The language abilities of children with autism spectrum disorder (ASD) are highly variable. More insight is needed into the mechanisms that underlie co-morbid language impairments (LI) in children with ASD (henceforth, ASD-LI) across complex lexical and/or grammatical phenomena, particularly for languages beyond English. The present study tested the comprehension and production of two-constituent compound words at the single-word level for Greek (e.g., pondikopayiδa 'mouse trap'). The Compound Word Test, measured on a range of psycholinguistic variables, was used to assess the (compound) constitutent recognition, comprehension/definition, and production of compositional noun-noun compounds in four school-aged children diagnosed with ASD-LI. Their results were compared to age-matched peers with typical language development as a group and as single cases. Comprehension was probed in relation to the word's constituents, for which semantic interpretation involved explaining the meaning of the compound. Production of compound words was tested using a picture confrontation naming task. The results revealed that the four children with ASD-LI who participated had less difficulty recognizing the compound constituents but showed a significant deficit in deriving the compound meaning. Naming compounds was exceptionally difficult despite generally intact comprehension of the object pictures. This leads us to suggest a dissociation between linguistic and conceptual knowledge about the constituents and the compound word. Moreover, the children with ASD-LI produced semantically infelicitous responses when explaining the meaning of compounds, an error not evident in their typically developing peers. Generalizing over the four single clinical cases, we hypothesize that children with ASD-LI have difficulties at the interface of (morpho)syntax with semantics and pragmatics, that is, at the conceptual-intentional system.

In this paper, reconstruction for Binding Principles A and C will be (re)considered in wh-slifting, a construction which appears to associate a wh-interrogative clause with a yes/no-interrogative clause, whose predicate typically selects propositions rather than questions. While the current view is that both binding principles bleed in wh-slifting, a thorough examination of Principle C and experimental pilot findings for Principle A reported here suggest the exact opposite conclusion: Binding Principles A and C do not bleed in wh-slifting. To the extent that this conclusion is valid, it favors the hypothesis that a wh-interrogative clause reconstructs to the complement position of a proposition-selecting predicate. This, in turn, raises non-trivial questions about the syntax and semantics of clausal complement selection, which we leave unanswered.