On the design and analysis of language acquisition studies Hugo Quen´e Utrecht inst of Linguistics OTS, Utrecht University hugo.quene@let.uu.nl draft 2008.06.03 1 Introduction Recent years have seen a large increase in the number of experimental studies into how children learn to speak a language. As in other sciences, clever ex- periments have been designed and conducted to answer questions that were inspired by linguistic theories. Most linguistic theories, however, focus on the abstract mental representation of language rather than on observed lan- guage behavior. As a consequence, there appears to be a somewhat strained relationship between linguistic theories and empirical observations, either be- havioral or neurological (Van Lancker Sidtis, 2006). In linguistics, empirical data are often used to adstruct theoretical constructs (themselves derived from introspective observations), but seldom to refute them. Even if there is ample evidence contradicting well-established theoretical constructs (e.g. the psychological reality of the phoneme, or the categorical distinction between 1