SLA: Second language acquisition research: is concerned with the different strategies followed by adult learners
The native speaker is exerting his verbal habits.
SLA is not concerned with the state (experimental psycholinguistics), it is concerned with the process language is learned.
Transfer: The learners transfers the parameters of the language he speaks natively in the target language context
We are concerned with universal strategies common to all foreign language learners.
I) Classical approach, major findings/ observation of SLA
There had been many shifts of interest.
There have been 5 major findings:
1- Language transfer/ interference
2- Developmental studies Patterns characterizing different stages in foreign language learning
3- Systematicity
4- Variability
5- Incompleteness
This domain is recent and is closely related to language pedagogy
Learning strategies are both systematic and variant.
Systematicity: developmental stages extracted from 1st language acquisition patterns
II) Modern approaches to the teaching of L2
Methods rest upon approach, design and procedures ‘syllabus’
Approach: to language (externalist or internalist), focus on the system seen from outside or the speaker’s knowledge about the language. It also includes your approach to learning (behavioristic explanation ‘habit formation process) (a cognitive activity which has nothing to do with behaviorism)
L1 studies: 2 types: longitudinal and cross sectional studies
Longitudinal studies were undertaken to highlight the process whereby L1 is acquired. They are undertaken over a period of time. You observe 1 subject or two
Cross-sectional studies of L1 development where the objective is to describe the grammatical system of children at a single time (large number of children at a single time)
L1 studies whether longitudinal or cross-sectional inform research in L2 acquisition. They provide it with methodological procedures. They also led to many theoretical issues; for example, whether L2 acquisition was determined by environmental (input given in the foreign language classroom) or innate factors (learner’s strategies).
SLA relies on such investigations and it borrowed methods used by L1 researchers. As a matter of facts, SLA focused on the same set of grammatical features which were considered in L1 studies, for example, the way questions/ negation/ passive voice is formed.
Rod Ellis identifies a set of issues for which SLA researchers have sought answers. These are the questions SLA researchers must answer:
1- What do 2nd language learners acquire? (The precise intake) The language acquired by the learner in his processing of the Target language input. The language ‘interlanguage’.
2nd language learners produce a language displaying obvious deviancies of the syllabus we have taught. It is the job of the SLA researcher to identify these deviancies and their description and to describe the rules which underline their production. Describe the grammar of mistakes to classify them. This is called error typology. The ultimate objective is to determine the Systematicity of the language.
2- Strategies followed by the learner in order to assimilate the structure of the 2nd language. The learners’ learning strategies
How do learners acquire a 2nd language? We have to explain this language and identify the cause of their major errors and their regularity.
The learner resorts to his L1 as his reference system. The learner resorts to his L1 in order to communicate: this has nothing to do with assimilation. Of course, those general learning strategies, the role of universal properties of language.
a- The role of social intuitions in which learning takes place and of course the process whereby the learners’ input leads to the learners’ output
b- The role of internal factors: those mental processes used to convert the input into knowledge: those cognitive strategies enabling the learner to convert the input into actual knowledge .Hence the difference between the actual systemization of the material question.
There is a difference between communicative and learning strategies. These variables combine together to determine the learners’ “interlanguage” (Selinker) Corder: “transitional competence” Nemser: “set of approximate systems”
3- What differences are there in the way in which individual learners acquire a 2nd language? This area in SLA is called “learning styles”
This is called variability. Learners’ vary in their rate of learning and this is determined by many factors, factors that contribute that variation. It is the objective of SLA to determine the precise factors which contribute to variation. Individual language learners vs universal learning processes. Among the obvious factors: we can mention: motivation, intelligence, interest, cultural background.
4- What effects does instruction have on our second language acquisition?
We know that learning does not depend on teaching at the same time we are in a foreign language classroom where the “total teaching operation” is independent of the learners’ cognitive learning strategies. We have a sort of opposition between naturalistic learning and instructed learning. The identity hypothesis sees no difference between SLA and 1st language acquisition. They recommend a particular teaching method called the silent way. They think that this approach gives the adult learner to process the input to which he is exposed naturally. VS instructive learning whose major objective is not to determine learning but to improve, facilitate, to expedite, learning.
The 1st findings of SLA
1) Transfer:
Behaviorist theories in the 2 decades following WWII claimed that the major difficulty to learning was interference from prior knowledge whether this relates to the learners’ native tongue or other languages. This is called “Proactive Inhibition”.
(Knowledge of L1 -> processing the TL input)
They claim that those linguistic patters which are similar will help the learner assimilate the TL thru positive transfer.
(Similar Patterns -> Positive transfer Eg: Le livre que j’ai acheté -> The book that I bought)
(Different Patterns -> Negative transfer Eg: الكتاب الذي اشتريته أمس -> *The book that I bought it yesterday).
ð This is the behaviorist explanation of learning. Behaviorists think that that learning is a simple process of habit formation which is based on a stimulus response bond. It is also called the “Habit Formation Theory”
ð Chomsky rejects this view of language learning: language is a creative process of a cognitive sort. Chomsky reviewed the works of the behaviorist Skinner and managed to show the inadequacies of Skinner’s theories. Among the strong arguments:
The child is exposed to a flow of speech not a set of rules in 1st language. The rules are not obvious. Owing to his Language Acquisition Device LAD he manages to construct a mental representation of the rules governing that language. The child, this way, manages to process novel sentences. This is creativity. He showed the fallacies of Skinner’s Verbal behavior.
* The book that I bought it tomorrow
- Externalists psychology
| Approach to language | language learning |
| Language as an inventory | Habit formation, which rests upon the stimulus response bond The CAH in its strong claim |
ð An obvious instance of interference: the learner has transferred the habits of his native tongue without being aware of it.
¹
- Internalist psychology
| Approach to language | language learning |
| LAD, UG | conceived of in terms of a creative construction model Error Analysis |
This mistake is an instance of an overgeneralization claiming that (the sequence I bought it is possible) the learner has overgeneralized the pattern over other patterns that do not support it.
I) Types of Transfer:
- Outright transfer (when the physical item is present)
The percentage of interference errors reported by various studies of L2 English grammar: learners who are studying English as a foreign language
| Studies | % of interference | Type of learner |
| Granberg (1979) Has analyzed errors | 36% were due to interference The remaining are of mentalistic nature | 1st language is German. They are adults and are advanced learners. |
| George (1972) | 33% of errors due to interference | Mixed 1st languages. Learners with different linguistic backgrounds. Adult graduate learners |
| Dulay & Burt (1973)(the proponents of identity hypothesis. 2nd language learner follow the same cognitive strategies as the child in 1st Language acquisition) | 3% of errors due to transfer | 1st language is Spanish. They are children of mixed levels. |
| Tran-Chi-Chau (1975) | 51% due to interference | His informants: 1st language is Chinese, adults, mixed levels. |
| Flick (1980) | 31% due to transfer | 1st language Spanish, adult learners, mixed levels. |
| Lott (1983) | 50% of errors due to interference | 1st language Italian, adult learners, students at university. |
We have a certain typology in transfer:
- Avoidance: Learners in general avoid using L2 structures. They avoid them as they find them difficult to assimilate. They are difficult because they are different of their counterparts of their native tongue. The effects of L1 are obvious in this type of interference not in what learners do but in what they do not do. We can explain the learner’s difficulty in terms of his linguistic background. In general avoidance is complex and its identification is not easy. In this context the psycholinguist Kellerman (1992) distinguishes 3 types of avoidance:
a- Avoidance occurs when learners know or anticipate that there is a problem and they have at least some sketchy idea of what the target form is like.
b- It occurs when the learner knows what the target is but find it too difficult to use in the particular circumstances as for example in free flowing conversation.
c- Learners know what to say and how to say it but are unwilling to actually say it because it is gives them the feeling that they are opposing their cultural system and their linguistic reflexes. The target form sounds strange to them.
- Over-use or over-indulgence (Levenstone)
It can occur as a result of intralingual (vs interlingual) processes such as overgeneralization. It can also result from transfer often as a consequence of the avoidance or underproduction of difficult patterns
Eg: Walk -> Walked bark -> barked go -> * goed
ð The 3 types of the learner’s reference to his native tongue show the necessity of considering the multiple ways in which L1 influence can exert itself. These types of influence which show the learner’s reference explain the emergence of the Contrastive Analysis Hypothesis CAH;
The Contrastive Analysis Hypothesis
It was first formulated by the linguist Lado (1957). It was based on the following principle: “the student who comes into contact with a foreign language will find some features of it quite easy and others extremely difficult. Those elements that are similar to his native tongue will be simple for him; and those elements that are different will be difficult.” Lado’s CAH shows that it is necessary to consider both systems in preparing teaching material.
We have interference which affects the long aspect (the system) and we have interference that affects performance. We are concerned with the 2nd type, with the learner performing. Lexical, syntactic borrowing: both types concern the system and not the learner. Performance is individual. The other type does not affect the individual speaker learner.
The Contrastive Analyst seeks to identify the differences of different language backgrounds. It is based on structuralist descriptions of the two languages and the procedure involves the following steps:
- A formal description of the two systems.
- Selection: Items of the two languages are chosen for detailed comparison. Eg: the subordinate clause in the two systems
- Comparison: entailing the identification of areas of differences and similarities
- Prediction: the CAH rests upon this step. Prediction: determining which areas are likely to cause errors. This is called the predictive power of the CAH => The strong version of the CAH -> it predicts the learner’s errors and claims that all L2 errors could be predicted.
There is a weaker version of the CAH which claims that only some errors can be traceable to transfer and in this weaker version the CAH serves to explain certain errors but not to predict. This version shows that the CAH needs to be used hand in hand with error analysis.
2) Developmental patterns
The results of the comparative studies that were undertaken in the 70s and 80s have urged researchers to recognize the need to consider the imparity of learner language in order to get insights about the system of rules or inter-languages that learners construct at different stages of development which are supposed to be similar to those developmental stages thru which a child goes in acquiring his native tongue. As a matter of facts, the description of how learner language (transitional competence) develops over time is important for a better understanding of 2nd language acquisition. It allows the researchers and the syllabus designer to adjust the teaching activity according to the learners developing grammar (pedagogical implications). One of the most powerful ideas that emerged from this type of investigation was that L2 acquisition proceeds in a regular systematic way. L2 learners and particularly adult learners have the capacity to engage in planned (vs. unplanned) language use by paying special attention to the language forms (vs. content) they choose, i.e. by using explicit (type of knowledge gained thru explicit analysis of the rules of the target language) (vs. implicit “makes part of the learner’s own cognitive strategies to have a mental representation of the rules”) knowledge of grammatical rules.
Unplanned and planned language use display different features and the idea of developmental patterns (staged development) is based on unplanned language use i.e. the learner’s transitional competence. We still do not know whether the concept of stage development (unplanned use) is applicable in planned language use.
The early stages of L2 acquisition in naturalistic settings are often characterized by a silent period, by the use of formulaic speech and by structural and semantic simplification. These are the specificities of the early stages of L2 acquisition in naturalist setting. These characteristics are also found in L1 acquisition.
ü The silent period: In 1st language acquisition, this period is necessary for the young child to discover what language is and what it does. In L2 acquisition the silent period is not obligatory as the learner already has the experience of language. And yet, many foreign language learners opt for a silent period despite the fact that they already know what language is and what it does. Of course, not all learners go thru a silent period. Classroom learners, for example, are obliged to speak. They have no alternative. But even when production in the foreign language classroom is not required some learners opt for it. Here the question arises as to why some learners opt for a silent period while others do not. The researcher Saville Troike suggests one interpretation. She suggests that the reason may reside in differences in the learners’ social and cognitive orientations. In this context, she distinguishes other-directed learners and inner-directed learners. The former, approaches language as an interpersonal social task with a predominant focus on the message they wish to convey i.e. they pay attention to the communicative aspect of language rather than the internal code governing that language. While the latter approaches language learning as an intrapersonal task with a predominant focus on the language code, i.e. on the language system (the rules of the target language). She suggests that while other-directed learners do not typically go thru a silent period, inner-directed learners of course do. In this respect they are similar to children acquiring their first language. There is some disagreement regarding this issue. Another interpretation is suggested by Stephen Krashen (1982). He argues that the silent period provides an opportunity for the learner to build up competence thru listening. According to this view, speaking ability emerges naturally after enough competence has been developed thru listening.
ü Formulaic speech consists of expressions which are learnt as unanalyzable wholes and employed on particular occasions. In this context, researchers distinguish routines and patterns to refer respectively to refer to whole utterances as memorized chunks and to utterances that are only partially unanalyzed and have one or more open slots. The psycholinguist Rod Ellis (1984) suggests that formulaic speech can consist of entire scripts such as greeting sequences which the learner can memorize because they are fixed and predictable. They operate as a set of stimuli that correspond to different situations.
ü Structural and semantic simplification: In comparison with formulaic speech the learner’s early creative utterances are typically truncated consisting of just 1 or 2 words with both grammatical morphemes and lexical items missing. These utterances show a strong similarity with those found in pidgin languages indicating that both structural and semantic simplifications are taking place. Structural simplification involves the omission of grammatical factors such as auxiliary verbs, plural ending… whereas semantic simplification involves the omission of lexical items. Both structural and semantic simplifications occur either because learners have not yet acquired the necessary linguistic forms (ignorance of the rules) or because they are unable to access them and put them in practice in production of specific utterance (in certain communicative and/or grammatical contexts). In this respect Pitt Corder (1981) rejects the concept of simplification in such utterances.
“If in the process of 1st or 2nd language acquisition the learner demonstrates that he is using a simple grammar or code as is well attested then he has not arrived at that code or grammar by a process of simplification of the TL code. In other words, you cannot simplify what you do not possess.”
1) Staged development and systematicity:
The acquisition of 2nd languages is typically staged. From this initial state grammar the L2 learners construct, they go thru stages of development towards the target language. In this respect L2 learners are similar to L1 learners who also typically go thru stages of development although more often than not starting from a different initial state grammar. A number of studies in the 70s and early 80s (comparative studies) found consistent patterns in the development of accuracy on grammatical morphology in English across a range of L2 learners (from different language backgrounds, of different ages and learning English under different conditions). Some example of the types of morphological phenomena:
The first morpheme to appear is the progressive then the morpheme expressing plurality, past regular and then past irregular, the possessive and the last one is the 3rd person singular. Dulay and Burt first noticed that L2 learners systematically produce some grammatical morphemes more accurately than others across 3 groups of L1 Spanish speaking 5 to 8 year old in the US. They subsequently replicated their findings with a group of 6 to 8 year old Cantonese speaking learners of English Finding a similar accuracy order. Krashen found the same order in a group of adult L2 learners of English from different L1 backgrounds. The psycholinguist Makino studies 777 adolescent Japanese classroom learners of English in Japan not only found a high correlation between his subjects and those of Dulay and Burt but also found that there was no relationship between this natural order and the order in which the Japanese learners we taught the grammatical morphemes in the classroom. This natural order is not determined by the inputè it does not depend on instructions.
What is important to retain from these examples: is that there is systematic development which is independent either of the 1st language the learner speaks or the type of input a learner receives. Given the mixed results available from comparative studies of L1 and L2 acquisition (L1 doesn’t seem to affect the learner’s developing grammar) It is not surprising to find different conclusion regarding L1 = L2 acquisition also called the ‘identity hypothesis’. What constitutes evidence for a developmental pattern has rarely been addressed in explicit manner but the following criteria seem to have been applied by most researchers.
- 1st: developmental patterns can be established by looking at, either the order in which different target structures are acquired or the sequences of stages thru which the learner goes to attain ultimate target competence.
- 2nd: In the case of transitional structures, a stage consists of a period during which learners use a particular form or structure systematically although not necessarily to the exclusion of other forms and structures.
- 3rd: The forms and structures that a learner produces at different points during the process of L2 acquisition can be ordered in such way that one form or structure always precedes another which explains staged development.
- 4th: Learners progress step by step along an order or sequence mastering one particular structure whether target language or transitional before another.
- 5th: Strong evidence for developmental patterns occurs when it can be shown that an order or a sequence is universal. (that the order or sequence does not depend on the learner’s particular L1 or on any other variable whether external ‘formal instructions’ or internal ‘cognitive reasons’) Weaker evidence is found if it shown that an order or a sequence applies only to specific L2 learners and/or to specific groups of learners. It seems that there is a systematic (because it is completely independent of the learner’s and of the selected data of L2) L1development. This independence led to the conclusion that learners acquire L2 the way a child acquires L1.
Learner errors and Error Analysis:
The learner’s difficulties were all thought as due to his L1. Transfer theory was the prevailing thesis. With the developmental stages and Systematicity there had been a shift of interest from contrastive studies to a cognitive explanation of the learner’s strategies. The study of errors is carried out by means of error analysis. It was Corder who first noted that errors could be significant in 3 ways:
1. They provide the teacher with information about how much the learner has learnt.
2. They provide the researcher with evidence of how language is learnt. (the process of acquisition)
3. They serve as devices by which the learner discovers the rules of the target language. They inform us about the learners’ cognitive strategies in processing the input of the target language.
The first point reflects the traditional role of EA (Error Analysis). The second point provides a new role that is very important for the researcher as it can explain and account for the
third point namely the process whereby the L2 is acquired. Corder suggests 5 steps in EA research
1. Collection of a sample of learner language (a corpus)
2. Identification of errors
3. Description of errors
4. Explanation of errors
5. Evaluation of errors
1. Collection of a sample of learner language (a corpus):
There are 3 main types according to the size of the sample.
A massive sample which consists of collecting several samples of language use or output from a large number of learners in order to establish a comprehensive list of errors representative of the entire population (of L2 learners).
The 2nd type is a specific sample consisting of 1 sample of language use collected from a limited number of learners.
The 3rd is an incidental sample involving only one sample of language use produced by a single learner.
The learners’ errors can be due to a variety of factors, that’s why it is important to collect well defined samples of learner language so that clear statements can be made regarding what kinds of error the learners make and under what conditions. Decisions also need to be made regarding the manner in which the samples are to be collected. An important distinction is whether the learner language reflects natural spontaneous language use or is elicited (the teacher leads the learner to produce the error) in some way.
Natural sample are generally preferred as they can inform us about the learner’s own strategies. Elicitation however, is not to be confused with testing which is concerned with measuring the learner’s knowledge to evaluate rather than to describe competence. In this context Corder identifies two kinds of elicitation:
- Clinical elicitation
It consists of getting the informant to produce data or output of any sort by means of general interview in general or by asking learners to write an essay or a paragraph (just to produce)
- Experimental elicitation
It consists of using special instruments. The researcher leads the learner to produce a specific output involving the linguistic feature the linguist wishes to investigate.
Another problem is whether the samples are collected cross-sectionally (at a single point in time) or longitudinally (at successive stages of development).
2. Identification of errors:
In the learner output the researcher identifies some errors providing him with insight about the learner’s knowledge at that stage. Corder makes a difference between errors and mistakes.
- Mistakes:
Mistakes are performance phenomena; they are not systematic which means that they are not due to the learner’s ignorance of the rules. They are similar to the mistakes made by the native speaker. (Chomsky’s ideal speaker/ hearer)
- Errors:
The error is systematic; it shows the learner’s way of systemizing that feature. The learner does not know the rule which governs that utterance.
Another problem considered by Corder is whether the error is overt or covert: surface or hidden. He claims that an overt error is easy to identify as it represents a clear deviation in form. A covert error occurs in utterances which are superficially well formed but this doesn’t express what the learner actually intended to say. Corder mentioned a typical error:
E.g. It was stopped.
This utterance is apparently grammatical until it becomes clear that ‘it’ refers to the wind. Furthermore, a superficially correct utterance may only be correct by chance. It depends on the context (communicative & grammatical). The existence of covert errors lead Corder to assert that every sentence is to be regarded as idiosyncratic until shown to be otherwise. Another question concerns whether the analysis should examine only deviations in correctness or also deviations in appropriateness. Correctness involves rules of usage and appropriateness involves rules of use (linguistic competence ‘usage’; communicative competence ‘use’)
- The description of errors:
The description of learner’s errors involves a systematic comparison of the learner’s idiosyncratic utterances with a reconstruction of those utterances in the target language; The researcher is required to focus on the learner’s errors to interpret them in the context and that context enables him to reconstruct it in the target language. Some researchers think that there is a difference between the description and explanation of errors. Dulay Burt and Krashen for example stress the need for descriptive taxonomies (classification) of errors that focus only on observable surface features of errors as a basis for subsequent explanation. The simplest type of descriptive classification is based on:
- linguistic categories
For example, the structure of the verbal system, the prepositional system, the auxiliary system in English and each general category is then broken down into further sub-systems (with the verbal system there is the modal system). An alternative to a linguistic classification of errors is to use
- Surface strategy taxonomy
This clarifies the ways surface structures are altered by means of such operations as omissions, additions, redundancies, and regularizations. The same researchers maintain that such an approach is reliable as it allows an identification of the cognitive processes underlying the learner’s reconstruction of the target language. But this seems doubtful as it presupposes that learners operate on the surface structures of the target language rather than create their own idiosyncratic or transitional grammar.
In this respect Corder’s model for describing errors is more promising as he distinguishes 3 types of errors according to Systematicity:
- Pre-systematic errors
They occur when the learner is unaware of the existence of a particular rule in TL. These errors are random. Some researchers call them ‘unique goofs’. The learner doesn’t seem to produce the output according to a particular rule.
- Systematic errors
This doesn’t mean the learner has applied the rule, but he has systematized the output without following the correct rules. The learner has discovered the rule but it’s not the right one. (For example when he overgeneralizes E.g. * He goed) The learner production is not random.
- Post-systematic
Occur when the learner knows the correct TL rule but he fails to apply it consistently and in order to identify these different kinds of error, it is necessary to interview the learner: to ask him why he made this or that mistake to account for the mistake.
Thus, pre-systematic errors are when the learner cannot give any account of why a particular form is chosen. Type 2 (systematic errors) happens when the learner is unable to correct the errors but can provide the researcher with the rule he has applied when making this mistake. Type 3 (post-systematic errors) occurs when the learner can explain the TL rule that is normally used but that he fails to apply.
4. Explanation of errors
It is different of the description of errors. The explanation of errors: we explain the cause of the error produced by the learner. Explanation is concerned with establishing the source of the error. It enables the researcher to establish the processes which are responsible for L2 acquisition. The psycholinguist Taylor explains that the error source maybe psycholinguistic, sociolinguistic, or epistemic (relating to general knowledge) or may reside in the discourse structure.
- Psycholinguistic sources concern the nature of the L3 knowledge system and the difficulties learners have in manipulating/using/processing/producing them.
- Sociolinguistic sources involve such variables as the learner’s ability to adjust their language in accordance with the social context.
- Epistemic sources concern the learner’s lack of general (world) knowledge.
- Discourse sources involve problems in the organization of utterances into a coherent text.
SLA is chiefly concerned with psycholinguistic sources. It ignores other sources. Investigators have identified many sources of competence errors. Richards for example identifies three main causes:
- Interference errors: i.e. transfer of linguistic properties from a speaker’s L1 into the L2. Transfer can affect all linguistic levels: the phonic and phonological levels, syntax, morphology, lexicon and discourse.
- Intra-lingual errors: “Intra-lingual errors reflect the general characteristics of rule learning such as faulty-generalization, incomplete application of rules and failure to learn conditions under which rules apply” Richards
- Developmental errors: these occur when the learner attempts to make hypotheses about the TL on the basis of limited experience/exposure.
Many investigators find it reasonable to classify the 3rd type as intra-lingual as they do not involve other systems. Interference errors can be further subdivided. Lott (1993) distinguishes 3 types of interference errors:
- Overextension of analogy (formal similarity): occurs when the learner misuses an item because it shares features of an item in the L1; for example, Italian learners are reported to use process to mean ‘trial’ because Italian ‘Processo’ has this meaning or French learners who use actually for ‘at present’.
- Transfer of structure or (outright/typical/surface transfer) occurs when the learner uses some L1 features (phonological, lexical, pragmatic…) rather than that of the TL.
- Inter-lingual & intra-lingual errors: occur when a particular distinction does not exist in the L1. For example, the use of make instead of do, by Italian learners because the make/do distinction does not exist in Italian. It is this third category that has caused so many of the problems in determining an error to transfer or intralingual.
Intra-lingual errors are also further subdivided and Richards 1971 distinguishes:
- Overgeneralization errors: occur when the learner creates a deviant structure on the basis of other structures in the TL. It generally involves the creation of one deviant structure in place of two TL structures. For example, the deviant structure: he may comes/ he can tries; where English allows: he may come and he comes.
- Ignorance of rules restrictions: involves the application of rules to contexts where they do not apply. For example: he made me to rest. Of course thru extension of the patter found with the majority of verbs that take infinitival complements as: he asked/invited/forced me to do something. This is due to the learner’s ignorance of the fact that to rest does not allow this type of compliment.
- Incomplete application of rules: Involves a failure to fully develop a structure; thus, learners of L2 English have been observed to use declarative word order in questions. This type of errors is often called “errors of transitional competence” showing the existence of the learner’s developing grammar.
Attempts to explain errors in learner language can be summarized in the following conclusions:
- 1st, a large number of the errors that learners produce are intra-lingual in origin rather than transfer; however, the precise proportion of the kinds of errors varies considerably from study to study.
- 2nd according to Taylor, learners at an elementary level produce more transfer errors than learners at an intermediate or advanced level.
- 3rd the proportions of transfer and intra-lingual errors vary in accordance with the task used to elicit samples of learner language (make the learner produce); thus, for example, translation tasks tend to result in more transfer errors than tasks that call for free composition.
- 4th Transfer errors are more common in the phonological and lexical levels of language than in the grammatical level. This being so, some areas of grammar acquisition are more likely to be influenced by the learner’s L1 than others. Graubreg found that interference/transfer accounted for 25% of the lexical errors produced by adult German learners of L2 English. 10% only of their syntactic errors are due to transfer and none of their morphological errors.
- 5th transfer errors are more common in adult learners than in child learners. For example, White found that 21% of the errors made by adult Spanish learners of English were due to transfer.
- 6th errors can have more than one source. For example, the no + verb errors as in (no look my car) is universal suggesting an intra-lingual explanation. But Spanish learners of L2 English have been reported to make this error more frequently and for a longer period of time suggesting that the L1 pattern for negatives (No + verb) has also an influence.
Remarks concerning the Contrastive Analysis Hypothesis:
Transfer (CAH) ¹ staged development Systematicity (Error Analysis)
It was an enterprise built in the 1950s and 1960s on the twin theories of structural linguistics and behaviorist psychology. Structural linguistics sought to provide detailed linguistic descriptions of particular languages from a collection of utterances produced by native speakers (the corpus observed). The rationale for this corpus-based approach is derived from the notion of signs namely that everything had to be demonstrated in terms of the visible behavior of whatever was being analyzed. Behaviorist psychology explains that the ability to behave verbally (or to acquire a 1st language) in terms of a set of habits that had been acquired by linking language forms with meanings thru reinforcement and reward (conditioning). The meaning of the two approaches (structural linguistics and behaviorist psychology) in researchers thinking about SLA gave rise to the CAH. In the study of foreign language learning the CAH aims at identifying those points of structural similarity and difference between the 1st language and the TL and it assumes that points of difference will constitute areas of potential difficulty. The main objective of the CA approach is to predict the specific difficulties of the learner and to provide teaching materials and methods to facilitate learning. The idea was that if acquisition of the L1 involved the formation of a set of habits, then the same process must also be involved in SLA with the difference that some of the habits appropriate to the L2 will already have been acquired in the L1. It follows that the basic information required to organize the learning of an L2 would be to know which habits were the same and which were different. This would allow a concentration on those which were different. It would simply remain to find ways and means of changing those habits which were different so that new appropriate L2 habits would be acquired and the purpose of CA was to compare the structures of languages and to identify the differences in general so that foreign language teachers and course designers could focus the content of their classes on those structures where the language differed and used methods of reinforcement and reward to change those habits in the L2.
3) Variability
2nd language learners seem to be systematic and follow well defined stage development but the learner-language show significant variability, in other words, errors are not systematically made. Sometimes the learner makes error X which can be structural, lexical or showing inadequate word order and sometimes he does not make it which shows the identification of a precise stage of development and this doesn’t mean that the learner will make use of a single form or pattern. In other words there is no consistency but a preference. This shows a theoretical contradiction: you claim that learners are systematic and they are variable at the same time! For some, this contradiction can be dispelled and they suggest 3 explanatory approaches to explain this contradiction:
- The first perspective is called the Chomskyan perspective. In this perspective the learners’ collective common competence, (systematicity) variation is considered as a feature of performance which does not affect competence. We have to distinguish between what learners know and what they do. That’s why this called the homogenous competence paradigm which of course suggests 3 types of idealization/ abstractions of competence from performance: (underlying sentences can be derived from concrete sentences)
· Regularization: entails the theoretical elimination of variation related to the speech disfluencies (Chomsky’s idealization of the speaker/hearer) Eg. False starts, hesitations, focus on communication at the expense of form; such phenomena are frequent in our everyday use of language.
· Standardization: or theoretical elimination of variation related to different dialect of the language; a discrepancy between what the speaker knows and what he produces.
· Decontextualization: or the elimination of variation related to the use of language in different social contexts/ situations. Within the framework of the homogenous explanatory solutions.
Researchers can gain access to invariable data (competence) they can have information about the learner’s competence, intuitions (what they THINK is correct in L2 rather than their actual utterances)
- The second approach is the socio-linguistic approach: its objective is to study language in relation to social context. It accounts for the variation which is related to different varieties of English for example. The type of variation related to differences in terms of general social factors, class, ethnicity, socio-cultural background for example. It also accounts for the type of variation within the speech of a single speaker as a result of change in situational context. Focus on the learner’s linguistic and sociolinguistic competence will inform the researcher about what learners know about the L2 system and what they know about how it is used in communication. In this approach variability is overcome by demonstrating that language use can be both variable and systematic.
- The third approach is the psycholinguistic approach: The objective of psycholinguistic processing models is to explain variability in terms of factors that affect the process whereby learners assimilate L2 structures under different conditions of use (for example whether the learner’s language is planned in which case we can expect more attention to form or unplanned when we have more focus on communication in spontaneous conversations.)
4) Incompleteness
Incompleteness in SLA is a failure on the part of L2 learners to achieve ultimate (native-like) competence in the TL. This failure is theoretically due to two main reasons:
- Learners have already set the parameters of variation in UG theory according to the specificities (particular features and organization) of the L1 system. They tend to transfer their parameter setting from their L1 into the target language. They can’t determine that their initial setting is inappropriate to the TL. They can’t be aware given the positive evidence they discover from those frequent instances of similarity between L1 and L2; hence, the controversy.
- Some subcomponents of UG which are fully available to child L1 learners become difficult to access for adolescents and adults showing the role of the learner’s age and the effect of age of 1st exposure to an L2 has been studied by Johnson and Newport who undertook a study of grammatical phenomena in the L2 English of 46 L1 Chinese and L1 Korean speakers and compared their performance with that of a group of 23 American born native speakers of English.
Post 7 year old L2 learners are typically incomplete in their ultimate knowledge despite large amount of exposure to the L2. (the validity of the critical age hypothesis). This study led to the conclusion that incompleteness grows progressively with age. In other words, the older you are at 1st exposure to TL the more incomplete your competence will be for that TL. Tsimply and Rousso 1981 suggest that L2 learners do not reset parameters at all but to use their own words
“rather look as if they do, because they misanalyse L2 input to make it conform where possible to the parameter settings of their L1.”
This explanation is supported by “constructivism” and the phenomenon which seems to be at work to misanalyse the input of the TL to make it match its equivalent in the L1 is called “assimilation”.
Krashen’s order: The Natural Order Hypothesis
Development of SLA: 1, Early approaches (5 items), 2, a shift from the behaviorally oriented explanation to mentalistic explanation (inter language), 3, the emergence of Krashen’s model (he extended LAD to SLA)
Inter language questions the validity of transfer and then comes Krashen’s model.
The discovery of clear examples of stage development and systematicity led a number of researchers to conclude that the CAH about SLA failed to explain the acquisition process and that the course of SLA, like the course of L1 acquisition, is determined by innate principles of linguistic knowledge what Pitt Corder called in relation to SLA a “built-in syllabus” (learner developing grammar corresponding to Corder transitional competence). This built-in syllabus: I provided the learners with a syllabus and the learners show clear deviancies from what I have provided which shows that the learners assimilate the input and provide something different that they construct: this is called the intake (the learner’s conversion of grammar). Krashen (1985) formulated the observation as a hypothesis: the Natural Order Hypothesis. This hypothesis as recognized by Krashen was first proposed to SLA by Pitt Corder. It states that we acquire the rules of language in a predictable way. Some rules tending to come early and others late. The order does not appear to be determined solely by formal simplicity and there is evidence that it is independent of the order in which rules are taught in the foreign language classroom. However, it soon became clear that the Natural Order Hypothesis needed to be completed by ancillary (subordinate) hypotheses. To begin with, it was found that while natural orders were often displayed in the spontaneous unplanned productions (learners concentrate on the content of the message) of L2 learners. When learners were allowed to plan their productions or when the task was not spontaneous those orders change. Typically, learners display different degrees of approximation to the TL norms under different task conditions. For example, Ellis in his study (1987) asked 17 adult learners of L2 English from different L1 backgrounds to perform 3 tasks. First, to write a story based on a picture composition where subjects were allowed 1 hour to complete the task. Second, to retell orally in the language lab the story they had written. The third task is to tell orally in the language lab a story about another picture composition without prior planning. One of the grammatical phenomena Ellis considered in the samples collected from his informants was the use of the regular past tense. He found that subjects were most target-like in their use of this morpheme (ed morpheme) in the planned written composition; less target like on the oral retelling and least target like on the unplanned telling of a different picture composition. Conclusion: task-based differences in accuracy of this sort are common in SLA. Similarly, if L2 speakers are asked to produce the L2 spontaneously, but are subsequently asked to review what they have said or written making corrections where necessary, it is found that they are aware (reminds of Corder’s distinction of errors: pre-systematic, systematic, and post-systematic) of those discrepancies between what they have said or written and what they know about the TL. (This explains the difference between knowledge of the language and knowledge about the language). Dulay and Burt and Krashen found that L2 subjects could increase their accuracy by between 6% and 47% with the kind of reviewing procedure. Variability in L2 performance of this sort (across tasks) led Krashen 1982/85/88 to supplement the NOH with the:
Acquisition Learning Hypothesis:
This hypothesis states that L2 learners are capable of developing two types of distinct grammatical knowledge about the TL. What he calls
- Acquired L2 Knowledge which develops subconsciously in learners as the result of exposure. (ample evidence of the assumption that learners process the language following their cognitive strategies that are independent of formal instructions and the learner’s intelligence, background…). The second type is
- Learnt L2 knowledge which learners acquire consciously (formal instructions) either thru learning about the language from textbooks or teachers or thru forming their own rules of thumb (error and trial).
Krashen’s view is that real L2 knowledge of this kind found in learners’ spontaneous meaningful productions can be initiated only by the acquired system which develops as the result of the learner being involved with the L2 in spontaneous meaningful interactions. The acquired system of knowledge is the one which displays natural orders of development. Under certain circumstance, however, knowledge about the L2 which has been consciously learnt that is the learner’s learnt language can be used by the learner to monitor the output initiated by the acquired system; hence, Krashen’s Monitor Hypothesis to check for discrepancies and to correct the output where such discrepancies are found. (discrepancies between the learner’s output and the norms of the TL). The circumstances in which such monitoring can occur according to Krashen’s early work are where the learners have enough time to access the learned knowledge (to think) which s/she possess and where the learner is focused on the form of what s/he is producing.
Monitor Hypothesis:
(Krashen defines the different roles of each type of knowledge): this hypothesis states how acquisition and learning are used in production. Our ability to produce utterances in another language comes form our acquired competence i.e. from our subconscious knowledge. Learning conscious knowledge serves only as an editor or monitor. We appeal to learning (we resort to our conscious knowledge) to make corrections to change the output of the acquired system. Before we speak or write or sometimes after we speak or write as in self correction. Krashen hypothesizes that two conditions need to be met in order to use the monitor (i.e. the monitor which is similar to our use of language doesn’t work systematically):
The performer must be consciously concerned about correctness and s/he must know the rule. Both these conditions are difficult to meet. (reminds of Corder’s distinction of errors: pre-systematic errors showing that the learner is not concerned about correctness, he doesn’t know that there is a rule governing the utterance he produced. The opposite is post-systematic errors when the learner knows the rule but fails to apply) The conditions are difficult to meet because the learner can focus on form at the expense of content or vise versa.
| Focusing on form may result on somewhat more grammatical accuracy but it does take more time |
. Of course this remark is supported by empirical evidence in that he claims that in a recent study using adult subjects it was reported that focusing on form took about 30% longer and resulted in about 40% less information transmitted. This may seriously disrupt communication in spontaneous conversation.
The Input Hypothesis:
The IH claims that humans acquire language in only one way by understanding messages or by receiving comprehensible input. We progress along the natural order (referring to the 2nd hypothesis: NOH) by understanding input that contains structures at our next stage (in other words the learner’s ability to contain an input containing structures that he doesn’t know), structures that are a bit beyond our current level of competence (from I to I +1) the next level along the natural order by understanding input containing I+1 we are able to understand language containing unacquired grammar with the help of context which includes extra-linguistic information, our knowledge of the world and previously acquired linguistic competence (partial language of L2 system). The caretaker provides extra-linguistic context by limiting speech to the child to the here and now in 1st language acquisition. The beginning language teacher provides contexts via visual aids and discussion of familiar topics. The IH has two complementary implications.
a- Speaking is a result of acquisition and not its cause. (¹behaviorist approach ) Speech cannot be taught directly but emerges on its own as a result of building competence. Via comprehensible input
b- If input is understood and there is enough of it, the necessary grammar is automatically provided (this implication reflects the child’s construction of the grammar of the surrounding language). The language teacher need not attempt to teach the next structure along the natural order. It will be provided in just the right quantities and automatically reviewed if the student receives a sufficient amount of comprehensible input.
The Affective Filter Hypothesis
Comprehensible input is necessary for acquisition but it is not sufficient. The acquirer needs to be ‘open’ to the input. The affective filter is a mental block that prevents acquirer from fully utilizing the comprehensible input he receives for language acquisition. When it is up, the acquirer may understand what he hears and reads but the input will not reach the LAD. The input can be processed but not converted into actual knowledge. This occurs when the acquirer is unmotivated, lacking in self-confidence or anxious, when he is on the defensive, when he considers the language class to be a place where his weaknesses will be revealed. The filter is down when the acquirer is not concerned with the possibility of failure in language acquisition and when he considers himself to be a potential member of the group speaking the target language. Krashen suggests that the filter is lowest when the acquirer is so involved in the message that he temporary forgets he is hearing or reading another language.
Evaluation of Krashen’s model
Krashen’s system of hypotheses attracted enormous interest and it was one of the 1st attempts to provide an integrated account/analysis of some of the main observations about SLA. Although the hypotheses that make up Krashen’s model fail in crucial ways to make the right predictions, the account was pioneering and helped to focus thinking about what the key issues requiring explanation in SLA are. If we consider how the system deals with the five main observations in SLA namely transfer, staged development, systematicity, variability and incompleteness; we notice:
- 1st Transfer has been completely ruled out. This was probably a historical consequence of the evolution in thinking from behaviorally oriented theories to mentally oriented ones that have taken place in SLA research. Indeed, once researchers had become disillusioned with the Contrastive Analysis Hypothesis which treated all L2 phenomena, (all the learners’ mistakes) as transfer effect, the tendency shifted to the other extreme namely, no L2 phenomena are the result of transfer effect. The learner’s mistakes are due to acquired or learnt effects. This is clearly implausible given the range of evidence already available.
- 2nd, Although staged development and systematicity have been given a label (acquired knowledge) this an entirely mysterious phenomenon in Krashen’s system. While the learner goes thru the stages he receives no explanation. To say that L2 knowledge is acquired thru meaningful interactions involving the L2 is to do no more than observe that there is a staged development and that learners are systematic in going thru those stages. An efficient approach to SLA needs to tell us much more about development and systematicity than this.
TGG -> (a) Conclusion: psychologically
transfer
ST development è the pedagogical implications of CA?
Lexical morpheme grammatical morpheme
5 hypotheses
- Acquisition/ Learning Hypothesis objective? (What is the relation between the acquisition learning hypothesis and the identity hypothesis? Acquisition Hypothesis concerns the Identity Hypothesis (L2acq=L1acq) as it confirmed it and clarified the actual differences between 1st language acquisition and 2nd language learning, ‘conscious vs unconscious learning’)
- The Input Hypothesis objective?
- Monitor Hypothesis objective?
- Natural Order Hypothesis objective?
