|dc.description.abstract||Since the times of Aristotle, metaphors have been the topic of various investigations. However, a critical
discussion of the classical view of metaphor was not up for discussion before the nineteen fifties. A
considerable number of studies on metaphor have since been carried out in the fields of literature theory,
psychology, philosophy, linguistics and other disciplines. A notable increase in terms of the number of
published works on metaphor can be found after the nineteen eighties when George Lakoff and Mark Johnson
introduced their "Conceptual Metaphor Theory" which is rooted in the framework of cognitive linguistics.
What is characteristic is that, in most of the studies, the notion of metaphor that is used is not made explicit.
This can lead to problems when comparing the results of different studies. In the present paper, I try to specify
the notion of metaphor. In the attempt to draw a line from Aristotle's notion of metaphor to the contemporary
notion of metaphor in cognitive linguistics, it will become apparent that there are at least three relevant
dimensions of analysis in applied approaches to metaphor: (i) metaphors can occur on the linguistic surface or
at the conceptual level; (ii) on both levels they can be novel or conventionalised, and (iii) they can be part of a
metaphoric model or stand isolated outside a model. Those metaphors that are part of a metaphoric model play
an important role in the advancement of science and the gain of knowledge. On the other hand, those
metaphors that are not part of a metaphoric model do not primarily organise knowledge in a systematic way.
They have a more associative potential.||nb_NO