If you can think of an instance of meaning, then you can think of a representation of it. The problem here is called the SymbolGroundingProblem. If all you do when you interpret a symbol is replace it with more symbols, then what have you actually acomplished? Each new symbol must then be interprested again, leading, it would seem, to an infinate recursion that never ends.
From Chapter 1 titled ![[WWW]](/wiki/classic/img/moin-www.png){kind=small}
Semantic Theory and the Representation of Meaning of the German book, by Nico Weber: Die Semantik von Bedeutungsexplikationen: "Directed paraphrasing" dubbed "predicate typing" was introduced as a "method for semantic classification" by Garvin/Brewer/Mathiot 1967. It consisted in having native speakers identify cue forms in sentences presented to them that pointed to one (or more) elements of a set of predetermined paraphrasing schemata. "The aim of this effort was the formulation of a semantic classification of pertinent linguistic units" (op. cit. 36). The "theoretical interpretation" (op. cit. 36 ff.) was not very forthcoming about the results and perspectives of the approach. Semantic paraphrases and descriptions of common and frequent lexical items, elicited from native informants with test lists and questionnaires were collected and classified by Casagrande/Hale 1967 as "Papago folk-definitions". Considering definitions as statements of semantic relations between the concepts defined and those used for defining, they "attempted to provide an exhaustive set of mutually exclusive types of semantic relationships, that is, a set under which all the types used in our sample of definitions can be subsumed" (op. cit. 167). The authors did not claim these types to be universal, nor even a priori applicable to other languages, or to lexicographical description (op. cit. 192). The following types were distinguished in Papago definitions (168 ff.):4
-
Attributive: X is defined with respect to one or more of its attributes Y, e.g. 'sun': 'that which goes over and thereby gives us light' (171).
-
Contingency: X is defined with relation to an antecedent or concomitant Y, e.g. 'to wash': 'if a person gets dirty, he washes himself' (172).
-
Function: X is defined as a means of effecting Y, e.g. 'ear': 'with which we hear when something is said' (175).
-
Spatial: X is defined with reference to its position in relation to Y, e.g. 'forehead': 'which is below our hair and above our eyes' (176).
-
Operational: X is defined with respect to a goal or recipient Y, e.g. 'saddle': 'which they put on a horse's back' (178).
-
Comparison: X is defined in terms of similarity and/or contrast with Y, e.g. 'bat': 'which looks like a mouse; but it has wings' (179).
-
Exemplification: X is defined in co-occurrence with a typical Y, e.g. 'to copulate with': 'a male will copulate with a female' (180).
-
Class Inclusion: X is defined as a member of a hyperonym Y, e.g. 'owl': 'which is a big bird' (182).
-
Synonymy: e.g. 'amusing': 'funny' (183).
-
Antonymy: e.g. 'low': 'not high' (183).
-
Provenience: X is defined with respect to a source Y, e.g. 'milk': 'we get it from a cow' (184).
-
Grading: X is defined with respect to a series or spectrum that also includes Y, e.g. 'Monday': 'the one following Sunday' (184).
-
Circularity: X is defined as X.
The method demonstrated here is certainly applicable to the analysis and classification of dictionary definitions. Although most of the types of semantic relations enumerated above are occasionally used in dictionaries, two of the most frequent types used in traditional language dictionaries are not represented: the “Aristotelian” type of definition per genus proximum et differentiam specificam (= GP&DS),5 and the “morpho-semantic” type exploiting morphological or morpho-syntactic parallelism between definiendum and definiens.6 This suggests that the latter two definition types result from lexicographers' efforts toward precise, rational, or technical descriptions of meaning.
