We investigate the syntax of verb-echo answers in Japanese. We first present two arguments showing that this answer form is best analyzed through overt V-to-T-to-C movement, followed by TP-ellipsis. We further show that verb-echo answers exhibit a scope reversal effect: the otherwise robust wide scope reading of focus-marked phrases with respect to negation is reversed in this construction, a pattern that holds across all grammatical positions. This ubiquitous scope reversal pattern indicates that certain instances of head movement in Japanese are syntactic, contrary to the view (Chomsky 2000, 2001) that head move-ment is to be relegated to the postsyntactic, phonological component.

1 Introduction

The theoretical status of head movement, a mainstay of generative theoretical syntax in the 1980s (Travis 1984, Baker 1988, Pollock 1989), has been the subject of considerable debate within the more recent Minimalist literature. The debate has been framed with particular reference to the mechanism and locus of head movement in the overall grammatical architecture as well as its interpretive consequences, with some researchers (e.g., Lechner 2007, 2017, Roberts 2010, Hartman 2011, Iatridou and Zeijlstra 2013, Keine and Bhatt 2016) continuing to take head movement as a syntactic operation and others (e.g., Chomsky 2000, 2001, Boeckx and Stjepanović 2001, Hale and Keyser 2002, Harley 2004) proposing that it is better analyzed as occurring in the postsyntactic, phonological component. One of the central motivations for the PF theory of head movement is that this operation is generally considered to lack appreciable semantic effects, unlike phrasal movements (particularly Ā-movements). This issue becomes even more challenging when one attempts to contribute to the ongoing debate from the perspective of head-final languages such as Japanese. Not only must one demonstrate that some cases of head movement, if any, are associated with interpretive outcomes, one must also show that head movement exists elsewhere in Japanese, to begin with. However, the latter task has proven one of the most difficult challenges facing researchers of Japanese syntax since the effect of such an operation is string-vacuous due to the head-final character of the language. Thus, Otani and Whitman (1991) and Koizumi (2000) have argued for the existence of string-vacuous head movement in Japanese on the basis of null object constructions and coordination, respectively. Despite the empirically rich discoveries made by these authors, their conclusion has been questioned and disputed in much subsequent work, particularly in Hoji 1998 and Fukui and Sakai 2003 (see also Fukui and Takano 1998, Sakai 1998, Takano 2004, Koopman 2005, Aoyagi 2006).

A notable contribution to this debate in Japanese syntax, however, has been made by Hayashi and Fujii (2015), who investigate the V-te moraw construction, illustrated in (1), in favor of their claim that this construction involves syntactic head movement.

(1)

  • Taroo-wa [Ziroo-ni piza-o tukut-te] morat-ta.

  • Taro-TOP Jiro-DAT pizza-ACC cook-CONT get-PAST

  • ‘Taro had Jiro cook pizza.’

  • (adapted from Hayashi and Fujii 2015:33)

In this example, the bracketed constituent, marked with the renyookei/continuative conjugation in traditional Japanese grammar, forms a syntactic complement with the matrix predicate moraw ‘to get’. Hayashi and Fujii draw on various data concerning the immobility of continuative-marked complements, the distribution of genitive marking within this construction, and the adverb-inclusive interpretation to show that the verbal head of the embedded complement clause—V-te in the T position, to be exact—undergoes head movement into the matrix verb in the manner depicted in (2).

(2)

graphic

Against this backdrop, the first goal of this article is to present additional evidence for head movement in Japanese from the previously unexplored perspective of verb-echo answers (VEAs), thereby further supporting the same position argued for by Hayashi and Fujii. In section 2, we will present two pieces of evidence, based on the adverb-inclusive interpretation and the impossibility of voice mismatches, to show that VEAs in Japanese are best analyzed as the consequence of string-vacuous V-to-T-to-C movement, followed by TP-ellipsis in the PF component. The other goal of this article is to use this finding as a critical probe into the locus of head movement in the generative grammatical architecture. In section 3, we present the hitherto unnoticed generalization that the otherwise straightforward surface-rigid scope interpretation between negation and various expressions marked with the disjunctive marker ka ‘or’ and the association-with-focus marker dake ‘only’ is reversed in the context of VEAs, a pattern that cuts across all grammatical positions, including subjects, direct objects, and indirect objects. We show that this ubiquitous scope reversal effect constitutes powerful support for the view that head movement in Japanese is associated with robust interpretive outcomes. Our observation, we show, has the significant implication that head movement per se is optional in Japanese, but that once it takes place, it comes with semantic consequences. Given these findings, we conclude that certain instances of head movement in Japanese are syntactic. Section 4 concludes the article.

2 VEAs in Japanese and String-Vacuous V-to-T-to-C Movement

VEAs are a type of fragmentary answer to polarity questions in which the verb is repeated. Typologically, VEAs are a widespread form of answer to a yes/no question, as documented in Holmberg’s (2016:64–72) large-scale crosslinguistic survey. Japanese, for one, allows a VEA to a yes/no question, as shown in (3A1). The other alternative is an answer form based on polarity particles, as shown in (3A2).

(3)

  • Q: Moo tuki-masi-ta-ka?

  • already arrive-POL-PAST-Q

  • ‘Did you arrive already?’

  • A1: Tuki-masi-ta-yo. (VEA)

  • arrive-POL-PAST-PRT

  • Intended: ‘I arrived already.’

  • A2: Hai./Iie. (polarity-based answer)

  • yes/no

  • ‘Yes./No.’

In this article, we will adopt Holmberg’s (2016:chap. 3) clausal ellipsis theory of VEAs as our analytical framework and argue that the derivation of VEAs in Japanese involves V-to-T-to-C movement, followed by TP-ellipsis at PF, as schematically depicted in (4). To avoid clutter, we will henceforth use the standard V-to-C notation.

(4)

graphic

It has been widely acknowledged in the literature dating back to Kuroda 1965 (see also Ohso 1976, Hoji 1985, Saito 1985) that Japanese allows radical pro-drop, a rather liberal omission of grammatically required arguments. A second way to analyze the fragment answer in (3A1), then, involves pros, as illustrated in (5).

(5)

  • VEA by pro-drop

  • [TPproSUBJproOBJ V]

A third analytical option is provided by a particular ellipsis strategy found in Japanese in which arguments are deleted individually instead of by ellipsis of the whole VP (Oku 1998, Saito 2007, Takahashi 2008a,b). Under this argument ellipsis (AE) analysis, VEAs are derived as shown in (6).

(6)

  • VEA by argument ellipsis

  • [TPSUBJOBJ V]

In this section, we will present two arguments for our head movement + ellipsis analysis.

2.1 Adverb-Inclusive Interpretations under VEAs

Our first argument for our clausal ellipsis analysis over the alternative small-size ellipsis analyses (pro and AE) is concerned with the so-called adverb-inclusive interpretation (Oku 1998). It has been commonly held since Oku 1998 that adjuncts themselves cannot be elliptic in Japanese. This observation is illustrated in (7), where the second clause cannot be interpreted as ‘Nobody was mending his glove in a careful manner’.

(7)

  • Sono hi-no siai-go Taroo-wa mitto-o teineini teiresi-te-ita-kedo, that day-GEN game-after Taro-TOP mitt-ACC carefully mend-CONT-PROG.PAST-but daremo guroobu-o teiresi-te-inakatta.

  • nobody glove-ACC mend-CONT-PROG.NEG.PAST

  • ‘After the baseball game on that day, Taro was mending his mitt carefully, but nobody was mending his glove.’

Imagine a scenario in which after a baseball game, a team of nine baseball players gathered in their clubhouse; Taro, the captain of the team, was mending his mitt in a careful manner, say, to prepare for his next game, but none of the other players was mending his glove. Sentence (7) is felicitous in this scenario. Crucially, the same sentence is infelicitous in a different scenario where Taro was mending his mitt in a careful manner while (one or more of) the other players were actually mending their glove but just not in the same careful manner as Taro was mending his mitt. In short, for (7) to be felicitous, none of the players should be mending gloves. This observation then constitutes solid evidence that manner adjuncts such as teineini ‘carefully’ cannot be elided on their own.

Suppose now that VEAs in Japanese are derived through small-scale ellipses targeting a series of individual constituents through pro-drop or AE. Given Oku’s (1998) observation that adjuncts themselves cannot undergo ellipsis, this analysis would predict that the adverb-inclusive interpretation should never be available in this construction. On the other hand, if the derivation of VEAs involves TP-ellipsis, as in our big-ellipsis theory, it is predicted that such an interpretation should, in principle, be available in this context. As first pointed out by Sugimura (2012:149–150), the prediction of our analysis, not that of the pro-drop/AE analysis, is correct. Consider (8A). Here, we have specifically used the manner adjunct zibun-no yarikatade ‘in one’s own way’, reflecting Saito’s (2007) observation that AE may target locative/temporal expressions but not manner or reason adverbials.

(8)

  • Q: Hanako-wa zibun-no yarikatade toodai-ni ukat-ta-no?

  • Hanako-TOP self-GEN way University.of.Tokyo-to pass-PAST-Q

  • ‘Did Hanako pass the entrance exam to the University of Tokyo in her own way?’

  • A: Ukat-ta-yo. Bekkaku-da-yo-ne. (✓Adverb-inclusive interpretation)

  • pass-PAST-PRT special-COP-PRT-PRT

  • Lit. ‘Passed. Truly special.’

The VEA in (8A) readily allows the interpretation that Hanako passed the entrance exam to the University of Tokyo in her own way. This pattern, while mysterious under the small-scale ellipsis theory, is straightforwardly predicted under our big-ellipsis theory: the adjunct is included within the TP, which in turn undergoes ellipsis in the same way that manner adverbials such as carefully are interpreted as being included in the ellipsis site in English sluicing, another case of TP-ellipsis (Merchant 2001).

(9) John washed something carefully, but I don’t know whati [TP John washed ti carefully].

Recall that our theory of VEAs in Japanese assumes head movement as a central analytical premise for TP-ellipsis to apply to yield the verb-stranding configuration. The results presented here thus show that Japanese is equipped with string-vacuous V-to-C movement.

2.2 Voice Mismatches under VEAs

Our second argument favoring the head movement analysis of VEAs in Japanese over the alternative pro/AE-based analyses comes from voice mismatches. Merchant (2001, 2008, 2013) and Chung (2013) point out that sluicing in English, an instance of TP-ellipsis, blocks voice mismatches between the antecedent clause and the sluiced clause, as illustrated in (10).

(10) *Someone shot Ben, but I don’t know [PP by who(m)] [TP Ben was shot tpp]. (Merchant 2001:35)

Merchant (2008, 2013) argues that voice mismatches are subject to a syntactic identity condition imposed on antecedent-elliptical TP pairs. Specifically, (10) is ungrammatical because the antecedent TP does not match the elliptical TP in terms of voice features. Sugisaki (2014) observes that sluicing in Japanese also does not tolerate this mismatch, as illustrated in (11a–b). Note that the mismatch does not affect grammaticality if the part elided in (11b) is pronounced, as in (11c).

(11)

  • Dareka-ga John-o yatot-ta-rasii-ga, boku-wa dare-ga-ka

    someone-NOM John-ACC hire-PAST-seem-but I-TOP who-NOM-Q

    sira-na-i. (active antecedent → active elliptical clause)

    know-NEG-PRES

    ‘Someone seems to have hired John, but I don’t know who.’

  • *Dareka-ga John-o yatot-ta-rasii-ga, boku-wa dare-ni-ka

    someone-NOM John-ACC hire-PAST-seem-but I-TOP who-by-Q

    sira-na-i. (active antecedent → passive elliptical clause)

    know-NEG-PRES

    ‘Someone seems to have hired John, but I don’t know by whom.’

  • Dareka-ga John-o yatot-ta-rasii-ga, boku-wa dare-ni yatow-are-ta-ka

    someone-NOM John-ACC hire-PAST-seem-but I-TOP who-by hire-PASS-PAST-Q

    sira-na-i. (active antecedent → passive full clause)

    know-NEG-PRES

    ‘Someone seems to have hired John, but I don’t know by whom he was hired.’

    ((11b–c) from Sugisaki 2014:6)

Our current analysis predicts that VEAs, derived here as an instance of TP-ellipsis, should disallow voice mismatches on par with sluicing because the latter is known to obey the aforementioned syntactic identity condition on voice. This prediction is borne out. The scene-setting question in (12Q) contains an active antecedent; the VEA in (12A1) contains an active response, whereas that in (12A2) contains a passive response.

(12)

  • Q: Anata-no kaisya-wa kotosi gonin-izyoo-no gakusei-o

  • you-GEN company-TOP this.year five-more.than-GEN student-ACC

  • konede saiyoosi-masi-ta-ne? (active antecedent)

  • through.personal.connection recruit-POL-PAST-PRT

  • ‘Did your company recruit more than five students this year through personal connections?’

  • A1: Saiyoosi-masi-ta-yo. (active elliptical clause)

  • recruit-POL-PAST-PRT

  • Lit. ‘Recruited.’

  • A2: *Saiyoos-are-masi-ta-yo. (passive elliptical clause)

  • recruit-PASS-POL-PAST-PRT

  • Lit. ‘Was recruited.’

We follow Merchant’s (2008, 2013) syntactic identity–based theory of voice mismatches in English sluicing noted above and assume that the response in (12A2) is ungrammatical because the passive voice feature within its elliptical clause does not match the active voice feature of the yes/no question in (12Q). Note also that the grammatical active VEA in (12A1) allows the adverb-inclusive interpretation whereby the company in question recruited more than five students through its personal connections this year. Given our discussion in section 2.1, the availability of this interpretation excludes both the pro-based analysis and the AE-based analysis. The contrast between (12A1) and (12A2) therefore indicates that Japanese VEAs block voice mismatches, a fact that falls into place if VEA is the result of V-to-C head movement and TP-ellipsis.1

3 Syntactic Head Movement in Japanese: Evidence from Negative Scope Reversal

In the previous section, we furnished new evidence from the adverb-inclusive interpretation associated with VEAs and the impossibility of voice mismatches to show that Japanese has string-vacuous V-to-C head movement. The purpose of this section is to demonstrate that this movement has semantic effects. We will present various data regarding the interaction of negation with dake ‘only’ and ka ‘or’ to show that the otherwise stable wide scope of the expressions marked with these particles vis-à-vis negation is reversed under VEAs. This scope reversal pattern supports the view that head movement in Japanese takes place in the narrow syntactic computation.

3.1 Obligatory Wide Scope Effects, Antireconstruction, and Acyclic Merger

In Japanese, quantified objects in direct object position can take scope either above or below negation, as shown in (13).

(13)

  • Taroo-wa [go-nin-izyoo-no gakusei]-o

  • Taro-TOP 5-CL-more.than-GEN student-ACC

  • sikar-anakat-ta. (Obj ≫ Neg, Neg ≫ Obj)

  • scold-NEG-PAST

  • ‘Taro didn’t scold more than five students.’

It is well-known that focus-sensitive particles such as mo ‘also’, dake ‘only’, and sae ‘even’, as well as the disjunctive marker ka ‘or’, can only take scope above negation. This observation is illustrated in (14) and (15) with dake and ka, respectively.

(14)

  • Taroo-wa pan-dake kaw-anak-atta. (only ≫ Neg, *Neg ≫ only)

  • Taro-TOP bread-only buy-NEG-PAST

  • ‘Taro didn’t buy only bread.’

  • (Shibata 2015a:73)

(15)

  • Taroo-wa pan ka kome-o kaw-anak-atta. (or ≫ Neg, *Neg ≫ or)

  • Taro-TOP bread or rice-ACC buy-NEG-PAST

  • ‘Taro didn’t buy bread or rice.’

  • (Shibata 2015a:73)

In this regard, Japanese behaves differently from English, in which focus-marked or disjunctive phrases can take scope below negation, as evidenced in (16a–b). The narrow scope reading of the relevant phrases with respect to negation in (16a–b) is expected if they are base-generated and remain in a position below negation, as is standardly assumed in the ordinarily postulated array of functional projections in English (i.e., V < v < Neg < T).

(16)

  • John didn’t buy only bread. (Neg ≫ only)

  • John didn’t buy bread or rice. (Neg ≫ or)

    (cf. Shibata 2015a:74)

Shibata (2015a,b) takes the scope ambiguity with regular quantified expressions in (13), together with the obligatory wide scope of the focus-marked/disjunctive DPs in (14) and (15), to indicate that object phrases in Japanese must undergo overt movement to a position above negation and that the attachment of a focus-sensitive or disjunctive particle somehow blocks the reconstruction process available in the case of regular quantifiers—what Shibata (2015a:67) calls the antireconstruction effect. Leaving technical details aside, Shibata proposes that the focus and disjunctive particles undergo obligatory acyclic adjunction to their host. This analysis follows in the footsteps of various proposals regarding late insertion of adjuncts, dating back to Lebeaux 1988 and much subsequent work (e.g., Stepanov 2001). Under this proposal, the derivation of the English sentence in (17a), for example, proceeds as shown in (17b–d).

(17)

  • Only that boy didn’t come.

  • [that boy] didn’t [that boy] come (movement of the subject)

  • [only [that boy]] didn’t [that boy] come (acyclic adjunction of only)

  • [only [that boy]] λx. didn’t [the boy identical to x] come (trace conversion)

In (17b), the subject that boy undergoes movement from Spec,vP to Spec,TP. Following the obligatory-late-insertion hypothesis, the modifier only attaches to the subject DP only after it has moved to the surface subject position, as shown in (17c). (17c) is then the input for (17d), a well-formed LF representation for the sentence in (17a).2 It follows then that the subject associated with this operator can only be interpreted above negation, thereby yielding the obligatory wide scope reading of the focus-sensitive phrase, as illustrated in (14).

Shibata (2015a:70) extends the same approach to DPs modified by the disjunction marker ka ‘or’, as illustrated in (18a). He adopts Chierchia, Fox, and Spector’s (2012) proposal that a disjunctive expression is interpreted with a silent exhaustive operator, as in (18b). OALT (S), which asserts S and negates all nonstronger alternatives of S, ensures that (18a) is assigned the weaker disjunctive denotation in (18c), not the stronger conjunctive denotation in (18d).

(18)

  • John ka Tom-ga kuru-daroo.

    John or Tom-NOM come-will

    ‘John or Tom will come.’

  • OALT (John or Tom) will come.

  • Either John will come or Tom will come.

  • Both John and Tom will come.

    (adapted from Shibata 2015a:70)

Given that the semantic computation of disjunction (like that of only) involves obligatory acyclic late merger, a ka-marked phrase must necessarily scope over negation.3

3.2 Negative Scope Reversal under VEAs

Let us now consider the facts about scope under VEAs. Interestingly, the VEA response to the polar question based on an example such as (14) shows the opposite scope relation. The VEA answer in (19) has the wide scope reading of negation with respect to the dake-marked phrase, but it blocks the opposite reading, which was available to (14).

(19)

  • Q: Taroo-wa pan-dake kat-ta-no?

  • Taro-TOP bread-only buy-PAST-Q

  • ‘Did Taro buy only bread?’

  • A: Kawa-nakat-ta-yo. (??only ≫ Neg, Neg ≫ only)

  • buy-NEG-PAST-PRT

  • Lit. ‘Didn’t buy.’

The negative scope reversal pattern is also manifested in the case of VEAs in response to disjunctive questions based on declarative sentences such as (15). Recall that in Japanese, disjunction necessarily takes scope over negation. Again, (20A), as a response to the disjunctive question (20Q), exhibits the scope reversal effect: what was the impossible Neg ≫ or interpretation in (15) is now the only interpretation under the VEA to the disjunctive yes/no question.

(20)

  • Q: Yoichiro-wa kinoo oyatu-ni aisu ka keeki-o tabe-ta-no?

  • Yoichiro-TOP yesterday snack-for ice.cream or cake-ACC eat-PAST-Q

  • ‘Did Yoichiro eat ice cream or cake for his snack yesterday?’

  • A: Tabe-naka-ta-yo. (??or ≫ Neg, Neg ≫ or)

  • eat-NEG-PAST-PRT

  • Lit. ‘Didn’t eat.’

Examples (19) and (20) illustrate the negative scope reversal effect with direct object positions. Below, we will show that the effect obtains quite generally under VEAs, regardless of grammatical function.

Let us start with the case where a phrase marked with dake ‘only’ or ka ‘or’ occurs in subject position in the antecedent question. As first noted by Saito (2010), the dake-marked subject must outscope negation, as shown in (21). Example (22) shows that this scope pattern is reversed in VEAs.

(21)

  • Sono toki kyoositu-ni-wa Taroo-dake-ga

  • that time classroom-in-TOP Taro-only-NOM

  • inak-atta. (only ≫ Neg, *Neg ≫ only)

  • be-NEG-PAST

  • ‘At that time, only Taro was not in the classroom.’

  • (Saito and Takita 2016:418)

(22)

  • Q: Sono toki kyoositu-ni-wa Taroo-dake-ga ita-no?

  • that time classroom-in-TOP Taro-only-NOM be.PAST-Q

  • ‘At that time, was only Taro in the classroom?’

  • A: I-nak-atta-yo. (??only ≫ Neg, Neg ≫ only)

  • be-NEG-PAST-PRT

  • Lit. ‘Wasn’t.’

The negative scope reversal effect is also observed with disjunctive arguments in subject positions. Example (23) has only the interpretation where Taroo ka Hanako ‘Taro or Hanako’ takes scope over negation. By contrast, the VEA response to a yes/no question of this sort admits only the opposite scope order. This point is illustrated in (24).

(23)

  • Sono toki kyoositu-ni-wa Taroo ka Hanako-ga

  • that time classroom-in-TOP Taro or Hanako-NOM

  • i-nak-atta-yo. (or ≫ Neg, *Neg ≫ or)

  • be-NEG-PAST-PRT

  • ‘At that time, either Taro or Hanako was not in the classroom.’

(24)

  • Q: Sono toki kyoositu-ni-wa Taroo ka Hanako-ga ita-no?

  • that time classroom-in-TOP Taro or Hanako-NOM be.PAST-Q

  • ‘At that time, was either Taro or Hanako in the classroom?’

  • A: I-nak-atta-yo. (??or ≫ Neg, Neg ≫ or)

  • be-NEG-PAST-PRT

  • Lit. ‘Wasn’t.’

The negative scope reversal effect is also observed when a focus or disjunctive phrase occupies the indirect object position of a ditransitive verb such as okur ‘to send’. In (25), the indirect object Taroo-ni ‘to Taro’ is marked with dake ‘only’ and takes scope over negation. The VEA has the reverse scope interpretation, as shown in (26A).

(25)

  • Hanako-wa Taroo-ni-dake nengazyoo-o

  • Hanako-TOP Taro-DAT-only New.Year’s.card-ACC

  • okura-nakat-ta. (only ≫ Neg, *Neg ≫ only)

  • send-NEG-PAST

  • ‘Hanako didn’t send her New Year’s card only to Taro.’

  • (Saito and Takita 2016:421)

(26)

  • Q: Hanako-wa Taroo-ni-dake nengazyoo-o okut-ta-no?

  • Hanako-TOP Taro-DAT-only New.Year’s.card-ACC send-PAST-Q

  • ‘Did Hanako send her New Year’s card only to Taro?’

  • A: Okura-nakat-ta-yo. (??only ≫ Neg, Neg ≫ only)

  • send-NEG-PAST-PRT

  • Lit. ‘Didn’t send.’

Examples (27) and (28) illustrate the scope reversal effect in indirect object position with the disjunctive phrase Taroo ka Hanako ‘Taro or Hanako’.

(27)

  • Yukiko-wa Taroo ka Hanako-ni nengazyoo-o

  • Yukiko-TOP Taro or Hanako-DAT New.Year’s.card-ACC

  • okura-nakat-ta. (or ≫ Neg, *Neg ≫ or)

  • send-NEG-PAST

  • ‘Yukiko didn’t send her New Year’s card to either Taro or Hanako.’

(28)

  • Q: Yukiko-wa Taroo ka Hanako-ni nengazyoo-o okut-ta-no?

  • Yukiko-TOP Taro or Hanako-DAT New.Year’s.card-ACC send-PAST-Q

  • ‘Did Yukiko send her New Year’s card to Taro or Hanako?’

  • A: Okura-nakat-ta-yo. (??or ≫ Neg, Neg ≫ or)

  • send-NEG-PAST-PRT

  • Lit. ‘Didn’t send.’

Finally, (29)–(32) show that the negative scope reversal effect is detectable in postpositional phrases. Examples (29) and (30) are control cases showing that such a phrase must scope over negation when it occurs with dake ‘only’ and ka ‘or’, respectively. Keeping this in mind, (31) and (32) are VEAs to the yes/no questions based on (29) and (30). Here, once again, the impossible scope relation in (29) and (30) is now available in (31) and (32).

(29)

  • Taroo-wa Hanako-to-dake hanasi-o si-nakat-ta. (only ≫ Neg, *Neg ≫ only)

  • Taro-TOP Hanako-with-only talk-ACC do-NEG-PAST

  • ‘Taro didn’t talk only with Hanako.’

(30)

  • Yukiko-wa Taroo ka Hanako-to hanasi-o si-nakat-ta. (or ≫ Neg, *Neg ≫ or)

  • Yukiko-TOP Taro or Hanako-with talk-ACC do-NEG-PAST

  • ‘Yukiko didn’t talk with either Taro or Hanako.’

(31)

  • Q: Anata-wa sono hi Hanako-to-dake hanasi-o si-ta-no?

  • you-TOP that day Hanako-with-only talk-ACC do-PAST-Q

  • ‘Did you talk only with Hanako on that day?’

  • A: Si-nakat-ta-yo. (??only ≫ Neg, Neg ≫ only)

  • do-NEG-PAST-PRT

  • Lit. ‘Didn’t do.’

(32)

  • Q: Anata-wa sono hi Taroo ka Hanako-to hanasi-o si-ta-no?

  • you-TOP that day Taro or Hanako-with talk-ACC do-PAST-Q

  • ‘Did you talk with Taro or Hanako on that day?’

  • A: Si-nakat-ta-yo. (??or ≫Neg, Neg ≫ or)

  • do-NEG-PAST-PRT

  • Lit. ‘Didn’t do.’

We summarize the data reported in this section in table 1.4

Table 1
Negative scope reversal under VEAs
Direct objectSubjectIndirect objectPP object
Base scope (Control case) ‘only’ only ≫ Neg; *Neg ≫ only (Antireconstruction effect) 
(14) (21) (25) (29) 
Disjunction or ≫ Neg; *Neg ≫ or (Antireconstruction effect) 
(15) (23) (27) (30) 
Derived scope (VEA) ‘only’ ??only ≫ Neg; Neg ≫ only (Negative scope reversal) 
(19) (22) (26) (31) 
Disjunction ??or ≫ Neg; Neg ≫ or (Negative scope reversal) 
(20) (24) (28) (32) 
Direct objectSubjectIndirect objectPP object
Base scope (Control case) ‘only’ only ≫ Neg; *Neg ≫ only (Antireconstruction effect) 
(14) (21) (25) (29) 
Disjunction or ≫ Neg; *Neg ≫ or (Antireconstruction effect) 
(15) (23) (27) (30) 
Derived scope (VEA) ‘only’ ??only ≫ Neg; Neg ≫ only (Negative scope reversal) 
(19) (22) (26) (31) 
Disjunction ??or ≫ Neg; Neg ≫ or (Negative scope reversal) 
(20) (24) (28) (32) 

3.3 Negative Scope Reversal Effects as Evidence for Syntactic Head Movement

On Shibata’s (2015a,b) analysis, the derivation of negative sentences that display the surface scope effect takes the form shown in (33). In this derivation, all vP-internal phrases undergo obligatory syntactic movement above negation into the TP region (see footnote 3). Coupled with obligatory late insertion of focus-sensitive and disjunctive operators, this derivation yields mandatory wide scope of the moved phrases with respect to negation.

(33)

graphic

On this account, the presence of the scope reversal effect indicates that the verbal complex containing negation must undergo head movement to a position that c-commands all arguments in TP. In other words, the verb undergoes successive-cyclic movement from the V position, picking up negation on its way, before it lands in the C position. Since movement to the C position changes the resulting scope relations, we can now conclude that the movement in question is a syntactic operation rather than a postsyntactic phenomenon in the phonological component.

In the rest of this section, we present one technical implementation of our central observation about the role of syntactic head movement in determining scope interpretation within a more general understanding of how negative scope works, as informed by previous literature. There is a long tradition of generative literature on negation (see, e.g., Klima 1964, Jacobs 1980, Ladusaw 1992, Rullmann 1995, Kayne 1998, Penka and von Stechow 2001, Potts 2002, Zeijlstra 2004, 2007, 2008a,b, 2011, Zeijlstra and Penka 2005, Schwarz and Bhatt 2006, Penka 2007, 2011, Iatridou and Sichel 2011, and many references cited therein) showing that there is a difference between morphological negation markers and the abstract syntactic position(s) in which semantic negativity is interpreted. Penka and von Stechow (2001), for instance, argue for the idea that semantic negation and morphological negative marking are independent on the basis of split-scope readings of negative indefinites. The split-scope reading is illustrated in the German example in (34).

(34)

  • Du musst keine Krawatte anziehen.

  • you must no tie wear

  • ‘It is not required that you wear a tie.’

  • (Zeijlstra 2011:113)

According to the salient reading of this example, negation takes wide scope over the modal, while the indefinite expression takes narrow scope. In other words, an operator is interpreted between semantic negation and the indefinite, an interpretation that would remain unexplained if the negative DP keine Krawatte ‘no tie’ were just a negative quantifier. Penka and von Stechow (2001) argue that negative indefinites are indefinite expressions that obligatorily cooccur with an abstract negative operator/feature that is not realized phonetically. In a similar vein, Zeijlstra (2004, 2007, 2008a,b, 2011) proposes that negative concord in languages such as Italian and Czech instantiates a syntactic agreement relation between one or more elements that carry an uninterpretable formal negative feature ([uNEG]) and a single negative operator that carries an interpretive formal negative feature ([iNEG]).

The family of approaches briefly outlined above, which argue for strict separation of morphological negation from negative operators/features/positions, allow for a straightforward implementation of our central observation regarding the scope-shifting potential of syntactic head movement in Japanese VEAs. Adopting Zeijlstra’s approach to negative concord in Japanese, we assume that there is an abstract agreement relationship between the morphological negation marker nai and the higher negation position/feature/operator in the derivation of negative sentences shown in (33). Recall that the VEA examples ((19), (20), (22), (24), (26), (28), (31), (32)) involve V- to-C movement of the verbal complex to outscope negation, whereas the baseline cases ((14), (15), (21), (23), (25), (27), (29), (30)) do not involve such movement and do not exhibit scope shift. We take this systematic contrast as evidence that V-to-C head movement in syntax is necessary to give rise to the scope shift in the VEA examples. Thus, we maintain that, as a follow-up operation to the abstract agreement relation, syntactic head movement still takes place in order to activate the semantic negativity of the negative position/feature/operator in (33).

4 Conclusion

One of the most vigorously contested issues in the Minimalist Program is the nature and locus of head movement. Strictly head-final languages like Japanese add a further complexity to the issue regarding the very existence of this phenomenon, given that any such movement yields string-vacuous outputs. The purpose of this article has been to shed light on these two thorny issues from the previously unexplored perspective of verb-echo answers.

We have argued for syntactic head movement in VEAs in two steps. First, we have proposed that VEAs in Japanese are best derived through successive-cyclic head movement of the echoed verb to the C position, followed by ellipsis of the TP at PF. Our evidence for this analysis was twofold. One type of evidence was concerned with the adverb-inclusive interpretation associated with VEAs. The availability of this interpretation, we have demonstrated, eliminates the otherwise plausible analyses of VEAs in terms of pros and argument ellipsis in favor of our TP-ellipsis analysis. The other type of evidence for our proposed analysis came from the fact that VEAs exhibit some hallmarks of other better-studied instances of TP-ellipsis such as sluicing. More specifically, VEAs do not accept voice mismatches. Given these pieces of evidence for TP-ellipsis, it follows that the verb sequence in VEAs must raise to the C head position to escape the TP-ellipsis.

Second, we have established a new generalization that the otherwise obligatory wide scope exhibited by focus-marked or disjunctive phrases with respect to negation is reversed in the context of VEAs, a pattern that cuts across all grammatical positions. We have shown how this generalization supports our conclusion that the head movement involved in VEAs is syntactic, contrary to the conjecture that it is a PF phenomenon (Chomsky 2001; see also the references cited in section 1). This argument is modeled after Lechner’s (2007, 2017) battery of empirical arguments, based on scope-splitting, comparatives, and coordination, for the conjecture that there are instances or manifestations of semantically active head movement. Our conclusion regarding the architectural locus of head movement in Japanese is also compatible with Kishimoto’s (2007, 2008) finding, based on negative polarity licensing, that head movement in this language occurs in syntax. Whether the Japanese grammar exhibits any other instances of semantically active head movement remains to be seen.

Notes

1 A reviewer asks whether the prohibition on voice mismatches with VEAs indicates TP-deletion or should rather be seen as an effect of the focus parallelism constraint outlined in Kertz 2008, 2010. We believe that Kertz’s findings have no direct impact on our working assumption that the prohibition of voice mismatches under VEA is due to TP-deletion. Kertz’s proposal is designed to condition the acceptability of voice mismatches under verbal ellipsis, but TP-ellipsis, as instantiated in sluicing and VEAs, uniformly prohibits any voice mismatch in any configuration under any information-structural accommodation in the first place.

2 Shibata (2015a,b) adopts Fox’s (2002, 2003) Trace Conversion, a two-step process consisting of variable insertion and determiner replacement, as defined in (ia–b).

    • Variable insertion

      (Det) Pred → (Det) [Pred λy (y = x)]

    • Determiner replacement

      (Det) [Pred λy (y = x)] → the [Pred λy (y = x)]

      (Fox 2002:67)

3 One question Shibata (2015a,b) addresses is why objects—or all vP-internal phrases, for that matter—must undergo movement above negation. Shibata (2015a) argues that the reason is that morphological merger may subsequently apply to create the complex predicate under the condition of structural adjacency: any vP-internal expression, such as a direct object, would necessarily block the morphological merger among V, v, and negation.

4 If we replace the affirmative yes/no questions in (19Q) and (20Q) with their negative counterparts, we obtain (iA) and (iiA). Significantly, the VEAs to the negative polarity questions do not exhibit the scope reversal effect.

  • Q: Taroo-wa pan-dake kawa-nakat-ta-no?

    Taro-TOP bread-only buy-NEG-PAST-Q

    ‘Didn’t Taro buy only bread?’

    A: Kawa-nakat-ta-yo. (only ≫ Neg, *Neg ≫ only)

    buy-NEG-PAST-PRT

    Lit. ‘Didn’t buy.’

  • Q: Yoichiro-wa kinoo oyatu-ni aisu ka keeki-o tabe-nakat-ta-no?

    Yoichiro-TOP yesterday snack-for ice.cream or cake-ACC eat-NEG-PAST-Q

    ‘Didn’t Yoichiro eat ice cream or cake for his snack yesterday?’

    A: Tabe-naka-ta-yo. (or ≫ Neg, *Neg ≫ or)

    eat-NEG-PAST-PRT

    Lit. ‘Didn’t eat.’

We will revisit the question that (i)–(ii) raise for our generalization in future work.

Various versions of this article were presented at the Linguistic Colloquium hosted by the Indian Institute of Technology Delhi (October 2016), the National University of Singapore (April 2018), the 41st GLOW Conference (April 2018), and the Linguistic Colloquium at Kobe University (May 2018). We thank the audience members on those occasions who provided helpful feedback on our project, too numerous to mention individually. Special thanks to two anonymous reviewers for their invaluable suggestions. Sato’s research is supported by the Singapore Ministry of Education Academic Research Fund Tier 1 (Project #: R-103-000-152-115), the Japan Society for the Promotion of Science, Grant-in-Aid for Scientific Research (C) Project #: JP19K0056), and the Seisen University Faculty Research Development Grant (April 2019–March 2020). Maeda’s research is supported by the Japan Society for the Promotion of Science, Grant-in-Aid for Young Scientists (Project #: 18K12412), Grant-in-Aid for Scientific Research (C) (Project #: 18K00574), and Grant-in-Aid for Scientific Research (C) (Project #: 18K00654).

References

Aoyagi,
Hiroshi
.
2006
.
On morphological merger
.
Nanzan Linguistics
3
:
1
32
.
Baker,
Mark
.
1988
.
Incorporation: A theory of grammatical function changing
.
Chicago
:
University of Chicago Press
.
Boeckx,
Cedric
, and
Sandra
Stjepanović
.
2001
.
Head-ing toward PF
.
Linguistic Inquiry
32
:
345
355
.
Chierchia,
Gennaro
,
Danny
Fox
, and
Benjamin
Spector
.
2012
. The grammatical view of scalar implicatures and the relationship between semantics and pragmatics. In
Semantics: An international handbook of natural language meaning
, ed. by
Claudia
Maienborn
,
Klaus
von Heusinger
, and
Paul
Portner
,
2297
2232
.
Berlin
:
Mouton de Gruyter
.
Chomsky,
Noam
.
2000
. Minimalist inquiries: The framework. In
Step by step: Essays on Minimalist syntax in honor of Howard Lasnik
, ed. by
Roger
Martin
,
David
Michaels
, and
Juan
Uriagereka
,
89
155
.
Cambridge, MA
:
MIT Press
.
Chomsky,
Noam
.
2001
. Derivation by phase. In
Ken Hale: A life in language
, ed. by
Michael
Kenstowicz
,
1
52
.
Cambridge, MA
:
MIT Press
.
Chung,
Sandra
.
2013
.
Syntactic identity in sluicing: How much and why
.
Linguistic Inquiry
44
:
1
44
.
Fox,
Danny
.
2002
.
Antecedent-contained deletion and the copy theory of movement
.
Linguistic Inquiry
33
:
63
96
.
Fox,
Danny
.
2003
. On Logical Form. In
Minimalist syntax
, ed. by
Randall
Hendrick
,
82
123
.
Oxford
:
Blackwell
.
Fukui,
Naoki
, and
Hiromu
Sakai
.
2003
.
The visibility guideline for functional categories: Verb raising in Japanese and related issues
.
Lingua
113
:
321
375
.
Fukui,
Naoki
, and
Yuji
Takano
.
1998
.
Symmetry in syntax: Merge and Demerge
.
Journal of East Asian Linguistics
7
:
27
86
.
Hale,
Ken
, and
Samuel Jay
Keyser
.
2002
.
Prolegomenon to a theory of argument structure
.
Cambridge, MA
:
MIT Press
.
Harley,
Heidi
.
2004
. Merge, conflation, and head movement: The First Sister Principle revisited. In
NELS 34: Proceedings of the 34th Annual Meeting of the North East Linguistic Society
, ed. by
Keir
Moulton
and
Matthew
Wolf
,
239
254
.
Amherst
:
University of Massachusetts, Graduate Linguistic Student Association
.
Hartman,
Jeremy
.
2011
.
The semantic uniformity of traces: Evidence from ellipsis parallelism
.
Linguistic Inquiry
42
:
367
388
.
Hayashi,
Shintaro
, and
Tomohiro
Fujii
.
2015
.
String vacuous head movement: The case of V-te in Japanese
.
Gengo Kenkyu
147
:
31
55
.
Hoji,
Hajime
.
1985
.
Logical Form constraints and configurational structures in Japanese
.
Doctoral dissertation, University of Washington
.
Hoji,
Hajime
.
1998
.
Null objects and sloppy identity in Japanese
.
Linguistic Inquiry
29
:
127
152
.
Holmberg,
Anders
.
2016
.
The syntax of yes and no
.
Cambridge
:
Cambridge University Press
.
Iatridou,
Sabine
, and
Ivy
Sichel
.
2011
.
Negative DPs, A-movement, and scope diminishment
.
Linguistic Inquiry
42
:
595
629
.
Iatridou,
Sabine
, and
Hedde
Zeijlstra
.
2013
.
Negation, polarity, and deontic modals
.
Linguistic Inquiry
44
:
529
568
.
Jacobs,
Joachim
.
1980
.
Lexical decomposition in Montague Grammar
.
Theoretical Linguistics
7
:
121
136
.
Kayne,
Richard
.
1998
.
Overt vs. covert movement
.
Syntax
1
:
128
191
.
Keine,
Stefan
, and
Rajesh
Bhatt
.
2016
.
Interpreting verb clusters
.
Natural Language and Linguistic Theory
34
:
1445
1492
.
Kertz,
Laura
.
2008
. Focus structure and acceptability in verb phrase ellipsis. In
WCCFL 27: Proceedings of the 27th West Coast Conference on Formal Linguistics
, ed. by
Natasha
Abner
and
Jason
Bishop
,
283
291
.
Somerville, MA
:
Cascadilla Proceedings Project
.
Kertz,
Laura
.
2010
.
Ellipsis reconsidered
.
Doctoral dissertation, University of California, San Diego
.
Kishimoto,
Hideki
.
2007
.
Negative scope and head raising in Japanese
.
Lingua
117
:
247
288
.
Kishimoto,
Hideki
.
2008
.
On the variability of negative scope in Japanese
.
Journal of Linguistics
44
:
379
435
.
Klima,
Edward
.
1964
. Negation in English. In
The structure of language
, ed. by
Jerry
Fodor
and
Jerrold
Katz
,
246
323
.
Englewood Cliffs, NJ
:
Prentice-Hall
.
Koizumi,
Masatoshi
.
2000
.
String vacuous overt verb-raising
.
Journal of East Asian Linguistics
9
:
227
285
.
Koopman,
Hilda
.
2005
.
Korean (and Japanese) morphology from a syntactic perspective
.
Linguistic Inquiry
36
:
601
633
.
Kuroda,
S.-Y.
1965
.
Generative grammatical studies in the Japanese language
.
Doctoral dissertation, MIT
.
Ladusaw,
William
.
1992
. Expressing negation. In
Proceedings of SALT II
, ed. by
Chris
Baker
and
David
Dowty
,
237
259
.
Columbus
:
Ohio State University, OSU Working Papers in Linguistics
.
Lebeaux,
David
.
1988
.
Language acquisition and the form of the grammar
.
Doctoral dissertation, University of Massachusetts, Amherst
.
Lechner,
Winfried
.
2007
.
Interpretive effects of head movement
.
Ms., University of Stuttgart. https://ling.auf.net/lingbuzz/000178
.
Lechner,
Winfried
.
2017
.
In defense of semantically active head movement
.
Paper presented at the Workshop for Martin Prinzhorn, Technical University, Vienna, 11 November. http://users.uoa.gr/~wlechner/Vienna%20HM%202017.pdf
.
Merchant,
Jason
.
2001
.
The syntax of silence: Sluicing, islands, and the theory of ellipsis
.
Oxford
:
Oxford University Press
.
Merchant,
Jason
.
2008
.
An asymmetry in voice mismatches in VP-ellipsis and pseudogapping
.
Linguistic Inquiry
39
:
169
179
.
Merchant,
Jason
.
2013
.
Voice and ellipsis
.
Linguistic Inquiry
44
:
77
108
.
Ohso,
Mieko
.
1976
.
A study of zero pronominalization in Japanese
.
Doctoral dissertation, Ohio State University
.
Oku,
Satoshi
.
1998
.
A theory of selection and reconstruction in the Minimalist perspective
.
Doctoral dissertation, University of Connecticut, Storrs
.
Otani,
Kazuyo
, and
John
Whitman
.
1991
.
V-raising and VP-ellipsis
Linguistic Inquiry
22
:
345
358
.
Penka,
Doris
.
2007
.
Negative indefinites
.
Doctoral dissertation, Universität Tübingen
.
Penka,
Doris
.
2011
.
Negative indefinites
.
Oxford
:
Oxford University Press
.
Penka,
Doris
, and
Arnim
von Stechow
.
2001
. Negative Indefinita unter Modalverben. In
Modalität und Modalverben im Deutschen: Linguistische Berichte, Sonderheft 9
, ed. by
Reimar
Müller
and
Marga
Reis
,
263
286
.
Hamburg
:
Helmut Buske
.
Pollock,
Jean-Yves
.
1989
.
Verb movement, Universal Grammar, and the structure of IP
.
Linguistic Inquiry
20
:
365
424
.
Potts,
Christopher
.
2002
.
The syntax and semantics of as-parentheticals
.
Natural Language and Linguistic Theory
20
:
623
689
.
Roberts,
Ian
.
2010
.
Agreement and head movement: Clitics, incorporation, and defective goals
.
Cambridge, MA
:
MIT Press
.
Rullmann,
Hotze
.
1995
.
Geen eenheid
.
Tabu
25
:
194
197
.
Saito,
Mamoru
.
1985
.
Some asymmetries in Japanese and their theoretical implications
.
Doctoral dissertation, MIT
.
Saito,
Mamoru
.
2007
.
Notes on East Asian argument ellipsis
.
Language Research
43
:
203
227
.
Saito,
Mamoru
.
2010
. On the scope properties of nominative phrases in Japanese. In
Universals and variation
, ed. by
Rajat
Mohanty
and
Mythili
Menon
,
313
333
.
Hyderabad
:
EFL University Press
.
Saito,
Mamoru
, and
Kensuke
Takita
.
2016
. Hiteiji to suuryoosi no sayooiki: Shibata Yoshiyuki-shi no kenkyu. In
Nihongo handobukku: Gengo riron to Gengo kakutoku no kanten kara
, ed. by
Keiko
Murasugi
,
Mamoru
Saito
,
Yoichi
Miyamoto
, and
Kensuke
Takita
,
408
443
.
Tokyo
:
Kaitakusha
.
Sakai,
Hiromu
.
1998
. Feature checking and morphological merger. In
Japanese/Korean Linguistics 8
, ed. by
David James
Silva
,
189
201
.
Stanford, CA
:
CSLI Publications
.
Schwarz,
Bernhard
, and
Rajesh
Bhatt
.
2006
. Light negation and polarity. In
Crosslinguistic research in syntax and semantics: Negation, tense, and clausal architecture
, ed. by
Raffaella
Zanuttini
,
Héctor
Campos
,
Elena
Herburger
, and
Paul
Portner
,
175
198
.
Washington, DC
:
Georgetown University Press
.
Shibata,
Yoshiyuki
.
2015a
.
Exploring syntax from the interfaces
.
Doctoral dissertation, University of Connecticut, Storrs
.
Shibata,
Yoshiyuki
.
2015b
.
Negative structure and object movement in Japanese
.
Journal of East Asian Linguistics
24
:
217
269
.
Stepanov,
Arthur
.
2001
.
Late adjunction and Minimalist phrase structure
.
Syntax
4
:
94
125
.
Sugimura,
Mina
.
2012
.
Domain extension: A study of restructuring predicates in Japanese and the role of head movement
.
Doctoral dissertation, McGill University
.
Sugisaki,
Koji
.
2014
.
Sluicing and its identity conditions in Child Japanese: A preliminary report
.
Paper presented at the Workshop on Korean-Japanese Linguistics, Seoul National University, 17–18 October
.
Takahashi,
Daiko
.
2008a
. Noun phrase ellipsis. In
The Oxford handbook of Japanese linguistics
, ed. by
Shigeru
Miyagawa
and
Mamoru
Saito
,
394
422
.
New York
:
Oxford University Press
.
Takahashi,
Daiko
.
2008b
.
Quantificational null objects and argument ellipsis
.
Linguistic Inquiry
39
:
307
326
.
Takano,
Yuji
.
2004
.
Coordination of verbs and two types of verbal inflection
.
Linguistic Inquiry
35
:
168
178
.
Travis,
Lisa
.
1984
.
Parameters and effects of word order variation
.
Doctoral dissertation, MIT
.
Zeijlstra,
Hedde
.
2004
.
Sentential negation and negative concord
.
Doctoral dissertation, University of Amsterdam
.
Zeijlstra,
Hedde
.
2007
.
Negation in natural language: On the form and meaning of negative elements
.
Language and Linguistics Compass
1
:
498
518
.
Zeijlstra,
Hedde
.
2008a
.
Negative concord is syntactic agreement
.
Ms., University of Amsterdam
.
Zeijlstra,
Hedde
.
2008b
. On the syntactic flexibility of formal features. In
The limits of syntactic variation
, ed. by
Theresa
Biberauer
,
143
174
.
Amsterdam
:
John Benjamins
.
Zeijlstra,
Hedde
.
2011
.
On the syntactically complex status of negative indefinites
.
Journal of Comparative Germanic Linguistics
14
:
111
138
.
Zeijlstra,
Hedde
, and
Doris
Penka
.
2005
.
Negative indefinites in Dutch and German
.
Ms., Universität Tübingen
.