This paper surveys the ratio of the part of speech of words that occur before conjunctive particles by using the “Balanced Corpus of Contemporary Written Japanese” (BCCWJ). The results reflect the grouping of the “Minami-model,” that is, the proportion of verbs is as high as 90% before conjunctive particles of Group B, and the proportion of verbs is around 60% before conjunctive particles of Group C.

In addition, Groups B and C show contrastive results. Not only adjectives or nouns but also stative verbs or verbs in the stative or negative form show less proportion before conjunctive particles of Group B. The number of affixes is different too: fewer before Group B, and more before Group C. From this perspective, conjunctive particles in Group C closely resemble sentence-final particles.

Moreover, the proportion of verbs is around 60% before some of the modality markers in the main clause, nearly the same as in Group C, but the proportion is as high as 90% before the modality markers derived from nouns, which shows their approximation to Group B. This is consistent with the proposal that the noun clause belongs to Group B.

A sentence which is mainly composed of indeclinable word(s) such as Bakamono!(Fool!), Hajishirazu!(Bounder!), or Usotsuki!(Liar!), has not been appropriately classified as a sentence in previous studies.

This study finds that these sentences, mainly composed of indeclinable word(s) with negative values, have different forms and functions from other sentences mainly composed of indeclinable word(s) such as exclamatory sentences (Kireina hana!), and addresses (Satō san!). Furthermore, the study categorizes these sentences composed of indeclinable word(s) with negative values into a new type of their own: “‘Letter’hari-bun(Labeling Sentence)”.

‘Letter’hari-bun is a sentence which is mainly composed of noun(s) categorized as indeclinable word(s), and has both element(s) which indicate nature, character and attribute, and element(s) indicating a person or an object.

‘Letter’hari-bun conveys emotional content such as anger, derision, ridicule, disrespect, hate, or contempt towards a person or an object. Therefore, ‘Letter’hari-bun always carries a negative intent. The speakers perform a speech act in which they express their negative appraisal and they assign a Label (letter) in a sentence form (Labeling Sentence, ‘Letter’hari-bun).

Kasu-type verbs are commonly employed in the Satsugu dialect recorded in the set of materials attributed to the Japanese castaway Gonza. In this study, I consider their morphological features and expressive characteristics in the light of research on the history of Japan's ‘central language.’ In doing so, I focus particularly upon such features as each kasu-type verb's relationship to the verb that can be understood as its source of derivation, the meaning of its Russian equivalent, and inconsistencies caused by use alongside other word forms employed in the materials. In terms of morphology, it has become clear that ra-line verbs most frequently take the form of “－ rakasu”; also, four-syllable examples are most common. In terms of their expressive characteristics, we can see that in almost all cases they possess transitivity, that they express completion or superfluity, more than conventional transitive verbs and that they more forcefully convey the position of the actor.

The dialects with two-pattern accentual systems in southwestern Kyushu regions follow the so-called “General accentual rule”, or Shiki-hozon, in which the accent patterns of compounds are determined by those of the first members of the compounds. However, it has recently been reported that in certain areas around Nagasaki and Saga prefectures, or the area around the Amakusa archipelago in Kumamoto prefecture, the distinction of the compound accentual patterns is neutralized under certain conditions. This paper proposes a diachronic hypothesis as to why and how such neutralization took place. The paper further proposes the proto-accentual system of the two-pattern accent systems in southwestern Kyushu.

The aim of this article is to clarify the meaning of a sentence-final article kashi through the examination of Early Middle Japanese materials. First, I focused on the characteristics of phrases preceding kashi. Then, I clarified that preceding phrases of kashi are limited to Unshared Items. Unshared items refer to the things in which facts and the grounds that can be confirmed as facts are not shared with others, such as personal opinions. Then, the meaning of kashi was examined. Kashi presupposes that the statement might differ from facts. In this article this is called Pseudo-Indetermination. In addition, I investigated a different distribution of auxiliaries on the examples of the termination of predicative form and the examples of kashi which are positioned afterward in order to ascertain these hypotheses. There were two findings. First, conjectural auxiliaries tend to be followed by kashi whereas estimated auxiliaries do not. It is explained that this difference is caused by whether the grounds that can be confirmed as facts can be shared or not. Second, ki can be followed by kashi whereas many past and perfect auxiliary verbs can not. These auxiliaries describe facts, therefore, it is difficult to become unshared items. Yet, ki can be followed by kashi because ki is seen only in examples which describe speaker's own memory. These facts confirm this article's hypothesis. It helps us to explain other conjugated form and examples in which particle zo was followed by kashi if we consider the meaning of kashi as pseudo-indetermination. It also becomes easy to explain its relationship with previous studies.