Open AccessHvad konnekterer konnektorerne?
Author(s) -
Henning Nølke
Publication year - 2006
Publication title -
tidsskrift for sprogforskning
Language(s) - English
Resource type - Journals
eISSN - 1902-8504
pISSN - 1603-5925
DOI - 10.7146/tfs.v4i1.316
Subject(s) - computer science , syllogism , argument (complex analysis) , linguistics , scope (computer science) , context (archaeology) , key (lock) , syntax , utterance , simple (philosophy) , cable gland , artificial intelligence , natural language processing , philosophy , programming language , epistemology , history , telecommunications , biochemistry , chemistry , computer security , archaeology
This article addresses the main question: Which elements do (pragmatic) connectors connect? Connectors are normally defined as linguistic items that combine simple utterances into more complex utterances. First I try to decircumscribe the class of connectors, which is extremely heterogeneous. Then I present the basic elements of a Connector Grammar distinguishing the connectors’ arguments, which are the semantic units they combine, from their (left and right) scopes, which are the formal units that they convey. Taking the French connector donc as my key example, empirical analysis reveals that very often there is no utterance functioning as left scope. I argue that the word (donc) should still be regarded as a connector in these cases because it retains its basic semantic (connecting) function, only the left argument is implicit and has to be inferred from the preceding context: the left scope is non-verbal. To preserve a unique logic-semantic description of the connector for all its uses, I introduce the generalized syllogism, which probably can be found behind most human reasoning