目录

  • 1 Understanding SLA
    • 1.1 Definition of SLA
    • 1.2 Objectives of SLA research
    • 1.3 Basic terminology in SLA
    • 1.4 The external and internal factors in SLA
    • 1.5 A review of first language acquisition
    • 1.6 Assignment
  • 2 The Study of Interlanguage
    • 2.1 Definition of interlanguage
    • 2.2 Characteristics of interlanguage
    • 2.3 Major findings in interlanguage studies
    • 2.4 Interlanguage pragmatics
    • 2.5 Assignment
  • 3 Linguistic Aspects of Second Language Acquisition
    • 3.1 The nature of human language
    • 3.2 Early approaches to SLA
    • 3.3 Universal grammar (UG)
    • 3.4 Typological universals: Accessibility hierarchy (AH)
    • 3.5 Functional approaches
    • 3.6 Assignment
  • 4 Psychological Aspects of Second Language Acquisition
    • 4.1 Language and the brain
    • 4.2 Behaviorist way of learning
Universal grammar (UG)

3.3 Universal grammar (UG)

 3.3.1 Definition

As for the source of language knowledge, there are two views which have confronted each other for over 2000 years. One view represented by Plato is that language knowledge comes from universal truths, and the other (which is not discussed here) represented by Aristotle is that language knowledge comes from convention. Today, some linguists still stand in Platos camp, but the names of the camps have changed. They are now called the nativists or mentalists, and they still believe that human beings do have an inborn knowledge of language which must be universally correct and acceptable. The location of such innate knowledge is just in our genes. Since all animals have some qualities genetically inherited from generation to generation, such as spiders inheriting skills of spinning webs, bees inheriting skills of collecting nectar and producing honey, why cant human beings inherit a certain universal logic of language to facilitate their language acquisition? Chomsky believed that there is a biological, physiological entity inside our brain which decides what we speak, and this entity is named universal grammar (UG). He has the following concepts about the knowledge of language:

 

(1) Every human being has language competence, because he has an inborn UG which other species lack.

(2) UG is the initial state of the human language faculty which alone cannot enable a human baby to speak. A baby needs to be exposed to the environment of a certain language and accumulate experience.

(3) Due to the effect of later experience, the babys mind develops from the initial state into the steady state, which corresponds to the competence of speaking a specific human language.

 

According to UG theory, every speaker is endowed with a set of principles which apply to all languages and also a set of parameters that can change from one language to another within certain limits. To have a better understanding, we can compare the principle to the universal traffic rule in every country that every vehicle must run on one side of the road, and the parameter to the specific rule in a specific countryIn one country, cars may be only allowed to go on the right side while in another on the left.

 

3.3.2 Principles and parameters

An example in language is that all human languages have subject, verb, and object, which may be decided by one of the universal principles. However, these three components are in different orders in the sentences of different languages. In Chinese and English, the dominant order is SVO; in Japanese, SOV; in Arabic, we can find VSO;  These differences are regarded as different values of the parameter of word order. For example, in Chinese we say 约翰踢球; in English, we also say John kicked the ball. But the Japanese version is:

 

John-wa booru-wo ketta. (John ball kicked)

 

Therefore, it can be said that acquiring a language means applying the principles of UG to a particular language, and learning which value is appropriate for each parameter.

A second example is co-reference. A general principle of language is that it permits co-reference by means of some form of reflexive. Let us look at the following English sentence:

 

Tom blamed himself.

 

In this sentence, the subject, Tom, is co-referential with the reflexive, himself. Both words refer to the same person. However, in another sentence such as:

 

Tom knew Gary blamed himself.

 

The reflective himself must be understood as referring to Gary, not Tom, because English only permits local binding. Long distance binding in which the reflexive co-refers to a subject in another clause, is prohibited. However, some other languages permit long-distance as well as local binding. For example, the Japanese version of the above sentence is ambiguous; the reflexive himself may refer to either Tom or Gary.

Other principles and parameter settings accounting for variations between languages include those that determine whether or not agreement between subject and verb must be explicitly expressed, and whether or not a subject must be overtly present. Lets compare the following three sentences with the same meaning in three different languages:

 

English: It is raining.

Chinese: Xia yu (down rain)

Spanish: Está lloviendo (Is raining)

(Saville-Troike, 2008:48-49)

 

The English speakers say It is raining, with a meaningless subject it, while the subject is omitted in the Chinese Xia yu and Spanish Está lloviendo.

There is no complete listing of principles and parameters, and there will never be one, since proposals concerning their identity keep changing as the theory evolves. Anyway, the specification of universal principles and parameters is related to theoretical developments and understandings and may have practical value in second language teaching and learning. It must be pointed out that children never use such a list, nor could they understand it if one were available. Principles and parameters are not, cannot, and need not be learned in the acquisition of any L1, because they are assumed to be built into the LAD we are born with. This may partially be true for adult L2 learners, though an awareness of parameter settings in an L2 may help learners focus perception on input and thus facilitate learning.

 

3.3.3 UG and SLA

We know that UG is the initial state of the human language faculty, and the reason why children can successfully pick up their mother tongue is that UG exists in their minds. What we are concerned about here and now is the relationship between UG and L2 acquisition. Does UG work for SLA? We focus on three key issues: the initial state in SLA, the nature and development of interlanguage, and the final state in SLA.

 

1. Initial state in SLA

When L2 acquisition begins, the learners already have the knowledge of L1. Guided by UG, they have made all the parametric choices that are appropriate for their L1. It is a normal phenomenon for some L1 knowledge to be transferred into L2. When L1 and L2 parameter settings for the same principle are the same, positive transfer is likely to take place. On the contrary, when L1 and L2 parameter settings are different, negative transfer or interference may occur. For example, an L1 Chinese learner of English might say kick the ball as a result of positive transfer of a parameter setting from Chinese to English; however, a Japanese beginning learner of English might say ball kick, which is clearly caused by negative transfer of parameter setting from Japanese into English.

Learners might still have access to UG in the initial state of SLA, but there are disagreements on this. Four possibilities have been proposed (see Ellis, 2000:69):

(1) Learners have full access to UG. It is argued that learners begin with the parameter settings of L1 but learn to switch to the L2 parameter settings. An assumption is that full target-language competence is possible.

(2) Learners have no access to UG. That is, UG is not available to adult L2 learners, who rely on general learning strategies. According to this position, L1 and L2 are fundamentally different. Learners will normally not be able to achieve full competence and their interlanguages may manifest impossible rule.

(3) Learners have partial access to UG. That is, they keep some of its components but not others. L2 acquisition is partly regulated by UG and partly by general learning strategies.

(4) Learners have dual access to UG. According to this position, L2 learners make use of both UG and general learning strategies. But the use of the latter can block the operation of UG, causing learners to produce impossible errors and fail to achieve full competence. It is argued that L2 learners can only be fully successful if they rely on UG.

It can be clearly seen from the above contradictory positions that the role of UG in SLA is still uncertain.

 

2. Nature and development of interlanguage

From an UG perspective, interlanguage can be defined as intermediate states of L2 development containing IL1, IL2, IL3, etc. If L2 learners retain some access to UG, the process of IL development is, in large part, one of resetting parameters based on the input of the new language. For example, a Japanese speaker learning English needs to reset the word order parameter from SOV to SVO. Learners change the parameter setting because the L2 input they receive does not match the L1 settings they have. If UG is still available, then that will limit their choices, and their IL grammars will not deviate from structures that are allowed by UG. Positive evidence, which is the input gained from experiencing L2 in natural use or formal instruction, will provide sufficient information for the learners to set necessary parameters. Negative evidence, which includes explicit correction, is often also provided to L2 learners, and this may play a role in parameter resetting for adult L2 learners.

From a constructionist perspective, an approach to SLA formulated within Chomskys Minimalist Program, IL development is considered as the progressive mastery of L2 vocabulary along with the morphological features (which specify word form) that are part of lexical knowledge. While the general principles and parameters do not need to be learned, morphological paradigms must gradually be added to the lexicon, just like words (White, 2003:194). There are stages and variations in IL development because these morphological  features in learners competence are not specified completely. In L1 acquisition, parameter setting and mastery of morphological features are linked; but for adult L2 learners, they are not necessarily linked. Therefore, failure to reach a state of full feature specification in the lexicon is seen as the main reason that many L2 learners fossilize at an intermediate level of development without gaining near-native competence.

In an UG approach, lexical acquisition plays an important role in providing information for parameter (re)setting and other aspects of grammar. This is quite different from the structuralists and behaviorists view that all of the basic grammatical structures of L2 can be learned in conjunction with minimal vocabulary.

If learners no longer have access to UG, IL development would need to be explained as a fundamentally different learning process than that taking place for L1.

 

3. Final state in SLA

The question of why some L2 learners are more successful than others is relevant for the second language acquisition, though it is not relevant for first language acquisition. The following are a few possibilities within the UG framework that account for the great difference found in the ultimate level of achievement made by L2 learners (Savolle-Troike,2008:52):

 

(1) All learners may not have the same degree of access to UG.

(2) Different relationships between various L1s and L2s may result in differential transfer or interference.

(3) Some learners may receive qualitatively different L2 input from others.

(4) Some learners may be more perceptive than others of mismatches between L2 input and existing L1 parameter settings.

(5) Different degrees of specification for lexical features may be achieved by different learners.