‘george-camilleri’

---

© Publications Committee, Faculty of Education, 2004 University of Malta, Msida, Malta MSD06 ISSN 1726-9725

George Camilleri

george.camilleri@gov.mt

George Camilleri (b.1952) studied at St Michael’s Teacher Training College, the University of Malta, the Institute of Education of the University of London, and attended further professional training at Leeds University. His research work Learner Autonomy: the Teachers’ Views (1999) was published by the Council of Europe and later translated into French. He taught English for 20 years in secondary and post-secondary schools, and since 1992 he has occupied the post of Education Officer for English with the Education Division, Malta. Camilleri has also published the anthology Realms of Fantasy: Folk Tales from Gozo, now translated into German, French and Hungarian. He is married to Helen, and has four children.

Introduction

The occurrence of so-called ‘Maltesisms’ in students’ language output, both in speaking and writing, is a familiar experience among teachers of English in Malta. Very often, such utterances come up as ‘howlers’ in staff room conversation and examiners’ reports. However, in spite of this widespread awareness, it appears that little reflection has taken place on the phenomenon of negative transfer in the context of L2 acquisition in Malta, in terms of its manifestation, its causes, and the way it operates.

Transfer

Transfer is one element in the wider phenomenon of language contact. When two or more languages come into contact, both at the level of the speech community and at the level of the individual, there is the possibility of cross-linguistic influences that may give rise to new dialects and language varieties, code-switching, transfer and other outcomes (Jacobsen, 2000). Transfer may operate in two directions: from L1 to L2 (‘substratum transfer’) and from L2 to L1 (‘borrowing transfer’). In the first case, the implication is that learners use their established knowledge of their first language in order to learn a second language. In the second, the implication is that the learning

Vol:2 No.1 2004 3-12

http://www.educ.um.edu.mt/jmer

Negative Transfer In

Maltese Students’ Writing In English

Abstract:

The influence of the mother tongue on the learning of other languages is a widely researched phenomenon. The operation of transfer is today a commonly acknowledged feature in second language acquisition. In the Maltese context, this field of inquiry addresses, among other issues, the learning of English by native Maltese speakers. In this paper, I describe briefly my recent research in the field, and propose a number of views and insights into how negative transfer operates in L2 acquisition in formal schooling at secondary level in Malta.

of a second language is bound to have an effect on one’s first language, for example, through code-switching and attrition, that is, the replacement of native language features by L2 features.

Transfer is not an entirely negative influence on learners’ L2 acquisition. Odlin’s definition underlines the ambivalent nature of the phenomenon:

Transfer is the influence resulting from similarities and differences between the target language and any other language that has been previously (and sometimes imperfectly) acquired. (1989: 27)

When the languages in play share linguistic features – such as an alphabetical system, particular grammatical categories, and rhetorical conventions – there is the likelihood of positive transfer, that is, the successful carrying of features of language from the learners’ L1 into their L2, as would be evident in their language output. In addition, positive transfer may also occur in the fields of metalanguage and paralanguage (Ellis, 1994:29). If there are similarities in the terms we use to describe the first language and the target language, the transfer of such technical terms will facilitate the understanding of the way the L2 works. Positive transfer may also occur in terms of skills related to language learning (such as reading strategies) and the way we use language (such as in turn-taking). It is thus highly probable that Maltese language learners use their knowledge of L1, and the experience of learning it, in order to learn their second and other languages.

Negative transfer

Transfer may, however, in certain circumstances, be a hindrance to L2 acquisition. This happens when transfer occurs in contexts where there is a ‘mismatch’ between, say, an L1 structure and the equivalent target structure. The outcome of such an occurrence would be a deviation from the target norm, what is termed as a ‘negative transfer’ error or ‘mother tongue interference’. Transfer errors are a typical feature of the interlanguage of the L2 learner, a kind of ‘intermediate’ language made up of L1 and L2 elements that are built into a systematic code by the L2 learner (Selinker, 1972).

Research project

The focus of my research was negative transfer in writing (Camilleri, 2003). My investigation consisted of a ‘focused description’, a research method whose eclectic nature gains from both qualitative and quantitative elements (Larsen-Freeman & Long, 1991: 15). The core data was obtained from 100 homework essay scripts written by Form 5 students preparing for the SEC examination in English Language. All the subjects in the sample were native Maltese speakers. The aim of the project was to find out whether and to what extent negative transfer was in operation in the students’ writing, to glean some insights into how it was manifested, and whether it was possible to develop some sort of taxonomy in terms of type and source of errors.

Such a ‘snapshot’ view of the phenomenon in the context of the Maltese classroom would, I believed, be of benefit to all of us involved in the field of teaching and learning in Malta.

My investigation followed a more or less typical pattern in error analysis studies (Ellis, 1994: 48) that are mainly inspired by the seminal work of Corder (1974). It consisted of five main stages: collection of data, identification of errors, description of errors, attribution of types and causes, and finally reflections on the findings.

Error Analysis

Error analysis, as many teachers would agree, is a highly complicated process. It is not easy to decide when an error has been committed, and even more difficult to specify its cause. There is therefore an element of arbitrariness in the analysis and evaluation of errors (Flick, 1979:60). Ideally, the analyst should interview the learner within a time-frame close to the commission of an error in order to find out the precise intention of the writer and the circumstances of the deviant writing. This opportunity is hardly possible to the researcher, and in any case, learners are sometimes unable to explain clearly what they want to say when errors are committed. In the identification and processing of errors, I therefore had to rely on my own 30 years’ experience of dealing with students’ scripts of the same type and level as those in the research sample.

Findings

The findings showed that negative transfer was indeed in operation, but probably less prevalent than what one may have expected, also when one keeps in mind the ‘annoyance factor’ such errors exert on teachers (Ellis, 1994: 67). The 125 negative transfer errors collected from the 100 scripts pertain to various types and sources but the incidence of 1.25 errors per script did not ring any alarm bells. In fact, the ratio of transfer errors tallies with that found by Borg (1994: 73) in her error analysis of 200 Form 1 JL English entrance examination scripts, in which she found 198 transfer errors, a ratio of 0.99 per script. However, for a full appreciation of the significance of these findings, one will have to look at the ratio of such errors in relation to all the errors in each script, in order to find out the relative proportion of transfer errors in students’ writing. This was not one of my aims, but it could indicate one of the directions of further research in the field.

When evaluating the frequency of errors, one should consider the significant characteristics of the sample that may have constrained the commission of such errors. The subjects of my sample had been learning English formally for 11 years, and the fact that they were preparing to take the SEC exam, though not sufficiently indicative, implied that they had reached a certain level of competence. In addition, since the writing was a homework task, the subjects had the opportunity of revising their work and thus reducing the number of errors in the final product.

Types of errors

In my analysis, I noticed that one way of characterizing the collected errors was in terms of their location in the surface structure of the sentence in which they occurred. This approach was consistent with that of other researchers in the field of error analysis (Dulay, Burt & Krashen, 1982). I subsequently developed a taxonomy based on the lexical and syntactical items affected by each error, such as ‘preposition’, ‘verb’ and ‘pronoun’. In many cases, errors affected more than a single lexical or syntactical item, and here decisions had to be taken on what constituted the core item. In other cases, the error concerned an entire utterance or string of lexical items. In such cases, I used the term ‘idiom’ to characterize such deviations. Using this surface taxonomy as a point of reference, the data yielded 13 different error types, as shown in Figure 1, grouped under the term ‘Language Area’.

Noun Adjective Adverb Verb

Verb form Idiom Preposition Pronoun Article Passive Voice Spelling Word Order Concord

Language Area

Error

Figure 1 – Types of errors

The most prevalent types of errors were those of ‘Preposition’, ‘Verb’ and ‘Idiom’, as can be seen in Figure 2. The first two often involved the incorrect substitution of single words, while the third type involved a deviation spread over a string of words.

0

5

10

15

20

25

30 Preposition Verb Idiom Pronoun Article Noun Adjective Adverb Word order Passive voice Spelling Verb form Concord

Figure 2 – Number of different types of errors

Number of errors

The two types of ‘Verb’ and ‘Preposition’ also yielded two L1 terms whose rendition in English turned out to be the most problematic language items for the L2 learners in the sample. These were the verb jagħmel and the preposition bi. The first occurred in 9.6% while the second in 4.8% of all errors in the data. In both cases, learners opted to translate the words into their perceived core meaning equivalent, namely make and with in instances when they were inappropriate. This may have happened because learners were ignorant of options other than the core meaning equivalent, as suggested by Kellerman (1984) and Larsen-Freeman and Long (1991:104). These errors therefore highlighted an area of interlangauge that may enable teachers to make predictions about learner difficulties. Figure 3 illustrates how the Maltese verb should have been rendered into English in those instances found in the data where each time it was erroneously rendered as ‘make’.

form (part of) spend (time) do (something) hold (an event) pluck up (courage) get (something done) give (a headache)

jaghmel

Figure 3 –Various English equivalents of ‘jagħmel’

Figure 4 shows the various English equivalents of ‘ bi ’ that were incorrectly rendered as ‘with’ in the sample data.

to play music ON a CD player to leave IN a car to travel ON a yacht TO the sound of IN an accident

bi

Figure 4 – Various English equivalents of ‘bi’

Taxonomy of sources

The next stage of my reflections concerned an attempt to establish a typology of sources or causes for the deviations. Although the one general cause was mother tongue influence, I wanted to explore the possibility of finding a lower-level class of causes based on common or different features of Maltese and English that might explain why a deviation has occurred. Typologies of causes or sources are in fact found in many error analyses that are concerned with transfer (Gass & Selinker, 2001: 77). The process of trying to find a source or cause of an error is an attempt to find an answer (or answers) to the question: ‘Why did the error occur ?’ In L1 negative transfer, the ‘source’ therefore identifies the nature of the ‘mismatch’ between L1 and

L2 features which happens every time an error occurs. As a result, the data produced a list of six sources, as shown in Figure 5 under the grouping ‘Mother Tongue’.

Literal translation L1 Form

Differentiation False friend

Coalescing New Category

Mother Tongue (Maltese)

Figure 5 – Sources of errors

In the classification, I found that I could borrow certain terms and notions from other taxonomies in the field that could be applied to the context I was investigating. These were five. The two terms ‘Differentiation’ and ‘New Category’, which appear in this study as possible ‘sources’ of errors, appear in a taxonomy of errors in Gass and Selinker (2001: 77). Another term, ‘Coalescing’, appears in a hierarchy of difficulty table drawn by Ellis (1994: 307), based on information given in Stockwell, Bowen, and Martin (1965). Two terms I borrowed to explain certain instances of errors were ‘Omissions’ and ‘Redundancy’; these are used in a surface strategy taxonomy designed by Dulay, Burt and Krashen (1982). The six sources or causes are explained below.

(i) ‘Literal translation’, as the term implies, refers to errors caused by a process of direct translation, whether they be single words (such as many errors involving verbs) or an entire string of words (as in the case of errors involving idioms). (ii) ‘L1 form’ also involves a process of direct translation. But apart from that, this process also involves the transfer of an underlying syntactic or orthographic feature from the mother tongue to the second language. An example is the redundant use of the definite article in L2 utterances. Maltese nouns are rarely used without the definite article, including abstract or generic nouns, which is not the case in English. (iii)‘Differentiation’ also known as ‘Splitting’ – concerns the correspondence of two or more L2 forms to one L1 form (Gass & Selinker, 2001: 76), such as do and make for the Maltese jagħmel. (iv) ‘Coalescing’ is the opposite process of differentiation, whereby two or more categories in the L1 collapse into one in the L2 (Stockwell, Bowen, & Martin, 1965). This may be exemplified by the case of two gender categories for objects in Maltese collapsing into one in English. (v) ‘New Category’ refers to the existence of a grammatical category in English which is not found in Maltese, such as the indefinite article. (vi) ‘False Friend’ consists of a L2 word or phrase with close orthographic and phonological similarities to a L1 word or phrase, but which is semantically different. An example is brave which looks and sounds similar to the Maltese term bravu (‘clever’).

0

5

10

15

20

25

30

35

40

45

50

Sources

Literal translation

L1 form

Differentiation

New Category

False friend

Coalescing

Figure 6 – Number of different sources of errors

In order to illustrate how the causes of L1 transfer are manifested in L2 writing, I will now present a number of examples of learners’ deviations as they occurred in the data. It can be seen from Figure 6 above that the three main causes of errors were attributed to Literal translation, L1 form and Differentiation. The first cause was more easily identifiable than the other two, since the bilingual reader could compare the parallel terms or string of words in the two languages, such as in the case of the utterance He was living under the eye of his parents. The other two causes addressed a deeper level of the sentence, since they brought into play grammatical and/or syntactic features and categories that were different in the two languages.

The deviation A black modern sunglasses , for example, appears to have occurred, at least partly, because of the different number grammatical category of the object being described while sunglasses is a plural noun in English, its Maltese equivalent nuċċali is a singular noun, hence the use of the indefinite article in the utterance.

An example of a Differentiation error was I didn’t tell anything about Charles. The deviation was located in the verb tell which probably should have been replaced by say. The learner here made the wrong choice between the two English equivalents of tell and say for the Maltese għidt.

An example of a deviation categorized under ‘False Friend’ is the utterance to fetch information on the internet, where the learner was misled into using fetch for “look for”, possibly because of the phonological and orthographic similarity between the English word and fetx , the Maltese word for ‘looked for’.

‘New Category’ errors, as stated earlier, concerned the difficulty learners encounter when they are required to process grammatical concepts or distinctions that are not found in their mother tongue. This difficulty is seen in errors like many people which… , where the learner has not yet mastered the distinction in the uses of which and who in the target language partly because such a distinction does not exist in the L1.

The error a bridge has his disadvantages is an example of a deviation caused by Coalescing. In this case, a grammatical distinction that exists in the L1 (two genders

for objects) is not replicated in the L2 (one new gender for objects), with the result that the learner made a choice where no choice existed.

It is interesting to note here that since the particular gender for objects in the target language does not exist in the L1, the error could also be considered as having been caused by a New Category. In fact, one of the findings of my research was the fact that in a number of cases a single deviation could be attributed to more than one cause.

Further reflections – stylistic transfer

My further reflections on the nature of the L1 transfer errors that I came across led me to conclude that it is possible that apart from lexical and syntactical transfer, other sorts of interlingual influences were in operation. One type of error, the avoidance of the Passive voice, could indicate that there is an element of ‘stylistic transfer’ in operation in L2 writing by Maltese learners. An example is if they built it… when the appropriate utterance should have been ‘if it was built…’. The avoidance of the passive voice in formal L2 writing (such as compositions) and its substitution by 3rd person plural structures where the agent is not defined may be attributed to the latter’s acceptability in formal writing in Maltese. Since such errors do not contain grammatical, orthographic or semantic flaws (hence, ‘covert’ errors), they may be considered as pertaining to stylistic transfer. Such influences have been noted to be in operation between other languages “The linguistic and rhetorical conventions of the first language interfere with writing in the second language” (Kaplan, 1966, cited in Connor, 1996).

Difficulty transfer

Another interesting reflection concerns the occurrence of errors in contexts where no apparent difficulty could be predicted and where the linguistic demands in the target language appeared to be attainable, judging from the level of performance of the entire script in which the deviation occurred. On examination, I realized that the errors occurred in structures whose L1 equivalents sometimes posed problems for learners in their L1 writing, such as when learners confuse two or more closely similar L1 terms. Errors such as She dressed the police uniform and It learns you discipline thus could be attributed to the ambiguity of the Maltese terms libset , (meaning both ‘she wore’ and ‘she dressed someone up’) and the close resemblance of jitgħallem (learns) and jgħallem (teaches) respectively. It appears therefore that sometimes Maltese learners of English carry an area of difficulty from their L1 into their L2 writing. The confusion caused by two similar Maltese words that is sometimes evident in the L1 writing of Maltese learners may be reflected in their writing in English, even when no such similarity exists in their L2 form.

Language and culture

My research also reinforced the notion of the close relationship between language and the culture it forms part of. In my error analysis, there was one particular case of an error which would probably defy understanding unless the analyst enjoyed a sound knowledge not only of the L1, but also of its culture. Connors (1996: 100) writes that research “points to the fact that written texts and the ways they are used vary according to cultural group… Writing as an activity is embedded in a culture”.

The error in question was “ Two were agony” , where the intended utterance was probably ‘Two were on the point of death’ or ‘Two were dying’. At surface level, there seems to be a fuzzy similarity between the erroneous and the native-like forms. A superficial interpretation of the error would suggest that it involves the omission of the preposition in before agony. That would at least make the utterance a grammatically correct and meaningful expression: Two were in agony. However, the intended meaning, inferred from the context of the entire composition, concerned two men who had been seriously injured in a traffic accident and who were now dying in hospital. In fact, the error draws back on an ancient custom, still extant in some Maltese and Gozitan villages, of tolling the death knell at the local parish church as a sign that one or more of the villagers was dying. A Maltese expression for both the death knell and the state of being on the point of death is agunija , and this may be literally translated as agony. The strange-looking and erroneous L2 form Two were agony is a literal translation of the correct L1 utterance Tnejn kienu agunija. This discussion of this particular error shows the importance for the interlanguage analyst to be familiar not only with the mother tongue of the subject, but also with its culture.

L3 transfer

Another finding which came out in the analysis, one that was not envisaged in the initial research queries, concerned L3 influence on L2. Errors like The job affascinates me may indicate the influence of L3, in this case Italian, over L2. This could happen particularly in those instances where L1 and L3 forms have closely similar features. This finding is corroborated by that of Stellini (1999) who also found evidence of L3 influence in the L2 writing of Maltese learners. Odlin states that “when individuals know two languages, knowledge of both may affect their acquisition of a third,” and “most probably, knowledge of three or more languages can lead to three or more different kinds of source language influence, although pinning down the exact influences in multilingual situations is often hard” (1989: 27). Since Maltese (L1) and Italian (L3) share many common cognates, it is possible that they mutually reinforce their influence on the English (L2) writing of Maltese learners.

Conclusion

This short paper was aimed at sharing with the reader some views and insights into how L1 transfer is in operation in the writing of Maltese learners of English at a particular level (Form 5). No doubt, educational practitioners working in the same

field will reflect on the validity or otherwise of my reflections, mediating them by their own experience and expertise. There are several avenues that future research in the field may take (cf. Camilleri, 2003:74), and it is hoped that eventually we will have a more solid and comprehensive view of the phenomenon. The importance of the learning of English in Malta can never be overstated, so they would be efforts well spent. Apart from language acquisition, language transfer addresses also the process of learning in general, because it helps us understand how learners use the known to learn the unknown.

References:

Borg, M. G. (1994). An error analysis of Junior Lyceum Entrance English Examination scripts. Unpublished B.Ed (Hons.) thesis. Malta: University of Malta.

Camilleri, G. (2003) The influence of learners’ knowledge of L1 (Maltese) on their learning of L2 (English). Unpublished Masters dissertation. Malta: University of Malta

Connor, U. (1996). Contrastive rhetoric: Cross-cultural aspects of second-language writing. Cambridge: Cambridge University Press.

Corder, S.P. (1967/1974). The significance of learners’ errors. In J. Richards (Ed.) Error analysis: Perspectives on second language acquisition. (pp.19-27). London: Longman.

Dulay, H., Burt, M. & Krashen, S. (1982). Language two. New York: Oxford University Press.

Ellis, R. (1994). The study of second language acquisition. Oxford: Oxford University Press.

Flick, W. (1979). A multiple component approach to research in second language acquisition. In R. Andersen (Ed.), The acquisition and use of Spanish and English as first and second languages. Washington D.C.: TESOL.

Gass, S. & Selinker, L. (2001). Second language acquisition: An introductory course. 2nd Ed., London: Lawrence Erlbaum Associates.

Jacobsen, B. (2000). Languages in contact – minds in contact. Paper delivered at the 12 th^ Innuit Studies Conference, Aberdeen, August 2000. Retrieved April 4, 2003, from: http://www.ilisimatusarfik .gl/dan/grl/bj_isc2000.html

Kaplan, R.B. (1966). Cultural thought patterns in intercultural education. Language learning 16 : 1-20.

Kellerman, E. (1984). The empirical evidence for the influence of L1 on interlanguage. In A. Davies, C. Criper & A.P.R. Howatt (Eds.), Interlanguage. Edinburgh: Edinburgh University Press.

Larsen-Freeman, D., & Long, M. (1991). An introduction to second language acquisition research. London: Longman.

Odlin, T. (1989). Language transfer: crosslinguistic influence in language learning. Cambridge: Cambridge University Press.

Selinker, L. (1972). Interlanguage. In J. Richards (Ed.), 1984/1997. Error analysis: Perspectives on second language acquisition. (pp. 31-54). London: Longman.

Stellini, S. (1999). Analysing spelling errors in Form 1 English compositions. Unpublished B.Ed (Hons) thesis. Malta: University of Malta.

Stockwell, R., Bowen, J., & Martin, J. (1965). The grammatical structures of English and Italian. Chicago: University of Chicago Press. Retrieved October 25, 2002 from: http://ejust.stanford.edu/method_interviews.html.