Based on the modern dialects and ancient documents I propose a proto-accent system of the mainland Japanese dialects. For the high-beginning series I hypothesize the falling pattern of the form HHMM-instead of the high-level pattern HHHH-. For the low-beginning series I incorporate the Kyosei tone—that is, words with the rising pitch—into the system. On the whole, the proto-system consists of the two tonal patterns, falling and low-level, and the two accent kernels, rising and lowering, and the system contains more oppositions than that previously proposed.
The morphological changes in 3 sg. mediopassive endings, -a → -ta and -a → -atta, were still operating during attested Hittite history. This fact, together with the nonexistence of -atta in Old Hittite manuscripts and the retention of original a-class status in 3 sg. imperatives of many ta-class mediopassives, shows that 3 sg. mediopassive verbs in -ta do not go back to a very early period. Contrary to the generally accepted view that both *-to and *-o must be reconstructed as 3 sg. mediopassive endings in the parent language, *-to cannot have been created when the Anatolian branch split off from the rest of the Indo-European family. The remodeled ending *-to was undoubtedly due to the influence of the corresponding active 3 sg. *-ti (primary ending) and *-t (secondary ending). The fact that -ta (< *-to) is overwhelmingly favored by preterite mediopassives in Hittite provides us with decisive evidence that many ta-class mediopassives were created after the affrication which occurred in pre-Hittite and applied to 3 sg active primary (i.e., present) *-ti (> *-tsi), but not to 3 sg. active secondary (i.e., preterite) *-t. The morphological history of Hittite mediopassive verbs clarified in this paper shows that Hittite still preserves an archaism of remarkable antiquity which plays an important role in reconstructing the Proto-Indo-European verbal system.
Many of the languages of Asia have families of expressions in which a verb meaning EAT exhibits Janus-faced behavior. In some cases the subject of EAT bears the semantic role of an agent while in others it bears the role of a theme, patient, or experiencer. A central concern of this paper is to characterize the geographic distribution in Asia of the semantic extensions of the experientially basic verb EAT in selected languages. We graph the areal distribution of representative EAT expressions as sets of data points on WALS maps and show that while some metaphoric extensions of usage of the verb EAT are probably found in every language, other particular types of extensions are found only in a specific area whose languages share a number of typological characteristics as shown by Masica (1976) in his seminal work on the SOV linguistic area of South, Central, and Northeast Asia. While language contact certainly has played and continues to play a significant role, we propose the independent operation of a radial network of semantic extensions as a possible complementary factor in the proliferation and convergence of EAT-expressions.
Case markers, as a nominal morphological feature, indicate the functions of NPs in a clause. The number of cases in individual languages is diverse, ranging from no case-marking in Chinese and Arabic to more than 10 cases in Hungarian and Nez Perce. In The World Atlas of Language Structures (WALS), this morphological feature is highlighted under the heading of “number of cases.” Selecting languages with more than 10 cases, I examine what kinds of cases each of them has, and then consider the reasons why these multiple-case languages do indeed have so many cases. Finally, I discuss whether the languages with rich case systems have other common typological features as described in WALS. The languages with no case marking are densely distributed in Africa and South Asia, and the languages with more than 10 cases are scattered in Eurasia (Basque, Finnish, and Evenki) and Australia (Gooniyandi, Martuthunira, and Kayardild). Cross-linguistic comparison shows that the languages with rich case systems are in fact rich in locative cases but poor in terms of the variety of cases. Finally, as a result of contrasting case with other WALS features, it is established that the languages discussed in this paper have other frequent grammatical tendencies, SOV word order, and postpositions.