Pastor-Cerezuela Gemma, Tordera Yllescas Juan C, González-Sala Francisco, Montagut-Asunción Maite, Fernández-Andrés María-Inmaculada
Basic Psychology, Facultad de Psicología, Universitat de València, Valencia, Spain.
Teaching of Language and Literature, Facultad de Magisterio, Universitat de València, Valencia, Spain.
Front Psychol. 2018 Mar 13;9:272. doi: 10.3389/fpsyg.2018.00272. eCollection 2018.
This study evaluates the comprehension of generalized conversational implicatures (GCI) in children with and without autism spectrum disorder (ASD), using a GCI test constructed based on the Levinson model, which distinguishes between three types of implicatures: type Q (or scalar: "what is not referred to does not occur"); type I ("by default, it is not necessary to say what can be assumed"); and type M ("if someone is expressing something in a not very simple or marked way, it is because s/he is describing a situation that is not very typical, frequent, or prototypical"). In addition to the ASD group ( = 22), two comparison groups were utilized: a group matched on chronological age with the ASD group, but with a higher linguistic age (TCD group, = 22), and a group matched on linguistic age with the ASD group, but with a lower chronological age (TLD group, = 22). In all cases, linguistic age was assessed with the Peabody test. The performance of the three groups on the GCI test was compared (overall and on each type of implicature), and performance on the three types of implicature was compared within each group. The ASD group obtained worse performance than the other two groups, both overall and for each implicature type, without also obtaining differences in performance on the three implicature types. The TCD group obtained better performance than the TLD group on overall performance, but not on each implicature type, and both groups obtained lower performance on the type M heuristics than on the type I. Based on these results, the children with ASD in our study presented limitations in the comprehension of the three types of GCI, but it was not possible to obtain evidence for an inferential continuum of the three types of GCI. However, in the two typical development groups, this evidence was obtained, leading us to propose an inferential continuum model based on the different levels of dependence on the context of each of the three types of implicatures, with type M implicatures being more contextually dependent.
本研究使用基于莱文森模型构建的一般会话含义(GCI)测试,评估患有和未患自闭症谱系障碍(ASD)的儿童对GCI的理解。该测试区分了三种类型的含义:Q型(或量级型:“未提及的情况不存在”);I型(“默认情况下,无需说出可被假设的内容”);以及M型(“如果某人以不太简单或有标记的方式表达某事,那是因为他/她正在描述一种不太典型、不常见或非原型的情况”)。除了ASD组(n = 22),还使用了两个对照组:一个与ASD组实足年龄匹配,但语言年龄较高的组(TCD组,n = 22),以及一个与ASD组语言年龄匹配,但实足年龄较低的组(TLD组,n = 22)。在所有情况下,语言年龄均通过皮博迪测试进行评估。比较了三组在GCI测试中的表现(总体表现以及每种含义类型的表现),并比较了每组内三种含义类型的表现。ASD组在总体表现以及每种含义类型上的表现均比其他两组差,且在三种含义类型的表现上也没有差异。TCD组在总体表现上比TLD组好,但在每种含义类型上并非如此,并且两组在M型启发式上的表现均低于I型。基于这些结果,我们研究中的ASD儿童在理解三种类型的GCI方面存在局限性,但无法获得关于三种类型GCI推理连续性的证据。然而,在两个典型发育组中,获得了这一证据,这使我们提出了一个基于三种含义类型中每种类型对语境依赖程度不同的推理连续体模型,其中M型含义对语境的依赖性更强。