Answers to a Discussion Note: On the 'Metaphor of the Metaphor'

Abstract

Should a debate of the choice(s) between metaphorical investigation and epistemological realism in organizational research be prioritized as Willy McCourt called for in Organization Studies? (McCourt 1997) We argue here against doing any such thing -- a 'realism' debate in organizational theory would merely be a 'red herring' (Hausman 1998). Theoretical investigation from Ricoeur to Derrida has liberated us from the need to re-visit the theme, but examination of Gareth Morgan's (and Gibson Burrell's) intellectual development, as begun by McCourt, is of interest because it reveals two very different 'realisms'. What is of interest about 'realism' is not an either/or of either 'realism' or 'constructivism', but a polyphony of the many voices ('selfs') of research.

Descriptors: metaphor, epistemology, Burrell and Morgan, realism, research methodology

Realism and Metaphor

Twenty-five years ago Paul Ricoeur amply discussed the issues raised when empirical research (in his case phenomenology) crossed swords with the metaphorical nature of language (Ricoeur 1978). While the researcher's goal is to assert that what he or she describes (has 'discovered') is 'real', epistemological rigor demands the admission of 'polysemy' and 'metaphor', which seem to threaten the researcher's ability to achieve his or her goal of asserting 'real-ness'. McCourt believes we need (close to) literal scientific language, in his terms the 'weak' version of the constructivist position, to be able to relate the 'real' (McCourt 1997). In his analysis of language and 'reality', McCourt moves back and forth between two different concepts of the role of metaphor. In the first place, metaphor can be identified in rhetoric with stylistic embellishment -- i.e. the creation of images and poetic descriptions that bring 'text' to life. The role of metaphor is to catch the reader's attention and create vividness. He re, metaphor is a matter of lexis (Aristotle: 'diction'). However, metaphor can also be seen epistemologically as the means by which vocabulary can grow or change. Thanks to metaphor, language can alter and renew itself. Here metaphor has to do with mimesis (Aristotle: duplication in text of reality, the expression of human action that is already there). Ricoeur argued for the second position, i.e. research reports and articles, like 'all linguistic creations, would be meaningless if they did not serve the general project of letting new worlds emerge' (Ricoeur 1978: 148). Words renew themselves easily because of polysemy; in natural languages they mean more than one thing -- flexibility is inherent to the semiotic entities (the 'signs' or 'expressions') that are attached to a multitude of semantic (the 'meanings' or 'content') entities (Beneviste 1973; Hjelmslev 1961). Semiotic entities 'are defined by their difference with regard to other units of the same (language, text, linguistic) system ... these entiti es are not related to extralinguistic realities such as things, events, properties, relations, actions, passions, or states of affairs. They are purely intralinguistic phenomena' (Ricoeur 1978: 121). Thus, meaning or the 'world revealed' is an event on the level of the sentence (the text) and not on the level of the word. The truth-claim' or 'reference' of research refers to the world (state of affairs) to which it tries to point. Scientific language can try and repress polysemy by delimiting the semantic field of the words it uses; it can try to control the meaning of the words used (for instance in mathesis). However, research in organizational studies rarely breaks the link with natural language -- it deals with companies, managements, workers, profit and loss, leadership, motives, etc., which are all present in natural language. The epistemological problem of natural science, that theory is often counter-intuitive and expressed in 'text' that is highly foreign to natural language, just does not arise (Hau sman 1998). Nor does the behaviourist rupture with natural language frequently occur: 'behaviourists insist that one stick with generalizations cast at the level of observation' (Hausman 1998: 17). Organizational studies try to explain and predict, making use of categories (though often under dispute) of social phenomena that really stick quite closely to the discourse of practitioners in the field. The study of organizations welcomes theorizing concerning matters of observation. Epistemologically, the theorizing can be as speculative and bizarre (i.e. as metaphorical) as the researcher wants, as long as it has explanatory value. Researchers in organization, including even the most radical social constructivist, will admit that a discourse about 'organization' (whatever that discourse might mean) exists/existed prior to investigation. Organizational researchers de facto assume that the concept(s) of 'organization' they 'observe' in the field exist independently of the researcher(s) and can be studied. Such (a minimal) realism is only challenged by those who assume 'underlying structures, powers, mechanisms and tendencies exist, whether or not detected, [that] govern or facilitate actual events' (Lawson 1997: 21). The dissenters are transcendental realists -- their assumption of depth level order is metaphysical. All the rest of us are local realists -- we assume that our texts (sentences) point to a circumstance, place and event, that we can (however partially) investigate. Thus, there is very little contextually neutral text in organizational studies. McCourt does not want to accept the logic of local realism -- of context-bound sentences (texts) which communicate pragmatically, i.e. of an instrumentalist OT which exists interactively, and in relationship to its stakeholders' (actors') goals and needs. McCourt wants a so-called langue bien faite 'ruled by the principle of a one-to-one relation between signs and entities, of one meaning for each word ... [aspiring to] a language which would be the exact picture o f the structure of facts' (Ricoeur 1978: 129). McCourt pretends that his problem is with polysemy and the creative role of the metaphor in generating ever more polysemy, but his problem is broader; he doesn't accept doing (social) science in ordinary language.