Second language gender system affects first language gender classification

Abstract

When you speak about the sea, in some languages you have to refer to it in the masculine gender (il mare in Italian), in others in feminine (la mer in French), and in still others in neuter (Mopemo in Bulgarian). Examples such as these are usually provided as arguments for the arbitrariness principle going back to Ferdinand de Saussure's (1916) postulate on the arbitrary nature of the relationship between the two aspects of the verbal sign - concepts and acoustic images. Grammatical gender is widely assumed to be a prime example of arbitrariness due to the fact that there is ample cross-linguistic variability in the genders of nouns referring to one and the same concept and little systematicity in the mapping between different gender categories (e.g., masculine, feminine, and neuter) and the meaning of nouns within languages. In the face of so much arbitrariness, a curious observer may ask whether the relationship between language(s) and thought may also be arbitrary or, alternatively, subject to certain constraints. While cognitive constraints on language are rarely in the focus of attention (however, see the literature on working memory and language processing, e.g., Caplan &Waters 1999; MacDonald & Christiansen 2002), there has been plenty of speculation and a growing amount of experimental data collected to address the possibility that cross-linguistic variation is associated with variation in mental processes. The latter view is usually subsumed under the Sapir-Whorf hypothesis (Whorf 1956) whose strong version of linguistic determinism has long been discredited but whose softer formulation in terms of linguistic relativity has undergone a revival in recent years. The two issues are indeed intertwined, especially if the general conceptual or mental apparatus of humans is held to be cross-linguistically and cross-culturally universal. The arbitrariness principle would then predict dissociation between the variability of linguistic competence and behavior and the universality of mental processes and representations. In this sense, it would preclude the possibility for a "Whorfian" influence of the language we speak on the way we view extra-linguistic reality. © 2007 Springer.