The problem of experience in the study of organizations

1990 'Ethnomethodology and the micro-macro order'. American Sociological Review 55: 794-808.

Homans, George C.

1950 The human group. New York: Harcourt and Brace.

Homans, George C.

1967 The nature of social science. New York: Harcourt and Brace.

Introduction

In science, as everywhere else, seeing is believing. According to James (1911), science consists in the substitution of a conceptual order for the perceptual order of experience. Russell (1921) insisted that scientific theory should be founded upon perceptual experience of some kind. In his terms, 'knowledge by description' presupposes 'knowledge by acquaintance'. To put this principle most colourfully, Wheelwright (1962) described the aim of theory as being to 'chop at the joints' -- that is, to mark off resemblances and differences where experience says it is most natural to do so.

For the field of organization studies, this principle is an embarrassment. Organizations cannot be perceived, and therefore cannot be unambiguously theorized. To the question: 'What is organization theory about?' there is no easy or ready reply. The research literature offers a dismaying multiplicity of possibilities, not one of which is recognizably definitive. To some authors organization theory is about bureaucracy, to others it is about systems of information and/or energy, to others it is about populations of social forms, to still others it is about modes of exchange (e.g. markets, hierarchies, and clans), and to still others, it is about patterns of meaning or culture. These stopgaps are questionable in the same measure; for about systems, markets, hierarchies, clans and cultures, there is an equal ration of indefiniteness. Disagreement is possible because perceptive experience offers little to agree about.

In this article, we show how the absence of perceptive experience of organizations makes it difficult to theorize about them scientifically. In fundamental respects, ours is an exposition and application of contemporary ethnomethodological critiques of social science (e.g., Garfinkel 1967; 1988; Hilbert 1990) to the field of organization studies. Parallelling those critiques, we argue that scientific conceptualization of organizations must be premised on phenomenological insights. Whereas science comes to life in the dynamic opposition of phenomenal experience and explanation; in the absence of this opposition, it dissipates to a desultory exercise of hopes. The fact that the 'organization' conforms to no definite experience (i.e., that it cannot be perceived, nor even imagined) means that the study of organizations is deprived of this life-giving dynamic. Perhaps the clearest sign of this is the strange inversion of discourse that occurs in organization studies where, amidst the lively discussion of concepts and theories, there is barely a whisper about how organizations should be conceived or operationalized. Finally, we discuss the prospects for a science of organizations and what might be done to improve them.

Organizations in Doubt

We find it natural to speak of organizations as objects in action. Indeed, our language is built upon a grammar of object and predicate (to the chagrin of advocates of a more fluid language of process, e.g., Whitehead 1933; Giddens 1979). Even though grammatical form is not the same thing as phenomenal form, we are likely to mistake the two when there is no concrete experience to keep them separate. As Wittgenstein (1953: 42-49) pointed out, our problems of thought are primarily grammatical. A 'main source of our failure to understand', he wrote, 'is that we do not command a clear view of the use of our words'.

Uncertain Substance

The problems of studying organizations begin with the difficulty of pointing them out -- to gesture and say, 'This is an organization.' This contrasts with the natural sciences where gesturing and saying are often easier and more convincing. Pointing makes definition easy because it frees us from having to translate what we can see into words. However, in the case of organizations, there is nothing obvious to point to, or to exclaim about. The word 'organization' names something that cannot be seen, and thus that cannot otherwise be confirmed as an object.

To be sure, the research literature on organizations offers almost no hint that this is, or could be, a problem. Although it is sometimes acknowledged that the boundaries of organizations are problematic (e.g., Katz and Kahn 1978; Weick 1979), there is little question that organizations are genuine objects. The overwhelming majority of empirical studies report no difficulties of identification or measurement. For example, in the last 5 years (1986-1990) of the Administrative Science Quarterly, a well-known and respected journal of empirical research on organizations, we counted 85 studies which referred to over 204,000 organizations or units thereof, ranging from Fortune 500 corporations to Finnish newspapers to small convenience stores to school districts and classes. Not one of these voices a serious misgiving about the existence of the organization(s) under study. Only two of these studies suggest that the idea of organization could be empirically problematic. Barnett and Carroll (1987) analyze the structure of competition and mutualism among early phone companies in Southeast Iowa and find that the dynamics of selection are based on communities of organizations, not individual organizations. Fiol (1989) notes that in letters to corporate shareholders, chemical companies that are engaged in joint ventures with other companies speak about their identity in more uncertain terms, and have a less well-defined sense of their external boundaries (although their sense of internal boundaries may be enhanced). However, even in these cases, there is no suggestion that the concept of organization should be abandoned, or indeed that it does not refer to an actual entity. The 'fact' of organization is sure enough for research to proceed.