Debunking the Trochaic Bias Myth: Evidence from Phonological Development

0. Introduction Since the advent of the theory of Generative Grammar (e.g. Chomsky 1957), a large number of researchers have been on a quest to uncover language universals, which are believed by many to be psychologically real, either because they represent basic language storage or processing mechanisms, or because they emerge as artifacts of our perception or production systems. In many different ways, this quest has offered positive, useful results. Typological studies have yielded generalizations that suggest the presence of basic structures or mechanisms underlying the functioning of human languages. Similar generalizations were obtained from studies of language acquisition. However, an undesirable effect of this quest has been, at times, somewhat hasty speculations that purported language universals should manifest themselves in language acquisition. Indeed, it is often the case that claimed universals end up representing cross-linguistic tendencies which themselves do not make any useful predictions in language acquisition. For example, while coronal consonants enjoy a special status in the phonological systems of a large number of the world’s languages (e.g. Paradis & Prunet 1991; Hall 1997), they display various behaviors in first language acquisition data that contradict the view that they should be universally unmarked or special in any way (e.g. Rose 2000). In this paper, we argue that one such purported universal, often referred to as the trochaic bias, should be abandoned from the theory on a number of grounds. First, the cross-linguistic facts supporting this construct are far from conclusive. Second, babbling and early acquisition data do not provide any independent evidence for such a bias in phonological development. Finally, even in contexts where everything is in place to favor the manifestation of a bias towards trochaic footing, no convincing evidence can be found. The paper is organized as follows. In section 1, we provide an overview of the predictions entailed by the trochaic bias as well as a brief survey of the literature on the topic in the context of child language phonology. As we will see, an overwhelming majority of the empirical evidence available fails to independently