Adult perception of emotion intensity in human infant cries: effects of infant age and cry acoustics.

The relation between adult perception of emotion intensity in the cries of 1- and 6-month-old infants and the acoustic characteristics of the cries was examined. In the first study, adults who were inexperienced in child care rated 40 cries on 3 emotion intensity scales: anger, fear, and distress. The cries of 6-month-olds were rated as being significantly more intense. Different acoustic variables accounted for emotion intensity ratings for the 2 infant ages. Peak amplitude and noisiness of the cry predicted adult judgments of intensity ratings of 1-month-olds' cries; a measure of amplitude ratio (in 2 frequency bands) was the best predictor of intensity ratings of 6-month-olds' cries. In the second study, parents of infants rated the same cries on the same scales. They also rated the older infants' cries as being more intense. The 2 adult groups did not differ on their ratings, and a regression equation derived from one adult group predicted the other adult group's rating of the same infant age better than it predicted its own ratings for the other infant age. Infant age, and its associated acoustic features, seems to be a more important determinant of adults' perception of emotion intensity than are such adult characteristics as gender or infant-care experience.

Contextual sources of information and responses to animal communication signals.

The response of signal recipients depends on information found in the signal itself and on other information sources known collectively as context. Context is the set of events, conditions, and changeable recipient characteristics that modify the effect of a signal on recipients' behavior. Sources of contextual information include (a) the characteristics of the recipient and (b) sources external to the recipient (the signaler and the setting). Contextual information is most commonly used when the signal itself does not provide enough information for the recipient to behave adaptively. The signal's referent may be ambiguous or deceptive, or recipients may need additional information to fine-tune their responses. When animal signals were thought to be information poor, contextual information was seen as being critically informative. Animal signals are now known to be much more informative, but even information-rich systems, such as human language, depend heavily on contextual sources of information.

Relation between food preference and food-elicited vocalizations in golden lion tamarins (Leontopithecus rosalia).

We report the results of 2 studies on food-elicited vocalizations in golden lion tamarins (Leontopithecus rosalia). First, we investigated the preferences of 10 golden lion tamarins for 6 foods. Tamarins prefer mealworms and raisins significantly more than apple, egg, carrot, or marmoset diet. Food preference rank was significantly and positively correlated with the rank of latency to choose a particular food. Second, we investigated the relation between food preference and 15 vocal parameters measured from the calls emitted by 5 tamarins to a subset of the foods. Only 1 parameter was significantly correlated with food preference across animals. Within-subjects multivariate analysis of variance showed that the vocalizations to foods are significantly different. Our results support an hypothesis that food-elicited vocalizations vary in ways that correspond to the caller's preference but not in a manner that labels food type.

Cross-gender perceptions of facial attributes and their relation to attractiveness: do we see them differently than they see us?

Examination of perceptions of human facial attributes revealed that individual attributes are similarly perceived by males and females. However, patterns of attribute interrelationships differ as a function of gender of the face. Undergraduate students (N = 280) rated pictures of 40 male or female Caucasians on 12 physical attributes (e.g., nose size, face width) and overall attractiveness. The four sets of attribute ratings (defined by rater gender and picture gender) were submitted to principal components analyses, and five-factor solutions were found for each condition (accounting for about 76% of the variance). Comparisons of the four component solutions using confirmatory factor procedures revealed that male and female raters share one factor structure when rating photographs of female faces and another factor structure when rating photographs of male faces. Multiple regression analyses revealed that the patterns of attribute interrelationships were not "perceptual units" in the perception of attractiveness, and that different "rules" are used to assess the attractiveness of male and female stimuli faces. The importance of these results for models of facial attractiveness and interfacial similarity judgements are discussed.

Sex differences in vocal structure in a callitrichid primate, Leontopithecus rosalia.

Four types of calls (long calls, chirps, chucks, and trills) from Golden Lion Tamarins (Leontopithecus rosalia) were analyzed for sex differences in structure. Six or more acoustic variables were analyzed for each call type. One or more variables from each call category differed significantly between the sexes. In addition, discriminant function analyses were conducted on male vs. female data for each call type. Long calls, chirps, and chucks from males and females were quite discriminable, but trills were not. That these differences exist is surprising given that callitrichid primates are otherwise quite monomorphic in structure and behavior. Possible functions served by sex differences in call structure include group censusing or assessing the sex of transient individuals.