Introduction

I would like to begin this chapter by drawing a distinction between identification and classification. At one level, there is no difference between the two: both processes concern the allocation of an item to a (usually) named category, or the process by which an object is placed in a class. Both entail the matching of perceptual images, words and concepts [Ohnuki-Tierney, 1981: 453], and may operate equally in terms of unmodified sense data or their cultural representations. One logically presupposes the existence of the other [Colless, 1970: 252]. A pragmatic distinction can be made, however, in terms of the way in which Nuaulu informants assign observed animal specimens to terminal categories and the way in which categories are arranged into more inclusive groups. This is reflected linguistically in the difference between the statements:

and

Both are relations of class inclusion but are distinguished in most languages, including Nuaulu. Thus the statement:

is one of identification, where the terminal -nea in this context denotes demonstrative emphasis; whereas the statement

is an abstract statement of classification.

Identifying animals in natural settings

The question as to what is a culturally appropriate identification is itself problematic. Why should the ethnographer accept one person's answer against another?

Background to the study

I first undertook fieldwork among the Nuaulu of south central Seram between 1969 and 1971. During this research, with its ecological orientation [Ellen, 1978b], it became increasingly evident that in order to analyse adequately how this eastern Indonesian people interacted with their environment, it was necessary to pay attention to the way in which it was apprehended and classified [Ellen, 1982]. The point, of course, had been made much earlier by Conklin [Conklin, 1957], and had been implicit in much of his subsequent work, as well as in that of others who owe him an intellectual debt. At the same time I was already intrigued and impressed by new ethnobiological reports then appearing, particularly the work of Berlin and his various associates in Mexico, and that of Bulmer and his associates working in Papua New Guinea.

Specifically, my interests were motivated on the one hand by a recognition that ethnobiology was a neglected (though fundamental) part of social anthropology, with implications for the study of subsistence behaviour, ecology, categorisation and belief; and on the other by a fascination with and an admiration for the rigorous techniques employed, and the detailed data obtained, by a handful of dedicated ethnographic enthusiasts. I judged it of some interest to attempt to replicate and evaluate these methods and results in a region – eastern Indonesia – for which they were at that time unknown.

Order and disorder in Nuaulu conceptions of nature

Simpson, of course, is speaking here of the taxonomic practice of biologists. Nevertheless, his remarks are highly pertinent to the examination of folk-biological classifications. I wish to argue in this chapter that even if we agree that it is possible to locate basic categories which have a strong likelihood of corresponding with cognitive prototypes contingent upon objective discontinuities in nature, and which therefore present a ready point of reference in cross-cultural studies, this is far less likely to be so when we are dealing with more inclusive categories. This is because it is more difficult to ensure that the categories we discover are fully contrastable, operating at the same ‘level’, regularly labelled, non-overlapping, and generally shared within a population. Paul Taylor [1990: 68] voices this scepticism when he rhetorically presumes that those who claim some distinctive characteristics for levels above basic categories [e.g. Brown, 1979; Witkowski and Brown, 1977] are able to identify those levels to which their generalisations apply. The difficulties seem to me to arise for two contradictory reasons.

How well does one remember the emotional events in one's life? Many studies indicate a strong correlation between the vividness with which an event is recalled, and the emotionality of the event, at the time it occurred (Bohannon, 1988; Brown & Kulik, 1977; Christianson & Loftus, 1990; Conway & Bekerian, 1988; Pillemer, 1984; Rubin & Kozin, 1984). Interestingly, this correlation seems independent of the type of emotion at stake (Reisberg, Heuer, McLean, & O'Shaughnessy, 1988; Robinson, 1980; White, 1989). Thus, in general, the stronger the emotion, the greater the vividness of subsequent recall. In this regard at least, “flashbulb” memories are continuous with remembering in general; that is, flashbulb memories simply represent the extreme of this affect–vividness relationship.

But how accurate are these vivid and detailed memories? If emotional events are recalled with great detail, is this evidence for some encoding or retrieval advantage associated with emotionality? Or, alternatively, is this evidence for high levels of construction and confabulation associated with emotion? In the former case, emotional memories would be both vivid and also veridical; in the latter case, emotional memories might be filled with errors and intrusions.

This question about accuracy is fueled by several concerns. As we describe below, several studies have documented conspicuous errors in the recall of emotional events. At the least, this implies that neither emotionality nor memory vividness provides any guarantee of memory accuracy.

In the classic paper on “flashbulb” memories, Brown and Kulik (1977) identified an aspect of autobiographical memory that had received scant attention from cognitive psychologists – memory for one's own personal circumstances (such as location, ongoing activities, and feelings) when receiving an important and shocking piece of news (such as first learning that President Kennedy had been shot), as opposed to memory for the newsworthy event itself. A second contribution of Brown and Kulik's original papers, and Neisser's (1982b) subsequent commentary, was a focus on memory function. What is the psychological value or adaptive significance of remembering personal details at such times?

In this chapter I examine memories of personal circumstances from a functional perspective. First, I argue that memories of personal circumstances are essential components of a fully functioning autobiographical memory system. Second, I critically examine the emphasis on memory accuracy in recent studies of flashbulb memories, and I conclude that the research agenda should be expanded to include memory functions that have heretofore fallen outside the realm of cognitive psychology. Third, I identify and describe three broad categories of memory function, none of which depends on absolute veridicality in recall:

1. Communicative functions: The act of recounting a detailed personal memory to others communicates meaning that transcends the surface content of the particular recollection, and this specialized form of communication appears to be rule governed.

2. Psychodynamic functions: Remembering personal circumstances in vivid detail can have a profound emotional and psychological impact on the rememberer, and on others with whom the memories are shared. Detailed recollection of specific episodes is frequently identified by clinicians as an essential component of psychotherapeutic process.

3. […]

Following the appearance of the now classic paper on “flashbulb” memory by Brown and Kulik (1977), Winograd and Killinger (1983) conducted a retrospective developmental study of the phenomenon. In reviewing Brown and Kulik's (1977) data, they noted that 24 of the 80 informants had been between 7 and 11 years of age at the time (1963) of the Kennedy assassination, yet their reports of the circumstances under which they discovered the news were no different from those informants who were adults at the time. A similar pattern was noted by Yarmey and Bull (1978), who found that only 15% of their subjects who had been between 6 and 10 years old in 1963 responded that they could not remember what they had been doing at the time they heard of Kennedy's assassination. Apparently, age at the time of encoding had little relation to subsequent recollections of the event's discovery. Winograd and Killinger (1983) thus set out to determine whether there were any differences between the memories of those who had been 7 or older at the time of the event versus those who had been younger. If age differences emerged, perhaps they could shed some light on both the development of autobiographical memory in general and the factors that contribute to flashbulb memories.

Winograd and Killinger (1983) collected questionnaires from 338 college students who had been 1 to 7 years old at the time of John F. Kennedy's assassination.