Leekam S R, Prior M
Institute of Social and Applied Psychology, University of Kent at Canterbury, U.K.
J Child Psychol Psychiatry. 1994 Jul;35(5):901-15. doi: 10.1111/j.1469-7610.1994.tb02301.x.
Previous studies show that autistic children fail tests of second-order belief attribution. They also fail tests of lying and deception. The present study used Leekam's (1988) joke-lie distinction task to test (a) understanding of second-order mental states (intention and belief) and (b) the ability to judge these acts as lies or jokes. Seventeen normal and 16 autistic children took part. Eight of the autistic children had previously passed a test of first-order false belief. Results showed that six autistic subjects (37.5%), all of whom are false belief "passers", gave consistently correct answers to second-order mental state questions. Neither normal nor autistic children found second-order intention easier than second-order belief. However, normal children found the ability to judge another person's mental state easier than labelling whether the person was lying or joking, supporting previous evidence. In contrast, there was no difference in these two judgements for autistic children. Overall these results qualify previous evidence by showing that autistic children can use second-order reasoning and can distinguish lies from jokes. Observational data on these children, however, suggest that their competence on the comprehension of these hypothetical situations was not matched by an ability to use lying and joking in real life. Methodological, language and diagnostic factors are discussed as providing possible explanations for the results.
先前的研究表明，自闭症儿童在二阶信念归因测试中表现不佳。他们在说谎和欺骗测试中也表现不佳。本研究使用利卡姆（1988年）的笑话-谎言区分任务来测试：（a）对二阶心理状态（意图和信念）的理解，以及（b）判断这些行为是谎言还是笑话的能力。17名正常儿童和16名自闭症儿童参与了研究。其中8名自闭症儿童之前通过了一阶错误信念测试。结果显示，6名自闭症受试者（37.5%），他们都是错误信念测试的“通过者”，对二阶心理状态问题给出了始终如一的正确答案。正常儿童和自闭症儿童都没有发现二阶意图比二阶信念更容易理解。然而，正常儿童发现判断他人心理状态的能力比判断这个人是在说谎还是在开玩笑更容易，这支持了先前的证据。相比之下，自闭症儿童在这两种判断上没有差异。总体而言，这些结果修正了先前的证据，表明自闭症儿童能够进行二阶推理，并且能够区分谎言和笑话。然而，对这些儿童的观察数据表明，他们对这些假设情境的理解能力并没有与在现实生活中说谎和开玩笑的能力相匹配。研究讨论了方法、语言和诊断因素，作为对结果的可能解释。
