Second Language Acquisition of Syntactic Movement in English by Turkish Adult Learners

There has been much discussion on the involvement of Universal Grammar (UG) in Second Language Acquisition (SLA) process. Despite growing research in the field, few precise answers to the problem have been suggested so far. Hence, recent L2 studies within the generative framework have shifted from investigating this issue to determining whether or not interlanguage grammars exhibit natural language characteristics (Can, Kilimci & Altunkol, 2007). The present study aimed to investigate L2 acquisition of syntactic movement in English noun clauses by Turkish adult learners. Accordingly, L1 involvement in SLA was sought through examining the upper intermediate Turkish learners’ knowledge about the movement in question. The study addressed the questions of whether or not Turkish adult ESL learners have problems, stemming from L1 interference, with the construction of the syntactic movement in English noun clauses, and whether or not there is any order of acquisition between the noun clauses in subject position and object position along with various wh-words. The study reported related findings, and concluded with a few pedagogical implications for practice, and a couple of suggestions for further directions.


Introduction
Several theories and approaches have been proposed up till now to account for the intriguing process of language acquisition.Chomsky's Linguistic Theory has been the most prevalent among them, particularly over the last few decades.Its major claim is that human beings are born with a capacity to acquire any natural languages they are exposed to, so long as normal circumstances are provided; thus, every individual has the opportunity to attain linguistic competence in a language on condition that s/he doesn't suffer a biological abnormality.Weissenborn and Höhle (1984) suggest such biological equipment enable human beings to produce unlimited number of novel meaningful utterances by means of relatively few linguistic rules.Constituting the core of Chomsky's theory, the Universal Grammar (UG) is a set of principles which are abstract values, common to all natural languages and parameters which are set on the basis of linguistic input and vary across languages (White, 1989;Canli & Canli, 2013;Rathert, 2012).The reason why it remains so pertinent is the fact that no grammars appear to violate its principles.Namely, over 6,000 languages are acquired in the same way, regardless of the gender or race of the acquirers, as illustrated in Figure 1.(Cook & Newson, 1996) Linguistic Input

Language
The present study sought an answer to the question of whether or not UG operates in adult second language acquisition in the same way that it does in first language acquisition.Accordingly, it investigated whether or not Turkish-speaking adult learners of English, whose first language displays no evidence of overt syntactic movement in the construction of noun clauses, have difficulty in recognizing such kind of operation which is very much welcomed within the target language.The following statement is intended to illustrate the difference between Turkish and English with respect to the requirements on syntactic movement in the formation of clauses in concern.
As wh-questions do not change in indirect forms, no overt grammatical operation is involved in the formation of Turkish noun clauses (Can et al., 2007).

Affirmative form
O yürüdü.
S/he walked.
Where did s/he walk?
a noun clause S/he asked where s/he walked.
Based on examples Can et al. ( 2007) The process of noun clause construction in English is described below.Subjacency is the principle of UG which limits the amount of movement that can take place within sentences (Gass & Selinker, 2008;Yılmaz, 2013).Can et al. (2007), and Kartal and Kocabas (2014) suggest that it is not operative in all languages and that those in which it is activated, the movement of wh-phrases is restricted into one bounding node at a time.In this regard, English falls into the language category which requires activation of the principle in question, while Turkish holds no activation of such kind.Besides, the latter does not permit any syntactic movement, neither in the construction of wh-questions nor in that of noun clauses.That is to say, Wh-words remain in their non-interrogative positions (Aygen-Tosun, 1999), as illustrated in the following statements adapted from Uzun (2000).-POSS.last week car buy-PAST-3SG -ACC.tell-PAST-3SG.
Within a sentence in Turkish, any constituent to emphasize is positioned well ahead of the main verb.Nonetheless, operation of this kind is not frequent.This holds true for noun clauses that appear in both object and subject positions within sentences, as illustrated below.

PROG-3SG
Ayşe does not remember whom she gave the book.
All in all, Turkish and English exhibit different behaviors with respect to the construction of noun clauses.Hence, Turkish-speaking adult learners of English are expected to have some difficulties in handling the formation of noun clauses in English.Considering this, the present study addressed two research questions:  Do Turkish adult learners have trouble acquiring English syntactic movement in the construction of noun clauses? Is there any order of acquisition for Turkish adult learners of English in wh-words involved in syntactic movement?
The following section offers findings of studies previously conducted on the second language acquisition of syntactic movement in English noun clauses by speakers of different language backgrounds.In a study conducted on the second language acquisition of whoperator movement in English, Hawkins, and Chan (1997) reported that Chinese-speaking learners of English whose L1 does not allow wh-operator movement in overt syntax successfully establish mental representations for English which involve pronominal binding rather than operator movement.A decade later, Hu and Liu (2007) examined the second language acquisition of restrictive relative clauses in Chinese by English-speaking and Korean-speaking learners.They reported that English-speaking learners distinguished between target-like restrictive relative clauses and non-target-like restrictive relative clauses earlier than the Korean-speaking learners.So, they concluded that the superficial similarity between Korean and Chinese regarding head direction leads learners not to restructure quickly, while the surface dissimilarity of English and Chinese gives rise to rapid restructuring in L2 grammars of learners.In a different study, Can et al. (2008) explored second language acquisition of syntactic movement in English noun clauses with the participation of intermediate and advanced learners of English with Turkish L1, and reported that the advanced group was significantly more competent in handling syntactic movement operations, but still had problems with noun clauses with an auxiliary verb.Another finding of their study is that both groups of non-native students used noun clauses in object position more successfully than those in subject position.Finally, the results of their corpora analysis indicated that native speakers of English also tend to use noun clauses in object position rather than subject position.Choi (2009) scrutinized the interpretation of wh-in-situ expressions in L2 Korean by adult native speakers of English with high-intermediate and advanced proficiency levels in their L2 where wh-in-situ words can receive multiple readings, and found that both groups performed statistically better on question reading than indefinite reading in both prosodic and morphological licensing environments.They also found that several advanced learners overcame incorrect question interpretation within the course of L2 development while nontarget-like interpretations persisted in their L2 productions.Cele and Gurel (2011) examined processing of wh-dependencies in L2 English by Turkishspeaking and Spanish-speaking advanced learners of English through an online grammaticality judgment task including grammatical and ungrammatical wh-extractions.They found a statistically significant difference between the native group and non-native groups in terms of accuracy, and that subject extraction from nonfinite clauses constituted the major problem for all three groups.They also revealed that the non-native groups did not significantly differ in the extraction, and that all groups had a subject-object asymmetry in wh-extraction from nonfinite clauses (Juffs & Harrington, 1995).
Lastly, they noted that the Spanish-speaking group displayed a lower performance than the Turkish-speaking group in this context, which they attributed to the obligatory complementizers in L1 Spanish wh-questions.In a recent study, Prentza (2012) investigated whether or not parameter resetting is possible in ESL acquisition of Greek adult learners on the assumption that the two languages exhibit parametrically different choices as regards the formation of restrictive relative clauses which are associated with abstract syntactic features.His findings obtained from a grammaticality judgement task administered to both native and non-native speakers of English have indicated that the former performed significantly better than the latter, and the researcher concluded that the advanced Greek learners tend to fail to acquire the feature specification of the English relative clauses.
Keeping the above-mentioned findings in mind, the present study is intended to reveal the second language acquisition of syntactic movement in English by Turkish-speaking adult learners of English, partially replicating the study by Can et al. (2008).The following section offers information about the participants and data collection procedure of the study.

Methodology
Being descriptive in design, the current study involved the administration of three tasks to a group of 17 Turkish-speaking adult learners with upper-intermediate level of English who were the participants of the present study.All of the participants were born and raised in Turkey, and at the time of testing, they were undergraduate students attending the English Language Teaching Department at Çukurova University, Turkey.Their ages range from 20 to 22, with a mean of 20.5.They were selected from those who had never lived in a country where English is the spoken language.All except three are Turkish monolinguals.Two subjects are bilingual in Arabic and Turkish, and one in Zaza and Turkish.The data were elicited through Grammaticality Judgment Task (GJT), Scrambled Sentence Task (SST) and Elicited Imitation Task (EIT) administered to the students.Tremblay (2005, p. 159) argues that the use of GJ tasks is essential in linguistic theory because it can provide crucial information about grammatical competence that elicited production tasks and naturalistic data collection cannot offer.She is of the opinion these tasks can provide empirical evidence, if designed and administered carefully, that may serve in the formulation, support, and refinement of theoretical claims in the study of language.Prior to the administration of the tasks, a vocabulary session was conducted to ensure that no items were unfamiliar to the participants.In the main test phase, they were asked to respond to a total of 54 items (18 items in each test).
GJT required the participants to decide whether or not the items were grammatically correct in English.Distribution of the items across the test in terms of their positions within the statements are displayed in Table 1.In order to clarify the task, a sample was provided along with the following simple instructions: You are supposed to look at each sentence and then decide whether it is "Correct" or "Incorrect" in English."Correct" means that you think a native speaker of English might produce the sentence on some occasions, and "Incorrect" means that you think s/he would never produce it.Here is an example: We did not see who she was talking to.
() Correct ( ) Incorrect SST (Wenzlaff, 1988(Wenzlaff, , 1993) ) was originally used to measure participants' tendency to interpret ambiguous information in positive or negative ways (Rude, Wenzlaff, Gibbs, Vane, & Whitney, 2002).In our study, it was used to elicit information about how English is represented as a foreign language in the minds of Turkish-speaking students.During the task, the participants were asked to unscramble the sentences in the target language, and to write grammatical sentences underneath.Table 2 provides the number and positions of whphrases within sentences.Instructions and an example for this task are as follows.
The following are scrambled sentences.You are supposed to unscramble them and write the grammatical sentences underneath.Here is an example: car / do / many /you / there/ in/people/ how / the / know / were Do you know how many people there were in the car?
Larsen-Freeman and Long (1991), and Kalayci & Humiston (2015), and Kalayci (2012) describe EIT as a procedure whereby the researcher reads to the participant a particular set of sentences containing examples of the structure under study.According to them, the procedure is based on the assumption that if the sentence is long enough, a participant's short-term memory will be taxed and consequently the participant will be unable to repeat the sentence by rote.In such a case, they advocate that participants will have to understand the sentence, and to reconstruct it using his or her own grammar.From this viewpoint, in the present study, the subjects were simply expected to repeat the sentences read to them.The items were prepared as slightly challenging in order to assess their exact performance in the construction of noun clauses in English.Namely, sentences including a number of morphemes, which is considered to exceed the limits of short term memory, were employed so that the participants would produce their own utterances on the basis of what they have heard.The following items are taken from the task in concern.
How he was going to pass the driving test developed a great interest among the trainers.(Item 10); He asked her which booklet she didn't want to bring to the meeting.(Item 2); We didn't even know who the mayor of Istanbul was.(Item 9) Each item was read to the participants only once and at a normal pace.During the task, they were filmed and never interrupted.Finally, their answers were transcribed prior to data analysis.Positions of wh-phrases within sentences utilized in the task are shown in Table 3. Subsequently, data obtained through the above-mentioned three tasks were analyzed in order to see how the knowledge of syntactic movement in English noun clauses is represented in the mind of Turkish-speaking adult learners of English.Findings of the study and related discussion on them are provided in the following section.

Findings
Preliminary findings of the study obtained from three sets of tasks have indicated that the students' performances vary significantly on the type of tasks they were assigned.Figure 2 depicts the related distribution.

Figure 2. Rate of success across tasks
As illustrated in Figure 3, GJT items were responded correctly more frequently than those in SST and EIT (GJT: 86.93%, SST: 78.76% and EIT: 75.16%).Indeed, the result in question was as anticipated, particularly when considering the fact that GJT would not involve the students making as much linguistic production as SST or EIT would.In other words, the subjects performed relatively better in cases where they were exclusively required to comprehend the items, and to judge upon their grammaticality than the ones which required them to comprehend what was produced in the target language, and to make some kind of production using their linguistic knowledge in L2.Namely, they were asked to repeat items with noun clauses in either subject or object positions read aloud to them in EIT, and to unscramble sentences and to formulate grammatically correct ones in SST.So, it could be convenient to say that the more productive skills the task required, the more the participants' performance decreased with regard to the syntactic movement in noun clauses.Their performances in the three tasks also varied depending on whether or not the items include an auxiliary, as shown in Figure 3.As indicated in Figure 3, the participants performed roughly equally in responding to the GJT items including and excluding auxiliary.It is noteworthy that all of the participants provided correct responses to the following three GJT items.
Item 2.She does not know where he works.Item 4. *She could not remember which task did she assign to her students.Item 6.She did not tell me why she was so angry.
Not surprisingly, two of the above-illustrated items contained no auxiliary verb (Items 2 and 6), which is in line with the finding Can et al. (2007) obtained in a similar study.They suggest that the subjects' relatively better performance on such items might stem from the fact that items without an auxiliary do not require syntactic movement operation.In return, the following are the ones mostly responded incorrectly by the participants.Namely, grammatically correct items tended to be judged as ungrammatical (Items 11 and 16), and grammatically incorrect ones as grammatical (Items 8 and 10).Likewise, SST results revealed that the items including auxiliary were responded correctly more frequently than those that did not contain it.Four items the students mostly failed to unscramble are illustrated as follows.

Whose books the class was going to read was important. (Expected Response)
It is remarkable that only one of the items responded incorrectly by most of the participants included a noun clause with an auxiliary verb (Item 13).Thus, it could be concluded that the students tended to perform relatively worse when they were required to unscramble the statements containing no auxiliary, indicating that auxiliary inclusion in a noun clause made it easier for them to handle the task as the structures of such kind entailed syntactic movement.
In EIT, they were simply asked to repeat the sentences read aloud to them at a normal pace.As stated earlier, no items were reread, and the students were never interrupted while responding.During the task, they were filmed, and their responses were transcribed for data analysis.Interestingly, one of the items was not responded correctly to by anyone in the group (Item 1. Whom she is going to dance with is important).In return, the following items were responded to correctly by all the participants.Item 3. The lady in the car wants to know where the restaurant is.Item 8.Where they were going to study was a wonder to everyone.

Item 13. The child on the playground wanted to know what the time was.
The reason why the subjects proved relatively less successful in EIT than GJT might be attributed to the fact that the latter provided them with the opportunity to think about the grammaticality of the items -spending at least a few seconds on each item -while the former to repeat what was read to them, with less thinking about the items.
All in all, contrary to GJT and SST, the items containing auxiliary were responded correctly less frequently in EIT.Nevertheless, it is impossible to come up with a conclusion in terms of auxiliary inclusion in the noun clauses similar to the one previously made for considering the results obtained from the former two tasks.
As for the positions of the noun clauses within the statements, the students showed better performance on the items in which wh-clause appears in object position than those in subject position.In addition, the items in which noun clauses appeared in subject position responded correctly more often than those occurring in object position.Figure 4 illustrates the distribution of correct responses regarding the position of noun clauses within sentences.

Figure 4. Success rate regarding the position of noun clauses
As depicted in Figure 4, there is almost no significant difference between the students' performance in GJT with respect to the position of the noun clause within sentences.Namely, they successfully responded to approximately slightly over 88% and approximately 87% respectively, of items containing noun clause in subject position, and object position.They displayed a relatively lower performance in EIT.Nevertheless, there does not seem to be a significant difference between their performance in repeating the items including noun clauses in object position and those including them in subject position.The most significant difference in this concern was found between their performances in responding to the SST items.That is, approximately 93% of the items including a noun clause in object position were responded to correctly by the participants, whereas less than 65% of those containing such clauses in subject position were responded to correctly.This particular finding might be attributed to the fact that such sort of structures are not so common in the students' L1.In other words, noun phrases tend to appear in the object position in Turkish sentences, rather than the subject position.Another finding of the study, was that two GJT items were responded to incorrectly by most of the students despite the fact that they contain no auxiliary in the noun clauses appearing in object position within the statements, as illustrated below.
Item 8.The manager asked him who he worked for.
Item 11. *She does not know who is the president of the Ivory Coast.
The exemplified cases in concern might stem from the fact that the subjects encountered wh-word "who", which typically appears in the subject position within interrogative statements, and in the object position where it was included in the noun clauses.Consequently, it is impossible for us to make a generalization with regard to the position of noun clauses within the sentences as the participants faced difficulty in responding to the items both with and without auxiliary.However, it is evident that most of the items which were frequently reproduced incorrectly included auxiliary.Lastly, the study present study scrutinized whether or not there is an order of acquisition for Turkish adult learners of English in wh-words involved in syntactic movement.Figure 5 illustrates the distribution of correct responses provided to the wh-words which appear in noun clauses in three sets of tasks.

Figure 5. Correct responses with respect to wh-words in noun clauses
As indicated in Figure 5, participants had trouble in responding the items which include wh-words who, when and how across all three tasks.In GJT, where they were supposed to decide whether items are grammatical or not, they provided most incorrect responses to those in which who appears in noun clauses (23.53%).Items including when in noun clauses revealed the ones second most incorrectly responded in GJT.By the same token, items with who and when in noun clauses were repeated most successfully in EIT (who: 58.83% and when: 29.41%).Among SST items which required the subjects to formulate grammatically correct sentences out of scrambled ones, those including when were provided no correct responses.Interestingly, the participants were able to successfully respond to 41.18% of the items including who in noun clauses, and most of them failed to respond the items including how in noun clauses (5.88%).

Conclusion
Turkish and English exhibit different behaviors with respect to syntactic movement in the formation of noun clauses.Namely, the former does not involve such an operation which is very widespread in the latter.Considering this, Turkish adult learners of ESL were expected to have some difficulty in the acquisition of syntactic movement which is central to the construction of English noun clauses.To investigate how they handle such kind of difficulty, which is considered to stem from the distinction between the languages at stake, two research questions were formulated.The first question investigated whether or not Turkish adult learners have trouble acquiring English syntactic movement in the construction of noun clauses.Results of the study have revealed that they are distinctly successful in recognition of the movement in concern particularly in cases where noun clauses include an auxiliary (85.86%), and that they encounter some sort of difficulty in cases whereby noun clauses do not contain an auxiliary (74.83%).Likewise, their performance in responding to the items largely differs with respect to the position of noun clauses within sentences.Namely, they were able to respond to the items in which noun clauses appeared in object position more efficiently than to those with noun clauses in subject position (Object: 83.6%, Subject: 77.06%).As a consequence, it would be inconvenient to say there is an order of acquisition for Turkish adult learners of English in wh-words involved in syntactic movement.Nonetheless, it is clear that they have some sort of disposition in recognizing certain structures in noun clauses.In our opinion, these should be carefully examined and language learning problems should be properly handled through consciousness-raising activities, as exemplified below.Suggestion 1. Change the question in the parenthesis to a noun clause.
(How old is he?)I do not know how old he is Suggestion 2. Make a question from the given sentence.The words in parenthesis should be the answer to the question you make.Use a question word (who what how when why etc.)Then, change the question into a noun clause.

Bea will come to Durham (tomorrow)
Question: when will Bea come to Durham?
Noun Clause: I want to know when Bea will come to Durham Suggestion 3. Error Correction (spoken or written): The teacher either reads the ungrammatical sentences to the learners and asks them to find the error and correct it, or hands out the following worksheet and asks them to write the correct form next to each sentence -in accordance with the proficiency level of students.

Correct the error.
Lady in the car wants to know where is the restaurant.
The present study is confined to the analysis of data obtained from three tasks administered to a limited number of students attending a state university in Turkey, who had upper-intermediate level of English proficiency at the time of the study.So, it would be beneficial to investigate second language acquisition of syntactic movement with a larger number of participants of various L1 backgrounds and with various proficiency levels in English.It might also be extended to students attending different levels of education which offer English (or another given language which displays different behaviors from their L1 in terms of syntactic movement) as a compulsory part of the related curricula.Finally, a further study might be conducted with the participation of non-native speaking child, adolescent or adult learners of English living in a country where English is the spoken language in order to see whether or not they significantly differ from their counterparts living in countries where English is spoken as a foreign language.

Notes
Corresponding author: REYHAN AGCAM The study in concern was orally presented with the title of 'We Like to Move It: L2 Acquisition of English Syntactic Movement in Noun Clauses by Turkish Adult Learners' at the 24th International Conference on Second Language Acquisition and Foreign Language Learning in Szczyrk, Poland.

Figure 3 .
Figure 3. Success rate with respect to auxiliary inclusion Item 11. *She does not know [who is the president of the Ivory Coast].Item 8.The manager asked him [who he worked for].Item 10.May is [when she takes her last examination].Item 16. *The professor on TV talked about [how important was the education].

Table 1 .
Distribution of wh-phrases in GJT Items

Table 2 .
Distribution of wh-phrases in SST items

Table 3 .
Distribution of wh-phrases in EIT Items