2.3 Major findings in IL studies

The research in interlanguage has yielded substantive findings. We firmly believe that these early findings in IL studies are invaluable in producing indirect implications for second language learning and teaching. In this part, we will summarize some main findings in terms of three principles governing IL developments: systematic change in grammar, acquisition orders and developmental sequences, and influence of learners’ L1.

 

2.3.1 Systematic change in grammar

First，look at the following examples:

 

No look my card.

Don’t look my card. 

(Ellis, 1984a)

 

The above two examples are from Ellis’ study of a 10-year-old Portuguese child. This boy produced two negative utterances within minutes during the same game of word bingo, for the same purpose and in the same context, while he was addressing the same person and with similar amount of planning time. Learners such as the boy begin to use two more forms of grammar freely, that is, in an non-systematic way. Other examples are:

 

Giving me the book.

Give me the book. 

(Wagner-Gough, 1975)

 

Obviously, this speaker uses v-ing form and simple verb form to perform the same function of request. The random changes in grammar are called free variation. It characterizes a learner’s interlanguage at the early stage of second language learning. However, later on, it can be found that the changes in learner grammar are not random, but systematic. This is because different changes take place at different stages of development. Initially, it can be assumed that learners begin by acquiring a single form (such as simple form of a verb) and use it for a variety of functions (such as referring to future, present, and past time). Later, they acquire other verb forms but use them interchangeably with the simple form. For example, when learners first acquire the past tense form of a verb (for example, ‘painted’) they are likely to use it in free variation with the simple form of the verb (‘paint’). Shortly after that, they begin to use the forms systematically. That is, they use ‘painted’ in planned discourse and ‘paint’ in unplanned discourse. Finally, they eliminate the non-target form and use the target-language form to perform the same function as native speakers, using ‘painted’ consistently to refer to past time.

By systematic change, we mean that the change is rule-governed. That is, we may be able to explain and even predict when learners use one form and when another. But what are the rules? Ellis (2000) has summarized three key points to illustrate the contexts upon which changes in IL grammar depend:

 

1. Linguistic context

Some research findings show that interlanguage changes are affected by linguistic context.

These changes may occur at the level of phonology. An early study (Dickerson, 1975) investigating phonological variation by native Japanese speakers learning English found that the sound /r/ was more frequently used before a low vowel, such as /a/ than before a mid vowel, such as /I/, and more frequently before a mid vowel than before a high vowel, such as /i/ or /u/.

More interlanguage changes could be found at the level of morphology. One of the findings is that learners’ choice of past tense marker depends partially on the type of the verb, e.g., whether it is an event verb, an activity verb or a state verb. Learners find it easier to mark verbs for past tense if the verb refers to events (for example, ‘arrive’), somewhat more difficult to mark verbs that refer to activities (for example, ‘sleep’), and most difficult to mark verbs referring to states (for example, ‘seem’). The kind of verbs affects the kind of errors made by learners. With activity verbs, learners tend to use the progressive form instead of the past tense. With state verbs, they tend to use the simple form of the verb. For example:

 

He arrived at noon. (event verb)

After that we sleeping in the tent. (activity verb )

Last night everything seem very quiet and peaceful. (state verb)

 

A similar interesting finding in an investigation (Wolfram, 1989) was that Vietnamese learners of L2 English in the United States were more likely to manifest past tense marking on suppletive forms (such as go/went) than on replacive forms (such as make/made). This change may be caused by a ‘principle of perceptual saliency’ as suggested by Wolfram. That is, the more distant the past tense form is phonetically from the present tense form, the more likely it is to be more marked for the past tense.

Another finding is that learners may behave differently depending on whether or not an adverb of frequency (for example, ‘every day’ or ‘usually’) occurs with an activity verb. In a sentence which refers to past time but which does not have an adverb of frequency, learners are likely to use a progressive marker. For example, a learner may say John watching TV to mean “John watched TV all the time”. However, in a sentence with such an adverb, learners are more likely to use the base form of the verb. For example, one may say John usually watch TV every day instead of “John usually watched TV every day”. From these two examples, we can see one linguistic form can trigger the use of another form.

Evidence of interlanguage variation caused by linguistic context can also be found at the level of syntax. For example, the effects of linguistic context are also felt in learners’ use of the verb ‘to be’. Learners sometimes use full ‘be’ (such as ‘is’), sometimes contracted ‘be’ (such as ‘’s’), and sometimes omit ‘be’ entirely. Which one is used is also determined by the linguistic context. It is found that the choice of the three forms is related to the type of subject of a sentence. If the subject of the sentence is a pronoun, the learner may use the full ‘be’ form (such as ‘is’) or its contracted form (such as ‘’s’), as in the sentence He is not here or He’s not here. If the subject is a noun subject, ‘be’ is more likely to be omitted (For example, ‘Teacher not here’), as can be seen in the following illustration:

 

--- Is Tom here?

--- He is not here. / He’s not here.

……

--- Some didn’t come. Can you name a few?

--- Tom not here. Jack not here…

 

2. Situational context

Learners change their use of language according to the situation they are involved in. Just like native speakers, L2 learners are influenced by situational factors, and in particular, by their addressee. That is, they are sensitive to the person they are talking to. For example, in a interview, the interviewees, Chinese-Thai bilinguals adapt their speech by using more Thai phonological variants with the Thai interviewer, and more Chinese variants with the Chinese interviewer. Another aspect of the situation to which learners are sensitive is topic. It is found by a study (for example, Selinker and Douglas, 1985) that Polish ESL learners vary their learner language noticeably on a number of linguistic features according to whether the discourse is concerned with an everyday topic or a specialized, technical topic. Besides, the attitude which learners have towards a topic also affects their language behavior. If they see themselves as experts, for example, they may be more likely to interrupt.

The study of interlanguage variation is greatly influenced by the study of style shifting. For example, according to Tarone (1983), L2 learners possess a continuum of styles, ranging from careful style to vernacular style. The former reflects the kind of language in formal situations that require careful language use, while the latter is evident in informal situations in which there is more spontaneous language use. Each style has its linguistic norms, and learners change their style in accordance with the demands of the situation. Generally speaking, learners are more likely to use the correct target-language forms in formal contexts and non-target forms in informal contexts.

 

3. Psycholinguistic context

Another factor which accounts for the systematic change in IL grammar is the psycholinguistic context. There are some important factors. One of them is attention. If learners focus more attention to linguistic forms, accuracy will increase. Another important factor is planning time. This is related to whether learners have the opportunity to plan their speech production or not. Generally, a second language learner will be more careful in selecting grammatical forms and is more likely to produce correct speech when he has time to prepare what he is going to say. On the other hand, when he is making a spontaneous speech or having an online talk, more errors would occur since he does not have time to plan his speech. However, if the effort and time is required to plan the propositional content or to produce complex sentences, learners may be inhibited from attending to specific linguistic forms.

The above findings may provide some insight into the process of SLA. It reveals the way form-function relationships in IL evolve over time, partly reflecting patterns of variable use at an earlier point of time. They also have implications for language teaching and testing. The fact that ILs are shown to be partly systematic, even in those areas where they are variable, means that they are potentially susceptible to systematic change through instruction.

 

2.3.2 Acquisition order and sequence of acquisition

Interlanguages also exhibit common acquisition order and the sequence of acquisition. Acquisition order concerns whether learners acquire the grammatical structures of an L2 in a definite order. For example, do they learn a feature before another? The sequence of acquisition, however, concerns the learning of a particular grammatical feature. The relevant question is whether learners acquire a grammatical structure in a single step or proceed through a number of interim stages before they master the target structure.

 

1. The order of acquisition

In order to “discover” the order of acquisition, researchers choose a number of grammatical structures to study. They collect samples of learner language and identify how accurately each feature is used by different learners. Then they get an accuracy order by ranking the features according to how accurately each feature is used by learners. It is argued that the accuracy order must be the same as the order of acquisition, for the more accurately learners are able to use a particular feature the more likely they are to have acquired that feature early. Roger Brown’s (1973) morpheme order studies within the context of child language acquisition became the cornerstone of early work in SLA. He observed that there was a predictable order of acquisition of certain inflectional morphemes in English. Studies in SLA have also shown that there is a definite accuracy order and that this order remains more or less the same regardless of the learners’ mother tongue, their age, and whether or not they have received formal language instruction. The result of such research can be roughly described as three layers of order as shown below, the first layer of inflections being the easiest, while third layer being the most difficult to learn:

 

Progressive –ing; auxiliary be; plural –s

Articles and irregular past

 

Regular past; and 3^rd person –s

 

It should be noted here that the above research on the order of acquisition has been criticized on a number of grounds. Some say it is wrong to assume that the order of accuracy is equal to the order of acquisition. Others have shown that the order does change somewhat according to the learner’s first language. There may also be a misunderstanding that the research treats acquisition as if it is a process of accumulating linguistic structures, with one to be learnt before or after another. The truth might be that even the learning of the simplest structure goes through a process of gradual development. This can be shown by the study of sequence of acquisition.

 

2. Sequence of acquisition

The process of learning a particular grammatical structure is seen as a process which involves reconstruction or reorganization. A well-known learning pattern during the process of learning an L2 is called a U-shaped pattern. A U-shaped pattern reflects three or more stages of linguistic use. Lightbown’s study (1983) of the use of the –ing form from French learners of English showed a three-stage U-shaped learning sequence. At the beginning, learners produced error-free forms like He is taking a cake. At Stage two, learners appear to lose what they knew at the first stage. A typical utterance is He take a cake, which is different from the TL norm. The final stage looks just like stage one; the correct TL usage appears again. The whole process can be illustrated in Figure 2.1.

 

                   Stage 1(error-free)           Stage 3(error-free)

 

 

 

 

 

 

 

 

                          

                             Stage 2 (erroneous)

TIME

Figure 2.1 U-shaped learning pattern

 

Ellis (2000) cited another example to show how L2 learners acquire irregular past tense form ‘ate’ and discovered a five-stage U-shaped learning process. At Stage one, learners fail to mark the verb for past tense, using ‘eat’ in all cases. At Stage two, they begin to produce the correct form ‘ate’. At Stage three, however, learners overgeneralize the regular past tense form, and begin to use ‘eated’. Next, some learners produce the hybrid form ‘ated’, and finally at Stage 5, they master the correct irregular past tense form. The whole process is shown in Figure 2.2.

The sequences of learning are instructive because they reveal that the correct initial use of a form, such as Stage 2, does not always mean that the form is ‘acquired’. Learners who produce ‘eated’ and ‘ated’, in fact, are more advanced than those at stage 2 who produce ‘ate’. Indeed, learners follow a U-shaped course of development. This is so because learners reorganize their existing knowledge in order to accommodate new knowledge. Thus stage 2 and 3 arise when learners have begun to acquire the regular –ed form such as jumped. Forms like ‘eated’ and ‘ated’ show an overgeneralization of the regular –ed past tense. This is called restructuring. When they restructure the grammatical system, they may seem to slip back. But in fact, they are advancing on the way of learning.

 

 

                   Stage 1(eat)               Stage 5(ate)

 

 

 

             Stage 2(ate)              Stage4(ated)

 

 

 

 

                          

                             Stage 3 (eated)

TIME

Figure 2.2 U-shaped learning process of ‘ate’

 

2.3.3 L1 influence

   L1 influence has been a key issue in SLA for a long time. According to Contrastive Analysis Hypothesis (CAH), developed early in the 1950s and 1960s, it was widely believed that L1 played a decisive and negative role in SLA. This was called ‘interference’, and this interference could be predicted by systematically comparing and contrasting the Learner’s L1 and L2, looking at the points of difference between the two. However, this strong view of CAH has not been supported by research findings. The following quotations may enable us to see whether language transfer is important in forming interlanguages:

 

Interference, or native to target language transfer, plays such a small role in language learning performance. (Whiteman and Jackson, 1972:40)

Direct interference from the mother tongue is not a useful assumption. (George, 1972:45) 

Language background did not have a significant effect on the way ESL learners order English morphemes. (Larsen-Freeman, 1978:372)

 

Since the late 1970s, research on the role of the native language has taken on a different view. That is, the assumption is that one can view transfer as a creative process. At this time, researchers interested in SLA were less interested in a wholesale acceptance or rejection of the role of the native language. Rather, the emphasis was on the determination of how and when L1 influences SLA and on the explanations for the phenomenon. The term cross-linguistic influence is suggested to refer to the general influence of L1 on learners’ IL. Here we will discuss some relevant issues related to L1 influence.

 

1. Avoidance

When speaking or writing in a second language, a learner may often try to avoid using a difficult word or structure, and will use a simpler word or structure instead. This is called avoidance in SLA. For example, a learner who is not sure of the use of the relative clause in English (such as That’s the building where I live) may avoid using it and use two short sentences instead: That’s my building. I live there.

How does this occur? There is evidence in SLA research that native language may influence which structures a learner produces and which are not produced. However, the source of avoidance is disputable. It may be that the differences between L1 and L2 are the major source of avoidance. There is also evidence that the opposite occurs. That is, when great similarities exist between L1 and L2, the learner may doubt these similarities are real. Another view holds that avoidance has less to do with NL-TL differences, but rather is based on the complexity of the L2 structures in question. This is supported by a study conducted by Dagut and Laufer (1985). They found that Hebrew-speaking learners of English (Hebrew has no phrasal verbs) generally preferred the one-word equivalent of the phrasal verbs. For example, they use enter, remove, save, stop, disappoint, confuse instead of come in, take away, lay aside, shut off, let down, mix up.

Although L1-L2 similarity and L2 complexity have a role, the only factor that consistently predicts avoidance is the L1-L2 difference.

 

2. Different paths

The influence of L1 can also be felt in direction of interlanguage development. Let’s review a study done by Zobl (1982) first. In Zobl’s study, there were two subjects, a Chinese-speaking child and a Spanish-speaking child, who were acquiring the definite article the. The Chinese child tended to use this to serve the function of a definitizer. When even there is native speaker modeling of the definite article the, it is deleted or changed to this, as can be seen from the following data:

 

NSChinese-speaking learner

Is this airplane your brother’s?This airplane…

Show me the airplane.Show me airplane

Put it on the chair.Chair? This one?

Is this table dirty?Yes, this is dirty.

Table is dirty.

Whose bike is this?This … Edmond’s

You want to push the penI want to push pen.

                                  (Gass & Selinker, 2008:141)

 

However, from the data with the Spanish-speaking child, both this and the were frequent, as can be seen in the samples below:

 

Hey hey this. Here the toy.

The car.

Lookit this. Lookit this cowboy.

Here. This cowboy.

Shut the door.

 (Gass & Selinker, 2008:142)

 

It can be seen that the differences between these two children suggest that facts of native languages lead them down two different paths --- the Chinese child through a stage in which this occurs before the definite article, and the Spanish child to a starting point in which the definite article and the demonstrative this co-occur.

 

3. Overproduction

Another role played by NL in interlanguage development is that there is an influence of NL function to L2 form. This can be illustrated by Chinese and Japanese learners’ overproduction of extraposed and existential sentences such as:

 

It is very unfortunate that …

There is a small village. People in the village are …

 

It is claimed that these structures are used to carry the weight of a particular discourse function, even though the TL makes use of other forms for the same function. The relative hypothesis made by Schachter and Rutherford (1979) is that there is an influence of NL function to L2 form. The learners’ L1s, Chinese and Japanese are of the type that rely heavily on the concept of topic. Sentences are organized around a topic-comment structure, as in:

 

As for meat[topic], we don’t eat it anymore[comment].

 

Therefore, we can see that overproduction of certain structures by L2 learners is the result of influence of L1 function.