Restoring Normalcy in Response to an Unwelcome Surprise:
When Relevance is Analyzable as an Interactional Matter, not a Cognitive One
Robert E. Sanders
Department of Communication
University at Albany, SUNY, USA
Harold Garfinkel designed a series of experiments in which he had students create an unwelcome surprise by acting in ways that disrupted others’ taken-for-granted background understandings (normalcy), with an interest in finding out what the others would do about that. Garfinkel found that such disruptions were met with emotion displays, and more importantly, efforts to restore normalcy, congruent with Sacks’ case for normalcy as an interactional achievement.
This study examines naturally occurring interactions in which one person reported, or did, something that was reacted to as an unwelcome surprise. Such reactions consisted of, or were followed by, an interactional effort to restore normalcy. Across instances in this corpus, three ways of restoring normalcy were exhibited. One is that the person responsible for creating the unwelcome surprise provides an excuse or justification. Another is that when the unwelcome surprise is disclosure of a “wrong thought” by the responsible person, the recipient “corrects” it. The third is that instead of an emotion display, the recipient probes for additional information to find out if what occurred was not the unwelcome surprise it seemed to be after all; it had been a false alarm.
One theoretical upshot of this study is that the data are at odds with a core tenet of Relevance Theory. Even though the “stimulus” in each case adds something new to the hearer’s existing “assumptions,” and should therefore result in the “positive cognitive effect” of making a “worthwhile difference to the individual’s representation of the world,” they are observably unwelcome. Contrary to Relevance Theory, speakers and/or hearers expend effort to cancel the cognitive effect by restoring jeopardized “assumptions” (normalcy).
A second theoretical upshot of this study is that it complicates, and broadens, the concepts of adjacency pair and contingent relevance. In these data, the first pair part is more complex than a simple action, such as a question or an offer. It is itself a pair, that which created the unwelcome surprise coupled with a response that marks it as such. And the second pair part may either be a single action (an account or correction), or a jointly produced sequence whose extent depends on the achievement of satisfaction by both parties that normalcy has been restored, or cannot be.