Centre for Brain and Cognitive Development, Department of Psychological Science, Birkbeck, University of London, London, United Kingdom.
PLoS One. 2012;7(5):e36428. doi: 10.1371/journal.pone.0036428. Epub 2012 May 15.
The language difficulties often seen in individuals with autism might stem from an inability to integrate audiovisual information, a skill important for language development. We investigated whether 9-month-old siblings of older children with autism, who are at an increased risk of developing autism, are able to integrate audiovisual speech cues. We used an eye-tracker to record where infants looked when shown a screen displaying two faces of the same model, where one face is articulating/ba/and the other/ga/, with one face congruent with the syllable sound being presented simultaneously, the other face incongruent. This method was successful in showing that infants at low risk can integrate audiovisual speech: they looked for the same amount of time at the mouths in both the fusible visual/ga/- audio/ba/and the congruent visual/ba/- audio/ba/displays, indicating that the auditory and visual streams fuse into a McGurk-type of syllabic percept in the incongruent condition. It also showed that low-risk infants could perceive a mismatch between auditory and visual cues: they looked longer at the mouth in the mismatched, non-fusible visual/ba/- audio/ga/display compared with the congruent visual/ga/- audio/ga/display, demonstrating that they perceive an uncommon, and therefore interesting, speech-like percept when looking at the incongruent mouth (repeated ANOVA: displays x fusion/mismatch conditions interaction: F(1,16) = 17.153, p = 0.001). The looking behaviour of high-risk infants did not differ according to the type of display, suggesting difficulties in matching auditory and visual information (repeated ANOVA, displays x conditions interaction: F(1,25) = 0.09, p = 0.767), in contrast to low-risk infants (repeated ANOVA: displays x conditions x low/high-risk groups interaction: F(1,41) = 4.466, p = 0.041). In some cases this reduced ability might lead to the poor communication skills characteristic of autism.
自闭症个体中常见的语言障碍可能源于他们无法整合视听信息,而这种信息整合能力对语言发展至关重要。我们研究了自闭症大龄儿童的 9 个月大的兄弟姐妹是否能够整合视听言语线索,因为他们有更高的自闭症发病风险。我们使用眼动追踪仪来记录婴儿在观看屏幕时的目光位置,屏幕上显示的是同一个模特的两张脸,一张脸在发/ba/音,另一张脸发/ga/音,其中一张脸与呈现的音节声音一致,另一张脸不一致。该方法成功地表明,低风险婴儿能够整合视听言语信息:他们在融合的视觉/ga/-听觉/ba/和一致的视觉/ba/-听觉/ba/显示中,观看两张脸的时间一样长,这表明在不一致的条件下,听觉和视觉流融合成一种麦格克型音节感知。它还表明,低风险婴儿能够感知到听觉和视觉线索之间的不匹配:他们在不匹配、不可融合的视觉/ba/-听觉/ga/显示中,观看嘴巴的时间比在一致的视觉/ga/-听觉/ga/显示中更长,这表明他们在观看不一致的嘴巴时感知到一种不常见的、因此有趣的类似言语的感知(重复方差分析:显示 x 融合/不匹配条件交互:F(1,16) = 17.153,p = 0.001)。高风险婴儿的注视行为并没有根据显示类型而有所不同,这表明他们在匹配听觉和视觉信息方面存在困难(重复方差分析,显示 x 条件交互:F(1,25) = 0.09,p = 0.767),这与低风险婴儿形成对比(重复方差分析:显示 x 条件 x 低/高风险组交互:F(1,41) = 4.466,p = 0.041)。在某些情况下,这种能力的下降可能导致自闭症患者沟通能力较差。