Abstract
This article outlines the different knowledge functions of metaphors in lay and scientific discourses and proposes a methodology for the development of metaphors to yield deeper organizational scientific knowledge. It argues that the traditional dichotomy between metaphorical and literal languages has led either to an overemphasis or a depreciation of the role of metaphors in organizational science. This dichotomy is unnecessary and unproductive because metaphorical language and literal language are different but not incompatible. Drawing on Beer's suggestions about scientific modeling, this article advances a transformational view of metaphors, which attempts to outline a methodology for the development of metaphorical insights to yield literal identities.