Asking the Wrong Questions about Conversation

Last night BBC News Technology Reporter Dave Lee put up an article on the Web site entitled “The quest for the online water-cooler experience.” It began with a critical assessment of the purported “solutions” of enterprise software technology regarding their failure to deal with the proper fundamental issues of effective work practices:

All these well-intentioned schemes hope to do is recreate the water-cooler effect: the incidental discussions, those informal links between departments that are the very foundation of sustaining success and encouraging great ideas.

The water-cooler has long been a Holy Grail for enterprise software, going back long before it was even called “enterprise software.” However, rather than pursuing the question of why this Grail has been so elusive, Lee then launches in to what is basically a promotion piece for two new technologies, Yammer and Spiceworks.

What Lee seems to have missed is the extent to which those encounters over the water-cooler are more than “links” or even “informal discussions.” Rather, they are instances of a general category, which Jürgen Habermas chose to call “communicative actions,” in his effort to explore a concept broader than the “speech acts” of John L. Austin. Habermas’ “theory of communicative action” (the title of his two-volume magnum opus) tried to get beyond technological thinking predicated upon reducing communication to the exchange of sentential forms that could be represented in some objective logical calculus (often called “transactions” by many of the software systems that have evolved). Rather, he was motivated by the anthropological efforts of Erving Goffman to view the fundamental unit of any conversational exchange as a “move,” a unit that is not strictly “answer-oriented” (Goffman’s language) but entails both psychological and sociological “baggage” beyond the need to have a question answered.

Those who place technology before solutions look at a water-cooler and see nothing more than a corpus of texts. They can then analyze those texts both grammatically and ontologically, even going so far as to build a “world model” to embrace what is being exchanged in those conversations. What is missing, however, is a phenomenological point of view, such as that proposed by Ernst Cassirer and Alfred Schutz. Rather than trying to home in objectively on what those words and sentential forms are, one must accept the broader view of how they appear to the participants in that encounter at the water-cooler.

This is not, as Georg Ell of Yammer seems to believe, simply a matter of being more “passionate about user-interface design,” however much I sympathize with Ell about the significance of such design and the consequences of its neglect. Rather, it is a recognition that conversations take place in the social world, rather than the objective world. Every now and then a system emerges that actually accepts the social world of its users and may even facilitate the moves that take place in that world. The Xerox Eureka system, which supported a global network of repair technicians, was such as system. However, the success of Eureka had less to do with the functionality of its software and more with how its design grew out of observations from anthropological field work concerning how these technicians exchanged “war stories” at the end of the working day.

Until purveyors of enterprise software learn how to address phenomenological questions about the social world in which their “solutions” are to be embedded, those “solutions” are likely to be no more effective than snake oil!