Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE (www.oxfordhandbooks.com). © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

Subscriber: null; date: 16 October 2019

Indeterminate Pronouns

Abstract and Keywords

This article presents a survey on issues surrounding the so-called indeterminate pronouns in Japanese such as dare “who” and nani “what.” It focuses on two contexts, interrogative and universal, which easily allow apparently long-distance association of indeterminates with the particles ka and mo. Two major approaches on how indeterminate pronouns associate with quantificational particles that occur higher in the structure—namely, the movement and nonmovement analyses—are reviewed from the perspective of the syntax–semantics interface. The movement analysis takes the presence of the wh-island effect to be indicative of a movement relation between indeterminates and higher particles, and attempts to provide an account for the lack of other island effects. The question of whether the reviewed proposals can be extended to handle the syntax and semantics of the Japanese indeterminate pronoun system, and the indeterminate pronoun system in natural language, remains to be examined.

Keywords: Japanese indeterminate pronoun, dare, nani, syntax, semantics, natural language

14.1 Introduction

This chapter surveys some of the major issues surrounding so-called indeterminate pronouns in Japanese such as dare ‘who’ and nani ‘what’. One of the questions that has attracted the most attention in this domain concerns how indeterminate pronouns associate with quantificational particles that occur higher in the structure. Two major approaches to this question—namely, the movement and nonmovement analyses—are reviewed from the perspective of the syntax-semantics interface, and the challenges that these analyses face are presented. Closely related to this issue is the question of what the semantics of indeterminate pronouns is. I survey two views on this and point out the advantages and disadvantages.

14.2 Brief Introduction of the Problem: Indeterminate Pronouns

Japanese expressions such as dare ‘who’, nani ‘what’, doko ‘where’, and so forth—although typically given English translations using wh-words—do not have interrogative meanings inherently. The quantificational force of these expressions depends on the particles that they co-occur with, as table 14.1 shows (the table is (p. 373) not intended to be an exhaustive list of indeterminate pronouns). The apostrophes in the table indicate the location of accent in the standard dialect.

Table 14.1 Japanese indeterminate pronouns and their quantificational meanings

a. Interrogative

b. Existential

c. Universal

d. NPI any

e. FC any

daʼre…ka

‘who’

daʼre-ka

daʼre-mo

dare-mo

dare-de-mo

naʼni…ka

‘what’

naʼni-ka

(naʼni-mo)

nani-mo

nan-de-mo

doʼre…ka

‘which (one)’

doʼre-ka

doʼre-mo

dore-mo

dore-de-mo

doʼno N…ka

‘whichDet

doʼno N-ka

doʼno N-mo

dono N-mo

dono N-de-mo

doʼtira…ka

‘which of the two’

doʼtira-ka

doʼtira-mo

dotira-mo

dotira-de-mo

doʼko…ka

‘where’

doʼko-ka

doʼko-mo

doko-mo

doko-de-mo

iʼtu…ka

‘when’

iʼtuka

iʼtu-mo

itu-de-mo

naʼze…ka

‘why’

naʼze-ka

doʼo…ka

‘how’

(doʼo-ka)

(doʼo-mo)

(doo-mo)

doo-de-mo

Because of their varying semantics, the expressions in question are referred to as huteigo ‘indeterminate words’ in traditional Japanese grammar and as indeterminate pronouns by Kuroda (1965:91). Indeterminate pronouns are interpreted as interrogative when they associate with the question particle ka that occurs at the right periphery of a clause as in column (a) of the table. When the indeterminates are immediately followed by ka as in column (b), they take on existential meaning. Columns (c) and (d) show that accented indeterminates and the particle mo give rise to universal meaning, whereas nonaccented indeterminates and mo give rise to meaning similar to the negative polarity item (NPI) any. Finally, column (e) shows that nonaccented indeterminates and -de-mo are interpreted as free choice any.

What we have just seen is by no means a phenomenon peculiar to Japanese, as can be seen in cross-linguistic studies such as Haspelmath 1997. To provide just one sample, table 14.2 shows a list of what we might call indeterminate pronouns in Basque and how they take on various quantificational meanings depending on the co-occurring particles (Haspelmath 1997:315).

When the common cores of all the series (nor, zer, non, etc.) occur on their own as in column (a), interrogative meaning arises. The -bait-series in column (b) are nonemphatic indefinites used in nonnegative polarity contexts, and the i-series in column (c) are negative polarity items. The edo-series and -nahi-series in columns (d) and (e) are found in free choice contexts.

Among the various uses of indeterminate pronouns in Japanese, the interrogative use, or more precisely its syntax, has received the most attention in the literature. As overviewed by Richards in this volume, Japanese indeterminates in the context of wh-questions have figured more prominently in the development of syntactic theory in the last few decades than those in any other contexts. However, given the way quantification is expressed in general in Japanese as observed in table 14.1 and the fact (p. 374) that the Japanese-type indeterminate pronouns are found cross-linguistically, we can say that looking at wh-questions in isolation may not necessarily give us an insight into learning how quantification in general works in Japanese. For any of the contexts where indeterminates show up, its analysis should be couched within a general theory of indeterminate pronoun quantification in Japanese. Of course, we would ultimately want to see a theory of quantification involving indeterminate pronouns across languages. First steps toward that direction have been taken by Kratzer and Shimoyama 2002 and Kratzer 2005, to which the reader is referred.

Table 14.2 Basque indeterminate pronouns and their quantificational meanings

a. Interrogative

b. -bait-series

c. i-series

d. edo-series

e. -nahi-series

nor ‘who’

nor-bait

i-nor

edo-nor

nor-nahi

zer ‘what’

zer-bait

i-zer

edo-zer

zer-nahi

non ‘where’

non-bait

i-non

edo-non

non-nahi

noiz ‘when’

noiz-bait

i-noiz

edo-noiz

noiz-nahi

nola ‘how’

nola-bait

i-nola

edo-nola

nola-nahi

zein ‘whichDet

edo-zein

zein-nahi

14.3 Description of the Core Data: Association of Indeterminate Pronouns and Particles

One of the central questions that has been addressed in the literature concerns how indeterminate pronouns associate with particles that do not occur locally. This section introduces the core data in this domain.

14.3.1 Nonlocal Association

Looking at table 14.1, one might get the impression that except for the interrogative case, Japanese is not that different from Basque, or English for that matter, because the particles from which indeterminates derive their quantificational force occur right next to them. This, however, is not always the case. The examples in (1) illustrate that the particle mo or ka can occur at either side of the postposition ni ‘to’ or kara ‘from’.

  1. (1)

    1. a.

      Daremo-ni/-ni-mo

      denwa-o

      kaketa.

      who-mo-to/-to-mo

      phone-acc

      rang

      ‘(I) called everyone.’

    2. b.

      Dare-ka-kara/-kara-ka

      denwa-ga

      atta.

      who-ka-from/from-ka

      phone-nom

      existed

      ‘There was a call from someone.’

(p. 375) These are cases of universal and existential quantification, but a similar observation can be made for the NPI and free choice contexts. In these contexts, however, the particles mo and de-mo must occur to the right of postpositions.

In the universal construction, the distance between indeterminates and associating particles can even be farther than just across postpositions. The example in (2a) illustrates that the universal particle mo can associate with dare ‘who’ across a noun phrase boundary. The example in (2b) is even more striking—the association takes place across a complex NP, which contains a relative clause.

  1. (2)

    1. a.

      [Dare-no

      hahaoya]-mo

      naita.

      who-gen

      mother-mo

      cried

      ‘For everyone x, the mother of x cried.

    2. b.

      [Dare-ga

      katta ie]-mo

      takakatta.

      who-nom

      bought house-mo

      was.expensive

      ‘For everyone x, a/the house(s) that x bought was/were expensive.

Such data have attracted the attention of Kuroda (1965), Hoji (1985), Nishigauchi (1986, 1990), Ohno (1989), von Stechow (1996), Shimoyama (1999, 2001, 2006a), and Takahashi (2002), among many others. Korean has a very similar construction, as shown in (3), taken from Ohno 1991.
  1. (3)

    [Enu sonye-ka pon koyangi]

    -na

    kwiyewessta.

    which girl-nom saw cat

    -na

    were.cute

    ‘The cats that each girl saw were cute.’

The universal mo (and perhaps the free choice de-mo) most freely allows the long-distance association of the kind we just observed. The existential particle ka, for instance, does not seem to allow nonlocal association as freely, as shown in (4) and (5), though Takahashi (2002) and Yatsushiro (2004) report on data from more permissive dialects.1

  1. (4)

    1. a.

      Taroo-wa

      dareka-no

      hon-o

      nakusi-ta.

      Taroo-top

      who-ka-gen

      book-acc

      lost

      ‘Taroo lost someone's book.’

    2. b.

      ?/??Taroo-wa

      dare-no

      hon-ka-o

      nakusi-ta.

      Taroo-top

      who-gen

      book-ka-acc

      lost

  2. (5)

    1. a.

      Taroo-wa

      [dareka-kara

      karita

      hon]-o

      nakusita.

      Taroo-top

      who-ka-from

      borrowed

      book-acc

      lost

      ‘Taroo lost a/the book that he had borrowed from someone.’

    2. b.

      ??/Taroo-wa

      [darekara

      karita

      hon]-ka-o

      nakusi-ta.

      Taroo-top

      who-from

      borrowed

      book-ka-acc

      lost

(p. 376) 14.3.2 Locality

Data such as (2b), where an indeterminate depends on the particle mo outside the complex NP it is embedded in, are familiar from indeterminates in the interrogative context. In (6), association of doko ‘where’ and the question particle ka is established across complex NP (see Richards, this volume).2

  1. (6)

    Hanako-wa

    [doko-kara

    kita

    hito]-ni

    aimasita

    ka?

    Hanako-top

    where-from

    came

    person-dat

    met

    q

    ‘Wherex did Hanako meet a/the person/people who came from x?’

This seemingly long-distance association in the interrogative context as well as in the universal context is in fact constrained by a certain locality condition. An important observation made by Nishigauchi (1986, 1990) is that the association does not result in full grammaticality in the presence of a wh-island. Starting from a wh-question example in (7), we can see that the sentence is most readily interpreted as a matrix yes/no-question and an embedded multiple question.3
  1. (7)

    Sabu-wa

    [Masako-ga

    dare-ni

    nani-o

    ageta

    ka]

    oboeteimasu

    ka?

    Sabu-top

    Masako-nom

    who-dat

    what-acc

    gave

    q

    remember

    q

    1. a. ‘Does Sabu remember what Masako gave to whom?’

    2. b. ?‘Whox does Sabu remember what Masako gave to x?’

    3. c. ‘Whatx does Sabu remember to whom Masako gave x?’

    4. d. ?‘Whox does Sabu remember whether Masako gave what to x?’

In a similar manner, the association between indeterminates and mo is blocked by a wh-island. Note first that the particle mo has a use in which it means ‘also’ or ‘even’. So, when there is no indeterminate pronoun in the scope of mo as in (8), ‘also’ or ‘even’ is the only possible interpretation.
  1. (8)

    Sono

    hito-mo

    Sabu-ni

    denwasita.

    That

    hito-mo

    Sabu-dat

    called

    ‘That person also called Sabu./Even that person called Sabu.’

In (9), we see that an attempt to interpret dare ‘who’ and/or nani ‘what’ as having a universal force fails.
  1. (9)

    [[[Masako-ga

    dare-ni

    nani-o

    ageta

    ka]

    siritagatteiru]

    Masako-nom

    who-dat

    what-acc

    gave

    q

    want.to.know

    hito]-mo

    Sabu-ni

    denwasita.

    person-mo

    Sabu-dat

    called

    1. a. ‘The person who wanted to know what Masako gave to whom also called Sabu.’

      ‘Even the person who wanted to know what Masako gave to whom called Sabu.’

    2. b. ‘For every person x, the person who wanted to know what Masako gave to x called Sabu.’

    3. (p. 377) c. ⋆⋆‘For every thing x, the person who wanted to know to whom Masako gave x called Sabu.’

    4. d. ‘For every person x, for every thing y, the person who wanted to know whether Masako gave y to x called Sabu.’

Based on observations of this kind, Nishigauchi (1986, 1990) has established that a uniform analysis should be sought for the two types of indeterminate-particle association.

To reinforce the need for such a uniform analysis, we can add here that a mo-marked phrase functions similarly to a wh-island in that it blocks indeterminate-particle association. Examples (10) and (11) from Shimoyama 2006a illustrate this point. Both show that an indeterminate cannot be construed with a higher particle (either ka or mo) when the lower particle, mo, is interpreted as universal.4

  1. (10)

    Yoko-wa

    [[[Taroo-ga

    nan-nen-ni

    nani-nituite

    kaita]

    Yoko-top

    Taroo-nom

    what-year-in

    what-about

    wrote

    ronbun]

    -mo

    yuu-datta

    ka]

    siritagatteiru.

    paper

    -mo

    A-was

    q

    want.to.know

    1. a. ‘Yoko wonders whether for every topic x, every year y, the paper that Taroo wrote on x in y got an A.’

    2. b. ?‘Yoko wonders for which year y, for every topic x, the paper that Taroo wrote on x in y got an A.’

    3. c. ‘Yoko wonders for which topic x, for every year y, the paper that Taroo wrote on x in y got an A.’

    4. d. (?)‘Yoko wonders for which topic x and for which year y, the paper that Taroo wrote on x in y also got an A.’

  2. (11)

    [[[[Taroo-ga

    nan-nen-ni

    nani-nituite

    kaita] ronbun]

    -mo

    yonda]

    Taroo-nom

    what-year-in

    what-about

    wrote paper

    -mo

    read

    sensei]-mo

    totemo

    tukareta

    teacher-mo

    very.much

    got.tired

    1. a. ‘The teacher who read, for every topic x, every year y, the paper that Taroo wrote on x in y also got very tired.’

    2. b. ‘For every year y, the teacher who read, for every topic x, the paper that Taroo wrote on x in y got very tired.’

    3. c. ⋆⋆‘For every topic x, the teacher who read, for every year y, the paper that Taroo wrote on x in y got very tired.’

    4. d. (?)‘For every topic x, every year y, the teacher who also read the paper that Taroo wrote on x in y got very tired.’

In summary, indeterminate pronouns can associate with higher particles ka and mo in a long-distance manner, except when these particles occur in the association path as shown in (12).

  1. (12) Indeterminate Pronouns

(p. 378) This locality pattern observed in Japanese has created an intriguing challenge to the theory of wh-dependencies. The pattern is different in interesting ways from the well-studied dependency between a wh-phrase and its trace in English, created by overt or covert movement. Overt wh-movement in English is subject to island constraints including not just wh-islands but also complex NP, adjunct, and subject islands, whereas covert wh-movement is not. Providing explanation to this cross-linguistic variation has been a central concern in syntactic research (see Richards, this volume). In the next two sections, I provide an overview of two major types of analyses—namely, the movement and nonmovement analyses—mainly from the perspective of the syntax-semantics interface. I mainly focus on how they fare in providing a uniform analysis of the interrogative and universal constructions.

14.4 Issues and Controversies Regarding Analysis: The Movement Analysis, Covert Pied-Piping, and the Syntax-Semantics Interface

The fact that the indeterminate-particle dependency is sensitive to wh-islands has been taken by many researchers to indicate that this dependency is mediated through movement (see, e.g., Nishigauchi 1986, 1990; Watanabe 1992a, 1992b; Maki 1995; von Stechow 1996; Richards 1997, 2000, 2001; Hagstrom 1998; Ochi 1999; Tanaka 1999). A challenge for the movement analysis, then, is to explain the lack of all other island constraints. An influential syntactic proposal in this area was made by Nishigauchi (1986, 1990) and Choe (1987), to which we now turn.

14.4.1 Covert Pied-Piping

A special mechanism that allows covert pied-piping of a whole island was proposed to account for the lack of island effects other than the wh-island effect (Nishigauchi 1986, 1990; Choe 1987). According to the pied-piping analysis, a whole island embedding an indeterminate pronoun counts as a wh-phrase due to feature percolation, and it is this big constituent that undergoes covert movement. The key in this proposal is that indeterminate pronouns within islands are never extracted out of them throughout the derivation, thus circumventing island violation. Though the authors cited in the previous paragraph do not necessarily agree on what moves at what level, most of them rely on some form of pied-piping analysis. For precise syntactic mechanisms of pied-piping, see Nishigauchi 1986, 1990 and Richards 2000.5

(p. 379) 14.4.2 Pied-Piping and Semantic Interpretation

From the perspective of syntax-semantics mapping, however, the pied-piping analysis comes with an undesirable consequence and therefore needs to be augmented with reconstruction of the pied-piped material, as discussed in detail in von Stechow 1996 (see also Chomsky 1977; Heim 1987; and Fiengo, Huang, Lasnik, and Reinhart 1988). Below I summarize von Stechow's main point—namely, that the LF structure with pied-piping is not suitable for deriving appropriate semantic interpretation for wh-questions, and I show how the revised pied-piping analysis solves the problem.6

According to the movement analysis, a wh-question such as (13a) has a Logical Form such as (13b).

  1. (13)

    1. a.

      Yoko-ga

      dono

      hon-o

      yomimasita

      ka?

      Yoko-nom

      which

      book-acc

      read

      q

      ‘Which book did Yoko read?’

    2. b. [cp [dono hon-o]1 [tp Yoko-ga t1 yomimasita] ka]

This LF has basically the same structure as the LF of the English wh-question Which book did Yoko read? and thus a standard interrogative semantics can be derived compositionally following a standard procedure proposed for its English counterpart. That is, we can assume, for example, that indeterminates introduce existential quantifiers that bind variables created by their movement, and that there is an interrogative operator, say in C0, that forms a set of propositions that are possible answers to the question, along the lines of Hamblin (1973) and Karttunen (1977).7 The formula in (14), a suitable semantics for the question in (13a), can then be read off from the LF in (13b).
  1. (14) λp∃x[book(x)(w) & p = λw′.read(x)(Yoko)(w′)]

The question denotes a set of propositions of the form Yoko read x, where x is a book in the actual world.

When an indeterminate pronoun is embedded in an island other than a wh-island, the whole island undergoes covert movement, according to the pied-piping analysis.8 In (15a), for example, where dare ‘who’ occurs inside the complex NP dare-ga kaita hon ‘book that who wrote’, the LF structure is derived by moving this entire object, as shown in (15b). The LF in (15b) also shows the movement of dare ‘who’ within the complex NP. It moves within the relative clause according to Nishigauchi (1986, 1990, 1999) and outside the relative clause according to Richards (2000). For Nishigauchi, the purpose of this type of internal movement is to make the wh-feature percolation from dare ‘who’ possible (see also van Riemsdijk 1984).

  1. (15)

    1. a.

      Kimi-wa

      [[dare-ga

      kaita]

      hon]-o

      yomimasita

      ka?

      you-top

      who-nom

      wrote

      book-acc

      read

      k

      ‘Whox did you read a book that x wrote?’

    2. b.

      [[[dare-ga]1

      t1

      kaita]

      hon-o]2

      kimi-wa

      t2

      yomimasita

      ka

      who-nom

      wrote

      book-acc

      you-top

      read

      q

(p. 380) It is proposed by Nishigauchi (1986, 1990) that variables introduced by both dare ‘who’ and hon ‘book’ are bound unselectively by ka and the question is claimed to receive the paraphrase in (16a), which can be translated into (16b).
  1. (16)

    1. a. For which x, y, x a book, y a person that wrote x, did you read x?

    2. b. λp∃x∃y[book(x)(w) & person (y)(w) & wrote(x)(y)(w) & p = λw′.read (x)(you)(w′]

As von Stechow (1996) points out, this semantic interpretation derived from the pied-piped LF structure incorrectly predicts that the sentence in (17a) is a possible answer to the question.
  1. (17)

    1. a.

      Umibe-no

      Kafuka-o

      yomimasita.

      shore-gen

      Kafka-acc

      read

      ‘(I) read Kafka on the Shore

    2. b.

      Murakami-ga

      kaita

      hon-o

      yomimasita.

      Murakami-nom

      wrote

      book-acc

      read

      ‘(I) read a book that Murakami wrote.’

A possible answer to this question should in fact have the form represented in (17b). Given this fact, the denotation of the question should be (18), which can be derived from the LF in (19).
  1. (18) λp.∃x[person (x)(w) & p = λw′.∃y[book(y)(w′) & wrote(y)(x)(w′) & read (y)(you)(w′)]]

  2. (19)

    [dare-ga]1

    [[[t1

    kaita]

    hon-o]2

    kimi-wa

    t2

    yomimasita

    ka]

    who-nom

    wrote

    book-acc

    you-top

    read

    q

According to von Stechow, a series of three covert movements are involved in the derivation of this LF structure. First, the entire island dare-ga kaita hon-o book that who wrote’ undergoes covert pied-piping. At the second step, where no island constraints are operative, dare-ga who’ is extracted out of the island and adjoined to it.9 Finally the pied-piped material [t kaita hon-o] is reconstructed to an interpretable position in the nucleus (this step does not leave a trace).

One might wonder why we even bother to covertly move the entire island to begin with, if most of it should be reconstructed for interpretational purposes. This is necessi-tated by the fundamental idea in the movement analysis that the locality effect we observed in section 14.3.2 is an indication of a movement relation between indeterminates and the particle ka or mo. One is then forced to assume that wh-movement takes place in Japanese at a level where island constraints do apply.10

14.4.3 A Consequence for the Universal Construction

So far in this section we have focused on indeterminate pronouns in the interrogative context. In particular, we have seen how the movement analysis, coupled with the amended pied-piping analysis, accounts for the puzzling pattern of (p. 381) island effects and derives appropriate interpretations. If we are to seek a general theory of indeterminate pronouns, and in particular, a uniform analysis of the interrogative and universal constructions, we want this analysis to extend to the universal construction. As becomes clear shortly, this turns out to be a difficult task. Therefore, if one insists on the movement analysis sketched for the interrogative construction, a uniform analysis of the interrogative and universal constructions has to be given up, which goes against Nishigauchi's (1986, 1990) original insight.

The main problem lies in the fact that the LF structure derived for the nonlocal universal construction following the successive steps assumed in the amended pied-piping analysis is not suitable for appropriate semantic interpretation. An example of the nonlocal universal construction is repeated here:

  1. (20)

    1. a.

      [Dare-ga

      katta

      ie]-mo

      takakatta.

      who-nom

      bought

      house-mo

      was.expensive

      ‘For everyone x, the house that x bought was expensive.

    2. b.

      [[dare-ga]1

      [[t1 katta

      ie]-mo]]

      takakatta.

      who-nom

      bought

      house-mo

      was.expensive

Following the steps described above, [dare-ga katta ie] ‘house that who bought’ first undergoes covert pied-piping to the specifier position of the phrase headed by mo. Then the indeterminate dare-ga ‘who’ is extracted out of the island and adjoined to it, and finally the pied-piped part undergoes reconstruction, which results in the LF in (20b).

The semantic interpretation we want for the quantified sentence in (20a) is (21), disregarding tense and possible worlds (Ohno 1989, von Stechow 1996).11

  1. (21) ∀x[person(x) → expensive(ɩy[house(y) & person(x) & buy(y)(x)])]

The LF structure in (20b), however, cannot be transparently mapped to the semantic representation (21) given that it does not have the typical tripartite structure that corresponds to the semantic interpretation. A technical device for the purpose of deriving (21) from the LF in (20b) is indeed proposed by von Stechow (1996), which I cannot go into here for reasons of space (see Shimoyama 2006a for a brief review). The proposed device, however, is not an attractive solution, relying on ad hoc assumptions made only for this type of sentence and hence not well motivated in the grammar of Japanese.

14.4.4 More Questions Surrounding Covert Pied-piping

There are many other issues surrounding the pied-piping analysis that cannot be fully discussed in this survey. For example, for the amended pied-piping analysis to be complete, pied-piped wh-islands should be blocked in a principled manner. Richards (2000, this volume) addresses this important question and tries to derive why pied-piping of wh-islands does not give rise to grammatical sentences. Readers (p. 382) are referred to his work, where a connection is made between the non-pied-pipability of wh-islands in overt wh-movement in Basque and the putative non-pied-pipability of wh-islands in covert wh-movement in Japanese (see also Ortiz de Urbina 1990 and Arregi 2003).

Richards (2000) also presents examples similar to (22) as an argument for covert pied-piping in Japanese. The example in (22) is taken from Watanabe (1992a:fn. 9), where the observation is attributed to Mamoru Saito (pers. comm.).

  1. (22)

    John-wa

    [Mary-ga

    [[nani-o

    doko-de

    katta]

    hito]-o

    John-top

    Mary-nom

    what-acc

    where-loc

    bought

    person-acc

    sagasiteiru

    ka]

    siritagatteiru

    no?

    looking.for

    q

    want.to.know

    q

    1. a. [cp…[cp…[cnp…indeterminate1…indeterminate2…] …q2] …q1]

      ‘For which x, x a thing, John wants to know for which y, y a place, Mary is looking for the person who bought x at y?’

    2. b. [cp…[cp…[cnp… indeterminate1…indeterminate2…] …q] …q1,2]

      ??/?‘For which x, x a thing, for which y, y a place, John wants to know whether Mary is looking for the person who bought x at y?’12

    3. c. [cp…[cp…[cnp…indeterminate1…indeterminate2…] …q1,2] …q]

      ‘Does John want to know for which x, x a thing, for which y, y a place, Mary is looking for the person who bought x at y?’

Readings (a) and (b) both involve wh-island violations. The observation is that reading (a), where the two indeterminates in the same complex NP are assigned distinct scopes, is worse than typical wh-island violations.

It is claimed in Watanabe 1992a and Richards 2000 that the contrast in grammaticality between (22a) and (22b) is accounted for straightforwardly by assuming that covert pied-piping takes place in Japanese. One crucial assumption there is that the reading in (22a) cannot be derived without extracting one indeterminate, in this case, nani ‘what’, out of the complex NP, whereas the reading in (22b) can be derived without extracting any indeterminate from the complex NP. Hence, an additional island effect is detected for (22a) but not for (22b), where only the wh-island effect is detected.

However, given that the original pied-piping analysis needs to be augmented with subsequent wh-extraction out of islands and reconstruction for the semantic reasons discussed above, extraction out of complex NPs always takes place, even in the same scope readings. That is, extraction out of the complex NP occurs not just in (22a) but in (22b) as well. Thus, the degradedness of the distinct scope reading does not follow simply from the assumption that covert pied-piping takes place in Japanese.

Consider, for instance, the derivation of the reading in (22a) in which the complex NP first moves to the lower scope position, and then indeterminate2 is extracted out of the complex NP. From this second step on, no island constraints are operative in the amended pied-piping analysis. Then, indeterminate1 can move (p. 383) out of the complex NP to the matrix scope position without any island violation (see also Richards, this volume). Thus, covert pied-piping and the distinct scope reading do not necessarily exclude each other in the amended pied-piping analysis.13

The contrast between (22a) and (22b) does not seem to follow, either, if we further augment the amended pied-piping analysis so that Subjacency needs to be satisfied only once per C0, as briefly touched on in note 9. Starting from a derivation of (22b), the movement of the complex NP to the matrix scope position incurs a wh-island violation. Because Subjacency has not yet been obeyed at the matrix Q level, the subsequent extraction of the two indeterminates incurs two complex NP island violations. In a derivation of (22a), it is possible to first move the complex NP to the lower scope position and extract indeterminate2 without incurring any violation. Then, we either further move the complex NP to the higher scope position, extract indeterminate1, and reconstruct the island, or move only indeterminate1 to the higher scope position out of the complex NP. Either way, we have at most one wh-island violation and one complex NP violation.14

The upshot is that in the amended pied-piping analysis, a possible source of the contrast between (22a) and (22b) cannot be as simple as the former having one extra island violation. That a possible source of the contrast is, in fact, of a more general nature than having to do with a complex NP island violation is suggested by the following example. Example (23) is a slight variation of example (22): The complex NP in the original example has now been replaced by a simple NP, dare-no nan-nen-no syasin ‘a photo of whom from what year’. The pattern of grammaticality stays more or less the same in this example.

  1. (23)

    John-wa

    [Mary-ga

    [dare-no

    nan-nen-no

    syasin]-o

    John-top

    Mary-nom

    who-gen

    what-year-gen

    photo-acc

    sagasiteiru

    ka]

    siritagatteiru

    no?

    looking.for

    q want.to.know

    q

    1. a. For which x, x a person, John wants to know for which y, y a year, Mary is looking for a photo of x from y?’

    2. b. ??/?‘For which x, x a person, for which y, y a year, John wants to know whether Mary is looking for a photo of x from y?’

    3. c. ‘Does John want to know for which x, x a person, for which y, y a year, Mary is looking for a photo of x from y?’

More research is needed to provide a satisfactory answer to the question as to where the contrast in the (a) and (b) examples stems from in (22) and (23).

(p. 384) 14.5 More Issues and Controversies Regarding Analysis: The Nonmovement Analysis and wh-island

As I have discussed, the movement analysis takes the presence of the wh-island effect to be indicative of a movement relation between indeterminates and higher particles and attempts to provide an account for the lack of other island effects. Alternatively, one could consider the lack of most island effects to be evidence that there is no such movement relation, as proposed in Toyoshima 1996 and Shimoyama 1999, 2001, for example. The challenge now is turned around: the presence of the wh-island effect requires an explanation.

14.5.1 The Hamblin Analysis and Locality

To see how the apparent wh-island effect can be accounted for without appealing to a movement relation between indeterminates and particles, we should first address what the semantic denotations of indeterminate pronouns are. Two proposals are found in this domain—one says that indeterminate pronouns denote sets of alternatives, and the other says that they denote open propositions. In what follows I first summarize the former proposal and present how it accounts for the apparent wh-island effect without wh-movement. I then summarize the latter proposal and evaluate its way of handling the apparent wh-island effect.

The idea that indeterminate pronouns denote sets of alternatives is based on a proposal in Hamblin 1973 that English wh-pronouns denote sets of alternatives. For example, who denotes the set of all persons. The application of Hamblin's semantics to indeterminate pronouns is found in Ramchand 1997 for Bengali, as well as in Lin 1996 for the Chinese wulun ‘no matter’-construction. For Japanese, a Hamblin semantics is adopted by Hagstrom (1998) for indeterminate pronouns that occur in the interrogative and existential contexts. Its extension to the universal context is suggested in Shimoyama 2001 and more fully explored in Kratzer and Shimoyama 2002. Starting from the interrogative context, dare ‘who’ in (24) denotes the set of all people (with appropriate contextual restriction) as in (25a). The VP utaimasu ‘sing’, in contrast, denotes a singleton set whose sole member is its ordinary denotation, as shown in (25b).

  1. (24)

    [Dare-ga

    utaimasu]

    ka?

    who-nom

    sing

    q

    ‘Who sings?’

  2. (25) For all possible worlds w and variable assignments g,

    1. a. 〚dare〛w,g = {x ∈ De: person(x)(w)}

    2. b. 〚utaimasu〛w,g = {λxλw′[sing(x)(w′)]}

(p. 385) In composing these two pieces, we apply the function λxλw′ [sing(x)(w′) to each of the members in the set denoted by dare ‘who’ and create a bigger set—namely, the set of propositions in (26).
  1. (26) 〚dare-ga utaimasu〛w,g = {f(x): f ∈ 〚utaimasu〛w,g, x ∈ 〚dare〛w,g} = {λw′[sing(x)(w′)]: person(x)(w)}

Thus we arrive at an appropriate basic form of interrogative semantics.

The question now is whether this view of wh-questions due to Hamblin can make a uniform analysis possible with the universal construction. A uniform analysis in fact becomes possible if we leave our standard conception of the nonlocal universal construction. The nonlocal universal construction is standardly assumed to involve quantification at a distance or some kind of inverse linking as represented in (28) (see, for example, Ohno 1989, von Stechow 1996, and Takahashi 2002).

  1. (27)

    [Dare-ga

    katta

    ie]-mo

    takakatta.

    who-nom

    bought

    house-mo

    was.expensive

    ‘For everyone x, the house that x bought was expensive.’

  2. (28) ∀x[person(x) → expensive(ιy[house(y) & person(x) & buy(y)(x)])]

Alternatively, one could conceive the nonlocal universal construction as involving a rather direct quantification, as proposed in Shimoyama 1999, 2001. More specifically, we could assume that the restriction of the universal quantification is not provided by the embedded indeterminate pronoun but is provided directly by the sister constituent as a whole as in (29).15
  1. (29) ∀x[x ∈ {ιy[house(y) &buy(y)(z): person(z)} → expensive(x)]

The domain of quantification in (29), the set of unique individuals that are houses each person bought, is exactly what is obtained through successive pointwise functional application. We thus obtain a uniform analysis of the interrogative and universal constructions based on a Hamblin semantics of indeterminate pronouns.

In the Hamblin analysis of indeterminate pronouns, the relation between indeterminates and particles is not mediated through movement. Rather, the relation is established by the fact that indeterminate pronouns introduce Hamblin alternatives, which expand as they compose with the denotations of other constituents, and the particles ka and mo select such denotations. In this mode of association, the lack of island effects is expected.

Crucially, the expansion of the Hamblin alternatives introduced by indeterminates is stopped when the first operator that selects Hamblin alternatives is encountered. The particles ka and mo take the Hamblin alternatives and return singleton sets, thereby making the alternatives created by indeterminates inaccessible to even higher particles.16 This is how the locality constraint illustrated in (12) follows, whether there is a single indeterminate or multiple indeterminates in the scope of the lower particle. Other long-distance associations across, for (p. 386) example, complex NP islands, are allowed as desired. Thus the puzzling locality pattern simply comes out as a consequence of the interpretation system.17

14.5.2 Indeterminate Pronouns as Open Propositions

A more widely held view on the semantics of wh-pronouns or indeterminate pronouns (other than the one in which they introduce existential quantifiers) is that they introduce free variables and denote open propositions. This type of proposal for English wh-phrases is explored by Baker (1970), Pesetsky (1987), Berman (1991), and Reinhart (1998) among many others. The idea builds on influential work by Kamp (1981) and Heim (1982) on indefinites in English. As for Japanese indeterminate pronouns, Kuroda (1965:101) was the first to observe that “the role of the indeterminate pronouns [is] very much like that of yet unbounded logical variables,” and his insight was pursued further by Nishigauchi (1986, 1990). The idea has been adopted for other languages, such as Chinese (Aoun and Li 1993, Tsai 1994, and Lin 1996, among others) and Malayalam (Jayaseelan 2001).

Although the Hamblin analysis of indeterminates introduced above made in-situ interpretation of indeterminates possible, the idea that indeterminates denote open propositions makes semantic interpretation possible with short movement of indeterminates and without movement to the specifier of CP or of the phrase headed by mo. As is discussed shortly, a uniform analysis of the interrogative and universal construction is achieved under the assumption that the semantics of the nonlocal universal construction is perceived as proposed in Shimoyama 1999, 2001, introduced in section 14.5.1. Thus similar results are achieved in the Hamblin analysis, on the one hand, and in the open proposition analysis, on the other hand. When we consider how the locality puzzle is dealt with within the latter analysis, a difference emerges. Let us first see the basic syntax-semantics mapping in the open proposition analysis and then consider how the locality puzzle is accounted for.

Following Heim (1982), we can assume that dare ‘who’ in (30a) and (31a), for example, receives index 1 and is interpreted as: the individual assigned to variable 1 is a person. Because this is not of the suitable semantic type to be combined with utaimasu ‘sing’ or katta ‘bought’, dare ‘who’ has to move to an interpretable position. For example, moving out of VP (normal subject movement) or adjoining to TP would do (see Heim 1982).

  1. (30)

    1. a.

      [Dare-ga

      utaimasu]

      ka?

      who-nom

      sing

      q

      ‘Who sings?’

    2. b.

      [[dare1-ga

      [TP/VP t1

      utaimasu]]

      Op1] ka

  2. (31)

    1. a.

      [Dare-ga

      katta

      ie]-mo

      takakatta.

      who-nom

      bought

      house-mo

      was.expensive

      ‘For everyone x, the house that x bought was expensive.’

    2. b.

      [[[dare1-ga

      [TP/VP t1 katta]]

      ie]

      Op1]-mo

      takakatta.

(p. 387) It is proposed in Shimoyama 2001 that interpretability forces an introduction of a set creating operator (Op) at LF, which gets copies of indices and is interpreted as in (32) along the lines of Heim 1982 and Berman 1991. The part g′ ≈ 1,…, ng in (32b) says that assignment g′ is just like assignment g, except possibly for values assigned to the variables 1,…, n.
  1. (32)

    1. a. Op-indexing: Copy the index of each indeterminate pronoun onto the c-commanding Op.

    2. b. 〚XP Op1,…, n]g = {〚XP〛g′ :g′ ≈ 1,…, ng} = {z: ∃g′[g′ ≈ 1,…, n g & z = 〚XP〛g′]}

In the LFs in (30b) and (31b), for instance, Op combines with its sister constituent and forms the set of propositions {that x is a person and x sings: x ∈ D} and the set of individuals {ιy[house(y) & person(x) & buy(y)(x)]: x ∈ D}, respectively, by abstracting over the free variable that it is coindexed with. The steps are shown in (33) and (34).18
  1. (33)〚 [dare1-ga [t1 utaimasu]] Op1g

    = {〚 [dare1-ga [t1 utaimasu]] Op1g′: g ≈1g′}

    = {p: ∃x[p = λw′[person(x)(w′) & sing(x)(w′)]}

    = {λw′[person(x)(w′) & sing(x)(w′)]: x ∈ D}

  2. (34) 〚 [[dare1-ga [t1 katta]] ie] Op1〛g

    = {〚[dare1-ga [t1 katta]] ie ]g′: g ≈ 1g′}

    = {z: ∃x[z = ιy[house(y) & person(x) & buy(y)(x)]]}

    = {ιy[house(y) & person(x) & buy(y)(x)]: x ∈ D}

Once the denotation in (34) is available, the universal quantifier mo can directly quantify over it, giving us the appropriate semantics.

So far, the open-proposition analysis does not seem to differ significantly from the Hamblin analysis—both derive appropriate semantics from LF structures. Now, though, I show a difference. Recall that in the Hamblin analysis, the puzzling locality pattern simply falls out from the interpretation mechanism. In the open-proposition analysis, however, a Relativized Minimality-type locality principle (Rizzi 1990) needs to be stated independently. That is, the Op-indexing rule in (32a) should be modified to: “Copy the index of each indeterminate pronoun onto the lowest c-commanding Op.” Note that Heim's (1982) rule of Quantifier Indexing, which the Op-indexing rule builds on, also has the locality clause built in. The purpose there is to block a certain LF representation that corresponds to an unavailable reading in sentences with indefinites and adverbs of quantification. This way of looking at the wh-island effect is specifically proposed in Toyoshima 1996 and Shimoyama 1999, 2001 and in a more general form in works such as Ochi 1999, Tanaka 1999, and Yoshida 1999. The Hamblin analysis fares better in this respect, because any independent statement of such locality principles is made unnecessary.19 For further exploration and consequences of the Hamblin analysis, see Kratzer and Shimoyama 2002 and Kratzer 2005.20

(p. 388) 14.6 Implications and Open Questions

I have reviewed two major approaches (i.e., movement vs. nonmovement) to the nonlocal association of indeterminates and particles. The focus of the discussion on association in the literature has been on the interrogative and universal contexts—two contexts where nonlocal association is most naturally allowed. The nature of limitations in nonlocal association in other series (existential, NPI, free choice) is yet to be investigated carefully, and the question whether the reviewed proposals can be extended to handle the syntax and semantics of the Japanese indeterminate pronoun system in general, as well as the indeterminate pronoun system in natural language, remains to be examined.

Another important question that has been addressed in the literature but needs more work is whether various uses of ka and mo should or can be related to one another. As is well known, ka is not just a question particle but also marks existential quantification (e.g., nani-ka ‘something’) as well as disjunction (e.g., Hanako-ka Taro[-ka] ‘Hanako or Taro’). Mo is used to mark universal quantification but can also be used to mean even or also when there is no indeterminate in its scope.21 Although works such as Hagstrom 1998 and Suzuki 2003 take up this question, we are still far from a good understanding of it.

References

Aoun, Joseph, and Yen-hui Audrey Li. 1993. Wh-elements in situ: Syntax or LF? Linguistic Inquiry 24:199–238.Find this resource:

    Arregi, Karlos. 2003. Clausal pied-piping. Natural Language Semantics 11:115–143.Find this resource:

      Baker, C. L. 1970. Notes on the description of English questions: The role of an abstract question morpheme. Foundations of Language 6:197–219.Find this resource:

        Beck, Sigrid. 2007. The grammar of focus interpretation. In Recursion + Interfaces = Language? Chomsky's minimalism and the view from syntax and semantics, ed. Uli Sauerland and Hans-Martin Gärtner, 255–280. Berlin: Mouton de Gruyter.Find this resource:

          Beck, Sigrid, and Hotze Rullmann. 1999. A flexible approach to exhaustivity in questions. Natural Language Semantics 7:249–298.Find this resource:

            Berman, Stephen. 1991. On the semantics and Logical Form of wh-clauses. Doctoral dissertation, University of Massachusetts, Amherst.Find this resource:

              Brody, Michael. 1995. Lexico-logical form: A radically minimalist theory. Cambridge, Mass.: MIT Press.Find this resource:

                Choe, Jae W. 1987. LF movement and pied-piping. Linguistic Inquiry 18:348–353.Find this resource:

                  Chomsky, Noam. 1977. On wh-movement. In Formal syntax, ed. Peter Culicover, Thomas Wasow, and Adrian Akmajian, 71–132. New York: Academic Press.Find this resource:

                    Chomsky, Noam. 1986. Barriers. Cambridge, Mass.: MIT Press.Find this resource:

                      (p. 391) Deguchi, Masanori, and Yoshihisa Kitagawa. 2002. Prosody and wh-questions. In Proceedings of NELS 32, ed. Masako Hirotani, 73–92. Amherst, Mass.: GLSA Publications.Find this resource:

                        Fiengo, Robert, C.-T. James Huang, Howard Lasnik, and Tanya Reinhart. 1988. The syntax of wh-in-situ. In Proceedings of WCCFL 7, ed. Hagit Borer, 81–98. Stanford, Calif.: CSLI Publications.Find this resource:

                          Groenendijk, Jeroen, and Martin Stokhof. 1982. Semantic analysis of wh-complements. Linguistics and Philosophy 5:175–233.Find this resource:

                            Hagstrom, Paul. 1998. Decomposing questions. Doctoral dissertation, MIT, Cambridge, Mass.Find this resource:

                              Hamblin, C. L. 1973. Questions in Montague English. Foundations of Language 10:41–53.Find this resource:

                                Haspelmath, Martin. 1997. Indefinite pronouns. Oxford: Oxford University Press.Find this resource:

                                  Heim, Irene. 1982. The semantics of definite and indefinite noun phrases. Doctoral dissertation, University of Massachusetts, Amherst.Find this resource:

                                    Heim, Irene. 1987. Where does the definiteness restriction apply? Evidence from the definiteness of variables. In The representation of (in)definiteness, ed. Alice ter Meulen and Eric Reuland, 21–42. Cambridge, Mass.: MIT Press.Find this resource:

                                      Hirotani, Masako. 2004. Prosody and LF: Processing Japanese wh-questions. Doctoral dissertation, University of Massachusetts, Amherst.Find this resource:

                                        Hoji, Hajime. 1985. Logical Form constraints and configurational structures in Japanese. Doctoral dissertation, University of Washington, Seattle.Find this resource:

                                          Huang, C.-T. James. 1982. Logical relations in Chinese and the theory of grammar. Doctoral dissertation, MIT, Cambridge, Mass.Find this resource:

                                            Ishihara, Shinichiro. 2003. Intonation and interface conditions. Doctoral dissertation, MIT, Cambridge, Mass.Find this resource:

                                              Jayaseelan, K. A. 2001. Questions and question-word incorporating quantifiers in Malayalam. Syntax 4:63–93.Find this resource:

                                                Kamp, Hans. 1981. A theory of truth and semantic representation. In Formal methods in the study of language (Mathematical Centre Tracts 135), ed. J. Groenendijk, T. Janssen, and M. Stokhof. 277–322. Amsterdam: Mathematisch Centrum.Find this resource:

                                                  (Reprinted in

                                                  Truth, interpretation, and information, ed. J. Groenendijk, T. Janssen, and M. Stokhof, 1–41. Dordrecht: Foris, 1984.Find this resource:

                                                    )

                                                    Karttunen, Lauri. 1977. Syntax and semantics of questions. Linguistics and Philosophy 1:3–44.Find this resource:

                                                      Ko, Heejeong. 2005. Syntax of wh-in-situ: Merge into [Spec,CP] in the overt syntax. Natural Language & Linguistic Theory 23:867–916.Find this resource:

                                                        Kratzer, Angelika. 2005. Indefinites and the operators they depend on: From Japanese to Salish. In Reference and quantification: The Partee effect, ed. Gregory N. Carlson and F. Jeffry Pelletier, 113–142. Stanford, Calif.: CSLI Publications.Find this resource:

                                                          Kratzer, Angelika and Junko Shimoyama. 2002. Indeterminate pronouns: The view from Japanese. In The Proceedings of the 3rd Tokyo Conference on Psycholinguistics, ed. Yukio Otsu, 1–25. Tokyo: Hituzi Syobo.Find this resource:

                                                            Kuroda, S.-Y. 1965. Generative grammatical studies in the Japanese language. Doctoral dissertation, MIT, Cambridge, Mass.Find this resource:

                                                              Lin, Jo-wang. 1996. Polarity licensing and wh-phrase quantification in Chinese. Doctoral dissertation, University of Massachusetts, Amherst.Find this resource:

                                                                Maki, Hideki. 1995. The syntax of particles. Doctoral dissertation, University of Connecticut, Storrs.Find this resource:

                                                                  (p. 392) Nishigauchi, Taisuke. 1986. Quantification in syntax. Doctoral dissertation, University of Massachusetts, Amherst.Find this resource:

                                                                    Nishigauchi, Taisuke. 1990. Quantification in the theory of grammar. Dordrecht: Kluwer.Find this resource:

                                                                      Nishigauchi, Taisuke. 1999. Quantification and wh-constructions. In The handbook of Japanese linguistics, ed. Natsuko Tsujimura, 269–296. Oxford: Blackwell.Find this resource:

                                                                        Ochi, Masao. 1999. Constraints on feature checking. Doctoral dissertation, University of Connecticut, Storrs.Find this resource:

                                                                          Ohno, Yutaka. 1989. Mo.In Papers on quantification, NSF Grant Report, Department of Linguistics, University of Massachusetts, Amherst, ed. Emmon Bach, Angelika Kratzer, and Barbara H. Partee, 224–250. Amherst, Mass.: GLSA Publications.Find this resource:

                                                                            Ohno, Yutaka. 1991. Arguments against unselective binding in Korean. In Harvard Studies in Korean Linguistics 4, ed. Susumu Kuno et al. 553–562. Seoul: Hanshin.Find this resource:

                                                                              Ortiz de Urbina, Jon. 1990. Operator feature percolation and clausal pied-piping. In MITWPL 13: Papers on wh-movement, ed. Lisa Cheng and Hamida Demirdache, 193–208. Cambridge, Mass.: MIT Working Papers in Linguistics.Find this resource:

                                                                                Pesetsky, David. 1987. Wh-in-situ: Movement and unselective binding. In The representation of (in)definiteness, ed. Eric Reuland and Alice ter Meulen, 98–129. Cambridge, Mass.: MIT Press.Find this resource:

                                                                                  Ramchand, Gillian C. 1997. Questions, polarity, and alternative semantics. In Proceedings of NELS 27, ed. Kiyomi Kusumoto, 383–396. Amherst, Mass.: GLSA Publications.Find this resource:

                                                                                    Reinhart, Tanya. 1998. Wh-in-situ in the framework of the Minimalist Program. Natural Language Semantics 6:29–56.Find this resource:

                                                                                      Richards, Norvin. 1997. What moves where when in which language? Doctoral dissertation, MIT, Cambridge, Mass.Find this resource:

                                                                                        Richards, Norvin. 2000. An island effect in Japanese. Journal of East Asian Linguistics 9:187–205.Find this resource:

                                                                                          Richards, Norvin. 2001. Movement in language: Interactions and architectures. Oxford: Oxford University Press.Find this resource:

                                                                                            Riemsdijk, Henk van. 1984. On pied-piped infinitives in German relative clauses. In Studies in German grammar, ed. Jindrich Toman, 165–192. Dordrecht: Foris.Find this resource:

                                                                                              Riemsdijk, Henk van, and Edwin Williams. 1981. NP-structure. The Linguistic Review 1:171–217.Find this resource:

                                                                                                Rizzi, Luigi. 1990. Relativized minimality. Cambridge, Mass.: MIT Press.Find this resource:

                                                                                                  Saito, Mamoru. 1992. Long distance scrambling in Japanese. Journal of East Asian Linguistics 1:69–118.Find this resource:

                                                                                                    Saito, Mamoru. 1994. Additional-wh effects and the adjunction site theory. Journal of East Asian Linguistics 3:195–240.Find this resource:

                                                                                                      Sauerland, Uli, and Fabian Heck. 2003. LF-intervention effects in pied-piping. In Proceedings of NELS 23, ed. Makoto Kadowaki and Shigeto Kawahara, 347–366. Amherst, Mass.: GLSA Publications.Find this resource:

                                                                                                        Sharvit, Yael. 1998. Possessive wh-expressions and reconstruction. In Proceedings of NELS 28, ed. Pius N. Tamanji and Kiyomi Kusumoto, 409–423. Amherst, Mass.: GLSA Publications.Find this resource:

                                                                                                          Shimoyama, Junko. 1999. Complex NPs and wh-quantification in Japanese. In Proceedings of NELS 29, ed. Pius N. Tamanji, Masako Hirotani, and Nancy Hall, 355–365. Amherst, Mass.: GLSA Publications.Find this resource:

                                                                                                            Shimoyama, Junko. 2001. Wh-constructions in Japanese. Doctoral dissertation, University of Massachusetts, Amherst.Find this resource:

                                                                                                              (p. 393) Shimoyama, Junko. 2006a. Indeterminate phrase quantification in Japanese. Natural Language Semantics 14:139–173.Find this resource:

                                                                                                                Shimoyama, Junko. 2006b. A QVE-like phenomenon in the Japanese universal construction and its implications. In In search of the essence of language science: Festschrift for Professor Heizo Nakajima on the occasion of his 60th birthday, ed. Yubun Suzuki, Keizo Mizuno, and Ken-Ichi Takami, 297–309. Tokyo: Hituzi Syobo.Find this resource:

                                                                                                                  Stechow, Arnim von. 1996. Against LF pied-piping. Natural Language Semantics 4:57–110.Find this resource:

                                                                                                                    Suzuki, Shogo. 2003. Additive mo. Ms., MIT, Cambridge, Mass.Find this resource:

                                                                                                                      Takahashi, Daiko. 2002. Determiner raising and scope shift. Linguistic Inquiry 33:575–615.Find this resource:

                                                                                                                        Tanaka, Hidekazu. 1999. LF wh-islands and the minimal scope principle. Natural Language & Linguistic Theory 17:371–402.Find this resource:

                                                                                                                          Torrego, Esther. 1985. On empty categories in nominals. Ms., University of Massachusetts, Boston.Find this resource:

                                                                                                                            Toyoshima, Takashi. 1996. LF Subjacency and stationary wh-in-situ. Ms., Cornell University, Ithaca, N.Y.Find this resource:

                                                                                                                              Tsai, Wei-tien Dylan. 1994. On economizing the theory of A-bar dependencies. Doctoral dissertation, MIT, Cambridge, Mass.Find this resource:

                                                                                                                                Watanabe, Akira. 1992a. Wh-in-situ, Subjacency, and chain formation. MIT occasional papers in linguistics 2. Cambridge, Mass.: MIT Working Papers in Linguistics.Find this resource:

                                                                                                                                  Watanabe, Akira. 1992b. Subjacency and s-structure movement of wh-in-situ. Journal of East Asian Linguistics 1:255–291.Find this resource:

                                                                                                                                    Watanabe, Akira. 2001. Wh-in-situ languages. In The handbook of contemporary syntactic theory, ed. Mark Baltin and Chris Collins, 203–225. Oxford: Blackwell.Find this resource:

                                                                                                                                      Wold, Dag E. 1996. Long distance selective binding: The case of focus. In Proceedings from SALT 6, ed. Teresa Galloway and Justin Spence, 311–328. Ithaca, N.Y.: CLC Publications.Find this resource:

                                                                                                                                        Yamashina, Miyuki and Christopher Tancredi. 2005. Degenerate plurals. In Proceedings of SuB 9, ed. Emar Maier, Corien Bary, and Janneke Huitink, 522–537. Nijmegen, the Netherlands: NCS.Find this resource:

                                                                                                                                          Yatsushiro, Kazuko. 2004. Mo and ka. Ms., University of Connecticut, Storrs.Find this resource:

                                                                                                                                            Yoshida, Tomoyuki. 1999. LF Subjacency effects revisited. In MITWPL 34: Papers on morphology and syntax: Cycle two, ed. Vivian Lin, Cornelia Krause, Benjamin Bruening, and Karlos Arregi, 1–34. Cambridge, Mass.: MIT Working Papers in Linguistics.Find this resource:

                                                                                                                                              Notes:

                                                                                                                                              (1.) Interestingly, somewhat similar speaker variations in Malayalam are reported by Jayaseelan (2001:72) with respect to the distance between a wh-word and a disjunction marker, which form an existential quantificational phrase.

                                                                                                                                              (2.) Naze ‘why’ cannot be associated with ka across complex NP or adjunct island for some reason, as shown in (i) and (ii). See Huang 1982, Saito 1994, Tsai 1994, Ochi 1999, and Ko 2005, for example.

                                                                                                                                              1. ((i))

                                                                                                                                                Hanako-wa

                                                                                                                                                [Kyoto-ni

                                                                                                                                                naze

                                                                                                                                                kita

                                                                                                                                                hito]-o

                                                                                                                                                yobimasi-ta

                                                                                                                                                ka?

                                                                                                                                                Hanako-top

                                                                                                                                                Kyoto -to

                                                                                                                                                why

                                                                                                                                                came

                                                                                                                                                person-acc

                                                                                                                                                invited

                                                                                                                                                q

                                                                                                                                                ‘Whyx did Hanako invite a person who came to Kyoto for reason x?’

                                                                                                                                              2. ((ii))

                                                                                                                                                Yumi-wa

                                                                                                                                                [Mika-ga

                                                                                                                                                naze

                                                                                                                                                kaetta-ra]

                                                                                                                                                okorimasu

                                                                                                                                                ka?

                                                                                                                                                Yumi-top

                                                                                                                                                Mika-nom

                                                                                                                                                why

                                                                                                                                                left-if

                                                                                                                                                get.angry

                                                                                                                                                q

                                                                                                                                                ‘Whyx would Yumi get angry if Mika left for reason x?’

                                                                                                                                              (3.) Certain cases of whether-island violations are known to improve with particular prosodic renditions (see, e.g., Deguchi and Kitagawa 2002, Ishihara 2003, and Hirotani 2004). However, prosody does not rescue all cases of wh-island violations, as can be seen from the examples to follow in the text. See Kratzer 2005 for speculation on how certain prosodic conditions might affect the scope of indeterminates.

                                                                                                                                              (4.) It is tempting to try to work out a unification of mo as universal and mo as ‘also’ (see Suzuki 2003). The (d)-readings in (10) and (11), however, are in principle available as long as the presupposition that comes with the lower mo as ‘also’ is satisfied. This suggests that the two instances of mo should be treated as distinct lexical items.

                                                                                                                                              (5.) It is proposed by Nishigauchi (1986, 1990) that examples such as (i) in note 2 are ruled out because of the category mismatch between naze ‘why’ and the head noun of the relative clause. It is not obvious how an analysis along those lines would extend to explaining why (ii) in note 2 is not possible, or why nominals such as dare ‘who’ and nani ‘what’ occur freely in adjunct clauses such as the if-clause in (ii).

                                                                                                                                              (6.) Sharvit (1998) and Watanabe (2001:222–223) attempt to assign semantic interpretation to the pied-piped structure. See Shimoyama 2006a (pp. 160–161) for comments on the latter attempt. See also Sauerland and Heck 2003.

                                                                                                                                              (7.) Alternatively, indeterminates may introduce free variables, which are bound by existential closure, as proposed for D-linked wh-in-situ in English and Japanese by Pesetsky (1987). Nishigauchi (1986, 1990) assumes that this interpretational device applies to covertly moved indeterminates. See von Stechow 1996 for the spelling out of this idea.

                                                                                                                                              (8.) This is not the case in Hagstrom's (1998) movement analysis of wh-questions in Japanese. There, it is the question particle ka that undergoes overt movement from the position next to a wh-phrase.

                                                                                                                                              (9.) The stipulation that island constraints do not apply at this second step—namely, in the derivation from what von Stechow refers to as WH-structure to LF—may not be necessary if we assume with Brody (1995) and Richards (1997, 2001) that Subjacency needs to be satisfied only once per C0 (as an effect of the Principle of Minimal Compliance in Richards's terms).

                                                                                                                                              (10.) In Watanabe's (1992a, 1992b) analysis, the movement relevant here is not the wh-movement at LF but the invisible operator movement at S-structure, whereas in Hagstrom's (1998) analysis, it is the overt movement of the particle ka/no. In the former analysis, subsequent LF wh-movement of the standard kind is assumed, including large-scale pied-piping, which raises the same problem as it does for Nishigauchi's (1986, 1990) LF.

                                                                                                                                              (11.) The domain of quantification is, in the end, narrower presumably due to accommodation of the existence presupposition of the definite description (see Shimoyama 2006a).

                                                                                                                                              (12.) This reading is marked “ok” in Watanabe (1992a:fn. 9), but it is at least as difficult to get as cases where there is no complex NP island in addition to a wh-island (see (7)). See also Richards 2000 (fn. 6) for a clarifying note on degrees of ungrammaticality.

                                                                                                                                              (13.) The following examples also show that pied-piping and distinct scope readings are, in principle, not incompatible with each other ((iii) is from van Riemsdijk and Williams 1981:196 and (iv) is from Fiengo, Huang, Lasnik, and Reinhart 1988:88). According to van Riemsdijk and Williams, whom in (iii) can take either matrix scope or embedded scope. (I thank Jon Nissenbaum for confirming this judgment.)

                                                                                                                                              1. ((iii)) Whoi ti knows [[which picture of whom]j Bill bought tj]?

                                                                                                                                              2. ((iv)) ?Whoi do you wonder [[which pictures of ti]j tj are on sale]?

                                                                                                                                              Fiengo and his colleagues also discuss Spanish examples similar to (iv) from Chomsky 1986 (pp. 25–26), originally observed by Torrego (1985).

                                                                                                                                              (14.) The first of the two subderivations is possibly blocked by an independent principle, given that there is a stage, before reconstruction, at which the trace ofindeterminate2 is not bound by its antecedent (see, e.g., Saito 1992).

                                                                                                                                              (15.) An extensional system is used here for simplicity. It remains to be seen whether truly parallel analyses can be given to the local and nonlocal cases of association with mo when the denotation is relativized to possible worlds.

                                                                                                                                              (16.) Ka presumably returns a singleton set whose sole member is a question denotation as proposed by Groenendijk and Stokhof (1982).

                                                                                                                                              (17.) For Hagstrom (1998), who adopts Hamblin denotations for indeterminates in the interrogative context, the source of the apparent wh-island effect is still movement.

                                                                                                                                              (18.) Note that, unlike in (26), the system derives the nontransparent interpretation of dare in (33). See Beck and Rullmann 1999 for discussion.

                                                                                                                                              (19.) See Wold 1996 for related discussion in the domain of association with focus. See also Beck 2007 for more recent discussion in a broader context.

                                                                                                                                              (20.) It does not necessarily follow that all indeterminate pronouns across languages should have a Hamblin semantics. For example, it is reported in Jayaseelan 2001 that Malayalam indeterminates can associate with a conjunction or disjunction particle across a closer particle. This suggests that this language presumably employs a selective version of the Op-indexing without the locality clause added.

                                                                                                                                              (21.) Nishigauchi (1990) and, more recently, Yamashina and Tancredi (2005) assume that mo lacks quantificational force on its own (see also Shimoyama 2006b).