*Keywords:* Maltese, tense and aspect, interlanguage, second language acquisition

‘antoinette-camilleri-grima’

---

(Re-)Narrating the Gigantic …

Volume 9, No.1, Special Issue, pp. 57 77 Faculty of Education©, UoM, 2015

Exploring the acquisition of tense and aspect

by adult foreign learners of Maltese

Antoinette Camilleri Grima

University of Malta

Abstract: This article describes an initial analysis of the acquisition of Maltese as a foreign language by adult learners. It specifically looks at four verbal tense and aspect forms. Three intermediate-toadvanced level learners were asked to carry out a set of tasks, and the resulting data was compared to that produced by three adult native speakers of Maltese. The main finding was that on production tasks (unanalysed knowledge) the most commonly used tense/aspect by foreign learners is the ‘imperfett’ (unrestricted habitual aspect) which native speakers (NSs) use only half as much as the foreign learners (NNSs). On the other hand, native speakers use the ‘perfett’ (past tense) twice as much as the foreign users of Maltese. Another interesting finding is that on tasks that require analysed knowledge the main difference between NSs and NNSs was that the NNSs had greater difficulty with the progressive and the restricted habitual, but did not show significant variability with regard to the ‘perfett’ and ‘imperfett’. Thus, the results indicate that (i) there is a difference in the performance of NNSs in quantity rather than quality with regard to the ‘perfett’ and ‘imperfett’; and (ii) that on tasks requiring analysed knowledge NNSs have difficulty with regard to the progressive and restricted habitual forms.

Keywords: Maltese, tense and aspect, interlanguage, second language acquisition

Introduction

This study explores the interlanguage of adult foreign learners of Maltese. The concept of interlanguage refers to “the type of language produced by secondand foreign-language learners who are in the process of learning a language” (Richards, Platt & Weber, 1985 p. 145). In the 1960s research into child first language acquisition showed that “children have mental grammars

__________________ Corresponding author: antoinette.camilleri-grima@um.edu.mt

of their own at each stage of language development” (Cook & Singleton, 2014, p. 59). This idea was followed up by researchers of Second Language Acquisition (SLA), and in 1967 Corder published a seminal article about the ‘transitional competence’ of second language learners (Corder, 1967; Selinker & Lakshmanan, 2001). This concept was renamed ‘interlanguage’ by Selinker (1972) and has continued to be widely researched under that name (Selinker, 1996; Cook, 2014). In fact, the concept of interlanguage is considered to be such a fundamental development in SLA research that Eckman (2012) divides the research history of second language phonology into two: preinterlanguage and post-interlanguage.

Indeed, since the 1970s it has been established that second language learners have an independent grammar of the foreign language, consisting of fragments of the language they are learning, elements of their native grammar, and other factors that cannot be easily accounted for, amongst which are those that derive from the immediate needs of communication (Corder, 1975). Widespread research has demonstrated that interlanguage is systematic, both intra-speaker and across speakers including learners with different first languages (O’Grady, Dobrovolsky & Katamba, 1996; Sorace 2005; Cook & Singleton, 2014).

The findings of SLA research are of fundamental relevance to language pedagogy. While formal grammars “are concerned with a specification of the formal properties of language, with the ‘code’ rather than the ‘use of the code’” (Allen, 1974, p. 59), a pedagogic grammar aims to provide the learner with material that is not only scientifically sound, but also appropriate for learning. This means that, for example, formal and pedagogic grammars vary in the formulation of grammatical complexity, sequencing of items, type, variety and quantity of examples, and various aspects of presentation. Applied linguistics is interdisciplinary (Widdowson, 2005), and a scientifically sound pedagogic grammar in addition to formal descriptions of language takes into account various factors such as the age of the learner, the language learning scenario, and SLA theory especially findings on learners’ interlanguage. Widdowson (2003) emphasises that it is up to applied linguists to make linguistic insights intelligible to language teachers, and Cook & Singleton (2014) state that SLA research is a valuable instrument for promoting greater teacher awareness of various aspects of the acquisition process of a foreign language.

This project takes the first step to unravelling the interlanguage of adult foreign learners of Maltese by focussing on verbal tense and aspect. It is an attempt at describing, even if it only scratches the surface, how knowledge of Maltese grammar in the mind of the foreign learner is represented through tasks that test analysed and unanalysed knowledge; and of how this differs from the representations of the native speaker. Our ultimate aim in applied

linguistics would be to draw the attention of syllabus writers, textbook authors and classroom teachers as to how they could better align the teaching of Maltese grammar to the foreign learner context. Since this study stems from theoretical considerations about verbal tense and aspect in Maltese, and the empirical research was conducted within the formal theory-based paradigm, we shall first outline how grammarians of Maltese express the significance of the chosen verb forms.

Maltese verbal tense and aspect

Tense expresses “the time at which an event occurred, or at which some state of affairs” took place (Hurford, 1994, p. 239), while aspect “is concerned with factors such as duration or completeness of events and states of affairs” (Hurford, 1994, p. 239). In Maltese, aspectual distinctions are also found to operate in nouns (de Waard, 1981) but the present study is limited to verbs. According to de Waard (1981) “the aspectual opposition is found to be the fundamental one in Maltese, with tense and modal associations deriving from it” (p. vii). Borg & Azzopardi-Alexander (1997) explain that “all finite verbs in Maltese distinguish two combined tense-aspect forms…in general the Perfect is associated with past events, while the Imperfect realizes timelessness and durativity” (p. 220). Examples of the ‘perfett’ (Perfect) would be sentences 1 and 2, and examples of the ‘imperfett’ (Imperfect) would be sentences 3 and 4 below.

1. Pawlu xtara karozza. ( Paul bought a car )

2. It-tfajla iggradwat mill-Università ta’ Malta. ( The young lady graduated from the University of Malta )

3. Pawlu jixtri l-ħobż mingħand Victor’s Bakery. ( Paul buys bread from Victor’s Bakery )

4. Eluf ta’ studenti jiggradwaw mill-Università ta’ Malta. ( Thousands of students graduate from the University of Malta )

The verbs ‘xtara’ ( he bought ) and ‘iggradwat’ ( she graduated ) in sentences 1 and 2 respectively illustrate an event that happened in the past, i.e. before the moment of speaking (therefore, past tense), and refer to an action that was started and completed (therefore, perfective aspect). On the other hand, the verbs ‘jixtri’ ( he buys ) and ‘jiggradwaw’ ( they graduate ) in sentences 3 and 4 respectively are actions that have happened before the moment of speaking, are happening at present and will continue happening in the future (hence, unrestresticted habitual aspect). In the case of unrestricted habitual aspect each action is started and completed numerous times, repeatedly, without specific reference to the moment of speaking. For instance, the action encoded in the verb ‘jixtri’ ( he buys ) in sentence 3, is started and completed each time that Paul buys bread, hence unrestricted habitual aspect, which would

become restricted habitual aspect if I use a time adverbial like ‘din il-ġimgħa’ ( this week ) and ‘qed’, as in sentence 5:

5. Din il-ġimgħa Pawlu qed jixtri l-ħobż mingħand Victor’s Bakery. ( This week Paul is buying bread at Victor’s Bakery ).

According to traditional Maltese grammar, including that which is still taught in schools and in other teaching scenarios, the ‘perfett’ is the base form and morphological processes are considered in relation to it (Borg, 1981). In Maltese there is no infinitive form, and the citation form is the ‘mamma’, i.e. the third person singular masculine form of the verb in the perfect form. The ‘mamma’ of verbs of non-Semitic origin, which have lost the infinitive form in Maltese, is similarly used as the citation form, e.g. to look up the meaning of a verb in the dictionary. To give but one example, ‘kiteb’ ( he wrote ) is the third person singular masculine in the perfect form (i.e. the ‘mamma’). If a learner comes across this verb in the ‘imperfett’ as in ‘tikteb’ ( she writes ) they would need to know that the ‘mamma’ is ‘kiteb’ in order for them to look up its meaning in the dictionary. While this would not pose any difficulty for the native speaking learner, it is an insurmountable problem for the foreigner. But this is not the focus of this paper.

I have limited my research project to four temporal/aspectual distinctions. The diagrams below illustrate the temporal and the aspectual characteristics of the four aspectual forms chosen for this study: the ‘perfett’ (past tense) and the ‘imperfett’ (unrestricted habitual) which are finite forms of the verb; the present participle (progressive aspect) which in certain cases occurs instead of a finite verb (Borg, 1981); and the construction of ‘qed + imperfett’.

In the diagrams below, the line T(-x) through T(0) to T(x) represents time, while the upper part of the diagrams represents aspectual distinctions. T(-x) represents a point in time prior to the moment of utterance, T(0) is the moment of utterance, and T(x) is a point in time subsequent to the moment of utterance.

(a) The unrestricted habitual aspect, i.e. the ‘imperfett’, e.g. ‘jimxi’ ( he walks ):

--------------------------

T(-x)------------------------------------------T(0)-------------------------------T(x)

The ‘imperfett’ encodes an infinite series of repeated, completed journeys (broken top-line in diagram above), i.e. unrestricted habituality which has happened prior to the moment of speaking and

will continue happening in the future, and in each occurrence the action is started and finished (Borg, 1981, p. 99).

(b) The progressive form, i.e. the active participle, e.g. ‘miexi’ ( he is walking ):

S-------------G

T(-x)-------------------------------------T(0)-----------------------------------T(x)

In this case, one particular journey is in progress. It started (S) some time in the past, i.e. prior to T(0) and will continue until the end of the journey (G) some time in the future.

(c) The restricted habitual form, constructed with the particle ‘qed’ followed by the Imperfect, e.g. ‘qed jimxi’ ( he walks these days ):

----------------------------

(S)---------------------(G)

T(-x)--------------------------------------T(0)---------------------------------T(x)

In the construction with ‘qed’ there is an interaction between the encoding of a bounded journey in progress (middle section of the diagram) and of an unbounded series of repetitions of such journeys (top section of the diagram) (Borg, 1981, p. 148). The use of time adverbials or other contextual cues would make it clear whether the ‘qed’ construction is referring to a bounded journey in progress or to a series of repetitions of journeys.

(d) The past tense, or ‘perfett’, e.g. ‘mexa’ ( he walked ):

S-----------------------G

T(-x)--------------------------------------------------T(o)-----------------------T(x)

Here the line S-G represents the completed journey encoded by the verb ‘mexa’ ( he walked ), which took place in the past, i.e. prior to the moment of speaking (Borg 1981, p. 98).

The investigation

There are various ways in which SLA researchers seek to describe the learners’ grammar, or interlanguage, such as by looking at frequency, variation and co-text using learner corpora collected from authentic communicative settings (Granger, 2012); by devising and analysing surveys that collect self-report data (Dornyei & Csizer, 2012); or by focusing on qualitative data such as ethnography and conversation analysis (Friedman, 2012), among others.

On the other hand, formal theory-based methodologies aim to describe and explain the nature of second language learners’ interlanguage grammar by eliciting data that is considered to represent their mental idiolects (Lardiere, 2012). The present investigation falls within formal theory-based methodologies in SLA research (Ionin, 2012; Lardiere, 2012), and as such it aims to make conjectures about the nature of the NNSs mental grammar by indirectly inferring it from observable performance. Ionin (2012) clarifies that “quantitative, empirical data on learners’ production and comprehension of the target language are used to draw conclusions about the underlying grammar (p. 30). Lardiere (2012) highlights that the acquisition of the grammar of a second language lies at the heart of the study of SLA. Norris & Ortega (2012) emphasise that SLA researchers make inferences about learners’ grammatical representations to describe what L2 features have been acquired (or not), or are being acquired over time. From a formal SLA perspective, what is acquired is grammar construed as mental representations of morphosyntactic rules (Norris & Ortega, 2012).

In my study, the research question stems from the theoretical considerations about tense and aspect in Maltese as already explained. This empirical study sets out to establish the frequency of occurrence and accuracy in the use of the four chosen verbal forms , on a set of five tasks, by three intermediate-toadvanced foreign learners of Maltese. These three learners were chosen on the basis of convenience sampling (Cohen, Manion & Morrison, 2007) and were considered to have reached an intermediate-to-advanced stage of learning because of their length of stay in Malta and their known efforts to learn Maltese both formally and informally. A control group of three adult native speakers of Maltese provides the comparative data because as Norris & Ortega (2012) explain most researchers working from a formal linguistic perspective find it imperative to adopt assessments that compare learners’ grammaticality judgements of particular rules with those of a baseline group of native speakers. The theoretical consideration under study refers to the traditional grammar axiom that the ‘perfett’ is the base form, the simplest, and therefore the one more commonly used by native speakers. Our research

question revolves around whether the same can be said of the interlanguage of foreign learners of Maltese.

The three learners, on which our analysis is based, were (names have been changed): (i) Maria, a native speaker of French, aged 33. She studied linguistics at postgraduate level in France and had been learning Maltese for six years. She lived in Malta for a few months at a time and learned Maltese mostly by recording speakers and transcribing the texts, by studying grammar books, by following lectures in Maltese at the University of Malta, and by asking for clarifications from students of Maltese. (ii) Jean, a native speaker of English, aged 21. She had lived in Malta for 12 years but only tried to learn Maltese after she left school in the last six years. She attended private lessons and practised Maltese within a drama group. (iii) Anna, a native speaker of German, aged 33. She had been living in Malta for five years during which time she had been trying to learn Maltese by interacting with Maltese speakers in a variety of social contexts.

Three native speakers of Maltese were asked to act as a control group. These were:

(a) Robert, a medical doctor, aged 25. (b) Alexia, a teacher, aged 35. (c) Josephine, a student of linguistics, aged 22.

Each subject was asked for his/her consent to participate in a study that was looking at the acquisition of Maltese as a foreign language, and was informed that names would be changed in the report. Both the foreign learners and the native speakers worked out the same tasks. Each participant worked on his own at home and was asked not to consult any book or seek help from other people. When they handed in the work they all said that they had worked on their own, and lamented that the tasks were very demanding and timeconsuming.

Data collection in SLA is not very frequently based on natural language use due to the difficulty of encountering enough occurrences of the item under study (Mackey & Gass, 2012). Therefore, researchers have to resort to clinical data which could be experimental in nature, such as fill-in-the-blanks, or more open ended as in written composition (Mackey & Gass, eds. 2012). Other data collection methods include picture-description and translation tasks (Nesselhauf, 2004). A great deal of SLA research which focuses on grammatical topics uses grammaticality judgement tests (Lardiere, 2012;

Norris & Ortega, 2012). Originally, grammatical acceptability tasks were used to identify native speakers’ linguistic competence: “A sentence which is judged as grammatical by a native speaker is part of that speaker’s mental grammar, while a sentence which is judged as ungrammatical is in violation of a linguistic rule of the speaker’s mental grammar” (Ionin, 2012, p. 30). In SLA, learners’ grammaticality judgements unravel the stage and nature of their interlanguage. Ionin (2012) specifies that “typically, formal SLA studies use experimental methodologies, such as elicited production, grammaticality judgement tasks, and interpretation tasks” (p. 30).

In my investigation the data was collected by means of a set of five tasks, which included methods ranging from grammaticality judgement and translation, through to two picture tasks, to a free writing exercise. The data collected from Tasks 1 and 2 is different in quality from that collected in Tasks 3 to 5. In SLA research a metacognitive distinction is made between analysed and unanalysed linguistic knowledge (Hulstijn, 1990; Bialystok & Bouchard Ryan, 1985). In simple terms the main difference between what is required from the learners in Tasks 1 and 2 on the one hand, and Tasks 3 to 5 on the other, relates to the kind of knowledge required to either know in a linguistic way what you are dealing with, or to ‘automatically’ apply your knowledge in a communicate event.

Task 1 aims to reveal the learners’ mental grammar and to elicit their explicit analysed knowledge, and Task 2 requires the use of analysed knowledge in translating from English to Maltese (Bialystok & Bouchard Ryan, 1985). When working on these two Tasks the learner needs to be aware of the language rules and to express this knowledge. In Tasks 3 to 5 the learners produce language that could be unanalysed, i.e. does not necessarily belong to their mental grammar but could consist of chunks learned and used routinely: “for deliberate speech monitoring and successful error correction to take place, second language learners need not have explicit, verbalizable knowledge of the language rule involved at their disposal” (Hulstijn, 1990, p. 43). For this reason it is useful to obtain both kinds of data (analysed and unanalysed knowledge).

Task 1 consisted of a grammaticality judgement test. Grammaticality judgement tests “are taken as primary evidence for the nature of speakers’ linguistic competence…L2 learners’ grammaticality judgements can similarly inform us about the learners’ linguistic competence in their target language” (Ionin, 2012, p.30). Furthermore, in order to access the foreign learners’ understanding, or analysed knowledge, they are also asked to give the reason why a construction is acceptable or not. The ability to reject ungrammatical sentences is considered stronger evidence of knowledge of linguistic constraints than simply accepting what is grammatical (Lardiere, 2012).

Our test consisted of a list of forty sentences (Table 1), with each verb form represented by at least eight sentences of which a minimum of four were acceptable or unacceptable (Table 2).

1. Fis-sajf immur ħdejn il-baħar.

2. It-tifel qed jixrob il-ħalib dan l- aħħar.

3. Arah l-ajruplan miexi bil-mod fis-sema.

4. Ix-xitwa li għaddiet immur għax-xogħol bil-karozza.

5. Siegħa ilu t-tifel qed jikteb ittra.

6. Il-ballun dieħel fix-xibka waqt il-logħba tal-bieraħ.

7. Il-bieraħ mort l-iskola.

8. Jien niekol ħafna ġobon.

9. Il-lotterija tielgħa l-Ħadd li għadda.

10. Il-bieraħ tant nixrob li żaqqi sa tinfaqa’.

11. Is-sena l-oħra Ġanni siefer tliet darbiet.

12. Il-karozza nieżla dik in-niżla.

13. Dalgħodu nieżla x-xita imma issa l-ajru għaddielu.

14. Tal-ħanut ibiegħ minn kollox.

15. Il-karozza riesqa l-bieraħ.

16. Il-missier qed imur jaħdem l- għalqa.

17. Is-sajf li għadda tagħmel ħafna sħana.

18. F’Diċembru li għadda niżlet ħafna xita.

19. Għada sifirt għall-vaganzi.

20. Fix-xitwa tagħmel ħafna kesħa.

21. Għada ħadem fl-uffiċċju.

22. Ix-xahar li ġej kiteb ktieb.

23. Ġanni jimxi kuljum mid-dar sal-iskola.

24. It-tifla tkellmet kmieni.

25. L-omm qed issajjar fil-kċina dalgħodu.

26. It-tifla qed iddur mal- kantuniera.

27. L-omm xtrat xirja kbira l- ġimgħa l-oħra.

28. Il-lejla Pietru qed imur it- teatru.

29. L-għalliem ħiereġ mill-iskola l-bieraħ.

30. Sa jumejn oħra kiel kollox.

31. Jien qed nikteb fuq din il- karta.

32. Is-sena l-oħra Marija qed iddur b’ommha.

33. Is-sagristan qabel miet qed idoqq il-qanpiena.

34. Il-qassis ħiereġ mill-knisja.

35. Il-bieraħ l-ajruplan nieżel f’salt.

36. Is-sena ddieħla mar jaħdem.

37. Illum ix-xogħol tiela’ mgħaġġel.

38. Is-sena l-oħra jgħaddi mill- eżamijiet kollha.

39. Il-karozza riesqa `l hawn.

40. Kittieb tas-seklu l-ieħor qed jippubblika ħafna kotba. Table 1: List of sentences in Task 1

Unrestricted habitual

Progressive Restricted habitual

Past tense

        1 8 14 20 23

4

10

17

38

3

12

34

37

39

6

9

13

15

29

35

2

16

25

26

31

5

28

32

33

40

7

11

18

24

27

19

21

22

30

36

Key:  grammatically acceptable sentence;  grammatically unacceptable sentence Table 2: Acceptable and unacceptable forms in Task 1

In designing Task 1 special attention was paid to make sure that the sentences were either acceptable or not by using time adverbials. For example, Sentence 19:

19. *Għada sifirt għall-vaganzi. ( Tomorrow I went abroad on holiday )

is clearly unacceptable because it starts with a time adverbial indicating the future ‘għada’ ( tomorrow ) which is followed by a verb in the past ‘sifirt’ ( I went abroad ).

The sentences were randomly presented to avoid that the subjects discover any pattern in what was being tested. The subjects were asked to read each sentence, decide whether it was grammatically acceptable or not, and mark it with a  or a  accordingly. For the sentences they considered unacceptable they were also asked to give a reason why they thought it was incorrect and to produce the acceptable form. So, for instance, for Sentence 19 they were expected to say ‘wrong verb form’ or ‘one cannot use the past tense with ‘tomorrow’. Each of these steps requires a very high degree of analysed language knowledge.

Task 2 was a translation task consisting of forty sentences in English to be translated into Maltese (Table 3).

1. Peter drove a car.

2. John goes to the cinema every week.

3. The boy is asleep.

4. You are drawing a picture.

5. The girl washed her clothes.

6. The bird flew in the sky.

7. The boy puts on his coat before going out.

8. That pianist plays classical music.

9. The workman is coming down the ladder.

10. The sun is rising.

11. The children are playing in the yard.

12. Our team is winning.

13. People are talking behind my door.

14. Mother is at home.

15. The fat man eats lots of food.

16. The caretaker opened the window.

17. The sun rises in the east.

18. The man is coming out of the office.

19. Philip is reading a book.

20. The boys are cooking.

21. The penitent is sorry for his sins.

22. The student thinks about his homework.

23. She whistled a tune.

24. You drank a cup of tea.

25. We are eating.

26. The monkey is standing on two feet.

27. Sandra writes many letters.

28. She cut her finger.

29. I ate in the kitchen.

30. This teacher travels every summer.

31. The sun is setting.

32. My son is growing up.

33. The sun sets in the west.

34. My telephone rings every morning.

35. The young lady is entering the shop.

36. The train is approaching.

37. The theatre is opening at the moment.

38. The girls are laughing.

39. The child sang in the festival.

40. He worked very hard. Table 3: List of sentences in Task 2

Translation has been used by SLA scholars to obtain linguistic data as in Khresheh (2012) who used translation over a year and a half to test whether learnt knowledge is transformed into acquired knowledge by practice. Task 2 was intended to generate a constrained linguistic response in the target language and required the activation of explicit knowledge. SLA researchers distinguish between implicit and explicit acquisition of knowledge: “implicit learning is acquisition of knowledge … by a process which takes place naturally, simply and without conscious operation…Explicit learning is a more conscious operation where the individual makes and tests hypotheses in a search for structure” (Ellis, 1994, p. 1). Task 2 is: (i) formal, i.e. it focuses on the code and refers to information the learner has about the properties of that code as opposed to functional use in communicative situations; and (ii) within the formal limits it also operates on already existing explicit knowledge (Krashen, 1985; Norris & Ortega, 2012). The sentences in Task 2 were designed very carefully to accommodate the research items (Table 4), and time adverbials were used for this purpose.

Unrestricted habitual

Progressive Restricted habitual

Past tense

2 imur 7 jilbes 8 idoqq 15 tikteb 18 isiefer 23 jiekol 25 titla’ 30 jaħseb 33 tinżel 34 idoqq

3 rieqed 9 nieżel 10 tielgħa 14 wieqfa/qiegħda 19 nieżla 22 qiegħda 26 ħiereġ 29 niedem 35 dieħla 36 riesqa

4 qed tpenġi 11 qed jilagħbu 12 qed jirbaħ 13 qed nieklu 20 qed jikber 21 qed jitkellmu 27 qed jaqra 28 qed isajru 37 qed jiftaħ 38 qed jidħku

1 saq 5 ħaslet 6 tar 16 qatgħet 17 kilt 24 fetaħ 31 saffret 32 xrobt 39 kanta 40 ħadem

Table 4: Verb forms in Task 2 (The numbers refer to the sentences)

Task 3 consisted in a series of eleven pictures, specified with time indicators to restrict the use of verb forms (Figure 1). The subjects were asked to answer the question, ‘What do these pictures tell you about John?’. Stimuli for foreign language (Lf) production could be auditory or visual, apart from the more direct linguistic prompts as in Task 1 and Task 2. Visual stimuli in the form of pictures are used in SLA research in various ways and linked to a range of goals. For example, Bialystok, McBride-Chang and Luk (2005) used pictures to elicit data in order to compare phonological awareness and decoding among three groups of children (monolingual, bilingual, and L2 learner). Joo (2003) used pictures and matching/non-matching sentences to study how Korean learners of English as L2 develop knowledge of the English locative alternative. The intention of Task 3 was to uncover ‘errors’ in verb production which would presumably reflect the stage of interlanguage development in this regard.

Figure 1: Task 3

Task 4 was more open-ended than Task 3 and consisted of 6 pictures (Figure 2). The subjects were asked to narrate the story illustrated in the pictures. This time they were not given time indications. They were allowed to choose the forms they knew best and which they found easier to use, as they would in a normal communicative situation.

Figure 2: Task 4 pictures (showing Anna’s writing)

In Task 5 the subjects were asked to produce a piece of writing of at least 10 sentences, entitled “F’Xatt il-Baħar” ( On the beach ). As in Task 4 this exercise required the use of unanalysed knowledge, i.e. verb forms the learners know and are able to use in a language production task. By comparing the production of learners with that of native speakers a clearer picture of interlanguage can be obtained. A sample of Anna’s work on Task 5 is presented in Fig. 3.

Figure 3: Sample of Anna’s work on Task 5

Results and Discussion

This section describes the results. It reports the individual learner results on each task, and compares the non-native speaker data with that obtained from native speakers.

Task 1 As expected, there was agreement by the NSs about the acceptability or otherwise of the sentences given, and disagreement among the NNSs. This tallies with results obtained internationally from tests conducted with NSs and NNSs (e.g. Sorace & Filiaci 2006). The NSs judgements were aligned with those of the author. On the other hand, the NNSs responses varied in some areas from those of NSs.

Maria’s explanations about the ungrammaticality of sentences were varied. She concentrated on meaning and prepositions rather than on verbs. She did not interpret correctly the time adverbials in sentences 2 and 5 and the past tense in 24. She judged the use of prepositions in sentences 1 and 14 unacceptable when in fact they were correct, and in sentences 18, 27 and 32 she re-wrote the sentence by changing the verb into the first person. The variability in her ‘errors’ reveals an element of uncertainty when dealing with Maltese grammar. However, she only made one mistake with reference to the ‘perfett’ and ‘imperfett’.

Jean’s grammaticality judgement was closest to that of native speakers. She identified all the unacceptable sentences, and correctly gave an explanation (e.g. ‘incorrect verb tense’), and wrote down the correct form. Her judgement differed from that of NSs in only one instance: for sentence 34 she produced the past tense instead of the progressive. Jean’s performance shows that she has reached a near-native level in terms of tense/aspect knowledge and use. Before taking the test Jean had lived in Malta for twelve years and had received formal instruction in preparation for the school leaving exam in Maltese, even though she actually sat for it a few years after leaving school. Her performance also seems to indicate she was test-wise as she seemed to be looking at verb forms, unlike the other NNSs who were sometimes sidetracked by pronouns, prepositions, adverbs and vocabulary. Indeed, this raises the question as to whether test-takers should have been informed beforehand about what exactly the task was testing.

Anna’s judgements differed in 11 sentences from those of NSs. Her problem lay in correctly interpreting the use of the progressive and the restricted habitual in sentences 5, 6, 13, 28, 29, 32, 33, and 40. While Anna did not seem to have problems identifying the correct usage of the unrestricted habitual and the past tense, she did not produce correct responses to the progressive

and restricted habitual verb forms. It seems that apart from the latter forms Anna has mastered the grammaticality of the ‘imperfett’ and ‘perfett’ forms.

Task 2 In Task 2, as in Task 1, NSs produced highly similar data. NSs gave different replies in two sentences: Robert translated Sentence 26 using the restricted habitual form, but this mistake is possibly due to fatigue or stress. In sentence 29 Robert and Maria used the verb ‘jissogħbih’ (unrestricted habitual) while Josephine used ‘niedem’ (progressive), but this is probably due to a difference in formulaic prayer learned in childhood. For this reason, with hindsight it seems to me that this sentence was not appropriate for this test.

Maria did not translate sentences 12 (‘qed’ construction) and 29 (problematic sentence anyway). In sentences 9, 26, 35 and 36 which required the active participle indicating progressive aspect, she produced inappropriate forms. In sentence 31 she confused the first person (‘saffart’) with the third person (‘saffret’). Maria’s performance on Task 2 is somewhat similar to her performance on Task 1 in that her ‘errors’ were variable, and she had the major difficulty with the use of the progressive and restricted habitual.

Jean’s translations match those of native speakers except in two cases which do not impinge on our research question. In sentence 29 she used an inexistent form of an uncommon verb ‘nissogħbi’, which again is likely due to a prayer she had learned but forgotten over time. In sentence 36 she translated ‘train’ as ‘vapur tal-art’ while all the other NSs and NNSs translated it as ‘tren’. Task 2 result shows once again that Jean’s analysed knowledge of tense/aspect is near-native.

Anna did not translate the verbs in sentences 6, 8, 14, 31, possibly due to lack of vocabulary knowledge. Sentences 9, 10 and 35, required the progressive form but Anna produced the ‘imperfett’/unrestricted habitual. In sentences 36 and 37 she used a future construction instead of the progressive or restricted habitual. Also, in sentence 33 Anna used the progressive instead of the unrestricted habitual. Anna’s performance on Task 2 tallies with that on Task 1 because in both cases she had major difficulty with the progressive and restricted habitual forms.

Overall, Tasks 1 and 2 reveal that NNSs are relatively comfortable with their knowledge about the ‘imperfett’ and ‘perfett’ tense/aspect forms, but find the progressive and unrestricted habitual structures problematic. Thus it seems that their knowledge of the active participle and the ‘qed’ construction is still lacking or variable in their interlanguage. Indeed, the main conclusions from the results of Tasks 1 and 2 which focussed on the learners’ analysed knowledge, can be summarised in this way: (i) The analysed knowledge of the NNSs with regard to the ‘imperfett’ and ‘perfett’ forms is apparently

comparable to that of NSs (admitting that statistical significance could not be carried out); and (ii) NNSs at an advanced stage of learning Maltese have not yet acquired analysed knowledge of the progressive and restricted habitual forms.

Tasks 3, 4 and 5 Tasks 3, 4 and 5 each consisted of a written exercise, and focussed on productive data which tap on unanalysed knowledge. Task 3 was slightly restrictive because it consisted of eleven pictures with a time reference against each picture. Task 4 was more open ended with only six pictures and no timereference, and Task 5 was completely open-ended although a title was given.

Tables 5 to 9 show the quantities of verb forms used by each test-taker, by type, i.e. the unrestricted habitual (‘imperfett’), the progressive (active participle), the restricted habitual (‘qed + imperfett’), and the past tense (‘perfett’). One immediately notices that the presence of the progressive and the restricted habitual forms is negligible. There were no occurrences at all in Task 5 (Table 7), and no instances of the progressive and only 4 instances in Task 3 (Table 5). In Task 4, the progressive features only twice, while the restricted habitual was used 5 times by one learner and 4 times by one native speaker. From the NNSs Maria seems to be the most at ease using the ‘qed’ construction. In fact, in Task 4, she used the restricted habitual in 5 sentences. Overall, our data shows that the progressive and restricted habitual forms were avoided by both NSs and NNSs on these production tasks.

Thus, the data obtained from Tasks 3, 4, and 5 indicates that we now need to turn our attention to the use of the ‘imperfett’ and the ‘perfett’ since they were the most frequently used forms on these tasks. Table 5 shows that with the exception of native speaker Robert who chose to narrate the story using the past tense, there are minor differences in the use of the ‘imperfett’ by NNSs and NSs on Task 3 (Table 8), but an important difference in the quantity of the ‘perfett’ (Table 9). Task 5 does not give us stark contrasts between the two groups.

Unrestricted habitual (‘imperfett’)

Progressive Restricted habitual

Past Tense ‘perfett’ Maria 4 0 1 7 Jean 4 0 0 5 Anna 6 0 0 5 Robert 3 0 1 26 Alexia 6 0 2 2 Josephine 1 0 0 6 Table 5: Occurrences of verb forms in Task 3 by NNs and NSs

Unrestricted habitual ‘imperfett’

Progressive Restricted habitual

Past tense ‘perfett’ Maria 8 1 5 2 Jean 2 0 0 16 Anna 20 0 0 1 Robert 5 0 0 20 Alexia 7 1 4 19 Josephine 7 0 0 13 Table 6: Occurrences of verb forms on Task 4 by NNSs and NSs

Unrestricted habitual ‘imperfett’

Progressive Restricted habitual

Past tense ‘perfett’ Maria 17 0 0 2 Jean 18 0 0 0 Anna 10 0 0 0 Robert 13 0 0 0 Alexia 18 0 0 1 Josephine 12 0 0 0 Table 7: Occurrences of verb forms on Task 5 by NNSs and NSs

The results obtained from Task 4 are interesting. In this case, the NNSs used the ‘imperfett’ almost twice as much as the NSs, and the ‘perfett’ features less than half the time in NNSs data when compared to NSs data (Tables 8 and 9). The total count of occurrences of the ‘perfett’/past tense on the three Tasks reveals that while NNSs used the ‘perfett’ 38 times, the NSs used it 83 times. Overall there is quite a clear indication that where native speakers prefer the ‘perfett’, the NNSs use the ‘imperfett’.

Task 3 Task 4 Task 5 NNSs NS NNSs NSs NNSs NSs 14 10 30 19 45 43 Table 8: Total of occurrences of the ‘imperfett’ (unrestricted habitual) on Tasks 3, 4 and 5 by NNSs and NSs

This has to be highlighted in view of the fact that in Tasks 1 and 2 NNSs did not seem to have major difficulties with either the ‘perfett’ or the ‘imperfett’, and therefore their significant preference for the ‘imperfett’ on production tasks must have some meaning. Our conjecture here would be that in spite of NNSs analysed knowledge of both the ‘imperfett’ and the ‘perfett’, their preference for the ‘imperfett’ on production tasks signifies that this is easier

for them to learn and to use. Furthermore, this contrasts with the inverse preference by NSs for the ‘perfett’ on the same tasks.

Task 3 Task 4 Task 5 NNSs NSs NNSs NSs NNSs NSs 17 34 19 52 2 1 Table 9: Table of occurrences of the past tense on Tasks 3, 4, and 5 by NNSs and NSs

Several questions, therefore, arise from this minor exploratory survey of the interlanguage of NNSs of Maltese. These are:

1. What is the relationship between analysed and unanalysed knowledge in the interlanguage of the foreign learner of Maltese?

2. Why is the ‘imperfett’ preferred by NNSs in those scenarios where NSs prefer the ‘perfett’?

3. Do NNSs of Maltese find the ‘imperfett’ easier to learn and/or to use?

4. Should foreign learners of Maltese be taught the ‘imperfett’ before the ‘perfett’ (if it is, indeed, easier and quicker for them to learn)?

Further questions could be asked as to why the progressive and restricted habitual forms are avoided by both NSs and NNSs on production tasks, and why NNSs at an advanced stage of learning still have difficulty with them.

Conclusion

This was a small-scale investigation of the interlanguage of adult foreign learners of Maltese, and therefore I cannot attempt any definitive conclusions in answer to the research question. Several difficulties were faced in the course of data collection. First of all the number of adult foreign learners of Maltese is not big enough to warrant a large sample of subjects that could allow for statistical analysis. Furthermore, although initially eight learners were asked to participate, only three handed in the work as requested. All three participants had practised Maltese with native speakers for real communicative purposes over a number of years, and hence had reached an advanced level of proficiency. Therefore, the results cannot be generalised to the whole foreign learner population.

Notwithstanding its limitations, being the first of its kind this study has shown that there is a clear difference in the frequency of use by NSs and NNSs of the ‘perfett’ and ‘imperfett’ tense/aspect forms. When the learner data is compared to native speaker data it becomes clear that NSs opt for a substantially larger use of the past tense as shown in two of the three open

ended writing tasks. The fact that three foreign learners of Maltese had different native languages, and had learned Maltese through different routes (e.g. they did not follow the same taught course, and did not have the same teacher, but interacted with native speakers on a large scale) might also be considered as a valid argument supporting the hypothesis that this result is not random but is, indeed, related to their interlanguage.

Larsen-Freeman (2009) emphasises that a description of the interlanguage of the foreign learner is an essential starting point for proper pedagogy. In this case, as in Rea Dickens and Woods (1988), the issue lies in learning to make a specific choice between two structures, which might be correct but not equally appropriate when evaluated against native speaker norms. Furthermore, as Larsen-Freeman (2003) notes, learners may know the grammar rules explicitly, but they fail to apply them in communication.

A clear implication emerges for pedagogical grammar. Larsen-Feeman (2013) suggests that what the foreign learners need to be able to do is not simply transfer or implement their knowledge to a communicative situation, but to actually transform and adapt their knowledge. In this sense, the teaching material that is used in the L1 classroom would not be suitable for the foreign learner of Maltese. The L1 student comes to the grammar lesson already equipped with a full repertoire of language and the presentation of, for instance, an analysis of the ‘perfett’ before that of the ‘imperfett’ does not interfere with his academic development or with use of structures in communicative contexts. The foreign learner, on the other hand, requires a specific approach due to their interlanguage, and one that would lead them to transforming and adapting their knowledge. If our hypothesis resulting from this pilot project that the ‘imperfett’ is easier for the foreign learner to acquire and use, and that notwithstanding their analysed knowledge of the ‘perfett’ they would still opt for the ‘imperfett’ in situations where the native speaker opts for the ‘perfett’, is substantiated by further research, then pedagogical grammars of Maltese as a foreign language need to take this into account. From the perspective of the native speaker grammar it is also interesting to find out why NSs opt for the ‘perfett’ to a larger degree, and whether this choice has any relation with the fact that the past tense of the verb is considered as the base form. One might wish to explore, if at all possible, whether traditional grammar correctly models the native speaker’s mental representation, or whether the frequency of use of the past tense by NSs is a result of traditional teaching methods in the native speaker classroom.

References

Allen, J. P. B. (1974). Pedagogic Grammar. In Allen, J.P.B. & Corder, S.P. (eds.) Techniques in Applied Linguistics (59-92). Oxford: Oxford University Press.

Bialystok, E. & Bouchard Ryan, E. (1985). A metacognitive framework for the development of first and second language skills. In Forrest-Pressley, D.L., MacKinnon, G.E. & Waller, T.G. (eds.) Metacognition, Cognition and Human Performance. Vol. 1. (207-252). New York: Academic Press. Bialystok, E., McBride-Chang, C. & Luk, G. (2005). Bilingualism, language proficiency, and learning to read in two writing systems. Journal of Educational Psychology , 97, 580-590. Borg, A. J. (1981). A study of aspect in Maltese. Karoma Pub. Inc. USA: Ann Arbor. Borg, A. & Azzopardi-Alexander, M. (1997). Maltese. Londong: Routledge. Cohen, L., Manion, L., & Morrison, K. (2007). Research methods in education (6th^ Ed.). London: Routledge. Cook, V. (2014). How Important is grammar in acquiring and using a second language? In Cook, V. & Singleton, D. Key Topics in Second Language Acquisition. (55-72). Bristol: Multilingual Matters. Cook, V. & Singleton, D. (2014). Key topics in second language acquisition. Bristol: Multilingual Matters. Corder, S.P. (1967). The significance of learners’ errors. International Review of Applied Linguistics in Language Teaching, 5, 161-170. Corder, S.P. (1975). The language of second-language learners: the broader issues. Modern Language Journal , December 1975, 409-415. de Ward, J. (1981). Foreward. In Borg, J.A. A Study of Aspect in Maltese. Karoma Pub. Inc. USA: Ann Arbor, vii-viii. Dornyei, Z. & Csizer, K. (2012). How to design and analyze surveys in second language acquisition research. In Mackey, A. & Gass, S. M. (eds) Research Methods in Second Language Acquisition (74-94). London: Wiley-Blackwell. Eckman, F. R. (2012). Second language phonology. In Gass, S. M. & Mackey, A. (eds.) The Routledge Handbook of Second Language Acquisition (91-105). Oxon: Routledge. Ellis, N. (1994). Introduction. Implicit and explicit language learning – an overview. In Ellis, N. (ed.) Implicit and Explicit learning of languages (1-32). London, Academic Press. Friedman, D. A. (2012). How to collect and analyze qualitative data. In Mackey, A. & Gass, S. M. (eds) Research Methods in Second Language Acquisition. (180-200). London: Wiley-Blackwell. Granger, S. (2012). How to use foreign and second language learner corpora. In Mackey, A. & Gass, S. M. (eds) Research Methods in Second Language Acquisition (7-29). London: Wiley-Blackwell. Hulstijn, J.H. (1990) A comparison between the information-processing and the anlysis/control approaches to language learning. Applied Linguistics , 11 (1), 30-45. Hurford, J. (1994). Grammar. A student’s guide. Cambridge: Cambridge University Press. Ionin, T. (2012). Formal theory-based methodologies. In Mackey, A. & Gass, S.M. (eds.) Research Methods in Second Language Acquisition. London: WileyBlackwell. Joo, H-R. (2003). Second language learnability and the acquisition of the argument structure of English locative verbs by Korean speakers. Second Language Research , 19, 305-328. Khresheh, A. (2012). Investigating performance of the third-person singular – s as a learnt property through orally repeated practice based on Krashen’s Monitor

conditions of Time and Form focuses: The case of Jordanian Arab trainees translating Arabic into English as a foreign language. International Journal of English Linguistics 2 (2), 96-112. Krashen, S. (1985). The input hypothesis: Issues and implications. London: Longman. Lardiere, D. (2012). Linguistic approaches to second language morphosyntax. In Mackey, A. & Gass, S. (eds.) Research Methods in Second Language Acquisition (106-126). Oxford: Wiley Blackwell. Larson-Freeman, D. (2003). Teaching language: From grammar to grammaring. Boston, MA: Thomson/Heinle. Larson-Freeman, D. (2013). Transfer of learning transformed. Language Learning, 63 (1), 107-129. Mackey, A. & Gass, S. M. (2012). Introduction. In Mackey, A. & Gass, S. M. (eds) Research Methods in Second Language Acquisition (1-4). London: WileyBlackwell. Mackey, A. & Gass, S. M. (eds.) (2012). Research methods in second language acquisition. Oxford: Wiley Blackwell. Nesselhauf, N. (2004). Learner corpora and their potential in language teaching. In Sinclair, J. (ed.) How to use corpora in language teaching (125-152). Amsterdam, Netherlands: John Benjamins. Norris, J. M. & Ortega, L. (2012). Assessing learner knowledge. In Mackey, A. & Gass, S. (eds.) Research Methods in Second Language Acquisition. Oxford: Wiley Blackwell, 573-589. O’Grady, W., Dobrovosky, M. & Katamba, F. (1996). Comtemporary linguistics. London: Longman. Rea Dickins, P. & Woods, E. (1988). Some criteria for the development of communicative grammar tasks. TESOL Quarterly, 22 (4), 623-46. Richards, J., Platt, J., & Weber, H. (eds.) (1985). Longman dictionary of applied linguistics. Essex: Longman. Selinker, L. (1972). Interlanguage. International Review of Applied Linguistics , 10, 201231. Selinker, L. (1996). On the notion ‘IL competence’ in early Second Language Acquisition research: an aid to understanding some baffling current issues. In Brown, G., Malmkjaer, K. & Williams, J. (eds.) Performance and Competence in Second Language Acquisition (89-113) .Cambridge: Cambridge University Press. Selinker, L. & Lakshmanan, U. (2001). How do we know what we know? Why do we believe what we believe? Second Language Research , 17 ,(4),323-328. Sorace, A. (2005). Selective optionality in language development. In Cornips, L. and Corrigan, K.P. (eds.) Syntax and Variation, Reconciling the Biological and the Social (55-80.). Amsterdam, Netherlands: John Benjamins Sorace, A. & Filiaci, F. (2006). Anaphora resolution in near-native speakers of Italian. Second Language Research, 22 (3), 339-368. Widdowson, H. G. (2003). Defining issues in english language teaching. Oxford: Oxford University Press. Widdowson, H. G. (2005). Applied Linguistics, interdisciplinarity, and disparate realities. In Bruthiaux, P., Atkinson, D., Eggington, W.G., Grabe, W. & Ramanathan, V. (eds.) Directions in Applied Linguistics (12-25). Clevedon: Multilingual Matters.