Organizer and Chair: Wesley M. Jacobsen, Harvard University
Discussant: Wako Tawa, Amherst College
Until recently, grammatical transitivity was treated as being essentially impervious to linguistic analysis, no more than a question of whether a verb by some inherent idiosyncrasy does or does not take a direct object. The complexity of the transitive structure of a language such as Japanese, with its myriad patterns of transitive and intransitive verb pairs, seeming inconsistencies between syntactic marking and verbal form, and apparently language-specific preference for intransitive modes of expression in contrast to English, did much to reinforce that impression. Understandably, therefore, attention was focused in linguistic analyses of Japanese on the more regular features of its grammar, such as case marking, causative and passive syntax, and word order within the sentence. Yet recent shifts in the attention of linguists from syntactic patterning between words to patterns of form and meaning existing within words themselves have given rise to a new appreciation of the fundamental role that the transitive/intransitive dichotomy plays not only in organizing the formal grammar of a language such as Japanese but in organizing the way we experience the phenomenal world itself, both universally and within particular cultures.
Each of the papers making up this panel represents a contribution to the ongoing effort to uncover the order that underlies the apparent idiosyncrasy and complexity of transitive expression in Japanese, especially as it contrasts with English. Tsujimura proposes a meaning-based account of why it is that certain Japanese verbs have distinct forms dedicated to the respectively transitive and intransitive occurrence of an event while other verbs make available only the transitive expression of an event. Kitagawa proposes an original mechanism by which striking regularities can be formulated in the heretofore puzzling behavior of "bidirectional" transitive suffixes and verbs with multiple transitive or intransitive forms. Miyagawa argues for a treatment of the causative marker in Japanese as a transitivizing form which comes into play to fill lexical "gaps" left when no word is independently available to express transitive meaning. Jacobsen explores the role of reflexive meaning as occupying an intermediate domain between transitive and intransitive meaning, providing a key to explaining the existence in both Japanese and English of verbal constructions with ambivalent transitive/intransitive characteristics.
Toward a Semantic Characterization of Transitive-Intransitive Pairs in Japanese
Natsuko Tsujimura, Indiana University
It is well known that Japanese exhibits a large number of transitive-intransitive verb pairs that are morphologically related (cf. Jacobsen, 1992). They include kowasu/kowareru ‘break,’ ageru/agaru ‘rise,’ katameru/katamaru ‘solidify’ and kawakasu/kawaku ‘dry,’ among many more. On the other hand, there are verbs that show only one variant: sagasu ‘look for,’ as in John-ga saihu-o sagasita ‘John looked for his wallet’ is a transitive verb and does not have an intransitive variant; while hasiru ‘run,’ as in John-ga hasitta ‘John ran,’ is exclusively intransitive.
While many researchers have focused on the morphological question of how one variant is derived from the other and what is the direction of the derivation, i.e., from transitive to intransitive, or vice versa, Hayatsu (1989) investigates semantic factors that may determine when a transitive verb can and cannot have its intransitive variant. She reports that verbs whose meaning focuses on the resulting state of an event tend to show doublets while verbs that denote the process of an action are predisposed to allow for one variant. These criteria, though promising, call for further elaboration in order to identify specific meaning components that are attributable to the availability of doublets. In this paper I will take up an in-depth examination by extending Hayatsu’s initial observations to intransitive verbs that do not exhibit transitive counterparts and by comparing the Japanese pattern with a similar phenomenon observed in English.
References
Hayatsu, E. (1989) "Yuutsui tadooshi to mutsui tadooshi no chigai-ni
tsuite," Gengo Kenkyuu 95:231–256.
Jacobsen, W. (1992) The Transitive Structure of Events in Japanese. Kuroshio
Shuppan.
Transitivity Alternations in Japanese
Chisato Kitagawa, University of Massachusetts, Amherst
This paper, essentially based on Kitagawa and Fujii (1998), presents a new dynamic model to account for the transitivity alternation in Japanese in terms of the three distinct in/transitivizer morphemes—AS, AR, and E—with their formal function clearly defined. It addresses in particular the issues identified in (i) and (ii) along the line of argumentation presented in (iii) and (iv). In so doing, the paper aims to contribute to the current theoretical interest in syntax—"morphology" (or, more strictly, word-formation) interface issues as indicated in (v).
(i) The behavior of the formative E, seemingly disparate manifestations of which, as exemplified in (1), have so far resisted any uniform treatment; and (ii) derivation of multiples, exemplified in (2).
Transitive Intransitive
(1) a. or ‘break’ or-E ‘break’
b. ak-E ‘open’ ak ‘open’
c. kog-AS ‘burn’ kog-E ‘burn’
d. mag-E ‘bend’ mag-AR ‘bend’
(2) a. tuta-ga karam -ta hei (karam -ta -> karanda)
‘the wall on which vines have grown’b. tuta-o karam-E -ta hei
‘the wall on which they grew vines’c. tuta-ga karam-AR -ta hei (karam-AR -ta -> karamatta)
‘the wall on which some vines were grown’d. tuta-o karam-AS -ta hei (karam-AS -ta -> karamasita)
‘the wall on which they had vines grow’
(iii) Various manifestations of formative E exemplified in (1) are to be considered as allomorphic variations of a single morpheme. This is a claim that has never been advanced in any satisfactory fashion, although, with respect to the E’s exemplified in (1 a and b), Kageyama (1996:198) provides a cogent semantic argument in terms of which these instances of E’s may be viewed as related in a significant fashion.
(iv) Multiples such as shown in (2) can be seen as derivationally related in terms of specific principles. In fact, the phenomenon of transitivity doublets spawning multiples has been noted by a number of grammarians (e.g., Nishio 1954, Suga 1980, Aoki 1997, Kageyama 1996:191–92). The present paper offers, for the first time to the best of my knowledge, a morphological mechanism to account for the phenomenon.
(v) Transitivity alternation in Japanese presents an interesting case study whereby the exact relation between the levels of syntax and word formation may be examined and delineated in a revealing manner. This is because two of the three formatives participating in it, namely AS and AR, clearly have their historically related "doubles" in syntactically "productive" causative (s)ase and passive (r)are. The findings of this paper thus contribute to the cause of modular morphology (Borer, 1991 and Kageyama, 1993), elucidating the dynamic nature of Japanese (in)transitivization beyond the level assumed by Shibatani (1973), Baker (1988:212), and Jacobsen (1992).
Causative Constructions in Japanese and English
Shigeru Miyagawa, Massachusetts Institute of Technology
The causative construction in Japanese formed with the bound morpheme -sase (e.g., tabe-sase-ru ‘make eat’) has received more attention in Japanese linguistics than any other construction. In this paper, I will argue that the causative construction is an instance of transitivization, in that the causative morpheme -sase adds an extra argument, and assigns case to the original subject. For example, starting with the verb aruku (walk), the causative verb -sase that attaches to this verb provides an extra argument (the causer), and assigns case to the subject of aruku. This way of viewing the causative construction explains a number of important properties noted in the literature. For example, a causative verb sometimes functions as a simple transitive verb. An example of this is the causative verb niow-ase-ru, which has the literal meaning of ‘cause to smell,’ but it has an idiomatic meaning of ‘hint.’
1. Taroo-ga zisyoku-o niow-ase-ta
Taro-Nom resignation-Acc smell-cause-past
‘Taro hinted resignation.’
This type of idiomatic reading is only possible with simple transitive verbs. That this is idiomatic, hence non-compositional, is seen by the fact that the verb stem, niou (smell), does not allow the idiomatic interpretation. The following has only the literal meaning, and is pragmatically inappropriate.
2. *Zisyoku-ga niou
resignation-Nom smell
I will show that this "transitivizing" approach to causatives, developed on the basis of Japanese, applies also to English lexical causatives and to the analytical "make" causative.
Reflexive Meaning and Result versus Action Orientation in Japanese and English
Wesley M. Jacobsen, Harvard University
English and Japanese are standardly viewed as differing fundamentally in the preferred "vantage point" from which events are linguistically expressed—English preferring to express events in terms of who brought the event about (action orientation) and Japanese preferring to express them in terms of what happened as a result (result orientation) (Ikegami 1981). The large number of intransitive forms in Japanese which can only be expressed transitively in English appears to bear this out (e.g., Apaato ga mitsukatta vs. "I found an apartment"). Yet we also find a richly developed corpus of resultative constructions in English without counterparts in Japanese and, on the other hand, many transitive construction types in Japanese which require intransitive expression in English, going against this general pattern.
The picture is further complicated by some surprising regularities that exist between the two languages in domains of meaning where the borderline between transitive and intransitive meaning (and therefore between action and result) is blurred. Reflexive meaning is just such a case, as it involves two distinct meaning "roles" (agent and object) which are nevertheless borne by a single entity, thus exhibiting aspects of both transitive and intransitive meaning. Kageyama (1996) presents evidence that forms exhibiting "ambivalent" transitivity in both Japanese and English are motivated by a reflexive type of meaning structure (for example, "ergative" verbs such as open in English, which can be used either transitively or intransitively, and Japanese transitive/intransitive verb couplets involving the form e-, which can work either to change a transitive form into an intransitive (yaku "burn (tr)" —> yakeru "burn (in)") or vice versa (aku "open (in)" akeru "open (in)").
We will show that reflexive meaning can also be seen to be responsible for cases where Japanese exhibits transitive "confusion"—where, for example, a direct object occurs with an intransitive verb form (e.g., Inu ga shippo o tarete (nigeta) "The dog lowered its tail (and ran away)"). Contra Kageyama, however, who argues that reflexive meaning is realized on the surface in "unaccusative" intransitive form (i.e., in intransitive clauses where the subject behaves like an object, not an agent), this paper shows that reflexive meaning finds its primary manifestation in "unergative" intransitive constructions (where the subject is primarily agentive in role). Evidence for this comes both from paraphrases of "unergative" intransitives involving transitive reflexive objects (okiru "get up" = karada o okosu "get oneself up") and from aspectual considerations involving the suffix te-iru of the sort discussed in Jacobsen (1992). Parallel tendencies will be observed for English, allowing us to posit a "core" area of meaning where the two languages exhibit fundamentally similar behavior, the traditionally observed differences between them occurring instead at peripheral zones of meaning at the extreme ends of a scale of transitive/intransitive meaning.