To understand how Saraba became a farewell salutation, it is necessary to elaborate on the process of conjunctivization, which signifies the preliminary stage of the transformation of a term to a farewell salutation. Conjunctivization represents the modification whereby Saraba lost the forward hypothetical condition and acquired the function of scene development. This study traces this transformation by focusing on the subsequent expressions and antecedent situations of Saraba and comparing it with Sareba, which expresses a forward definite condition, as well as the mizen form (imperfective form) + ba, which denotes a forward hypothetical condition. The results reveal that, first, Saraba often co-occurs with expressions of will and command, unlike Sareba and the mizen form + ba; since the Kamakura period, 70–90% of the use of Saraba has occurred with such expressions. Second, Saraba is frequently subjected to antecedent external situations, and this practice exceeded 80% of Saraba during the Muromachi period. These findings suggest that Saraba was conjunctivized and acquired the function of scene development during the Muromachi period. Comparisons with Sareba and the mizen form + ba reveal that the previously mentioned characteristics of Saraba are unique to this sa-system conjunctive expression of a forward hypothetical condition.

The Wagokuhen (倭玉篇) is a general term for kanji dictionaries, compiled and established in the Muromachi period. A lot of printed editions of the Wagokuhen are still existent but only a very few of its manuscripts survive. In the study of the Wagokuhen the emphasis has traditionally been placed on the compositional aspects of the text, such as the arrangement of sections, while the study of wakun (和訓), i.e. Japanese equivalent words given to individual kanji, has been neglected. However, it is by investigating the sources of the individual wakun that we can gain a better understanding of the system of various texts in the Wagokuhen and the process of their formation.

From this perspective, the paper investigates the wakun in five representative texts of Wagokuhen, namely Kōjininen-bon (弘治二年本), Gokuhen-yōryakushū (玉篇要略集), Gokuhenryaku (玉篇略), Shūhenmokushū (拾篇目集), and Yonezawabunko-bon (米沢文庫本), and shows that the wakun in the Kōjininen-bon, unlike that in the other books, was greatly influenced by the wakun of Monzen (文選) and Irohajiruishō (色葉字類抄). The Monzen was extremely respected as an anthology of the Chinese classics in Japan. By examining individual cases, moreover, we prove that there is no possibility that the wakun of Monzen derived from elsewhere.

In Japanese, the adjectival predicate suki da ([I/Someone] likes X) is typically used with the object case particle ga (noun ga suki da), but under certain conditions, it can be used with the particle o (noun o suki da). Prior research has shown that the acceptability of o is based on factors such as avoiding particle duplication (i.e., o is acceptable when ga has already been used in same clause), degree of transitivity, clause type, and modality. This paper considers other factors that may be at play in the choice between ga and o preceding suki da by focusing on ga and o in relation to the information structure (old vs. new). Data were obtained from The Balanced Corpus of Contemporary Written Japanese and analyzed using the referential distance (RD) method. The results show that the particle o is more likely to be used in contexts with a high degree of givenness of object. Although previous research identified modality as a key factor in the use of o with suki da, this paper argues that its use is tied to the information structure.

This paper aims to illuminate the historical development of the subject markers zero particle (ø), no, and ga in Japanese, focusing on structural changes from the Heian period to the present day. The points discussed are as follows:

(1) Until the end of the medieval period, the use of zero particles (ø) tended to be associated with subordinate clauses and main clauses, while no was predominantly employed in adnominal clauses and quasi-nominal clauses, based on the size of the structure.

(2) The use of the zero particle (ø) resulted in an unclear structural boundary in the subject-predicate relationship, especially in sentences with a large structure featuring a finite-form predicate. In contrast, no and ga created clear structural boundaries in sentences with smaller structures and attributive-form predicates.

(3) The expansion of ga at the expense of zero particles can be attributed to its ability to mark a clear structural boundary between the subject and predicate, particularly in sentences with short structures and attributive-form predicates.

(4) The decline of the quasi-nominal clauses conveying the meaning of people or things is considered a factor contributing to the maintenance of no's role in subject marking in nominal clauses.