1.1 32 info:srw/schema/1/dc-v1.1 XML Authors/Titles Chomsky Noam Chomsky<br /> Literatur<br /> Chomsky I<br /> Noam Chomsky<br /> "Linguistics and Philosophy", in: Language and Philosophy, (Ed) Sidney Hook New York 1969 pp. 51-94<br /> In<br /> Linguistik und Philosophie, G. Grewendorf/G. Meggle Frankfurt/M. 1974/1995<br /> <br /> Chomsky II<br /> Noam Chomsky<br /> "Some empirical assumptions in modern philosophy of language" in: Philosophy, Science, and Method, Essays in Honor of E. Nagel (Eds. S. Morgenbesser, P. Suppes and M- White) New York 1969, pp. 260-285<br /> In<br /> Linguistik und Philosophie, G. Grewendorf/G. Meggle Frankfurt/M. 1974/1995<br /> <br /> Chomsky IV<br /> N. Chomsky<br /> Aspekte der Syntaxtheorie Frankfurt 1978<br /> <br /> Chomsky V<br /> N. Chomsky<br /> Language and Mind Cambridge 2006 1 info:srw/schema/1/dc-v1.1 XML Active/passive Chomsky Lyons I 258<br /> Active/Passive/Transformational Grammar/Chomsky/Lyons: although subject and object are interchanged, identity or similarity prevails between the two corresponding sentences in the deep structure. However, this is also a prerequisite for determining that subject and object can be exchanged. <br /> Problem: there is disagreement about whether dissimilation prevails here or not. <br /> For example, suppose that "the shooting of the hunters" is not ambiguous. <br /> Problem: then we would still require the grammar to be written in the following way that<br /> a) between "the shooting of the hunters" and the transitive sentence NP1 "shoot the hunters", and <br /> b) between "the hunters shooting" and the intransitive "the hunters shoot" relationships are established. <br /> Lyons I 261<br /> Active/Passive/Transformational Grammar/Chomsky/Lyons: (N. Chomsky, Syntactic Structures, Berlin, New York 1957)<br /> Passive (optional)<br /> <br /> Structural descriptions: NP – Aux – V NP<br /> Structural change: X1 – X2 - X3 – X4 > X4 – X2 + be + en – V – by + X1<br /> <br /> (Notation: concatenation: sometimes „+“, sometimes “–„ (is not explained here). <br /> different: (formally more precise): <br /> <br /> NP1 – Aux – V – NP2 > Aux + be + en – V – by + NP1<br /> <br /> T rule: contains two parts: the description (analysis, notation: SA, SB) and the change (structural change, notation: SC, SV). <br /> By definition, T-rules are only effective in chains that can be analyzed by means of the elements indicated in their structural description. <br /> Transformation: its result is exactly what has already been described in the alternative representation of the rule: <br /> <br /> NP1 – Aux – V – NP2 > Aux + be + en – V – by + NP1<br /> <br /> What does it mean that the chain can be analyzed using four elements (NP, Aux, V and NP)?<br /> I 262<br /> The following chains resulted from the rules: <br /> from (1): NP + VP<br /> (2): NP + Verb + NP<br /> (3) : NP sing + Verb + NP sing<br /> (4): T + N + 0 + Verb + T + N + 0<br /> (6): T + N + 0 + Aux+ V + T + N + 0<br /> (7) T + N + 0 + C + M + have + en + V + T ü N + 0.<br /> <br /> Rule (3) n d(4) was applied twice (4), because NP sing f both positions were selected in the output of rule (2). <br /> Rule (5) was not applicable. <br /> Rule (7): Au has been replaced by C + M + have + en. <br /> The edition of (7) is a core chain which is underlied by the type of corresponding active and passive sentences, e.g. "The man will have read the book" and "The book will have been read by the man".<br /> Passive transformation: now we apply them to the chain: where none of the elements specified in the structural description with respect to the passive transformation occur in the core chain. <br /> Furthermore, we did not come across the chain NP + Aux + V + Np at any stage of deriving the core chain through the PS rules. Therefore, we review the rules again to create the constituent structure of the desired core chain: <br /> <br /> By rule<br /> (1): ∑ (NP + VP)<br /> (2): ∑ (NP + VP)(Verb + NP))<br /> (3): ∑ (NP (NP sing) + VP(Verb + NP(NP sing)))<br /> (4): ∑ (NP (NP sing (T + N + 0)) + VP(Verb + NP(NP sing(T + N + 0))))<br /> (6): ∑ (NP (NP sing (T + N + 0)) + VP(Verb (Aux + V)+ NP(NP sing(T + N + 0))))<br /> (7): ∑ (NP (NP sing (T + N + 0)) + VP(Verb (Aux (C + M + have + en) + V)+ NP(NP sing(T + N + 0)))).<br /> This is the constituent structure of sentences that is the basis of sentences such as "The man will have read the book" and "The book will have been read by the man." (active/passive). <br /> Lyons I 262<br /> Definition Phrase Marker/P-Marker/Grammar/Chomsky/Lyons: if a chain is represented with constituent parentheses and parentheses indices (labelled-bracketing), this is referred to as a formation marker or P-marker. <br /> Definition parenthesis index: labelled-bracketing/terminology: Designation of a node in the tree diagram or symbol in front of a parenthesis. <br /> I 263<br /> Definition Dominate/Dominance/Chomsky/Lyons: a symbol dominates an entire parentheses expression when the parentheses in the P marker is opened immediately after this symbol. In the tree diagram: The symbol dominates everything that is derived from the node indicated by the symbol. <br /> Definition (structural) analyzability/grammar/Chomsky/Lyons: (is a condition for the application of T-rules): if a chain without residual elements can be broken down into subchains, each of which is dominated by a symbol given in the structural description of the T-rules, then the chain satisfies the conditions defined by the structural description (SB).<br /> <br /> Passive transformation/Chomsky/Lyons: (is optional) and looks like this: <br /> <br /> {T + N + 0} + {C + M + have + en} + {V} + {T + N + 0}<br /> <br /> NP1 - Aux - V - NP2<br /> <br /> Transformation: due to the operation of the actual T-Rule (in the structural change), a further chain (no more core chain) results as output, which then serves with its P-marker as input for further T-Rules. 2 info:srw/schema/1/dc-v1.1 XML Innateness Chomsky Katz I 380<br /> Innate ideas are the crux of the dispute between empiricists and rationalists. 3 info:srw/schema/1/dc-v1.1 XML Thinking Chomsky McGinn I 19f<br /> Chomsky thesis: richness in one direction is accompanied by indigence in the other direction, and vice versa. McGinn per. 4 info:srw/schema/1/dc-v1.1 XML Simplicity Chomsky I 295<br /> Simplicity/Chomsky: mathematics has no concept of the "easiness" of a calculation - therefore one needs an optimal computational procedure.<br /> ---<br /> I 305 <br /> Chomsky refers to choice between grammars, not between theories - E of G would be empirically like a physical constant. 5 info:srw/schema/1/dc-v1.1 XML Explanation Chomsky Strawson VI 392<br /> Explanation/Chomsky: Chomsky admits that an "explanatory adequate" grammar does not have to be "descriptive-adequate" - we need a theory of linguistic universals and why our grammar was selected. 6 info:srw/schema/1/dc-v1.1 XML Forms Chomsky I 310<br /> Chomsky: interested in whether there are ideas and principles which determine the form of the knowledge acquired (rationalist variant) or (Vs) whether the structure of the acquisition mechanism is limited to simple processing mechanisms (empiricist version) - Empiricism/Chomsky: is not that important. 7 info:srw/schema/1/dc-v1.1 XML Grue Chomsky I 290<br /> Grue/ChomskyVsGoodman: Grueness is amarginal problem - the initial question is much too vague. - You can easily find a property of language "grue bleen" which is not a property of a "languange like German" - e.g. the predicate "being similar", only applied to objects rather than to qualia.<br /> Chomsky: there is no point in time t such that we can predict of objects that they will not be similar - they could be the similar if both were green - it is a property of natural languages ​​that they behave more like German than like "grue bleen" - but language concepts such as "German" are too vague to satisfy Goodman’s criterion - we cannot explain why the learner does not acquire grue as basis for generalisation - that certainly follows from the sensory system. 8 info:srw/schema/1/dc-v1.1 XML Grammar Chomsky Searle VIII 414<br /> ChomskyVsStructuralism: phrase structure rules alone cannot resolve ambiguities - e.g. Active/Passive - Solution/Chomsky: transformation rules, transformation phrase markers by permutation, insertion, eradication of elements in other phrase markers - then the syntax consists of two components: base and transformation.<br /> ---<br /> VIII 418<br /> Deep structure/Chomsky: determines the meaning - Surface structure: determines the phonetic form (late works: sometimes the meaning) - Syntax/Chomsky: is to be separated from semantics - (according to Searle): man is a syntactic creature, the brain is syntactic.<br /> ---<br /> VIII 421 <br /> SearleVsChomsky: from this it would follow that if one day we had syntactically modified forms, we would have no language anymore, but something else.<br /> ---<br /> VIII 421<br /> Generative grammar/NeogrammariansVsChomsky: semantics crucial for the formation of syntactic structures. 9 info:srw/schema/1/dc-v1.1 XML Ideas Chomsky Katz II 380<br /> Innate ideas are the crux of the dispute between empiricists and rationalists. 10 info:srw/schema/1/dc-v1.1 XML Competence Chomsky I 307<br /> Competence/ChomskyVsHarman: I do not claim that they consist in "knowing-that", that language is described by the rules of grammar - Competence/ChomskyVsHarman: not a number of habits, no reference to the ability of the cyclist - instead the mastery of generative grammar - (non-formulated knowledge) - less than the ability to speak a language.<br /> ---<br /> Searle VIII 404<br /> Competence/performance/Chomsky: Thesis: performance is just the peak of the iceberg of competence.<br /> ---<br /> VIII 437 <br /> SearleVsChomsky: the distinction is wrong: he assumes that a theory of speech acts must be more like a theory of performance than one of competence - he does not see that ultimately competence is a performance competence - ChomskyVsSpeech act theory: suspects behaviorism behind it. SearleVs: not true, because speech act theory involves intention.<br /> ---<br /> Searle VIII 409<br /> Chomsky: new: object of study is the language skills - old: random number of sentences, classifications. ChomskyVsStructuralism: a theory must be able to explain which chains represent sentences and which do not.<br /> --- <br /> VIII 414 <br /> SearleVsChomsky: not clear how the grammatical theory provides the knowledge of the speaker. 11 info:srw/schema/1/dc-v1.1 XML Constants Chomsky Lyons I 266<br /> Constituent Structure Rule/Transformation Rule/Constant/Variable/Chomsky/Lyons: if the formalization of a system requires that the T-rules should come according to the PS-rules (phrase structure rules) (as in N. Chomsky, Syntactic Structures, Berlin, New York 1957), a constituent structure rule would have to be reclassified as a T-rule. This is possible because a constant can always be considered as a variable with only one value. <br /> This again demonstrates the heterogeneity of the T-rules. 12 info:srw/schema/1/dc-v1.1 XML Criteria Chomsky II 345<br /> Criteria/mental states/Wittgenstein: mental states or the "inner workings of the mind" do not provide a criterion for the proper use of an expression.<br /> II 346<br /> ChomskyVsWittgenstein: here it is not about a "real statement" e.g. if someone reads something, but about a legitimate claim - e.g. mirage: can provoke a legitimate (incorrect) assertion. 13 info:srw/schema/1/dc-v1.1 XML Artificial Intelligence Chomsky Putnam III 29/30<br /> Artificial Intelligence/Chomsky: Chomsky believes that the computer model of the mind is right, but doubts the success of the artificial intelligence. 14 info:srw/schema/1/dc-v1.1 XML Method Chomsky I 278<br /> Method/theory/Chomsky: requirement; we must be able to describe what the person receives - the percept itself is a construction of the first order - its properties are determined experiment. Grammar: construction of the second-order - for this one must abstract from the other factors involved in the use and understanding of language and refer to internalized knowledge of the speaker - VsBehaviorismus: excludes the concept of "what is perceived" and "what is learned" from the outset.<br /> I 297ff<br /> Method/theory: PutnamVsChomsky: certain ambiguities can only be discovered through routine, therefore their postulated explanation by Chomsky's grammar is not that impressive - ChomskyVsPutnam: he misunderstands it, in fact this refers to competence and not to performance - routine does not matter here, but the inherent correlation between sound and meaning.<br /> I 303 <br /> Chomsky: my universal grammar is not a "theory of language acquisition", but one element of it - my thesis is an "all-at-once" proposal and does not try to capture the interplay between the tentative hypotheses constructed by the child and new data interpreted with them.<br /> ---<br /> II 316<br /> Method/theory/Chomsky: "association", "reinforcement", "random mutation ": hide our ignorance - (s) something dissimilar may also be associated.<br /> II 321<br /> Method/theory/ChomskyVsQuine: his concept of "reinforcement" is almost empty - if reinforcement is needed for learning, it means that learning cannot happen without data.<br /> II 323 <br /> Language Learning/ChomskyVsQuine: he does not explain it: if only association and conditioning, then the result is merely a finite language.<br /> II 324 <br /> VsQuine: concept of probability of a sentence is empty: the fact that I utter a particular German sentence is as unlikely as a particular Japanese sentence from me. 15 info:srw/schema/1/dc-v1.1 XML Surface Structure Chomsky I 269F <br /> Surface Structure/Chomsky: finding a hierarchy of phrases that belong to certain categories: noun phrase, verb phrase, adjective phrase, etc. E.g. John is certain Bill wants to leave - John is certain to leave: similar surface structure, different deep structure. <br /> ---<br /> I 273 <br /> Surface Structure/Chomsky: Assumption: it contributes nothing to the meaning - what contribution a term makes to the sentence, is adjusted by the deep structure (> <a href="https://philosophy-science-humanities-controversies.com/listview-list.php?concept=Compositionality">compositionality</a>). <br /> ChomskyVsAnalytic Philosophy: if different intensions after substitution should change the meaning, there would have to be a corresponding difference in the deep structure, which is unlikely. 16 info:srw/schema/1/dc-v1.1 XML Rules Chomsky Lyons I 259<br /> Phrase structure rules/PS-Rules/Chomsky: (N. Chomsky, Syntactic Structures, Berlin, New York 1957):<br /> (1) S > NP + VP<br /> (2) VP > Verb + NP<br /> (3) NP > NP sing<br /> or<br /> NP plur<br /> (4) NP sing > T * N + 0 (null)<br /> (5) NP plur > T + N + s<br /> (6) Verb > Aux + V<br /> (7) Aux > C(M) (have + en) (be + ing) (M = auxiliary)<br /> <br /> I 260<br /> Lexical substitution/Lyons: we have ignored it here, but instead <br /> Lexicon/Chomsky: Example<br /> <br /> T = {the}<br /> N = {man, ball,…}<br /> V = {hit, take, walk, read,…}<br /> M = {will, can, shall, must}<br /> <br /> Rule (7): develops Aux into a chain with up to four elements. Of these, C is obligatory, (his task to regulate congruence relations is interpreted by a transformation rule). The other rules on the right side of (7) are optional. Each element is independent of the other two. <br /> The output of rule (7) will therefore contain one of the following eight chains (in Aux's position in the output of rule (6)):<br /> <br /> (1) C<br /> (2) C + M<br /> (3) C + have + en<br /> (4) C + be + ing<br /> (5) C + M + have + en<br /> (6) C + M + be + ing<br /> (7) C + have + en + be + ing<br /> (8) C + M + have + en + be + ing<br /> <br /> Definition Core Chain/Transformational Grammar/Chomsky/Lyons: this is the output of phrase structure rules (kernel string). <br /> Definition core sentence/kernel sentence: is any sentence created from a single core chain without the use of any optional transformations. <br /> <br /> Transformation rules/Chomsky: no sentence is created without applying at least a limited number of mandatory T-rules. It is wrong (how often assumed) that core sentences would be generated by phrase structure rules alone. <br /> I 264<br /> T-rules/Chomsky/Lyons: are heterogeneous. The construction of a partial transformational grammar for a language itself is an enormously complex matter. Many factors influence the decision whether to apply a T-rule or a basic rule (Phrase structure rule, constituent structure rule) to a certain element. <br /> T-rules: not all of them are transformational by their very nature!<br /> Transformational: two criteria for "inherent" transformational rules:<br /> 1. Any rule that does not meet the conditions imposed by a constituent structure rule is transformational. <br /> I 265<br /> 2. Definition T-rule: is the one that in the symbol chain to the left of the replacement arrow contains at least one symbol that acts as a variable, which assumes as a value any one of the whole class of subchains that are dominated by this symbol in the P-marker. This P-marker belongs to the chain serving as a rule input. <br /> For example, all symbols (except V) are in this sense variables in the structural descriptions of the above passive transformation: <br /> <br /> SB: NP - Aux – V – NP<br /> <br /> V: on the other hand, is a constant: it is an end symbol that does not dominate any substring other than itself ("self-dominance" is therefore a formal requirement of the system). <br /> Transformation/Chomsky: the difference between constant and variable is fundamental for its definition. 17 info:srw/schema/1/dc-v1.1 XML Semantics Chomsky I 272 <br /> Semantics/Chomsky: here the surface structures hardly help, the deep structure helps even more. <br /> --- <br /> Strawson V 393 <br /> StrawsonVsChomsky: hardly deals with semantics - its lexicon contains much fewer entries than our dictionaries. 18 info:srw/schema/1/dc-v1.1 XML Language Chomsky I 279ff<br /> Language/Chomsky: apart from its mental representation, it has no objective existence. Therefore, we do not need to distinguish here between "systems of beliefs" and "knowledge".<br /> ---<br /> II 319<br /> Language/ChomskyVsQuine: must separate language and theory - otherwise, two speakers of the same language could have no disagreement.<br /> II 330<br /> Language/Chomsky/Quine: no frame of a tentative theory as in physics - several analytical hypotheses not only possible but necessary - ChomskyVsQuine: Vs "property space": not sure whether the concepts of the language can be explained with physical dimensions - Aristotle: rather associated with actions - VsQuine: not evident that similarities can be localized in a room - principles, not "learned sentences".<br /> II 333 <br /> VsQuine: cannot be dependent on "disposition for reaction", otherwise moods, eye injuries, nutritional status, etc. would be essential.<br /> II 343 <br /> Perhaps language does not have to be taught.<br /> ---<br /> Graeser I 121f<br /> Language/ChomskyVsGrice: Question: should the main aspect really be communication? - Searle: rather representation, but not as opposite - Meaning/VsGrice: most of the sentences of a language have never been uttered, so anyone can hardly ever have meant something by them - Meaning/VsGrice: We can only ever find out speaker meanings, because we know what the sentence means. - Students of Grice: Strawson and Searle.<br /> ---<br /> Münch III 320<br /> Language/Chomsky/Holenstein: no natural kind.<br /> <br /> <br /> Elmar Holenstein, Mentale Gebilde, in: Dieter Münch (Hg) Kognitionswissenschaft, Frankfurt 1992 19 info:srw/schema/1/dc-v1.1 XML Language Acquisition Chomsky 20 info:srw/schema/1/dc-v1.1 XML Structures Chomsky Lyons I 259<br /> Chomsky: (N. Chomsky, Syntactic Structures, Berlin, New York 1957): Thesis: The notion of the constituent structure (formational structure) corresponds to a limited section of the language and the rest of the language can be derived from the fact that a fairly simple class of transformations is repeatedly applied to the chains given by the constituent structure grammar. If we were to expand the grammar of constituent structures in such a way that they directly describe the entire language, we would give up simplicity. <br /> Syntax/Chomsky: should be split into two parts: <br /> 1. basic component: constituent structure component (phrase-structure component, base component, phrase structure component, phrase structure rules (PS rules) (s) constituents).<br /> 2. transformational component with additional rules. <br /> Transformational rules/Chomsky: the entire transformational rules should be understood as additional rules. 21 info:srw/schema/1/dc-v1.1 XML Synonymy Chomsky II 335 <br /> Synonymy/ChomskyVsQuine: false idealization: not "equality in the terms" causes synonymous expressions - not assertibility conditions (circumstances) but it is about distinguishing between langue and parole, between competence and performance. 22 info:srw/schema/1/dc-v1.1 XML Terminology Chomsky Lyons I 262<br /> Definition Phrase Marker/P-Marker/Grammar/Chomsky/Lyons: if a chain is represented with constituent parentheses and parentheses indices (labelled-bracketing), this is referred to as a formation marker or P-marker. <br /> Definition parenthesis index: labelled-bracketing/terminology: designation of a node in the tree diagram or symbol in front of a parenthesis. <br /> I 263<br /> Definition Dominate/Dominance/Chomsky/Lyons: a symbol dominates an entire parentheses expression when the parentheses in the P-marker is opened immediately after this symbol. In the tree diagram: The symbol dominates everything that is derived from the node indicated by the symbol. <br /> Definition (structural) analyzability/Grammar/Chomsky/Lyons: (is a condition for the application of T-rules): if a chain without residual elements can be broken down into subchains, each of which is dominated by a symbol given in the structural description of the T-rules, then the chain satisfies the conditions defined by the structural description (SB).<br /> <br /> Passive transformation/Chomsky/Lyons: (is optional) then looks like this: <br /> <br /> {T + N + 0} + {C + M + have + en} + {V} + {T + N + 0}<br /> <br /> NP1 - Aux - V - NP2<br /> <br /> Transformation: due to the operation of the actual T-Rule (in SV), a further chain (no more core chain) results as output, which then serves with its P-marker as input for further T-Rules. 23 info:srw/schema/1/dc-v1.1 XML Deep Structure Chomsky I 269F<br /> Surface Structure/Chomsky: Determination of a hierarchy of parts of sentences that belong to specific categories: noun phrase, verb phrase, adjective phrase, etc. E.g. John is certain that Bill will leave. John is certain to leave: - similar surface structure, different deep structure.<br /> ---<br /> I 273<br /> Surface Structure/Chomsky: Assumption: it contributes nothing to the meaning. The contribution an expression makes to the sentence is defined by the deep structure (> <a href="https://philosophy-science-humanities-controversies.com/listview-list.php?concept=Compositionality">compositionality</a>). <br /> ChomskyVsAnalytic Philosophy: if different intensions were to change their meaning after substitution, there would have to be a corresponding difference in the deep structure, which is unlikely.<br /> ---<br /> I 276f<br /> Deep Structure/Chomsky: plays a role in the mental representation of sentences. 24 info:srw/schema/1/dc-v1.1 XML Transformational Grammar Chomsky Chomsky I 271<br /> Chomsky: thesis: in any language, surface structures are produced by "grammatical transformation" from "deep structures" - Definition transformation: Representation of an indexed bracket on an indexed bracket, e.g. [S[NPJohn][VP is [AP Certain] [VP ...] - deep structure: even an indexed bracket - the large class of deep structure is specified by basic rules - deep structure: subject and predicate may be exchanged - deep structures are limited in their variance.<br /> Chomsky I 296<br /> Transformation/Grammar/ChomskyVsPutnam: Transformations are not rules but operations - (for creating surface structures from deep structures).<br /> ---<br /> Strawson VI 395<br /> Transformational grammar Vs traditional grammar: it is supposed to be too unsystematic, no explanation with the traditional concepts "verb" , "noun", "object" is possible - transformational grammer Vs formal logic.<br /> ---<br /> Strawson VI 397<br /> Grammar/Strawson: must distinguish between essential and non-essential connections.<br /> --- <br /> Lyons I 269<br /> Generalized Transformation/Chomsky/Lyons: up to now we only had one end chain as input in the transformational component. However, the system also allows the combination of two or more end chains (by concatenating chain pairs = by means of optional generalized transformations, these are also called<br /> Definition transformations with double base/double-based/Chomsky/Lyons: if two or more end chains serve as input for the transformation. = "generalized transformation"). <br /> Transformation/Chomsky/Lyons: here there are two classes:<br /> a) Embedding rules <br /> b) Conjunction rules. <br /> Tradition/Lyons: this does not quite correspond to the traditional distinction between complex sentence and compound sentence. <br /> Lyons I 269<br /> Surface texture/Lyons: e.g. flying planes has the same surface texture as e.g. supersonic planes (adjective + noun). <br /> Deep structure: e.g. flying plane has a transformational relationship to the deep structure of e.g. plane fly and to planes are flying. <br /> Grammar: thus it generates a matrix string of the form NP - be - A) and a constituent string of the form NP - V intr. <br /> Lyons I 269<br /> Embeding/embedding rules/Chomsky/Lyons: were merely suggested in "Syntactic Structures" (N. Chomsky, Syntactic Structures, Berlin, New York 1957). The important thing is that an embedded structure...<br /> Lyons I 270<br /> ...is the transformation of a chain which could also be the underlying structure of an entire sentence, but which functions as a constituent of another sentence. It is a sentence in another sentence. <br /> The P-marker of the matrix sentences dominated by S therefore contains another S, which is dominated by the corresponding symbol with regard to the function of the constituent sentence in the overall structure. <br /> Definition clause/Terminology/Linguistics/Lyons: Subset<br /> Definition phrase/Terminology/Linguistics/Lyons: Complex of words. <br /> Conjunction transformation: on the other hand, also connects sentences within a larger sentence. <br /> In this case, however, no sentence is subordinated, but both retain their sentence status. The P-marker for the larger sentence will therefore contain two (or more) co-ordinated ∑ below the uppermost ∑.<br /> Transformational Grammar/Chomsky/Lyons: does not actually connect sentences, but rather the underlying structures of the sentences. 25