Describing and analysing the present state of knowledge in this field is the goal of Chapter 7 under the heading of "Learner Strategies": how does the learner work on the input to turn it into intake? Ellis differentiates between learning strategies and communication strategies. Among the varied strategies that the learner uses for learning, the most discussed are the following: overgeneralization, memorization, inferencing intra- and extralingual , formulaic speech memorizing unanalyzed chunks , and hypothesis formation and testing.
One of the puzzling contradictions the reader will run into is the role of the imitating adoption of prefabricated patterns within creative rule-learning in the Chomskyan sense. Some researchers hold that formulaic speech and rule-created speech are unrelated; others claim that the formulas are gradually? Reduction strategies have to do with the avoidance of the problem; achievement strategies have to do with compensation e.
The role of the very frequently observed strategy of simplification is still not clearly defined: it can be a general learning strategy falsifying incorrect hypotheses , a strategy of production, or a communication strategy. Realizing how important paraphrasing for the learners' growing communicative competence is, the foreign language teacher will certainly be confirmed in his conviction that it is important to explain words and expressions in the target language rather than to have recourse to El.
The concept of the "Universal Hypothesis", forming the theme of Chapter 8, is controversial. In my opinion the issue of how language learning difficulties can be seen in the light of universal language properties should be of the greatest interest to language teachers--even though there is arguably a great deal of uncertainty and speculation involved.
Here again Ll-acquisition research is basic to the understanding of SLA research. One of the issues to be discussed in connection with the universal hypothesis is whether developmental principles and strategies must be seen as a part of general cognitive procedures or as an independent linguistic faculty.
System 16 1 , pp. More recent research stresses the possibility that environmental factors may play an important role.
It might be that there is a continual interaction process between the Universal Grammar unfolding gradually in the child and developmental and environmental impulses activating hypothesis testing and restructuring.
But if we adopt the theory of Universal Grammar being itself subject to an innately specified developmental process, we can see one important difference between L1 and L2 acquisition, since the L2 learner must be assumed to have the full Universal Grammar at his disposal. With regard to L2 learning the crucial question concerns learning difficulty.
It has long been known that not all the linguistic differences between L1 and L2 result in learning difficulty. Can the psychological reality of linguistic learning difficulty be explored by defining whether linguistic properties of the L2 are universal or language-specific? One outcome of such considerations is the markedness theory. Properties that can be found in all or many? However, every language also contains elements that are not constrained by Universal Grammar: these comprise the periphery.
Core rules are u n m a r k e d and periphery rules are marked, i. Different stimuli produced different responses from a learner.
These responses could be haphazard in the sense that they could not be predicted , or they could be regular. The association of a particular response with a particular stimulus constituted a habit, and it was this type of regular behaviour that psychologists such as Watson or Skinner set out to investigate. They wanted to know how habits were established. Behaviourist psychologists attributed two important characteristics to habits. The first was that they were observable.
As Watson argued, the true basis for psychological enquiry existed only in objects that could be touched and actions that could be observed. Watson denied the existence of internal mental processes, dismissing them as 'superstition' and 'magic'.
The second noteworthy characteristic was that habits were automatic. That is, they were performed spontaneously without awareness and were difficult to eradicate unless environmental changes led to the extinction of the stimuli upon which they were built. The role of the -first language A habit was formed when a particular stimulus became regularly linked with a particular response.
There were various theories about how this association could take place. In the classical behaviourism of Watson, the stimulus was said to 'elicit' the response. That is, the presence of the stimulus called forth a response. If the stimulus occurred sufficiently frequently, the response became practised and therefore automatic. In the neo-behaviourism of Skinner a rather different account of how habits developed can be found.
Skinner played down the importance of the stimulus, on the grounds that it was not always possible to state what stimulus was responsible for a particular response. Instead he emphasized the consequences of the response. He argued that it was the behaviour that followed a response which reinforced it and thus helped to strengthen the association. The learning of a habit, then, could occur through imitation i.
Theories of habit formation were theories of learning in general. They could be and were applied to language learning. In L1 acquisition children were said to master their mother tongue by imitating utterances produced by adults and having their efforts at using language either rewarded or corrected.
In this way children were supposed to build up a knowledge of the patterns or habits that constituted the language they were trying to learn. It was also believed that SLA could proceed in a similar way. Imitation and reinforcement were the means by which the learner identified the stimulus-response associations that constituted the habits of the L2.
Language learning, first and second, was most successful when the task was broken down into a number of stimulus-response links, which could be systematically practised and mastered one at a time. Irrespective of whether the type of learning behaviour described by psychologists working within the frameworks provided by Watson and Skinner actually occurred, habit-formation theory dominated discussion of both first and second language acquisition up to the s. One of its major attractions was that it provided a theoretical account of how the learner's Ll intruded into the process of SLA.
In other words, in addition to offering a general picture of SLA as habit-formation, it also explained why the L2 learner made errors. According to behaviourist learning theory, old habits get in the way of learning new habits.
Where SLA is concerned, therefore, 'the grammatical apparatus programmed into the mind as the first language interferes. The notion of interference has a central place in behaviourist accounts of SLA. Interference was the result of what was called proactive inhibition. This is concerned with the way in which previous learning prevents or inhibits the learning of new habits. In SLA it works as follows. Where the first and second language share a meaning but express it in different ways, an error is likely to arise in the L2 because the learner will transfer the realization device from his first language into the second.
Learning a L2 involves developing new habits wherever the stimulus-response links of the L2 differ from those of the LL In order to develop these new habits, the learner has to overcome proactive inhibition. Of course, not all the patterns or habits of the Ll are different from those of the L2.
It is quite possible that the means for expressing a shared meaning are the same in the first and second language. For example, the idea of being cold, which is a shared meaning between English and German, is also expressed using identical formal devices - 'Ich bin kalt' is analogous with 'I am cold'. In cases such as this it is possible to transfer the means used to realize a given meaning in the Ll into the L2.
When this is possible, the only learning that has to take place is the discovery that the realization devices are the same in the two languages. The learner does not need to overcome proactive inhibition by mastering a different realization device.
Behaviourist learning theory predicts that transfer will take place from the first to the second language. Transfer will be negative when there is proactive inhibition. In this case errors will result. Transfer will be positive when the first and second language habits are the same. In this case no errors will occur.
Thus differences between the first and second language create learning difficulty which results in errors, while the similarities between the first and second language facilitate rapid and easy learning. In behaviourist accounts of SLA, errors were considered undesirable. They were evidence of non-learning, of the failure to overcome proactive inhibition.
Some language teaching theorists even suggested that there was a danger of errors becoming habits in their own right if they were tolerated. Brooks , for instance, wrote: 'Like sin, error is to be. However, as errors were the.
Errors, according to behaviourist theory, were the. But in either case. The role of the first language To this end attempts were made to predict when they would occur. By comparing the learner's native language with the target language, differences could be identified and used to predict areas of potential error. In this way classroom practice could be directed on the problem areas in order to help the learner overcome the negative effects of first language transfer.
Having examined the main principles of behaviourist learning theory as it was applied to SLA, it is time to consider the means that were used to predict potential errors.
These were contained in the procedure known as Contrastive Analysis. Contrastive Analysis. Contrastive Analysis was rooted in the practical need to teach a L2 in the most efficient way possible. As Lado , one of the prime movers of Contrastive Analysis, makes clear, 'The teacher who has made a comparison of the foreign language with the native language of the students will know better what the real problems are and can provide for teaching them'.
The origins of Contrastive Analysis, therefore, were pedagogic. This was reflected in comparisons of several pairs of languages by scholars in the United States, all directed at establishing the areas of learning difficulty that were likely to be experienced by English speakers learning other languages.
In addition to these pedagogically oriented studies, there have been a number of more theoretical contrastive studies carried out in Europe, some of which have not been concerned with SLA at all. Clearly Contrastive Analysis is an area of considerable theoretical interest for general linguistics, but I shall concern myself only with those studies that are concerned with SLA.
Contrastive Analysis had both a psychological aspect and a linguistic aspect. The psychological aspect was based on behaviourist learning theory, and the linguistic aspect, in the first place at least, on structuralist linguistics. The psychological aspect of Contrastive Analysis.
The psychological rationale takes the form of the Contrastive Analysis Hypothesis. This exists in a strong and a weak form Wardhaugh The strong form claims that all L2 errors can be predicted by identifying the differences between the target language and the learner's first language. As Lee notes, it stipulates that 'the prime cause, or even the sole cause, of difficulty and error in foreign language learning is interference coming from the learner's native language'.
The strong form of the hypothesis was common before research began to show that many of the errors produced by L2 learners could not be traced to the Ll see next section. The weak form of the hypothesis claims only to be diagnostic. A constrastive analysis can be used to identify which errors are the result of interference.
Thus according to the weak hypothesis, Contrastive Analysis needs to work hand in hand with an Error Analysis. First actual errors must be identified by analysing a corpus of learner language. Then a contrastive analysis can be used to establish which errors in the corpus can be put down to differences between the first and second language. Implicit in the weak version is the assumption that not all errors are the result of interference. The weak form claims a less powerful role for the L1 than the strong form of the hypothesis.
The strong form of the hypothesis has few supporters today. It is now evident that the Ll is not the sole and probably not even the prime cause of grammatical errors. Nevertheless, the weak form is not very satisfying.
It makes little sense to undertake a lengthy comparison of two languages simply to confirm that errors suspected of being interference errors are indeed so. As James points out, this is a 'pseudo procedure'. In order to hypothesize that the errors in a corpus are interference errors, a de facto contrastive analysis must have taken place.
It makes little sense to conduct a complicated contrastive analysis simply to confirm what a de facto analysis suggested. If Contrastive Analysis is to be worth while, it should be predictive. Diagnosis will then remain the job of Error Analysis. Ideally the psychological aspect of Contrastive Analysis should deal with the conditions under which interference takes place.
That is, it should account for instances when linguistic differences between the first and second languages lead to transfer errors and instances when they do not.
It is because it is not possible to predict or explain the presence or absence of transfer errors solely in terms of linguistic differences between the first and second languages that a psychological explanation is necessary.
What are the non-linguistic variables that help to determine whether and when interference occurs? One possible variable is the setting in which SLA takes place. Marton argues that whereas interference need not be a major factor in naturalistic SLA, it will always be present in classroom or foreign language learning.
In naturalistic SLA learners have the chance of extensive and intensive contacts with the target language, but in classroom SLA learners will always use their Ll between classes, and this strengthens proactive inhibition. The difference of opinion represented in the two quotations in the introduction t o this chapter can be explained in terms of this variable. Another variable may be the learner's stage of development. Taylor argues that there are quantitative differences in errors produced by elementary and intermediate students.
Whereas the former rely on transfer, the latter rely to a greater extent on overgeneralization of target. There is, however, no clearly articulated theory that explains how such variables as type of learning and stage of development affect the mechanisms of transfer. A major failing of Contrastive Analysis has been the lack of a well-developed psychological theory.
This has been one of the major sources of criticism of contrastive analysis. The linguistic aspect of Contrastive Analysis. A comparison of two languages can be carried out using any of several. Initially the model used was that of.
Bloomfield ; Fries This emphasized the importance of detailed 'scientific description' of languages based on. These categories were defined in formal terms and they were established inductively. The differences among languages were em-. The differences among languages are great enough to prevent our setting up any system of classification that would fit all languages.
Bloomfield It is clear that Contrastive Analysis and structuralist linguistics made. How can an effective comparison be executed if languages do not have any categories in common?
This problem was ignored, however, in the spate of contrastive studies that were carried out in the United States e. Stockwell and Bowen ; Stockwell, Bowen and Martin These studies compared languages from within the same language family e.
English and Spanish , so the problem of identifying a set of categories which were common to both languages was not acute. However, although for practical purposes the problem of establishing the linguistic basis for comparison could be overlooked, the theoretical problem remained.
Ideally Contrastive Analysis needs to be based on universal categories. Chomsky's theory of grammar proposed just such a model and.
However, most of the contrastive studies carried out have been based on surface structure characteristics, such as those described by the structuralists. The procedure followed was 1 description i. In 3 , comparison, the simplest procedure was to identify which aspects of the two languages were similar and which were different.
However, contrastive analysts soon realized that there were degrees of similarity and difference. Here are some of the possibilities that a comparison might reveal:.
The contracted form 'J'ai' in French is mirrored by the contracted form 'I've' in English. Where the L2 is English, German 'kennen' and 'wissen' coalesce into 'know'. In German, subordinate clauses require a different word order from main clauses, whereas in English the word order is the same in both clause types. In many African languages [lJ] occurs word-initially, but in English it only occurs word medially or finally e.
In Spanish, negation is preverbal 'Nose' , whereas in English it is postverbal 'I don't know'. In addition English negation involves the use of the auxiliary system, whereas Spanish negation does not. Where the L2 is French, English 'the' diverges into 'le' and 'la'. It is one thing to develop categories, such as 1 to 6 above, for classifying the ways in which two languages differ.
I t is quite another, however, to relate these linguistic differences to learning difficulty. Differences can be identified linguistically, but difficulty involves psychological considerations.
Stockwell, Bowen and Martin and Prator have proposed that linguistic differences can be arranged in a 'hierarchy of difficulty'. Prator, for example, suggests that 1 to 6 above are ordered from zero to greatest difficulty. This claim is not based, however, either on a psycholinguistic theory which explains why some differences create more learning difficulty than others, or on empirical research. It is based only on the conviction that the degree of linguistic difference corresponds to the degree of learning difficulty.
Most contrastive analyses have compared phonological systems, probably as a recognition of the role that the L1 plays in 'foreign' accents. However, the Contrastive Structure Series Stockwell, Bowen and Martin provided full-length studies of the contrastive syntax of the major European languages and English, while the s saw a number of studies in Europe see James for a list.
As Sridhar notes, there have been relatively few studies of vocabulary,. There are several problems concerning the linguistic aspect of Constrastive Analysis. One of these-the descriptive basis of the comparison-has already been briefly considered.
Other problems are considered in the next section. However, if the problems with Constrastive Analysis were only linguistic, they would be amenable to a linguistic solution. As the tools of contrastive linguistics grow more refined, the problems would recede. The major problems, however, have to do with the relationship between the psychological and the linguistic aspects of Contrastive Analysis. There is little point in comparing languages if learners make only limited use of their first languages in SLA.
The accuracy of prediction will always be open to doubt if Contrastive Analysis fails to specify the conditions that determine if and when interference takes place.
The 'hierarchy of difficulty' was an attempt to solve this problem linguistically, but unless the solution has psychological validity i. Contrastive Analysis constituted a hypothesis, and like all hypotheses was open to empirical investigation.
The real failure of the s was to rely on extrapolation from a general learning theory instead of getting down to the business of testing out theory by examining the language that learners produce. Criticisms of the Contrastive Analysis hypothesis. The criticisms that gathered force in the early s were of three major types.
First, there were the doubts concerning the ability of Contrastive Analysis to predict errors. These doubts arose when researchers began to examine language-learner language in depth.
Second, there were a number of theoretical criticisms regarding the feasibility of comparing languages and the methodology of Contrastive Analysis. Third, there were reservations about whether Contrastive Analysis had anything relevant to offer to language teaching. The 'crisis' in Contrastive Analysis was the result, therefore, of empirical, theoretical, and practical considerations.
I shall consider the major criticisms under these three headings. Empirical research and the predictability of errors. The existence of non-interference errors was always recognized, except by the staunchest of supporters of the Contrastive Analysis Hypothesis. Brooks , for instance, gives four causes of learner error. The issue was not, therefore, whether interference could account for a l l errors, but whether it could account for most.
Dulay and Burt , a set out to examine this issue empirically. They identified four types of error according to their psycholinguistic origins:. Dulay and Burt calculated the frequencies of these error types in the speech data of Spanish-speaking children learning English. They examined morphological features like past tense inflections.
After eliminating ambiguous errors they claimed that 85 per cent were developmental, 12 per cent unique, and only 3 per cent interference. On the basis of this and similar studies, Dulay and Burt argued that children do not organize a L2 on the basis of transfer or comparison with their L1, but rely on their ability to construct the L2 as an independent system, in much the same way as in L1 acquisition.
They suggested that interference may be a major factor only in phonology. Dulay and Burt's research constituted a powerful attack on the Contrastive Analysis Hypothesis. Clearly, if only 3 per cent of all learners' errors were the result of interference, then a comparison of the learner's native and target languages could not help to predict or explain very much about the process of SLA.
However, other research does not bear out Dulay and Burt's findings and there has been little agreement as to exactly what proportion of errors can be put down to transfer.
Table 2. Two points are worth making about Table 2. It does not reveal anything new in the field, but gives a large-scale view of different areas in second language acquisition. Having said that, we must not forget that this work is meant to serve as an introduction not as a close study. In my opinion, what make the text important and valuable are its accessibility to people other than linguists, and its meticulous analysis of the essentials of this rapidly expanding discipline.
Therefore, this book becomes an excellent tool for acquiring knowledge in SLA. References Corder, S. Pit, Error analysis and interlanguage. Oxford: Oxford University Press. Kellerman, Eric, In: S. Class and L. Selinker eds. Rowley, MA: Newbury House. Krashen, Stephen, The input hypothesis. New York: Longman.
Long, Michael. Maturational constraints on language development. In: Studies in Second Language Acquisition 12, White, Lydia, Video Audio icon An illustration of an audio speaker. Audio Software icon An illustration of a 3. Software Images icon An illustration of two photographs. Images Donate icon An illustration of a heart shape Donate Ellipses icon An illustration of text ellipses.