Abstract
Self-continuity – the sense that one’s past, present, and future are meaningfully connected – is considered a defining feature of personal identity. However, bases of self-continuity may depend on cultural beliefs about personhood. In multilevel analyses of data from 7287 adults from 55 cultural groups in 33 nations, we tested a new tripartite theoretical model of bases of self-continuity. As expected, perceptions of stability, sense of narrative, and associative links to one’s past each contributed to predicting the extent to which people derived a sense of self-continuity from different aspects of their identities. Ways of constructing self-continuity were moderated by cultural and individual differences in mutable (vs. immutable) personhood beliefs – the belief that human attributes are malleable. Individuals with lower mutability beliefs based self-continuity more on stability; members of cultures where mutability beliefs were higher based self-continuity more on narrative. Bases of self-continuity were also moderated by cultural variation in contextualized (vs. decontextualized) personhood beliefs, indicating a link to cultural individualism-collectivism. Our results illustrate the cultural flexibility of the motive for self-continuity.
Notes
* Study materials and data have been deposited with the UK Data Service (Vignoles & Brown, Citation2016).
1. Similar estimates for the moderation effects of mutable (vs. immutable) personhood beliefs were found in a model without controls for contextualism beliefs, gender and age.
2. This pattern of findings would not offer an alternative explanation of the differences observed by Chandler et al. (Citation2003) between First Nations and European-heritage Canadians, where the former relied more on narrative and the latter on stability. Here, stability was a stronger basis for continuity not only in cultural groups with more contextualized personhood beliefs, but also among individuals with more decontextualized personhood beliefs. Although it is intriguing to find opposing effects at two different levels of analysis, the individual-level moderation especially should not be over-interpreted, considering that it only very narrowly reached statistical significance at the conventional .05 level.