1.8. Reading as construction and interpretation
The interpretation of linguistic narrative rests on the general basis of overall linguistic interpretation. Linguistic meaning is a complex phenomenon which exists at different phenomenological levels and can accordingly be described at a number of levels of abstraction. Part of that meaning is coded in the lexicon of a language and its syntactic structures. We can refer to this level as linguistic meaning proper, since it is that part of meaning which can be studied once we have abstracted the structure of a language from actual speech occurrences. This procedure is perhaps best defined in Saussure's Course in General Linguistics. It is useful in many respects, although it inevitably involves a simplification of actual linguistic use and contextual varieties of language. We should carefully avoid identifying this grammatical meaning or dictionary meaning of language with meaning as a whole.
If our aim is to study actual linguistic usage then we pass from linguistics to discourse analysis, and the pragmatic aspect of meaning. Here we study speech acts, discourse varieties, illocutions and perlocutions. We go beyond the study of structures to the study of processes, of the dynamic use and transformation of those structures in order to adapt language to situation and action. The speaker uses the language system in a creative way to articulate his expressive intention. The listener likewise knows that the concrete speech act plays with linguistic meaning, and that words used in context acquire a specific meaning which requires an overall (and not merely grammatical) act of interpretation. Pragmatics is perhaps best defined as the study of signs, sign formation and interpretation, codes, and code formation and interpretation in actual semiotic behaviour.
There is, as we see, a wide area left for linguistic study even if we take linguistics proper for granted. When we take into account the complexity of this study and, we can easily see that it too can be undertaken at several levels of abstraction. We have already noted that the "speech act theories" developed during the sixties were just such an abstraction: Austin and Searle studied not actual speech acts but rather idealized speech act models, what is more, working within the paradigm of a sentence grammar which has long since been exploded as a sufficient one for an approach to pragmatics.