How do parents talk about analogy?

• In Event: Cultural perspectives on relational learning: Variation in parental input and children’s performance

Abstract

Language fosters analogical reasoning in at least two fundamental ways. First, relational labels make relations salient and portable, making it easier for children to notice common relations. For example, hearing “daddy-mommy-baby” helped 3-year-olds perceived the relation “big-medium-small” (Gentner & Medina, 1998). Second, language acts as an invitation to compare: when two events are labelled the same way, children are prompted to think of the events’ similarity. Since comparison highlights relational commonalities (Christie & Gentner, 2010) language-invites-comparison serves as a ubiquitous, effective toolkit fostering analogical learning. But despite extensive laboratory evidence showing that language fosters analogical reasoning, we currently do not have a clear picture of how much and what kind of analogy-related language children hear in their everyday learning.
To address this gap, we probed how parents—the main source of linguistic input for young children—talk about relations. We observed parental talk in a semi-natural setting—book reading—and measured 1) relational vs object talk: how much parents talk about relations between objects vs. talking about the objects per se, and 2) comparison talk: whether parents highlighted common relations and objects across the book’s scenes. We gave parents of 4-year-olds (M=50.84 months, range 36-65 months; SD=9.048) a book without words, asking parents to read freely to their children. The book contained an adapted version of the scene analogy task (Richland, Morrison, Holyoak, 2006)—scenes with matching objects and relations that appeared side-by-side. For example, parents saw [a boy chasing a girl] on page 1 and [a dog chasing a cat; a boy standing on the side] on opposite page 2 (Figure 1). The boy on page 1 was identical to the page 2-boy (object match), and the same chasing relation (relational match) was present. That is, the book was designed to afford talking about objects, relations, as well as object and relational matches. All participants are from Beijing, China, and they speak Mandarin Chinese.
We coded parental utterance, classifying them into Relational (e.g., talking about chasing), Object (e.g., talking about the boy), or Other utterances (e.g., talking about the background clouds). Data from 31 parent-child reading showed that in average parents mentioned relations 32% of the time, object 49%, and others 19%. Aligned with prior works using scene analogy task where children found object matches salient, parents also mostly talked about objects. Next, we analyzed parents’ comparison talk—whether they mentioned any similarities across two matching scenes. While parents always saw two comparative scenes side-by-side, they rarely compared them, doing so in only 12% of their total book-reading utterances. Within these comparative utterances, we found no difference between object match (6%) and relational match talk (6%). This non-differential is somewhat surprising given prior assumptions; it’s possible that the pattern of comparison talk may be different in other cultures (Christie et al., 2020).
Overall, this work provides an important window into analogy-language input in young children’s natural environment. We expect to scale up this paradigm to quantify how parents talk about relations in various cultures.

Authors

• Yang Gao, Tsinghua University
  Presenting Author

• Stella Christie, Tsinghua University
  Non-Presenting Author