The use of verbal morphology in Turkish as a third language: The case of Russian–English

Abstract

Aims and Objectives:

Several studies suggest that third language acquisition (TLA) is marked with complex patterns of language interaction. However, it is not clear yet to what extent multilinguals activate each of their background languages in TLA, as various factors may trigger the activation of one of the previously learnt languages. This study aims to contribute to the discussion by examining the use of verbal morphology in third language (L3) Turkish of Russian–English–Turkish trilinguals. We investigate whether the use of verbal morphology in L3 Turkish of Russian–English–Turkish trilinguals differs from that of Turkish native speakers and in the case of a deviation, which of the background languages can account for it.

Design/Methodology/Approach:

The study is done within the framework of cross-linguistic influence.

Data and Analysis:

The data are collected from eight native speakers of Russian who are highly proficient in their L2 English and L3 Turkish, and use their three languages more or less equally every day. Fictional narratives are employed in the study as the tool for data collection. The use of finite and non-finite verbal forms in Turkish is compared with the baseline data coming from native speakers of Turkish.

Findings/Conclusions:

The results reveal that the trilingual participants mainly use the verbal morphology consistently with native speakers of Turkish; nonetheless, two deviations from the Turkish baseline stand out: tendency to use finite verbal forms in their noun clauses and inconsistency in the use of temporal-aspectual markers in finite verbal forms. The former can be attributed to the L1 and L2 interlanguage, while the latter can be attributed to the influence of particularly L1 Russian.

Originality, and significance/Implications:

Thus, the study provides evidence that the source of interference into L3 production is likely to occur due to the cross-linguistic influence from L1, which in turn might imply the predominant position of L1 over sequentially learnt languages at the morphosyntactic level in the language processing.

Keywords

Turkish third language cross-linguistic influence Russian–English interlanguage verbal morphology

Introduction

Research studies on cross-linguistic influence in third language acquisition (TLA) have shown that there is a growing interest in investigating how three languages interact with one another during the third language (L3) learning process. While in the second language (L2) learning process, the direction and route of interaction is quite straightforward, in the case of the L3 learning process, interaction patterns might be much more complicated and diverse in nature. The chief reason for such intricacies is that in TLA the learner’s both native and non-native languages can be the sources of influence (Cenoz, 2001; Hammarberg, 2001; Möhle, 1989; Ringbom, 1987, 2001). Besides, L3 learners are acknowledged to possess “unique and specific linguistic configurations” (De Angelis & Selinker, 2001), the state of which is conceptualized with the term “multicompetence” by Cook (1995). There are a number of studies investigating the roles of already known languages on acquisition of L3. However, these studies reveal contradicting results. While some studies demonstrate that the role of the native language (L1) is negligible particularly when the L2 shares common linguistic features with the L3, some others acknowledge that while producing in the L3, trilinguals activate their L1 and L2 knowledge to considerable extents (Dewaele, 1998; Williams & Hammarberg, 1998). Along with this discussion, scholars acknowledge that cross-linguistic influence in L3 acquisition seems to be influenced by several factors, such as typological similarity between languages (Anderson, 1983; Kellerman, 1983, 1995; Ringbom, 2001), learners’ proficiency level in the languages they know (De Angelis & Selinker, 2001; Hammarberg, 2001; Kellerman, 1983; Odlin, 1989), the age of acquisition (Cenoz, 2001), the amount of language exposure (Dewaele, 2001; Jarvis, 2000), frequency of use (Hammarberg, 2001; Mägiste, 1986) and cognitive mode of learners, namely monolingual and bilingual modes (Dewaele, 1998, 2001; Grosjean, 1995, 2001).

In line with these studies, we aim to contribute to these discussions, by investigating L3 Turkish acquisition by adult native speakers of Russian who are highly proficient speakers of English, using the language for professional purposes. More specifically, within the framework of cross-linguistic influence, our study aims to answer the following research questions:

Does the use of verbal morphology in the written production of Russian–English–Turkish trilinguals in their L3 Turkish differ from that of native speakers of Turkish?

In the case of such a difference, which of the background languages (Russian or English) can account for the deviation?

We believe that this study will be innovative from several points of view. In terms of independent variables we tried to control typological similarity by choosing participants with an L1, L2 and L3 that are typologically distant from each other. The Russian language is an East-Slavic language that retains syntactic-inflectional structure and a rich set of morphological categories, has a complex inflectional, primarily highly fusional, morphology and is marked with so-called free word order (Comrie & Corbett, 1993; De Bray, 1951, 1980; Kondrasov, 1986). The English language is a Germanic language that has an analytic structure with minimal inflectional morphology and exhibits subject–verb–object (SVO) order. Turkish belongs to Turkic languages and it is an agglutinating language in which verbs and nouns are richly inflected with suffixes. As for its word order, Turkish is considered to have a so-called rigid subject–object–verb (SOV) order that also permits inverted word orders (Göksel & Kerslake, 2005; Kornfilt, 1997). As for the language proficiency of the participants, frequency of use and the amount of language exposure, we collected the data from participants who are highly proficient in their L2 English and L3 Turkish and use both L1 Russian and L2 English more or less equally in formal and informal settings every day. Since the participants of this study consist only of adult multilinguals, the role of the age factor in the TLA is left beyond the scope of this study.

The focus of this study is centred on the use of verbal morphology in L3 Turkish production, as Turkish verbal morphology is reported to be difficult for bi-/multilinguals due to its rich inflectional system. Therefore, we will first give a brief description of verbal morphology in the Turkish language and its acquisition in multilingual contexts. Next, we will present the design of the study, including method, participants and data analysis. Finally, our findings and discussion will be presented.

Verbal morphology in the Turkish language

Verbal morphology in the Turkish language is traditionally divided into two major classes: finite verbal forms and non-finite verbal forms. Finite verbal forms make a simple/main clause in Turkish, while non-finite forms appear in the subordinate clauses, as it is illustrated by examples below.

(1)

Dün yeni bir araba aldım. [MC (finite)] Dün yeni bir araba al-dı-m Yesterday new a car buy-PAST-1stSG

(2)

Dün [okula gelmediğini] [duydum]. [Sub. NC (non-finite)] [MC(finite)] Dün [okul-a gel-me-diğ-i-ni] [duy-du-m] Yesterday [school-DAT come-NEG-NOM-POSS-ACC] [hear-PAST-1stSG] “I heard that you did not come to the school yesterday.”

(3)

Her gün [okula giderken] [arkadaşımı görüyorum]. [Sub. ADVC. (non-finite)] [MC (finite)] Her gün [okul-a gid-erken] [arkadaş-ım-ı gör-üyor-um] Every day [school-DAT go-CON] [friend-POSS-ACC see- PRE IMPF-1stSG] “While going to the school, I see my friend every day.”

(4)

Yarın [görmediğim] [bir arkadaşımı ziyaret edeceğim]. [Sub. AC. (non-finite)] [MC (finite)] Yarın [gör-me-diğ-im] [bir arkadaş-ım-ı ziyaret edeceğ-im] Tomorrow [see-NEG-NOM-1stSG] [a friend-POSS-ACC visit- FUT-1stSG] “Tomorrow I will visit a friend of mine whom I haven’t seen for a long time.”

In Turkish, verbal markers may express tense/aspect and/or mood simultaneously. Because differentiating these functions is not always straightforward, Johanson (1994) merges these two categories into the hybrid category of “aspect-temporal” in analysing the verb paradigm in Turkish. In this study we will use TAM to refer to tense/aspect and/or modality markers of the verbs.

To briefly present functions of these verbal markers, finite verbal forms are formed by using simple verbal markers -DI (the past perfective), -mIş (the evidential/perfective), -(A/I)r/z (the aorist), -(y)AcAK (the future) and -(I)yor (the imperfective with an aspectual meaning) or by using a simple form together with a copula (-DI, -mIş and -sA.). Formation of non-finite verbal forms, on the other hand, requires some other suffixes to be used on the verbs. For instance, in adverbial clauses -kEn, -IncE, -ErEk, -mEk, -Ip, in relative clauses -(y)An, -DIK or -(y)AcAK and in noun clauses -mAK, -mA, -DIK, -(y)AcAK or -(y)Iş are used.

It is important to note that although the most typical type of subordinate clauses in Turkish is non-finite, there is a limited usage of subordinate clauses where finite verbal forms incorporating typically the subordinators ki, diye and gibi are used (Göksel & Kerslake, 2005; Kornfilt, 1997).

(5)

[Bir kızınız olmuş diye ] duyduk. [Sub. NC (finite)] [MC (finite)] [Bir kız-ınız ol-muş diye ] duy-duk. [a girl-POSS-2ndPL be- EVD-PF-3rdSG that] hear- PAST-1stPL “We heard [that you have had a daughter].”

(6)

Sanıyorum [ ki eşinden ayrılmak isti-yor]. [MC (finite)] [Sub. NC (finite)] San-ıyor-um [ ki eş-in-den ayrılmak isti-yor]. [Think-PRE-1stSG] [that spouse-POSS-ABL leave-INF want-PRE-3rdSG]. “I think [(that) s/he wants to leave his/her spouse].”

(7)

Sana [ilgilenmiyorum gibi] gel- ebil-ir. [Sub. NC (finite)] [MC (finite)] Sana [ilgilen-mi-yor-um gibi] gel- ebil-ir. [care-PRE-NEG-1stSG like] [come-PROB-AORİST-3rdSG] “It may seem to you [that I do not care].”

Taking into consideration that the formation and use of finite and non-finite verbal forms in the Turkish language is incompatible, it was decided to analyse them separately in this study.

The acquisition of finite forms in Turkish is looked into by Aksu-Koç (1994), who examines the use of TAM in Turkish narratives based on the data obtained from monolingual children aged between 3 and 5 and monolingual adults within an age group of 20–24. Based on her analysis of the aspects as “temporal deixes”,¹ it is stated that in Turkish narratives one finite form functions as “anchor tense” or “favoured tense”. While the adult participants show consistency in their use of anchoring tense, shifts are observed in the children’s data.

Rehbein and Karakoç (2004) investigate finite elements in the retellings of Turkish monolingual and Turkish–German bilingual children. They report that the language behaviour related to the use of aspectotemporal elements of the bilingual children differ from that of their monolingual participants. The researchers state that the bilinguals shift between aspectotemporal elements, which is not consistent with the monolingual Turkish data (2004, p. 142).

In another study, Karakoç (2007) studies the connective role of aspectotemporal elements in Turkish used by monolingual Turkish children and Turkish children growing up as bilinguals in Germany. The analysis of the spoken data show that the bilingual children’s production differs from their monolingual counterparts. The bilingual children have the tendency to use -(I)yor (present imperfective) and -DI (past perfective) forms in their narratives to maintain connectivity, while the monolingual children also use -mIş (perfective aspect/evidential modality) form. Moreover, Karakoç reports a highly frequent use of temporal-deictic expressions, such as o zaman (in that time), sonra (than), ondan sonra (after that), etc., in the bilingual group.

Aarssen (2001) examines how Turkish monolingual and Turkish–Dutch bilingual children express temporal forms in their narratives. He reports that while both groups had difficulties in establishing “a stable temporal frame”, the bilingual children, echoing the above-mentioned studies, experience more problems. Nevertheless, the researcher points out that with an increasing level of proficiency in the language and the maturation of the participants, these difficulties in establishing a temporal frame decline.

Similar results are reported by Akıncı (1998) from the French context. Turkish–French children aged 5–10 born of immigrant parents in France reveal no clear and consistent “anchor tense”. However, at the age of 9–10 the children begin to use either present or past tense as the dominant one. In a later study, Akıncı (2003) investigates the developmental profile of temporal anchoring of texts of first- and second-generation Turkish immigrants in France, relying on data coming from a narrative and a free language production. The researcher finds that the majority of the monolinguals prefer the present as dominant, while the majority of the first-generation immigrants are inconsistent. However, the bilingual second-generation subjects after the age of 14 “use tense just as the monolingual high-educated participants do” (2003, p. 296). The researcher argues that social class attitude and the level of literacy play an important role in the development of the tense use in the language.

To summarize the findings of the above-mentioned studies, it can be inferred that bilingual populations in general seem to differ from native speakers of Turkish in terms of using finite and non-finite verbal forms in the Turkish language in multilingual contexts.

The study

Stimuli and method

As is identified in the introductory section, the aim of our study is to investigate whether L3 Turkish written production of the adult Russian–English–Turkish trilinguals in their L3 Turkish differs from that of native speakers of Turkish in the ways they use verbal morphology and, in case of such a difference, which of their background languages, Russian or English, might account for the deviation. The research questions of our study determine the choice of the material for eliciting data. Fictional narratives are employed into the study as the tool for data collection. Narratives are commonly used in the multilingual research, as they are able to provide the researcher with data in context, which is authentic, and allow studying language properties that occur only in connected speech (Pavlenko, 2008). In addition, narratives have also proved to be a feasible tool for answering research questions involving morphological development of multilinguals and cross-linguistic influence in their language use, in particular (Berman, 1999; Kupersmitt & Berman, 2001; Lanza, 2001; Pavlenko, 2003; Polinsky, 2008; Viberg, 2001, among others). Thus, we ask our participants to produce a fictional narrative after watching a suggested film in the Turkish language. The name of the film is “Juno”. To briefly summarize the film, Juno, the protagonist, at the age of 15, falls pregnant and decides to give the baby up for adoption. For this purpose, she finds a couple and spends some time with them. Since the husband does not feel ready to become a father, he decides to get divorced. In the end, Juno gives the baby to the woman/wife, who is still willing to become a mother.

The choice of the film used in the study is determined by two major factors: first of all, the story, which should be up to date and appealing for modern young women; secondly, the language spoken in the film, which should be contemporary and comprehensible for a wider audience. The participants are interviewed individually and given the following task.

“Here is a film named ‘Juno’. Would you, please, watch the film and write down what happened in that film in Turkish”. Upon receiving the narratives in Turkish, in order to be able to examine the background roles of the already known languages, the participants are asked to reproduce their narratives in Russian and English. The instruction is given in the Turkish language to activate the monolingual Turkish mode of the participants (Grosjean, 2001). The participants are asked to write their narratives in pen. The subjects do not have any time limitation for their writing.

Participants

The subjects of the study are eight working Russian adult females in their 30s (M = 34.1), who had been living in Turkey for between 6 and 15 years (M = 12.3) at the moment of the data collection. All of them arrived in Turkey for a permanent residence in their 20s (M = 22.4). During the first year after their arrival in Turkey, all of the participants started learning Turkish at various language courses, at the end of which they took the final comprehensive proficiency exam and, having scored above 90 out of 100, were certified as advanced users of the Turkish language. All the participants come from the upper-middle-class families and have pedagogical, linguistic educational backgrounds. The participants are married to non-Russian-speaking Turkish males and use Turkish in their everyday communication mostly. Along with the Turkish language, all the participants have native-like proficiency in English; all of them have, at least, an undergraduate degree in the English language teaching or philology and use the language professionally at work, where English is the medium of instruction. Although the group can be considered as homogeneous, it is necessary to state that only one of the participants, along with the English language, also uses the Turkish language in the working environment for official communication. The data coming from the Russian participants are compared with the baseline data received from eight working Turkish native-speaking females who are of a similar age (M = 35). All the Turkish participants also come from upper-middle-class families and are married. The baseline group is given the analogous instruction to fulfil the same task.

Method of data analysis

First the data are transcribed using the CHAT format of CHILDES. As the focus of the study is on deviations from the Turkish monolingual norms observed in the use of verbal morphology by the Russian–English–Turkish trilinguals, error analysis is used as the main method of data analysis.

Two native speakers of Turkish, who have degrees in Turkish linguistics, are asked to judge the narratives produced by both main and baseline groups in terms of grammaticality and acceptability. The judges are asked to identify ungrammatically and/or unacceptably used forms and provide an explanation for their evaluations. The judges are not informed about the background languages of the participants. Further, all the identified forms are categorized and the frequencies of the grammatical/acceptable and ungrammatical/unacceptable verbal forms are counted. Then, the data obtained from the trilingual participants are compared with that of the native speakers of Turkish. As the last step, in order to be able to speak about the influence of the background languages of the participants on their L3 Turkish acquisition, the data produced by the Russian–English–Turkish trilinguals in their L3 Turkish are compared with their production in L1 Russian and L2 English.

The results

As has been justified above, the results related to the use of non-finite and finite verbal forms in Turkish of Russian–English–Turkish trilinguals are presented separately. Initially, we discuss the use of non-finite forms in their written narrative production; further, the finite form will be dwelled on.

Non-finite verbal forms in the use of the Russian–English–Turkish trilinguals

The analysis of the data obtained from Russian–English–Turkish trilinguals and Turkish native participants show that both groups make use of finite-main and non-finite subordinate clauses in their narratives (Table 1). The analysis of how Russian–English–Turkish trilinguals use non-finite forms in their Turkish in relation to Turkish baseline data reveal that there are 199 instances when the Russian–English–Turkish trilinguals use non-finite verbal forms in various subordinate clauses and that, by and large, the trilingual participants do not experience difficulties in using non-finite verbal forms. Among the total 199 non-finite forms, only four are identified as ungrammatical. The insignificant percent (2%) of ungrammatical cases observed in the use of non-finite verbal forms in the data of the trilinguals allows us to infer that the participants have acquired the use of non-finite verbal forms in various types of subordinate clauses in their L3 Turkish.

Table 1.

Types of clauses formed by the participants.

Participants	Russian–English–Turkish trilinguals	Turkish native speakers
N of clauses	364	372
N of main clauses	166	182
N of subordinate clauses (in total)	199	190

However, an interesting tendency in the use of non-finite verbal morphology of the Russian–English–Turkish participants is observed when their narrative productions are contrasted with those of the native speakers of Turkish. Along with non-finite verbal forms, five out of eight Russian–English–Turkish trilinguals appear to use finite forms in the subordinate noun clauses, introducing them with the subordinator ki (total 18, which is 18.1 % of the whole subordinate noun clauses as presented in Table 2), while no single use of ki is observed in the Turkish baseline data.

Table 2.

Types of subordinate clauses formed by Russian–English–Turkish trilinguals.

Participants	1	2	3	4	5	6	7	8	Total
N of subordinate clauses with non-finite forms	25	16	49	29	16	20	29	15	199
N of subordinate clauses with finite forms	0	2	6	3	0	3	0	4	18
N of noun clauses (with non-finite forms )	20	9	20	17	8	9	10	6	99
N of noun clauses (introduced with ki)	0	2	6	3	0	3	0	4	18
N of adj.& adv. clauses	5	7	29	12	8	11	19	9	100

Examples below demonstrate Russian–English–Turkish trilinguals’ tendency to use finite forms in subordinate noun clauses in place of non-finite noun clauses. It is noteworthy to state that the use of ki to introduce a noun clause is not observed in the written narratives of the Turkish-native participants.

(8)

Belli oluyor ki [bu cümle onun fikrini değiştirmediyse bile…]. [MC (finite)] [Sub. NC (finite)] Belli ol-uyor ki [bu cümle o-nun fikr-ini Evident be-PRE-3rdSG that this sentence-NOM her-POSS-3rdSG idea-POSS-ACC değiştir-me-di-yse bile…]. change-NEG-PAST-3rdSG-COND even…] It is getting evident that if this sentence has not changed her idea even…

(9)

Çinli kız bağırıyor ki [çocuğunuzun tırnağı var]! [MC (finite)] [Sub. NC (finite)] Çinli kız bağır-ıyor ki [çocuğ-unu-zun tırnağ-ı var]! Chinese girl shout-PRE-3rdSG that child- POSS-GEN nail-POSS-EX PR! The Chinese girl shouts that your child has nails!

Examples 8 and 9 above, which are taken from different participants’ narratives, are characterized as “being atypical” and “sounding strange” by the native Turkish-speaking judges, who report that the use of non-finite forms rather than finite forms in the noun clauses would provide more natural and smoother flow of narration. As stated by Göksel and Kerslake (2005), in Turkish non-finite verbal forms are more acceptable than their finite equivalents introduced by ki, especially in the written production. A plausible explanation for such a tendency in Turkish is related to ki’s being a syntactic pattern borrowed from Persian and having a limited use in the modern standard Turkish (Kornfilt, 1997). It is important to underline that no finite forms in the subordinate clauses are found in the narrative production of any of the native speakers of Turkish.

It is presumably the case that the existence of these structures in Russian and English leads the participants to use these forms in their L3 Turkish. Sentences below demonstrate Russian and English equivalents of the subordinate clauses introduced by ki from the data of the trilingual participants.

(10) English equivalence of Example 8:

It is getting evident that if this sentence has not changed her idea even, it…

(11) Russian equivalence of Example 8:

Становится очевидно, что даже если это предложение не изменило ее мнение, оно…

Stanovitsja očevidno, čto daže esli èto predloženie ne izmenilo ee mnenie, ono…

Both in Russian and English equivalents, finite verbal forms are used in the main sentence followed by a subordinating conjunction and a subordinate clause with a finite verb. Thus, the similarity of the corresponding surface structures between the Russian and English languages does not make it possible to disentangle which of the background languages could be responsible for the replacements of non-finite subordinate clauses with finite ones; therefore, we would like to restrict ourselves here by recording the tendency related to the use of finite verbal forms in the subordinate noun clauses of the trilingual participants, which is not consistent with the Turkish baseline data and which could be related to the influence of their Russian–English interlanguage.

Finite verbal forms in the use of Russian–English–Turkish trilinguals

The analysis of the Russian–English–Turkish trilinguals’ narratives and their evaluation of the native Turkish-speaking judges reveal that the participants use verbal finite markers of number, person, TAM in accordance with the rules of the Turkish language and no single mistake in the formation of the finite verbs was found in their data. Regarding the use of the anchoring TAM markers in the finite verbal forms in their narratives, five participants use the past perfective -DI and three of them rely mainly on the present imperfective -(I)yor to anchor their narratives. In this respect, the trilingual group is similar to the baseline Turkish data of our study and other native speakers of Turkish, who are reported to show no clear preference in the use of the anchoring TAM markers in their narratives (Aksu-Koc, 1994, p. 334). However, the striking peculiarity of the finite form use in the Turkish L3 narratives of the Russian–English–Turkish trilinguals is related to the inconsistency in TAM markers use in finite verbal forms in their production. The absolute majority of the trilingual participants, rather than anchoring one of the TAM markers and sticking to it throughout their narratives, tend to shift from past to present and vice versa. Table 3 presents the number of tense markers used in finite verbs and TAM shifts recorded in the narratives of each trilingual participant.

Table 3.

The use of TAM markers and shifts.

Participants	1	2	3	4	5	6	7	8
Number of present imperfective -(I)yor	40	9	46	13	27	9	27	29
Number of past perfective uses -DI	2	25	16	34	0	23	11	17
Number of shifts	2	6	14	10	0	8	9	8

As it is evident from Table 3, only one trilingual participant is consistent in the TAM marker use in her finite verbal forms, choosing the present imperfective as anchoring for her narrative. This only narrative occurs to be consistent with the production of the native speakers of Turkish, who demonstrate consistency in the TAM markers use while narrating. The performance of the other trilingual participants is marked with numerous shifts (M = 1 shift per five finite forms), which is not found in the narratives of the native speakers of Turkish.

The examples (12 and 13) given below illustrate TAM shifts observed in the trilingual participants’ L3 Turkish narratives.

(12) Kocasıyla zaman geçiriyor (PRE IMPF) ve

“She (Juno) spends time with the husband.”

kocası eşini bırakacağını söyledi (PAST PF).

“He told Juno that he would divorce his wife.”

Juno ağlaya ağlaya dışarıya çıkıyor (PRE IMPF),

“Juno starts to cry and runs out of the house.”

kadınla karşılaşıyor (PRE IMPF),

“She meets the woman.”

o anda kocası ona da gideceğini söyledi (PAST PF).

“At that moment the husband told her that he was going to leave her.”

Kadın şoktayken Juno eve dönüyor (PRE IMPF).

“When the woman was still under the shock, Juno left for her house.”

In Example 12 the participant talks about Juno and states that she spends some time with the husband by stating “Kocasıyla zaman geçir iyor ” using the imperfective present -(I)yor as the anchoring TAM marker (geç-ir-iyor “pass-DRV-PRE IMP.”). In the following clause “kocası eşini bırakacağını söyle di ” she shifts to past perfective -DI form (söyle-di “say-PAST”). In the next clause, the participant moves back to imperfective present -(I)yor and says “Juno ağlaya ağlaya dışarıya çık ıyor ” (çık-ıyor “get out- PRE IMP”). The use of TAM markers in the verbs in this example are all considered to be grammatical by the Turkish native-speaking judges. However, the TAM shifts from one to the other TAM are reported to be unnatural and unacceptable.

(13) Juno’nun hamileliği çok zor değildi (PAST PF).

“Juno’s pregnancy wasn’t very difficult.”

Fakat cok sıkıcı ve moral bozucuydu (PAST PF),

“But it was very boring and annoying.”

çünkü okulda onunla olay ediyorlar (PRE IMPF).

“because everybody is making fun of her.”

Karnı kocaman oldu (PAST PF).

“Her belly was getting very big.”

Birkaç hafta sonra, Juno çifte gidiyor… (PRE IMPF)

“In a few weeks, Juno goes to the couple…”

Example 13 exemplifies another typical TAM shift. The participant anchors the tense with past perfective -DI form in “Juno’nun hamileliği çok zor değil di ” (değil-di “not-PAST”) and then shifts to present imperfective -(I)yor form in “çünkü okulda onunla olay ed iyor lar” (alay ed-iyor-lar “tease-PRE IMPF”). In this example, it is possible to observe constant shifts from one TAM to another in each statement. The Turkish native-speaking judges consider these shifts as unacceptable and additionally state that such shifts are causing difficulties in understanding and, thus, are tiring. These random shifts show that the Russian–English–Turkish trilinguals do not maintain the TAM elements they initially anchor through their narratives in their L3 Turkish.

The standard Turkish and English languages characteristically follow a linear presentation of events and do not allow TAM shifting within narratives (Karakoç, 2007; Kornfilt, 1997). The example taken from the baseline Turkish data illustrates the consistency in the use of the aorist as the anchoring TAM.

(14) Juno bebeği evlatlık vermeye karar verir (AORIST).

“Juno decides to give her baby up for adoption.”

Bunun için arkadaşı ile bir aile arar (AORIST).

“For this reason she tries to find a couple.”

Sonra da durumu ailesine haber verir (AORIST).

“Then she informs her family about the situation.”

Ailesi durumu olgunlukla karşılar (AORIST).

“Her family shows maturity in understanding her situation.”

The Russian language, on the other hand, does not require the tense consistency in the narratives and native speakers of Russian often employ tense shifts in their narratives (Bondarko, 2005; Paducheva, 2011). The examples taken from the Russian classics below are demonstrative:

(15) Повернулся (PAST) и уходит(PRE). Я за ним, кричу (PRE). (Dostoyevsky, 1879)

Povernulsja (PAST) i uhodit (PRE). Ja za nim, kriču (PRE)…

Turned (PAST) and leave (PRE). I behind him, shout (PRE)…

“He turned and was about to leave. I followed him shouting…”

(16) Барон встал (PAST) и говорит (PRE)… (Tolstoy, 1912)

Baron vstal (PAST) i govorit (PRE)…

Baron stood up (PAST) and say (PRE)…

“Baron stood up and said…”

(17) Сплю (PRE) я сегодня ночью и приснилось (PAST) мне… (Rekemchuk, 1962)

Splju (PRE) ja segodnja noč’ju i prisnilos’ (PAST) mne…

Sleep (PRE) I today at night and dreamt (PAST) me…

“Today I was sleeping at night and saw in my dream…”

There might be several reasons justifying tense shifts in the Russian language, such as distinguishing between the topic and the focus and between the foreground and background, underlining the consequence of actions in the narrative, defining two temporal dimensions, and emotional-expressive actualization (Bondarko, 2005; Paducheva, 2011). Since the detailed analysis of the shifts in the narrative structure is beyond the scope of the present study, we would restrict ourselves by just recording the deviation from the native Turkish baseline in the form of TAM shifts in the finite verbs within the same narrative.

It is necessary to say that TAM shifting in narratives is also recorded by several studies that investigate TAM markers in the production of monolingual and bilingual speakers. For example, Aarssen (2001) found out that Turkish–Dutch bilingual children at the age of 4–10 tend to use unmotivated tense shifts in their narrative production. These findings are consistent with Bos (2001), who also recorded the use of tense shifts in the production of monolingual and bilingual children speaking Moroccan Arabic and Dutch. However, these tense shifts are defined as developmental because at a later age and increasing level of proficiency, there is a tendency towards the steady use of one anchoring tense in the narrative. In our case, the shifts of the finite verbal TAM forms recorded in the Turkish L3 written narratives of the Russian–English–Turkish trilinguals can hardly be defined as developmental because the participants are advanced users of the Turkish language living in a Turkish-dominant environment for rather a long period of time. The data of this study demonstrate that the Russian–English–Turkish participants do shift between the TAM markers notwithstanding the high proficiency level of their L3 Turkish. These TAM shifts in the Turkish narratives seem to be fossilized and triggered by the interference from their L1 Russian.

To confirm our presupposition, we examine the narratives written in the Russian and English languages produced by the same participants. The qualitative data analysis of the Russian narratives reveals that the participants use numerous shifts when narrating in their native language Russian from past to present and vice versa, which is a characteristic of the Russian language. Of interest, the analysis of the English data reveals that the native speakers of Russian vary in terms of the anchoring tense for their English narratives: three of them use present tenses and five of them prefer past tenses when narrating; however, none of the participants shifts tenses when narrating in the English language. These observations validate our presupposition that the native speakers of Russian seem to resort to their L1 Russian but not to their L2 English in their use of TAM markers in the finite verbs while narrating in their L3 Turkish. In addition, such language behaviour of the Russian participants raises a question of why most of them transfer the tense shifts from their L1 to their L3 but do not do that when they write in their L2 English. A plausible explanation for such language behaviour of the participants could lie in the social context of the language use. As described above, seven out of eight participants mainly use the English language in the official setting and the Turkish language in the informal one. The formal use of the language implies strict obedience to the rules of the standard language and the monolingual mode of a speaker; the informal use of the language, on the other hand, is known to be more flexible to deviations from the norms of the standard language and is commonly characterized by the mixed mode (Dewaele, 2001). Only one of the Russian participants reports that she uses the Turkish language in the formal setting and she happens to be the only one who follows the rule of the TAM consistency in the Turkish narrative.

To summarize the findings related to the deviations in the use of finite verbal forms in the data of the Russian–English–Turkish trilinguals if compared with native speakers of Turkish, it can be inferred that the trilingual participants are found to be inconsistent in the use of TAM markers in the finite verbal forms within their narratives, and this deviation could to be attributed to the influence from their L1 Russian.

General discussion and conclusion

The purpose of this study is to investigate whether or not the use of verbal morphology in Turkish L3 written production of the Russian–English–Turkish trilinguals differs from that of native speakers of Turkish and, in the case of a recorded deviation, which of the participants’ background languages might account for this. The findings of the study show that at some domains of the Turkish verbal morphology, the trilingual participants reveal language behaviour that is not consistent with the Turkish baseline data. Thus, in the use of the non-finite forms, the multilingual participants are prone to replace non-finite noun clauses with finite clauses with the subordinator ki. Such a tendency is also defined as atypical and unusual for native speakers of Turkish, although it is compatible with norms of the standard Turkish grammar (Göksel & Kerslake, 2005). Relying on the contrastive analysis of the surface structures corresponding to Turkish non-finite noun clauses in the participants’ L1 Russian and L2 English languages, it is suggested that the recorded tendency to substitute non-finite forms with finite forms in the Turkish L3 narratives might be linked to the influence of the Russian–English interlanguage of the participants. However, since the surface structures corresponding to the Turkish non-finite noun clauses are similar syntactically in Russian and English, it does not seem to be possible to disentangle which of the background languages might be responsible for such language behaviour. The data related to the use of finite forms of the Russian–English–Turkish trilinguals, on the other hand, allow us to suggest that the trilingual participants resort to their L1 Russian when using TAM markers in the finite forms in their L3 Turkish narratives, as their performance is marked with numerous TAM shifts, which are commonly used in Russian narratives but are not allowed in the Turkish and English languages. It is important to highlight that the study shows that even proficient non-native speakers of Turkish tend to experience unmotivated TAM shifts. As also stated by the anonymous reviewer, although according to a widespread view of multilingual language acquisition the influence of L1 decreases with the number of additional languages (Bardel & Falk, 2007; Flynn, Claire, & Inna, 2004), the results obtained in this study show that the influence of the L1 is still predominant. These findings suggest that under conditions when independent variables that might account for the cross-linguistic influence in L3 language acquisition, such as closeness of the L1, L2 and L3 and the frequency and recency of L1 and L2 use, are controlled and the participants possess a very high proficiency in their L2, the source of the interference into L3 production is most likely to occur due to the cross-linguistic influence from the native language, which in turn might imply the predominant position of the L1 over sequentially learnt languages at the morphosyntactic level in the language processing.

Abbreviations

ABL Ablative case

AC Adjectival clause

ACC Accusative case

ADC Adverbial clause

CON Converb

COND Conditional

DAT Dative case

DRV Derivation

EVD Evidential

EX PR Existential predicate

FUT Future tense

IMPF Imperfective

INF Infinitive

MC Main clause

NC Noun clause

NEG Negator

NOM Nominalizer

PAST Past tense

PF Perfective

PL Plural

POSS Possessive marker

PRE Present

PROB Probability

PROG Progressive

SG Singular

TAM tense/aspect and/or modality markers

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

Notes

Author biographies

Elena Antonova-Ünlü completed her Ph.D thesis on acquisition of the Russian morphology by a bilingual Russian-Turkish child in a Turkish-dominant environment. Her main research interests include bi-multilingualism, bilingual first language acquisition, second language acquisition, interlanguage and testing.

Çiğdem Sağın-Şimşek is an Associate Professor at Middle East Technical University, Department of Foreign Language Education. Her domains of research include bi-multilingualism, language contact, second language acquisition, third language acquisition and Turkish Linguistics.

References

Aarssen

(2001). Development of temporal relations in narratives by Turkish–Dutch bilingual children. In Verhoeven

Strömqvist

(Eds.), Narrative development in a multilingual context (pp. 209–233). Amsterdam, The Netherlands/Philadelphia, PA: John Benjamins Publishing Company.

Akıncı

(1998). Le development de la temporalite en ture chrz les enfants bilingues ture-francais en France. LINX, 38, 19–34.

Akıncı

(2003). Temporal anchoring of texts in Turkish of first and second generation Turkish immigrants in France. In Ozsoy

Akar

Nkipoglu-Demiralp

Erguvanli-Taylan

Aksu-Koc

(Eds.), Studies in Turkish linguistics. Istanbul, Turkey: Bogazici University Press.

Aksu-Koc

(1994). Development of linguistic forms: Turkish. In Slobin

D. I.

Berman

R. A.

(Eds.), Relating events in narratives: A Crosslinguistic developmental study (pp. 329–388). Hillsdale, NJ: Erlbaum Associates.

Anderson

(1983). Transfer to somewhere. In Gass

Selinker

(Eds.), Language transfer in language learning (pp. 177–201). Rowley, MA: Newbury House.

Bardel

Falk

(2007). The role of the second language in third language acquisition: The case of Germanic syntax. Second Language Research, 23, 459–484.

Berman

R. A.

(1999). Bilingual proficiency/proficient bilingualism. Insights from narrative texts. In Extra

Verhoeven

(Eds.), Bilingualism and migration (pp. 187–208). Berlin, Germany: Mouton de Gruyter.

Bondarko

A. V.

(2005). Teorija morfologicheskih kategorij i aspektologicheskie issledovaniya [The theory of morphological categories and aspectological investigations]. Moscow, Russia: Jazyki Slavjanskih Kul’tur.

Bos

(2001). Temporality issues in Moroccan Arabic and Dutch. . In Verhoeven

Strömqvist

(Eds.), Narrative development in a multilingual context (pp. 233–255). Amsterdam, The Netherlands/Philadelphia, PA: John Benjamins Publishing Company.

10.

Cenoz

(2001). The effect of linguistic distance, L2 status and age on cross-linguistic influence in third language acquisition. In Cenoz

Hufeisen

Jessner

(Eds.), Crosslinguistic influence in third language acquisition: Psycholinguistic perspectives (pp. 8–20). Clevedon, UK: Multilingual Matters.

11.

Comrie

Corbett

(1993). The Slavonic languages. London, UK: Routledge.

12.

Cook

(1995). Multi-competence and the learning of many languages. In Bensoussan

Kreindler

Aogáin

(Eds.), Multilingualism and language learning: language, culture and curriculum (pp. 93–98). Clevedon, UK: Multilingual Matters.

13.

De Angelis

Selinker

(2001). Interlanguage transfer and competing linguistic systems in the multilingual mind. In Cenoz

Hufeisen

Jessner

(Eds.), Cross-linguistic influence in third language acquisition: Psycholinguistic perspectives (pp. 42–58). Clevedon, UK: Multilingual Matters.

14.

De Bray

R. G. A.

(1951). Guide to Slavonic languages. London, UK/New York, NY: Dent, Dutton.

15.

De Bray

R. G. A.

(1980). Guide to the East Slavonic languages. Columbus, OH: Slavica Publishers.

16.

Dewaele

(1998). Lexical inventions: French interlanguage as L2 versus L3. Applied Linguistics, 19, 471–490.

17.

Dewaele

(2001). Activation or inhibition? The interaction of L1, L2 and L3 on the language mode continuum. In Cenoz

Hufeisen

Jessner

(Eds.), Cross-linguistic influence in third language acquisition: Psycholinguistic perspectives (pp. 69–89). Clevedon, UK: Multilingual Matters.

18.

Dostoyevsky

F. M.

(1879). Brat’ja Karamazovy. Retrieved from http://www.magister.msk.ru/library/dostoevs/karamazo.htm.

19.

Flynn

S. L.

Claire

Inna

(2004). The Cumulative Enhancement Model for Language Acquisition. International Journal of Multilingualism, 1, 3–17.

20.

Göksel

Kerslake

(2005). [Third reprint 2010]. Turkish, a comprehensive grammar. London, UK: Routledge.

21.

Grosjean

(1995). A psycholinguistic approach to code-switching: The recognition of guest words by bilinguals. In Milroy

Muysken

(Eds.), One Speaker, Two Languages: Cross-Disciplinary Perspectives on Code-Switching (pp. 259–275). Cambridge: Cambridge University Press.

22.

Grosjean

(2001). The bilingual’s language modes. In Nicol

(Eds.), One mind, two languages: Bilingual language processing (pp. 1–22). Oxford, UK: Blackwell.

23.

Hammarberg

(2001). Roles of L1 and L2 in L3 production and acquisition. In Cenoz

Hufeisen

Jessner

(Eds.), Cross-linguistic influence in third language acquisition: Psycholinguistic perspectives (pp. 21–41). Clevedon, UK: Multilingual Matters.

24.

Jarvis

(2000). Methodological rigor in the study of transfer: Identifying L1 influence on the interlanguage lexicon. Language Learning, 50, 245–309.

25.

Johanson

L. A

(1994). Turkeiturkische Aspectotempora. In Thierroff

Ballweg

(Eds.), Tense system in European languages. Tubingen, Germany: Max Niemeyer.

26.

Karakoç

(2007). The use of finite verb forms in Turkish-German bilingual setting. Conference on code-copying in Turkish as spoken in North-West Europe. Multilingualism in multicultural society.

27.

Kellerman

(1983). Now you see it, now you don’t. In Gass

Selinker

(Eds.), Language transfer in language learning (pp. 112–134). Rowley, MA: Newbury House.

28.

Kellerman

(1995). Crosslinguistic influence: Transfer to nowhere? Annual Review of Applied Linguistics, 15, 125–150.

29.

Kondrasov

N. A.

(1986). Slavjanskie jazyki [Slavonic languages]. 3rd edition. Moscow, Russia: Prosvjaschenie.

30.

Kornfilt

(1997). Turkish. London, UK/New York, NY: Routledge.

31.

Kupersmitt

Berman

(2001). Linguistic features of Spanish-Hebrew children’s narratives. In Verhoeven

Strömqvist

(Eds.), Narrative development in a multilingual context (pp. 277–319). Amsterdam, The Netherlands/Philadelphia, PA: John Benjamins Publishing Company.

32.

Lanza

(2001). Temporality and language contact in narratives by children bilingual in Norwegian and English. In Verhoeven

Strömqvist

(Eds.), Narrative development in a multilingual context (pp. 15–55). Amsterdam, The Netherlands/Philadelphia, PA: John Benjamins Publishing Company.

33.

Mägiste

(1986). Selected issues in second and third language learning. In Vaid

(Ed.), Language processing in bilinguals: Psycholinguistic and neurolinguistic perspectives (pp. 97–122). Hillsdale, NJ: Lawrence Erlbaum Associates.

34.

Möhle

(1989). Multilingual interaction in foreign language production. In Dechert

H. W.

Raupach

(Eds.), Interlingual processes (pp. 179–194). Tubingen, Germany: Gunter Narr.

35.

Odlin

(1989). Language transfer. Cambridge, UK: Cambridge University Press.

36.

Paducheva

(2011). The linguistics of narrative the case of Russian. Moscow: Lambert Academic Publishing.

37.

Pavlenko

(2003) "I feel clumsy speaking Russian"; L2 influence on L1 in narratives of Russian L2 users of English. In Cook

(Ed.), Effects of the second language on the first (pp. 32–61). Clevedon, UK: Multilingual Matters.

38.

Pavlenko

(2008) Narrative analysis in the study of bi- and multilingualism. In Moyer

Wei

(Eds.), The Blackwell guide to research methods in bilingualism (pp. 311–325). Oxford, UK: Blackwell.

39.

Polinsky

(2008). Heritage Language Narratives. In Brinton

D. D.

Kagan

Bauckus

(Eds.), Heritage language education: A new field emerging. New York, NY: Routledge.

40.

Rehbein

Karakoç

(2004). On contact-induced language change of Turkish aspects: Languaging and language practices. Copenhagen Studies in Bilingualism, 36, 125–149.

41.

Rekemchuk

A. E.

(1962). Molodo-zeleno. Retrieved from http://www.rulit.net/books/molodo-zeleno-read-87308-1.html

42.

Ringbom

(1987). The role of first language in foreign language acquisition. Clevedon, UK: Multilingual Matters.

43.

Ringbom

(2001). Lexical transfer in L3 production. In Cenoz

Hufeisen

Jessner

(Eds.), Cross-linguistic influence in third language acquisition: Psycholinguistic perspectives (pp. 59–69). Clevedon, UK: Multilingual Matters.

44.

Tolstoy

A. N.

(1912). Baron. Retrieved from http://ruslit.traumlibrary.net//book/tolstoyan-ss10-02/tolstoyan-ss10-02.html#work001006

45.

Viberg

(2001). Age-related and L2-related features in bilingual narrative development in Sweden. In Verhoeven

Strömqvist

(Eds.), Narrative development in a multilingual context (pp. 15–55). Amsterdam, The Netherlands/Philadelphia, PA: John Benjamins Publishing Company.

46.

Williams

Hammarberg

(1998). Language switches in L3 production: Implications for a polyglot speaking model. Applied Linguistics, 19, 295–333.

The use of verbal morphology in Turkish as a third language: The case of Russian–English–Turkish trilinguals

Abstract

Aims and Objectives:

Design/Methodology/Approach:

Data and Analysis:

Findings/Conclusions:

Originality, and significance/Implications:

Keywords

Introduction

Verbal morphology in the Turkish language

The study

Stimuli and method

Participants

Method of data analysis

The results

Non-finite verbal forms in the use of the Russian–English–Turkish trilinguals

Finite verbal forms in the use of Russian–English–Turkish trilinguals

General discussion and conclusion

Abbreviations

Footnotes

Funding

Notes

Author biographies

References