Do Children Recall Numbers as Generic? A Strong Test of the Generics-As-Default Hypothesis

ABSTRACT

A striking characteristic of human thought is that we form representations about abstract kinds (Giraffes have purple tongues), despite experiencing only particular individuals (This giraffe has a purple tongue). These generic generalizations have been hypothesized to be a cognitive default, that is, more basic and automatic than other forms of generalization, including those involving quantifiers such as “all” or “most.” In support of this hypothesis, children often recall quantified statements (e.g., “All/many/most bears climb trees”) as generic (“Bears climb trees”), and do so more frequently than the reverse error of recalling generics as quantified. The present study provides a strong test of the generics-as-default position by testing whether even numerically quantified statements (e.g., “Five giraffes have purple tongues”) are recalled as generic. Two groups of three-year-old children (N = 74) were tested: those who held a correct numerical interpretation of five (“5-knowers”), and those who did not (“non-5-knowers”). Results indicate that non-5-knowers often defaulted to the generic in recall after hearing a number (e.g., recalling “Five giraffes.” as “Giraffes.”), whereas 5-knowers did not. Thus, consistent with the generic-as-default hypothesis, before children have a specific numerosity assigned to “five,” they display a tendency to recall numerically quantified phrases as generic.

Acknowledgment

The research was supported by NICHD grant HD-36043 to S. Gelman. We are grateful to the children, parents, and teachers of Annie’s Preschool and Gretchen’s House for participating in the research. We also thank Natalie Davidson, Megan Martinez, Merranda McLaughlin, and members of the Conceptual Development Lab at the University of Michigan for their research assistance.

Disclosure

No potential conflict of interest was reported by the authors.

Notes

1. If that were the case, then we would expect better performance on the Give-A-Number task for those in the Generic/Five condition than those in the Generic/Some condition. However, no such differences were obtained (including counting consecutively to at least 5 without the researcher’s help, spontaneously mentioning 5, and highest consecutive number reached without the researcher’s help), all ps > .22.

2. For one child, test trials were administered in the incorrect order; this child nonetheless was classified as a 1-knower, due to two successes on “1” and two failures on “2.”

3. When reviewing the transcribed responses, we noticed that some children referred to the pictures during the recall phase (e.g., pointing to each animal in turn while saying, “This one and this one and this one and this one have long mouths”). Referring to the set of instances in the picture is plausibly an alternate way of expressing the 5-statements provided during the teaching phase. We therefore conducted a supplementary set of analyses in which the specific plural responses (also including, e.g., “These guys have strong tongues”; “The snakes swallow food in holes”) were coded as Number responses, and thus correct when in response to 5-sentences, and opposite when in response to generic-sentences in the generic/five condition. We found that precisely the same patterns obtained, whether or not the specific-plural responses were included, and thus these responses are not considered further.

Additional information

Funding

This work was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development [HD-36043].