Research Project in Language Teaching and Learning AN INVESTIGATION OF INDONESIAN STUDENTS’ ABILITY IN PRODUCING THE THIRD PERSON SINGULAR /S/ IN SPEAKING. Fitriah (fitriahasad@yahoo.com) UIN Sunan Ampel Surabaya. ABSTRACT This study examined the ability of Indonesian students in producing the third person /s/ in speaking. Seven respondents were presented with two speaking tasks and questions related to their daily activity and picture descriptions. Overall, results showed that the majority of the respondents were able to produce agreement in speaking, although only five respondents could produce agreement correctly above 30%. Therefore, the study suggested that strategies other than speaking should be examined if they could better facilitate students’ production of the third person singular /s/. Key words: subject-verb agreement, third person singular /s/, speaking INTRODUCTION In the last decade, psycho linguistics has shown strong interest in the production of subject-verb agreement. Several studies have identified the types of errors that most frequently occur in language production. These studies report observations on agreement errors in oral English (Haskel & Macdonald 2005; Hartsuiker & Barkhusyen 2006; Johnson, Villiers & Seymour 2005; Johnson 2005), written French (Hupet, Fayol & Schelstraete 1998; Fayol, Hupet &Largy 1999; Largy & Fayol 2001) and written German (Hemforth & Konieczny 2003). One example of a study which observed agreement errors in oral English was conducted by Haskel and Fitriah 30| IJEE, Vol. 1, No. 1, 2014 Macdonald (2005). They examined agreement with disjunctive subjects which contained plurality nouns (singular-plural noun and plural- singular noun, e.g. have/has the president or the senators read the documents yet?). The research proved that in relation to agreement with disjunctions like „the president or the senators‟, English speakers tend to prefer a verb form that agrees with the nearer of the two nouns (Haskel & Macdonald, 2005). An investigation of working memory affecting the production of agreement errors in speaking was conducted by Hartsuiker & Barkhusyen (2006). To manipulate the availability of working memory, half of the participants had to remember the list of words while performing the primary (load condition) and half of the others performed the primary task without a memory load. All participants were given the speaking span test and had to perform under specific time constraints. The researchers assumed that agreement errors would occur more frequently in the load condition than in the no-load condition. In this study, there were 64 subjects from the University of Nijmegen participating; they were all native speakers of Dutch. In the presentation of the speaking span test in the load condition, the participants were presented with an adjective (e.g. large) that was followed by a sentence fragment (e.g. the cup for the winners). In the no-load condition, the adjective and sentence fragment were presented at the same time. Then, the participants were instructed to repeat and complete each fragment so they had a full sentence, using the adjective (e.g. the cup for the winners was large) before the deadline. In this experiment, the result confirmed the researchers’ hypothesis that agreement errors were more common in the load condition than in the no-load condition, and the errors occurred more frequently when the head noun was mismatched in the An Investigation of Indonesian Students’ Ability in Producing… IJEE, Vol. 1, No. 1, 2014 |31 number with the local noun (e.g. the colour on the canvasses). Though there are many research studies on verb agreement, in my study, a different attempt was made to elicit the respondents’ ability in producing the third person /s/ in speaking. I chose not to use complex subjects (which have plurality of the head noun and the local noun). I preferred obvious subjects, for example, ‘she, he, the girl, the boy’ (from which the number of the subjects is clear). The goal of the study was to examine the ability of Indonesian students to produce third person singular /s/ in speaking. I analyzed only the productive of agreement markers (there is /s/ inflection), such as, she walks, the girl works, he cleans. In conjunction with this stated purpose, the following research questions guide the current study: 1. Do advanced learners produce third person /s/ in speaking? 2. How well do they produce the third person /s/ in speaking? THEORETICAL FRAMEWORK In the traditional description, the rule of English subject-verb agreement is that a singular subject takes a singular verb (e.g. the girl walks home) and a plural subject takes a plural verb (e.g. the girls walk home) (Huddleston & Pullum, 2005). To implement the rule, people require the knowledge of three aspects, (1) how to mark number on verbs and nouns, (2) how to identify the number of a subject and (3) how to identify the subject of a verb (Bock & Miller, 1991). However, there are some nouns that are notionally singular but grammatically plural and take plural marking on the verb (e.g. trousers, binoculars, scissors), nouns which are uninflected for plurality (e.g. people, women, police, cattle), nouns which are notionally plural (though they are uninflected) and take singular Fitriah 32| IJEE, Vol. 1, No. 1, 2014 marking on the verb (e.g. team, committee) (Bock & Miller, 1991). These examples of nouns may lead to confusion in regard to number, specifically whether they belong to singular or plural. Bearing this in mind, it seems that the placement of these nouns in sentences will cause errors in language production as learners may not be able to differentiate between them. Pienemann (1998) pointed out that the implementation of subject-verb agreement procedures will be learned as one block if the subject and verb are not very complex lexically and morphologically in the interlanguage. With regard to this, learners will acquire subject- verb structures easily if the construction of subject and verb is simple and the position of the two remains fixed in the interlanguage. Regarding subject-verb agreement, a series of experiments was carried out using complex subjects with the agreement of the copula (is, are, was, were) (Fayol et al. 1999; largy and Fayol 2001; Hemforth & Konieczny 2003; Haskel & Macdonald 2005; Hartsuiker & Barkhusyen 2006). In all of these experiments, an attempt was made to provoke agreement errors in an experimental setting. The researchers utilised plurality subjects (e.g. the hat of the woman or the gloves of the woman) to find out the probability of errors being made by the learners. In a study involving 17 native French second graders, Fayol et al. (1999) discovered agreement errors presented more frequently when a large number of head nouns and the local nouns were mismatched. In this study, the experimental materials also included subject noun phrases with embedded prepositional phrase, but using varied head nouns (singular and plural, e.g. the gloves of the woman or the wheels of the wagons). The plurality of nouns was used in combination with lexical verbs (e.g. ‘the wheels of the wagons move’ or ‘the father of the children sings’) not with the copula as Hartsuiker’s and Barkhusyen’s study. In the previous study, the An Investigation of Indonesian Students’ Ability in Producing… IJEE, Vol. 1, No. 1, 2014 |33 subjects had to produce the fragment in speaking, but in this study, the participants had to transcribe or write it after hearing several sentences. In both studies it seems that memorizing is the main thing in producing agreement; the more students can memorize the better they can produce subject verb agreement. It would appear that the methodology used in both studies did not really test students’ understanding of subject-verb agreement, but instead, tested students’ memorization. In this case, the learners with high level of memorization would be able to produce subject-verb agreement well. Similar results to those of Fayol et al. (1999) and Hartsuiker and Barkhusyen (2006) have been reported by researchers studying German. A study by Hemfort and Konieczny (2003) investigated the effect of mismatch in noun phrase (NP) and prepositional phrase (PP) constructions, and the experiments were conducted on German constructions. The study involved 62 native speakers of German. In this experiment, the subjects received a booklet with constructions missing and auxiliaries which they had to fill in (e.g. the colour on the canvasses …..dry or the colour on the canvas ……dry). The findings noted that the number of agreements increased when the head noun and the local noun were dissimilar. In this study, the researchers did not provide the copula in sentences (as an agreement marker) as in the previous research, but instead, the participants had to determine which copula suited the subject. In this task, students’ comprehension of subject-verb agreement is essential to complete stimuli questions. If the students are not able to distinguish the number between the head and the local nouns, it seems agreement errors will repeatedly appear. METHODS The subjects for this study were seven Indonesian students who were doing various majors at the Fitriah 34| IJEE, Vol. 1, No. 1, 2014 University of Canberra. They were categorized as advanced learners because their IELTS scores were 6.0 and above. They were aware that the tasks were conducted for research purposes but they did not know the exact focus of the tasks. To collect the data on the production of third person singular /s/, the subjects were assigned two speaking tasks. In the first task, the questions related to participants’ daily activity as in the following examples. 1. What does your husband/wife/da- ughter do every morning? 2. What does your husband/wife do on the weekend? 3. What does your daughter usually do after getting up? 4. How about your husband/wife/frie- nd, does he/she have any special activity? Before starting the main questions, the respondents were given question prompts (e.g. it seems that now you are happy because your husband is here, you can do something together with him. Do you notice ‘what does your husband do every morning?’). The question prompts were different for each participant because some participants lived with their family and others shared a house with friends or lived in university residences. The aim of the question prompts was to create a natural context as well as a comfortable atmosphere. The second task was describing a picture. Through the description of the picture, it was hoped that the participants would be able to generate agreement markers. If the participants could not describe the picture using the present tense, the researcher provided stimuli questions, as in the following examples: 1. What do they wear or what does the boy/the girl wear? 2. Does the man/woman look happy? 3. What does she/he look like? 4. Can you describe the girl’s/boy’s dress? An Investigation of Indonesian Students’ Ability in Producing… IJEE, Vol. 1, No. 1, 2014 |35 The tasks took about three to five minutes, depending on the participants’ responses. If they could produce agreement markers in their answers, the tasks lasted about three minutes. However, if they could not produce the third person /s/, the tasks took longer. All the participants’ responses were recorded using a tape recorder. From seven subjects participating in the study, the researcher included only seven participants in the data collection due to the recording problem. Analysis of the data began by transcribing the participants’ responses in speaking tasks, followed by underlining in red the degree of agreement that the participants should be able to produce. After highlighting the entire agreement /s/ marker, the researcher calculated the correct agreement (if the respondents produced the same agreement, it was counted only once). FINDINGS AND DISCUSSION Table 1. Calculation of the third person /s/ produced by respondents Regarding the first research question whether advanced learners produce the third person /s/ in speaking, the findings revealed that from seven respondents, only five were able to generate the third person /s/ in speaking. Two respondents, participants 2 and 3, could not produce /s/ marker in their responses as indicated in table 1. The table shows that participants 5 was the only one who was able to produce the third /s/ inflection correctly above 50% of the time. Three participants (participants 1, 4 and 6) could generate agreement correctly more than 30% of the time, and Participants Total agreement that the subjects should be able to produce Total correct agreement Percentage 1 6 2 33% 2 7 0 0% 3 2 0 0% 4 5 2 40% 5 7 4 57% 6 8 3 37% 7 10 1 10% Fitriah 36| IJEE, Vol. 1, No. 1, 2014 participants 2 and 3 could not produce a single agreement in speaking. Interestingly, although participant 7 produced more frequent production of the third person /s/ in context, she only got one agreement (10%) correct. Thus, regarding the second question ‘how well do they produce the third person /s/ in speaking?’, it shows that the majority of the respondents produce the third person /s/ in their speaking less than 50% of the time and only one respondent could achieve the production of agreement marker above 50% (participant 5). Compared with other studies which examine agreement errors both in speaking (Hartsuiker & Barkhusyen, 2006) and writing (Fayol et al. 1999), for the most part errors occurred more frequently when the plurality of subjects (the head noun and the local noun) was dissimilar in number. Agreement represents a classic case of syntactic dependency where information that controls the form of one element of a sentence may be separated from it (Bock & Miller, 1991). This is the case when in one sentence there are two subjects (the head noun and the local noun), specifically, if both subjects are mismatched in number. Thus, agreement errors occur more frequently. However, in my study I did not use the plurality of subjects. Nevertheless, most of the respondents could not produce correctly the third person /s/ inflection in their speaking more than 40% of the time. Moreover, two respondents could not even generate a single agreement. Haskell and MacDonald (2003, cited in Hartsuiker & Barkhusyen, 2006) mentioned that the production of verb agreement is sensitive to many different types of information and it requires the simultaneous storage and processing of a relatively large amount of information. In spite of this, all the respondents were identified as advanced learners. Therefore, it was predicted that when they were asked about their daily activity, (for instance, ‘what An Investigation of Indonesian Students’ Ability in Producing… IJEE, Vol. 1, No. 1, 2014 |37 does your husband/wife do every morning, what does your daughter usually do after getting up or what does your husband do on the weekend?’) they would answer using the present tense since the context was clear. Interestingly, some respondents answered using the past tense, past continuous and future to express their daily activity, as in the following example: “My mom she was getting up at 5 o‟clock she was praying subuh she was reading Koran and then she was watching Indonesian program……” It seems that the respondents’ proficiency or spontaneity is not developed enough to answer the questions since they were not able to respond using the same tense as in the questions. Another possible explanation for the results of this study is that the tasks designed were not effective for eliciting the production of agreement markers. It would seem that the tasks did not force the learners to produce the agreement, particularly in the second task ‘describing a picture’, the respondents might answer or respond using other tenses (not the present tense). However, in the previous studies the researchers could elicit agreement errors in their stimuli tasks constantly as they controlled the subjects. Thus, the respondents could not use another tense (should use the present tense). This was the limitation of my study but should provide the basis for future research on subject-verb agreement to design the tasks which are more effective for eliciting the production of agreement marker. CONCLUSION This study examined the ability of Indonesian students in producing the third person /s/ in speaking. The result indicated that the majority of the respondents were able to produce agreement marker in their responses. However, only one respondent was able to generate the agreement correctly above 50% of the time, the rest could produce the agreement Fitriah 38| IJEE, Vol. 1, No. 1, 2014 marker 40% and below. From the results, it would seem that the use of speaking tasks is ineffective for eliciting the production of agreement marker and further research in this area is certainly justified. REFERENCES Bock, K. & Miller, A. 1991. Broken agreement. Cognitive Psychology, (23): 45-93. Fayol, M., Hupet, M. & Largy, P. 1999. The acquisition of subject-verb agreement in written French: from novice to experts’ errors. Reading and Writing: An Interdisplinary Journal, (11): 153-174. Huddleston, R., & Pullum, G.K. 2005. A student‟s introduction to English grammar. Cambridge University Press: United Kingdom. Hartsuiker, R.J., &Barkhuysen, P.N. 2006. Language production and working memory: the case of subject-verb agreement. Language and Cognitive Processes, (21): 181- 204. Haskell, T.R., & Macdonald, M.C. 2005. Constituent structure and linear orden in language production: evidence from subject-verb agreement. Journal of Experimental Psychology, (31): 891-904. Hemforth, B. & Koenieczny, L. 2003. Proximity in agreement errors. Proceeding of the Conference of the Cognitive Science Society in Boston, July/August. Hupet, M., Fayol, M. & Schelstraete, M. 1998. Effect of semantic variables on subject verb agreement processes in writing. British Journal of Psychology, (89): 59-75. Johnson, V.E. 2005. Comprehension of third person singular /s/ in AAE-speaking English. American Speech-Language- Hearing Association (36), 116- 124. Johnson, V.E., Villiers, J.G., & Seymour, H.N. 2005. Agreement without understanding? The case of third person singular /s/. Sage Journal Online and Highwires Press Platforms, (25): 317-330. Largy, P., & Fayol, M. 2001. Oral cues improve subject-verb agreement in written French. International Journal of Psychology, 36 (2): 121-132. An Investigation of Indonesian Students’ Ability in Producing… IJEE, Vol. 1, No. 1, 2014 |39 Pienemann, M. 1998. Language processing and second language development: processability theory. ANU. Fitriah 40| IJEE, Vol. 1, No. 1, 2014