1. Refer to the two sense involved by the word "hybrid" as used by James.
2. What is the psychological basis of contrastive analysis.
3. State some of the problems in defining S and R in second language acquisition.
4. Provide an account of George's cross-association.
5. What are the weaknesses associated with the ignorance hypothesis.
2. What is the psychological basis of contrastive analysis.
3. State some of the problems in defining S and R in second language acquisition.
4. Provide an account of George's cross-association.
5. What are the weaknesses associated with the ignorance hypothesis.
A number of proposals have been developed in order to account for the interplay of the components which shape a specific language: phonology, morphology, syntax, lexicon, semantics, and pragmatics. Theoretical Linguistics places these components against a general background of language and cognition. Descriptive Linguistics studies how these components are manifested in particular languages, and in order to do so considers data from native speakers. Lastly, Applied Linguistics takes the insights of theorical and descriptive linguistics and applies them to areas such as translation, language instruction, etc.
Contrastive analysis
Contrastive analysis is a linguistic enterprise applied to language instruction. It compares and contrasts two language systems that, for the sake of any meaningful comparison at all, are assumed to have some level of similarity. By noting similarities and differences, contrastive analysis predicts problems in language acquisition and proposes strategies to overcome them.
The three correspondence categories
Contrastive analysis recognizes three categories of correspondences between two languages:
1) Language A has a feature x matched rather closely by x in language B. This equivalence can come about in one of four different ways: a) x may be a universal of language; b) x may be shared because A and B are linguistic relatives; c) x may be shared due to cultural cross-fertilization (directly exchanging some words); d) a coincidence.
2) Language A has a feature x that resembles an x in B to some extent, but differs in details of form, function or usage.
3) Language A has a feature x which B lacks or which can be rendered only in terms of B's, which operates according to different principles.
Pedagogical implications of the correspondence
Any given language A has features of positive transfer value and negative transfer value when compared with a language B. The first are relatively easy to learn (category 1), while the latter tend to pose difficulties in the acquisition of B (categories 2 and 3). Under this premise, CA has arrived to the following conclusions:
1) Distinguishing degrees of difficulty. If, within category 3, A has two different units that in B can be represented through one unit, then B learners will not have much trouble learning that unit. However, sometimes it works the other way around, and when so happens learners will need more practice to learn two units that in their native language are represented by just one unit. >> Negative inversion/comparative.
2) Identifying specif problems. Any given language to be learned as a second language cannot be covered in its entirety. For this reason, CA suggests addressing the points identified as posing special problems, rather than spending time in aspects that the learner could learn intuitively.
3) Determining how a unit works within its own system. Linguists since Saussure have argued that the value of any unit of language must be defined according to the role it plays in its own system. Units that at first glance appear to be equals - and therefore of category 1 - may actually have very differing uses in their specific systems, and belong to category 2 or 3. >> Ratón=mouse (=x) mice / ¿Está lloviendo?=¿Llueve? / Is it raining?=Does it rain?
4) Assigning priorities of the basis of functional load. Though there may be some correspondence between units in A and B, in one of the two languages that unit may be more fundamental, more frequent or more indispensable. > Ordinal numbers / Subjunctive mode.
The limit of linguistics in language pedagogy
The predictions are not always going to be 100% accurate. There is only so much information linguistics in language pedagogy can provide that the many variables particular to each individual student cannot be covered in the attempts to make generalizations about interference between linguistic systems. Among the special problems noted are the following:
1) A category 2 or category 3 aspect of a language might indicate that it might be especially hard to learn for a L2 student, but contrary to these predictions some students will actually absorb the new information with relative ease. >>'s possessive in Spanish nouns/adjectives preceding nouns.
2) Many errors are intralingual rather than interlingual. >> separating subject from predicate with a comma.
3) Instead of projecting L1 patterns into L2, students may just avoid those L2 patters that give them trouble. >> "causative have"/question tags/phrasal verbs.
4) General cognitive processes, social interaction, and even personal idiosyncrasies in learning a language are factors teachers should also consider. These factors are the reason why learners tend to make mistakes at the level of grammar even when the same structures exist in their L1.
The predictions are not always going to be 100% accurate. There is only so much information linguistics in language pedagogy can provide that the many variables particular to each individual student cannot be covered in the attempts to make generalizations about interference between linguistic systems. Among the special problems noted are the following:
1) A category 2 or category 3 aspect of a language might indicate that it might be especially hard to learn for a L2 student, but contrary to these predictions some students will actually absorb the new information with relative ease. >>'s possessive in Spanish nouns/adjectives preceding nouns.
2) Many errors are intralingual rather than interlingual. >> separating subject from predicate with a comma.
3) Instead of projecting L1 patterns into L2, students may just avoid those L2 patters that give them trouble. >> "causative have"/question tags/phrasal verbs.
4) General cognitive processes, social interaction, and even personal idiosyncrasies in learning a language are factors teachers should also consider. These factors are the reason why learners tend to make mistakes at the level of grammar even when the same structures exist in their L1.
A place of CA in Linguists
There are three dimensions or axes in which a linguistic enterprise can be classified into:
1) There are two broad approaches to linguistics, the particularist and the generalist. Particularists focus on individual languages, while generalists consider the general phenomenon of human language.
2) Languages can be studied in isolation with the purpose of discovering the unique features that make it unlike any other language. Conversely, they can be studied in a comparative basis, proceeding from the assumption that all languages have enough in common to be compared and classified into types according to their preferred grammatical devices. Thus, they talk of "synthetic", "analytic", "inflectional", "agglutinating" and "tone" languages.
3) This is a distinction made by De Saussure. He recognizes synchronic and diachronic languages, and explains it as "everything that relates to the static side of our science is synchronic; everything that has to do with evolution is diachronic."
There are three dimensions or axes in which a linguistic enterprise can be classified into:
1) There are two broad approaches to linguistics, the particularist and the generalist. Particularists focus on individual languages, while generalists consider the general phenomenon of human language.
2) Languages can be studied in isolation with the purpose of discovering the unique features that make it unlike any other language. Conversely, they can be studied in a comparative basis, proceeding from the assumption that all languages have enough in common to be compared and classified into types according to their preferred grammatical devices. Thus, they talk of "synthetic", "analytic", "inflectional", "agglutinating" and "tone" languages.
3) This is a distinction made by De Saussure. He recognizes synchronic and diachronic languages, and explains it as "everything that relates to the static side of our science is synchronic; everything that has to do with evolution is diachronic."
"In terms of the three criteria discussed, CA can be defined as a linguistic enterprise aimed at producing inverted, two-valued typologies, and founded on the assumption that languages can be compared."
CA as interlanguage study
Interlanguage study -just like phonetics or dialectology- is yet another branch of Linguistics that studies human language. It is not primarily concerned with languages in the conventional sense, as it is interested in the emergence of the language that comes into being from the interaction of the learner trying to master a L2. Due to its dynamic nature, CA as interlanguage study is viewed as diachronic rather than synchronic. However, its diachronic orientation is sightly different from the mentioned by De Saussure since it reflects the change within the individual rather than its evolution through history.
Aside note: translation theory talks of interlingua and error analysis of approximate systems when referring to the emerging language.
CA as "pure" or "applied" linguistics
As a hybrid discipline, applied linguistics is not only constituted of linguistics but also of psychology and sociology. Such is the case that there is not any single branch of applied linguistics that relies exclusively on "pure" linguistics. Any pure linguistic statement must, therefore, be assessed taking into consideration the insights from these two additional disciplines.
"Pure" linguistics, on the other hand, is interested in discovering and establishing universals of language. To this respect Chomsky holds that "Real progress in linguistics consists in the discovery that certain features of given languages can be reduced to universal properties of language, and explained in terms of these deeper aspects of linguistic form."
The psychological basis of contrastive analysis
CA is a hybrid drawing on the sciences of linguistics and psychology. This is so because linguistics is concerned with the formal properties of language and not directly with learning, which is a psychological matter. Since CA is concerned with L2 learning, it needs a psychological component.
Transfer in learning psychology
Ellis defines transfer as "the hypothesis that the learning of task A will affect the subsequent learning of task B." Because this definition holds true in language learning, it can be argued that transfer theory is the psychological foundation of CA, for gains made in a skill like speaking will affect the gains made in a skill like writing.
On the other hand, the psychological basis of CA are two psychological enterprises: associationism and S-R theory. Associationism has to do with the fact that learning involves the association of two entities (L1 and L2 in CA). These two entities are a stimulus (S) and a response (R).
Some problems of definition
1) What non-verbal learning shares with verbal learning is that both activities involve learning the associations to a certain stimulus. However, what differentiates these two types of learning is that non-verbal learners know how to respond to a stimulus, while verbal learners (L2 learners) have to learn the responses.
2) CA is concerned with teaching rather than learning.
3) The unclear nature of an S or an R in L2 learning. Jackovists considers S as "...the environmental conditions that are antecedent to linguistic utterances." For Richterich S is a "communicative need". Another problem defining S is that language is a two-way process. In Bloomfield's example about Jill and Jack, Jill's speech (her R) becomes in turn an S to Jack.
4) Focusing on form and disregarding substance. There is a one-to-many relationship between sentences and utterances. Linguists may be equipped to describe both form and substance, they can make predictions only about form, but they're utterly unprepared to specify which member of a form-class are likely to appear.
Cross-association by H.V. George
Cross-association (H. V. George) is an alternative to L1 transfer that reconstructs the mental processes of induction and generalization that lead the L2 learner to make mistakes in the second language. George observes that the L1 German learner of English is often subjected to the mistake of referring to "female spouse" by the word woman, on the basis the Frau means woman as well as female spouse in German. George argues that redundancy of the L2 is the direct cause of such errors.
Ignorance hypothesis by Newmark and Reibel
This is another alternative to L1 transfer. Newmark and Reibel say that "The adult can want to say what he does not yet know how to say in the L2, and he uses whatever means he has at his disposal." Selinker has referred to it as a precondition to ignorance.
Ignorance-without-interference: certain structures do not cause problems because hardly ever learners will use it spontaneously (avoidance strategy).
Interference-without-ignorance: Learners will just stumble upon the same errors on repeated occasions even though they are well aware the structure or term they are using is incorrect. They cannot avoid it. (background interference from L2 to L1 will normally have to be accepted as constitution ignorance without ignorance.)
Weaknesses:
1) For the interference theory to function requires the learner to assess his knowledge with a particular target language feature and find it lacking. In other words, he realizes about his ignorance by self-evaluation. But how can he make this decision?
2)
3) Ignorance hypothesis considers learners being called upon to produce L2 patterns which they do not know. When this demand is made the learner cannot by use "whatever means he has at his disposal". Yet, no language-teaching theory has ever imagined asking learners to perform specific L2 items before giving them some reasonable access or exposure to the L2 item in question.
Microlinguistic is the strict description of the linguistic code, without making reference to how messages carried out by this code are modified by the contexts in which they occur. However, more recently there has been a shift of emphasis towards macrolinguistics, which is concerned with the language in a broader sense, taking notice of contextual determination of messages and their interpretation.
CA places either one of these description within the same linguistic framework. That is breaking down the language into three smaller and more manageable areas: phonology, grammar, and lexis, and then describing the language through the linguistic categories of unit, structure, class and system.
*Procedural orientation: In the description of a language through descriptive statements, the phonology is traditionally described before the morphology and the syntax given the fact that it is an easier and more finite system to describe.
*Mixing words: It used to be considered a good practice not mix the descriptive levels with each other (a statement about phonology should only be concerned about that and nothing else). Nowadays, mixing is allowed, and in some cases even necessary.
Any CA first describes each of the two language on the appropriate level, and then goes on to a juxtaposition stage for comparison, where it is sometimes necessary to cross levels. The degree to which it is necessary to cross levels is good indication of the degree of interlingual non-correspondence of the two languages involved.
> We knew where it was [Sabíamos donde estaba.]
> We found out where it was [Supimos donde estaba.]
What is a lexical distinction in English is expressed through morphological contrast in Spanish (an interlingual level shift from lexis to grammar).
Categories of grammar
UNIT: Arranged from the largest to the smallest, the units are sentence, clause, phrase, word, and morpheme. Each one of these units consists of the unit immediately below them. This order of direct inclusion is called rank scale. In traditional linguistics, a single sentence in L1 will always correspond on a 1:1 basis with a single sentence in L2. There may be however, for example, a 2:1 clause correspondence (a 2:1 interlingual rank shift).
STRUCTURE: Halliday defines it as an arrangement of elements ordered in places. These elements are the subject, predicator, complement and adjunct.
CLASS: Class regards the restrictions on which certain structures can only be composed of certain units. For example, a predicator can only be made with a verb phrase.
SYSTEM: System regards the possibility of choices the speaker has at his disposal when composing a sentence. By choice it is meant "the selection of one particular place on the chain in preference to another term or other terms which are also possible at that place." For example, on the slot of a predicator, we must use a verb phrase, but we are free to choose between past and present tense forms, and between perfect and non-perfect, as well as between progressive and non-progressive forms.
Structural model
Structuralists base their analysis on the claim that any grammatical construction consisting of more than one element can be reduced to a pair of constituents. Accordingly, the technique they use to this end is known as Immediate Constituent analysis.
Such analysis takes no account of the meaning of constructions; rather, it depends on the notion of distribution (or what normally goes with what), and omissibility (the favoring of syntactically-sound constructions over non-constructions).
Furthermore, this type of analysis is structured on two axes: the syntagmatic one, which delineates construction types, and the paradigmatic one, which defines sets of possible fillers for each position.
There are, however, certain weakness that the drawing of an IC boundary cannot account for. For example:
a) She is a beautiful dancer.
b) I want that so bad.
In a) we cannot tell for certain whether the sentence means she is beautiful as a dancer or as a person. In b), given the IC's disregard for colloquial usage of the language (descriptive grammar), IC will expect the verb want to be modified by an adverb (badly), and therefore find a non-construction in bad.
This weakness stems from the fact that IC analysis only considers the surface structure - the relative position of elements in the structure -, and not the deep structure -the functional relations between constituents-.
How to proceed with a structuralist CA
There are three kinds of formal devices operating at the level of grammar: morphological markers, function words, and suprasegmentals. To conduct a CA, we first inquire whether the two languages employ the same formal devices. It is possible to find cases in which L1 carries a certain meaning by one device, while L2 conveys the same meaning by another device. When there are contrasts emanating from the preference by each language for a different medium of grammar, we talk about the existence of a medium shift.
Transformational-Generative Grammar
The most notable feature of Transformational-Generative Grammar is that it recognizes a level of deep structure and a level of surface structure which are related by sets of transformations. The term "generative" combines two senses: (1) it is understood to be "projective" (or predictive) because it projects a corpus of sentences upon a larger set of sentences, and because it establishes as grammatical not only actual sentences but also potential sentences; (2) it is "explicit" because it says which sentences are possible in the language by specifying them: ungrammatical ones are omitted.
Reasons for using TGG
(1) First, its explicitness; for each step in deriving surface from deep structures an explicit rule must be formulated.
(2) Second, a reason particularly attractive to CA is that it has been claimed that deep structures are "universal."
(3) Third, the transformations applied to deep structures are taken from a universal stock, which Chomsky call the "formal universals".
CA in the framework of TGG
It is standard practice to move predicative adjectives to a attributive position. However, as opposed to languages like English or German, this does not apply to Spanish or French because they do not allow adjective shifting. In passing from deep structure to surface structure three transformations are involved:
(a) relativization
(b) wh/relative deletion
(c) adjective shifting
Deep structure >> I have a house + The house is green.
(a) I have a house which is green. [relativization]
(b) I have a house green. [wh deletion]
(c) I have a green house. [adjective shifting] << Surface structure
This TGG approach provides the contrastive analyst a kind measure of degree of difference between compared constructions in L1 and L2. Deep structure is common to all languages; where they mostly differ is in the surface structures.
Structuralist model vs TGG in CA
Structuralists look for surface-structural correspondence, whereas TGG linguists look for correspondence between transformational rules.
Types of difference in rule application
Although according to Marton a rule of L1 can perfectly correspond with a rule of L2, the contrastive analyst is more interested in how rules differ in their applicability to congruent deep structures of two languages. There are several types of difference in rule application:
(1) One of the languages applies the rules, whereas the other either does not, or does so less generally. >> in converting a deep structure sentence into a surface structure sentence, in Spanish as opposed to English there is no adjective shifting.
(2) In language a rule is obligatory but in another one it is optional. >> The insertion of an object relative pronoun is mandatory in Spanish but not in English.
(3) ---
(4) Some rules imply others. >> The rule known as raising generates structures which can undergo passivization in English: (a) They believe that John is a clever boy; (b) They believe John to be a clever boy; (c) John is believed {by them} to be a clever boy.
* What are the salient features of TGG?
* What is the idea conveyed by the term "generative"?
* State some reasons for using TGG in CA.
* How can sentential ambiguities be explained?
* There are several types of difference in rule application. Mention at least three. Exemplify.
Interlanguage study -just like phonetics or dialectology- is yet another branch of Linguistics that studies human language. It is not primarily concerned with languages in the conventional sense, as it is interested in the emergence of the language that comes into being from the interaction of the learner trying to master a L2. Due to its dynamic nature, CA as interlanguage study is viewed as diachronic rather than synchronic. However, its diachronic orientation is sightly different from the mentioned by De Saussure since it reflects the change within the individual rather than its evolution through history.
Aside note: translation theory talks of interlingua and error analysis of approximate systems when referring to the emerging language.
CA as "pure" or "applied" linguistics
As a hybrid discipline, applied linguistics is not only constituted of linguistics but also of psychology and sociology. Such is the case that there is not any single branch of applied linguistics that relies exclusively on "pure" linguistics. Any pure linguistic statement must, therefore, be assessed taking into consideration the insights from these two additional disciplines.
"Pure" linguistics, on the other hand, is interested in discovering and establishing universals of language. To this respect Chomsky holds that "Real progress in linguistics consists in the discovery that certain features of given languages can be reduced to universal properties of language, and explained in terms of these deeper aspects of linguistic form."
The psychological basis of contrastive analysis
CA is a hybrid drawing on the sciences of linguistics and psychology. This is so because linguistics is concerned with the formal properties of language and not directly with learning, which is a psychological matter. Since CA is concerned with L2 learning, it needs a psychological component.
Transfer in learning psychology
Ellis defines transfer as "the hypothesis that the learning of task A will affect the subsequent learning of task B." Because this definition holds true in language learning, it can be argued that transfer theory is the psychological foundation of CA, for gains made in a skill like speaking will affect the gains made in a skill like writing.
On the other hand, the psychological basis of CA are two psychological enterprises: associationism and S-R theory. Associationism has to do with the fact that learning involves the association of two entities (L1 and L2 in CA). These two entities are a stimulus (S) and a response (R).
Some problems of definition
1) What non-verbal learning shares with verbal learning is that both activities involve learning the associations to a certain stimulus. However, what differentiates these two types of learning is that non-verbal learners know how to respond to a stimulus, while verbal learners (L2 learners) have to learn the responses.
2) CA is concerned with teaching rather than learning.
3) The unclear nature of an S or an R in L2 learning. Jackovists considers S as "...the environmental conditions that are antecedent to linguistic utterances." For Richterich S is a "communicative need". Another problem defining S is that language is a two-way process. In Bloomfield's example about Jill and Jack, Jill's speech (her R) becomes in turn an S to Jack.
4) Focusing on form and disregarding substance. There is a one-to-many relationship between sentences and utterances. Linguists may be equipped to describe both form and substance, they can make predictions only about form, but they're utterly unprepared to specify which member of a form-class are likely to appear.
Cross-association by H.V. George
Cross-association (H. V. George) is an alternative to L1 transfer that reconstructs the mental processes of induction and generalization that lead the L2 learner to make mistakes in the second language. George observes that the L1 German learner of English is often subjected to the mistake of referring to "female spouse" by the word woman, on the basis the Frau means woman as well as female spouse in German. George argues that redundancy of the L2 is the direct cause of such errors.
Ignorance hypothesis by Newmark and Reibel
This is another alternative to L1 transfer. Newmark and Reibel say that "The adult can want to say what he does not yet know how to say in the L2, and he uses whatever means he has at his disposal." Selinker has referred to it as a precondition to ignorance.
Ignorance-without-interference: certain structures do not cause problems because hardly ever learners will use it spontaneously (avoidance strategy).
Interference-without-ignorance: Learners will just stumble upon the same errors on repeated occasions even though they are well aware the structure or term they are using is incorrect. They cannot avoid it. (background interference from L2 to L1 will normally have to be accepted as constitution ignorance without ignorance.)
Weaknesses:
1) For the interference theory to function requires the learner to assess his knowledge with a particular target language feature and find it lacking. In other words, he realizes about his ignorance by self-evaluation. But how can he make this decision?
2)
3) Ignorance hypothesis considers learners being called upon to produce L2 patterns which they do not know. When this demand is made the learner cannot by use "whatever means he has at his disposal". Yet, no language-teaching theory has ever imagined asking learners to perform specific L2 items before giving them some reasonable access or exposure to the L2 item in question.
Unit 4
Microlinguistics and macrolinguisticsMicrolinguistic is the strict description of the linguistic code, without making reference to how messages carried out by this code are modified by the contexts in which they occur. However, more recently there has been a shift of emphasis towards macrolinguistics, which is concerned with the language in a broader sense, taking notice of contextual determination of messages and their interpretation.
CA places either one of these description within the same linguistic framework. That is breaking down the language into three smaller and more manageable areas: phonology, grammar, and lexis, and then describing the language through the linguistic categories of unit, structure, class and system.
*Procedural orientation: In the description of a language through descriptive statements, the phonology is traditionally described before the morphology and the syntax given the fact that it is an easier and more finite system to describe.
*Mixing words: It used to be considered a good practice not mix the descriptive levels with each other (a statement about phonology should only be concerned about that and nothing else). Nowadays, mixing is allowed, and in some cases even necessary.
Any CA first describes each of the two language on the appropriate level, and then goes on to a juxtaposition stage for comparison, where it is sometimes necessary to cross levels. The degree to which it is necessary to cross levels is good indication of the degree of interlingual non-correspondence of the two languages involved.
> We knew where it was [Sabíamos donde estaba.]
> We found out where it was [Supimos donde estaba.]
What is a lexical distinction in English is expressed through morphological contrast in Spanish (an interlingual level shift from lexis to grammar).
Categories of grammar
UNIT: Arranged from the largest to the smallest, the units are sentence, clause, phrase, word, and morpheme. Each one of these units consists of the unit immediately below them. This order of direct inclusion is called rank scale. In traditional linguistics, a single sentence in L1 will always correspond on a 1:1 basis with a single sentence in L2. There may be however, for example, a 2:1 clause correspondence (a 2:1 interlingual rank shift).
STRUCTURE: Halliday defines it as an arrangement of elements ordered in places. These elements are the subject, predicator, complement and adjunct.
CLASS: Class regards the restrictions on which certain structures can only be composed of certain units. For example, a predicator can only be made with a verb phrase.
SYSTEM: System regards the possibility of choices the speaker has at his disposal when composing a sentence. By choice it is meant "the selection of one particular place on the chain in preference to another term or other terms which are also possible at that place." For example, on the slot of a predicator, we must use a verb phrase, but we are free to choose between past and present tense forms, and between perfect and non-perfect, as well as between progressive and non-progressive forms.
Unit 5
Structural model
Structuralists base their analysis on the claim that any grammatical construction consisting of more than one element can be reduced to a pair of constituents. Accordingly, the technique they use to this end is known as Immediate Constituent analysis.
Such analysis takes no account of the meaning of constructions; rather, it depends on the notion of distribution (or what normally goes with what), and omissibility (the favoring of syntactically-sound constructions over non-constructions).
Furthermore, this type of analysis is structured on two axes: the syntagmatic one, which delineates construction types, and the paradigmatic one, which defines sets of possible fillers for each position.
There are, however, certain weakness that the drawing of an IC boundary cannot account for. For example:
a) She is a beautiful dancer.
b) I want that so bad.
In a) we cannot tell for certain whether the sentence means she is beautiful as a dancer or as a person. In b), given the IC's disregard for colloquial usage of the language (descriptive grammar), IC will expect the verb want to be modified by an adverb (badly), and therefore find a non-construction in bad.
This weakness stems from the fact that IC analysis only considers the surface structure - the relative position of elements in the structure -, and not the deep structure -the functional relations between constituents-.
How to proceed with a structuralist CA
There are three kinds of formal devices operating at the level of grammar: morphological markers, function words, and suprasegmentals. To conduct a CA, we first inquire whether the two languages employ the same formal devices. It is possible to find cases in which L1 carries a certain meaning by one device, while L2 conveys the same meaning by another device. When there are contrasts emanating from the preference by each language for a different medium of grammar, we talk about the existence of a medium shift.
Unit 6
The most notable feature of Transformational-Generative Grammar is that it recognizes a level of deep structure and a level of surface structure which are related by sets of transformations. The term "generative" combines two senses: (1) it is understood to be "projective" (or predictive) because it projects a corpus of sentences upon a larger set of sentences, and because it establishes as grammatical not only actual sentences but also potential sentences; (2) it is "explicit" because it says which sentences are possible in the language by specifying them: ungrammatical ones are omitted.
Reasons for using TGG
(1) First, its explicitness; for each step in deriving surface from deep structures an explicit rule must be formulated.
(2) Second, a reason particularly attractive to CA is that it has been claimed that deep structures are "universal."
(3) Third, the transformations applied to deep structures are taken from a universal stock, which Chomsky call the "formal universals".
CA in the framework of TGG
It is standard practice to move predicative adjectives to a attributive position. However, as opposed to languages like English or German, this does not apply to Spanish or French because they do not allow adjective shifting. In passing from deep structure to surface structure three transformations are involved:
(a) relativization
(b) wh/relative deletion
(c) adjective shifting
Deep structure >> I have a house + The house is green.
(a) I have a house which is green. [relativization]
(b) I have a house green. [wh deletion]
(c) I have a green house. [adjective shifting] << Surface structure
This TGG approach provides the contrastive analyst a kind measure of degree of difference between compared constructions in L1 and L2. Deep structure is common to all languages; where they mostly differ is in the surface structures.
Structuralist model vs TGG in CA
Structuralists look for surface-structural correspondence, whereas TGG linguists look for correspondence between transformational rules.
Types of difference in rule application
Although according to Marton a rule of L1 can perfectly correspond with a rule of L2, the contrastive analyst is more interested in how rules differ in their applicability to congruent deep structures of two languages. There are several types of difference in rule application:
(1) One of the languages applies the rules, whereas the other either does not, or does so less generally. >> in converting a deep structure sentence into a surface structure sentence, in Spanish as opposed to English there is no adjective shifting.
(2) In language a rule is obligatory but in another one it is optional. >> The insertion of an object relative pronoun is mandatory in Spanish but not in English.
(3) ---
(4) Some rules imply others. >> The rule known as raising generates structures which can undergo passivization in English: (a) They believe that John is a clever boy; (b) They believe John to be a clever boy; (c) John is believed {by them} to be a clever boy.
* What are the salient features of TGG?
* What is the idea conveyed by the term "generative"?
* State some reasons for using TGG in CA.
* How can sentential ambiguities be explained?
* There are several types of difference in rule application. Mention at least three. Exemplify.
Unit 7
Contrastive Generative Grammar
Krzeszowski's CGG is different from the Structuralist and the TGG models in the sense that it is a vertical, single-phased CA in which L1 and L2 are generated from some common base and then compared and contrasted during this process of generation. Its defining characteristics are:
(i) It is not based on the confluence of two monolingual grammars (as classical CA is), but is a single bilingual grammar. K. justifies this by saying that the function of CA is to provide an account of the intuitions of an "ideal" bilingual about the relatedness of his two languages. James does not agree with K. on this regard for the reason that a balanced bilingual has solved the L1:L2 mismatch, which essentially means he has solved the very problems CA addresses itself to.
(ii) CGG proceeds by deriving in five stages universal semantic INPUTS to language specific surface structure OUTPUTS: [INPUT] Semantic > Categorical > Syntactic > Lexical > Post-lexical [OUTPUT]
Krzeszowski's CGG is different from the Structuralist and the TGG models in the sense that it is a vertical, single-phased CA in which L1 and L2 are generated from some common base and then compared and contrasted during this process of generation. Its defining characteristics are:
(i) It is not based on the confluence of two monolingual grammars (as classical CA is), but is a single bilingual grammar. K. justifies this by saying that the function of CA is to provide an account of the intuitions of an "ideal" bilingual about the relatedness of his two languages. James does not agree with K. on this regard for the reason that a balanced bilingual has solved the L1:L2 mismatch, which essentially means he has solved the very problems CA addresses itself to.
(ii) CGG proceeds by deriving in five stages universal semantic INPUTS to language specific surface structure OUTPUTS: [INPUT] Semantic > Categorical > Syntactic > Lexical > Post-lexical [OUTPUT]
- Emphasizing on language neutrality, CGG's process starts on a "universal semantic", which is conceptual input based on primitive notions such as agent, patient, and all sorts of specifications of location in time and space.
- Each language categorizes the input from stage 1 into Hilliday's unit, structure, class and system. Some categories are language specific and some are not.
- The categories are syntactically arranged into permissible order in actual sentences. Function words are introduced here, which K calls "minor lexicalizations."
- Lexical elements from the dictionary are inserted into the syntactic frames specified in stage 3. This is "major lexicalization."
- Post-lexical or "cosmetic" transformations are applied, providing outputs with inflections and markers.
CGG and learner-strategies
Several applied linguists have said that learners, irrespectively of their L1 or their target language, tend to produce simple versions of the language they are learning. Initially learners reduce L2 to its bare communicative essentials, and once they do this they embark on the process of re-elaboration, adding specific features of the particular L2.
K. claims that CGG can account for both these processes of simplification and re-elaboration.
CGG provides a five-point scale to locate any manifestation of a simplified utterance. For example, native speaker's utterances have passed through all five stages of CGG, while the degree of simplicity with which learners communicate can be measured in terms of which of the later stages have been left out. A sentence like "me cold," with no proper use of the pronoun, no verb (and no tense) fails to go beyond the third stage of syntactic configuration. The point of elaboration starts at stage 3 where minor lexicalizations are introduced.
James's vision on CGG
Even though James does not agree with K. in how he describes the salient features of CGG, he points out that this model can account for some facts that are being discovered about the learning of a second language, but because it can't explain certain phenomena CGG has to be approached with caution.
