8 133 45 33 3 41 4 5 2 The analysis of the Turkish monolinguals’ use of null and overt subjects allowed us to conclude that the Turkish monolingual children had acquired the syntactic and pragmatic constraints of subject realization in Turkish around the age of seven and use both null and over subjects accurately and appropriately. Journal of Child Language Acquisition and Development – JCLAD Vol: 5 Issue: 4 193-207, 2017, December ISSN: 2148-1997 201 3.2. Subject realization of the German-Turkish bilinguals Analysis of the German-Turkish bilingual data revealed differences between the subject realization of the Turkish monolingual and German-Turkish bilingual children. The results presented in Table 3 demonstrate that the German-Turkish bilingual children also used patterns of null-subject for the same functions as their monolingual counterparts did although with relatively higher number of misuses. Table 3 Distribution of subject realization in the German-Turkish bilingual data Distribution of null-subject Distribution of overt subjects Age Utterance Same subject Context recover- able Misuse New info. Contrast Emphasis Overuse C1 6,5 96 39 15 5 7 8 4 18 C2 7,3 101 36 18 6 11 2 9 19 C3 8,7 97 29 21 5 21 6 13 22 When the distribution of the overt subjects was examined, it was observed that the German-Turkish bilinguals used overt subjects in order to indicate their pragmatic intentions like Turkish monolingual children. However, the data revealed that the frequency of the inaccurate use of overt subjects was considerably higher in the German-Turkish bilingual data than the Turkish monolingual data. These findings were in line with our expectations. Since Turkish is a null-subject language with syntactic and pragmatic constraints and German does not allow subject omission in general, acquisition of German and Turkish demanded the bilingual children to acquire different constraints of different domains. Examples (7a-7i) illustrate one of the bilingual participants’ tendency to overuse overt subjects, specifically the first person pronoun, though the use of overt subject is not syntactically obligatory (as recoverable via the person marker on the verb) and pragmatically necessary. (7a) Ben yaz tatil-im-de Türkiye’ye git-ti-m. I summer holiday-POSS-LOC Turkey-DAT go-PAST-1.sg ‘I went to Turkey last summer.’ (7b) Orada akraba-lar-ım var. There relative-3.pl-POSS exist ‘There I have relatives.’ (7c) #Biz onlar-ı ziyaret et-ti-k ve beraber çok eğlen-di-k. We they-ACC visit-do-PAST-3.pl and together very fun-PAST-3.pl ‘We visited them and we had fun together.’ (7d) #Ben kuzen-ler-im-le oyun oyna-dı-m, deniz-de yüz-dü-m. I cousin-3.pl-POSS-AC game play-PAST-1.sg sea-LOC swim-PAST-1.sg Subject realization in bilingual Turkish children Sağın-Şimşek, Antonova-Ünlü 202 ‘I played gamed with my cousins and swam in the sea.’ (7e) #Biz büyük kuzen-im için düğün yap-tı-k. We old cousin-POSS for wedding make-PAST-3.pl ‘We gave a wedding party for my elder cousin.’ (7f) Akşam düğün-de eğlen-di-k, dans et-ti-k. Evening wedding-LOC fun-PAST-3.pl dance-PAST-3.pl ‘We had fun and dance at the wedding that evening.’ (7g) #Ben mutlu bir tatil yap-tı-m. I happy one holiday make-PAST-1.sg ‘I had a happy holiday.’ (7h) Sonra aile-m-le Almanya’-ya geri dön-dü-m. Then family-POSS-AC Germany-DAT back return-PAST-1.sg ‘Then I came back to Germany with my family.’ ve #ben burada tatil yap-tı-m. and I here holiday make-PAST-1.sg ‘and I made a holiday here.’ (7i) #Ben tekrar Türkiye’-ye git-mek isti-yor-um. I again Turkey-DAT go-INF want-PROG-1.sg ‘I want to go to Turkey again.’ It is possible to hypothesize that the inappropriately use of overt subjects in null-subject Turkish observed in the German-Turkish bilingual data can be attributed to cross-linguistic influence from German, since the latter is a non-null-subject language. 3.3. Subject realization of the Russian-Turkish bilinguals Analysis of the data obtained from the Russian-Turkish bilingual children revealed that the Russian-Turkish bilingual children use both null-subjects and overt subjects in their utterances. Only in one instance one of the bilingual children misused the null-subject in Turkish. This result was not unexpected as both Russian and Turkish languages are null-subject languages. Table 4 Distribution of subject realization in the Russian-Turkish bilingual data Distribution of null-subjects Distribution of overt subjects Age Utterance Same subject Context recover- able Misuse New info. Contrast Emphasis Overuse C1 7;0 86 36 2 1 6 0 0 41 C2 8;6 72 24 18 0 11 0 0 19 However, when the distribution of the overt use of subjects were examined, as presented in Table 4, unexpected results were revealed. We predicted that Journal of Child Language Acquisition and Development – JCLAD Vol: 5 Issue: 4 193-207, 2017, December ISSN: 2148-1997 203 the combination of Russian-Turkish would not display observable differences between the Turkish monolinguals and the Russian-Turkish bilinguals regarding the subject realization patterns because both Turkish and Russian are null-subject languages allowing the omission of subjects. Thus, we expected more or less similar subject realization patterns to be used by the Turkish monolingual and Russian-Turkish bilingual children. However, contrary to our expectations the Russian-Turkish bilinguals’ overt subject realization patterns were quite different from those of the Turkish monolinguals. The bilinguals had the tendency to make use of overt subjects more extensively than the use of null-subjects. In cases when the subject that was introduced in the discourse was also the doer of the following utterances and therefore, discourse-recoverable via the agreement markers on the verbs, the Russian-Turkish bilingual children preferred using overt subjects rather than omitting them. The overt use of the subjects when they were syntax and discourse predictable was an unexpected pattern. As presented in Table 4, of the 158 utterances of the two Russian-Turkish bilingual children, 60 included overused, inappropriate overt subjects. (8a) İlk önce #ben git-ti-m Konya’-ya. First I go-PAST-1.sg Konya-DAT ‘First I went to Konya.’ (8b) Konya’-da #ben çok güzel vakit geçir-di-m akraba-lar-ım-la Konya-LOC I very good time spend-PAST-1.sg relative-PL-POSS-with ve arkadaş-lar-ım-la. and friend-PL-POSS-with ‘I spent good time with my relatives and friends in Konya.’ (8c) Her gün #ben ve benim aile-m bir yere gid-iyor-du-k. Every day I and my family-POSS a place-DAT go-PROG-PAST-3.pl ‘Every day we used to go somewhere with my family.’ (8d) Sonra #ben ve #benim aile-m Kapadokya’ya git-ti-k. Then I and my family-POSS Kapadokya-DAT go-PAST-3.pl ‘Then we went to Kapadokya with my family.’ (8e) Her gün #biz orada güzel gez-di-k. Every day we there nice visit-PAST-3.sg ‘Everyday we visited nice places there.’ (8f).Sonra #ben Rusya’-ya dön-dü-m. Then I Russia-DAT return-PAST-1.sg ‘Then I came back to Russia.’ The inappropriately overuse of overt subjects in Turkish that are exemplified in the examples (8a-8f) observed in the Russian-Turkish bilingual data cannot be attributed to the cross-linguistic influence from Russian since the Subject realization in bilingual Turkish children Sağın-Şimşek, Antonova-Ünlü 204 latter is also a null-subject language. Moreover, contrary to our expectations, the similarities between Russian and Turkish regarding subject realization seemed to have no facilitating effect on subject realization in Turkish. 4. Discussion and Conclusions The analysis of the data revealed that the language behaviour of the German-Turkish and the Russian-Turkish bilingual children was different from the Turkish monolingual children in that both the German-Turkish and the Russian-Turkish bilingual children overused overt subject pronouns in their Turkish. These results are consistent with the findings of the studies that examined subject realization in null-subject languages of bilinguals suggesting that the bilinguals tended to overuse subjects notwithstanding the fact that their other language is a non-null or null-subject one. Research on the issue commonly presented cross-linguistic influence as a factor that accounts for vulnerability of subject realization as a syntax-pragmatics interface phenomenon. This vulnerability is related to the defined conditions of cross-linguistic influence (Hulk & Müller, 2000; Müller & Hulk, 2000; 2001) according to which a grammatical property is a vulnerable grammatical phenomenon when it is similar at the surface level in both languages and when it is at the interface between syntax and pragmatics in one language. However, our findings cannot be merely explained on the basis of cross-linguistic influence since both the German-Turkish and Russian- Turkish bilingual children demonstrated similar subject realization patterns; namely, they both overused overt subjects. If cross-linguistic influence were the only factor that might account for the deviations in subject realizations, different patterns would be expected from the two bilingual groups in their realization of subjects in Turkish; with the Russian-Turkish participants performing better in using the null-subjects. However, our data revealed that similar to the German-Turkish bilinguals, the Russian-Turkish bilingual children overused overt subjects and subject pronouns, particularly the 1st person pronoun, even though both Turkish and Russian are null-subject languages allowing subject omission. Hence, the inappropriate subject realizations of the Russian-Turkish bilingual children cannot be explained as evidence for cross-linguistic influence only but conceivably as a language processing problem. Serratrice et al. (2004) in their article where they compared subjects and objects in the English-Italian bilingual and monolingual acquisition suggested that languages with pragmatic constraints can be considered more complex than others which do not have any pragmatic constraints and that “the coordination of syntactic and pragmatic knowledge is a demanding task for young children in general” (p.201). Similarly, we believe that the Russian- Turkish bilingual children who use null-subjects in their Russian had to figure out the extent of null-subject realization in their Turkish as the choice between whether to use or to omit the subjects is not only regulated by syntactic but also pragmatic constraints in Turkish. Therefore, it may be hypothesized that if the acquisition of pragmatic constraints in addition to the invariant syntactic rules required additional demand from the bilinguals, and therefore perceived as more complex, then the bilingual children might have preferred to rely on the underlying syntactic structure of Turkish. Journal of Child Language Acquisition and Development – JCLAD Vol: 5 Issue: 4 193-207, 2017, December ISSN: 2148-1997 205 Turkish is syntactically a SOV language but due to pragmatic intentions of the speakers it is possible to use other word orders with or without omitting subjects. It might be assumed that in cases when a language of bilinguals necessitates acquisition of pragmatic constraints which are context-bound and require appropriate evaluation of the context, a tendency to use the default option in the languages might be a solution for overcoming the processing difficulty. Accordingly, it is possible to assume that inappropriate subject realizations might not only characterize subject realization of bilingual children, but also monolingual children. However, the monolingual data analysed in the present study and in the other studies that compared subject realization of the Turkish monolinguals with the Turkish-English bilingual children (Haznedar, 2010) provided evidence that Turkish monolinguals acquire pragmatic constraints of their languages quite early and start using subject omission at very early ages. Therefore, the inappropriate subject realizations should be interpreted as a peculiarity of the bilingual language use and most probably is related not only to the availability of two different linguistic systems in the repertoire but also to the limited input bilinguals might receive in one of their languages. However, our data do not allow us to draw conclusions in this respect. Unquestionably, considering the number of children involved in the study, we accept that our results are not indisputable and further research with more participants is needed to verify our data. References Avrutin, S. (1999). Development of the syntax–discourse interface. Kluwer: Dordrecht. Belletti, A., Bennati, E. & Sorace, A. (2007). Theoretical and developmental issues in the syntax of subjects: evidence from near-native Italian. Natural Language and Linguistic Theory, 25, 657-689. Enç, M. (1986). Topic switching and pronominal subjects in Turkish. In: Slobin DI and Zimmer K (eds) Studies in Turkish linguistics. Amsterdam: John Benjamins, 195-209. Erguvanlı, E. (1984). The function of word order in Turkish grammar. Berkeley, CA: University of California Press. Hacohen, A. & Schaeffer, J. (2007). Subject realization in early Hebrew/English bilingual acquisition: The role of crosslinguistic influence. Bilingualism: Language and Cognition, 10, 333-44. Haznedar, B. (2010). Transfer at the syntax–pragmatics interface: Pronominal subjects in bilingual Turkish. Second Language Research, 26, 3, 355–378 Holmberg, A. (2005). Is there a little pro? Evidence from Finnish. Linguistic Inquiry 36, 533–64. Hulk, A., & Müller, N. (2000). Cross-linguistic influence at the interface between syntax and pragmatics. Bilingualism: Language and Cognition 3, 227–244. Subject realization in bilingual Turkish children Sağın-Şimşek, Antonova-Ünlü 206 Kornfilt, J. (1984). Case marking, agreement and empty categories in Turkish. Unpublished PhD thesis, Harvard University, Cambridge, MA. Kornfilt, J. (1997). Turkish. London and New York: Routledge. Lozano, C. (2006). The Development of the Syntax-Information Structure Interface: Greek Learners of Spanish. In V. Torrens & L. Escobar (Eds.), The Acquisition of Syntax in Romance Languages, 371-399. Amsterdam/Philadelphia: John Benjamins. Margaza, P. & Bel, A. (2006). Null Subjects at the Syntax–Pragmatics Interface: Evidence from Spanish Interlanguage of Greek Speakers. In M. G. O’Brien, C. Shea & J.Archibald (Eds.), Proceedings of GASLA 2006, 88–97. Somerville, MA: Cascadilla Press. Müller, N. (2007). Some notes on the syntax–pragmatics interface in bilingual children. German in contact with French. In J. Rehbein, C. Hohenstein & L. Pietsch (Eds.), Connectivity in Grammar and Discourse, 101-135. Amsterdam/Philadelphia: John Benjamins [Hamburg Studies on Multilingualism]. Müller, N., & Hulk, A. (2001). Cross-linguistic influence in bilingual language acquisition: Italian and French as recipient languages. Bilingualism: Language and Cognition, 4, 1–21. Müller, N., Kupisch, T., Schmitz, K. & Cantone, K. (2006). Einführung in die Mehrsprachigkeitsforschung. Tübingen, Gunter Narr Verlag. Özsoy, S. (1987). Null subject parameter in Turkish. In: Boeschoten HE and Verhoeven LT (eds) Studies on modern Turkish: Proceedings of the third conference on Turkish linguistics. Tilburg: Tilburg University Press, 82–91. Paradis, J. & Navarro, S. (2003). Subject realization and crosslinguistic interference in the bilingual acquisition of Spanish and English. Journal of Child Language, 30, 1–23. Pinto, M. (2006). Subject pronouns in bilinguals: Interference or maturation? In V. Torrens & L. Escobar (Eds.) The acquisition of syntax in Romance languages. Amsterdam & Philadelphia: John Benjamins Publishing Company, 331–350. Sağın Şimşek, Ç. (2009). Adıl düşürme değiştirgeninin Türkçe tek dilli ve Türkçe-Almanca iki dilliler tarafından işletimi. 23. Ulusal Dlbilim Kurultayı Bildirileri. 64-77. Schmitz, K., Patuto, M. & Müller, N. (2011). The null-subject parameter at the interface between syntax and pragmatics: Evidence from bilingual German–Italian, German–French and Italian–French children. First Language, 32, 1-2, 205–238. Serratrice, L. & Sorace, A. (2003). Overt and null subjects in monolingual and bilingual Italian acquisition. In: Beachley B, Brown A, and Conlin F (eds) Proceedings of the 27th annual Boston University Conference On Language Development. Somerville, MA: Cascadilla Press, 739–50. Sorace, A., Serratrice, L., Filiaci, F. & Baldo, M. (2009). Discourse Conditions on Subject Pronoun Realization: Testing the Linguistic Intuitions of Bilingual Children. Lingua, 119, 460–477. Serratrice, L., Sorace, A., & Paoli, S. (2004). Cross-linguistic influence at the syntax–pragmatic interface: Subjects and objects in English–Italian bilingual and monolingual acquisition. Bilingualism: Language and Cognition, 7, 183–205. Journal of Child Language Acquisition and Development – JCLAD Vol: 5 Issue: 4 193-207, 2017, December ISSN: 2148-1997 207 Sorace, A., Filiaci, F., (2006). Anaphora resolution in near-native speakers of Italian. Second Language Research, 22, 339–368 Tsimpli, T., Sorace, A., Heycock, C. & Filiaci, F. (2004). First language attrition and syntactic subjects: a study of Greek and Italian near-native speakers of English. International Journal of Bilingualism, 8, 257–277.