Permalink : http://hdl.handle.net/2115/38300

On the Map 26 in Haspelmath, M. et al. eds. (2005), Ainu is valued as an "approximately equal amounts of suffixing and prefixing" language, which is an uncommon trait along with Ket in the northern Eurasia, where "strongly suffixing" languages are predominant. The recalculation of the "affix indexes", however, shows that Ainu has more preference for prefixing to be counted between "predominantly prefixing" and "moderate preference for prefixing", than the one M.Dryer, the author of the map 26, estimated. One of the reasons of this disagreement is probably that Dryer considers Ainu having "plural affixes on nouns" (affix type iv), but, in fact, Ainu can't be regarded as having such affixes. This trait of prefix predominance can be seen more in the southern area of Eurasia than in the northern area, where Ainu is spoken. Moreover while Ainu has OV and NP + Po order, Hawkins and Gilligan (1988) said that statistically OV order is rarely co-occurs with prefix predominance. I propose here a diachronic explanation for this discrepancy. Once Japanese had not been spoken in Japanese archipelago and Ainu had been situated side by side with Austronesian languages, which were VO order and prefix predominance. Ainu had formed a linguistic area with these languages and had the same traits of them. Then Japanese intruded into the islands, between Ainu and the southern languages, so that Ainu was pushed toward north and surrounded by Japanese, Nivkh or Tungusic languages, which were OV order. Then Ainu formed new linguistic area with them and changed its word order into OV, while unchanged on its prefix preference.