摘要
It is hypothesized that people possess implicit theories regarding the inherent consistency of their attributes, as well as a set of principles concerning the conditions that are likely to promote personal change or stability. The nature of these theories is discussed in the context of a study of beliefs about life-span development. It is then suggested that people use their implicit theories of self to construct their personal histories. This formulation is used to interpret the results of a wide-ranging set of studies of memory of personal attributes. It is concluded that implicit theories of stability and change can lead to biases in recall. The extent and practical implications of these biases are discussed. Personal memories play an important role in people's everyday lives. Individuals dwell on their pasts for a variety of reasons including entertainment (for others' amusement, people fashion stories out of their lives), curiosity (gazing at their teenage children, parents might wonder what they, themselves, were like as teenagers), and the need to achieve self-understanding. People can study the past to learn about their preferences, abilities, and so forth. Personal recollections are also used to control and manipulate public images. Published autobiographies have served this function for hundreds of years (Korda, 1987). In short, people's personal memories are relevant to some of the traditional concerns of social psychologists, including selfunderstanding and self-presentation. Furthermore, much psychological research depends on personal recall. For example, not so long ago, psychologists formulated theories of development on the basis of parents' retrospective descriptions of their child-rearing practices. We now know that such descriptions may be invalid, and seek more direct evidence (Yarrow, Campbell, & Burton, 1970). Nonetheless, researchers and practitioners continue to make considerable use of retrospective self-reports. These include reports of voting, medical care, purchases, and finances. On the basis of such self-reports, social scientists evaluate theories of human behavior and offer advice on public policy. For a number of reasons, then, it is important to know how personal memories are formed and how accurate they maybe. In 1972, Tulving proposed what has become a widely accepted distinction between episodic and semantic memory. Episodic memory contains information that is coded both temporally and with reference to the rememberer, semantic memory stores general world knowledge that carries neither temporal nor autobiographical codes. In the present article, I examine a