Young Children’s Understanding of the Deductive Inference as a Source of Knowledge

Abstract

Over the past decade, developmental psychology has extensively examined young children’s reasoning abilities, yielding contradictory and puzzling findings. While five-year-old children are known to successfully apply the deductive inference (“A or B”, “Not A”, “Therefore, B”), three-year-olds have difficulty using such operations. These findings are in contrast with a set of positive evidence in similar tasks with infants, suggesting that some foundational reasoning abilities may be present earlier. This discrepancy highlights a developmental gap in how and when these concepts become available during a human child’s growth trajectory. According to a recent theoretical attempt aiming to explain these findings, children younger than four lack modal concepts (distinguishing what is possibly true from what is necessarily true), which prevents them from representing alternatives (as possible or necessary), but they are able to generate expectations based on a simulation (minimal representation of possibilities); whereas children over four have access to modal concepts, enabling them to represent multiple possibilities simultaneously [1].

Our project aims to investigate whether young children consider deductive inference as a reliable source of knowledge. It also seeks to verify the predictions of the minimal theory on the acquisition of modal concepts. To achieve this, we plan to investigate whether young children are able to evaluate the validity of deductive conclusions in a reasoning task, asking them to judge whether some conclusions necessarily follow from a set of premises. Importantly, we assess how effectively young children gauge the necessity of the conclusions drawn.

Our experiment adapts the invisible hiding paradigm where one target object is hidden in one of three possible locations (A, B, or C), while a fourth location (D) is never involved in the hiding. Each trial starts with the elimination of an alternative, and finishes with a conclusion. We ask children to evaluate the certainty of different conclusions by using a discrete scale (1: not sure / 2: somewhat sure / 3: very sure) to assess whether children understand that the valid conclusions necessarily follow.

After being shown that a target object is not in location A, participants are presented with three scenarios: i) Guessing (“Not A,” “Therefore, B”), ii) Correct Inference (“Not A”, “Therefore B or C”), or iii) Impossible Inference (“Not A”, “Therefore, D”). While three-year-olds are likely to reject a case referring to physical impossibility, they are expected to accept the conclusion in the Guessing condition as valid if they lack modal concepts and cannot entertain multiple possibilities, as predicted by the minimal representation account. This is not the case for the five-year-olds; instead, they should choose the Correct Inference conclusion with high confidence.

The project’s significance stems from exploring whether children understand that deductive inferences yield necessary conclusions, rather than just likely or contingent ones. Its results would also deepen our understanding of the developmental origins of modal concepts.

References

[1] B. P. Leahy and S. E. Carey, “The Acquisition of Modal Concepts,” Trends Cogn. Sci., vol. 24, no. 1, pp. 65–78, Jan. 2020. doi: 10.1016/j.tics.2019.11.004.