Crossing the Boundary: No Catastrophic Limits on Infants' Capacity to Represent Linguistic Sequences

Abstract

The boundary effect, namely the infants' failures to compare small and large numerosities, is well documented in studies using visual stimuli. The prevailing explanation is that the numerical system used to process sets up to 3 is incompatible with the system employed for numbers >3. This study investigates the boundary effect in 10-month-old infants presented with linguistic sequences. In Condition 1 (2 vs. 3), infants can differentiate small syllable sequences (2 vs. 3), with better performance for the 2-syllable sequence, which imposes a lower memory load. Condition 2 (2 vs. 4) revealed that infants are capable of discriminating across bounds, with relatively higher performance for the 4-syllable sequence, possibly encoded as one large ensemble. This study offers evidence that, when processing linguistic sounds, infants flexibly deal with small and large numerical representations with no boundaries or incompatibilities between them. Simultaneously encoding units of different magnitudes might aid early speech processing beyond memory limits.

Keywords: auditory discrimination; boundary effect; language processing; number representation; syllables.

FIGURE 1

Illustration of the experimental procedure and timeline. At the bottom, the flow of the entire experiment is shown, highlighting the structure of each block (each color represents one block). In the upper part, inside the blue square, the flow for each familiarization and test trial is depicted. The screenshot showing the child illustrates the online setup.

FIGURE 2

Main results of Condition 1—infant's ability to discriminate between sets of 2 and 3 syllables. These findings are summarized as follows: (A) Normalized difference scores based on first‐look duration. (B) Individual distribution of the normalized difference scores calculated on the longest look: Colored dots represent individual participants; black dots indicate the group means, and bars the standard error. Chance is determined by the dotted black line in the middle. (C) Proportion of gaze duration (correct side).

FIGURE 3

Main results of Condition 2—infants’ ability to discriminate between sets of 2 and 4 syllables. These findings are summarized as follows: (A) Normalized difference scores based on the first‐look measures. (B) Individual distribution of the normalized difference scores calculated on the longest look: Colored dots represent individual participants; black dots indicate the group means and bars the standard error. Chance is determined by the dotted black line in the middle. (C) Proportion of looking time (correct side).

FIGURE 4

Comparison of the infants’ performance in Conditions 1 and 2. The y‐axis depicts mean difference scores. Error bars show standard errors of the mean.