Infants' and adults' perception of scale structure.

Adults and 9-month-old infants were required to detect mistuned tones in multitone sequences. When 7-tone versions of a common nursery tune were generated from the Western major scale (unequal scale steps) or from an alternative scale (equal steps), infants detected the mistuned tones more accurately in the unequal-step context than in the equal-step context (Experiment 1). Infants and adults were subsequently tested with 1 of 3 ascending-descending scales (15 tones): (a) a potentially familiar scale (major) with unequal steps, (b) an unfamiliar scale with unequal steps, and (c) an unfamiliar scale with equal steps. Infants detected mistuned tones only in the scales with unequal steps (Experiment 2). Adults performed better on the familiar (major) unequal-step scale and equally poorly on both unfamiliar scales (Experiments 3 and 4). These findings are indicative of an inherent processing bias favoring unequal-step scales.

Mothers' and fathers' singing to infants.

Mothers and fathers sang a song of their choice, once to their infant and once as if to their infant (simulated). The pitch of songs was higher and the tempo slower for infant-directed than for simulated versions. Listeners varying in child-care experience, musical background, and cultural origin reliably identified which of the two versions was infant-directed (Experiment 1). Identification accuracy was enhanced by musical training, knowledge of the singers' language and culture, and by greater pitch and tempo differences between versions. Other adult listeners who rated the singer's emotional engagement assigned significantly higher ratings to infant-directed than to simulated versions (Experiment 2). Differences in pitch and tempo between both versions predicted emotional engagement ratings. Finally, rating differences between infant-directed and simulated versions were highly correlated with identification accuracy.

Parents' sung performances for infants.

Naive listeners rated the style of singing in mothers' and fathers' sung performances for infants and their simulations of those performances (Experiment 1). Performances in an infant's presence were judged as more expressive--either more playful or more soothing--than were simulations. Parents' style of singing, as reflected in these ratings, differed as a function of the sex of singer and listener. Both parents sang more playfully for same-sex infants than for opposite-sex infants. Independent listeners rated the manner in which parents enunciated the lyrics of their songs (Experiment 2). Parents rendered the lyrics of songs more expressively in infant-present than in infant-absent contexts. Moreover, this expressiveness was greater for same-sex infants than for opposite-sex infants. These findings are consistent with parents' greater attachment to same-sex infants. Discrepancies between parents' choice of songs and their manner of singing lend credence to functional rather than nominal classifications of songs for infants.