The articles in this special collection give us an opportunity to further reflect on a central concern for any discipline dedicated to the study of human action, namely, the role that introspection plays in giving us insights into what people think, feel, and want. The starting point of this discussion is the observation that members of a number of Pacific societies have been said to claim—or to imply through their behavior—that it is impossible to know what goes on in another person's mind. This claim, called the "doctrine of the opacity of other minds" by the editors (hereafter "opacity doctrine"), is often extended to the self, as when native consultants refuse to provide motivations for their own actions or resist any kind of intentional reading of what they have just done. In treating this type of attitude as a puzzle that needs our empirical and theoretical attention, the authors in this collection, like those of us who were dealing with it some twenty plus years ago, are making a set of analytical choices based on assumptions that should be re-examined. In the following comments I review some potential problems that emerge in the arguments presented in this collection and also suggest some ways of integrating philosophical and anthropological perspectives that could help develop more precise research tools and hypotheses about the opacity doctrine. [End Page 483]
Levels of Argumentation
A careful reading of the articles in this collection shows that the opacity doctrine has been dealt with on two main levels, observational and explanatory, each of which relies on a different order of generalizations and implies different kinds of analytical categories and types of argumentation. The first level—which we might call "observational"—draws from fieldwork situations in which ethnographers try to get their consultants, field assistants, hosts, and friends to tell them about other members' actions, motivations, explanations, and emotions. These attempts, as we know from the ethnographic literature, which include the articles in this collection, are sometimes met with surprise, suspicion, or flat rejection, as in "How would I know?" or "Why should I know?" Such responses are subsequently interpreted as consistent with a particular "local" theory of interpretation, to be distinguished from the ethnographer's. This move transforms the earlier "observational" level into a first stage "explanatory" level. With the confidence that comes with labeling, ethnographers could begin to search for more evidence to support their hypothesis about the local theory. At this point ethnographers also face the question of whether they themselves are creating the problem, and the theory, by asking certain kinds of questions (Briggs 1984). Could it be that a different way of asking about introspection would bring about a different kind of answer? This line of inquiry, as far as I know, has never been fully developed, even though the next step could be seen in part as a way of addressing this issue. Armed with the hypothesis that the natives have a way of thinking about what we can know about others that is different from what is usually assumed in western theories of meaning, ethnographers are able to search for naturally occurring situations in which the same reluctance to read other minds might manifest itself spontaneously, that is, without any outsider's prompting. One important source of evidence for this line of investigation is provided by language socialization studies, as shown by Bambi Schiefflin in this volume and elsewhere (Schieffelin 1990:72–3). The fact that Bosavi adults do not try to guess what an infant is trying to say—Bosavi, like Samoans (Ochs 1982), do not "expand" children's elliptical utterances—shows that the opacity doctrine works across a number of activities and contexts.
Another opportunity to look for further evidence of the local theory is presented through inter-cultural contact situations and the introduction of new activities that typically accompany contact. For example, it makes sense to hypothesize that the opacity doctrine would inhibit or in some way [End Page 484] affect participation in introduced Christian practices such as confession, which relies on the assumption of the existence of an inner self that is accessible to introspection...