Albert J. Caron

Infant discrimination of naturalistic emotional expressions: The role of face and voice.

The ability of infants to discriminate dynamic, multimodal expressions of emotion was assessed in a series of 5 experiments. In Experiment 1, 48 infants of 4 and 5 months (total N = 96) were habituated to color/sound videotapes of 6 women speaking the same script sadly or happily. Following habituation, 2 new women were presented, each speaking once in the familiarized emotion and once in the novel emotion. Order of stimulus presentation (Sad----Happy, Happy----Sad) was counterbalanced. 5-month-olds were able to discriminate the expressions in both directions, whereas 4-month-olds could discriminate them only in the Sad----Happy direction. In Experiment 2, the ability of 5- and 7-month-olds to discriminate happy and angry expressions was examined using the Happy----Angry stimulus order alone. Only the 7-month-olds could differentiate these stimuli. In Experiment 3, it was shown that 7-month-olds could not distinguish these same Happy----Angry stimuli without vocal accompaniment. The purpose of the fourth experiment was to determine whether the voice played an equally important role in the Sad----Happy discrimination of Experiment 1. Surprisingly, a 5-month group tested without voice readily discriminated these stimuli. Finally, the fifth experiment sought to determine whether an Angry----Happy comparison might also be discriminable without voice. A 7-month group tested in this manner could not discriminate these expressions, while a group tested with voice could. The results indicate that infants can differentiate dynamic, multimodal expressions as early as 5 months, that they distinguish dynamically distinct expressions earlier than more similarly animated expressions, and that they seem to rely more on the voice than the face in making these discriminations.

Abstraction of Invariant Face Expressions in Infancy.

To determine whether infants can abstract invariant face expressions across different persons (i.e., can form face expression categories), groups of 18-, 24-, and 30-week-old infants (18 boys and 18 girls per group) were habituated by the infant control procedure to photographs of 4 different female faces all wearing an identical expression (happy or surprise). In an immediately following test phase, categorization was inferred from greater generalization of habituation (less recovery of fixation) to 2 new female faces in the familiarized expression than to the same new faces in the altered (novel) expression. To rule out the possibility that generalization at test might be due to failure to discriminate the new persons, control groups of 18 boys and 18 girls at each age saw the same test faces following repeated presentations of only 1 of the 4 habituation faces. The results indicated that not until 30 weeks could infants differentiate happy and surprise expressions on a categorical basis. At 24 weeks they could distinguish a surprise expression following habituation to happy faces, but could not do the reverse. At 18 weeks they could do neither. Overall, the performance of girls was superior to that of boys. The findings are consistent with recent evidence suggesting that the ability to extract invariant configural information relative to the human face does not emerge until about 7 months of age.

Do Infants See Emotional Expressions in Static Faces

To determine whether young infants discriminate photographs of different emotions on an affect-relevant basis or on the basis of isolated features unrelated to emotion, groups of 17-, 23-, and 29-week-olds were habituated to slides of 8 women posing either Toothy Angry, Nontoothy Angry, or Nontoothy Smiling facial expressions and were then shown 2 new women in the familiarized expression and in a novel Toothy Smiling expression. At all 3 ages, recovery to the novel Toothy Smiling faces occurred only after habituation to Nontoothy faces (whether smiling or angry), not after habituation to Toothy Angry faces, indicating that infants had been responsive to nonspecific features of the photographs (presence or absence of bared teeth) rather than to affectively relevant configurations of features. In a second experiment, 2 older age groups (35 and 41 weeks) also proved to be insensitive to affect-related aspects of still faces, though more so for angry than for happy expressions. It is suggested that the young infants difficulty in extracting emotional information from static stimuli may be attributable to the absence of the critical invariants (dynamic, multimodally specified) that characterize naturalistic expressions of emotion.

Infant Understanding of the Referential Nature of Looking

To determine whether infants follow the gaze of adults because they understand the referential nature of looking or because they use the adult turn as a predictive cue for the location of interesting events, the gaze-following behavior of 14- and 18-month-olds was examined in the joint visual attention paradigm under varying visual obstruction conditions: (a) when the experimenters line of sight was obstructed by opaque screens (screen condition), (b) when the experimenters view was not obstructed (no-screen condition), and (c) when the opaque screens contained a large transparent window (window condition). It was assumed that infants who simply use adult turns as predictive cues would turn equally in all 3 conditions but infants who comprehend the referential nature of looking would turn maximally when the experimenters vision was not blocked and minimally when her vision was blocked. Eighteen-month-olds responded in accord with the referential position (turning much more in the no-screen and window conditions than in the screen condition). However, 14-month-olds yielded a mixed response pattern (turning less in the screen than the no-screen condition but turning still less in the window condition). The results suggest that, unlike 18-month-olds, 14-month-olds do not understand the intentional nature of looking and are unclear about the requirements for successful looking.

Gaze following at 12 and 14 months: Do the eyes matter?

Two questions were addressed: (1) Is the gaze following of infants under 18 months sensitive to eye status? (2) Do they construe looking as referential behaviour? Corkum and Moore (1995) concluded that, prior to 18 months, gaze following is responsive to the head turn alone (H), because infants followed such turns as frequently as conjoint head and eye turns (H/E). Since their results may have been compromised by an absence of targets and a relatively lengthy response time, we retested a 12- and 14- month group with H/E, H and eyes closed (H/Ecl) cues in the presence of targets and with reduced response time. To examine comprehension of referentiality, two more H/E cues were shown—(1) saying ‘oh wow’ while turning and (2) actively scanning the targets—each intended to increase gaze following if infants regard looking as seeing something. Fourteen-month-olds, but not 12-month-olds, responded significantly more to the standard H/E cue than to the H and H/Ecl cues, indicating the importance of eyes for gaze following at this age. Neither age group, however, responded more to the two ‘enhanced’ H/E cues than to the standard. In a second experiment, a new 14-month group was tested without targets, and again, responding was significantly greater to H/E than to H and H/Ecl. It was concluded that by 14 months, the eyes are co-equal with the head in controlling gaze following, but whether such head/eye turns are understood as object-directed is problematic.

Infant Sensitivity to Deviations in Dynamic Facial--Vocal Displays: The Role of Eye Regard.

Do young infants appreciate the intentionality of adult interactors? In view of recent speculation that infants are innately sensitive to eye direction and that communicative intent is conveyed in part by attentional cues, the reactions of 3- and 5-month infants were compared to video episodes of normally responsive women who either appeared or did not appear to make eye contact. Across three experiments, lack of eye contact was achieved by either averting the eyes (E), averting the head and eyes (H&E), closing the eyes (ECL), or averting the head alone (H). Three-month-olds smiled less at H&E, H, and ECL, but not at E, relative to frontal faces, indicating sensitivity to head but not to eye orientation. By contrast, 5-month-olds smiled less at H&E, E, and ECL, but not at H, indicating sensitivity to both head and eye orientation. The implications of the data for mentalist views of infant social behavior are discussed.

Effects of ecologically relevant manipulations on infant discrimination learning

In our laboratory, persistent attempts to establish an operant discrimination in 4-month-old infants met with limited success, even though the procedures involved readily conditionable head-movement responses, highly discriminable audiovisual stimuli, and potent sensory reinforcers. On the assumption that discriminative learning might be facilitated by experimental arrangements which were less arbitrary and better aligned with the infants natural endowment, modifications designed to take advantage of the young babys presumed sensitivity to the locus of incoming stimulation were introduced. Thus, (1) a more naturalistic response-consequence relationship was achieved by placing reinforcement at the end of the heat-rotation are rather than at midline and by emphasizing the orientational rather than the manipulative character of the response, and (2) the discriminanda were presented in separate locations in the vertical plane rather than in a single midline position. As a consequence of these topographical manipulations, rapid and accurate discrimination learning was obtained. Almost all subjects ( N = 32) exhibited significant levels of correct responding on random test trials, with half the sample scoring at least 80% correct in the second of two discrimination sessions. Moreover, many infants exhibited long runs of consecutively correct responses. It remains to be determined whether the results are a function of change in the form of the response, the enhanced salience of the stimuli, or the natural predisposition of infants to associate spatially differentiated cues with spatially differentiated responses. Since, under the rearranged conditions, many infants displayed behaviors characteristic of problem-solving, it is suggested that a properly constructed discrimination learning task may be useful in examining the emergence of higher-order cognitive functions.

Do infants see objects or retinal images? Shape constancy revisited

Although earlier research by Bower on the spatial constancies indicated that 2-month-olds perceive a veridical three-dimensional world, recent evidence based on distance-appropriate behavior suggests that infants may be responding to differential depth cues per se, devoid of spatial significance. To shed light on this apparent contradiction, a test of the infants ability to perceive real shape-at-a-slant was conducted using recovery from habituation as an index of discrimination. Groups of 80-day-old babies (total N = 120) were each exposed to one of six treatment conditions. The same figure in the fronto-parallel plane served as the recovery stimulus in all conditions. The to-be-habituated stimulus differed between groups according to whether it deviated from the standard in terms of various combinations of objective shape, projective shape, and slant. While the results were consistent with Bowers conclusion that the momentary retinal projection plays no role in infant perception, they were discrepant from Bowers findings in that subjects appeared to respond exclusively to variation in slant and gave no evidence of perceiving constant shape across changing slant. Whether the ability to perceive real shape was present but masked by salient slant cues or whether infants can detect only the proximal retinal cues specifying slant could not be conclusively determined. A second experiment involving 60-day-olds (N = 20) and an additional 80-day-old group (N = 18) indicated that the findings could not be attributed to sampling error, the age differential between Bowers subjects and our own, or tendencies to view the slanted figures projectively.

Motivation as a determinant of retroactive inhibition

Summary The present study explored the effects of motivational factors upon retention in an A-B,A-C retroactive inhibition (RI) paradigm. High- and Low-incentive instructions were distributed between the original (OL) and interpolated (IL) learnings (with relearning, RL, always under the OL condition) so that four experimental groups were established: High-High-High (HHH), High-Low-High (HLH), Low-Low-Low (LLL), and Low-High-Low (LHL). All S s were brought to the same learning criterion (8/10) on OL and IL lists, which consisted of paired two-syllable adjectives. Two predictions were tested: (1) that instructional change (from OL to IL) would reduce RI, and (2) that High instructions (during OL and IL) would facilitate recall. The major finding was a clearly significant reduction in RI for the HHH as compared with the other three groups. This occurred despite the fact that there were no differences between the four groups in rate of OL and IL, in interlist intrusions, or in number of items recalled during the early phase of IL. A control group (LLH) was no different from the LLL group in either learning or retention and was also significantly poorer than the HHH group in recall. Thus, neither instructional change nor the occurrence of High instructions during OL or IL alone enhanced recall, but rather the occurrence of High instructions during IL combined with OL, RL, or both. It is suggested that High instructions at these points may have elicited learning-to-learn responses which served either to differentiate the two lists or to mediate between them.