The random and the non-random in intra-individual L2 variation

Abstract When intra-individual variation in language use is discussed it tends to be seen as not having much significance. The recognizability of the relation of variants to each other is (usually tacitly) acknowledged; intra-individual variation is often attributed to influence across different varieties of the language in question (and, in L2 variation, also across languages); in addition such variation is frequently seen as a concomitant of the way in which the given linguistic feature develops. None of these commentaries actually has much to do with randomness ; all seem to relate to identifiable complexes of correlation. The present article begins by exploring how the concept of randomness is to be understood and defined. It proceeds to explore the relationship of intra-individual linguistic variation to the recognizability of the connections between variables, to cross-varietal influence and to acquisitional trajectories. It probes the question of whether, in the light of the putatively explanatory factors for intra-individual variation considered, much room remains for its treatment as straightforwardly random.