The effects of language and priming on the relative accessibility of the private self and the collective self.

Several researchers (e.g., Greenwald & Pratkanis, 1984; Trafimow, Triandis, & Goto, 1991; Triandis, 1989) have argued for a distinction between the private self and the collective self. The private self includes cognitions that involve traits, states, or behaviors (e.g., "I am honest"). The collective self contains cognitions about group membership (e.g., I am a member of my family). According to Triandis (1989; also see Hofstede, 1980; Hsu, 1981, 1983, 1985), there is a great deal of accumulated data suggesting that the private self is emphasized more in individualistic cultures than in collectivist ones, but the reverse is true concerning the collective self. For example, Triandis, McCusker, and Hui (1990) found that people from individualistic cultures (e.g., the United States) tend to retrieve more private self-cognitions and fewer collective self-cognitions than do people from collectivist cultures (e.g., China). Further, Bochner (1994) found that Malaysians retrieved more collectivist responses than a combined Australian/British sample. Trafimow et al. (1991) argued that there are at least two possible types of cognitive organizations that can account for these findings which they called the one-basket and two-baskets theories, as indicated below:

There are at least two ways that private and collective self-cognitions

could be organized in memory. First, these cognitions may be stored in a

single cognitive structure (i.e., the same place in memory), but with no

internal organization. Thus, the likelihood of retrieving a self-cognition

of a particular type (e.g., a private self-cognition) is determined by the

frequency with which that type of self-cognition is represented, relative

to the totality of self-cognitions. To use a rough analogy, consider a

basket containing red and blue marbles. The probability of retrieving a

red marble depends upon the number of red marbles, relative to the total

number of marbles in the basket.

Another possibility is that self-cognitions are organized according to

whether they pertain to the private or collective self. More specifically,

private self-cognitions are organized around a general private self-concept,

but collective self-cognitions are organized around a general collective

self-concept. Thus, accessibility of different types of self-cognitions is

largely determined by the relative accessibility of the private and

collective self-concepts. To use the basket analogy again, consider two

baskets of marbles, a red and a blue basket containing red and blue marbles,

respectively. The probability of retrieving a red or blue marble depends

upon which basket that person samples from. (p. 649)

Trafimow et al. (1991) obtained two types of findings that supported the two-baskets theory but contradicted the one-basket theory. First, they found that priming the private self increased the retrieval of private self-cognitions, but priming the collective self increased the retrieval of collective self-cognitions. According to the one-basket theory, people do not have separate private and collective self-concepts to prime, so the priming manipulations should have had no effect. Second, retrieval of self-cognitions was clustered by type. Specifically, the probability of retrieving a private self-cognition was greater following the retrieval of another private self-cognition than following the retrieval of a collective self-cognition, and the probability of retrieving a collective self-cognition was greater following the retrieval of another collective self-cognition than following the retrieval of a private self-cognition. If the one-basket theory were true, then the probability of retrieving a particular type of self-cognition should not depend upon the type of self-cognition that was previously retrieved. Thus these data are inconsistent with the notion that private and collective self-cognitions are stored in the same location in memory, so they must be stored in different ones. Finally, it is worth noting that the predictions described above were confirmed not just for individualistic subjects (i.e., American university students), but for collectivist ones as well (i.e., Chinese students studying in America), which suggests that the two-baskets theory generalizes across cultures.

There are, however, some limitations concerning the Trafimow et al. (1991) data. First, because the collectivist subjects used in their experiment (Experiment 1) were all in America at the time, it is possible that exposure to Americans caused the formation of a distinct private self-concept that would otherwise not have been formed. A second limitation is that due to practical considerations, they failed to use a control group that did not receive either a private or collective self-prime. Two sources of ambiguity arise from this failure to use a "no-prime" condition. One source of ambiguity resides in the possibility that the prime itself caused people to distinguish between private and collective self-cognitions, which in turn caused the clustering of self-cognitions that were elicited from the subjects. To the extent that this is plausible, it compromises an important portion of the support for the two-baskets theory. A replication of the conditional probabilities under a no-prime control condition would resolve this difficulty. A second source of ambiguity is the difficulty in knowing whether it was the private self-prime, the collective self-prime, or both that caused the obtained differences in the proportions of private and collective self-cognitions that were retrieved. A comparison between these groups and a no-prime control condition would solve this problem. One purpose of the present experiment is to eliminate these limitations.

EVIDENCE THAT LANGUAGE CAN ACT AS A PRIME

There are reasons to suspect that the language one uses can increase the accessibility of either the private or the collective self. Consider this from a developmental perspective. It is not clear when the private self and the collective self develop as distinct concepts in memory that are associated with private self-cognitions and collective self-cognitions, respectively. However, it obviously must happen some time or else the two-baskets theory would be wrong. In addition, it seems reasonable to assume, in general, that such development occurs in people's native culture where the native language is spoken. In that case, there might be "language …