Development of infants' communication and expression of emotion through vocalization.

通过对10名2、6、9、12和17个月大的婴儿的语音样本进行感知评级实验，研究了婴儿通过发声表达情绪的能力。听力被试为听力正常的四组，即学生、母亲、照顾者和儿童。得到以下结果：(1)不同听者群体的语境依赖存在差异。学生的分数会因情境而有所偏差，但评分量表之间的相互关系没有显著变化。另一方面，母亲们根据不同的环境改变了评分量表之间的相互关系。母亲们似乎改变了感知权重，在此基础上，她们会听到某一特定方面的情绪，而不是简单地在特定背景下对评级得分产生偏见。对情绪的感知似乎受到听者育儿经历差异的影响。(2)成人和儿童在6个月时都能从语音样本中感知到丰富的情感内容。通过因素分析，从儿童给出的评分分数中提取的因素比成人给出的因素多。儿童对婴儿发声的情绪感知比成人更为敏感。儿童似乎比成人从婴儿的声音中感知到更丰富的情感。(3)通过对200个2月龄婴儿语音样本进行因子分析，提取出高兴与害怕、平静与惊讶、说话与唱歌三个情绪对比因子。声学分析表明，这些感知因素与F0范围、F0变化类型、F0最小值、F0最大值、发声长度和分段数等声学参数显著相关。这些结果表明，即使是2个月大的婴儿也能产生表达情感对比所需的声音元素，而这些情感对比对于成年听众来说是可识别的。

Infants' ability to express emotin through vocalizations is studied through perceptual rating experiments for voice samples recorded from ten infants at 2,6,9,12 and 17 months of age. Listening subjects were four groups with normal hearing, i.e., students, mothers, care takers, and children. Following results were obtained.(1) The context dependency was different among the listener groups. The students biased their scores depending on the contexts without significant changes in inter-relationships between the rating scales. On the other hand, the mothers changed inter-relationships between the rating scales depending on the contexts. The mothers seemed to change perceptual weights, based on which they would hear a specific aspect of emotion, rather than simply biasing the rating scores under a certain context. Perception of emotion seemed being influenced by differences in listeners' child-rearing experiences. (2) Both the adult and child listeners perceived rich contents of emotions from the voice samples recorded even at 6 month of age. More factors were extracted from the rating scores given by the children than by the adults via factor analyzes. Children are more sensitive in perception of emotions from infants' vocalization than the adults. Children seem to perceive richer emotions from infants'vocalizations than the adults. (3) By a factor analysis for 200 voice samples recorded at 2 months of age, three factors representing emotional contrast of pleased vs. frightened, calm vs. surprising, and speaking vs. singing were extracted. Acoustical analyzes showed that these perceptual factors significantly correlate with the acoustical parameters such as F0 range, F0 variation type, F0 minimum, F0 maximum, vocalization length, and the number of segments.These results suggest that infants even at 2 months of age can produce vocal elements necessary to express emotional contrasts that are identifiable for adult listeners.