Newborns' Cries are Unique to Individuals (But Not to Language Environment)

Human neonates exhibit individually distinctive patterns of variation—sometimes called cry “signatures”—in the acoustics of their cries. For tests of group differences (e.g., age, health status, or ambient language differences), the uniqueness of cry sounds to individuals means that multiple cries from same infant cannot appropriately be treated as independent events for analysis. This approach violates the assumption of independence that underlies many statistical techniques and runs the risk of uncovering “significant” group effects that do not actually exist. Nonetheless, publications continue to appear in which multiple cries from the same infant have been treated as independent events for analysis. To demonstrate the perils, we have compared the cries of infants born into a Mandarin Chinese language environment with cries of infants born into an American English language environment. We show (1) that treating each cry sound as an independent event yields numerous “significant” group effects and (2) that these effects all but disappear when cries are treated appropriately, as “nested” within individual infants. The latter outcome is in keeping with prevailing models of neonatal cry production and casts doubt on the claim that neonates alter their cries to match features of the language spoken in their environment.