Semantic Elaboration through Integration: Hints Both Facilitate and Inform the Process

Abstract

Semantic knowledge can be extended in a variety of ways, including self-generation of new facts through integration of separate yet related episodes. We sought to promote integration and self-generation by providing “hints” to help 6-year-olds (Experiment 1) and 4-year-olds (Experiment 2) see the relevance of separate episodes to one another. Varying the timing of presentation of the hints—whether provided between episodes or before test—afforded insights into the processes responsible for self-generation of new semantic knowledge through integration. In both experiments, hints provided immediately before testing facilitated performance relative to a control condition. In contrast, hints provided between episodes did not. The findings imply that integration of distinct experiences is not an automatic process set into motion based on shared elements. Even when integration is primed through a hint that a prior episode is relevant to a forthcoming experience, shared elements apparently are not sufficient to promote self-generation of new semantic knowledge. Instead, the process seems to occur upon demand, such as in response to a question. These findings can be used to guide interventions designed to promote enrichment of semantic memory and also further our understanding of the basic cognitive processes involved in extension of knowledge.

ACKNOWLEDGMENTS

The authors thank members of the Memory at Emory laboratory group for assistance with various aspects of the research. They extend a special note of thanks to the children and families who gave of their time to take part in the study.

Notes

¹In prior related research (Bauer et al., 2012; Bauer & San Souci, 2010; Varga & Bauer, 2013), we have scored children's recall and recognition of the stem facts presented in the context of the passages of text and then analyzed relations between recall and total performance for the stem facts and self-generation and recognition of the integration facts. Memory for stem facts has been shown to be necessary but not sufficient for self-generation and recognition of integration facts. That is, children who self-generate integration facts have high levels of recall of stem facts, yet high levels of recall of stem facts do not ensure self-generation and recognition of integration facts. In the present research, parallel analyses yielded a replication of this general effect; no new insights into integration and self-generation processes were provided by analysis of memory for the stem facts either in Experiment 1 or in Experiment 2. Because the analyses yielded no unique information relative to prior related research and because relations between stem-fact recall and self-generation under different hint conditions were not the focus of the research, we do not present the details of the analyses (specific results are available from the first author).