The Project

The project aims to give a detailed account of verb semantics and argument realization in pre-modern Japanese. Argument realization is a fundamentally important aspect of the syntax of a language as it concerns the way in which verb meaning determines the number of arguments and their morpho-syntactic and semantic properties. The project has a synchronic and a diachronic part, each with theoretical, descriptive, and practical implications of relevance to Japanese studies generally, Japanese linguistics, and historical and general linguistic theory. For a short description of the project in Japanese, click here (日本語).

Synchronic part
The synchronic part of the project investigates verb semantics and argument realization in the following three stages in the history of Japanese; these stages are those in which the written source material best and most fully reflects the contemporary language and which are the best suited for syntactic studies of pre-modern Japanese.

Old Japanese (mainly 700 - 800)
Early Middle Japanese (800-1200, but especially 900 - 1100)
Late Middle Japanese  (1200 - 1600)

In each of these periods the range of argument realization for each individual verb and semantically based verb class will be described, and relevant parameters of verb semantics investigated, addressing questions such as:

  • Which lexical semantic features are relevant in the classification of verbs (for example, in no particular order, semantic roles, argument structure, lexical aspect, telicity, type of motion, manner/result, causation, unaccusativity/unergativity)?
  • What argument realization patterns emerge?
  • Which verb classes share common argument realization patterns?
  • If there are commonalities in argument realization across verb classes, what are they?
  • And if there are differences between members of verb classes which appear to be similar, what are they?
  • Which other grammatical phenomena than argument realization are explained by verb semantics (for example, aspect, auxiliary selection, ellipsis, or specification of adjuncts or non-subcategorized arguments)?

The immediate result of this will be a detailed synchronic description of verb semantics and argument realization in Old, Early Middle, and Early Modern Japanese. Apart from the intrinsic descriptive value, this will have both important practical implications for the interpretation of pre-modern Japanese texts and serve as a platform for further research on pre-modern Japanese syntax.

From these essentially descriptive points a number of general questions arise, which have significant theoretical interest: (a) In historical linguistics – which may be viewed as an extreme kind of corpus linguistics, with a closed corpus (‘dead language’) and no recourse to native speaker intuition or any kind of directed testing or informant work – the application of any version of current syntactic theory raises theoretical and practical methodological issues, which this project will be the first to address on a large practical scale. (b) In lexical semantic representation: Which part of verb meaning determines argument realization and other grammatically relevant phenomena? Although the framework adopted to study these phenomena is general (see §2), it has mainly been developed in the context of English and has not been applied in detail to a typologically very different language such as Japanese which, for example, has extensive argument ellipsis (pro-drop), fairly free word order, and frequently drops case markers (a feature reflected far more widely in pre-modern than in written modern Japanese). Consequently, its application to Japanese might be expected to yield new insights.

Diachronic part
The diachronic part of the project will provide an account of the development through time of verb semantics and argument realization, based on the results of the synchronic investigation. Comparison of the lexical semantics of individual verbs and of verb classes, and the accompanying argument realization patterns through the three stages of Japanese under investigation, will give: (a) an inventory of changes through the history of Japanese (i) in verb semantics for both individual verbs and for verb classes, and (ii) in argument realization; and (b) a first description of development pathways for verb meanings and argument realization. The diachronic part of the project advances our knowledge of pre-modern Japanese and its development through time. More generally, it will bring to light possible development pathways for verb meanings and argument realization (incorporating insights from grammaticalization theory (Traugott & Dasher, 2005) about universals or tendencies in the direction of semantic change) and will also contribute hypotheses about how and why such changes take place (informed by Henning Andersen’s widely acknowledged, but rarely applied, theory of language change (e.g. 1989), which is distinguished by being user-focused and realistic). This type of study has never before been undertaken for Japanese or any other language.

Research Context
While modern Japanese is among the most intensively investigated languages from the standpoint of theoretical morpho-syntax, the traditional focus of studies on pre-modern Japanese language has been phonology and morphology. Recently, interest in the syntax of earlier Japanese from the point of view of current syntactic theory has emerged, with studies dealing with various specific aspects of the syntax of Old and Middle Japanese, such as complementation, negation, focus constructions, case marking  and word order (e.g. Kuroda, forthcoming; Miyagawa & Ekida, 2003; Watanabe, 2002; Whitman, 1997; Wrona, 2005; Yanagida, 2006). However, no comprehensive study of the very basic features of earlier Japanese syntax, such as that proposed here, has yet been undertaken, and relatively little attention has been paid to the historical development of Japanese morpho-syntax, and almost none to changes in verb semantics and argument realization. This may overall be linked to the complexity and quantity of the materials which must be studied: Few people have had the necessary philological and theoretical syntactic background, and in any case the task is too big for any individual. By contrast the current research group combines – partly individually, but particularly collectively – the skills and specializations in philology, Japanese and general historical linguistics, and syntactic theory which are necessary to ensure the quality of both the empirical and theoretical work. The group will also be large enough to conduct the proposed project both in the required detail and breadth of coverage.

In generative syntax (e.g. Government and Binding, Minimalism) arguments and many of their properties are determined by a verb’s lexical information. An alternative to this “projectionist” approach is a “constructional” approach in which argument realization in conjunction with verb semantics contributes in non-trivial ways to sentence meaning (e.g. Lexical Functional Grammar; see Dalrymple, 2001). One major reason for choosing this alternative is that detailed studies of English have revealed that the range of argument realization for individual verbs is much broader than usually recognized in purely syntactic studies. In particular, the work of Levin & Rappaport Hovav (2005) has shown that an aspectual approach to event structure has powerful explanatory value.

While much research has been done on aspect and related phenomena in modern Japanese, with a strong native tradition (e.g. Kindaichi), previous studies (including those of earlier language stages) have simply assumed the correctness of the aspectual classes proposed by Kindaichi or, in the western tradition, Vendler. Rather than assuming Kindaichi’s or Vendler’s aspectual classes a priori, the present project investigates aspect and aspectual classes in pre-modern Japanese on their own merit.

More generally, most research into historical Japanese grammar, including syntax, implicitly projects the categories of the modern language onto earlier stages. This project will provide an immanent, comprehensive account of each of the earlier stages and will also significantly impact theories of how Japanese has changed and how the stages are related.

Research Methods
Texts from the 8th to the 16th century form the empirical backbone of the project. For the Old Japanese period, the entire text corpus will be investigated. Of special note is the inclusion – in addition to the poetry of the Man’yôshû, Kojiki, Nihon shoki, Fudoki, Shoku Nihongi, and the Bussokuseki-uta – of the prose texts from the period (Senmyô and Norito) which are usually disregarded in studies of OJ syntax despite their unique and crucial contribution. Much preliminary work on the OJ texts has already been done by Frellesvig and Wrona.

For Early Middle Japanese, especially in the period 900-1100, a large corpus of prose texts is available, reflecting the vernacular of the court nobility and officials, allowing a much clearer picture of the language than for any other time in pre-modern Japanese. The following texts will be used, many in full, but some in excerpts: Kokinwakashû kanajo (c. 920), Ise monogatari (900-), Tosa nikki (935), Kagerô nikki (950-), Makura no sôshi (c. 1000), Genji monogatari (1001-10), Tsutsumi chûnagon monogatari (c. 1055), Sarashina nikki (1059-60), Konjaku monogatari (c. 1100), Ôkagami (c. 1119).

The end of the Late Middle Japanese period is well attested in materials produced by missionaries, in the contemporary vernacular and presented in alphabet writing. The project will first of all use two sizeable texts: the Feiqe Monogatari (a vernacular retelling of a popular classic originally from the 13th century) and the Esopono Fabvlas (a translation of Aesop’s Fables), incorporating also the grammars and dictionaries produced by the missionaries.

Finally, the project will incorporate, as appropriate, additional materials supplementary to those described above, both within and between the studied stages of Japanese, for example the so-called kunten-materials (scholastic annotations).

The intial part of the project consists in building an electronic database of the se texts which will be transcribed and tagged for syntactic function.

The electronic texts and databases will organize the empirical material for the close synchronic syntactic and semantic analysis of the attested verbs and their argument realization patterns, and for the diachronic establishment of correspondences between verbs and verb classes through time and the identification of development pathways, as set out in §§1.1-1.2.

Time frame
The project runs for 4.5 years and is divided into 4 main stages:




Stage 1


[Jan.2009 - March 2010]

establishment of common philological, methodological, and theoretical ground, including literature review

data collection; transcription; text tagging

completion of electronic text corpus

completion of text tagging

Stage 2



[April 2010 - June 2011]

extraction of verbs and their argument patterns

synchronic syntactic and semantic analysis

completion of electronic databases

Stage 3


[July 2011 - June 2012]

synchronic syntactic and semantic analysis

preliminary completion of synchronic analysis

Stage 4


[July 2012 - Sept. 2013]

diachronic analysis:
correspondences between verbs and verb classes through time; development pathways

feedback of the diachronic analysis into the synchronic analysis

full completion of synchronic and diachronic analysis

Andersen, Henning (1989). Understanding linguistic innovations. In Breivik & Jahr. (Eds.), Language Change: Contributions to the Study of its Causes. Mouton. 5-28.
Dalrymple, Mary (2001). Lexical Functional Grammar. Academic Press.
Kuroda, S.-Y. (forthcoming). On the Syntax of Old Japanese. In Frellesvig, Shibatani & Smith. (Eds.), Current Issues in the History and Structure of Japanese. Kurosio.
Levin, Beth & Rappaport Hovav, Malka (2005). Argument Realization. CUP
Miyagawa, Shigeru & Ekida, Fusae (2003). Historical Development of the Accusative Case Marking in Japanese as Seen in Classical Literary Texts. Journal of Japanese Linguistics 19: 1-97.
Traugott, Elizabeth Closs & Dasher, Richard (2005). Regularity in Semantic Change. CUP.
Watanabe, Akira (2002). The loss of overt Wh-movement in Old Japanese. In Lightfoot, D. (Ed.), Syntactic effects of morphological change. OUP. 179-195.
Whitman, John. (1997). Kakarimusubi from a comparative perspective. Japanese/Korean Linguistics. Vol. 6. CSLI. 161-178.
Wrona, Janick (2005). Specificational pseudo-clefts in Old Japanese. Folia Linguistica Historica XXVI(1/2): 139-157.
Yanagida, Yuko (2006). Word Order and Clause Structure in Early Old Japanese. Journal of East Asian Linguistics 15(1): 37-67.