Functional approaches
While UG has been the dominant linguistic approach to SLA for many years, many researchers have rather chosen to take an external focus on language learning. The more influential of these approaches are based on the framework of Functionalism.
Functional models of analysis date back to the early twentieth century and have their roots in the Prague School of linguistics that originated in Eastern Europe. They differ from structuralist and early generative models by emphasizing the information content of utterances, and in considering language primarily as a system of communication rather than as a set of rules.
The term function has several meanings in linguistics, including both structural function (such as the role which elements of language structure play as a subject or object, or as an actor or goal) and pragmatic function (what the use of language can accomplish, such as convey information, control others’ behavior, or express emotion). Approaches to SLA which are characterized as functional differ in emphasis and definition but share the following characteristics in general opposition to those in the Chomskyan tradition:
• Focus is on the use of language in real situations (performance) as well as underlying knowledge (competence). No sharp distinction is made between the two.
• Study of SLA begins with the assumption that the purpose of language is communication, and that development of linguistic knowledge (in L1 or L2) requires communicative use.
• Scope of concern goes beyond the sentence to include discourse structure and how language is used in interaction, and to include aspects of communication beyond language (Tomlin 1990).
Four of the functional approaches which have been influential in SLA are Systemic Linguistics, Functional Typology, function-to-form mapping, and information organization.
Systemic Linguistics
Systemic Linguistics has been developed by M. A. K. Halliday, beginning in the late 1950s. This is a model for analyzing language in terms of the interrelated systems of choices that are available for expressing meaning. Basic to the approach is the notion, ultimately derived from the anthropologist Malinowski, that language structures cannot be idealized and studied without taking into account the circumstances of their use, including the extralinguistic social context.
From this functional view,
To relate this notion to the question about what language learners essentially acquire, in Halliday’s view it is not a system of rules which govern language structure, but rather “meaning potential”: “what the speaker/ hearer can (what he can mean, if you like), not what he knows” (1973 :346). The process of acquisition consists of “mastering certain basic functions of language and developing a meaning potential for each” (1975 :33).
Halliday (1975) describes the evolution of the following pragmatic functions in early L1 acquisition (he calls them “functions of language as a whole”), which are universal for children:
• Instrumental – language used as a means of getting things done (one of the first to be evolved): the “I want” function.
• Regulatory – language used to regulate the behavior of others: the “do as I tell you” Function.
• Interactional – use of language in interaction between self and others: the “me and you” function.
• Personal – awareness of language as a form of one’s own identity: the “here I come” function. • Heuristic – language as a way of learning about things: the “tell me why” function.
• Imagination – creation through language of a world of one’s own making: the “let’s pretend” function.
• Representational – means of expressing propositions, or communicating about something (one of the last to appear): the “I’ve got something to tell you” Function.
Linguistic structures which are mastered in the developmental process are “direct reflections” of the functions that language serves; their development is closely related to the social and personal needs they are used to convey.
One application of Halliday’s model to the study of SLA comes with seeing L2 learning as a process of adding multilingual meaning potential to what has already been achieved in L1. This is an approach that some of my colleagues and I have taken in our research. We have concluded that “Second language acquisition is largely a matter of learning new linguistic forms to fulfill the same functions [as already acquired and used in L1] within a different social milieu” (Saville-Troike, McClure, and Fritz 1984:60). In studying children who had just arrived in the USA from several different countries, for instance, we found that all of them could accomplish a wide range of communicative functions even while they still had very limited English means at their disposal. What we observed and recorded over a period of several months for every child in our study was not the emergence of new functions (as we would expect in early L1 development), but emergence of new language structures to augment existing choices for expressing them. This structural emergence follows the same general sequence for each function (not unlike early stages of L1). For example:
Other applications of Halliday’s model can be found in the study of SLA in relation to social contexts of learning and use. That perspective is dis cussed in Chapter 5.
Functional Typology
Another approach within the functional framework is Functional Typology, which is based on the comparative study of a wide range of the world’s languages. This study involves the classification of languages and their features into categories (or “types”; hence “typology”), with a major goal being to describe patterns of similarities and differences among them, and to determine which types and patterns occur more/less frequently or are universal in distribution. The approach is called “functional” because analysis integrates considerations of language structure, meaning, and use.
Functional Typology has been applied to the study of SLA most fruit fully in accounting for developmental stages of L2 acquisition, for why some L2 constructions are more or less difficult than others for learners to acquire, and for the selectivity of cross-linguistic influence or transfer (i.e. for why some elements of L1 transfer to L2 and some do not). A particularly important concept which is tied to these accounts is markedness – the notion of markedness deals with whether any specific feature of a language is “marked” or “unmarked.” A feature is “unmarked” if it occurs more frequently than a contrasting element in the same category, if it is less complex structurally or conceptually, or if it is more “normal” or “expected” along some other dimension. The concept applies to all levels of linguistic analysis. For example:
• In phonology, the most common syllable structure which occurs in languages of the world is CV (consonant + vowel, as in me and ba-na na), so this structure is “unmarked.” It is much less common to have a sequence of consonants at the beginning or end of syllables; English sequences like street [ str i:t] and fence [fεnts] are “marked” in this respect.
• In vocabulary, the preposition in denotes location while the preposition into is more complex, denoting both location and directionality. Into is thus “marked” in contrast with in because it is both structural ly and conceptually more complex.
• In syntax, the basic word order in sentences of SVO (subject–verb object) is more common in languages of the world than is SOV. SVO is thus relatively “unmarked” and SOV relatively “marked.”
• In discourse, the expected “unmarked” response to the English for mulaic greeting How are you? is Fine. How are you? (no matter how the respondent is actually feeling). A response which reports information about one’s health or other personal conditions is not expected in this routine exchange, and is “marked.” Similarly, the “unmarked” response to a question requesting information is an answer about the same topic. Silence or a comment on a different topic is a “marked” response because it is not in accord with “normal” conversational practice.
In accounting for order and relative difficulty for acquisition, unmarked elements are likely to be acquired before marked ones in children’s L1 (Jakobson 1941), and to be easier for a learner to master in L2. In phonology, for instance, the babbling and first words of a child in L1 are likely to have an unmarked CV syllabic structure (no matter what the native language), and marked CC sequences appear only at a later stage of development. It is also likely that L2 learners will find marked CC sequences more difficult to produce, especially if they do not occur at all in the speakers’ L1. A markedness account of selective transfer from L1 to L2 (proposed as the Markedness Differential Hypothesis by Eckman 1977) predicts that unmarked features in L1 are more likely to transfer, as well as that marked features in L2 will be harder to learn. A simplified summary of this hypothesis is shown in Table 3.2.
For example, the pronunciation of the marked consonant sequence [sk] in school should be difficult for Spanish L1 speakers, whose native phonological system is “simpler” than English in this respect because it does not allow two voiceless consonants to occur together. It is indeed common for beginning Spanish L1 learners of English L2 to break this [sk] combination apart into two syllables and pronounce the word as [εs-kul], thus avoiding the marked structure. In reverse, learners of Spanish L2 should have no comparable problem pronouncing escuela [ εs-kwe-la] ‘school,’ since it contains no consonant cluster in any syllable.
Functional Typology resembles Contrastive Analysis in comparing elements of different languages in order to predict or explain transfer from L1 to L2, but it goes beyond the surface-level structural contrasts of CA to more abstract patterns, principles, and constraints. The Markedness Differential Hypothesis is also an advance over the traditional CA approach in that:
One implication that we might draw from this approach is that some aspects of some languages are more difficult to learn than others, in spite of the traditional claim within linguistics that all languages are equally complex. Another issue that we might speculate about is why some types and patterns of features are more or less frequent than others in both native and second languages. Functional explanations tend to refer to extralinguistic factors, or elements outside of language. Certain factors that have been suggested are: perceptual salience, ease of cognitive processing, physical constraints (e.g. the shape of the human vocal tract), and communicative needs (see Ramat 2003 for a collection of articles which include consideration of all levels of linguistic analysis).
Function-to-form mapping
Another functional approach which has been applied to the description and analysis of interlanguage emphasizes function-to-form mapping in the acquisitional sequence. A basic concept from this perspective is that acquisition of both L1 and L2 involves a process of grammaticalization in which a grammatical function (such as the expression of past time) is first conveyed by shared extralinguistic knowledge and inferencing based on the context of discourse, then by a lexical word (such as yesterday), and only later by a grammatical marker (such as the suffix -ed). For example, if you ask a beginning learner of English what he did the day before he might say I play soccer, relying on context to convey the meaning of past time; a somewhat more advanced learner might say Yesterday I play soccer, using an adverb to convey the meaning of past; and a still more advanced learner might say I played soccer, using the grammatical inflection -ed.
The general principle of increasing reliance on grammatical forms and reducing reliance on context and lexical words to express functions such as time is followed in all languages. In Chinese L2, for example, learners tend to use the lexical adverb jiu ‘then’ to express temporal sequencing of events before they use the grammatical marker le ‘finished’ in expressing this notion. The following utterances were produced by a beginning learner (a) and a more advanced learner (b) who were retelling the same event in a film (The Pear Story) that they had viewed (Yang 2002):
Talmy Givón (1979) proposed the distinction between a style of expressing meaning which relies heavily on context (which he calls a pragmatic mode) and a style which relies more on formal grammatical elements (a syntactic mode), and the notion that change from one to the other is evolutionary in nature. He lists a number of contrasts in addition to the evolution from no use of grammatical morphology to elaborate use of grammatical morphology, which I illustrated above. Additional develop mental contrasts include:
• From topic-comment to subject-predicate structure. A subject- predicate structure involves more grammatical marking because of the agreement it requires between sentence elements, while a topic comment structure requires no such marking in stating what the topic is and then giving some information about it.
• From loose conjunction (with elements merely juxtaposed or connected with and) to tight subordination (with elements connected by words like since or because)
• From slow rate of delivery (under several intonation contours) to fast rate of delivery (under a single intonational contour).
• From word order governed mostly by the pragmatic principle of old information first, followed by new information (as in topic-comment structures) to word order used to signal semantic case functions (such as subject or object).
• From roughly one-to-one ratio of verbs to nouns in discourse to a larger ratio of nouns over verbs. The increase in the ratio of nouns to verbs indicates that more semantic case functions are being expressed: e.g. not just subject (only one noun with one verb), but also object and indirect object (a total of three nouns).
According to this approach, language acquisition importantly involves developing linguistic forms to fulfill semantic or pragmatic functions. Grammaticalization is driven by communicative need and use and is related to the development of more efficient cognitive processing (e.g. via automatization) as part of language learning. This aspect of language acquisition will be considered in Chapter 4.
Information organization Information organization refers to a functional approach which focuses on utterance structure, or “the way in which learners put their words together” (Klein and Perdue 1993 :3). The task of studying SLA from this perspective includes describing the structures of interlanguage (called learner varieties by Klein and Perdue), discovering what organizational principles guide learners’ production at various stages of development, and analyzing how these principles interact with one another.
The evidence for this description and analysis comes primarily from the European Science Foundation (ESF) project (e.g. Klein and Perdue 1992; Perdue 1993). Over a period of almost three years, Klein, Perdue, and other linguists regularly recorded the L2 production of speakers of six L1s who were learning five different L2s. All of the learners were adult immigrants in Europe who needed to use the L2 to communicate but did not receive a significant amount of formal instruction in that language.
The number of L1s and L2s in this study is important because it allows the researchers to make generalizations about the nature of interlanguage (or learner varieties) which would not be possible if all of the participants were speakers of the same L1, or if all were learning the same L2. The combinations of native and target languages are shown in Figure 3.2 (adapted from Klein and Perdue 1992 :5).
This list indicates that the participants are native speakers of both Punjabi and Italian learning English, of Italian and Turkish learning German, of Turkish and Arabic learning Dutch, of Arabic and Spanish learning French, and of Spanish and Finnish learning Swedish. Most of the L2s are related Germanic languages, but the L1s represent several very different language families: Turkic (Turkish), Semitic (Arabic), Indo Iranian (Punjabi), Romance (Italian and Spanish), and Finno-Ugric (Finnish).
Developmental levels
All of the learners in this study, no matter what their L1 and L2, go through a remarkably similar sequence of development in their interlanguage. The examples are from narratives about a Charlie Chaplin film that were told by learners in English L2 (as reported in Huebner, Carroll, and Perdue 1992).
• Nominal Utterance Organization (NUO). Learners generally begin with the seemingly unconnected naming of subjects and objects (i.e. with nouns and pronouns, or “nominals”). They may also use adverbs and adjectives or other elements but seldom use a verb to help organize an utterance.
• Infinite Utterance Organization (IUO). Learners increasingly add verbs to their utterances, but they seldom use grammatical morphemes to convey the meaning of tense, person, or number (i.e. the verb is uninflected, or “infinite”). There is also increasing use of grammatical relators such as prepositions. At this stage, learners have constructed an interlanguage grammar which is called the Basic Variety. They may be able to express themselves adequately at this stage in some contexts, and not all continue development beyond this level.
• Finite Utterance Organization (FUO). Learners who continue interlanguage development beyond the IUO level next add grammatical morphemes to the verb (i.e. the verb becomes inflected, or “finite”). This is the process of progressive grammaticalization, which was described in the previous section on function-to-form mapping.
The sequence of structural development shows minimal cross-linguistic influence for the NUO and IUO levels; speakers of all languages follow the same pattern. More L1 transfer occurs as learners increase their L2 resources and produce more complicated utterances (Perdue 2000).
Organizing principles
There is a limited set of principles which learners make use of for organizing information. These interact, and the balance or weight of use among them shifts during the process of interlanguage development. These principles may be classified as follows:
• Phrasal constraints, or restrictions on the phrasal patterns which may be used. Once the verb has emerged, for example, a basic pattern is noun phrase plus verb (NP + V), with a second NP after the verb possible. There are also restrictions on the composition and complexity of each phrasal category. For example, at one stage of development a noun phrase (NP) may consist only of a noun (N) or a pronoun. At the next stage of development, it may consist of a determiner (e.g. the) plus noun (D + N) or an adjective plus noun (Adj + N), but not D + Adj + N. Possible phrasal composition increases in complexity with developmental level.
• Semantic constraints, or features of categories like NP which determine their position in a sentence and what case role they are assigned (e.g. agent or “doer” of the action, or patient or recipient of the action). When an utterance has more than one NP, learners use such semantic factors to decide which one should come first. The principle that learners follow is to put the agent first, or the NP that refers to the thing that is most likely to be in control of other referents.
• Pragmatic constraints, including restrictions that relate to what has been said previously, or to what the speaker assumes that the hearer already knows. The general pragmatic principle is to put what is known (the topic) first, and new information or what the speaker is focusing on last.
While all learners follow essentially the same principles in organizing their utterances, there is individual variation, in part attributable to how the principles apply in their L1 and influence interlanguage use. These constraints are therefore not seen as deterministic, but as “something like ‘guiding forces’ whose interplay shapes the utterance” (Perdue 1993 :25).
In summarizing results, Klein and Perdue (1993 :261–66) offer four “bundles of explanations” for the sequence of acquisition they find, and for why some L2 learners are more successful than others:
• Communicative needs. Discourse tasks push the organization of utterances, in part to overcome communicative inadequacies. Linguistic means are acquired to overcome limitations of earlier levels or stages of expression.
• Cross-linguistic influence. Influence from L1 affects rate of interlanguage development and ultimate level of success, although not order of acquisition. L1 influence is a factor in rate and achievement because it more or less facilitates learners’ analysis of L2 input and plays a role in their selection from among possible L2 organizational devices.
• Extrinsic factors. Progress beyond the basic variety is dependent both on “propensity” factors such as attitudes and motivation, and on “environmental” factors such as extent and nature of learners’ exposure to L2. The everyday environment has more influence on progress at this level than does classroom learning.
• Limits on processing. Learners’ current internalized interlanguage system must be ready to integrate new linguistic features or they cannot be put to immediate use in communication. Learners cannot attend to all communicative needs at the same time.
Klein and Perdue conclude:
In addition to understanding the information organization of the developmental structures of learner language, linguists at least since Talmy (1975) have considered the typological classification of language according to how semantic components of events are encoded in verbs or other grammatical structures. Languages have different preferred pat terns for conveying such concepts as (1) what manner of motion occurs in an event, (2) how the time of speaking about an events relates to the time of its occurrence, and (3) what spatial perspective the speaker takes on an event.
Examples of different patterns of encoding manner of motion include the English preference for lexicalizing the manner of motion in the verb and the path in a prepositional phrase (e.g. The girl danced into the room ) contrasted with the Japanese preference for lexicalizing the path of motion in the verb and the path in an adjunct or satellite structure (e.g. Shojo-ga heya-ni odori-nagara haitta ‘The girl entered the room while she was dancing’).
Decisions on which component(s) of a concept should be lexically anchored and which should be expressed in a more peripheral structure are normally expected only of very advanced L2 learners. (A discussion of different patterns and of factors involved in L1 and L2 selection are found in Carroll and von Stutterheim 2003).
Implications of this functional approach for teaching are discussed by Cadierno and Robinson (2009). Learning to use appropriate L2 linguistic forms for some events (such as those expressing types of motion) may require learning to think about them differently, such as learning to attend to different details. Likelihood of success in instruction is clearly dependent on the typological relation of L1 and L2, but the level of learn er L2 proficiency is more problematic.
All of the functional approaches discussed here basically agree on the following:
• what is being acquired in SLA is a system for conveying meaning,
• how language is acquired importantly involves creative learner involvement in communication, and
• understanding of SLA processes is impossible if they are isolated from circumstances of use.
However, for many who take a functional approach, concern with communicative meaning and context does not preclude belief in the existence of an innate (and possibly language-specific) faculty as an explanatory mechanism, nor does it rule out concern with addressing the “logical problem,” that learners somehow know much more about language than can be accounted for by the input they receive.
الاكثر قراءة في Linguistics fields
اخر الاخبار
اخبار العتبة العباسية المقدسة
