Notice in the table that more 4-month-olds exhibited their first negative vocalization when positioned toward their hands than did 7-month-olds, z = 3.54, p < .001. More 7-month-olds expressed their first negative vocalization toward their mother than did the 4-month-olds, z = 2.99, p < .003. Four- and 7-month- olds did not differ significantly from each other on any other targets. Looking at these results in a slightly different manner, 7-month-olds expressed their first negative vocalization exclusively in the direction of persons, X2 = 12, p < .001. More 4-month-olds expressed their first negative vocalization to the site of the frustration (hands) than they did to the figures in the room, although not significantly more, X2 = 3, p < .10. The same overall pattern of results was observed when the first negative vocalizations of the infants were tabulated irrespective of period of onset. The vocalic expressions of 7-month-olds, then, were decidedly person specific; those of four-month-olds were task specific. One-month-olds were indiscriminate. Not only were 1-month-olds' first negative vocalizations not discriminatively expressed in period 2, but even when they were first observed these infants were facing the miscellaneous target areas as often as persons or hands (7 vs. 7), X2 = 0, p = ns.
The presence of crying did not account for the anger facial behaviors reported previously. During the initial postrestraint facial expression scoring segment among 4- and 7-month-olds, there were 18 subjects who showed the "semidiscrete template," and 10 who showed the "discrete anger template" (see Table 10.3). Only 2 of the 18 subjects cried while manifesting the "semidiscrete template," and only I of 10 babies cried while showing the "discrete anger template."
Nor did the data conform to the hypothesis that the infants were posturing their faces in the predicted manner in anticipation of crying (i.e., a precry face). The expression of anger was not more prevalent in crying versus noncrying children. During the prolonged facial scoring epoch, when more than half the subjects were crying, the presence of crying was not significantly related to the expression of either type of template.
This study lends strong support to the claim that the capacity to express anger through coordinated facial patterning emerges before 4 months of age. In addition, it reveals that by at least 4 months anger facial displays may function as discrete social signals. These signals are at first directed proximally to the immediate source of frustration, but by 7 months they become expressed directly to