| (in English) |
Language makes use of a number of coordinators to connect clauses, and a subset of these are ?logical? in nature, which is to say that they are analogous to the truth-conditional operators of logic, even if they can behave differently in actual linguistic communication (due to pragmatic factors, which can be neutralised by the grammar). Remarkably, only two/three of the possible 16 binary connectives of logic have been lexicalised in the world?s languages: and, (inclusive) or, and, in some accounts, the joint denial nor (Horn, 2012). However, connectives such as material implication or alternative denial seem easily entertainable in a mental logic-like system, and this raises the question of whether the language faculty allows any of the unlexicalised connectives at the interface between narrow syntax and the thought systems. In this talk, I provide a panoramic view of the type of research that has been conducted on language?s logical connectives and argue that there is a theoretical split between the ?lexicalisation? of connectives and the ?learnability? of invented words corresponding to unlexicalised logical connectives. In support of this position, I describe the ?visual world? data we have obtained at our laboratory, which show that experimental participants: a) interpret the connective or inclusively, and b) are capable of learning a made-up word for the alternative denial connective, also known as the nand operator (or Sheffer?s stroke). The last result indicates that the language faculty allows for more conceptual distinctions than are lexicalised in the world?s languages, and entirely independently of the input available during the acquisition of a native language. A potential reason for this is that the language faculty is probably more closely connected to the thought systems than usually appreciated, and thus (some) linguistic diversity ? lexicalised or not ? may be the result of our ability to entertain conceptual distinctions. |