The paper touches upon the relations between grammar-oriented and communication-oriented ways of analysing language in the light of some recent related developments in theoretical linguistics, congnitive psychology and artificial intelligence. The integration of (linguistic) meaning, situational interpretation and background knowledge is discussed. In Particular, different types of linguistically relevant common-sense knowledge and higher level cognitive procedures are considered.

This paper argues that Jürgen Habermas' theory of universal pragmatics is valuable for semantic and text-linguistic theory. Habermas' distinctions between communicative action (Kommunikatives Handeln) and discourse (Diskurs; debate on the validity of statements) are reflected in linguistic structure, as is his distinction between the four dimensions involved in every speech, act: comprehensibility, truth, sincerity, and correctness. The primary question is: Given that Habermas' distinctions are universal, what means of signalling them do natural languages contain? The present article limits itself to contrastive focus and epistemic verbs. The application of contrastive focus, to epistemic verbs enables speakers to mark the distinction between communicative action and discourse (the level of debate) and to clarify the dimension, under debate: comprehensibility, truth, sincerity, or correctness.

This paper outlines a model designed for the description of East Norwegian sencetece intonation. It is argued that there is a specific kind of intonational structure in Norwegian that is restricted to assertions and which will therefore normally be perceived as a prosodic clue to the indentification of assertive clauses, whether embedded ro syntactically independent.

The paper is devoted to the thesis that all Sanskrit aspirates (voiceless as well: as voiced) should be represented as clusters of a consonantal segment /C/ plus a segment /h/, the latter being [-consonantal, -vocalic], and not as single consonantal segments with the feature [+aspirate], which is the usual analysis. The arguments for the correctness of this thesis are based on formal considerations, i.e. on the fact that a cluster analysis yields a simpler formul ation of a 1arge number of different phonological processes in the language. The paper compares the cluster analysis of aspirates with a non-cluster analysis with respect to ten different rules.