Published online by Cambridge University Press: 17 February 2009
Elsewhere we have argued on the basis of cross linguistic studies of directionality effects on anaphora in child language, that there is no universal ‘forward directionality preference (FDP)’; rather such a preference is linked to languages with specific grammatical properties. Although such a preference has been attested in English acquisition, matched experimental designs in Japanese, Chinese and Sinhalese, for example, do not show this effect. In this paper we argue that current attempts to show that forward directionality effects can also be induced in Japanese acquisition do not succeed in supporting the FDP. Specifics of the design of stimulus sentences in these experiments vary previous cross-linguistic designs so as to favour forward directionality on independent grounds, and confound cross linguistic comparisons. They in fact support a universal structure dependence in children's hypotheses about directionality of anaphora’.
This paper is prepared with the partial support of NSF grant No. 8318983. We thank James Gair, Akio Kamio, Susumu Kuno, Kazuyo Otani, Lou Mangione, Tatsuko Wakayama, Shin Oshima, Wendy Snyder, Yu-Chin Chien and Suzanne Flynn for discussion of the issues involved in this paper. We also thank an anonymous reviewer for provocative and clarifying questions regarding our arguments.