Argument-Adjunct_Asymmetries.pdf

Viewer
Transcript

Interface Conditions on “Gaps”: Argument-Adjunct Asymmetries and Scope Reconstruction Jun Abe Workshop on Modality and Embedded Clauses YNU, December 23, 2015 1.

Introduction

- Argument-Adjunct Asymmetry (I): (1) a.??Who do you wonder whether John said t solved the problem? b. *How do you wonder whether John said [Mary solved the problem t]? (Chomsky and Lasnik 1993; Chomsky 1995, p. 87) è Chomsky and Lasnik (1993): “why violation of the economy condition is more severe for adjuncts than arguments” (p. 90) - Argument-Adjunct Asymmetry (II): cf. Huang (1982) (2) a.

[Hanako-ga

[Taroo-ga

Hanako-Nom

nani-o

te-ni iretta tte]

itta] koto-o

Taroo-Nom what-Acc obtained Comp

sonnani

okotteru no?

so much

be-angry Q

said fact-Acc

‘Lit. What are you so angry about [the fact that Hanako said that Taroo obtained t]?’ b. *[Hanako-ga

[Taroo-ga

Hanako-Nom

naze sore-o

Taroo-Nom why it-Acc

sonnani

okotteru no?

so much

be-angry Q

te-ni iretta tte]

itta] koto-o

obtained

said fact-Acc

Comp

‘Lit. Why are you so angry about [the fact that Hanako said that Taroo obtained it t]?’

(Lasnik and Saito 1984, p. 245)

Assuming the mechanism of *-marking proposed by Chomsky and Lasnik (1993), I propose a condition that requires that gaps be “licensed” at both AP and CI interfaces. (i) AP interface: only “real” gaps are subject to locality conditions. è These locality conditions could be evaded at PF by not creating gaps by string vacuous movement and resumption. 1

(ii) CI interface: only non-referential gaps are subject to locality conditions a la Rizzi (1990) 2.

Argument-Adjunct Asymmetry (I)

- Chomsky and Lasnik’s (1993) *-marking: (3) When a chain link formed by Move violates a locality condition, the bottom copy is assigned *. #The sentences in (1) are violations of Minimize chain links (MCL): (4) a. b.

[CP who do you wonder [CP whether John said * solved the problem]] [CP how do you wonder [CP whether John said Mary solved the problem *]]

#AP interface: cf. Abe and Hornstein (2012) (5) * is retained when the top copy of the resulting chain is pronounced; otherwise, remove *. #CI interface: (6) An argument copy may be replaced by pro. (7) What did you see yesterday? (8) [CP what did [TP you see yesterday]] ↓ replacement of with pro [CP what did [TP you see pro yesterday]] è Here pro insertion may be taken as a rule alternative to Fox’s (2002) trace conversion rule. (9) * is removed when the bottom copy of the resulting chain is replaced by pro. - Derivation of (1a): (10) [CP who do you wonder [CP whether John said * solved the problem]] AP interface: * CI interface: * -> pro - Derivation of (1b): (11) [CP how do you wonder [CP whether John said Mary solved the problem *]] AP interface: * CI interface: *

2

(12) a.

the man whoi they think that if Mary marries *(himi), then everyone will be happy.

b. I wonder whoi they think that if Mary marries *(himi), then everyone will be happy.

(Chomsky 1982, p. 11)

è Under this theory, overt pronoun insertion is a remedy strategy for both AP and CI interface conditions. - DP vs. non-DP asymmetries: cf. Cinque (1990) (13) a. ?Who did you leave without speaking to? b. *To whom did you leave without speaking? - Referential vs. non-referential: cf. Rizzi (1990) (14) a. ?What did John wonder how to weigh? b. Apples. /*200bls. 3.

String-Vacuous Movement and Locality: Abe and Hornstein (2012)

(15) * is retained when the top copy of the resulting chain is pronounced; otherwise, remove *. (16) John likes

and Bill hates

, the linguistic professor teaching Ling 101.

- Ross’s (1967) ATB movement analysis: (17) [Clause [Clause John likes

] and [Clause Bill hates

] [the linguistic professor

teaching Ling 101]] - Immune to locality conditions: (18) *Max said [that he was going to return

to the library] yesterday each of the

books that he checked out last week.

(Sabbagh 2007, p. 350)

(19) Josh said that he thought that I should sell that she might want to buy

, and Jamie said that she thought

, each of the Rambaldi artifacts that I have in

my attic.

(ibid., p. 358)

(20) I know someone who wants to buy sell

, and you know someone who wants to

, a copy of this manuscript.

(21) Josh wonders who bought

(ibid., p. 352)

, and Bill will find out who sold

Fred.

, pictures of (ibid., p. 382)

3

(22) Politicians win when they defend

, and lose when they attack

of a woman to an abortion.

, the right (ibid., p. 382)

(23) The head of a chain produced by Move cannot be pronounced unless it has an effect on PF output. (24) [ … DP1] and [… DP2] DP3

è Abe and Hornstein propose that while RNR involves overt ATB-movement as a default case, hence DP3 being the target of pronunciation, the PF condition stated in (23) dictates that DP2 must be pronounced when the chain (DP3, DP2) involves string-vacuous movement. (25) Locality conditions such as the RRC and the island conditions apply only to ‘overt’ movement. (26) *Joss said [that he was going to donate claimed [that she would donate

to the library] yesterday, and Jamie

to the museum] last week, a large

collection of ancient texts.

(ibid., p. 355)

- No exemption from locality violations in the case of adjunct RNR: (27) a.

John got fired

and Bill had his salary reduced

, because he

talked back/because of office politics. b.

*Mary helped the person who got fired who had his salary reduced

, because he talked back/because of

office politics.

(Abe and Hornstein 2012, p. 201)

(28) [ … *] and [… *Arg2] AP interface: * -> , *Arg2 -> Arg2 CI interface: * -> pro, *Arg2 -> pro (29) [ … *] and [… *Adj2] AP interface: * -> , *Adj2 -> Adj2 CI interface: *, *Adj2 4.

Argument-Adjunct Asymmetry (II)

(30) Anata-wa you-Top

kinoo

and comforted the person

dare-ni

aimasita ka?

yesterday who-Dat saw

‘Who did you see yesterday?’ 4

Q

- Covert wh-movement analysis: (31) [CP [TP anata-wa kinoo dare-ni aimasita] ka] - Derivation of (2a): (32) [CP [TP … [Complex NP … *nani-o …] …] Q] AP interface: *nani-o -> nani-o CI interface: *nani-o -> pro - Derivation of (2b): (33) [CP [TP … [Complex NP … *naze …] …] Q] AP interface: *naze -> naze CI interface: *naze 5.

Scope Reconstruction

(34) What did you see yesterday? (35) [CP what did [TP you see yesterday]] ↓ replacement of with pro [CP what did [TP you see pro yesterday]] - Scope reconstruction: (36) Who do you think everyone saw at the rally?

(May 1985, p. 45)

- Abe (1993): (i) interrogative phrases are composed of WH and indefinites. (ii) existentials can carry [Scope] optionally and if they do not, they are interpreted as referential in the sense of Fodor and Sag (1982), which means that they take the widest scope. (37) Who did everyone see? (38) a.

[CP who did [TP [TP everyone see ]]] [WH]

b.

[Scope]

[Scope]

[CP who did [TP everyone see ]] [WH]

[Scope]

(39) α c-commands β iff every category that dominates α contains β. where α contains β iff some segment of α dominates β. (40) a.

[WHx] [ ∀ y: person(y)] [ ∃ x: person(x)] y saw x

b. [WHx] [ ∃ x: person(x)] [ ∀ y: person(y)] y saw x 5

- Derivation of (36): (41) [CP who do [TP you think [CP [TP [TP everyone saw at the [WH]

[Scope]

[Scope]

rally]]]]] (42) [CP who did [TP everyone see ]] [WH]

[Scope]

(43) Pro replacement is possible for non-quantificational wh-phrases, i.e., those that lack [Scope] features. if an island intervenes in a wh-chain, then scope reconstruction should be impossible. -> Longobardi’s (1987) generalization (44) ??Who do you wonder whether everyone saw at the rally? (Lasnik and Saito 1992, p. 175) è This sentence “has the status of a Subjacency violation. But … the wide scope reading for everyone is impossible. (p. 176) (45) a.

[CP who do [TP you wonder [CP whether [TP * [TP everyone saw at [Scope]

[Scope]

the rally]]]]] AP interface: * CI interface: * b.

[CP who do [TP you wonder [CP whether [TP everyone saw * at the rally]]]]]

[Scope]

AP interface: * CI interface: * -> pro 6.

Semantic Approach to Scope Reconstruction: Ruys (2015)

- Cresti (1995), Rullman (1995): (46) Someone is likely [t to arrive]. (47) a.

someone (λx[likely(^arrive(x))])

t is x of type e

b. (^someone) λX[likely(^vX(arrive))] = likely (^someone(arrive))

6

t is X of type ,t>>

(48) Condition on Trace Typing (CTT) If D is a trace, then D is translated as a variable of some type τ. If D is attracted to a phase-accessible target B, and D is identical to the sister of B, then τ can be the type of any nontrace interpretation of D. Alternatively, τ can default to type e. (Ruys 2015, p. 461) (49) How many people do you think I should talk to t? a. for what n: there are n-many people x such that you think I should talk to x b. for what n: you think there should be n-many people that I talk to (50) [CP how many people do you [vP [vP think [CP I should [vP [vP talk to ]]]]]] (51) How many people do you wonder whether I should talk to t? a. for what n: there are n-many people x such that you wonder whether I should talk to x b. *for what n: you wonder whether there should be n-many people that I talk to (52) [CP how many people do you [vP [vP wonder [CP whether I should [vP [vP talk to ]]]]]] è “the underlined trace cannot access the type of the constituent that composes with its binder. It therefore defaults to type e.” (p. 463) - Derivation of (49): (53) a.

[CP how many people do you think [TP [TP I should talk to ]]]

[Scope]

(reading (49b))

b. [CP how many people do you think [TP I should talk to ]] (reading (49a)) - Derivation of (51): (54) a.

[CP how many people do you wonder [CP whether [TP * [TP I should talk to ]]]]

[Scope] (reading (51b))

AP interace: * CI interface: *

7

b.

[CP how many people do you wonder [CP whether [TP I should talk to ]]]

(reading (51a))

AP interace: * CI interface: * -> pro (55) Which book do you wonder whether every man dislikes?

(Ruys 2015, p. 467)

è This sentence “does not allow the functional reading for which book paraphrased in [56], where the trace is arguably a function variable, type ” (p. 467): (56) *for which function f mapping men to a book, do you wonder whether every man x dislikes f(x) [possible answers: his oldest, his most expensive, …] - Abe (2015): (57) Hutari-no onna-oi

John-ga [sannin-no sensei-ga ti kiniitteiru to]

two-Gen woman-Acc John-Nom three-Gen teacher-Nom like ‘Two women, John said that three teachers like.’

itta.

Comp said

(two > three, three > two)

(58) [TP hutari-no onna-o [TP John-ga [CP [TP [TP sannin-no [Scope]

[Scope]

sensei-ga kiniitteiru]] to] itta]] (59) Hutari-no onna-oi

John-ga [sannin-no sensei-ga ti kiniitteiru ka] sitteiru.

two-Gen woman-Acc John-Nom three-Gen teacher-Nom like ‘Two women, John knows whether three teachers like.’ (60) a.

Q

know

(*three > two)

[TP hutari-no onna-o [TP John-ga [CP [TP * [TP sannin-no [Scope]

[Scope]

sensei-ga kiniitteiru]] ka] sitteiru]] b. [TP hutari-no onna-o [TP John-ga [CP [TP sannin-no sensei-ga * kiniitteiru]] ka] sitteiru]] - QPs scrambled long distance: (61) Hutari-no onna-oi

sannin-no sensei-ga

[John-ga ti kiniitteiru to]

two-Gen woman-Acc three-Gen teacher-Nom John-Nom like ‘Two women, three teachers said that John liked.’

8

itta.

Comp said (*two > three)

(62) [TP hutari-no onna-o [TP sannin-no sensei-ga [CP [TP [Scope]

[Scope]

[TP John-ga kiniitteiru]] to] itta]] - The availability of functional readings in wh-island contexts: (63) a.

Dono hutari-no onna-oi

John-ga [sannin-no sensei-ga ti kiniitteiru

which two-Gen woman-Acc

John-Nom three-Gen teacher-Nom like

to]

itta no?

Comp said Q ‘Which two women did John said that three teachers like?’ b. Dono hutari-no onna-oi which two-Gen woman-Acc ka]

sitteiru

no?

Q

know

Q

John-ga [sannin-no sensei-ga ti kiniitteiru John-Nom three-Gen teacher-Nom like

‘Which two women did John know whether three teachers like?’ (64) Zibun-no hutari-no zyosi-gakusei-o

desu.

self-Gen two-Gen female student-Acc be ‘Two female students of self’s.’ è This functional answer is available to (63b) as well as (63a). 7.

Scope Reconstruction and Condition C

- Fox (1999): (65) a. #How many houses in Johni’s city does hei think you should build t? b. How many houses in Johni’s city does hei think you should demolish t? (Fox 1999, p. 167-168) (66) a.

for what n: there are n-many houses x in John’s city such that he thinks you should demolish x

b. *for what n: he thinks there should be n-many houses in John’s city that you demolish è “[65b] does not violate Condition C, since it can have a wide scope reading; in [65a], the embedded verb forces a narrow scope reading, which renders the structure unacceptable.” (p. 477)

9

è Assuming the late merge analysis of adjuncts proposed by Lebeaux (1988), which circumvents Condition C violations in (65), Ruys (2015) claims that in that case, the narrow scope reading is suppressed since “the trace/variable … defaults to type e because its λ-binder does not compose with a constituent identical to the trace.” (p. 479-480) - Derivation of (65b): (67) [CP how many houses in Johni’s city does hei [vP [vP think [CP you should [vP [vP demolish ]]]]]] (68) a.

[CP how many houses in Johni’s city does [TP hei think [TP [TP you should demolish ]]]]

[Scope]

b. [CP how many houses in Johni’s city does [TP hei think [TP you should demolish ]]]] è The illegitimacy of (68a) is due to the non-identity of the copy in the wh-operator position and the one that licenses the [Scope] feature; Operator-Scope Identity Condition (OSIC) (69) a.

[Mary-ga

atta] ooku-no

hito-o

John-ga

kasikoi to

Mary-Nom saw many-Gen person-Acc John-Nom smart ‘Many people Mary saw, John thought to be smart.’ b.

[Maryi-ga atta] ooku-no

hito-o

Comp thought (think >< many)

kanozyoi-ga kasikoi to

Mary-Nom saw many-Gen person-Acc she-Nom

omotta.

omotta.

smart Comp thought

‘Many people Maryi saw, shei thought to be smart.’ (many > think, *think > many) (70) a.

there were many people x seen by Mary such that John thought x was smart

b. John thought there were many people seen by Mary that were smart (71) a.

[TP ooku-no hito-o [TP kanozyoi-ga kasikoi to omotta]] ↓ late merge of [Maryi-ga atta]

b. [TP [Maryi-ga atta] ooku-no hito-o [TP kanozyoi-ga kasikoi to omotta]] (72) a.

[Zibuni-ga atta] ooku-no

hito-o

Maryi-ga kasikoi to

self-Nom saw many-Gen person-Acc Mary-Nom smart ‘Many people shei saw, Maryi thought to be smart.’ 10

omotta.

Comp thought (think > many)

b.

[Maryi-ga atta] ooku-no

hito-o

kanozyoi-no

Mary-Nom saw many-Gen person-Acc she-Gen to

omotta.

Comp

thought

otooto-ga

kasikoi

brother-Nom smart

‘Many people Maryi saw, heri brother thought to be smart.’

(many >< think)

(73) [TP [zibuni-ga atta] ooku-no hito-o [TP Maryi-ga [CP [TP <[zibuni-ga atta] [Scope] ooku-no hito-o> [TP <[zibuni-ga atta] ooku-no hito-o> kasikoi]] to] omotta]] (74) [TP [Maryi-ga atta] ooku-no hito-o [TP kanozyoi-no otooto-ga [CP [TP <[Maryi-ga [Scope] atta] ooku-no hito-o> [TP <[Maryi-ga atta] ooku-no hito-o> kasikoi]] to] omotta]] (75) a.

How many slides of Jonathani’s trip to Kamchatka did hei decide to show t at the party?

(many > decide; *decide > many)

b. How many slides of Jonathani’s trip to Kamchatka did Susi decide that hei would show t at the party?

(many >< decide) (Fox 1999, p. 171)

(76) a.

How many houses in Johni’s city do you think hei should build t?

b. How many houses in Johni’s city do you think hei should demolish t? (77) a.

[TP hei would show how many slides at the party] [Scope]

b.

[TP how many slides of Jonathani’s trip to Kamchatka [TP hei would show [Scope] at the party]]

c.

[CP how many slides of Jonathani’s trip to Kamchatka did [TP Susi decide [CP [TP [TP hei would show [Scope] at the party]]]]] 11

(78) α c-commands β iff every segement that dominates α dominates β. (79) [CP how many slides of Jonathani’s trip to Kamchatka did Susi [vP [vP decide [CP hei would [vP [vP show at the party]]]]]] è Probably, Ruys’s (2015) semantic theory also accommodates this fact. (80) a.

[Maryi-ga atta] ooku-no

hito-o

kanozyoi-ga [John-ga t kiniitteiru

Mary-Nom saw many-Gen person-Acc she-Nom to]

John-Nom like

omotteiru.

Comp think ‘Many people Maryi saw, shei thinks John likes. (many > think, *think > many) b.

[Maryi-ga atta] ooku-no

hito-o

John-ga [kanozyoi-ga t kiniitteiru

Mary-Nom saw many-Gen person-Acc John-Nom she-Nom to]

like

omotteiru.

Comp think ‘Many people Maryi saw, John thinks shei likes.’

(many >< think)

(81) *[TP [Maryi-ga atta] ooku-no hito-o [TP kanozyoi-ga [CP [TP [Scope] [TP John-ga kiniitteiru]] to] omotteiru]] (82) [TP [Maryi-ga atta] ooku-no hito-o [TP John-ga [CP [TP <[Maryi-ga atta] ooku-no [Scope] hito-o> [TP kanozyoi-ga kiniitteiru]] to] omotteiru]] (83) [Maryi-ga Mary-Nom

atta] ooku-no

hito-o

kanozyoi-ga t kiniitteiru.

saw many-Gen person-Acc

she-Nom

like

‘Many people Maryi saw, shei likes.’ (84) [Zibuni-o kiniitteiru] self-Acc like to]

zyosei-o

Johni-ga [daremo-ga t kiniitteiru

woman-Acc

John-Nom everyone-Nom like

omotteiru.

Comp think

(a woman >< everyone)

‘A woman who likes himi, Johni thinks everyone likes.’ 12

(85) [TP [zibuni-o kiniitteiru] zyosei-o [TP Johni-ga [CP [TP <[zibuni-o kiniitteiru] [Scope] zyosei-o> [TP daremo-ga <[zibuni-o kiniitteiru] zyosei-o> kiratteiru]] to] [Scope] omotteiru]] (86) [Johni-o kiniitteiru] zyosei-o John-Acc like

karei-ga [daremo-ga t kiniitteiru to]

woman-Acc he-Nom everyone-Nom like

Comp

omotteiru. think

(a woman > everyone, *everyone > a woman)

‘A woman who likes Johni, hei thinks everyone likes.’ (87) *[TP [Johni-o kiniitteiru] zyosei-o [TP karei-ga [CP [TP [TP daremo-ga [Scope]

[Scope]

Argument-Adjunct_Asymmetries.pdf

Page 1 of 18. 1. Interface Conditions on âGapsâ: Argument-Adjunct Asymmetries and. Scope Reconstruction. Jun Abe. Workshop on Modality and Embedded Clauses. YNU, December 23, 2015. 1. Introduction. - Argument-Adjunct Asymmetry (I):. (1) a.??Who do you wonder whether John said t solved the problem? b.

Download PDF

264KB Sizes 2 Downloads 292 Views

Report

Argument-Adjunct_Asymmetries.pdf

Recommend Documents