Preschoolers’ Speaker-specific Interpretations of Emotional Prosody During Online Language Processing

• In Event: 1-022 - Affective Cognition: Bidirectional Influences of Emotion and Social Cognition

Integrative Statement

In order to gauge communicative intent, listeners draw upon cues beyond the linguistic content itself. One relevant cue is the emotional prosody that often accompanies speech. For example, “Get out!” means something very different if spoken with excited vs. angry emotional prosody. Recent research has demonstrated that preschoolers are remarkably successful at using emotional prosody to determine the communicative intent of a speaker. Here, we examined whether social-cognitive processes give rise to this ability (vs. more associative processes) by examining whether preschoolers’ interpretation of emotional prosody reflects highly flexible reasoning about the basis for a particular speaker’s use of emotional cues.

In Experiment 1, 110 4- and 5-year-olds were introduced to either a conventional speaker who demonstrated congruent use of linguistic and affective cues (e.g., “My dog ran away” said using negative intonation) or an unconventional speaker who demonstrated incongruent use of these cues (e.g., “My dog ran away” said using positive intonation; see Figure 1). Test trials consisted of displays containing of pairs of objects that belonged to the same category, but that differed in terms of their likelihood to be associated with negative or positive emotional prosody (e.g., broken ball/intact ball) and were accompanied by referentially ambiguous instructions (“Look at the ball”) spoken in either a positive- or negative-sounding voice (see Figure 2). In the Conventional Speaker condition, children looked more to the broken object when the utterance was spoken with negative emotional prosody compared to positive emotional prosody (p<.001). This effect emerged as soon as the noun began to unfold. In contrast, there was no effect of emotional prosody in the Unconventional Speaker condition (p=.98).

In Experiments 2 and 3, we further examined the extent to which children will suspend their use of the emotional prosody cue for an unconventional speaker. In Experiment 2, 53 4- and 5-year-olds were introduced to a speaker who made unconventional emotional evaluations of stimuli in a neutral tone of voice (e.g., “That butterfly looks ugly”). During test trials, children’s on-line sensitivity to emotional prosody was delayed, emerging during the post-noun interval. In Experiment 3, 55 4- and 5-year-olds were introduced to a speaker who provided overly informative descriptions (e.g., referring to an image of a banana as a “long yellow fruit you can peel that monkeys like to eat”), thus demonstrating unconventionality unrelated to emotion. Here, similar to the conventional speaker condition (Experiment 1), 5-year-old children directed a greater proportion of looks to the negative object during negative emotional prosody trials, compared to positive emotional prosody trials following the onset of the noun, p<.001. Conversely, 4-year-old children’s looking patterns did not vary as a function of the speaker’s emotional prosody.

These findings demonstrate that relevant types of speaker unconventionality can disrupt preschoolers’ use of emotional prosody, with important developmental differences occurring between the ages of 4 and 5 years with respect to the specificity of children’s speaker evaluations. Taken together, these results provide compelling evidence that social-pragmatic reasoning plays a role in children’s flexible use of emotional cues during language processing.

Authors

• Justine Thacker, University of Calgary
  Presenting Author

• Craig G Chambers, University of Toronto
  Non-Presenting Author

• Susan Graham, University of Calgary
  Non-Presenting Author