Show Summary Details

Page of

PRINTED FROM the OXFORD RESEARCH ENCYCLOPEDIA, LINGUISTICS (oxfordre.com/linguistics). (c) Oxford University Press USA, 2019. All Rights Reserved. Personal use only; commercial use is strictly prohibited (for details see Privacy Policy and Legal Notice).

Subscriber: null; date: 15 February 2019

Compound and Complex Predicates in Japanese

Summary and Keywords

Compound and complex predicates—predicates that consist of two or more lexical items and function as the predicate of a single sentence—present an important class of linguistic objects that pertain to an enormously wide range of issues in the interactions of morphology, phonology, syntax, and semantics. Japanese makes extensive use of compounding to expand a single verb into a complex one. These compounding processes range over multiple modules of the grammatical system, thus straddling the borders between morphology, syntax, phonology, and semantics. In terms of degree of phonological integration, two types of compound predicates can be distinguished. In the first type, called tight compound predicates, two elements from the native lexical stratum are tightly fused and inflect as a whole for tense. In this group, Verb-Verb compound verbs such as arai-nagasu [wash-let.flow] ‘to wash away’ and hare-agaru [sky.be.clear-go.up] ‘for the sky to clear up entirely’ are preponderant in numbers and productivity over Noun-Verb compound verbs such as tema-doru [time-take] ‘to take a lot of time (to finish).’

The second type, called loose compound predicates, takes the form of “Noun + Predicate (Verbal Noun [VN] or Adjectival Noun [AN]),” as in post-syntactic compounds like [sinsya : koonyuu] no okyakusama ([new.car : purchase] GEN customers) ‘customer(s) who purchase(d) a new car,’ where the symbol “:” stands for a short phonological break. Remarkably, loose compounding allows combinations of a transitive VN with its agent subject (external argument), as in [Supirubaagu : seisaku] no eiga ([Spielberg : produce] GEN film) ‘a film/films that Spielberg produces/produced’—a pattern that is illegitimate in tight compounds and has in fact been considered universally impossible in the world’s languages in verbal compounding and noun incorporation.

In addition to a huge variety of tight and loose compound predicates, Japanese has an additional class of syntactic constructions that as a whole function as complex predicates. Typical examples are the light verb construction, where a clause headed by a VN is followed by the light verb suru ‘do,’ as in Tomodati wa sinsya o koonyuu (sae) sita [friend TOP new.car ACC purchase (even) did] ‘My friend (even) bought a new car’ and the human physical attribute construction, as in Sensei wa aoi me o site-iru [teacher TOP blue eye ACC do-ing] ‘My teacher has blue eyes.’ In these constructions, the nominal phrases immediately preceding the verb suru are semantically characterized as indefinite and non-referential and reject syntactic operations such as movement and deletion. The semantic indefiniteness and syntactic immobility of the NPs involved are also observed with a construction composed of a human subject and the verb aru ‘be,’ as Gakkai ni wa oozei no sankasya ga atta ‘There was a large number of participants at the conference.’ The constellation of such “word-like” properties shared by these compound and complex predicates poses challenging problems for current theories of morphology-syntax-semantics interactions with regard to such topics as lexical integrity, morphological compounding, syntactic incorporation, semantic incorporation, pseudo-incorporation, and indefinite/non-referential NPs.

Keywords: compounding, incorporation, indefiniteness, lexical aspect, lexical integrity, grammaticalization, individual-level predication, syntax-morphology-semantics interface

1. Varieties of Compound and Complex Predicates

As seen from a recent explosion of publications such as Borik and Gehrke (2015), Amberber, Baker, and Harvey (2010), Nash and Samvelian (2016), and Nolan and Diedrichsen (2017), compound and complex predicates—predicates that consist of two or more lexical items and function as the predicate of a single sentence—present an important class of linguistic objects that pertain to an enormously wide range of theoretical issues in the interactions of morphology, phonology, syntax, and semantics. Languages offer different morphosyntactic mechanisms for the creation of such complex predicates. In English, for example, addition of affixes (e.g., connectdisconnect) and particles (e.g., looklook up) is primarily responsible for extension of verbal concepts, while concatenation of multiple verbs is rare or impossible. By contrast, Japanese makes extensive use of compounding to expand a single predicate to a multipart predicate involving two or more verbs. The preponderance of such multi-verb predicates in Japanese might be readily associated with its verb-final word order and agglutinating character, where multiple auxiliaries and inflections designating tense, aspect, modality, and politeness are combined one after another into an increasingly larger predicate sequence, as in tabete simai soo desi ta [eat finish look-like POLITE PAST] ‘It looked like he was about to eat it all up.’ Creation of such a complex sequence, however, is not limited to notions of TAM (tense, aspect, modality) but extends into the main verb itself, thus giving rise to clusters of two or more verbs, as in tabe-nokosu [eat-leave] ‘leave some food uneaten’ and tabe-nokosi-kakeru [eat-leave-be.about.to] ‘be about to leave some food uneaten.’ In fact, Japanese boasts a startling variety of compound predicates consisting of a head predicate of various categories (verb [V], adjective [A], verbal noun [VN], or adjectival noun [AN]) preceded by another element of diverse kinds. Combinations like V-V, N-V, N-A, N-VN, and N-AN are thus available as compound predicates, though productivity varies with individual types. Besides, Japanese has productive syntactic suffixes for passive (rare), causative (sase), potential ((ra)re), and desiderative (tai) which deserve special treatment in a separate paper. These are excluded from this article because they are bound affixes rather than compounding elements.

This article provides a compact and yet comprehensive survey of the varieties of compound and complex predicates available in contemporary Japanese, viewed from the standpoint of morphology, semantics, syntax, and language typology. Specifically, it aims to open up a new perspective on the understanding of their mutual relationships as well as their common principles by partitioning them into three categories according to the strength of cohesion between their members:

  1. I. Morphologically tight compound predicates (section 2);

  2. II. Morphologically loose compound predicates (section 3); and

  3. III. Syntactically complex predicates (section 4).

Tight compounds can be identified by the standard phonological criteria, including (i) accentual unity (a word, whether simplex, derived, or complex, is pronounced with a single stretch of accent, called “lexical accent”: see Kubozono, 2016) and (ii) sandhi phenomena that are observed at compound-internal boundaries, such as rendaku (sequential voicing of the initial voiceless obstruent of the second member of a compound: see Vance, 2016) and vowel mutations as relics of earlier Japanese. For example, the noun sake [saKE] ‘liquor, Japanese sake’ is compounded with the noun kura [kuRA] ‘vault’ to yield the compound noun saka-gura ‘liquor vault,’ where the last vowel [e] in sake is realized as [a], the initial consonant [k] of kura is changed to [g] by rendaku, and the whole compound is pronounced as [saKAGURA] by lexical accentuation (capital letters stand for high pitches).

Morphologically tight compounds are contrasted with morphologically loose compounds. The phonological rules that characterize tight compounds are by and large restricted to members of the native (or Wago) lexical stratum. On the other hand, morphologically loose compounds, mostly originating from the Sino-Japanese vocabulary stratum, do not observe the native lexical accent or sandhi rules but behave phonologically as if they were sequences of independent words. Notable examples are a set of Sino-Japanese prefixes like hi- ‘non,’ zen- ‘ex-,’ and kaku- ‘each’ that exhibit phrase-like accent (Poser, 1990; Kageyama, 2001). These prefixes are similar to English non- (e.g., non-European) and ex- (e.g., ex-husband) in that they are pronounced with a short break after them, as in hi : yooroppa-gengo ‘non-European languages,’ zen : daitooryoo ‘ex-president,’ and kaku : daigaku ‘each university’ (the colon (:) is intended to indicate the characteristic break associated with these prefixes). The same phonological break is observed with Sino-Japanese compound words as well, e.g., sekai : saidai [world : biggest] ‘the biggest in the world’ and Okkusufoodo-daigaku : gakutyoo [Oxford-university : president] ‘President of the University of Oxford.’ Although the phonological break might suggest a phrasal status, the standard diagnoses for “lexical integrity,” such as the impossibility of inserting a syntactic element or moving a component syntactically, indicate that these sequences indeed count as “words” (Kageyama, 2001).

The tight-loose distinction of compounds does not exhaust the inventory of multipart predicates in Japanese. There are several classes of complex predicates that have the appearance of syntactic phrases and yet behave functionally as single predicates on a par with compound predicates. These complex predicates involve not only the well-known light verb construction composed of a Verbal Noun [VN] and suru ‘do,’ but also less well-known constructions consisting of particular types of nouns followed by suru ‘do’ or aru ‘be.’ While these verbal constructions tend to be regarded as matters of syntax, section 4 will make a new attempt to analyze them as complex predicates involving “semantic incorporation” of indefinite/non-referential elements.

Integration of the three types of multipart predicates reveals that Japanese exploits an amazingly vast array of predicate formations that are not easily found in other languages. The constellation of properties inherent in such multipart predicates poses challenging problems for current theories of morphology-syntax-semantics interface relating to such phenomena as lexical integrity, morphological compounding, syntactic incorporation, semantic incorporation, pseudo-incorporation, and indefinite/non-referential NPs (section 5).

2. Morphologically Tight Compound Predicates

Tight compound predicates in Japanese are typified by the Verb-Verb (V-V) compound verbs and Noun-Verb (N-V) compound verbs that are pronounced as one word with a single stretch of accent, as in aRUKI-TUKAREru [walk-get.tired] ‘get tired from walking’ and teMA-DOru [time-take] ‘take lots of time’ (capital letters indicate high pitch). Their word status is confirmed by tests for lexical integrity (“The syntax neither manipulates nor has access to the internal structure of words”: Anderson, 1992, p. 84). For example, both V-V and N-V compound verbs resist insertion of syntactic elements such as focus particles inside them, as shown by the total unacceptability of *aruki-sae-tukareru [walk-even-get.tired] and *tema-sae-doru [time-even-take]. This section introduces four types of tight compound predicates, together with V-te V concatenations:

  • Syntactic Verb-Verb compound verbs (sections 2.1 and 2.2);

  • Lexical Verb-Verb compound verbs (sections 2.1 and 2.3);

  • Noun-Verb compound verbs (section 2.4);

  • Noun-Adjective compound adjectives (section 2.5); and

  • V-te V concatenations (section 2.6).

All of these inflect for tense and can be used to close a finite sentence. The presence or absence of tense inflection plays a key role in distinguishing them from the tenseless or non-finite compound predicates to be addressed in section 3.

2.1 Two Types of V-V Compound Verbs: Lexical Versus Syntactic

V-V compound verbs share the morphological composition of a first verb (V1) in the “infinitive” form (Bloch, 1946; Martin, 1975)—called ren’yōkei (‘predicate-modifying form’) in traditional Japanese grammar—plus a second verb, and the V1-V2 amalgam as a whole is subject to tense and other inflections. The infinitive in V1 is identical in shape to the verb stem if it ends in a vowel, as in hazime- < stem hazime- ‘to begin’; otherwise, it has a supplementary vowel i added at the end of a stem, as in huri- < stem hur- ‘(for rain or snow) to fall.’

There is good reason to divide the V-V compound verbs into two classes, as illustrated in (1) and (2).

(1)

Compound and Complex Predicates in Japanese

(2)

Compound and Complex Predicates in Japanese

The lexical-syntactic distinction, first proposed by Kageyama (1989, 1993) and developed by Yumoto (2005, 2009), is based on syntactic diagnoses testing whether the V1 position can accommodate syntactically motivated elements such as passive verbs with -rare, causative verbs with -sase, honorific verbs with o-V-ni naru, light verb constructions based on a Verbal Noun and the verb suru ‘do,’ and idiom chunks consisting of a verb and an object or other element. Because of their syntactic nature, these elements are expected not to occur in the V1 position inside compound verbs. This prediction is borne out in the class of lexical compound verbs of (1). Surprisingly, however, such syntactic elements can legitimately occur in the V1 position in the class of syntactic compound verbs of (2). The contrast between the two classes is illustrated in Table 1.

Table 1. Syntactic Criteria for the Lexical-Syntactic Distinction

Lexical Compound Verbs

Syntactic Compound Verbs

Passives in V1

*[os-are]-aku ([push-PASS]-openvi) Lit. ‘open (vi.) by being pushed’

[ais-are]-tuzukeru ([love-PASS]-continue) ‘continue to be loved’

Honorific verbs in V1

*[o-uke ni nari]-toru ([HON-get DAT become]-take) ‘receive’

[o-utai ni nari]-kakeru ([HON-sing DAT become]-be.about.to) ‘be about to sing’

Light verb constructions in V1

*[rakka-si]-yamu ([drop-do]-stop) ‘stop falling’

[rakka-si]-owaru ([fall-do]-stop) ‘stop falling’

Idiom chunks in V1 (e.g., abura o uru ‘shoot the breeze’)

#[abura o uri]-siburu ([oil ACC sell]-hesitate) ‘hesitate to sell oil/*shoot the breeze’

[abura o uri]-sokonau ([oilACC sell]-fail) ‘fail to sell oil/shoot the breeze’

The inclusion or exclusion of such syntactically motivated elements is observed holistically and systematically between the two classes (Kageyama, 1993). The systematicity is robust evidence for distinguishing the two classes as qualitatively distinct categories. The distinction is further corroborated by their different degrees of productivity. Using C. E. Shannon’s theory of entropy and redundancy, Tamaoka, Lim, and Sakai (2004) observe that syntactic compound verbs are significantly higher in entropy than lexical ones though they do not differ markedly in redundancy.

On the assumption that the formation of lexical compounds precedes the formation of syntactic compounds in the architecture of grammar, it is predicted that a lexical compound verb may precede (i.e., be embedded in) a syntactic compound verb, but not the other way around. The prediction proves accurate in (3a) and (3b).

(3)

Compound and Complex Predicates in Japanese

Note that the lexical-syntactic distinction is not intended to assign each compound verb uniquely to either the lexical or the syntactic class; instead, there are several head verbs that have a dual membership, a well-known example being -kakeru ‘hang’ in hanasi-kakeru [speak-hang], which means either ‘talk to (a person)’ (as a lexical compound) or ‘be about to talk’ (as a syntactic compound).

2.2 Syntactic V-V Compound Verbs

Syntactic compound verbs can take syntactically motivated elements in the V1 position as long as the last element is a verb in the infinitive form. This leads to an apparent contradiction: on the one hand, the infinitive verb in V1 makes up a morphological compound word with the verb in V2, while on the other hand, the elements associated with V1 are syntactically motivated.

2.2.1 Syntactic Complementation Structure

It is commonly assumed in the literature that syntactic V-V compound verbs have an embedded structure in syntax, where the head verb (V2) takes a complement clause containing the first verb (V1) as its head, as schematically shown in (4).

(4)

Compound and Complex Predicates in Japanese

This syntactic structure alone cannot explain the morphological status of the V1-V2 sequence as a compound word. A plausible solution (Kageyama, 1993) is to assume that the infinitive verb (V1) in the complement clause is compounded with the head verb (V2) by the rule of head movement called “Incorporation” (Baker, 1988). For example, when the head verb kakeru [lit. ‘hang’] ‘be about to’ occurs with the idiom abura o uru [oil ACC sell] ‘shoot the breeze,’ the base structure in syntax looks like (5a), from which the infinitive verb uri ‘sell’ is adjoined to the head verb kake-ta ‘was about to.’

(5)

Compound and Complex Predicates in Japanese

Alternative analyses include restructuring (Fukuda, 2012) and reanalysis (Kishimoto, 2013).

2.2.2 Semantic Bleaching in Head Verbs

There are 30 head verbs in V2 position that fulfill the syntactic conditions introduced in section 2.1 (Kageyama, 1993; Himeno, 1999). These verbs, while maintaining the lexical category of “verb,” are de-lexicalized and semantically bleached. Table 2 (adapted from Kageyama, 2016a) provides a comprehensive list of the 30 head verbs classified roughly into semantic groups.

Table 2. Syntactic Compound Verbs Classified Semantically

Semantics

Examples

inception

V-kakeru [hang, vt.] ‘be about to V,’ V-dasu [take out, vt.] ‘begin to V,’ V-hazimeru [begin, vt.] ‘begin to V,’ V-kakaru [set about, vi.] ‘set about V-ing’

duration, repetition

V-tuzukeru [continue, vt.] ‘continue to V,’ V-makuru [roll up, vt.] ‘V on and on intensely’

completion

V-oeru [finish, vt.] ‘finish V-ing,’ V-owaru [end, vi.] ‘stop V-ing,’ V-tukusu [exhaust, vt.] ‘V exhaustively,’ V-kiru [cut, vt.] ‘V completely,’ V-toosu [let through, vt.] ‘V to the end,’ V-nuku [pull out, vt.] ‘V to the end,’ V-hateru [come to an end, vi.] ‘V utterly’

incompletion, failure

V-sokonau [harm, vt.] ‘miss V-ing,’ V-sokoneru [harm, vt.] ‘fail to V,’ V-sonziru [damage, vt.] ‘fail to V,’ V-sobireru [miss (auxiliary use only)] ‘fail to V,’ V-kaneru [be.unable] ‘be unable to V,’ V-wasureru [forget] ‘forget to V,’ V-nokosu [leave undone] ‘leave something without V-ing completely,’ V-ayamaru [err, vt.] ‘make a mistake in V-ing,’ V-okureru [be late, vi.] ‘be delayed in V-ing,’ V-aguneru [Classical Japanese ‘be satiated’] ‘hesitate to V’

excessiveness

V-sugiru [go past, vi.] ‘V excessively’

retrial

V-naosu [repair, vt.] ‘V again (to obtain a desired result)’

repetition, habituation

V-tukeru [attach, vt.] ‘be used to V-ing,’ V-nareru [get accustomed, vi.] ‘be accustomed to V-ing,’ V-akiru [get weary, vi.] ‘get weary of V-ing’

reciprocity

V-au [meet, vi.] ‘V reciprocally’

likelihood

V-eru or V-uru [obtain, be possible] ‘be likely to V’

Most of the de-lexicalized meanings are subsumed under the category of “grammatical (viewpoint) aspect,” such as inception, continuation, and completion, but meanings that are not commonly deemed aspectual are also observed, such as reciprocity (au lit. ‘meet’) and excessiveness (sugiru lit. ‘go beyond’).

The degrees of grammaticalization are not uniform. Verbs like hazimeru ‘begin,’ tuzukeru ‘continue,’ and akiru ‘get weary’ retain their respective original meanings, while others have developed aspectual meanings, as in dasu (lit. ‘take out’), which signifies the inception of an event presumably due to the outward movement of its original meaning, kiru (lit. ‘cut’), which expresses full completion of an action leaving nothing undone, and naosu (lit. ‘repair’), which indicates retrial of an action for the purpose of attaining an intended goal; still others have not only undergone semantic change but have also lost the lexical linkage with the original independent verbs and now function only as the V2s of compound verbs, as in aguneru (originally ‘be at a loss’) ‘hesitate to’ and sobireru (originally ‘miss a chance’) ‘fail to.’

2.2.3 Productivity

Although the head verbs in V2 constitute a closed class, the formation of syntactic V-V compound verbs is deemed fully productive. Any infinitive verb can fit in with the V1 position as long as it is semantically congruent with the head verb.

2.2.4 Case-Marking and Argument Structure

While sharing the canonical complementation structure of (4), three subclasses can be distinguished depending on how the argument relations of the whole compound are determined. Of the three subclasses, two are well known: (i) “Control type,” where the subject must be a human (or sentient) noun and bears the semantic role of Agent or Experiencer (e.g., V-sobireru ‘miss V-ing,’ V-tukeru ‘be used to V-ing,’ V-kaneru ‘be unable (not in a position) to V,’ V-aguneru ‘hesitate to V,’ V-nareru ‘be accustomed to V-ing,’ V-akiru ‘get weary of V-ing’); (ii) “Raising type,” where no animacy condition is imposed on the subject (e.g., V-kakeru ‘be about to V,’ V-dasu ‘start to V,’ V-sugiru ‘V excessively,’ V-kanenai ‘something bad is likely to happen’). For example, the clause ame ga huru ‘fain falls’ with an inanimate subject is compatible with the Raising type, as in ame ga huri-kaketa ‘It was about to rain,’ but not with the Control type, as shown by the ungrammaticality of *ame ga huri-sobireta lit. ‘Rain missed falling.’

Kageyama (1993) points out a third type of V2, which imposes a semantic restriction on both the subject and the object simultaneously. Examples include V-oeru ‘finish V-ing,’ V-naosu ‘V again,’ V-tukusu ‘V exhaustively,’ and V-wasureru ‘forget to V.’ An outstanding property of these V2s is that they are susceptible to what Nishigauchi (1993) called “long distance passive,” where the object of an embedded clause (V1) can be passivized crossing the clause boundary of the complementation structure, as in (6).

(6)

Compound and Complex Predicates in Japanese

The phenomenon of long distance passive is truly idiosyncratic in view of the fact that normally, passive may apply only to V1 in both the control type and the raising type, and the sequence “V1-V2-passive” is ill-formed. Different analyses have been proposed in the framework of generative syntax by Nishigauchi (1993), Kageyama (1993), Yumoto (2005), and Kishimoto (2013).

As seen from the preceding discussion, syntactic V-V compound verbs present a great number of interesting issues in the realms of morphology, syntax, semantics, and grammaticalization (see Kageyama (2016a) for a more detailed exposition).

2.3 Lexical V-V Compound Verbs

Lexical V-V compound verbs are not capable of accommodating syntactically motivated elements in V1 position. This means that an infinitive verb (V1) is directly adjoined to the head verb (V2) to make up a morphological unit of compound word, as in V1 + V2 → [V1-V2].

2.3.1 Productivity

Unlike syntactic V-V compound verbs, lexical V-V compound verbs form a more or less closed set. The Compound Verb Lexicon (Kageyama & Kanzaki, 2014), a free online dictionary on the website of the National Institute for Japanese Language and Linguistics, contains 2,756 lexical compound verbs commonly used in contemporary standard Japanese, and the number will increase slightly if uncommon verbs used in literary works are included. Nonetheless, they are not freely productive, though new compounds can be coined sporadically. The number of actual compounds heavily depends on individual V2 verbs. The Compound Verb Lexicon displays, for example, as many as 255 compounds ending with komu ‘go in’ as V2 but only one compound ending with tagiru ‘boil violently’ as V2 (nie-tagiru ‘boil over violently’).

2.3.2 Combinatory Restrictions

A recalcitrant problem about lexical V-V compound verbs has to do with how the combinatory restrictions on two verbs are determined. Broadly speaking, two approaches have been proposed to pinpoint the nature of the restrictions. One is a cognitive/semantic approach. Under the action chain hypothesis (verbs express an event as a chain of several phases in the flow of time, as in “action → causation → change → result”), it is natural that many Japanese V-V compound verbs express an agent’s action as the first verb (V1) and the change of state/location resulting from that action as the second verb (V2) (Ishii, 1983; Hayatsu, 1989). For example, the act of pushing a door open is represented by the compound verb osi-akeru [push-open(tr.)] with ‘pushing’ before ‘opening,’ and not by *ake-osu [open(tr.)-push] with the opposite order.

The other approach, which is actually complementary with the cognitive-semantic approach, invokes the notion of argument structure. In the literature on lexical semantics based on the Unaccusative Hypothesis, three verb classes are distinguished: (i) transitive verbs, which take both an external argument (x) and an internal argument (y); (ii) unaccusative verbs, which take only an internal argument (y); and (iii) unergative (intransitive) verbs, which take only an external argument (x). Capitalizing on this tripartition, Kageyama (1993) stipulates the Transitivity Harmony Principle, which says that the two verbs that compose a lexical compound verb must be of the same type in argument structure, where the sameness of argument structure is determined by the presence or absence of an external argument (or agent). Transitive verbs and unergative verbs are of the same type, whereas unaccusative verbs, lacking an external argument, are a different type. The Transitivity Harmony Principle makes a specific prediction that a transitive verb may be combined with another transitive verb or with an unergative verb, but not with an unaccusative verb, as summarized in Table 3. Kageyama (2016d) claims that this principle is cognitively motivated to guarantee the identity of two agents involved in a single event.

Table 3. Possible and Impossible Combinations Predicted by the Transitivity Harmony Principle

V2: Transitive

V2: Unergative

V2: Unaccusative

V1: transitive

*

V1: unergative

*

V1: unaccusative

*

*

As predicted, combinations of two transitive verbs such as osi-akeru [push-open(tr.)] are amply attested, whereas combinations of a transitive verb and an unaccusative verb, such as *osi-aku [push-open(intr.)], are not. Frellesvig et al. (2010) report that Old Japanese did not have full-fledged morphological compound verbs, and consequently transitivity harmony was not observed. In contemporary Japanese, on the other hand, the Transitivity Harmony Principle is a robust rule supported by the statistical data of about 83.4% of the lexical V-V compounds in the Compound Verb Lexicon being congruent with it. The percentage is amazingly high in view of the fact that the online dictionary contains pseudo-compound verbs whose first members are grammaticalized to prefixes.

The Transitivity Harmony Principle is pervasive in lexically derived compound verbs but is irrelevant to syntactic compound verbs (section 2.2). Its validity and nature have been debated by a number of researchers (see Nishiyama (2008) for a thorough review of relevant literature).

2.3.3 Semantic Relations Between V1 and V2

While the Transitivity Harmony Principle captures the matching relation of the two agents associated with the member verbs, it says nothing about the semantic relations between the two events expressed by V1 and V2. At least six different semantic relations have to be postulated, as shown in (7).

(7)

Compound and Complex Predicates in Japanese

A variety of analyses have been proposed in a variety of theoretical frameworks, such as Kageyama (1993) and Yumoto (2005) in Conceptual Semantics, Matsumoto (1996, 1998) in Lexical-Functional Grammar, Toratani (2002) in Role-and-Reference Grammar, Naumann and Gamerschlag (2003) in formal semantics, and Fukushima (2005) in macro-role semantics.

While the previous studies treated all of the six relations more or less on an equal footing, Kageyama (2016a) makes a novel proposal to divide them into two large groups, depending on which of V1 and V2 determines the argument structure of the whole compound: the class of thematic compound verbs and the class of aspectual compound verbs. According to this new classification, the semantic core of a compound event lies in V2 (morphologically a head) in the thematic class but in V1 (morphologically a modifier) in the aspectual class. A heuristic for the distinction is paraphrasability by the construction “V1-te V2” (‘V1 and V2’), where V1-te functions as a converb modifying the head verb V2. Thus, if a given compound can be paraphrased by this construction, its semantic core is the V2 and hence it is a thematic compound verb; otherwise, it belongs to the class of aspectual compounds. For example, doa o osi-akeru ‘push the door open’ can be paraphrased by doa o osi-te akeru ‘push the door and open it’ and therefore osi-akeru is a thematic compound verb. All the semantic relations summarized in (7a) through (7d) converge on this type. By contrast, compounds holding the complementation relation (7e) and adverbial relation (7f) do not fit in with such a paraphrase, as seen from the oddity of mi-te nogasu lit ‘see and miss’ for mi-nogasu ‘overlook’ and sikat-te tukeru lit. ‘scold and attach’ for sikari-tukeru ‘scold severely.’ The inappropriateness of such a paraphrase for (7e) and (7f) suggests that their two members do not have a relation of left-to-right modification. On the contrary, many examples of the types (7e) and (7f) are susceptible to a paraphrase where the modification relation is reversed in such a way that V2 semantically modifies V1, as in isoi-de sinu for sini-isogu ‘try to hurry death.’

The distinction between thematic and aspectual compound verbs is reflected in a notable disparity in the applicability of the Transitivity Harmony Principle. Kageyama (2016d) thus points out that the degree of THP congruence among thematic compounds (89.45%) is significantly higher than among aspectual compounds (71.39%) [χ‎2(1)=128.01, p<.001].

2.3.4 Argument Realization

Compound and Complex Predicates in JapaneseClick to view larger

Figure 1. Lexical thematic compound verbs.

Compound and Complex Predicates in JapaneseClick to view larger

Figure 2. Lexical aspectual compound verbs.

The disparity in the locus of the semantic core between the two classes of lexical compound verbs suggests that different mechanisms are involved in their semantic interpretations, as schematically represented in Figures 1 and 2 (Kageyama, 2016a).

In thematic compound verbs (Figure 1), V2 is identified as the morphological and semantic head that provides the core lexical meaning and argument structure for the whole compound, with V1 supplying a semantic modification to it. In aspectual compound verbs (Figure 2), on the contrary, V1 serves as the semantic head that provides the core lexical meaning and argument structure, into which V2 feeds an additional Aktionsarten meaning.

2.3.5 De-Lexicalized Verbs in Aspectual Compound Verbs

While Masica (1976) identified compound and complex verbs as an areal feature of Asian languages, only Japanese and possibly Korean appear to have a robust class of lexical aspectual compound verbs. Although this class is apparently similar to the class of syntactic compound verbs, their aspectual meanings are different: syntactic compound verbs represent grammatical (viewpoint) aspect, whereas lexical aspectual compound verbs elaborate on the internal structure of V1’s Aktionsarten or situation aspect. Compare, for example, a syntactic compound verb (Hana ga) saki-dasita or saki-hazimeta ‘Flowers began to bloom’ with a lexical aspectual compound verb (Hana ga) saki-someta (same translation in English). Although it is hard to pinpoint their semantic difference precisely, many native speakers agree with the following intuition: the former is an objective report of the beginning of blooming and simply states that the event of flowers blooming began; by contrast, the latter focuses on the initial phase of the blooming event, implying the beautiful state of newborn flowers. As shown in Table 4, lexical aspectual compounds cover a far wider range of fine-grained aspectual meanings than the syntactic compound verbs.

Table 4. Bleached Meanings in Lexical Aspectual Compound Verbs (Kageyama, 2016a)

a. completive

huri-yamu (vi.) ‘stop falling, said of snow and rain,’ kaki-ageru (vt.) ‘finish writing,’ kaki-agaru (vi) ‘writing is finished,’ ni-tumeru (vt.) ‘boil down’

b. incompletive

ii-sasu (vt.) ‘stop speaking halfway’

c. intensive result

ne-komu (vi.) ‘fall sound asleep,’ komari-hateru (vt.) ‘be completely at a loss,’    sizumari-kaeru (vi.) ‘become completely silent,’ saki-midareru (vi.) ‘bloom in profusion,’ otituki-harau (vi.) ‘be perfectly composed’

d. inception

ake-someru (vi.) ‘begin to dawn,’ saki-someru (vi.) ‘begin to bloom’

e. continuative

huri-sikiru (vi.) ‘(rain) fall incessantly,’ naki-kurasu (vi.) ‘cry all day’

f. iterative

hozikuri-kaesu (vt.) ‘dig again,’ tukai-komu (vt.) ‘use repeatedly,’ tate-kaeru (vt.) ‘rebuild,’ ii-narawasu (vt.) ‘commonly say’

g. intensive action

sawagi-tateru (vi.) ‘fuss about,’ izikuri-mawasu (vt.) ‘fumble about,’ waki-kaeru (vi.) ‘boil violently,’ home-tigiru (vt.) ‘praise highly’

h. ineffective

kasi-siburu (vt.) ‘hesitate to lend,’ nobi-nayamu (vi.) ‘do not make expected progress,’ sagasi-aguneru (vt.) ‘be unable to find,’ kiki-tigau (vt.) ‘hear wrongly’

i. reciprocal

i-awaseru (vi.) ‘happen to be at the same place,’ tukai-wakeru (vt.) ‘use different things according to the needs’

2.4 N-V Compound Verbs

Compared with V-V compound verbs, Noun-Verb compound verbs (with tense inflection) are not productive, with only 60 or 70 verbs actually occurring in contemporary Japanese. Lack of productivity is correlated with the paucity of the literature on them (e.g., Kageyama, 1980, 2016b; Watanabe, 1999). In contrast to V-V compound verbs, which are said to be absent in Old Japanese (Frellesvig et al., 2010), N-V compound verbs have been attested since the earliest period (Sakakura, 1966; Kinuhata, 2010), although their formation has never been fully productive throughout the history of Japanese. Some examples from contemporary Japanese are shown in (8).

(8)

Compound and Complex Predicates in Japanese

The word status of these combinations is indicated by their lexical accent, coupled with the occasional application of sandhi rules such as rendaku (e.g., datu for tatu ‘stand’ in nami-datu) and vowel mutation (e.g., ta for te ‘hand’ in ta-basamu).

In morphological studies, it has been established that verbal compounding and noun incorporation may apply to direct objects of transitive verbs, subjects of unaccusative verbs, and adjuncts, but never to subjects of transitive verbs (Lieber, 1983; Mithun, 1984; Baker, 1988). As seen from (8), the possible range of combinations in Japanese N-V compound verbs conforms in full with this universal tendency, whereby no unequivocal compound verbs involving a transitive verb and its subject are found. One possible counterexample, musi-bamu [bug-bite] ‘to affect adversely,’ is actually not problematic, as it is interpreted as a kind of manner adverb meaning ‘like a bug, as if by a bug.’

Watanabe (1999) proposes to derive these N-V compound verbs by a syntactic head-movement in line with Hale and Keyser’s lexical syntactic structure with layered VP shells. However, lack of productivity and the abundance of lexical irregularities such as vowel mutation, bound morphemes, and idiomatization point to lexical derivation.

2.5 N-A Compound Adjectives

Like tensed N-V compound verbs (section 2.4), tensed Noun-Adjective compound adjectives, exemplified in (9) below, exhibit lexical accent and rendaku but are lexically restricted and hardly productive.

(9)

Compound and Complex Predicates in Japanese

Kishimoto and Booij (2014) bring to light interesting sets of compound adjectives that are headed by the negative adjective nai ‘nonexistent, null.’

(10)

Compound and Complex Predicates in Japanese

(11)

Compound and Complex Predicates in Japanese

All of these compound adjectives are paraphrased by representing the subject arguments with the nominative case particle (ga), which is presented above as an optional element in parentheses. On the premise that the presence of a case marker signals phrasal status, it is expected that if the nominative marker is present, the combination of a noun and the negative adjective nai is a syntactic phrase; if not, it is a compound. This is indeed correct in the case of the examples in (10), where an adverb like mattaku ‘(not) at all’ may be inserted between the subject and nai only if the case particle is present.

(12)

Compound and Complex Predicates in Japanese

The expressions in (10) thus parallel tensed N-V compound verbs, which can be paraphrased with case particles, as in nami ga tatu (wave NOM rise) versus [nami-datu].

Now, Kishimoto and Booij (2014) observe unexpected behavior with the examples in (11), which resist adverb insertion regardless of the presence or absence of the case particle, as shown in (13).

(13)

Compound and Complex Predicates in Japanese

From such observations, Kishimoto and Booij (2014) conclude that the entire complexes with case-marked nouns in (11) “behave as lexical units, while their components are visible syntactically” (p. 59) and analyze them as a case of “quasi-noun incorporation,” where the case-marked noun is adjoined to the head adjective nai to form a compound adjective.

While Kishimoto and Booij (2014) regard the class of compound adjectives in (10) as lexically derived, Kageyama (2016b) suggests the possibility that they are indeed formed in syntax. Kageyama’s evidence concerns the deletion of their noun members.

(14)

Compound and Complex Predicates in Japanese

In B’s responses, nai takes a null subject (Ø) that refers to the N elements (atogusare ‘trouble arising later’ and nukari ‘oversight’) in the compound adjectives used in A’s questions. Notably, such deletion of noun elements is prohibited from the Noun-Verb compounds discussed in section 2.4.

(15)

Compound and Complex Predicates in Japanese

The impossibility of noun deletion holds for the N-A compound adjectives in (11) as well. The N-V compound verbs and the N-A compound adjectives in (11) are thus plausibly derived by simple concatenation of a noun and a verb/adjective. On the other hand, the compound negatives in (10), which allow noun deletion, call for special treatment, where a noun and the negative adjective nai are given as separate constituents in syntactic structure and are later fused into a compound word.

2.6 V-te V Concatenations

Discussion of Japanese V-V complexes cannot be complete without mentioning what might be called “V-te V concatenations,” where the morpheme -te, analyzed as a kind of verbal inflection (gerundive), is a connective morpheme that links a verb (or more precisely a clause headed by a verb) to the following verb or clause. It thus functions typically as a linker for coordinate clauses or as a marker of converbs (verbs used as adverbials) (see Hasegawa, 1996 for details). For our discussion, it is important that the V-te form can be used to make a complex predicate of the form “V1-te V2” in combination with a set of designated (auxiliary-like) verbs, as shown in Table 5 (simplified from Nakatani, 2016).

Table 5. List of Grammaticalized Verbs That Combine With V1-te

V2 Verbs

Original Meanings

Grammaticalized Meanings

Kuru

iku

‘to come’

‘to go’

Coming/going motion, a gradual change/advancement of state/event, etc.

iru

‘(for an animate) to exist’

Progressive or perfect aspect

aru

‘(for an inanimate) to exist’

Perfect aspect, with a connotation of preparation

oku

‘to put’

Preparation

simau

‘to put away’

Emphasis on completion

miru

‘to see’

Attempt

miseru

‘to show’

Demonstration

Kureru

ageru

‘to give (to the speaker)’

‘to give (to a non-speaker)’

Benefactive activity

morau

‘to receive/be given’

Receipt of a benefit

Most of these auxiliary-like verbs have acquired aspectual meanings such as gradual change, result, completion with some later effects, and attempt, but the last three have the meanings of benefactivity (giving or receiving of an action).

Although morphologically, the possibility of inserting focus particles like sae ‘even’ denies the compound status of V1-te V2 concatenations, syntactically they can be regarded as coherent units of predicates representing the occurrence of complex events. Their syntactic unity can be justified by the fact that the first member (V1-te) can be neither moved away from the auxiliary-like verbs nor deleted leaving only the V2s behind. This functional unity is often strengthened by the phonological unity, where V1-te and V2 are coalesced in colloquial speech, as in V-te simau → V-tyau and V-te oku → V-toku. The V1-te V2 concatenations thus bear a great resemblance to the syntactic V-V compound verbs discussed in section 2.2. Historically, it is known that the -te forms developed later than V-infinitive V compound verbs, but it remains an open question what kinds of factors were responsible for the bifurcation of the two functionally similar classes.

Finally, V-te forms may be used with motion verbs iku ‘go,’ kuru ‘come,’ and kaeru ‘return’ to form complex predicates like mot-te kuru [have-GER come] ‘bring,’ koi-de iku [row-GER go] ‘go by rowing,’ and ture-te kaeru [accompany-GER return] ‘bring/take home’ (see Shibatani, 2007 for details). Some of these may alternate with lexical V-V compound verbs (section 2.3), as in ture-te kaeru [accompany-GER return] and ture-kaeru [accompanyinfinitive-return], with the stylistic difference that the compound version sounds slightly archaic.

This section has surveyed the inventory of Japanese compound predicates that can conclude a finite sentence in their tensed inflectional forms. Among the various types, productivity varies significantly depending on whether the first member is a verb or a noun. V-V compounds are highly productive, while N-V and N-A compounds are very much less so. It remains to be seen why the difference in the lexical categories of first members brings about such an asymmetry in productivity. One factor may be that Japanese has an agglutinative character in the domain of predicates, producing a conglomeration of verb, aspect, tense, modality, and politeness, while lacking integration by agglutination in the realm of noun phrases (for example, Japanese case particles, unlike the Korean counterparts, are not morphologically integrated with the preceding nouns: Kageyama, 2016c). The next section will introduce tenseless compound predicates used in non-finite clauses, as opposed to the tensed compounds discussed in section 2.

3. Morphologically Loose Compound Predicates

Japanese has two special classes of lexical categories that do not directly inflect for tense and yet have argument structure and case. The two special classes, originating primarily from the Sino-Japanese vocabulary, are (i) Verbal Nouns (VN) such as koonyuu ‘purchase’ (transitive), sanpo ‘take a walk’ (unergative), and hassei ‘occur’ (unaccusative), and (ii) Adjectival Nouns (AN) such as tokuyuu ‘peculiar’ and huzyuubun ‘inadequate.’ When put in certain non-finite constructions, these categories behave as full-fledged predicates equivalent to native verbs and adjectives and undergo compounding with their preceding nouns. As contrasted with the tight compounds involving native verbs/adjectives, Sino-Japanese predicates can give rise to morphologically loose compounds, which will be introduced in the following order:

  • Post-syntactic compounds in subordinate clauses (section 3.1);

  • Post-syntactic compounds in noun phrase structure (section 3.2);

  • Agent (external argument) compounds (section 3.3); and

  • N-AN compounds (section 3.4).

3.1 Post-Syntactic Compounds in Subordinate Clauses

Shibatani and Kageyama (1988) discovered a peculiar class of compounds, named post-syntactic compounds, whose formation is best located at a late stage of syntactic derivation. Post-syntactic compounds always have corresponding clausal structures, as exemplified in (16a) and (16b).

(16)

Compound and Complex Predicates in Japanese

Tests for lexical integrity such as the absence of a case marker after the noun and the impossibility of inserting adverbials inside show that the elements in the brackets of (16b) make up a compound word, where the colon (:) stands for the characteristic phonological break observed with loose compounds in general (section 1).

Their derivation in syntax is confirmed by the structural conditions imposed on the types of predicates and the grammatical relations of the compounded nouns. As demonstrated by Kageyama (1993), only the direct internal argument of a predicate (VN) is susceptible to compounding. More concretely, only three types of arguments are qualified: the direct object of a transitive VN (17a), the subject of an unaccusative VN (17b), and the goal/locative/source complement of motion VN (17c).

(17)

Compound and Complex Predicates in Japanese

The direct-internal argument restriction automatically rules out combinations of a transitive VN and its subject (18b) as well as combinations of any type of VN and an adjunct (18b).

(18)

Compound and Complex Predicates in Japanese

The syntactic nature of these loose compounds is reinforced by the optional appearance of the honorific prefix on a VN.

(19)

Compound and Complex Predicates in Japanese

The honorific prefix is normally attached to a predicate only in syntax and never appears inside morphologically tight compounds.

The essential feature of post-syntactic compounds is that they are tenseless, despite the fact that they may case-mark their arguments. Addition of tense by light verbs (suru ‘do,’ sita ‘did,’ site iru ‘be doing’) to the examples in (17) and (19) renders them ungrammatical. As long as the grammatical conditions on argument relations and tense are satisfied, post-syntactic compounding is fully productive.

3.2 Post-Syntactic Compounds in Noun Phrases

Right after the discovery of post-syntactic compounds in tenseless subordinate clauses (Shibatani & Kageyama, 1988), parallel phenomena were observed in noun phrase structures as well (Kageyama & Shibatani, 1989; Kageyama, 1993). Compare the subordinate clause structures discussed in (17) with the noun phrase counterparts in (20).

(20)

Compound and Complex Predicates in Japanese

In (20), the nouns in the brackets, i.e., ‘adventure film,’ ‘earthquake,’ and ‘London,’ are marked in the genitive if they appear in NP structures. The impossibility of compounding an agent (external argument) and an adjunct in subordinate clauses in (18) is mirrored by the ungrammaticality of the same compounding in NP structures in (21).

(21)

Compound and Complex Predicates in Japanese

The honorific prefix on VNs as in (19) is also acceptable in the post-syntactic compounding in NP structure.

3.3 Agent Compounding

As evidenced by the ungrammaticality of (18a) and (21a), agent arguments (or external arguments) are strictly prohibited from participating in post-syntactic compounding. The exclusion of agent (or external) arguments has in fact been held as a hallmark of verbal compounding in English (Selkirk, 1982) and noun incorporation in the world’s languages (Mithun, 1984; Baker, 1988). This is a natural consequence of the fact that an external argument is structurally distanced from the predicate in both argument structure and syntactic structure. Kageyama (2006), however, brought to light a class of systematic exceptions, where VNs are compounded with their external arguments to form loose compounds in breach of the universal ban on agent incorporation. The nature of agent compounding will be best appreciated by contrasting the fully grammatical expressions in the brackets of (22) with the ungrammatical post-syntactic compounds given earlier in (18a) and (21a).

(22)

Compound and Complex Predicates in Japanese

In (22a), ‘Spielberg’ is the agent (external argument) of ‘produce,’ while ‘adventure film’ is its Theme argument. Likewise, in (22b), ‘a female pilot’ is the agent and ‘the plane’ is the Theme. The tenselessness condition, however, holds for agent compounding as well, as it is impossible to add the light suru ‘do’ to the VNs in (22).

The prominent feature of such agent-containing compounds is that they function as modifiers of the Theme nouns to characterize their noteworthy properties. Thus, the adventure film in (22a) is noteworthy because it is a Spielberg production, and the plane in (22b) deserves attention just because its pilot is female. In other words, such agent compounds present “individual-level predications” as opposed to the “stage-level predications” that are associated with post-syntactic compounds. It is thus inappropriate to add to (22) a temporal adverbial pinpointing a particular time or an expression indicating the ongoing progress of an event.

(23)

Compound and Complex Predicates in Japanese

Kageyama (2006, in press) further discusses the differences between (22a) and (22b), observing that only the former is amenable to paraphrase by the Topic-predicate construction as in Sono eiga wa [Supirubaagu : seisaku] da ‘The film is such that Spielberg produced it.’ According to Kageyama, individual-level predications are divided into two types, one characterizing an inherent attribute of the modified noun (Theme) that will remain permanently, as in (22a), and one describing an acquired characteristic of the Theme that will hold temporarily, as in (22b). Only those agent compounds that represent the former type of individual-level predication can appear in the Topic-predicate construction.

3.4 N-AN Compounds

Just as Verbal Nouns have a dual function of verb and noun, Adjectival Nouns (ANs) have a dual function of adjective and noun and may take a subject argument and optionally a dative complement. Namiki (1988), Yumoto (1990), and Kageyama (1993) observe that a compounding rule analogous to post-syntactic compounding takes place in the constructions illustrated in (24).

(24)

Compound and Complex Predicates in Japanese

Like post-syntactic compounds, these N-AN compounds are tenseless and are pronounced with the characteristic phonological break inside. In their corresponding syntactic constructions, the noun members correspond to the AN’s internal arguments represented by the nominative ga or the dative ni. This compounding is also fairly productive.

3.5 Characteristics of Loose Compound Predicates

By way of summary, major differences observed between tight and loose compound predicates are listed in Table 6.

Table 6. Tight and Loose Compound Predicates in Contrast

Tight N-V Compounds

Loose N-Pred Compounds

internal composition

Predicate-Predicate, Noun-Predicate

Noun-Predicate

lexical strata

native

Sino-Japanese

tense and other inflections

Yes.

No.

productivity

Pred-Pred is high;

Noun-Predicate is low.

High

honorification on Pred

No.

Yes.

Two additional comments are in order concerning the compounds of Noun-Predicate composition. One is that while the N member of the tight N-V compound verbs is strictly limited to nouns or bound nominal morphemes, that of loose compound predicates may sometimes be expanded to short nominal phrases such as sono gengo ‘that language’ instead of Nihongo in (24a) or Amerika-sei no sinsya ‘a new car made in America’ instead of sinsya ‘new car’ in (16b). Just because the noun members can be expanded to phrase-like categories, however, does not mean that the loose compounds are not really compounds or are due to pseudo-noun incorporation in the sense of Massam (2001), because they obey other tests for lexical integrity. In fact, expansion of a noun in modifier position to a phrasal category is attested outside the class of loose compounds as well (Kageyama, 2016c).

The other difference has to do with finiteness/non-finiteness. In contradistinction to the tight N-V compound verbs, which carry tense inflections, the VNs and ANs used in loose compound predicates are tenseless and reject manifestation of tense by suru support. The tenseless or non-finite nature of VNs and ANs appears to be tied in with their high degree of productivity. There is thus a general tendency in Japanese that tensed N-V compound verbs are hardly productive while tenseless N-Pred compounds are extremely productive. Kageyama (2016b) further elaborates on the relation between the finite/non-finite distinction and productivity, proposing a non-finiteness continuum where the tensed and tenseless compound types occupy the two poles and a new class of “semi-finite” compounds is situated in between. The semi-finite compounds include N-V compounds whose V member takes non-finite inflections, as illustrated in (25).

(25)

Compound and Complex Predicates in Japanese

Adnominal and converb (-te) inflections are semi-finite because they are not completely nominal like VNs and ANs but are not capable of concluding a sentence, either. Interestingly, these semi-finite compounds appear to contain a slight pause inside, though not so conspicuously as the Sino-Japanese loose compounds. Such converb and adnominal compounds have been neglected in the literature and call for detailed investigation.

4. Syntactically Complex Predicates

Following tight and loose compound predicates, this section presents a bird’s-eye view of syntactically complex predicates that are not qualified as morphological compounds but nonetheless display word-like behavior in rejecting syntactic movement of their constituents. These complex predicates take the form of an indefinite/non-referential element + suru ‘do’/aru ‘be.’

4.1 The Light Verb Constructions

Following Grimshaw and Mester (1988), the term “light verb construction” in Japanese linguistics refers to sentences like (26).

(26)

Compound and Complex Predicates in Japanese

In this construction, the syntactic sequence of a VN (e.g., aiseki ‘to share a table (with a stranger)’ followed by the verb suru ‘do’ serves as a complex predicate. Suru ‘do’ in this construction is called a “light verb” because its sole function is to carry tense; it does not convey its original meaning any more because it can take not only VNs of volitional action but also VNs depicting non-volitional events and states such as hassei ‘occur’ and sonzai ‘exist.’

Among its numerous theoretically interesting properties, two deserve special mention here. First, the VN, for example aiseki in (26), has an argument structure that determines the argument relations of the whole sentence. Since aiseki is intransitive, the whole construction is intransitive, with the partner argument (‘a traveler’) represented with the comitative particle to. While Grimshaw and Mester (1988) assumed that the accusative marker on the VN is automatically assigned by the light verb, Dubinsky (1989), Miyagawa (1989a), Tsujimura (1990), and Kageyama (1991) independently discovered that the possibility of accusative marking on VNs hinges on the argument structure types of individual VNs—only VNs of volitional action may take an accusative marker, while non-volitonal VNs, in particular unaccusative VNs, cannot be marked in the accusative (cf. the Unaccusativity Hypothesis). This characterization, however, raised a problem with transitive VNs, which, though volitional, are generally judged incompatible with the accusative marker on them. Compare the two sentences in (27), one with an accusative on the VN kenkyuu ‘research’ and the other without.

(27)

Compound and Complex Predicates in Japanese

While it is generally considered that the ungrammaticality of (27a) stems from the occurrence of two accusative markers in a single sentence (the so-called “double accusative constraint”), Sells (1989) argued on the basis of scrambling and topicalization that the double accusative construction is theoretically possible in Japanese, and Kageyama (1993) further observed that the phenomenon is empirically attested in certain formal statements like politicians’ speeches, as in (28).

(28)

Compound and Complex Predicates in Japanese

Such examples show that the double accusative constraint is a superficial restriction dependent on style.

Leaving aside numerous issues involving this construction (see Miyamoto & Kishimoto, 2016 for an overview), the issue that is directly relevant to this article is this: Why do the VNs, regardless of the presence or absence of the accusative marker, resist syntactic movement such as clefting and topicalization?

(29)

Compound and Complex Predicates in Japanese

As contrasted with (29a), where the referential NP ‘endangered language’ is focused, (29b) is totally unacceptable. Moreover, the VN cannot be modified by a determiner as in (29c), regardless of the presence or absence of the accusative marking. These observations indicate that the VNs in the light verb constructions are non-referential.

The non-referentiality of the VNs will be accounted for by assuming that the VN o in the light verb construction is semantically incorporated with suru to constitute a functional unit as a complex predicate, as shown schematically in (30) (Kageyama, 1993).

(30)

Compound and Complex Predicates in Japanese

The semantic incorporation has a morphological counterpart in (31), where the accusative marker is dropped and the VN is morphologically combined with suru.

(31)

Compound and Complex Predicates in Japanese

In the literature, the morphological status of the VN suru sequence (without the accusative marker) has been a longstanding issue. Matsumoto (1996) regards it as a syntactic rather than morphological sequence, and Booij (2010) identifies it as quasi noun incorporation, while Kageyama (2016c) maintains that VN suru forms a morphological word.

4.2 Physical Attribute Construction

The verb suru ‘do’ has another special usage in what we might call the physical attribute construction, as schematically shown in (32) with an example sentence.

(32)

Compound and Complex Predicates in Japanese

As contrasted with its regular usage as a volitional activity verb, the verb suru ‘do’ in this construction expresses a stable property of the subject with the aid of the resultative auxiliary -te iru in concluding a sentence, as in (32), or the resultative -ta inflection in prenominal position, as in aoi me o si-ta onnanoko ‘a girl with blue eyes.’

In addition to the humanness requirement of the subject, this construction imposes certain special conditions on the object NPs marked in the accusative. First, they must take the syntactic form of “adjective + noun,” where the noun designates a physical, psychological, or intellectual attribute inherent in a human being and the adjective expresses its current state, as in aoi me ‘blue eyes,’ hosoi yubi ‘slender finger,’ utukusii koe ‘beautiful voice,’ and yookina seikaku ‘cheerful character.’ Because of space limitation, this article will not address the question of why an adjective is syntactically mandatory before the noun (see Tsujioka (2002) and Kageyama (in press) for concrete analyses). The second prominent feature of the object NPs is that they must be indefinite and non-referential. It is thus not possible to add to them a determiner that delimits their referents, as in (33).

(33)

Compound and Complex Predicates in Japanese

Also, it is impossible to replace the whole object NP with a pronoun, as in *sore o si-te iru ‘has it/them,’ or delete it under identity with another referential NP.

The indefiniteness and non-referentiality suggested by the exclusion of specific determiners are correlated with their syntactic immobility, as in (34).

(34)

Compound and Complex Predicates in Japanese

The parallelism between (34) and the light verb construction in (29b) strongly suggests that the physical attribute construction in Japanese involves semantic incorporation, whereby an indefinite object NP is semantically connected with the verb suru to give rise to a functional unit of complex predicate.

As opposed to the light verb construction, which describes temporary (or stage-level) events and states that may change as time passes, the physical attribute construction depicts a stable (or individual-level) property of the subject person that will not be easily affected by the flow of time. Kageyama (2006, in press) shows the relevance of “property (individual-level) predication” by contrasting it with sentences like (35) that are identical in form as the physical attribute construction but represents a temporary (i.e., stage-level) property.

(35)

Compound and Complex Predicates in Japanese

Since the state of affairs represented by (35a) is volitionally controlled and holds only temporarily, semantic incorporation does not take place, as shown by the grammaticality of (35b).

4.3 Other Complex Predicate Constructions With suru ‘Do’

The verb suru ‘do’ is versatile enough to create complex predicate constructions in conjunction with a variety of elements in front of it. Besides the two constructions discussed in sections 4.1 and 4.2, two other constructions are briefly mentioned here. One is the mimetic verb construction, where suru is combined with mimetic (or sound-symbolic) words that characteristically represent sounds, voices, and manners of actions and events as perceived by human five senses (see Kageyama, 2007 for details). Two examples are given in (36).

(36)

Compound and Complex Predicates in Japanese

(36a) represents a particular banging action, and (35b) a motion event with the accusative particle designating the traversal path of motion (see Haig, 1981). It is important to note that neither example allows the mimetic words to be syntactically moved away from the verb suru or get deleted, thus indicating that the concatenations of mimetic words and suru functionally make up a predicate unit as a whole.

The other example is the reduplicated noun + suru construction, which depicts an inherent property of the subject (see Oho & Yamada, 2011 and Ono, 2015 for details).

(37)

Compound and Complex Predicates in Japanese

The syntactic immobility observed with the other constructions in this section holds true also for the reduplicated noun phrases bottyan-bottyan and gengogaku-gengogaku, which are indefinite bare nouns that cannot be modified by any determiner or quantifier or get deleted.

4.4 Human Noun + aru ‘Be’ Construction

In closing this section, we briefly touch on a peculiar construction involving a human subject NP and the existential verb aru ‘be.’ Contemporary standard Japanese has two verbs of existence, aru and iru, which are strictly distinguished by the animacy of their subject NPs: aru for inanimate nouns including not only inanimate entities but also happenings, and iru for animate nouns such as human and animal/insect nouns. Of these two, only aru exhibits peculiar behavior in the following construction.

(38)

Compound and Complex Predicates in Japanese

(38a) and (38b) are identical except that they realize the quantifiers ‘many’ and ‘one hundred’ in different positions due to the so-called “quantifier float” (see Miyagawa, 1989b). Disregarding this positional difference, then, the striking feature of these constructions is that the verb aru, generally limited to inanimate subjects, is perfectly compatible with the human noun sankasya ‘participant.’ However, the human nouns that enter into this construction must have the lexical meaning of the occurrence of a happening in addition to the biological humanness, as in zyookyaku ‘passenger,’ raizyoosya ‘attendant,’ sisya ‘fatality,’ and husyoosya ‘injured person,’ and the verb aru in this construction denotes the occurrence of a particular event involving the persons in the subject. Sentences of the type (38) are parallel to the other constructions discussed in sections 4.1, 4.2, and 4.3 in that the subject NPs are syntactically “frozen.” Thus, it is not possible to topicalize or cleft the subject NP from its original pre-verbal position or delete it under identity with another instance of the same NP, nor is it possible to add determiners like ‘the’ or ‘these’ to it. From these observations, Kageyama (2004) claims that an invisible semantic incorporation is operative on the subject and the verb aru. In an alternative approach, Kishimoto (2016) argues that the indefiniteness effects stem from the nature of existential sentences.

5. Theoretical Approaches to Compounding and Incorporation

Despite their direct relevance to syntax, multipart predicates have been largely neglected in the literature of traditional Japanese grammar. Concentrated studies on the phenomena started in the 1980s, when linguists began to direct attention to the relationship and interplay of morphology, syntax, and semantics in the grammatical architecture. Kageyama (1980) and Sugioka (1986) are among the seminal works in this area, which sparked a surge of research in the morphosyntax and lexical semantics of such multipart predicates in the 1990s up until the early 21st century, with Kageyama (1993) and Matsumoto (1996) being the cornerstones of the current research. Unfortunately, many of the important monographs and papers are not easily accessible to international readers because they are written in Japanese. The situation, however, is beginning to improve with the publication of handbooks in English, such as Tsujimura (2002), Miyagawa and Saito (2008), Kageyama and Kishimoto (2016), and Hasegawa (in press).

The compound and complex predicates discussed in this article are so complex and so diverse that they present a variety of issues that are directly relevant to the current theories of syntax, morphology, and semantics. Since various modules of grammar are involved in their formation, it can safely be said that an extreme theory that pays attention only to the lexicon or only to the syntax, as in the strong lexicalist hypothesis or the strong syntactic hypothesis in the 1970s and 1980s, is not powerful enough to account for the diversity and relatedness of these phenomena. Obviously, then, broad and multifaceted theories that allow predicate formation in various modules of grammar need to be developed, perhaps along the lines of Modular Morphology (Shibatani & Kageyama, 1988), Distributed Morphology (Halle & Marantz, 1993; Nishiyama, 2008), or Construction Morphology (Booij, 2010).

Concerning Verb-Verb complexes, the rich repertoire of Japanese V-V compound verbs and V-te V concatenations should be used to test the validity of Masica’s (1976) claim that V-V complexes are an areal feature of Asian languages. The international conference on this topic held at the National Institute for Japanese Language and Linguistics, Tokyo, in 2013 suggests that Japanese and Korean abound in V-V compound verbs, whereas languages of South and Central Asia have only V-V complexes mediated by a linker, as in the Japanese V-te V concatenations.

Noun-Predicate complexes are another major area of further research. On the morphological side, the Japanese N-V compound verbs of the tensed type seem to be different in nature from N-V complexes due to Noun Incorporation that have been extensively studied in polysynthetic and other languages (Mithun, 1984; Baker, 1988; Mathieu, 2009). The relationship between compounding and incorporation must be investigated in greater depth (cf. Mithun, 2000). On the semantic side, the Japanese complex predicates discussed in section 4 form a lacuna of research. These constructions need to be investigated in light of recent advancements on the typology, syntax, and semantics of incorporation, pseudo incorporation, indefinites, and other related phenomena (Van Geenhoven, 1998; Carlson, 2006 Borik & Gehrke, 2015, and many others).

Further Reading

Frellesvig, B., Horn, S., Russell, K., & Sells, P. (2010). Verb semantics and argument realization in pre-modern Japanese: A preliminary study of compound verbs in Old Japanese. Gengo Kenkyu, 138, 25–65.Find this resource:

    Kageyama, T. (1982). Word formation in Japanese. Lingua, 57, 215–258.Find this resource:

      Kageyama, T. (1989). The place of morphology in the grammar: Verb-Verb compounds in Japanese. In G. Booij & J. van Marle (Eds.), Yearbook of morphology (Vol. 2, pp. 73–94). Dordrecht: Foris.Find this resource:

        Kageyama, T. (1993). Bunpō to gokeisei. Tokyo: Hituzi Syobo.Find this resource:

          Kageyama, T. (1999). Word formation. In N. Tsujimura (Ed.), The handbook of Japanese linguistics (pp. 297–325). Oxford: Blackwell.Find this resource:

            Kageyama, T. (2009). Isolate: Japanese. In R. Lieber & P. Štekauer (Eds.). The Oxford handbook of compounding (pp. 512–526). Oxford: Oxford University Press.Find this resource:

              Kageyama, T. (in press). Events and properties in morphology and syntax. In Y. Hasegawa (Ed.), The Cambridge handbook of Japanese linguistics. Cambridge, UK: Cambridge University Press.Find this resource:

                Kageyama, T., & Kanzaki, K. (2014). Compound verb lexicon (online dictionary). Tokyo: National Institute for Japanese Language and Linguistics.Find this resource:

                  Kageyama, T., & Kishimoto, H. (Eds.). (2016). Handbook of Japanese lexicon and word formation. Berlin: De Gruyter Mouton.Find this resource:

                    Kishimoto, H., & Booij, G. (2014). Complex negative adjectives in Japanese: The relation between syntactic and morphological constructions. Word Structure, 7, 55–87.Find this resource:

                      Matsumoto, Y. (1996). Complex predicates in Japanese: A syntactic and semantic study of the notion ‘word.’ Stanford: CSLI.Find this resource:

                        Nakatani, K. (2016). Complex predicates with -te gerundive verbs. In T. Kageyama & H. Kishimoto (Eds.), Handbook of Japanese lexicon and word formation (pp. 387–423). Berlin: De Gruyer Mouton.Find this resource:

                          Nishiyama, K. (1998). V-V compounds as serialization. Journal of East Asian Linguistics, 7, 175–217.Find this resource:

                            Nishiyama, K. (2008). V-V compound verbs. In S. Miyagawa & M. Saito (Eds.), The Oxford handbook of Japanese linguistics (pp. 320–347). Oxford: Oxford University Press.Find this resource:

                              Sugioka, Y. (1986). Interaction of derivational morphology and syntax in Japanese and English. New York: Garland.Find this resource:

                                Toratani, K. (2002). The morphosyntactic structure and logical structures of compound verbs in Japanese (Unpublished doctoral dissertation). State University of New York.Find this resource:

                                  Yumoto, Y. (2005). Fukugōdōshi haseidōshi no imi to tōgo. Tokyo: Hituzi Syobo.Find this resource:

                                    Yumoto, Y. (2009). Modularity of word formation: Differences between two types of Japanese compound verbs. In H. Hoshi (Ed.), The dynamics of the language faculty (pp. 203–230). Tokyo: Kurosio.Find this resource:

                                      References

                                      Amberber, M., Baker, B., & Harvey, M. (Eds.). (2010). Complex predicates: Cross-linguistic perspectives on event structure. Cambridge, UK: Cambridge University Press.Find this resource:

                                        Anderson, S. (1992). A-morphous morphology. Cambridge, UK: Cambridge University Press.Find this resource:

                                          Baker, M. (1988). Incorporation: A theory of grammatical function changing. Chicago: University of Chicago Press.Find this resource:

                                            Bloch, B. (1946). Studies in colloquial Japanese I: Inflection. Journal of the American Oriental Society, 66, 97–109.Find this resource:

                                              Booij, G. (2010). Construction morphology. Oxford: Oxford University Press.Find this resource:

                                                Borik, O., & Gehrke, B. (Eds.) (2015). The syntax and semantics of pseudo-incorporation. Leiden, The Netherlands: Brill.Find this resource:

                                                  Carlson, G. (2006). The meaningful bounds of incorporation. In S. Vogeleer & L. Tasmowski (Eds.), Non-definiteness and plurality (pp. 35–50). Amsterdam: Benjamins.Find this resource:

                                                    Dubinsky, S. (1989). Compound suru verbs and evidence for unaccusativity in Japanese. In C. Wiltshire, R. Graczyk, & B. Music (Eds.), Papers from the 25th Annual Regional Meeting of the Chicago Linguistic Society. Part one, The general session (pp. 98–111). Chicago: The Society.Find this resource:

                                                      Fukuda, S. (2012). Aspectual verbs as functional heads: Evidence from Japanese aspectual verbs. Natural Language & Linguistic Theory, 30(4), 965–1026.Find this resource:

                                                        Fukushima, K. (2005). Lexical V-V compounds in Japanese: Lexicon vs. syntax. Language, 81, 568–612.Find this resource:

                                                          Gamerschlag, T. (2001). Complex predicate formation and argument structure of Japanese V-V compounds. In N. M. Akatsuka & S. Strauss (Eds.), Japanese/Korean linguistics 10 (pp. 532–544). Stanford: CSLI.Find this resource:

                                                            Grimshaw, J., & Mester, A. (1988). Light verbs and θ‎-marking. Linguistic Inquiry, 19, 205–232.Find this resource:

                                                              Haig, J. (1981). Are traversal objects objects? Papers in Linguistics: International Journal of Human Communication, 14(1), 69–101.Find this resource:

                                                                Halle, M., & Marantz, A. (1993). Distributed Morphology and pieces of inflection. In K. Hale & S. J. Keyser (Eds.), The view from building 20: Essays in linguistics in honor of Sylvain Bromberger (pp. 111–176). Cambridge, MA: MIT Press.Find this resource:

                                                                  Hasegawa, Y. (1996). A study of Japanese clause linkage. Stanford: CSLI.Find this resource:

                                                                    Hasegawa, Y. (Ed.). (in press). The Cambridge handbook of Japanese linguistics. Cambridge, UK: Cambridge University Press.Find this resource:

                                                                      Hayatsu, E. (1989). Yūtsui-tadōshi to mutsui-tadōshi no chigai nitsuite. Gengo Kenkyu, 95, 231–256.Find this resource:

                                                                        Himeno, M. (1999) Fukugōdōshi no kōzō to imi-yōhō. Tokyo: Hituzi Syobo.Find this resource:

                                                                          Ishii, M. (1983). Gendaigo fukugōdōshi niokeru ichi-kanten. Nihongogaku, 2(8), 79–90.Find this resource:

                                                                            Kageyama, T. (1980). Goi no kōzō. Tokyo: Shōhakusha.Find this resource:

                                                                              Kageyama, T. (1989). The place of morphology in the grammar: Verb-Verb compounds in Japanese. In G. Booij & J. van Marle (Eds.), Yearbook of morphology (Vol. 2, pp. 73–94). Dordrecht: Foris.Find this resource:

                                                                                Kageyama, T. (1991). Light verb constructions and the syntax-morphology interface. In H. Nakajima (Ed.), Current English linguistics in Japan (pp. 169–203). Berlin: Mouton de Gruyter.Find this resource:

                                                                                  Kageyama, T. (1993). Bunpō to gokeisei. Tokyo: Hituzi Syobo.Find this resource:

                                                                                    Kageyama, T. (1999). Word formation. In N. Tsujimura (Ed.), The handbook of Japanese linguistics (pp. 297–325). Oxford: Blackwell.Find this resource:

                                                                                      Kageyama, T. (2001). Word plus: The intersection of words and phrases. In J. van de Weijer & T. Nishihara (Eds.), Issues in Japanese phonology and morphology (pp. 245–276). Berlin: Mouton de Gruyter.Find this resource:

                                                                                        Kageyama, T. (2004). Sonzai/syoyū no keidōsi-kōbun to imi-hennyū. In T. Kageyama & H. Kishimoto (Eds.), Nihongo no bunseki to gengo-ruikei (pp. 3–23). Tokyo: Kurosio.Find this resource:

                                                                                          Kageyama, T. (2006). Gaikō-fukugōgo to jojutsu no taipu. In T. Masuoka, H. Noda, & T. Moriyama (Eds.) Nihongo-bunpō no shinchihei I (pp. 1–21). Tokyo: Kurosio.Find this resource:

                                                                                            Kageyama, T. (2007). Explorations in the conceptual semantics of mimetic verbs. In B. Frellesvig, M. Shibatani, & J. C. Smith (Eds.), Current issues in the history and structures of Japanese (pp. 27–82). Tokyo: Kurosio.Find this resource:

                                                                                              Kageyama, T. (2013) Post Post-syntactic compounds and semantic head-marking in Japanese. In B. Frellesvig & P. Sells (Eds.), Japanese/Korean linguistics (Vol. 20, pp. 363–382). Stanford: CSLI.Find this resource:

                                                                                                Kageyama, T. (2016a). Noun-compounding and noun-incorporation. In T. Kageyama & H. Kishimoto (Eds.), Handbook of Japanese lexicon and word formation (pp. 237–272). Berlin: De Gruyter Mouton.Find this resource:

                                                                                                  Kageyama, T. (2016b). Verb-compounding and verb-incorporation. In T. Kageyama & H. Kishimoto (Eds.), Handbook of Japanese lexicon and word formation (pp. 273–310). Berlin: De Gruyter Mouton.Find this resource:

                                                                                                    Kageyama, T. (2016c). Lexical integrity and the morphology-syntax interface. In T. Kageyama & H. Kishimoto (Eds.), Handbook of Japanese lexicon and word formation (pp. 489–528). Berlin: De Gruyter Mouton.Find this resource:

                                                                                                      Kageyama, T. (2016d). Agents in anticausative and decausative compound verbs. In T. Kageyama & W. M. Jacobsen (Eds.), Transitivity and valency alternations: Studies on Japanese and beyond (pp. 89–124). Berlin: De Gruyter Mouton.Find this resource:

                                                                                                        Kageyama, T. (in press). Events and properties in morphology and syntax. In Y. Hasegawa (Ed.), The Cambridge handbook of Japanese linguistics. Cambridge, UK: Cambridge University Press.Find this resource:

                                                                                                          Kageyama, T., & Shibatani, M. (1989). Meishiku kara no fukugōgo keisei. In S. Kuno & M. Shibatani (Eds.), Nihongogaku no shintenkai (pp. 139–166). Tokyo: Kurosio.Find this resource:

                                                                                                            Kageyama, T. & Kanzaki, K. (2014). Compound verb lexicon (online dictionary). Tokyo: National Institute for Japanese Language and Linguistics.Find this resource:

                                                                                                              Kageyama, T., & Kishimoto, H. (2016). Handbook of Japanese lexicon and word formation. Berlin: De Gruyter Mouton.Find this resource:

                                                                                                                Kishimoto, H. (2013). Tōgoteki fukugōdōshi no kaku to tōgososei. In T. Kageyama (Ed.), Fukugōdōshi kenkyū no saizensen (pp. 143–183). Tokyo: Hituzi Syobo.Find this resource:

                                                                                                                  Kishimoto, H., & Booij, G. (2014). Complex negative adjectives in Japanese: The relation between syntactic and morphological constructions. Word Structure, 7, 55–87.Find this resource:

                                                                                                                    Kishimoto, H. (2016). Stative and existential/possessive predicates. In T. Kageyama & H. Kishimoto (Eds.), Handbook of Japanese lexicon and word formation (pp. 559–598). Berlin: De Gruyer Mouton.Find this resource:

                                                                                                                      Kinuhata, T. (2010). Jōdaigo no meishi-hōgō nituite. Gobun, 92/93, 34–44.Find this resource:

                                                                                                                        Kubozono, H. (2016). Accent in Japanese phonology. Oxford Research Encyclopedia of Linguistics.

                                                                                                                        Lieber, R. (1983). Argument linking and compounds in English. Linguistic Inquiry, 14, 251–286.Find this resource:

                                                                                                                          Martin, S. (1975). A reference grammar of Japanese. New Haven: Yale University Press.Find this resource:

                                                                                                                            Masica, C. P. (1976). Defining a linguistic area: South Asia. Chicago: University of Chicago Press.Find this resource:

                                                                                                                              Massam, D. (2001). Pseudo noun incorporation in Niuean. Natural Language & Linguistic Theory, 19, 153–197.Find this resource:

                                                                                                                                Mathieu, E. (Ed.). (2009). Noun incorporation and its kind [Special issue]. Lingua, 119.Find this resource:

                                                                                                                                  Matsumoto, Y. (1996). Complex predicates in Japanese: A syntactic and semantic study of the notion ‘word.’ Stanford: CSLI.Find this resource:

                                                                                                                                    Matsumoto, Y. (1998). Nihongo no goiteki-fukugōdōshi niokeru dōshi no kumiawase. Gengo Kenkyu, 114, 37–83.Find this resource:

                                                                                                                                      Mithun, M. (1984). The evolution of noun incorporation. Language, 60, 847–894.Find this resource:

                                                                                                                                        Mithun, M. (2000). Constraints on compounding and incorporation. In I. Vogel & S. Scalise (Eds.), Compounding (pp. 37–56). Amsterdam: John Benjamins.Find this resource:

                                                                                                                                          Miyagawa, S. (1989a). Light verbs and the ergative hypothesis. Linguistic Inquiry, 20, 659–668.Find this resource:

                                                                                                                                            Miyagawa, S. (1989b). Structure and case marking in Japanese. San Diego: Academic Press.Find this resource:

                                                                                                                                              Miyagawa, S., & Saito, M. (Eds.). (2008). The Oxford handbook of Japanese linguistics. Oxford: Oxford University Press.Find this resource:

                                                                                                                                                Miyamoto, T., & Kishimoto, H. (2016). Light verb constructions with verbal nouns. In T. Kageyama & H. Kishimoto (Eds.), Handbook of Japanese lexicon and word formation (pp. 425–457). Berlin: De Gruyter Mouton.Find this resource:

                                                                                                                                                  Nakatani, K. (2016). Complex predicates with -te gerundive verbs. In T. Kageyama & H. Kishimoto (Eds.), Handbook of Japanese lexicon and word formation (pp. 387–423). Berlin: De Gruyer Mouton.Find this resource:

                                                                                                                                                    Namiki, T. (1988). Fukugōgo no nichi-ei taishō. Nihongogaku, 7(5), 68–78. Tokyo: Meiji Shoin.Find this resource:

                                                                                                                                                      Nash, L., & Samvelian P. (Eds.). (2016). Approaches to complex predicates. Leiden, The Netherlands: Brill.Find this resource:

                                                                                                                                                        Naumann, R., & Gamerschlag, T. (2003). Constraining the combinatorial patterns of Japanese V-V compounds: An analysis in dynamic event semantics. Journal of Semantics, 20(3), 275–296.Find this resource:

                                                                                                                                                          Nishigauchi, T. (1993). Long distance passive. In N. Hasegawa (Ed.), Japanese syntax in comparative grammar (pp. 79–114). Tokyo: Kurosio.Find this resource:

                                                                                                                                                            Nishiyama, K. (1998a). The morphosyntax and morphophonology of Japanese predicates (Unpublished doctoral dissertation). Cornell University, Ithaca, NY.Find this resource:

                                                                                                                                                              Nishiyama, K. (1998b). V-V compounds as serialization. Journal of East Asian Linguistics, 7, 175–217.Find this resource:

                                                                                                                                                                Nishiyama, K. (2008). V-V compound verbs. In S. Miyagawa & M. Saito (Eds.), The Oxford handbook of Japanese linguistics (pp. 320–347). Oxford: Oxford University Press.Find this resource:

                                                                                                                                                                  Nolan, B., & Diedrichsen, E. (Eds.). (2017). Argument realization in complex predicates and complex events: Verb-verb constructions at the syntax-semantics interface. Amsterdam: John Benjamins.Find this resource:

                                                                                                                                                                    Oho, A., & Yamada, M. (2011). A Japanese salad-salad paper. Paper presented at GLOW in Asia Workshop for Young Researchers, Mie University.Find this resource:

                                                                                                                                                                      Ono, N. (2015). Kōbunteki chōfukugo keisei. In Y. Yumoto & N. Ono (Eds.), Goi-imiron no aratana kanōsei o sagutte (pp. 463–489). Tokyo: Kaitakusha.Find this resource:

                                                                                                                                                                        Poser, W. (1990). Word-internal phrase boundary in Japanese. In S. Inkelas & D. Zec (Eds.), The phonology-syntax connection (pp. 279–288). Chicago: University of Chicago Press.Find this resource:

                                                                                                                                                                          Sakakura, A. (1966). Gokōsei no kenkyū. Tokyo: Kadokawa Shoten.Find this resource:

                                                                                                                                                                            Selkirk, E. O. (1982). The syntax of words. Cambridge, MA: MIT Press.Find this resource:

                                                                                                                                                                              Sells, P. (1989). More on light verbs and theta marking. Unpublished paper, Stanford University.Find this resource:

                                                                                                                                                                                Sells, P. (1995). Korean and Japanese morphology from a lexical perspective. Linguistic Inquiry, 26, 237–325.Find this resource:

                                                                                                                                                                                  Shibatani, M. (2007). Grammaticalization of motion verbs. In B. Frellesvig, M. Shibatani, & J. C. Smith (Eds.), Current issues in the history and structure of Japanese (pp. 107–133). Tokyo: Kurosio.Find this resource:

                                                                                                                                                                                    Shibatani, M., & Kageyama, T. (1988). Word formation in a modular theory of grammar: Post-syntactic compounds in Japanese. Language, 64, 451–484.Find this resource:

                                                                                                                                                                                      Tamaoka, K., Lim, H., & Sakai, H. (2004). Entropy and redundancy of Japanese lexical and syntactic compound verbs. Journal of Quantitative Linguistics, 11(3), 233–250.Find this resource:

                                                                                                                                                                                        Toratani, K. (2002). The morphosyntactic structure and logical structures of compound verbs in Japanese (Unpublished doctoral dissertation). State University of New York at Buffalo.Find this resource:

                                                                                                                                                                                          Tsujimura, N. (1990). Ergativity of nouns and case assignment. Linguistic Inquiry, 21, 277–287.Find this resource:

                                                                                                                                                                                            Tsujimura, N. (Ed.). (2002). The handbook of Japanese linguistics. New York: John Wiley & Sons.Find this resource:

                                                                                                                                                                                              Tsujioka, T. (2002). The syntax of possession in Japanese. New York: Routledge.Find this resource:

                                                                                                                                                                                                van Geenhoven, V. (1998). Semantic incorporation and indefinite descriptions: Semantic and syntactic aspects of noun incorporation in West Greenlandic. Stanford: CSLI.Find this resource:

                                                                                                                                                                                                  Vance, T. (2001). Semantic bifurcation in Japanese compound verbs. In N. M. Akatsuka & S. Strauss (Eds.), Japanese/Korean Linguistics 10 (pp. 365–377). Stanford: CSLI.Find this resource:

                                                                                                                                                                                                    Vance, T. (2016). Rendaku or sequential voicing in Japanese phonology. Oxford Research Encyclopedia of Linguistics.

                                                                                                                                                                                                    Watanabe, A. (1999). Remarks on head movement within VP shell. In Nanzan GLOW Local Committee, Proceedings of the Nanzan GLOW 1999 (pp. 461–485). Nagoya: Nanzan University.Find this resource:

                                                                                                                                                                                                      Yumoto, Y. (1990). Nichi-ei taishō fukugō-keiyōshi no kōzō. Gengobunka Kenkyū, 16, 353–370. Toyonaka: Osaka University.Find this resource:

                                                                                                                                                                                                        Yumoto, Y. (2005). Fukugōdōshi haseidōshi no imi to tōgo. Tokyo: Hituzi Syobo.Find this resource:

                                                                                                                                                                                                          Yumoto, Y. (2009). Modularity of word formation: Differences between two types of Japanese compound verbs. In H. Hoshi (Ed.), The dynamics of the language faculty (pp. 203–230). Tokyo: Kurosio.Find this resource: