THE PLIGHT OF THE LESS PROFICIENT SECOND LANGUAGE LEARNER: BETWEEN DUCKS AND PARROTS

This article investigates the English language ability of a cross section of Malaysian second language learners. Using cross sectional data taken from a cartoon completion task given to one hundred and twenty respondents, the study investigates learner ability in terms of language knowledge and language use of three groups namely, initial, intermediate and finite stage. The goal of the project was to determine: a) how L2 learners use the past tense forms and pronouns; b) how second language (L2) learners use the knowledge acquired and c) how this knowledge is put to use in sentence constructions. The findings revealed that L2 learners at all levels continue to the exhibit performance errors which include: a) identical nominal gender in L1 overriding the different gender cues; b) some principles available only at the later stages of the process, and c) memory restrictions. Given the limited language proficiency of many of the subjects it was difficult to establish whether these errors were due to competing or shared representations or false assumptions that learning has taken place based on restricted insights into learners word use in classrooms.


Introduction
"Globalization has brought about new changes and challenges to language teaching in terms of how, where, what and more importantly whom we teach" (Chong & Gopinathan, 2008). In other words global economic opportunities and the need to keep up with 21 st . century skills have raised the academic bar for educationist committed to raising teaching standards and learning goals. Taken together, the field of second language acquisition (SLA) has often noted from the first language (L1) to second language (Odlin, 1989;Barto-Sisamout, Nicol and Witzel, 2009) (either facilitation or interference) to occur at a number of levels. In fact, SLA studies have repeated addressed the occurrence of this phenomenon at various levels such as the L2/FL learning process (Dulay & Burt, 1974;Dulay, Burt & Krashen, 1982), L2 phonology (Fledge & Davidian, 1984, Eckman,1981, Lexical (Jiang, 2002, 2004, Hall & Ecke, 2003, syntactic (Helms-Park, 2001, Gass, 1980White, 1908), discourse/pragmatic ( Yoon & Kellog, 2002;Yu,2004) and reading skills (Brown & Haynes, 1985;Koda, 1990). Though the evidence for the level of transfer appears mixed, one subcomponent of sentence processing -the processing of grammatical morphology has been stated to be particularly sensitive to L1 influence (Jian, 2004;2007, Sisamout et.al., 2009. The present study seeks to investigate the L1 to L2 transfer in the processing of L2 grammatical morphology from within the Malaysian context.

Malaysian L2 learners' processing preference
Given the emphasis paid to raising standards of English taught in classrooms and higher institutions (Zuraidah Don, 2008) and the growing concern that Malaysian L2 learners despite learning English as a second language in schools for eleven years continue to struggle to speak and write beyond the literalness of textbook English (Isarji et. al., 2008), a number of researchers have raised concerns that L2 learners' performance gaps may have more to do with L1 interference. Marlyna et al. (2007) systematically indicates that a number of Malay learners' sentence processing errors to be direct transference from their L1 and calls for more attention to purposeful teaching materials. Elsewhere, Nor Hashimah, et al. (2008) views most Malay students' errors to be influenced by differences in grammatical structures of both Malay and English and suggests that rather than "… focus on the pedagogy or teaching of the target language " … English teachers should be exposed to linguistic knowledge to better equip them in teaching the language" (p. 116). Yet, despite these growing concerns and given the existing L2 environment in Malaysia where most learners come to the L2 classroom with full knowledge in at least one or two mother tongues/ L1, there surprisingly little evidence of classroom related investigation on the sentence processing ability of Malaysian learners. To take for instance, the L2 sentence involving agreement of verbs in English. Existing research continue to indicate that rural Malay students have problems with verbs since there is ".. no such thing as a subject verb agreement rule that requires inflections based on the number of subjects" for Malay (Marylyna et. al., 2007;Nik Safiah Karim et al., 2004). Nevertheless, studies that have looked at Malay learners' word use and grammar errors have found some weak but inconsistent indications of L1 interference. Therefore, it is not clear if all rural Malay learners will continue to demonstrate this state of inconsistencies even when they enter higher learning institutions or if some errors become more prominent compared to others. The study of Nadarajan and Amalina (2008) found proficient Malay learners at the end of the school going years to continue omitting and over generalizing the -s when attempting to agree with the number or status of the subject (p.63). Similarly, the study of Nor Hashimah et al. (2008) noted the absence of "affixes that indicate or produce negative meanings (e.g. im-, dis-, mal-) and b) absence of plural inflectional forms in the L1 (Malay) to be affecting L2 learners' ability to perform in the cloze test and suggested that that "… comparative analysis of Malay and English may be one mode of understanding and solving persistent problems of acquiring the target language" (p. 116). Nadarajan and Amalina (2008) however, caution against relying on contrastive analysis given the variations between existing dialectal variations and differences between East and West Malaysian Malay languages. The writers appear more concerns over the detrimental effect of stereotyping L2 learners according to language groups and dichotomies especially within a multilingual setting like Malaysia where teachers might overlook the individual within the learner. Meanwhile, Gan (2012) posits that some L2 learners do manage to learn difficult L2 grammar features with more innovative teaching strategies rather than rely on transfers. This study, therefore, attempts to look at Malay L2 learners sentence processing difficulties in terms of cues that help promote and hinder the L2 acquisition process.

The competition model
According to MacWhinney's Competition Model (1997, 2002, various cues to sentence interpretation including word order, agreement and case marking differ in terms of their

THE PLIGHT OF THE LESS PROFICIENT SECOND LANGUAGE LEARNER: BETWEEN DUCKS AND PARROTS
importance/reliability across languages, and the relative weightings of these cues in the L1 are taken over to the processing of the L2. Taken together with Marlyna et al. (2010) which highlighted at least three grammar errors in the L2 that are viewed as caused by a direct transfer from the L1 syntax (e.g. determiners, pronouns, tense markers and plural forms), it is increasingly apparent that some cues in the L1 tend to be more salient compared to others. Currently, what is known within SLA about transfer is that English speakers 'weigh' word order as the primary cues to sentence interpretation, whereas German, a language with freer word order tend to give subject verb agreement more weight (e.g. Bates & MacWhinney, 1981).Research in eye movement patterns of English L1 speakers and L2 learners have attested to these interference patterns (French-Mestre & Pynte (1997). However, comparable findings (e.g. Dussias & Cramer (2006), Jiang (2007 and Juffs (1998)) suggest that bilinguals (at least at higher levels of proficiency) tend to converge on native like processing of L2 verbs. Most of these works have been on European languages but the works of Hahne (2001), Jiang (2004) looked at English speakers and highly proficient Chinese learners of English reading ability on four sentences in a self -paced reading task; 1a. The key to the cabinet was rusty from many years of disuse. 1b. The key to the cabinets was rusty from many years of disuse. 2a. The bridge to the island was about ten miles away. 2b. * the bridges to the island was about ten miles away.
Jiang's work revealed that the native speakers were sensitive to the plural -s morpheme but showed processing difficulty on sentence (1b) relative to (1a) as well on the ungrammatical sentence (2b) relative to the grammatical (2a) and immediately after the copula (was). The result for sentence type (1) has been outlined as a "broken agreement effect" found to exist in production (Bock & Cutting, 1992) and comprehension (Nicol, Forster & Veres, 1997) studies, whereby a plural noun (cabinets) interferes with agreement processing between a singular head noun (key) and the subsequent agreement 'was'. The result for sentence type (2), can be attributed to the detection of ungrammaticality when there is a mismatch between plurality based agreement relationships (e.g. bridges and was). Meanwhile, the Chinese learners were unable to integrate plurality and, specifically, the plural morpheme -s automatically during the comprehension process suggesting that there are bound to be interference when there are differences or similar/near similar relationships between a particular component of the L1 and L2 even for proficient learners. Given the findings, it can be argued that the grammar errors made by Malay learners may not be peculiar to less proficient learners alone but also to a wider population and this needs to be investigated for raising standards of L2 teaching and learning.

Comparison of Malay and English
Transfer of Malay phonology, lexicon and syntax into the L2 is to be expected given that there are significant differences between Malay and English irrespective of the environment in which they are being used. In terms of morphology, English is seen as an impoverished language, marked by a strict word order which is used to indicate subject-object or agent patient relationship. Compared to English, Malay is seen as having a more impoverished morphological system without case distinctions or verb agreement. Nevertheless, Malay speakers rely on word order cues just as much as English native speakers do. As explained by Marylyna (2008), where possessive determiners are concerned, the expression my ball would be expressed as 'bola saya', with 'saya' taking on the similar form as the nominative case of 'I'. In addition, Malay is not marked by grammatical gender, with a neutral word 'dia' being used to replace he, she, his and her. Verbs are also not inflected for person or numerals but there are a wide range of prefixes and tenses denoted by time adverbs. Malay lacks a grammatical subjects in the sense English does and in intransitive clauses, the noun comes before the verb and when there are both an agent and object, these are separated by a verb. It is possible to assume at this stage that given the greater variation in word order, the validity of word order is likely to be lower for Malay. Nevertheless, Marylyna Maros et al. (2007) highlighted three persistent problems with Malay speakers'L2 grammar which requires further investigation. These problems have been examined via numerous settings where learners continue to demonstrate incorrect or omit determiners, the third person singular -s, subject verb agreement, and copula 'be' in a number of school related studies. A likely explanation for this phenomenon being that while determiners in English are restricted to the numbers or countability of the head nouns with which they co-occur, the Malay equivalent which are the words 'itu' (that) and 'ini (this) only occur as final elements in Malay noun phrases, therefore making it necessary for any modifier that comes after the head noun to appear between the head noun (on the left) and the kata penentu (on the right) (Abdullah Hassan, 1993 p.54). Similarly, Malay grammar does not recognize the English subject verb agreement rule that requires inflections to morph based on the number of subjects. This in turn can result in the third person singular being omitted or inaccurately used. In terms of Copula 'be' it is also assumed that learners tend to over generalize the copula 'be' due to an existing equivalent meaning in the L1 -'ialah' and 'adalah'-which is said to play a comparable function in both the L1 and the L2. So, the leading assumption being that in situations where the form is not required in the L1, the L2 learners are more likely to omit the copula 'be' altogether. The present study seeks to examine whether Malay learners continue to commit these errors as they enter higher learning institutions and the workforce.

The experiment
Crucially, this study intends to demonstrate that L2 learners' sentence constructions would resemble native English sentences as learners gain in L2 proficiency. Therefore, we predicted that L2 learners as they advanced should be able to construct simple grammatically accurate sentences especially in terms of determiners, subject verb agreements and the copula 'be'. We used evidence from the findings from SLA research and Marylyna Maros (2012) to investigate and support our assumptions.
The following sentence types were constructed to assess the targeted language errors.

Omission of determiners
To obtain a better understanding of the actual learning problems of the learners, it was first necessary to detail the relevant structural contrast between Malay and English and our prediction for each sentence type. Three types of sentences were selected in order to test L1-L2 transfer between groups under the similar but different relationships. For the first sentence type, the Malay determiner appear as the final element in Malay noun phrases and therefore the Determiner will not be used before the noun phrase. The sentences would be as in (1a -3b).
(absence of 'the ' for specific locations) 1a. an old lady in the rain 1b. *an old in lady in _rain (absence for 'the' for instruments) 2a. she place the newspaper over Sarah's head 2b. *she placed _newspaper over Sarah's head Possessive determiner error 3a Her umbrella was blown away. 3b.*She's umbrella was blown away

Subject verb agreement
If "similar but different" relationships between components of the L1 and L2 morphosyntactic systems caused interference resulting in misplaced NPs in English sentences, then it is assumed that learners will exhibit processing difficulties in sentences 4a-b when the number of the possessive pronoun fails to match the number of the noun that it modifies. Similar mismatch events could take place when one form in the L1 could have many equivalents in the L2 resulting the L2 learner facing difficulty in selecting the appropriate SVA as in 5a-b.
Agreement with numbers 4a. she holds the newspaper 4b. *she hold the newspaper Agreement and (non) inflections 5a they share the car 5b. *they shares the car

Copula 'be'
Choosing the copula 'be' is said to be the most problematic feature of the L2 for Malay learners due to its multiplicity of forms and absence in the Malay language. Given that the copula 'be' possess the same meaning when used as subject or predicate and often seen as 'not' essential in Malay (Nik Safih, 1995, p.212), there is a possibility that learners would include them as in 6(a-b) or omit them altogether as in 6(c-d).
Assumed to be equivalent to ialah/adalah with be.
6a. The old lady waiting in the rain is wet.
(NP + AP) 6a. The car is from the lady's home.
(NP + PP) (Possessive determiner error) 6c. The old lady waiting in the rain wet 6d. The car from the lady's home. Mismatch in tense 6e. The old lady are wet and cold. 6f. The old lady was wet and cold.
Mismatch between Singular/Plural form 6g. The ladies was wet and cold. 6h. The ladies were wet and cold.
The variations in the copula 'be' due to tense and plural forms have also been described as an additional reason for incorrect use of the form as in 6 (e-h).In addition to the sentence types described above, we also assessed the sentences for creative constructions and the ability to think beyond the literalness of the text as an aspect of real life sentence construction. To assess the various differences, two research questions were formulated.
1) Is there an increase in Malay speakers' use of accurate L2 forms across ages? 2) Is there a relationship between Malay speakers use of L2 forms across ages? 3) Is there a difference between Malay L2 learners ability to use accurate grammar forms?

THE PLIGHT OF THE LESS PROFICIENT SECOND LANGUAGE LEARNER: BETWEEN DUCKS AND PARROTS
The study Three groups of L2 learners were tested. A total of 112Malay speakers responded to the study. Three groups comprising rural beginners, urban beginners and young adults were given a writing task in real time and the sentences were analysed for differences. In order to allow for the clearest indication of L2 language use, the performance of three sets of primary learners performance were compared to two other beginner groups from the urban setting and a university setting (Primary Rural ,Year 6), according to the 3 settings (Primary, University, Workplace) with all other factors being equal. All subjects were from schools and a public university in Sarawak. The subjects had 10 minutes to complete the task (Refer Appendix 1) in class after which the researchers collected the task from their instructors.

Subjects
The rural beginner group were 23 Primary six pupils (12 year olds) from a rural school in Long San, Sarawak. The urban beginner groups were 22 Primary Four (10 year olds) and 37 Primary Five pupils (11 year olds) from Kuching, Sarawak. The young adults were undergraduates who had enrolled into the university a weak before the study (ages 19-21 years).

Instrument
An existing instrument was recreated and used to simulate authentic language use, constraint the length and variations in sentences produced and to allow subjects to use their own language. A three frame cartoon (Appendix 1) adapted from Yoon and Kellog's cartoon Discourse Completion Test (2002) was used given the fact that Pictorial compositions are increasingly becoming a regular feature of many writing classrooms in schools and language institutions. Each frame was followed by two lines that were used for describing the pictures. It was assumed that variations in task would be difficult to interpret due to task complexity. The first frame began with a "Sarah saw…" The settings and characters though initially designed for a Korean context were since it was possible to imagine a parallel action taking place in another part of Malaysia.

Procedures
All participants were given a paper and pencil test. The class teachers/lecturers were requested to conduct the test. All respondents were given 10 minutes to complete the task. All subjects were told to describe each event in their own words. There was no other instruction given for all settings.
The data for each group of subjects were analysed using ANOVA with repeated measures. The design was extended to include a between subject factor when comparing the performance of the beginner group with the adult L2 learner group and a within subject factor when comparing performance on L1 and L2 within the advanced group. Effects were considered significant when they were at or below the .01 level. Table 1 outlines the performance of the subjects according to child and adult learners.

Results
Complex sentences which reflected accurate use at the initial phrase but inaccurate form in the subsequent phrase in the same sentence were eliminated from the analysis. The beginners constructed

THE PLIGHT OF THE LESS PROFICIENT SECOND LANGUAGE LEARNER: BETWEEN DUCKS AND PARROTS
sentences that were largely simple with few determiners, pronouns and inflectional features. The number of accurate and consistent use of grammar forms for the sentences and the percentages for the various word uses as reflected in table 1.
1) There was a difference in L2 learners' ability to use accurate L2 forms. While the beginner subjects did not use many determiners, there beginners in Groups 2 and 3 were able to use more accurate forms compared to Group 1. The adult learners however were using more forms but the forms were largely incorrect.
We set out to determine whether there will be a relationship between Malay speakers use of L2 forms across ages. Our findings show that there was a significant difference across the subjects' use of L2 forms for possessive determiner errors, verb agreement in numbers and incorrect inflections as indicated in Table 2. *. Correlation is significant at the 0.05 level (2-tailed). **. Correlation is significant at the 0.01 level (2-tailed).

G1-(Rural -Primary) G2 & 3 (Urban -Primary) G4 -(Undergraduates)
In determining whether there was a difference between Malay speakers use of L2 forms, the analysis shows that there was a significant difference between the inflections, possessive pronouns and third person singular.  Table 3 presents the ANOVA results for the Malay learners' performance in the test. As seen from the table missing inflections due to the third person and possessive determiners appear to be the strongest cue for Malay learners across all levels.

Discussion
The results of the present experiment can be summarized as follows: All three beginner groups -the rural beginner group, the Urban primary 4 and 5 groups-showed a consistent interference effect in their use of a) articles, b) subject verb agreement and c) copula 'be'. Specifically, the rural students did not insert the article 'the' as in (i-ii). This is in line with Marylyna's (2012) findings. The urban beginners however, despite being younger evidently were more aware of the language as in (iii-iv) indicating that proficient L2 learners were capable of noticing and using determiners even at an early age.
(i) * Sarah saw old women time rain (12 year Old) (ii) Sarah saw old woman in raining (12 year old) (iii) ... covering her head with a newspaper in the rain (11 year old) (iv) The newspaper covered her head from the rain (10 year old) (v) … the wind blow Sarah umbrella. Then, Sara umbrella broken. * Suddenly, the wind blow to the Sara's umbrella. And then, Sara's umbrella is broken. The old woman use a newspaper to fold on the Sara head.
However, the additional processing load imposed by plurality and pronouns also showed up among the beginner learners for all levels. Most beginner learners appeared to substitute the 2 nd . Person singular pronoun with the Subject " Sarah" suggesting that learners tend to apply coping strategies such as 'ducking' in which information is ignored or reinterpreted in an easier manner such as in (vix). This appears to be a carried over effect until the university (refer x). Similarly, there was evidence of 'parroting' whereby language from one frame was recycled to another for a number of levels.

Implications of the study
So far, the study has considered two important issues in language learning: 1) what constitutes learning an L2 English in the Malaysian context, and 2) how does this learning take place. As for i) to know the language form is to be in a certain stage which comprises knowing the core and periphery (Anshen, 1986). The findings of the study have consequences for teaching English to L2 environments where there are more Malays students. In fact, much has been said and will be said about the language learning difficulties of L2 learners in general and Malaysian learners in particular.
The study provides additional evidence to support the fact that Malay students in particular and Malaysian students in general have problems even with differentiating between plural and singular forms. Though English was used as a medium of instruction, most students in the study regardless of child and adult learners continue to be influenced by their mother tongue when lost for words or revert to a simpler word form rather than use a difficult word wrongly and were not willing to communicate beyond the literalness of their textbooks. The learners in the study clearly had problems with subject verb agreement especially with plural nouns. In fact, given the number of years many of the subjects in this study may have spent on learning English, and the number of times, they got the words right and wrong, it is possible to assume that some learners may have reached some stage of fossilisation where they did not realize the error while others simple did not know when to use it or omit it. Getting learners to notice accurate forms is possible and can have a positive effect on the learners' language ability, but this can only come about if the instructor or better learner is able to detect the gaps on time and able to provide corrective feedback. By raising attention to the incorrect forms, it is possible for instructors to get other students to indirectly benefit from the input and gradually improve their interlanguage.