NTT Basic Res. Labs., 3-9-11 Midori-cho, Musashino, Tokyo, 180 Japan
Word-familiarity databases have usually been developed for visually presented words, not for auditorily presented words. The purpose of this study is to develop a familiarity database for both auditorily and visually presented words, and to investigate the relationship between both representations. Using a 7-point scale, ten Japanese adults rated the familiarity of every word in the Shinmeikai Japanese dictionary (Tokyo: Sansei-do, 1974)---about 62 000 words. The words were auditorily presented to each subject in one session and visually presented in another. The Pearson correlation coefficient between the auditorily and visually presented words was 0.76 when the mean rating of each word was calculated over the ten subjects. The results show that for a large part, familiarity is basically the same whether a word is presented auditorily or visually. However, a significant number of words received quite different familiarity ratings, suggesting that the auditory mental lexicon and the visual mental lexicon are organized differently to some extent. These differences, at least in part, could be attributed to the idiosyncratic properties of Japanese orthography. The experiment will be continued using more subjects to obtain more reliable databases.