Figure 5.

Imitation versus contingent hypotheses. (a) Schematic illustration of the imitation and contingent hypotheses. In the imitation hypothesis, irrespective of the infant vocalization dynamics, changes in infant call acoustics are solely driven by parental calls. In the contingent hypothesis, the dynamics of the infant call acoustics both modulates and is influenced by the dynamics of parental vocalization. (b) Power spectrum of the temporal dynamics of Wiener entropy and duration of infant calls. Powers were measured in normalized units (n.u.) (c) Probability density distribution of the phase of the infant call entropy dynamics at which the parents vocalize. More negative Wiener entropy values indicate more adult-like calls. The top panel is a representation of the phases. Zero and 180° represent the trough and peak, respectively. The bottom panel represents, on a polar coordinate (degrees), the probability density distribution of the most probable phases for each session. The solid red line indicates the values of the probability distribution. (d) Probability density distribution of the phase of the infant call duration dynamics at which the parents vocalize. Larger values indicate more adult-like calls. Convention is the same as in (c).