(A-)symmetric Epistemic Intervention in Quanti cation and

(A-)symmetric Epistemic Intervention inQuantification and Wh-construals

Shi-yue Jeff LinNational Tsing Hua University & National Taiwan Normal University

[email protected]

TEAL-7@Hiroshima University, February 19, 2012

1 Introduction

• Epistemic Containment Principle (von Fintel and Iatridou 2003) refers tothe descriptive generalization that a quantifier cannot have scope over an epistemicmodal.

(1) Epistemic Containment Principle (ECP)

A quantifier cannot have scope over an epistemic modal=⇒ ∗[Quantifieri � Epistemic Modal [. . .x i. . .]]

(2) Necessity Epistemic Modal:

Most of our students must be home by now.a. ∗Most of our students � must/ (most � 2).b. Must � most of our students/ (2 � most).

(3) Possibility Epistemic Modal:

Every student may have left.a. ∗Every student � may/ (∀ � 3)b. May � every student/ (3 � ∀)

• In this article the interaction between epistemic modals and four types of scope bear-ing element (SBE) is taken into consideration and examination, including (i) quan-tificational phrase with dou (QP. . .dou construction)1 (ii) focus operators (zhiyou‘only’), (iii) negation (bu ‘not’), and (iv) wh-construals.

• Goals of this paper:

– Show an asymmetry between epistemic modal auxiliaries (EMAUX = keneng‘may’) and epistemic modal adverbials (EMADV = yexu ‘perhaps’ ) with re-spect to ECP.

1In this article I follow Lin’s (1996) insightful study to assume that the quantificational force inQP. . .dou construction resided in dou(P), rather than in QP like mei -CL ‘every’, which acts more likea plural noun, in Chinese. See also Yang (2002) for an alternative analysis, which assumes mei to be avariable bound by dou.

1

(4) Asymmetric ECP Effect:

A scope bearing element (SBE), including QP-dou, negation, and focusoperator, can scope over an EMAUX , but not an EMADV .

a.∗ [QPi/SBE � EMADV [. . . x i . . .]]:

∗Mei-geEvery-CL

xueshengstudent

douall

yexuEMADV

perhaps

likai-le.leave-SFP

‘For every student x, it is perhaps that x has left/ ∀ � 3.’

b. [QPi/SBE � EMAUX [. . . x i . . .]]:

Mei-geEvery-CL

xueshengstudent

douall

kenengEMAUX

may

likai-le.leave-SFP

‘For every student x, it may be the case that x has left/ ∀ � 3.’

– Show more than one phenomenon is behind ECP with respect to the wh-construals.

(5) asymmetric epistemic intervention/ECP effect:

A Q-operator cannot bind a variable across an EMADV , but can acrossan EMAUX ; a reason wh-operator (reason how and why) cannot scopeover an EMADV , but can scope over an EMAUX .

a.∗ [Qi-op � EMADV [. . . wh(x i) . . .]]

JohnJohn

yexuperhaps

mai-guobuy-Prt

sheme.what

‘∗What might John have bought? (∗Interrogative)’‘John perhaps had bought something. (Existential/Indefinite).’

b. [Qi-op � EMAUX [. . . wh(x i) . . .]]

JohnJohn

kenengpossible

mai-guobuy-Prt

sheme?what

‘What might John have bought? (Interrogative)’‘It may be the case that John had bought something(Existential/Indefinite).’

c.∗ [QREASON � EMADV [. . .]]

∗Weishemewhy

JohnJohn

yexuperhaps

likai-le?leave

‘Why is it possible that John left?’

d. [QREASON � EMAUX [. . .]]

Weishemewhy

JohnJohn

kenengmay

likai-le?leave

‘Why may John have left?’

2

(6) symmetric epistemic intervention effect:

Covert movement of a wh-adverbial (manner how) across either EMAUX

or EMADV is generally prohibited.

a. ∗[Qi-op � EMADV [. . . t i . . .]]

∗JohnJohn

kenengmay

zenmeMANNER

howMANNER

chulihandle

zhe-jianthis-CL

shi?matter

‘In what manner may John handle this matter?’

b. ∗[Qi-op � EMAUX [. . . t i . . .]]

∗JohnJohn

yexuperhaps

zenmeMANNER

howMANNER

chulihandle

zhe-jianthis-CL

shi?matter


(7) symmetric anti-ECP effect:

Wh-construals are prohibited when an epistemic modal occupies thesentence initial position, which is a contrary effect to the claim of theECP.

a. anti-ECP effect (unselective binding):∗[CP Epistemic modal � Qi-op [TP . . . wh(x i) . . .]]

Yexu/kenengperhaps/may

JohnJohn

mai-guobuy-Prt

sheme.what

‘∗What may John have bought? (∗Interrogative)’‘John may have bought something. (Existential/Indefinite).’

b. anti-ECP effect (covert movement):∗[CP Epistemic modal � Qi-op [TP . . . t i . . .]]

∗Yexu/∗kenengperhaps/may

JohnJohn

zenmeMANNER

howMANNER

chulihandle

zhe-jianthis-CL

shi?matter


c. anti-ECP effect (reason wh-operator):∗[CP Epistemic modal � QREASON [TP . . .]]

∗Yexu/∗kenengperhaps/may

weishemeREASON

whyJohnJohn

likai-le?leave


– Propose an intervention account:

† Generalized Relativized Minimality theory in Rizzi (2004) is adopted, butit is proposed here that [Epistemic] feature should be also classified withinthe Quantificational class, according to its quantificational nature.

3

(8) Feature types:

a. Argumental: person, number, gender, case. . .

b. Quantificational: Wh, Neg, measure, focus. . .[Epistemic]

c. Modifier: evaluative, epistemic, Neg, frequentative, celerative,measure, manner. . .

d. Topic

† According to Lyons (1977)2,

→ a subjective modal is infused into the illocutionary force ofthe utterance, while

→ an objective modal is infused into the truth-conditional con-tent.

→ Accordingly, two different syntax-semantics structural requirementsare proposed: subjective EMADV modifies the utterance andrequires the whole CP as its complement, while an EMAUX ,under its objective reading, requires only a truth-conditionalmaximal projection (= TP) (cf. Huitink 2008).

(9) Syntax-semantics Structrual Requirement:

a. J EMADV [CP] K = subjective conjecture

b. J EMAUX

{[CP][TP]

}K =

{subjective conjectureobjective conjecture

}

† As a consequence,

→ the [Epistemic] feature of an EMADV must undergo feature movementto check/value [uForce] in ForceP as [iEpistemic] in order to modifythe whole CP and to express the conjecture of the speaker. On thecontrary,

→ an EMAUX can be interpreted objectively in-situ because its structuralrequirement can be satisfied by external merger with TP, and ForcePcan be valued as [iDeclarative].

2See also Papafragou (2006), Tancredi (2007), Huitink (2008), Anand and Hacquard (2009), vonFintel and Gillies (2011), and among others for the subjectivity-objectivity distinction.

4

(10) Feature Movement of [Epistemic] Feature to ForceP:

6

ForceP

TopP

Top′

ModalP

Modal′

TP

. . .

Modal

EMAUX[Epistemic

]

EMADV

yexu[Epistemic

]

Top

DP

Subj

[uForce]

† Finally, based on the insight of Clausal-Typing Hypothesis (Cheng 1991),I assume that a wh-question would require the ForceP to be valued as[iInterrogative] to express questioning (c.f. Pesetsky and Torrego 2004).

(11) Interrogative sentence:

[ForceP [iInterrogative] [IntP Qi-op [TP John mai-guo shemei?]]]=⇒ Questioning:λx . John bought x & tell me the information about x.

† Consequently, the Quantificational [Epistemic] feature would induce threedifferent syntactic effect with respect to the SBEs.

→ First of all, [Epistemic] feature movement would be blocked by an SBEwith Quantificational feature, which is the asymmetric ECP effect.


∗[ForceP [uForce] [ QPi/SBE[Quant.

] [ModP EMADV[Epistemic

] [TP . . .]]]]

6

×

→ Secondly, [Epistemic] feature would contradictorily compete ForcePwith [Interrogative] for checking/valuation, which results in the block-ing of interrogative reading in wh-construals.

5

(13) Asymmetric Epistemic Intervention Effect:

∗[ForceP [uForce] [ Q− op[Int

] [ModalP EMADV[Epistemic

] [TP . . .]]]]]]

6

×

6

×

→ Finally, a Quantificational [Epistemic] feature in both EMADV andEMAUX becomes a strong intervener which would block another Quan-tificational feature movement, such as [Wh] feature, of a wh-adverbial.

(14) Symmetric Epistemic Intervention Effect:

∗[CP [ ] [ModalP EMADV/AUX[Epistemic

] [TP . . . wh− adverbial[Wh

] ]]

6

×

2 Literature Review

2.1 Drubig (2001)

• Assume the framework in Westmoreland (1998):

i. Epistemic must (in contrast to root modals) is not a modal but must beanalyzed as an evidential marker labelling the proposition in its scope as adeduction. It relates a proposition ϕ to some other information that serves asevidence for ϕ.

ii. The traditional view that may ϕ or might ϕ is true (or acceptable) just aslong as ϕ is consistent with the context does not suffice to capture its meaning;rather: an expression such as might ϕ is used to mean that the contextcontains causal factors that make ϕ plausible.

– In general we may say: just as a question marker takes a proposition andderives a question, an epistemic modal takes a proposition and yields an evi-dentially labelled proposition.

– An epistemic modal, such as must or may, according to this view,is part ofthe metalogical vocabulary, i.e. not equivalent to the square 2 but more likethe sign marking the deduction of an inference.

– Beyond that, the extrapropositional status of an epistemic modal implies thatit takes scope over all propositional operators.

(Drubig 2001, p.2 - p.5)

=⇒ Epistemic modal �{

negation / tense / aspectquestion / focus operator

}

6

2.2 von Fintel and Iatridou (2003)

• A specific descriptive condition on Quantifier Raising (QR):

(15) Epistemic Containment Principle (Refined Version):

At LF, a quantifier cannot bind its trace across an epistemic modal.=⇒ ∗[Quantifieri � Epistemic Modal [. . .t i. . .]].

2.3 Tancredi (2007)

• Two types of epistemic modal:

{subjectiveobjective

}

=⇒ only subjective epistemic modal induced the ECP effect.

(16) Subjective ECP Effect in English:

a. #(As far as I know) Every student is perhaps Jones.

b. #(As far as I know) Most students are perhaps Jones

(17) Subjective ECP Effect in Japanese:

a.# Subete-noEvery-GEN

gakusei-gastudent-NOM

moshikashitaraperhaps

JonesJones

de aruCOP

‘Every student is perhaps Jones.’

b.# Hotondo-noMost-GEN


moshikashitaraperhaps

JonesJones

de aruCOP

‘Most students are perhaps Jones.’

=⇒ objective epistemic modal could cancel the ECP effect.

(18) Objective ECP canceling Effect in English:

a. (Objectively speaking,) Every student may be Jones.

b. (Objectively speaking,) Most students may be Jones.

(19) Objective ECP canceling Effect in Japanese:

a. (Kyakkanteki-niObjectively

mite)looking

Subete-noEvery-GEN


JonesJones

de aruCOP

kamoshirenaimay

‘(Objectively speaking,) every student may be Jones’

7

b. (Kyakkanteki-niObjectively

mite)looking

Hotondo-noMost-GEN


JonesJones

de aruCOP

kamoshirenaimay

‘(Objectively speaking,) most students may be Jones’

• Analysis: multi-model modal theory of I-semantics

2.4 Huitink (2008)

• Similar pattern in Dutch: where epistemic modal auxiliaries can go, but epistemicmodal adverbials cannot

(20) Scope Over Epistemic Modal Auxiliaries in Dutch:

Iedereevery

studentstudent

kanmay

vertrokkenleft

zijn.be

‘Every student may have left. (∀ � 3 / 3 � ∀)’

(21) ECP Effect in Dutch (Epistemic Modal Adverbials):

Iedereeneveryone

heefthas

hetit

misschienperhaps

gedaan.done

‘Perhaps everyone has done it. (∗∀ � 3 / 3 � ∀)’

• Subjective modality indicates the strength of the speech act, while objective modal-ity contributes to content.

– Default speech act for declaratives: assertion.

– Subjective modals change the speech act.

• Quantifiers contribute to the propositional level of meaning=⇒ they scope under subjective epistemic modals.

Table 1: Modifying Speech Act v.s. Modifying Content

Force ContentAssert rain It is raining

Conjecture rain It mayS be raining.Assert rain It mayO be raining

2.5 Gagnon and Wellwood (2010)

• To follow Beghelli (1997) and Beghelli and Stowell (1997) to propose an topologicalaccount based on the quantificational left-periphery theory:

8

(22) [CP Each � Epistemic Modal � Every � [TP . . .]]

(23) Each Cancels the ECP Effect; Every doesn’t:

a.# Every girl might be in love with John, but some of them aren’t.

b. Each girl might be in love with John, but some of them aren’t

3 Case Study in Chinese

3.1 QP. . .Dou Construction

• A counterexample to the ECP can be immediately detected with a direct translationof the English examples into Chinese.

(24) Counterexample to the ECP Effect:

a. Mei-geEvery-CL

xueshengstudent

douall

kenengmay

likai-le.leave-SFP

‘For every student x, it may be the case that x has left/ ∀ � 3.’

b. Mei-geEvery-CL

xueshengstudent

kenengmay

douall

likai-le.leave-SFP

‘It may be the case that every student have left/ 3 � ∀.’

• However, it cannot be determined a priori that ECP is inert in Chinese since theECP effect is still observed once modal expressions such as yexu ‘perhaps’ or dagai‘probably’ are present.

(25) ECP effect in Chinese:

a.∗ Mei-geEvery-CL

xueshengstudent

douall

yexuperhaps

likai-le.leave-SFP

‘For every student x, it is perhaps that x has left/ ∀ � 3.’

b. Mei-geEvery-CL

xueshengstudent

yexuperhaps

douall

likai-le.leave-SFP

‘It is perhaps that every student have left/ 3 � ∀.’

(26) ECP effect in Chinese:

a.∗ Mei-geEvery-CL

xueshengstudent

douall

dagaiprobably

likai-le.leave-SFP

‘For every student x, it is probably that x has left/ ∀ � 3.’

b. Mei-geEvery-CL

xueshengstudent

dagaiprobably

douall

likai-le.leave-SFP

‘It is probably that every student have left/ 3 � ∀.’

• Generalization: asymmetric ECP effect.

9



a.∗ [CP . . . [DouP QPi dou � EMADV [. . . x i . . .]]]

b. [CP . . . [DouP QPi dou � EMAUX [. . . x i . . .]]]

3.2 Focus Operators

• In Chinese the asymmetry between EMAUX and EMADV would result in an asym-metric distribution also with respect to the focus operator. Specifically, as we haveseen in the previous subsection, we would predict that a focus operator can takescope beyond an EMAUX , but can’t beyond an EMADV

(28) Focus operator zhiyou ‘only’ can scope beyond or behind an EMAUX :

a. Zhiyouonly

kenengmay

JohnJohn

mai-guobuy-Prt

zhe-benthis-CL

shu.book

‘The only possibility is that John bought this book.’

b. Zhiyouonly

JohnJohn

kenengmay

mai-guobuy-Prt

zhe-benthis-CL

shu.book

‘Only John may have bought this book.’

c. Kenengmay

zhiyouonly

JohnJohn

mai-guobuy-Prt

zhe-benthis-CL

shu.book

‘It may be the case that only John bought this book.’

d. Kenengmay

JohnJohn

zhiyouonly

mai-guobuy-Prt

zhe-benthis-CL

shu.book

‘It may be the case that John only bought this book.’

(29) Focus operator zhiyou ‘only’ cannot scope over an EMADV :

a.∗ Zhiyouonly

yexuperhaps

JohnJohn

mai-guobuy-Prt

zhe-benthis-CL

shu.book

‘Intended: The only possibility is that John bought this book.’

b.∗ Zhiyouonly

JohnJohn

yexuperhaps

mai-guobuy-Prt

zhe-benthis-CL

shubook

‘Only John, perhaps, had bought this book.’

c. Yexuperhaps

zhiyouonly

JohnJohn

mai-guobuy-Prt

zhe-benthis-CL

shu.book

‘It is perhaps the case that only John bought this book.’

d. Yexuperhaps

JohnJohn

zhiyouonly

mai-guobuy-Prt

zhe-benthis-CL

shu.book

‘It is perhaps the case that John only bought this book.’

10




a.∗ [CP . . . [FocusP zhiyou[Focus− op

] � EMADV [. . . x i . . .]]]

b. [CP . . . [FocusP zhiyou[Focus− op

] � EMAUX [. . . x i . . .]]]

3.3 Negation

• There are two types of negative expressions in Chinese: irrealis bu ‘irrealis not’and realis mei(you) ‘realis not’. As noted in Tsai (2009), irrealis bu can be locatedeither beyond or below an epistemic modal, for example:

(31) a. JohnJohn

kenengmay

bunot

qugo

Taipei.Taipei

‘John may not go to Taipei.’

b. JohnJohn

bunot

kenengmay

qugo

Taipei.Taipei

‘It is not possible that John goes to Taipei.’

• However, bu cannot scope over and negate the epistemic modal adverbial, for in-stance:

(32) ∗bu � 3

a. JohnJohn

yexuperhaps

bunot

qugo

Taipei.Taipei

‘It is perhaps the case that John doesn’t go to Taipei.’

b.∗ JohnJohn

bunot

yexuperhaps

qugo

Taipei.Taipei

‘It is not the case that John perhaps goes to Taipei.’

(33) ∗bu � 3

a. JohnJohn

dagaiprobably

bunot

qugo

Taipei.Taipei

‘It is maybe the case that John doesn’t go to Taipei.’

b.∗ JohnJohn

bunot

dagaiprobably

qugo

Taipei.Taipei

‘It is not the case that John maybe doesn’t go to Taipei.’

11

(34) ∗bu � 2

a. JohnJohn

jueduidefinitely

bunot

qugo

Taipei.Taipei

‘It is definitely the case that John doesn’t go to Taipei.’

b.∗ JohnJohn

bunot

jueduidefinitely

qugo

Taipei.Taipei

‘It is not the case that John definitely go to Taipei.’


(35) Asymmetric ECP Effect:A scope bearing element (SBE), including QP-dou, negation, and focusoperator, can scope over an EMAUX , but not an EMADV .

a.∗ [CP . . . [NegP bu[negation

] � EMADV [. . . x i . . .]]]

b. [CP . . . [NegP bu[negation

] � EMAUX [. . . x i . . .]]]

Table 2: Generalization for ECP effect

QP-dou construction Focus operator NegationEMADV

∗QP-dou � EMADV∗Focus � EMADV

∗Negation � EMADV

EMAUX QP-dou � EMAUX Focus � EMAUX Negation � EMAUX

• Interim summary:



a.∗ [QPi/SBE � EMADV [. . . x i . . .]].

b. [QPi/SBE � EMAUX [. . . x i . . .]].

4 Analysis

• Puzzle: why SBEs cannot scope over an EMADV , but can scope over an EMAUX ?

• Speculation: SBEs block the covert [Epistemic] feature movement.

12

4.1 Generalized Relativized Minimality

• A modified version of Rizzi’s (2004) generalized relativized minimality theory: toclassify [Epistemic] feature within the Quantificational class.

(37) Feature Types:

a. Argumental: person, number, gender, case. . .

b. Quantificational: Wh, Neg, measure, focus. . .[Epistemic]

c. Modifier: evaluative, epistemic, Neg, frequentative, celerative, measure,manner. . .

d. Topic

• Two supportive views for this classification:(i) Traditional Kratzerian modal semantics and(ii) The ability of wh-indefinite licensing.

4.2 Covert Movement and Intervention Effect

• It has been argued in Pesetsky (2000) that there are two types of covert movement:feature movement and phrasal movement. Furthermore, he also proposed thatintervention effect is induced by feature movement, but not by phrasal movement,for instance:

(38) Wh1-In-Situ Undergoes Feature Movement (Intervention Effect):

∗Which booki didn’t which student read ti?

(39) Wh2-In-Situ Undergoes Phrasal Movement (No Intervention Effect):

Which student didn’t read which book?

• Yang (2008) extended Pesetsky’s speculation and develop a comprehensive studiesin Chinese (see also Yang 2011). Specifically, Yang proposed that wh-adverbialswould undergo feature movement to obtain their scope position, and thus inducethe intervention effect.

(40) Wh-adverbial Intervention Effect:

a.∗ Mei-geevery-CL

renperson

douall

weishemewhy

cizhi?quit

‘Why does everyone quit?’

b.∗ Mei-geevery-CL

renperson

douall

zenmehow

cizhi?quit

‘How does everyone quit?’

• I think that Yang has provided several convincing arguments and natural expla-nations to the intervention effect, based on the generalized relativized minimalityapproach. Thus, I follow these promising works to assume that feature movementwould induce intervention effects, once a intervener with the same feature type werepresent in the syntactic structure.

13

4.3 The Syntax of ForceP Valuation

• Question: where and why did [Epistemic] feature move to?

– [Epistemic] feature moves to check/value the [uForce] feature in ForceP toexpress speaker’s subjective conjecture.

(41) [Epistemic] Feature Movement:

ForceP

ModalP

Modal′

TP

. . .

Modal

EMAUX[Epistemic

]

EMADV[Epistemic

][uForce]

6

• What’s the distinction between EMADV and EMAUX?

– An EMADV would always undergo feature movement to check/value ForceP,but an EMAUX could optionally do so.

– This distinction is attributed to the syntactic structure of ModalP. More specif-ically, following standard assumption of the Kratzerian modal semantics, anepistemic modal would take a proposition as (one of) its argument (Kratzer1977, 1981, and 1991). Putting this assumption in the minimalist terms, Ipropose that an epistemic modal feature [epistemic], being Quantificational,would require a propositional maximal projection as its sister node.

– The structural requirement of an EMAUX can be satisfied with external merger(Chomsky 2006). That is, through syntactic derivation, an EMAUX , witha [Epistemic] feature, always merged with an propositional maximal projec-tion inherently, and the structural requirement can be subsequently satis-fied. On the contrary, an EMADV , being hosted inherently in the specifier ofModalP, cannot satisfied the structural requirement. Thus, [epistemic] featureof EMADV would undergo feature movement to ForceP since ForceP alwaysdominates a propositional maximal projection, the CP phase.

• Some semantic consequence:

– As noted in Lyon (1977), an objective epistemic modal contributes to thetruth conditional content of the utterance, but a subjective epistemic modalcontributes to the illocutionary force of the utterance.

14

– In addition, Tancredi (2007) and Huitink (2008) also argued quite convincinglyfor a distinction between subjective epistemic modal and objective epistemicmodal, and proposed that an EMADV is commonly subjective.

– These arguments probably can be correlated nicely to the assumption thatEMADV always undergo feature movement to ForceP for the structural re-quirements. That is, a sentence would express subjectivity once its ForcePwere valued as [iEpistemic]. On the contrary, an EMAUX optionally valuesthe ForceP and the sentence could be interpreted as Force([iDeclarative]) +EMAUX([Epistemic]), which denotes objectivity.

4.4 The Feature Movement Account

• Given our assumptions, the asymmetrical ECP effect is induced because of theobligatory feature movement of [Epistemic] feature and its structural requirement.Specifically, EMADV must undergo feature movement to ForceP because it sisternode, an intermediate Modal′ is not a propositional maximal projection. On thecontrary, EMAUX can stay in-situ since its structural requirement has been satisfiedwith external merger.

(42) EMADV Induced ECP effect:

∗ForceP

DouP

Dou′

ModalP

Modal′

TP

. . .

Modal

φ

EMADV

yexu[Epistemic

]

Dou[Quant.

]QP

mei-ge xuesheng

[uForce]

6

×

15

(43) EMAUX Obviated ECP effect (Under Objective Reading):

ForceP

DouP

Dou′

ModalP

TP

. . .

Modal

keneng[Epistemic

]

Dou[Quant.

]QP

mei-ge xuesheng

[iDeclarative]

5 Wh-construals

• Epistemic modal would also interact with wh-construals. Specifically, three types ofwh-construals are investigated, including unselective binding, covert wh-movement,and the reason-wh-phrase in CP. We will see various epistemic intervention effectsand propose how to capture these within our framework.

5.1 Unselective Binding

• As indicated in the translations, yexu would be compatible with a wh-argument onlyif this wh-phrase were used as an indefinite wh-expression. By contrast, keneng isnot so restricted and compatible with both interrogative reading and existentialreading.

(44) EMADV within Unselective Binding:

JohnJohn

yexuperhaps

mai-guobuy-Prt

sheme.what

‘∗What might John have bought? (∗Interrogative)’‘John perhaps had bought something. (Existential/Indefinite).’

(45) EMAUX within Unselective Binding:

JohnJohn

kenengmay

mai-guobuy-Prt

sheme?what

‘What might John have bought? (Interrogative)’‘It is possible that John bought something (Existential/Indefinite).’

• The similar point can be made clear by a further supportive evidence from instrumental-zenme(-yang) ‘by what means’, which is also sensitive to unselective binding (Tsai1999, 2000, 2008):

16

(46) EMADV with zenme-(yang):

∗JohnJohn

yexumaybe

zenme(-yang)INSTRUMENTAL

how-wayINSTRUMENTAL

qugo

xuexiao?school

‘By what means may John go to school?’

(47) EMAUX with zenme-(yang):

JohnJohn

kenengmay

zenme(-yang)INSTRUMENTAL

how-wayINSTRUMENTAL

qugo

xuexiao?school

‘By what means may John go to school?’

• Besides, the phenomena can become even more complicated when the relative po-sition between the subject and epistemic modal are inverse.

• When an epistemic modal precedes a subject and stays at a sentence initial position,the interrogative reading is blocked with respect to both EMAUX and EMADV .

(48) Sentence-Initial EMAUX Blocks Interrogative Reading:

Kenengmay

JohnJohn

mai-guobuy-Prt

sheme.what

‘∗What may John have bought? (∗Interrogative)’‘John may have bought something. (Existential/Indefinite).’

(49) Sentence-Initial EMADV Blocks Interrogative Reading:

Yexuperhaps

JohnJohn

mai-guobuy-Prt

sheme.what

‘∗What may John have bought? (∗Interrogative)’‘John perhaps bought something. (Existential/Indefinite)’

• An interim summary here is that an epistemic modal seems to display differenteffect with respect to its syntactic position.

– On the one hand, within a sentence, it asymmetrically interferes the unselectivebinding construals: only EMADV blocks the interrogative reading.

(50) Asymmetric Epistemic Intervention Effect:

A Q-operator cannot bind a variable across an EMADV , but can acrossan EMAUX ; a reason wh-operator (reason how and why) cannot scopeover an EMADV , but can scope over an EMAUX .

a.∗ [Qi-op � EMADV [. . . wh(x i) . . .]]

b. [Qi-op � EMAUX [. . . wh(x i) . . .]]

– On the other hand, at the sentence initial position, both EMADV and EMAUX

symmetrically blocked the interrogative reading.

17

(51) Symmetric Anti-ECP Effect (Unselective Binding):

a.∗ [EMADV � Qi-op . . . [Subject . . . [. . . wh(x i) . . .]]]

b.∗ [EMAUX � Qi-op . . . [Subject . . . [. . . wh(x i) . . .]]

5.2 Covert Wh-Movement

• let us consider the case of manner-zenme ‘in what manner’ and its interaction withepistemic modal.

(52) EMAUX Blocks the Wh-Construal:

∗JohnJohn

kenengmay

zenmeMANNER

howMANNER

chulihandle

zhe-jianthis-CL

shi?matter


(53) EMADV Blocks the Wh-Construal:

∗JohnJohn

yexuperhaps

zenmeMANNER

howMANNER

chulihandle

zhe-jianthis-CL

shi?matter


• Surprisingly, the asymmetric property is absent. That is, as noted in Tsai (2008),post-modal zenme can only be interpreted as instrumental-how, but not as manner-how.

• On the other hand, EMADV and EMAUX both induced the identical effect; namely,they rule out the sentence with manner-zenme.

(54) Sentence-Initial EMAUX Blocks the Wh-Construal:

∗Kenengmay

JohnJohn

zenmeMANNER

howMANNER

chulihandle

zhe-jianthis-CL

shi?matter


(55) Sentence-Initial EMADV Blocks the Wh-Construal:

∗Yexuperhaps

JohnJohn

zenmeMANNER

howMANNER

chulihandle

zhe-jianthis-CL

shi?matter


• Generalization: epistemic modal is generally incompatible with covert wh-movement:

(56) Symmetric Epistemic Intervention (Covert Wh-Movement):

Covert movement of a wh-adverbial (manner how) across either EMAUX orEMADV is generally prohibited.

18

a. ∗[Qi-op � EMADV [. . . t i . . .]]

b. ∗[Qi-op � EMAUX [. . . t i . . .]]

(57) Symmetric Anti-ECP Effect (Covert Wh-Movement):

a.∗ [EMADV � Qi-op . . . [Subject . . . [. . . t i . . .]]]

b.∗ [EMAUX � Qi-op . . . [Subject . . . [. . . t i . . .]]

5.3 How-Why in CP

• Finally, in addition to the unselective binding construal and the movement con-strual, the structure involving weisheme ‘why’ and reason-zenme ‘reason how’should be taken into consideration, since both of them can be externally mergedhighly in the CP domain (Stepanov and Tsai 2008).

(58) Weisheme ‘why’:

a. Weishemewhy

JohnJohn

kenengmay

likai-le?leave


b. ∗Weishemewhy

JohnJohn

yexuperhaps

likai-le?leave


(59) Zenme ‘reason how’

a. JohnJohn

zenmeREASON

howkenengmay

zhucook

wancan.dinner

‘How come that John cooked for dinner.’

b. ∗JohnJohn

zenmeREASON

howyexuperhaps

zhucook

wancandinner

‘How can it be possible that John cook for dinner.’

• The inverse order between epistemic modal and the reason wh-operator patternswith both unselective binding and covert movement construal. Both EMADV andEMAUX rule out the structure, displaying the anti-epistemic containment effect.

(60) 3 � Weisheme ‘Reason Why’:

a.∗ Kenengmay

weishemeREASON

whyJohnJohn

ikai-le?leave


b.∗ Yexuperhaps

weishemeREASON

whyJohnJohn

likai-le?leave


19

(61) 3 � Zenme ‘Reason How ’

a.∗ Kenengmay

zenmeREASON

howJohnJohn

zhucook

wancan.dinner

‘How come that John cooked for dinner.’

b.∗ Yexuperhaps

zenmeREASON

howJohnJohn

zhucook

wancandinner

‘How can it be possible that John cook for dinner?’

• Generalization: how-why in CP shows the similar pattern with the unselectivebinding construal:

(62) Asymmetric ECP Effect (Reason Wh-Operator):

c.∗ [QREASON � EMADV [TP . . .]]

d. [QREASON � EMAUX [TP . . .]]

(63) Symmetric Anti-ECP Effect (Reason Wh-Operator):

a.∗ [EMADV � QREASON [TP . . .]]

b.∗ [EMAUX � QREASON [TP . . .]]

5.4 The Comprehensive Picture

5.4.1 Interim Summary

• Table for different wh-construal with different patterns:

(64) Table for Wh-Construal Generalization:

Table 3: Epistemic Modal within the Three Types of Wh-Construals

Unselective Covert How-WhyBinding Movement in CP

[Q(i)-op � EMADV [. . . ti/ x i . . .]] ∗ ∗ ∗[Q(i)-op � EMAUX [. . . ti/ x i . . .]] ok ∗ ok[EMADV � Q(i)-op [. . . ti/ x i . . .]] ∗ ∗ ∗[EMAUX � Q(i)-op [. . . ti/ x i . . .]] ∗ ∗ ∗

5.4.2 Feature Interpretability Contradiction in ForceP

• Explain the various (a-)symmetric epistemic intervention/containment effect withinthe [Epistemic] feature movement framework:

– Feature interpretability contradiction in ForceP:both wh-phrase and EMADV intend to type the whole clause as [iInterrogative]and [iEpistemic], respectively.

20

(65) Feature Contradiction in ForceP:∗ForceP

IntP

TopP

ModalP

Modal′

TP

mai-guo sheme?

EMAUX

keneng[Epistemic

]

EMADV

yexu[Epistemic

]

DP

John

Q− op[Int

][uForce]

6

×

6

×

– Quantificational [Epistemic] feature blocked [Wh] feature movement:

(66) Quantificational [Epistemic] Intervention Effect:∗ForceP

IntP

TopP

ModalP

Modal′

TP

zenme[[Wh][Int]

] . . .?EMAUX

keneng[Epistemic

]

EMADV

yexu[Epistemic

]

DP

John

[ ]

[uForce]

6

×

21

– Symmetric anti-epistemic containment effect: wh-construals are prohibitedwhen an epistemic modal occupies the sentence initial position.

– Epistemic modal is externally merged in ForceP:↪→ [uForce] feature is checked/valued as [iEpistemic]↪→ No place for the [Interrogative] feature↪→ The clause type is incompatible with a wh-phrase.

(67) Anti-ECP Effect (Unselective Binding):

∗ForceP

Force′

IntP

TopP

TP

mai-guo shemei?

DP

John

Int

Qi − op[Int

]

EMAUX

keneng[iEpistemic

]

EMADV

Yexu

6

×

(68) Anti-ECP Effect (Covert Wh-Movement):

∗ForceP

Force′

IntP

TopP

TP

zenme[Int

] chuli zhe-jian shi?

DP

John

Int

[Wh]

EMAUX

keneng[iEpistemic

]

EMADV

Yexu

6

×

22

(69) Anti-ECP Effect (Reason Wh-Operator):

ForceP

Force′

IntP

TopP

TP

likai-le?

DP

John

Int

weisheme[Int

]

EMAUX

keneng[iEpistemic

]

EMADV

Yexu

6

×

6 Concluding Remarks

• To conclude, this article presents the investigation to the ECP effect (von Fintel andIatridou 2003). From many different perspectives, I think Chinese can bring us theEMADV -EMAUX dichotomy, which could possibly be a clue for the comprehensivepicture of the ECP effect.

• In addition to the quantifier-modal interaction, many quantificational constructionincluding focus operator, negation, and wh-construals are discussed and examinedin this paper, and some speculation is provided.

• Specifically, I propose that the ECP effect is a consequence of three independentworking hypotheses: generalized relativized minimality, feature-induced interven-tion effect, and the syntax of ForceP valuation.

• This speculation, however, may be still far from the final destination since manyaspects of this issue still call for explanations and further evidences.

23

References

Anand, Pranav, and Dannis Tang. 2004. Redistributing dou: cleaving exhaustivity fromdistributivity. In Proceedings of the 23rd WCCFL, ed. Vineeta Chand et al., 15-28.

Anand, Pranav, and Valentine Hacquard. 2009. Epistemics with attitude. In Proceedingsof SALT 18 ed. Tova Friedman and Satoshi Ito. 19-36.

Aoun, Joseph and Y.-H. Audrey Li. 1993. Syntax of scope. Cambridge, MA: MIT Press.

Butler, Jonny. 2003. A minimalist treatment of modality. Lingua 113: 967-996.

Cheng, L.-S. Lisa. 1991. On the typology of wh-questions. Ph.D. diss., MIT.

Cheng, L.-S. Lisa. 1995. On dou quantification. Journal of East Asian Linguistics 4:197-234.

Chiu, Bonnie. 1990. A case of quantifier floating in Mandarin Chinese. Paper presentedat the Northeast Conference on Chinese Linguistics (NACCL), Cornell University.

Chiu, Bonnie. 1993. The inflectional structure of Mandarin Chinese. Ph.D. diss., UCLA.

Cinque, Guglielmo. 1999. Adverbs of functional heads: a cross-linguistic perspective.New York: Oxford University Press.

Cinque, Guglielmo. 2006. Restructuring and functional heads. In Structures and beyond:The cartography of syntactic structures, ed. Adriana Belleti, vol. 3, 132-191. NewYork: Oxford University Press.

Chomsky, Noam. 2000. Minimalist inquiries. In Step by step: Essays on minimalistsyntax in honor of Howard Lasnik, ed. R. Martin, D. Michaels, and J. Uriagereka,89156. Cambridge, MA: MIT Press.

Chomsky, Noam. 2001. Derivation by phase. In Ken Hale: A life in language, ed.Michael Kenstowicz, 152. Cambridge, MA: MIT Press.

Chomsky, Noam. 2004. Beyond explanatory adequacy. In Structures and beyond: Thecartography of syntactic structures, ed. Adriana Belleti, vol. 3, 104-131. NewYork: Oxford University Press.

Chomsky, Noam. 2005. Three factors in language design. Linguistics Inquiry 36: 1-22.

Chomsky, Noam. 2006. On phases. In Foundational issues in linguistic theory, ed.Robert Freidin, Carlos Otero, and Maria-Luisa Zubizaretta. Cambridge, MA:MIT Press.

Chomsky, Noam. 2007. Approaching UG from Below. In Interfaces + recursion =language?: Chomsky’s minimalism and the view from syntax-semantics, ed. UliSauerland and Hans-Martin Gartner, 1-29. New York: Mouton de Gruyter.

Drubig, Hans Bernhard. 2001. On the syntactic form of epistemic modality. Ms. Avail-able at http://www.sfb441.uni-tuebingen.de/b2/papers/DrubigModa-lity.pdf

Gagnon, Michael and Alexis Wellwood. 2010. The topology of distributivity and epis-temic containment phenomena. In the Proceedings of the 2010 annual conferenceof the Canadian Linguistic Association, ed. Melinda Heijl, 115-127.

Huang, C.-T. James. 1982a. Move wh in a language without wh-movement. The Lin-guistic Review 1: 369-416.

24

Huang, C.-T. James. 1982b. Logical relations in Chinese and the theory of grammar.Ph.D. diss., MIT.

Huang, C.-T. James. 1983. On the representation of scope in Chinese. Journal of ChineseLinguistics 1: 37-91.

Huitink, Janneke. 2006. Can quantifier scope over epistemic modals? Ms. University ofNijmegen.

Huitink, Janneke. 2008. Scope over epistemics in English and Dutch. Ms. University ofNijimegen.

Kratzer, Angelika. 1977. What ‘must’ and ‘can’ must and can mean. Linguistics andPhilosophy 1: 337-355.

Kratzer, Angelika. 1981. The notional category of modality. In Words, worlds, andcontexts: New approaches in word semantics, ed. Eikmeyer, H. J. and Rieser, H.,38-74. Berlin: de Gruyter.

Kratzer, Angelika. 1989. An investigation of the lumps of thought. Linguistics andPhilosophy 12: 607-653.

Kratzer, Angelika. 1991. Modality. In Semantics: An international handbook of con-temporary research, ed. Von Stechow, A and Wunderlich, D., 639-650. Berlin: deGruyter.

Lee, H.-T. Thomas. 1986. Studies on quantifications in Mandarin Chinese. Ph.D. diss.,UCLA.

Li, Y.-H. Audrey. 1990. Order and constituency in Mandarin Chinese. Dordrecht:Kluwer Academic Publishers.

Li, Y.-H. Audrey. 1992. Indefinite wh in Mandarin Chinese. Journal of East AsianLinguistics 1: 125-155.

Lin, Jo-wang. 1996. Polarity licensing and wh-phrase quantification in Chinese. Ph.D.diss., UMass.

Lin, Jo-wang. 1998a. Distributivity in Chinese and its implications. Natural LanguageSemantics 6: 201-243.

Lin, Jo-wang. 1998b. On existential polarity wh-phrases in Chinese. Journal of EastAsian Linguistics 7. 219-255.

Lin, T.-H. Jonah. 1998. On ge and other related problems. In The referential propertiesof Chinese noun phrases, ed. Liejiong Xu, 209-253. Paris.

Lyons, John, 1977. Semantics. Cambridge: Cambridge University Press.

May, Robert. 1985. Logical form: its structure and derivation. Cambridge, MA: MITPress.

Miyagawa, Shigeru. 2010. Why agree? Why move? Unifying agreement-based anddiscourse-configurational languages. Cambridge, MA: MIT Press.

Papafragou, Anna. 2006. Epistemic modality and truth conditions. Lingua 116: 1688-1702.

Pesetsky, David. 1987. Wh-in-situ: movement and unselective binding. In The repre-sentation of (in)definiteness, ed. Eric J. Reuland and Alice ter Meulen, 98-129.Cambridge, MA: MIT Press.

25

Pesetsky, David. 2000. Phrasal movement and its kin. Cambridge, MA: MIT Press.

Pesetsky, David and Esther Torrego. 2007. The syntax of valuation and the interpretabil-ity of features. In Phrasal and clausal architecture: Syntactic derivation and in-terpretation, ed. S. Karimi, V. Samiian and W.Wilkins, 262-294. Amsterdam:Benjamins.

Rizzi, Luigi. 1997. The fine structure of the left periphery. In Elements of grammar, ed.Liliane Haegeman, 281-337. Dordrecht: Kluwer Academic Publishers.

Rizzi, Luigi. 2001. On the position “Int(errogative)” in the left periphery of the clause. InCurrent studies in Italian syntax, essays offered to Lorenzo Renzi, eds. GuglielmoCinque and Giampaolo Salvi, 287-296. Amsterdam: Elsevier.

Rizzi, Luigi. 2004. Locality and left periphery. In Structures and beyond: The cartographyof syntactic structures, ed. Adriana Belletti, vol. 3, 223-251. New York: OxfordUniversity Press.

Rizzi, Luigi. 2006a. On the form of chains: Criterial positions and ECP effects. In Whmovement: Moving on, ed. L. Cheng and N. Corver, 97-133. Cambridge, MA:MIT Press.

Rizzi, Luigi. 2006b. On intermediate positions: Intervention and impenetrability. Paperpresented at EALing 2006, Paris.

Schwarzschild, Roger. 1991. On the meaning of definite plural noun phrases. Ph.D. diss.,UMass.

Schwarzschild, Roger. 1996. Pluralities. Dordrecht: Kluwer.

Stepanov, Arthur, and Wei-Tien Dylan Tsai. 2008. Cartography and licensing of wh-adjuncts: A crosslinguistic perspective. Natural Language and Linguistic Theory26: 589-638.

Swanson, Eric. 2010. On scope relations between quantifiers and epistemic modals.Journal of Semantics 27: 529-540.

Tancredi, Christopher. 2007. A multi-model modal theory of I-semantics. Ms. Universityof Tokyo.

Tsai, W.-T. Dylan. 1999. On economizing the theory of A-bar dependencies. Ph.D. diss.,MIT.

Tsai, W.-T. Dylan. 2001. On subject specificity and theory of syntax-semantics interface.Journal of East Asian Linguistics 10: 129-168.

Tsai, W.-T. Dylan. 2003. Three types of existential quantification in Chinese. In Func-tional structure(s), form and interpretation: Pespectives from Asian Languages,ed. Audrey Li and Andrew Simpson, 161-179, London, Routledge Curzon.

Tsai, W.-T. Dylan. 2008. Left-periphery and how -why alternation. Journal of EastAsian Linguistics 17: 83-115.

Tsai, W.-T. Dylan. 2009. Actuality entailments and the topography of Chinese modals.Paper presented in the 7th GLOW in Asia, English and Foreign Languages Uni-versity, Hyderabad, India.

Tsai, W.-T. Dylan. 2010. EPP as a Topic feature: evidence from Chinese applicatives,modals, and refutory whs. Ms. National Tsing Hua University.

26

von Fintel, Kai, and Anthony Gillies. 2011. ‘Might ’ made right. In Epistemic modality,ed. Andy Egan & Brian Weatherson, 108-130. Oxford: Oxford University Press.

von Fintel, Kai, and Sabine Iatridou. 2003. Epistemic containment. Linguistic Inquiry34: 173-198.

Westmoreland, Robert. 1998. Information and intonation in natural language modality.Ph.D. diss., Indiana University.

Wu, J.-X.. 1999. The syntax and semantics of quantification in Chinese. Ph.D. Diss.,University of Maryland.

Yang, C.-Y. Barry. 2002. Quantification and its scope interpretation in Mandarin Chi-nese. MA Thesis, National Tsing Hua University.

Yang, C.-Y. Barry. 2008. Intervention effects and the covert component of grammar.Ph.D. diss., National Tsing Hua University.

Yang, C.-Y. Barry. To appear. Intervention effect and wh-construal. Journal of EastAsian Linguistics.

Zhang, N. Niina. 1997. Syntactic dependencies in Mandarin Chinese. Ph.D. Diss. Uni-versity of Toronto.

Zhang, N. Niina. 2008. Encoding exhaustivity. UST Working Papers in Linguistics 4:133-143.

27

Documents

(A-)symmetric Epistemic Intervention in Quanti cation and