Dataset for Investigating Menzerath's Law in Crows and Humans During Cued Vocal 'Counting'

To uncover universal principles of vocal behavior, studies have examined whether linguistic laws apply across species. Menzerath’s law, for instance, posits that larger constructs have shorter constituent parts, reflecting an efficiency principle that reduces articulatory and energetic demands. Another proposed universal is final lengthening, where end-of-sequence vocalizations are extended to potentially signal sequence boundaries. While studies have examined these temporal patterns in ecological communicative contexts, it remains unclear whether they also emerge in more constrained, externally instructed vocal responses. This study examines temporal patterns in vocal sequences during a numerically cued production task in crows and humans. While both species performed the same task, they differ in numerical representation: crows rely on an approximate number system, whereas humans possess a symbolic understanding of number. Crows showed a negative correlation between sequence length and vocalization duration—consistent with Menzerath’s law—and a positive correlation between sequence position and duration, consistent with final lengthening. In contrast, humans exhibited the opposite: vocalization duration increased with sequence size and decreased with position. These findings suggest that Menzerath’s law is not a universal principle but shaped by species-specific cognition, motor constraints, and task demands. This highlights the importance of considering context and cognitive capacities when interpreting temporal patterns in vocal behavior across species.