Several recent investigations of children's cognition of musical pitch
have examined the nature of children's sensitivity to the ''tonal hie
rarchy'' identified by Krumhansl (1990a). These studies presented chil
dren with musical ''contexts,'' asking them to make judgments about su
bsequent pitches, and have produced strikingly divergent results. Fact
ors of age and of type of musical material used in context appear to p
lay significant roles in determining subjects' sensitivities. This pap
er describes two experiments that examine the time course of the devel
opment of children's cognitive representations of pitch relations, tak
ing into account the contributions made to such representations by str
uctural and by temporal factors (following West & Fryer, 1990). A prob
e-tone technique was used with two contrasting context types, one bein
g a ''typical'' cadential sequence and the other consisting of differe
nt randomizations of the diatonic collection. This experiment was cond
ucted on 285 children ranging between 6 and 11 years old, from two dif
ferent single-sex schools. The results were further investigated in a
game-playing experiment-using chime bars-with children from each age g
roup represented in the first experiment. These experiments appear to
indicate that children's early representations of pitch relations are
remarkably stable and that development may take the form of an increas
ing sensitivity to time-dependent characteristics of the musical surfa
ce leading to an internalization of the tonal hierarchy. Despite the d
ifferent methodologies used here, results are broadly in line with tho
se suggested by Krumhansl and Keil (1982), although children's represe
ntations of musical pitch as exhibited here appear to be more sophisti
cated than would be implied in that study.