ACQUIRING LANGUAGE

PART I

INTERLANGUAGE AND LANGUAGE TRANSFER

By

Casimirus Andy Fenanlampir

12706251052

A.      Background, Definition, Terminology, and Scope

If someone wants to learn a foreign language, he will obviously meet with many kinds of learning problems dealing with its sound system, vocabulary, structure, etc. This is understandable since the student learning the foreign language has spoken his own native language, which has been deeply implanted in him as part of his habit. Very often, he transfers his habit into the target language he learns, which perhaps will cause errors. Linguists try to find out the causes of the problems to be applied in language teaching, to minimize the problems. It was Larry Selinker (1972) and Uril Weinreich (1953)  formulation of Interlanguage theory which came into standard use. The interlanguage theory is proposed as a critic to Contranstive Analysis theory which is assumed that learners' errors were caused by the difference between their first language and their second language.

An interlanguage is the term for a dynamic linguistic system that has been developed by a learner of a second language (or L2) who has not become fully proficient yet but is approximating the target language: preserving some features of their first language (or L1), or overgeneralizing target language rules in speaking or writing the target language and creating innovations.

Interlanguage is based on the theory that there is a "psychological structure latent in the brain" which is activated when one attempts to learn a second language. Selinker noted that in a given situation the utterances produced by the learner are different from those native speakers would produce had they attempted to convey the same meaning. This comparison reveals a separate linguistic system. This system can be observed when studying the utterances of the learner who attempts to produce meaning in using the target language. To study the psychological processes involved one should compare the interlanguage utterances of the learner with two things:

1.       Utterances in the native language to convey the same message produced by the learner

2.       Utterances in the target language to convey the same message, produced by a native speaker of that language.

Meanwhile, Language transfer (also known as L1 interference, linguistic interference, and crossmeaning) refers to speakers or writers applying knowledge from their native language to a second language. Odlin (1989), in Splosky (2008: 411) stated that:

Transfer (or cross-linguistic influence) is the influence resulting from the similarities and differences between the target language and any other language that has been previously (and perhaps imperfectly) acquired.

The definition brings out a number of features. First, the neutral term influence enables us to consider that positive and negative impact may arise. Second, the source of these impacts is the similarities and differences between the language systems. Third, the definition is consistent with multiple language learning, so that L2–L3 influences (and so on) are also covered by transfer. Regarding interlanguage, Ellis (1994), in Spolsky (2008: 412) state that:

The term has come to be used with different but related meanings (i) to refer to the series of interlocking systems which characterize acquisition, (ii) to refer to the system that is observed at a single stage of development (“an interlanguage”), and (iii) to refer to particular L1/L2 combinations (for example, L1 French/L2 English vs. L1 Japanese/L2 English).

                As we will see, the two areas, transfer and interlanguage, are considerably intertwined, and can be considered together most of the time, with L1 influences (transfer) having a strong influence on interlanguage.

B.      Constraints on Interlanguage Development and Transfer

There are six constraints to be covered fall into two main types, as follows:

1.       The first type (function more at a descriptive level) consists of language level, sociolingiustic influence and task, and language distance and psycotypology.

2.       The second type (more theoritically motivated accounts of what occurs in language change and transfer) consists of linguistics influences, psycholinguistic influences, and developmental influences.

1.1. Language level

Ellis (1994) and Hansen (2006) state that the strong effect on transfer is on phonology level. It is because the sound systems of a first language being particularly deep seated and difficult to change. In addition, Ellis (1994) proposes that syntax is slightly less affected by transfer than other areas because metalinguistic factors can have a dampening influence. The potential availability of clear feedback, as well as a clear focus on the area within teaching, may also reduce the amount of transfer at this level. Kellerman (1978) state that Lexis is affected and interacts in interesting ways with learner perceptions of difficulty and plausible transferability. There is also the area of discourse, which is possibly more difficult to study. Olshtain (1983) was able to show interesting over-use effects with apologies used by US English speakers learning Hebrew. These English L1 learners transferred their direct apology expressions into Hebrew in ways which were not native-like.

1.2. Sociolinguistic influence and task

The point to consider here is that performance may be partly a function of elicitation task or language context. As a result, transfer may manifest itself differently in different contexts, and indeed transfer may not always occur in categorical fashion. Transfer is, at least in part, a performance phenomenon, and second language speakers have to resort to the L1 when they are dealing with greater communicative problems.

1.3. Language distance and Psychotypology

Languages vary considerably in how close they are to one another, and this has been of interest in two ways. First, greater language distance may have effects on the nature of transfer. Second, language effects may have influences on perceptions of transferability where these are linked to psychological factors which impact on performance.

1.4. Linguistic factors

The most well-known theoretically-motivated account of transfer derives from Universal Grammar (UG). One should say that there are, in fact, a range of positions on transfer, since there are a range of positions on the role of UG in L2 learning. A starting point to look at SLA is the principles and parameter setting model which is the most promising advancement in L2 acquisition research. Flynn presents three hypotheses to explain role of UG in SLA (Logical problem):

Ø  “No Access” Hypothesis

UG is totally inaccessible to the adult L2 learner; learning takes place in terms of non linguistic learning strategies.

Ø  “Partial Access” Hypothesis

UG is partially available to the learner; only those parametric values characterising the L1 grammar are available, the rest must be learnt in terms of non-linguistic learning strategies.

Ø  “Full Access” Hypothesis

UG is fully available; differences in patterns of acquisition between L1 and L2 learners and the lack of completeness can be accounted for in other ways.

               

UG Theory of the human capacity for language and its acquisition

Principles Universally invariant properties of grammar construction àinnate

Parameters Determine those properties of language relevant to the construction of a specific grammar à must be learned

Parametric Values Determine what forms the variation can take.

 

Universal             Language Specific

(Source: Article on Seminar Language Acquisition and Universal Grammar)

Parameter settings, are available as a starting point or an influence on second language learning; and how completely the different positions operate. In principle, there may still be partial access directly to UG through what is called the minimal trees hypothesis, or alternatively there may still be complete access, so that second language learners interact with data in the same way as first language learners do (Schwartz & Sprouse, 1996 in Spolsky, 2008: 416). UG researchers are assuming that the end state of the process of first language acquisition is where the second language learner starts, so that whether from direct access and complete transfer, or from some diluted version of these, the learner is equipped at the starting point with whatever has been learned before.

1.5. Psycholinguistic accounts: The Competition Model

The Competition Model does not regard language as a special domain, but instead portrays the human brain as a general purpose learning device, strongly influenced by frequency of occurrence, cue reliability and availability. The model uses a connectionist approach to second language development which emphasizes communicative functions, as these are encoded in the surface elements of language. For example, the model focuses on the agent identification in a sentence. This role is seen as identifiable through surface cues such as pre-verbal positioning, agreement marking, use of article, animacy, presence of by and other cues of this nature. the key point with the Competition Model is that different surface cues are linked with decisions on e.g., agency, in different ways in different languages. In English, for example, the role of actor is strongly associated with pre-verbal positioning. The central issue for transfer then is what the learner of an L2 does regarding the cues that have been learned to be important in the L1. As MacWhinney (2001: 80) in Spolsky (2008: 417) says: “Initially, the learning of the L2 is highly parasitic on the structures of the L1 in both lexicon . . . and phonology . . .” Later he says: “This means that initially the L2 system has no separate conceptual structure and that its formal structure relies on the structure of the L1. The learner’s goal is to build up L2 representations as a separate system.”

1.6. Processing accounts

This approach emphasises the consequences of processing constraints for second language development. Roger Andersen (1983) in Spolsky (2008: 417) proposes that there are processing constraints which mean that:

Ø  learners may need to reach a certain stage of development before transfer becomes possible.

Ø  language development is delayed when an L1 structure coincides with a developmental structure.

Andersen (1983) also proposes his ‘Transfer to Somewhere Principle’, at a very general level: “Transfer can only function in conjunction with the operating principles that guide language learners and users in their choice of linguistic forms to express the intended meaning.” The principles, with children, are meant to reflect general cognitive operations, but linked to how children are able to process language. Andersen then applies them to the second language case, arguing that when they apply transparently in particular language combinations, transfer is more likely to occur.

A more recent and extensive processing viewpoint is associated with the work of Manfred Pienemann. Drawing on Lexical-Functional Grammar (Bresnan, 1982) and also Levelt’s Speech Production Model (1989, 1999), he offers an incremental (i.e., cumulative) account of language development. This moves:

• from the lemma,

• to the category procedure (i.e., involving the lexical category of the lemma),

• to the phrasal procedure (which presupposes knowledge of the category of

            the head of the phrase),

• to the S-procedure and target language word order rules,

• ending with the subordinate clause procedure.

The order is the consequence of what information needs to be exchanged in order to achieve successful performance at each level, and as such, reflects the increasing development of the L2 system, and also the constraints introduced by limited capacity memory systems. In other words, each of the stages is a pre-requisite for more complex stages, and more complex stages cannot be used unless the earlier stages have been achieved. Pienemann (2005) contains a number of articles which apply Processability Theory to a range of different L1–L2 combinations. 

Processability Theory (PT) is important for transfer. Pienemann et al. (2005) write: “The theoretical assumptions underlying our approach . . . include the following two hypotheses: (1) that L1 transfer is contrained by the processability of the given structure and (2) that the initial state of the L2 does not necessarily equal the final state of the L1 . . . because there is no guarantee that the given L1 structure is processable by the under-developed L2 parser.” In other words, for transfer to be possible, two conditions have to be met: (1) there is something which could be transferred, and (2) the L2 learner has to be sufficiently advanced to be able to process that something effectively.

This stands in strong contrast to the sort of claim made by the Competition Model, such as“the early second language learner should experience a massive amount of transfer from L1 to L2 . . . connectionist models . . . predict that all aspects of the first language that can possibly transfer to L2 will transfer” (MacWhinney, 1997: 119).

Pienemann (1998) characterizes this position as ‘bulk transfer’ and argues that it will lead to problems if underlying processing conditions are not met, and also lead to the proliferation of unwieldy hypotheses. Pienemann et al. (2005) suggest: “it is hypothesized that such cases of L1 transfer occur as part of the overall reconstruction process. This means that L1 transfer is developmentally moderated and will occur when the structure to be transferred is processable within the developing L2 system.” Pienemann et al. (2005) cite studies by Haberzettl (2000), showing the way in which advantages occur through typological similarity after a processing condition is met, and Kawaguchi (1999), showing how typological dissimilarity is not a disadvantage, for similar reasons.

C.      Methodology

The central problem for methodology with interlanguage and transfer is to establish convincing explanations of data, where particular interpretations are shown not only to account for data, but also to be the best or only account of that data. In an attempt to address this range of issues, Jarvis (2000) proposes three criteria that need to be met for transfer to be established:

Ø  Similarities between native and L2 performance: This is the most basic and most widely applied criterion, and requires that L2 performance look as if it is connected to the L1 system.

Ø  Inter-group heterogeneity: This means that L2 learners of a particular language who have different L1s should show differences in performance from one another, e.g., French and Japanese learners of English (say) should make (some) different errors. Meeting this criterion is important to rule out simplification or developmental explanations of an error.

Ø  Intra-group homogeneity: This focuses on the consistency of L2 performance

of a group of learners who do share an L1, e.g., there should be some consistency in the Mandarin errors (say) of L1 Spanish learners of this language.

Regarding interlanguage studies, methodology has a slightly different role. Earlier studies were data driven, and the focus was on finding systematicity and progression in interlanguage development. This line of inquiry has continued to some degree, but now broader theoretical accounts have, as we have seen, assumed greater importance. As a result, theory and prediction have become more prominent, and simply identifying patterns in data is less central. One of the key concepts in interlanguage, is to be established, rather than simply assumed.

D.      Interlanguage, Transfer and Performance Models

An interesting development in recent years has been the more widespread adoption within SLA studies of psycholinguistic models of first language speech production. The most influential of these has been Levelt’s (1989, 1999) Speech Production Model, and its direct adaptations to the second language case by, for example, De Bot (1992) and Kormos (2006). The models provide a general framework for speech production, especially with Levelt’s stages of Conceptualization, Formulation, and Articulation.

Speech Production Processes (Levelt 1989)

CONCEPTUALIZATION (MESSAGE LEVEL OF REPRESENTATION) Ø Involves determining what to say Ø Speaker conceives an intention Ø Speaker selects relevant information in preparation for construction of intended utterance Ø The product is a preverbal message

FORMULATION Ø Involves translating the conceptual representation into a linguistic form Ø Includes the process of lexicalization, where the words that speeaker wants to say are selected Ø Includes the process of syntactic planning where words are put together to form a sentence Ø Involves detailed phonetic and articulatory planning Ø Include the process of phonological encoding, where words are turn into sounds

ARTICULATION Ø Involves retrieval of chunks of internal speech from buffer Ø Invloves motor execution

 

The model may not currently explain very much about interlanguage or transfer, but it enables different effects to be located in a more satisfying manner, and linked to underlying knowledge systems. We can, in this way, perhaps relate pre-emptive avoidance to the Conceptualizer stage, and most transfer and interlanguage stages to the Formulator and Articulator stages. It is also possible, in this way, to reinterpret, slightly, performance accounts of transfer, in that the Conceptualizer may deliver to the Formulator communicative requirements which the Formulator cannot handle and so therefore the improvisation that the Formulator has to achieve draws on L1 sources, and transfer is the consequence. De Bot (1992) also makes proposals about common lexical stores which are drawn on in performance, and also that parallel speech plans may be produced by the bilingual, and so when difficulties in one plan are encountered the parallel L1 plan may be drawn on. Proposals such as these have not been elaborated into specific hypotheses about transfer, but it would seem likely that this approach will have increasing importance in the future, with the two areas being integrated in a more satisfactory manner.

E.       Pedagogy

A central point with such research (interlanguage and language transfer) is that it can be used to inform pedagogic decisions. A concern for issues at this level may enable more effective macro decision making, which may impact upon national educational systems. This might include national coursebook series which are attuned to local knowledge about particular L1–L2 transfer problems and micro issues that individual learners may be having difficulty and it may help a teacher to have resources to analyze error more effectively. This would be a particularly acute problem in contexts where learners do not share an L1, so that although the teacher will be dealing with a common L2, the range of L1s and therefore different potential transfer errors can be very large. There are publications available for this situation, at least for the learning of English, which provide resources on the difficulties which can be expected from speakers of different L1s. Swan and Smith (2001), for example, provide information for teachers of English outlining transfer difficulties for a wide range of different L1s. The assumption is that errors which are made by particular learners may have their origin in the L1, and therefore can only be intelligently handled if the teacher has knowledge of what is presumed to be the source of the error.

PART II

SECOND LANGUAGE ACQUISITION AND ULTIMATE ATTAINMENT

In second language acquisition (L2A) research, ultimate attainment refers to the outcome or end point of acquisition, and is used interchangeably with the terms final state, end state, and asymptote. “Ultimate” is not to be thought of as synonymous with “native-like,” although native-likeness is one of the observed outcomes of L2A. In short, the term “ultimate attainment”is properly used in a neutral sense in reference to the outcome of second language acquisition (L2A), irrespective of whether this outcome is similar to or different from nativelikeness.

A central concern of L2A research is to identify, characterize, and understand putative constraints on L2 learning. The study of ultimate attainment is not uncontroversial, however. Larsen-Freeman (2005: 196) argues that the “static view of finite linguistic competence”implied by the notion of ultimate attainment is conceptually and empirically at odds with the view that an individual’s language (the L1 as well as the L2) is a dynamic system.

Comparisons with adult natives reveal convergences, divergences, and shortcomings – findings that contribute to the overall picture of attainment potential in L2A. This knowledge has driven research into L2 learning mechanisms, experiential and biological factors, and inter-individual differences in L2 learning style and ability. This knowledge also serves as the empirical basis of L2A theory. In comparing L2A with L1A – the most basic enterprise in L2A research – facts about outcomes simply cannot be ignored.

A.      Nativelike Ultimate Attainment

There is a negative and generally linear correlation between age of acquisition (AoA) – usually understood as age of immersion in the L2 context – and ultimate level of L2 attainment. Some researchers claim that nativelike attainment for post-pubertal AoA is impossible, or is so rare as to be irrelevant (Bley-Vroman, 1989 in Spolsky, 2008: 426). Low rates of authentic (unaccented) pronunciation are common in studies of late L2 learners at end state.For example:

Ø  Flege, Munro, and MacKay (1995) studied 120 Italian natives with post-pubertal immersion in English; of these subjects, only 6 percent had no detectable accent, and none had begun speaking English after age 16.

Ø  Birdsong (2003) studied speech samples of 22 Anglophone late learners (AoA > 12 years) of French, along with samples from 17 native controls. From word-list reading tasks, acoustic measurements for Voice Onset Time (VOT) and vowel length were taken. For recordings of subjects reading aloud paragraphs in French, global ratings were given by three judges. Two of the Anglophone late learners performed like French natives on the acoustic measures and in global accent ratings.

Long (1990), Scovel (1988), and Hyltenstam and Abrahamsson (2000, 2003) have claimed that nativelikeness is observed in restricted domains of the target language. For example, a learner may display nativelike accuracy in certain areas of the grammar but not others; or the learner may have mastered surface morphosyntactic features of the language but have accented pronunciation. Certain types and domains of processing in the L2 appear to be resistant to nativelike attainment. Recent studies comparing highly proficient late L2 learners with monolingual natives have found non-nativelike lexical retrieval, structural ambiguity resolution, and perception of acoustic features such as consonant voicing and syllable stress.

The major strands of investigation (ERP methodology and fMRI) have tended to focus on two factors as potential predictors of degree of L1-like processing of the L2: AoA and level of proficiency. In the context of the present contribution, the most relevant generalization emerging from these studies is that both the temporal patterns as well as the regional brain activity patterns of highly proficient late L2 learners are largely congruent with those of early bilinguals or with the same subjects’patterns when processing their L1 (Green, 2005 in Spolsky, 2008: 427).

B.      Factors in Ultimate Attainment

Researchers have suggested a variety factors related to ultimate attainment. They are loss of neural plasticity (Penfield & Roberts, 1959); lateralization of cognitive neurofunction (Lenneberg, 1967); cognitive-developmental factors such as decline of implicit learning ability (DeKeyser, 2003); and affective-motivational factors such as psycho-social identification and acculturation with the L2 population (Moyer, 2004). Spolsky (2008: 428), explain two important explaination, at this point. First, it must be recognized that, among users of multiple languages, not only does the L1 influence the L2 but the L2 also influences the L1 which can be found in VOT, collocations, grammaticality judgments, and syntactic processing. Secondly, experiential factors relating to L2 use and interaction with L2 speakers may be implicated in the ultimate attainment of learners.

In general, however, the literature suggests that L2 input and/or use accounts for a small percentage of the variance in L2 ultimate attainment. As Flege (in press) points out, however, the apparently small role of L2 input and use may in part be an artifact of comparisons of this factor with the “macrovariable” of Age of Acquisition. AoA is a proxy for several variables that have been hypothesized to affect ultimate L2 attainment, such as state of neural development; state of cognitive development; state of L1 phonetic category development; level of L1 proficiency; level of L1 representational entrenchment; and proportion of contact with native-speaker versus non-native L2 users.

C.      The Age Function and Ultimate Attainment

A critical period for L2A implicates specific temporal and geometric features of the function that relates AoA to attainment. Some proponents of the critical period ( Johnson & Newport, 1989; Pinker, 1994) posit an end to this decline, at the point where neurocognitive maturation is complete. The idea is that, once maturation is complete, there are no longer any neurological developments that would cause the L2 attainment function to depart from a horizontal trajectory.

D.      Cognitive Aging and Biological Aging

Birdsong (2006b) examines the linkage of these declines with declines in L2 processing and acquisition over age. Three general areas of cognitive aging are identified: decreases in processing speed, decreases in the ability to suppress irrelevant information, and declines in working memory capacity. Each of these capacities is implicated in language acquisition and use. (Because much of L1 processing is automatized, the effects of these declines in L1 use are likely to be less pronounced than in the L2 case.) Cognitive declines are not observed to the same degree across the board. Significant declines beginning in young adulthood are seen in working memory, associative memory, episodic memory, and incremental learning. In contrast, relatively mild declines are found in tasks involving priming, implicit memory, procedural memory, recent memory, and semantic memory. When processing new information, and under demands of speed and accuracy, performance declines begin in the early twenties and the decline over the remainder of the lifespan is continuous and typically linear. A range of inter-individual differences is observed. In general, these facts are compatible with behavioral evidence in L2 use and acquisition.

In the biological aging literature, researchers have identified changes in neurochemical and hormonal levels that are associated with declines in cognitive performance. Acetylcholine molecules mediate a variety of neural functions, including learning, memory, and attention. In normal aging, declines of acetylcholine and cholinergic receptors start in about the fourth decade of life and progress thereafter. With age and stress, abnormally high levels of cortisol are created in the brain. These increases have been explicitly linked to hippocampal atrophy, resulting in declines in ability to lay down new memories, particularly declarative memories. Age-related changes in estrogen levels have been associated with functional declines in verbal processing and production. Of particular note are declines over age in dopamine receptors, starting in the early twenties. Dopamine is known to mediate a number of motoric and higher-order cognitive functions, some of which are involved in language learning and processing (Backman & Farde, 2005). Schumann et al. (2004) argue that dopamine may be involved in motivation to learn a second language and in reinforcement of learning. It is also thought to be necessary for “defossilization,” the process by which automated non-targetlike performance is undone, thus removing a barrier to nativelike attainment.

E.       Psycho-Social Variables and Nativelikeness

Different levels of L2 ultimate attainment have been linked to individual differences along psycho-social and sociocultural dimensions. People vary in their experiences of society and culture, their ideologies of the L1 and the L2, and in their reasons for learning the L2. Individuals’ goals for the outcome of learning vary as well. To a certain degree, therefore, the level of attainment and the way L2 knowledge is implemented in L2 use are determined by the learner (Gillette, 1994 in Spolsky, 2008: 431).

Conversely, the desire to assimilate is often associated with high L2 proficiency, near-nativelikeness, and nativelikeness. Piller (2002) looked at L2 speakers who desire to pass for native speakers of the L2. Conversations with natives involve psychologically motivated “identity play”; those who are most eager to be identified as natives, and who in some sense take on an identity associated with the L2, are often able to fool their native interlocutors. In her sociolinguistic study of the linguistic practices of bilingual couples, Piller (2002) found that some 40 percent of the individual subjects (with AoAs between 15 and 29) claimed to have attained high-level proficiency and were able to pass for natives on certain occasions. The degree to which L2 learners emulate the performance of natives when speaking with them can vary from one interactional context to the next. Nativelikeness also depends on whether the L2 speaker wishes to be dealt with as a foreigner (which sometimes provokes stereotypical attitudes from interlocutors). Many L2 users seem to weigh, consciously or not, the benefits and disadvantages of the passing-for-native act; this calculus plays out in their L2 speech, with resulting variations in perceived nativelikeness.

F.       Future Directions

Additional concerns that may figure in future research include the following:

Ø  second language dominance

Birdsong (2006a) suggests that studies (both qualitative and quantitative) of L2–L1 attainment differences with samples of highly proficient learners  should not necessarily be construed as meaning that L2–L1 processing differences are inevitable, because it is not clear that L2 high-proficients, as a group, represent the upper limits of L2 attainment. Under-represented in L2A research are individuals whose L2, particularly if learned late, is their dominant language. For a given individual, dominance can be operationalized psycholinguistically, for example by comparing recall or recognition of words heard under noise in the L1 with this performance in the L2.

Ø  approximating the first language context

If one were interested in determining the upper bounds of L2 attainment, it would be important to study L2 learners under conditions of immersion and interaction with natives that are known to favor learning. One candidate approach would be to approximate, within the limits of practicality, the external conditions of the L1 learner: full immersion in the L2, linguistic interactions solely with L2 speakers, and no contact whatsoever with the L1.

Ø  ultimate understanding of ultimate attainment

As we have seen, the candidate factors may relate to the neurobiology of the species, cognitive aging, socio-psychological orientations toward learning, and experiential variables such as amount of L2 use. It will be useful to further specify the components of L2-learning aptitude and the role of each component in determining L2 ultimate attainment (Robinson, 2002). In addition, we expect continued interest in the possibility that training can contribute to nativelikeness in low-level processes such as auditory discrimination (McClelland, Fiez, & McCandliss, 2002) and imitation ability (Bongaerts, 1999).

PART III

EXPLICIT FORM-FOCUSED INSTRUCTION AND SECOND LANGUAGE ACQUISITION

There are two good reasons for examining the effect that explicit formfocused instruction (FFI) has on second language (L2) acquisition. The first is pedagogical. The second reason is theoretical. Theories of second language acquisition distinguished two types of linguistc knowledge. They are implicit knowledge (is unconscious (i.e., we are not aware of what we know implicitly), procedural, accessible for automatic processing, not verbalizable (except as an explicit representation), “acquirable” (i.e., can be internalized implicitly), and typically employed in unproblematic, freeflowing communication), and explicit knowledge (as conscious, declarative, accessible only through controlled processing, verbalizable, learnable (in the sense that any piece of factual information is learnable), and typically employed when learners experience some kind of linguistic problem). In this part wil discuss and focus on explicit form-focused instruction, however.

Theoretical differences exist with regard to the potential for explicit instruction to affect these two types of knowledge. Recent studies in second language acquisition (SLA) have suggested that instruction taking psycholinguistic and cognitive factors into consideration is highly beneficial to second language teaching and learning. To examine this issue in Instructed SLA, an approach called “focus on form” has been proposed (Abe, 2011: 184). DeKeyser (1998), for example, adopts a strong interface position, arguing that instruction consisting of explicit rule-presentation followed by communicative practice can guide the learner from a declarative representation of a linguistic feature to a procedural one. According to Long (in Abe, 2011: 184), focus on form is defined as a type of instruction drawing “students' attention to linguistic elements as they arise incidentally in lessons whose overriding focus is on meaning, or communication”.

A.      Definitions

The term instruction will be used narrowly to refer to attempts to intervene in the process of interlanguage development. Spolsky (2008: 437) distinguishes two kinds of instruction: Communication- Focused Instruction and Form-Focused Instruction (FFI). The former involves the use of tasks that focus learners’ attention on meaning. The latter refers to “any pedagogical effort used to draw the learner’s attention to language form” (Spada, 1997: 73). Our concern here is with one type of FFI, that is explicit FFI.

               

Implicit and explicit forms of form-focused instruction

Implicit FFI	Explicit FFI
·         attracts attention to target form ·         is delivered spontaneously (e.g., in an otherwise communication-oriented activity) ·         is unobtrusive (minimal interruption of communication of meaning) ·         presents target forms in context ·         makes no use of metalanguage ·         encourages free use of the target form	·         directs attention to target form ·         is predetermined and planned (e.g., as the main focus and goal of a teaching activity) ·         is obtrusive (interruption of communicative meaning) ·         presents target forms in isolation ·         uses metalinguistic terminology (e.g., rule explanation) ·         involves controlled practice of target form

Spolsky (2008: 438), following DeKeyser (2003), distinguish explicit/implicit instruction and deductive/inductive instruction.

Ø  Explicit FFI involves “some sort of rule being thought about during the learning process. It means learners are encouraged to develop metalinguistic awareness of the rule. This can be achieved deductively, as when a rule is given to the learners, or inductively, as when the learners are asked to work out a rule for themselves from an array of data illustrating the rule.

Ø  Implicit instruction is directed at enabling learners to infer rules without awareness.

The distinction between explicit and implicit FFI needs to be considered in relation to another common distinction. Long (1991) distinguished “focus on forms” and “focus on form” instruction. Focus-on-forms is evident in the traditional approach to grammar teaching based on a synthetic syllabus. The underlying assumption is that language learning is a process of accumulating distinct entities. In such an approach, learners are required to treat language primarily as an “object” to be studied and practised bit by bit and to function as “students” rather than as “users” of the language. In contrast, focus-onform“draws students’ attention to linguistic elements as they arise incidentally in lessons whose overriding focus is on meaning or communication” (Long, 1991: 45–46). Such an approach, according to Long and Robinson (1998), is to be distinguished not only from focus-on-forms but also from focus-onmeaning, where there is no attempt to induce attention to linguistic form at all. It is clear that, as defined, focus-on-forms entails explicit language teaching of the deductive or inductive kind.

The terms explicit and implicit instruction can only be defined from a perspective external to the learner. That is, it is the teacher, materials writer, or course designer who determines whether the instruction is explicit or implicit (or, more likely, a mixture of the two). In contrast, the terms implicit/explicit learning and intentional/incidental learning can only be considered in relation to the learner’s perspective. Thus, implicit learning takes place when the learner has internalized a linguistic feature without awareness of having done so while explicit learning involves awareness.

The goal of explicit instruction is not just explicit knowledge but rather implicit knowledge, with explicit knowledge seen just as a starting point. In other words explicit instruction is premised on either a strong or a weak version of the interface hypothesis. It is also possible that implicit instruction will result in explicit knowledge. This might occur if learners are not developmentally ready to incorporate the target of instruction into their interlanguage systems and thus temporarily store information about the target as explicit knowledge (see Gass, 1997). It should also be noted that the effects of instruction on learners’ ability to use the target structure in unplanned language use may not be immediately evident; they may only emerge later when explicit knowledge is put to work as “pattern recognizers for linguistic constructions” (N. Ellis, 2005 in spolsky, 2008: 440-441).

B.      Types of Explicit Instruction

Types of explicit instruction can be distinguished as follows:

	Deductive	Inductive
Proactive	Metalinguistic explanation	Consciousness-raising tasks Practice activities • production based • comprehension based
Reactive	Explicit correction Metalinguistic feedback	Repetition Corrective recasts

Ø  Proactive/Deduktive explicit FFI

This type of explicit FFI is realized by means of metalinguistic explanations. These  typically consist of information about a specific linguistic property supported by examples. Metalinguistic explanations can be provided orally by the teacher or in written form in a textbook or reference grammar.

Ø  Proactive/Inductive explicit FFI

Proactive/inductive explicit FFI involves either consciousness-raising (CR) tasks or practice exercises. Spolsky in Ellis (1991)in Spolsky (2008: 442) defined a CR task as “a pedagogic activity where the learners are provided with L2 data in some form and required to perform some operation on or with it, the purpose of which is to arrive at an explicit understanding of some regularity in the data”.  Practice activities constitute a proactive/inductive type of explicit FFI only when the students are either told or implicitly expected to derive metalinguistic awareness of the target feature. That is, they invite intentional rather than incidental learning. Practice activities can involve production, in which case they can be “text-manipulating” (i.e., involve what Norris and Ortega (2000) called “constrained constructed response”) or “text-creating” (i.e., involve the use of tasks that require learners to employ their own linguistic resources). Production activities can also be error-avoiding (most commonly) or errorinducing, as in Tomsello and Herron’s (1988) study. In the latter case, learners are led into making overgeneralization errors and then receive corrective feedback. Practice activities can also be comprehension based. In this case they take the form of “interpretation tasks” (Ellis, 1995) consisting of structured input (i.e., input that has been seeded with the target structure) and some form of operation (e.g., carrying out an action or pointing at an object in a picture) to demonstrate comprehension.

Ø  Reactive/Deductive explicit FFI

Lyster and Ranta (1997) in Spolsky (2008: 442) define explicit correction “as the explicit provision of the correct form” accompanied by a clear indication that what the learner said was incorrect. They define metalinguistic feedback as follows:

Metalinguistic feedback contains either comments, information, or questions related to the well-formedness of the student’s utterance, without explicitly providing the correct form.

Ø  Reactive/Inductive explicit FFI

Two kinds of corrective feedback manifest this characteristic: repetition and corrective recasts. The former involves the repetition of the student’s erroneous utterance with the location of the error signaled by means of emphatic stress. A corrective recast reformulates the learner’s erroneous utterance with the correct form highlighted intonationally.

C.      A Review of Explicit FFI Studies

Proactive Explicit FFI Studies

Spolsky (2008: 443) distinguish between explicit proactive FFI based on these questions below:

1.       What is the effet of different ways of providing metalinguistics information on L2 learning?

Sharwood Smith (1981) proposes that explicit teaching techniques can vary in terms of the degree of elaboration or conciseness with which the explicit information is presented and the degree of explicitness or intensity of the information.

2.       What are the relative effects of deductive and inductive FFI on L2 acquiition?

Erlam’s own study investigated th effects of these two types of instruction on the acquisition of direct object pronouns in French as a foreign language. She reported a distinct advantage as can be seen in table below:

Types of metalinguistic explanation

  10      Elaboration

Type A Covert but elaborate guidance (e.g., through the use of “summarizers”)	Type B Elaborated and explicit guidance (e.g., in the form of an algorithm)
Type C Brief indirect “clues” that hint at a regularity	Type D Concise prescriptions using simple metalanguage

   0                                                              Explicitness                                                   10

3.       Does explicit deductive instruction result in the acquisition of L2 implicit knowledge?

It is noticeable that in all the studies where a positive effect was found the instruction was prolonged (i.e., the learners continued to receive both metalinguistic information and practice activities over a period of several weeks). A tentative conclusion, therefore, is that explicit instruction involving metalinguistic information and practice activities is effective if it is substantial.

4.       Do practice activities work bwst with or without accompanying metalinguistic information?

This question has been addressed by studies based on VanPatten’s theory of Input Processing Instruction as follows:

Processing instruction is a type of grammar instruction whose purpose is to affect the ways in which learners attend to input data. It is input-based rather than output-based.

VanPatten and Oikennon (1996) compared three groups: (1) received explicit information about the target structure followed by structured input activities, (2) received only explicit information, and (3) just completed the structured input activities. Acquisition was measured by means of both comprehension and production tests. In the comprehension test, significant gains were evident in groups (1) and (3) but not (2). In the production test, group (1) did better than group (2). VanPatten and Oikennon interpreted these results as showing that it was the structured input rather than the explicit information that was important for acquisition.

5.       Do input-based and production-based practice have differential effects on L2 acquisition?

VanPatten’s theory of input processing predicts that input-based practice that draws attention to form–meaning mappings will prove more effective than traditional, production practice. The first is that the instructional treatments generally included metalinguistic explanations. The second is that, with a few exceptions, the tests measuring acquisition did not measure learners’ ability to process the target structures in unplanned language use.

6.       Is there a relationship between the quality of practice opportunities and L2 acquisition?

Since the correlational statistic do not address cause and effect, Spolsky made the point that “practice” cannot be considered a monolithic phenomenon. It is, in fact, highly varied, subject to a whole host of social and personal factors.

Reactive Explicit FFI Studies

A number of studies  examined the effects of explicit forms of corrective feedback on learners’ acquisition of specific linguistic features by comparing the relative effects of implicit and explicit types of feedback. The implicit feedback typically took the form of recasts or requests for clarification while the explicit feedback consisted of explicit rejection, explicit correction, metalinguistic information, or some combination of these. In short, these study provides convincing evidence that reactive metalinguistic activity assists development.

D.      Conclusion

In this chapter, I have shown that there is ample evidence that both proactive and reactive explicit FFI assist acquisition and I have also produced some evidence to show that this assistance can be seen even in measures of unplanned language use, which are hypothesized to tap L2 implicit knowledge. I have also suggested some of the characteristics of explicit FFI that appear to be especially facilitative. In this respect, the characteristic that emerges as especially noteworthy is metalinguistic activity involving such instructional strategies as providing learners with metalinguistic information (proactively or reactively), inviting them to discover grammatical rules for themselves, and encouraging reflection on and self-repair of their errors. However, there is no evidence that such strategies work in isolation; rather, the evidence indicates that they work when learners are either subsequently or concurrently engaged in practice activities, which in many of the studies were communicative in nature.

REFERENCES

Spolsky, B. & Hult, F.M. 2008. The Handbook of Educational Linguistics. Oxford: Blackwell Publishing.

Harley. 2001. LanguageProduction. www.ehu.es/HEB/Mikel/Adam%26Mikel_Master2011_12/Lecture%25203%2520biblio/Harley2001production.pdf+levelt's+speech+production+model. 23 April 2013.

Wäber, Kirstie & Wanda Czendlik. 2002. Second Language Acquisition and Theories of Universal Grammar (Dr. Pius Ten Hacken). Seminar Language Acquistition and Universal Grammar. WS 02/03.