Is arbitrariness a design-feature of the sign?

Throughout the history of linguistics, there has been a tendency to believe that there is no relationship between the form of linguistic signs and their meanings; in other words, that their relationship is arbitrary. This view was specifically popularised by de Saussure (1916), although it has been sustained in the philosophy of semiotics since Aristotle, and was supposedly cemented by Hockett (1960) in his Origins of Speech, calling ‘arbitrariness’ the eighth of thirteen ‘design-features’ of language. While recent trends in the study of nonarbitrariness have set out to overthrow the Saussurean precedent, this paper aims suggest that, in fact, arbitrariness is not a design-feature of the sign. Starting from the beliefs of Cratylus in Plato’s Cratylus, a philosophical, nonarbitrary language (called Nonarbitrer) is constructed to test whether there could be a language that had the ‘design-feature’ of nonarbitrariness, or iconicity, such that the language might function just as any other natural one. By measuring the accuracy and time taken for 33 participants in a memory recall task for lexical items in three ‘Levels’ of Nonarbitrer, alternating in Level by degree of iconicity (compared with analogous data from Mandarin and Basque, selected for their genetic and orthographic resemblances with, or not with, English) it is found that Nonarbitrer could be considered functional, and that, therefore, Hockett (1960)’s position should be reconsidered.