Language Change in Progress: Evidence from Computer- Mediated Communication

0. Introduction Computer-Mediated Communication (CMC) has become an increasingly well investigated research area (cf. Herring 2001), and has led to the uncovering of interesting sociolinguistic findings involving variation (e.g., Herring 2003, 2004). For example, in their investigation of the relationship among language, gender, and genre in weblogs, Herring and Paolillo (2006) observe that the style difference of female vs. male features varies depending on the sub-genre of diary and filter (cf. Herring et al. 2004), regardless of the gender of the author. They find that the diary entries exhibit more female features while the filter entries more male features. The results shown in their study cast a fresh look at a genderindexing distinct from the one that is generally characterized in spoken and written language. Research in CMC is not limited to English-based on-line communication. Nishimura (2003a, b) examines the nature of Japanese websites. She reports that an unconventional way of written scripts gives rise to, for example, puns based on Kanji (=Chinese) characters and playful coinage of new words resulting from replacement of two similar-looking characters. These phenomena together create a unique communication environment that is characteristic of the CMC mode. In this paper I will present a preliminary observation from CMC in Japanese that offers an intriguing data source that can be interpreted to suggest a possible language change in progress. The type of data to be discussed below is not conventional by nature in that they have traditionally been considered “ungrammatical” or speakers’ “errors” in more conventional modes of communication (Miyaji 1956). However, such ungrammatical or erroneous patterns regarding certain constructions that I shall discuss have turned out to be amply instantiated on the Japanese Internet. The large number of such patterns