Generative transformation grammar

from Wikipedia, the free encyclopedia

The generative transformational grammar ( TG , nor generation grammar ) is a generative grammar with transformation rules . The model was designed by Noam Chomsky in the 1950s and later revised and expanded several times. The discussion about the semantic component within the transformation grammar led to a broad scientific debate (in which George Lakoff appeared as a critic of Chomsky in particular ), in the course of which Chomsky and his colleagues developed their ideas in various versions.

Motivation of the theory

According to Noam Chomsky's theory, generative transformation grammar represents an overcoming of taxonomic structuralism . Chomsky asked how a speaker of any language with a finite number of rules can produce an infinite number of sentences and how a listener understands sentences that he has never heard before Has.

A generative transformation grammar is therefore a model of the description of the dynamic process of speech production and speech reception, the ability of the ideal speaker / listener to generate and understand grammatical expressions ( competence ).

This idea of ​​the ideal speaker is connected with the design of a natural logic . The structural grammar theories were based on Rudolf Carnap, among others . As a representative of logical empiricism , he worked on a logical analysis of language based on the model of physical language, which he regarded as the universal language of science.

Another prerequisite for such an approach is the postulate that the language system in the brain functions similarly to a computer . By Jerry Fodor's view of the diverse structures and meanings may linguistic utterances ( surface structure ) on a grammatical rule apparatus (the deep structures are returned), on the one hand by forming operations ( transforms ) the use of language created (generated) and on the other hand allows understanding. Accordingly, the linguists adopted the mathematical symbols of graph theory - used in computer science - in connection with algorithms for their notations : the basic form for the constituent analysis is the tree graph .

Jerry Fodor describes the abstract basic structures as the language of the mind , which is localized in individual regions of the brain and can be reproduced through causal sequences and rules. Since he, like Chomsky (using the catchphrase “Cartesian Linguistics”) , assumes a genetic disposition, he assumes that every person has this language competence and that it is possible to model a universal basic language across the sub- languages for an ideal speaker / listener. When learning to speak, the child only needs to acquire the lexical units and morphemes and connect them to the structures. Chomsky et al. Implemented this approach in generative transformation grammar : The listener understands the linguistic utterances by deducing (interpreting) the meaning of the sentence from the meaning of the individual components, the formative . The idea of ​​an ideal speaker / listener and the transformation rule system are being challenged by new research in cybernetics and cognitive science .

Chomsky, on the other hand, describes the use of language as performance . The standard version is to him from a generating part (base), the deep structures generated in the transformation part (English z. B., German) according to individual languages different transformations in the surface structures are transferred, thereby a semantic and a phonological interpretation experienced. The basis of this grammar is syntactic. Such a grammar provides for every sentence that it generates (produces), a depth and a surface structure as well as the meaning and the phonetic realization.

The Lakoff variant

→ See also Jerrold Katz : The Semantic Theory 1.1 Prehistory 1.2 The model of semantic interpretation 1.3 Discussion 1.4 Literature

In response to the criticism of the Aspects version , Chomsky expanded his model of transformation grammar to include the semantic component , while Lakoff et al. a. chose a different approach with their generative semantics :

  • In contrast to Chomsky's model, the abstract basic components ( rules of formation ) on which the sentences of a language are based no longer generate syntactic deep structures in Generative Semantics , but semantic sentence representations , also called deep structures (but not in Chomsky's definition), which fully describe the meaning of the sentence. This is done using the smallest meaningful building blocks (= atomic predicates ), which are written in capital letters. That is why generative semantics - in contrast to transformation grammar - do without semantic components operating on the deep structures .
  • Chomsky's transformational grammar, on the other hand, assumes a syntactic component for the derivation stages , which produces the basic component ( formation rules + lexicon ) and the deep structures . This is expanded by the semantic component (= semantic rules ), which generates the semantic representations in the deep structure . The surface structures are created with the help of transformation rules ( transformation components ) .
  • In generative semantics, the transformation component following the semantic sentence representation replaces the abstract atomic predicates , which already carry meaning, with formatives . These only need to be equipped with the previously missing phonological and syntactic properties and produce correctly formed - normal language - syntactic surface structures .

Synopsis as a summary

There is a fundamentally different understanding of the relationship between syntax and semantics.

  • The transformation grammar is made up of two different sets of rules: generative syntax and semantics, which interpret the structures built up by the syntax. Structuring and semantic relationships of linguistic expressions are therefore two different linguistic aspects.
  • In generative semantics, on the other hand, there is no fundamental difference between semantic and syntactic phenomena and therefore only a single semantic set of rules.
  • Proponents of transformation grammar criticize the fact that generative semantics explains the central semantic phenomena not with its rule apparatus, but with the additional postulates of meaning . So it is just an addition to the syntax description.
  • Generative semantics does not organize the tree graph into noun phrases (NP = subject) and verb phrases (VP with V + NP = verb / predicate + object) as immediate constituents of the sentence (S), but rather represents the verb with the NP (subject and object) equal. This gives S a predicate-argument structure as in predicate logic .
  • In addition, generative semantics formalizes sentence requirements ( presuppositions ) and implications according to Gottlob Frege's notations, which cannot be modeled in the generative syntax of transformation grammar.
       /   |   \
    NP     V    NP
   / |     |    | \
Die Katze fraß den Kuchen
  • In contrast to transformation grammar, generative semantics also claims to include linguistic contextual relationships and speaking situations in its model, but it comes up against the limits of mathematical formalizability and is therefore criticized by representatives of linguistic pragmatics .

See also


  • Klaus Baumgärtner, Hugo Steger (Ed.): Funkkolleg Language. An introduction to modern linguistics. Volume 9, pp. 33–38 (criticism of the question and premises of semantic theories within transformation grammar), volume 10, pp. 7 ff. (Criticism of the language theory of transformation grammar). Beltz Weinheim 1972.
  • Noam Chomsky: Syntactic Structure . 1957.
  • Noam Chomsky: aspects of syntactic theory (Translation of: Aspects of the Theory of Syntax, 1965). Frankfurt 1969.
  • Noam Chomsky: Cartesian Linguistics. A chapter in the history of rationalism . Tübingen 1971. Translation (R. Kruse) by: Noam Chomsky: Cartesian linguistics: a chapter in the history of rationalist thought . University Press of America, Lanham, Maryland 1965. Reprint: University Press, Cambridge 2009.
  • Jerrold J. Katz, Jerry A. Fodor: The structure of a semantic theory. In: Hugo Steger (Ed.): Suggestions for a structural grammar of German. Darmstadt 1970, pp. 202-268.
  • George Lakoff: Linguistics and Natural Logic. Frankfurt 1971.
  • Hugo Steger (ed.): Suggestions for a structural grammar of German. (= Paths of Research . Volume 146). Scientific Book Society, Darmstadt 1970.

Web links

Wiktionary: generative transformation grammar  - explanations of meanings, word origins, synonyms, translations
Wiktionary: ideal speaker-listener  - explanations of meanings, word origins, synonyms, translations

Individual evidence

  1. ^ Noam Chomsky: Cartesian Linguistics. A chapter in the history of rationalism . Tübingen 1971. Translation (R. Kruse) by: Noam Chomsky: Cartesian linguistics: a chapter in the history of rationalist thought . University Press of America, Lanham, Maryland 1965. Reprint: University Press, Cambridge 2009.