|
|
|
|
|
|
dict for a given expression to occur, while I reserved the name of cotext for the actual environment of an expression in the course of an actual process of communication. Thus I would say that the expression I order you can normally occur in those contexts (or classes of texts) where the sender is characterized by a position of superiority in respect to the addressee, or in circumstances where the same social relationship holds, and that it occurs in the co-text of the novel So and So. |
|
|
|
|
|
|
|
|
In Eco 1976 (2.11) I outlined a semantic model supposed to predict differences in meaning that depend on possible usual contexts and circumstances; in Eco 1984 (2.3.1) I tried a representation of propositions and adverbs where contextual selections interact with the topic (as a cotextual conjecture that a semantic theory cannot predict but must ideally take into account). |
|
|
|
|
|
|
|
|
Also Greimas's (1966) notion of "classème" enriches semantic representations with a contextual selection. |
|
|
|
|
|
|
|
|
2.4. Felicity Conditions and Illocutionary Force |
|
|
|
|
|
|
|
|
In the framework of generative semantics, many authors have felt the need to provide a context-oriented representation. Lakoff (1975) suggests that felicity conditions must be given as meaning postulates, for instance, |
|
|
|
|
|
|
|
|
Request (x,y,P)Attempt (x,cause(y,P)) |
|
|
|
|
|
|
|
|
Many other felicity conditions can be semantically recorded. For example, in the representation of a verb such as scolding, it should, and can be, registered a pragmatic-oriented marker such as "S>A," where S is the Sender, A is the addressee, and > stands for a relation of social superiority, or a hierarchical operator. |
|
|
|
|
|
|
|
|
Fillmore's case grammar, by introducing into the lexical representation such cases as Agent, Goal, Instrument, Result, and so on, links the interpretation of the lexical item, from its very inside, to the co-occurrence of a contextthis context being virtually given by the systematic representation of meanings, hence not depending on mere extralexical world knowledge. In other words, general schemes of world knowledge are assumed as a part of the lexical information. In the same line of thought, let me rank the semantic models of Bierwisch (1970 and 1971), for example, the representation of "kill": |
|
|
|
|
|