Find counter arguments by entering NameVs… or …VsName.
The author or concept searched is found in the following 4 entries.
| Disputed term/author/ism | Author |
Entry |
Reference |
|---|---|---|---|
| Context/Context Dependence | Lyons | I 229 Context-independent constituent structure grammar/linguistics/Lyons: until now, all constituent structure rules had the form A > B ("replace A by B") No matter the context. >Constituent structure grammar. Chomsky: studied the effects of organizing rules within the system, the consequences of introducing optional, alternative subrules and recursive rules. Conditions so far: 1. A and B must not be identical (i.e. A must not be replaced by itself), 2. A must be a simple symbol, B may be complex, which is usually the case. I 239 Context-dependent/Grammar/Lyons: Terminology: Context-dependent: context-sensitive, (c-dependent, c-restricted). Old: until now, all grammars we considered were context-independent. This means that the symbol to the left of the arrow in the output of the rule has been replaced by the symbol chain to the right. example N > N + and + N only condition: that N was in the input of the rule. E.g. alternative chains: (I) X + N + Y (II) W + N + Z then we get once X + N + and + N + Y W + N + and + N + Z Thus context-independent. (This was required!). I 239 Context-sensitive grammar/Chomsky: new: Instead of "Replace A with B": now: e.g. N > N + and + N/in the context X + ...+ Y I 241 Context-independent ones are special cases of dependent ones. New: Context Dependency: Assuming we have new rules: N > N + and + N/in the context X + ...+ Y Everyday language: "N is only to be replaced (optional or obligatory) if it is appears in the input chain... I 240 in such a way that X immediately follows on the left and Y immediately on the right. Then this rule would apply to (I) but not to (II). >Everyday language. Context-dependet rules: different types: (we limit ourselves to the options introduced so far: optional/obligatory, recursive/non-recursive, coordinated/subordinated): Variants: X and Y can represent one or more symbols. Assuming that the class of context-dependent grammars we are dealing with here is defined by the fact that in a rule of the type A > B/in the context X + ... + Y X and Y can (each individually) refer to any finite number of concatenated symbols, but that A must be a singular symbol. B must neither be identical to A nor zero. >Recursion. Then the following rules would be well-formed: a) P > Q/in the context E + F + ... + G b) P > Q + R/in the context E + ... + G + H + K + L c) P > R + S + T/in the context G +... + H etc. I 241 Context-independent grammar/Lyons: can be seen as a subclass of the (newly introduced) context-dependent grammars ((s) as special cases). Def Context-Independent/Lyons: if a rule has contextual variables X and Y with an unlimited value (i.e. they can be positive or zero), then the rule is context independent. Otherwise context-dependent. I 242 context-dependet: e.g. f) P > Q/in the context 0 (zero) + ... + 0 P may only be replaced by Q if there is no other sign to the left and right of P in the input chain. This normally only applies to the character . g) P > Q/in the context 0 (zero) + ... + R + S P may only be replaced by Q if the input chain is P + R + S. h) P > Q/in the context T + ... + 0 P may only be replaced by Q if it is at the last position in the input chain: T + P. General form with variables: X + A + Y > X + B + Y Then a context-free rule of the form A > B is a special case of a context-dependent rule in which there are no restrictions for the values of X and Y. Context-dependent and context-independent rules can be placed in the same formal framework. I 245 Congruence/Subject-verb-congruence/context-independent/Lyons: Example (1a) The dog bites the man (1b) The dog bites the men I 246 (1c) The dogs bite the man (1d) The dogs bite the men (2a) The chimpanzee eats the banana etc. Context-independent Grammar/Lyons: e.g. (1) ∑ > NP sing + VP sing or NP plur + VP plur. (2) VP sing > V sing + NP (3) VP plur > V plur + NP (4) NP > NP sing or NP plur (5) NP sing > T + N sing (6) NP plur > T + N plur (7) N sing > N + 0 (Null) (8) N plur > N + s (9) V sing > V + s (10) V plur > V + 0 Here, more than one symbol is replaced at a time. Lexical substitutions/Lyons: here we assume that their rules are outside grammar. >Lexicon, >Grammar. I 247 Number/context-independent grammar/Lyons: is defined here by rule (1) as the category of the sentence for the subject-verb congruence. However, it is also introduced in the object nominal expression by rule (4). Singular/Plural: so the alternative is something that is completely independent and different from the same alternative in the object position. The grammar does not make everything visible here, not even that the choice in subject and object position is independent, and that the verb, if the subject is once determined as singular or plural, is determined according to congruence. >Correctness. I 249 Context Dependence/Rules/Economy/Lyons: the rule growth to cover all other congruence ratios would be small. On the other hand, it would be significant in context-independent grammar. Here, context-dependent grammars are more economical. Correctness/lyons: both types of grammars formalize the congruence ratios correctly. I 250 Def Weak Adequacy/Grammar/Lyons: a grammar is weakly adequate when it generates the desired class of sentences. Def strongly adequate/Lyons: it is strongly adequate when it also assigns the correct structural description to each sentence. >Adequacy/Lyons. Correctness/Theory/Lyons: our definition of strong/weak adequacy implies in no way an interpretation of "correct". It does not even make an assumption as to whether there are any norms of "correctness". >Terminology/Lyons. However, we determine that it is possible, at least in certain cases, to say that one description is more correct than another. We just do not claim that we can decide what is "absolutely correct". Context-dependent/context-independent/grammar/adequacy/equivalence/Lyons: the two grammars are probably weak, but not strongly equivalent. The context-dependent is more appropriate. Comparability/equivalence/Lyons: since the two systems are weakly equivalent, they are at least comparable. |
Ly II John Lyons Semantics Cambridge, MA 1977 Lyons I John Lyons Introduction to Theoretical Lingustics, Cambridge/MA 1968 German Edition: Einführung in die moderne Linguistik München 1995 |
| Recursion | Tarski | Skirbekk I 156 Recursion/recursive method/Tarski: starting from simple propositional calculus specifying the operations with which we construct composite functions. >Functions/Tarski, >Recursive rules. Skirbekk I 157 Recursion/Tarski: problem: composite statements are constructed from simpler propositional functions, but not always from simpler statements. >Propositional functions. Hence no general recursion is possible. Recursive definition of satisfaction is only possible in a much richer metalanguage (i.e. in metalanguage we have variables of a higher logical type than the in the object language.(1) >Expressivity, >Richness. 1. A.Tarski, „Die semantische Konzeption der Wahrheit und die Grundlagen der Semantik“ (1944) in: G. Skirbekk (ed.) Wahrheitstheorien, Frankfurt 1996 |
Tarski I A. Tarski Logic, Semantics, Metamathematics: Papers from 1923-38 Indianapolis 1983 Skirbekk I G. Skirbekk (Hg) Wahrheitstheorien In Wahrheitstheorien, Gunnar Skirbekk Frankfurt 1977 |
| Rules | Lyons | I 157 Rules/Grammar/Transformational Grammar/Chomsky/Lyons: Chomsky seems to reject this. In his opinion: ChomskyVsGrammatical rules: Thesis: The grammatical structure of the language is determined ((s) not according to the above rules) and is "intuitively" (unconsciously) mastered by the native speaker. (ChomskyVsRules due to the consequence of "uncertainty of grammar"/ChomskyVsUncertainty of grammar). Lyons: the differences in opinion here are exaggerated. Not all grammar is uncertain. I 219 Phrase structure grammar/Constituent grammar/Rules/Chomsky/Lyons: Each rule brackets the constituents that form the construction defined by it and also describes them. >Constituent grammar, >Phrase structure grammar. Layers: (of the structure) are determined by the order in which the rules are applied. Def Initial symbol/Terminology/Grammar/Chomsky/Lyons: Example ∑ for sentence ((s) stands farthest left or above a branch). Grammar: produces a chain of symbols by applying the rules. >Lexicon. Def End Symbol/Grammar/Lyons: specifies the class of elements of the lexicon e.g. adjective. Def End chain/terminal string/grammar/terminology/Lyons: consists of end symbols. I 220 Sentence/Grammar/Chomsky/Lyons: occurs when we replace the end symbols from the end chain with an element of the lexical class they describe. Its constituent structure is fully determined by the replacement rules that create the end chain. >Terminology/Lyons. I 220 Replacement rules/Grammar/Alternative rules/Extension/Chomsky/Lyons: to distinguish transitive and intransitive verbs, we introduce: (1) ∑ > NP + Vp (2a) VP > V intr + Adv 2b) VP > V tr + Adv (3) Np > A + N. I 221 If we introduce the option between (2a) and (2b), we must change the word classification in the lexicon: V intr = [{ran, etc.} V ir = {love, kill, etc.}. >Word classes. Grammar/problem: it is still unsatisfactory: 1. It still produces illegal sentences such as Poor John kill old women ((s) no special form for 3rd person singular). Solution: we must consider the congruence between the "subject" and the verb. >Congruence/Lyons. 3 (we leave that out here). 2. as it stands now, we can only produce sentences with five words like "Old men love young women" or sentences with four words like "Poor John ran away". The following sentences are not possible: e.g. John ran away, e.g. Men love young women, e.g. Old men love women, e.g. Old men love young women passionately. Optional Rule/Extension/Grammar/Replacement Rules/Lyons: For example: we extend rules (3) by making two rules out of one: (3) NP > N (4) N > A + N We say that (3) is obligatory, but (4) optional. New: then we also get: e.g. John ran away, e.g. Men love young women, e.g. Old men love women etc. All these sentences are subtypes of the sentence type. ∑(NP + VP). This means that their structures are identical at a certain level of analysis. Family Tree/Structure Tree: Example (I) John ran away (II) Poor John ran away (III) Men love women (IV) Old men love women (V) Men love young women, (VI) Old men love young women I 223 Rules/Replacement Rules/Order/Grammar/Chomsky/Lyons: the priority of a certain order of rules over another can significantly change the result of the grammar. optional: e.g. (1) ∑ > NP + VP (2a) VP > V intr + Adv 2b) VP > V tr + Adv (3) Np > A + N. (4) N > T + N (5) N > Adj + N Rules (4) and (5) are optional. New: therefore the grammar now generates men, the men, good men, and the good men. Order: if (5) should come before (4), there would be e.g. good the men. Order: also that of (3) is essential: if it were in front (2b), it would have to be repeated afterwards to guarantee the extension for the complex resulting from VP > V tr + NP. The sequence can therefore prevent inadmissible sentences and reduce the scope of the rule corpus. Order: Assumed, (6) N > N + and + N If (6) operates before (5), we get for example (old men) and women and men and (old women) If (6) operates after (5), we get for example old (men and women). I 224 Semantically, it is the same, despite the different brackets. Def Recursive Rules/recursive/Recursion/Lyons: allow infinitely repeated application (only in infinite cases they are called recursive). Example (6b) N > N + and + N + and + N (6c) N > N + and + N + and + N + and + N (6d) … E.g. This is how you can tell stories: e.g. He came in and he sat down and he said that ...and he... Recursion/Grammar/Lyons: a "realistic" model of grammar I 225 will be designed in such a way that there are more examples of recursive structures with two constituents than with three, more with three than with four, etc. ((s) the simplest forms should be the most likely ones). >Constituent Structure Grammar. Probability/Grammar/Correctness/Lyons: the probability of an occurrence must not be confused with its correctness. Coordination/Recursive rules/Grammar/Lyons: Problem: Coordination using a recursive rule: ambiguity by different possible brackets e.g. Tom and Dick and Harry, (Tom and Dick) and Harry, Tom and (Dick and Harry). Dilemma: a) Intuition: recursive rules do not indicate what the intuitively perceived structural description is. b) and yet recursive rules are necessary. I 227 Formation rules/phrase structure rules/constituent structure grammar/Chomsky/Lyons: Spelling/Terminology: PSG - phrase structure grammar. PS rules - Phrase structure rules. a) Formation rules = phrase structure rules b) Transformation rules: specify how the end chains are transformed into real sentences. I 249 Context Dependence/Rules/Economy/Lyons: the rule growth to cover all other congruence ratios would be small. >Context/Lyons. On the other hand, it would be significant in context-independent grammar. Here, context-dependent grammars are more economical. Correctness/Lyons: both types of grammars formalize the congruence ratios correctly. >Correctness/Lyons. |
Ly II John Lyons Semantics Cambridge, MA 1977 Lyons I John Lyons Introduction to Theoretical Lingustics, Cambridge/MA 1968 German Edition: Einführung in die moderne Linguistik München 1995 |
| Valuation | Bigelow | I 125 Valuation function V/Bigelow/Pargetter: its definition is complex because it has to be recursive. It assigns an interpretation or a semantic value. (To each expression of the language). >Recursion. Valuation: First, semantic values are assigned to the non-logical constants. >Semantic value. Rules are then created for semantic values from compound expressions. Logical constants: their valuation is specified by recursive rules. >Logical constants. Domain: can also be restricted, e.g. if you want to exclude the Barcan formula. >Domains. For example, restriction: for each world w you can assume a separate individual domain DW. Which, for example, consists only of the possibilia of this possible world. >Possibilia, >Possible worlds. I 126 Def partition/Bigelow/Pargetter: is a family of individual domains that do not overlap. I.e. no individual is in more than one possible world. That would correspond to Lewis's counterpart theory. >Counterpart theory. I 129 Counterfactual Conditional/Valuation/Valuation Function/Valuation Rules/Bigelow/Pargetter: V9 If a = (ß would be γ) then V (a) is the set of all possible worlds w ε w so that there is a possible world u where ß is true and γ is true and every possible world v in which ß is true and γ is false, is less accessible from w than from u. >Similarity metrics, >Counterfactual conditional. Similarity/possible worlds/similarity metrics/counterfactual conditional/Bigelow/Pargetter: Rule V9 states that a counterfactual conditional (ß would be > would be γ) is true in a possible world if the next ß-worlds are all γ-worlds. |
Big I J. Bigelow, R. Pargetter Science and Necessity Cambridge 1990 |
| |||
No results. Please choose an author or concept or try a different keyword-search.