A pragmatic explanation of dou -support Mingming Liu Tsinghua - - PowerPoint PPT Presentation

A pragmatic explanation of dou-support

Mingming Liu

Tsinghua University

CLSW 2020

1

Where we are

The puzzle and a solution Mei-NPs are quantificational Non-quantificational dou Obligatory presuppositions A more nuanced characterization of obligatory-dou

2

Mandarin universal quantifiers and dou-support

◮ Mandarin, universally quantified DPs in preverbal positions need the support

• f dou, which is usually glossed as ‘all’ in this environment.

(1) a. Obligatory-dou Mei.ge/Suoyou every/all san.nianji third.grade xuesheng student ∗(dou) dou lai.le. come “Every/all third-grade student(s) came.” b. No additional all Every/all third-grade student(s) (∗all) came.

◮ This is puzzling, since if mei/suoyou-NPs are universal quantifiers like English

every/all-NPs, it is unclear why an additional all is required (or even possible)

3

An influential solution: Lin (1998); Zhang & Pan (2019)

◮ Mandarin mei-NPs are referential, synonymous with the-NPs.

(2) a. mei.ge san.nianji xuesheng =

• third.grade.student

b. mei =

• c.

mei.ge san.nianji xueshengc1=zhexie san.nianji xueshengc1 =zs ⊕ ls ⊕ ww

4

An influential solution: Lin (1998); Zhang & Pan (2019)

◮ Mandarin mei-NPs are referential, synonymous with the-NPs.

(2) a. mei.ge san.nianji xuesheng =

• third.grade.student

b. mei =

• c.

mei.ge san.nianji xueshengc1=zhexie san.nianji xueshengc1 =zs ⊕ ls ⊕ ww

◮ Dou is a distributivity operator, similar to English each.

(3) douLin = λPλx∀y[y ≤atom x → P(y)] (cf. Link (1987) D operator) (4) (dou forces dist-reading) Zhangsan Zhangsan he and Lisi Lisi dou dou hua.le draw.asp liang.fu two.cl hua. pictures “Zhangsan and Lisi each drew two pictures.”

4

An influential solution: Lin (1998); Zhang & Pan (2019)

◮ Mandarin mei-NPs are referential, synonymous with the-NPs.

(2) a. mei.ge san.nianji xuesheng =

• third.grade.student

b. mei =

• c.

mei.ge san.nianji xueshengc1=zhexie san.nianji xueshengc1 =zs ⊕ ls ⊕ ww

◮ Dou is a distributivity operator, similar to English each.

(3) douLin = λPλx∀y[y ≤atom x → P(y)] (cf. Link (1987) D operator) (4) (dou forces dist-reading) Zhangsan Zhangsan he and Lisi Lisi dou dou hua.le draw.asp liang.fu two.cl hua. pictures “Zhangsan and Lisi each drew two pictures.”

◮ Mei-NP requires dou in order to express a quantificational meaning ◮ Compositionally, ∀y[y ≤atom

• third.grade.student → came(y)]

≡ ∀y[third.grade.student(y) → came(y)]

◮ Lin decomposes ∀-statement: ≈ maximization + distributivity

4

Problems

Many have followed Lin (1998) in taking dou to be some sort of universal quantifier; but there are problems.

◮ Conceptually, assigning mei-NP a (plural) definite semantics does not really

explain why it needs dou-support: cf. the students came.

5

Problems

Many have followed Lin (1998) in taking dou to be some sort of universal quantifier; but there are problems.

◮ Conceptually, assigning mei-NP a (plural) definite semantics does not really

explain why it needs dou-support: cf. the students came.

◮ Empirically, there is evidence that mei-NPs are genuine distributive universal

quantifiers.

5

Problems

Many have followed Lin (1998) in taking dou to be some sort of universal quantifier; but there are problems.

◮ Conceptually, assigning mei-NP a (plural) definite semantics does not really

explain why it needs dou-support: cf. the students came.

◮ Empirically, there is evidence that mei-NPs are genuine distributive universal

quantifiers.

◮ Explaining dou-support as a syntactic-semantic requirement of mei-NPs is

both too strong and too weak.

◮ Too strong: many mei-NPs in preverbal positions do not need (or even

cannot have) dou (Huang, 1996; Chen & Liu, 2019; Liu, 2019).

◮ Too weak: the phenomenon of obligatory-dou goes beyond mei-NPs:

if the context is right, conjunction of proper names also requires dou. Crucially, Dou-support is sensitive to discourse contexts. A pragmatic story.

5

Where we are

The puzzle and a solution Mei-NPs are quantificational Non-quantificational dou Obligatory presuppositions A more nuanced characterization of obligatory-dou

6

Two types of facts

◮ mei-NPs without dou are very different from plural definites without dou. ◮ mei-NPs with dou are very different from plural definites with dou.

Post-verbal mei-NP’s cannot have dou (an adverb that associates to its lef).

7

Mei-NPs without dou: Non-homogeneous and maximal

Mei-NPs without dou under ¬ retain their ∀, while Plural definites under ¬ are ∃. (5) Non-homogeneous a. Wo I meiyou not ba ba zhe.jian.shi this.thing gaosu tell mei.yi.ge every san.nianji third.grade xuesheng. student “I didn’t tell this to every third-grade student.” ¬ > ∀ b. Wo I meiyou not ba ba zhe.jian.shi this.thing gaosu tell zhe.xie these san.nianji third.grade xuesheng. student “I didn’t tell this to these third-grade students.” ≈ I didn’t tell this to any of these third-grade students. ¬ > ∃

8

Mei-NPs without dou: Non-homogeneous and maximal

Mei-NPs without dou under ¬ retain their ∀, while Plural definites under ¬ are ∃. (5) Non-homogeneous a. Wo I meiyou not ba ba zhe.jian.shi this.thing gaosu tell mei.yi.ge every san.nianji third.grade xuesheng. student “I didn’t tell this to every third-grade student.” ¬ > ∀ b. Wo I meiyou not ba ba zhe.jian.shi this.thing gaosu tell zhe.xie these san.nianji third.grade xuesheng. student “I didn’t tell this to these third-grade students.” ≈ I didn’t tell this to any of these third-grade students. ¬ > ∃ (6) Maximal a. Lisi Lisi qing.le invite.asp mei.yi.ge every san.nianji third.grade xuesheng. student “Lisi invited every third-grade student.” (Maximal only) b. Lisi Lisi qing.le invite.asp zhe.xie these san.nianji third.grade xuesheng. student “Lisi invited these third-grade students.” (Non-maximal allowed) Mei-NPs do not exhibit homogeneity and non-maximality, two well-known properties of plural definties across many languages, and they always retain their maximal universal quantificational force, in both positive and negative contexts.

8

Mei-NPs without dou: Q-sensitive expressions

(7) Q-sensitive expressions a. I invited {every boy/#the boys} but John. b. I invited almost {every boy/#the boys}.

9

Mei-NPs without dou: Q-sensitive expressions

(7) Q-sensitive expressions a. I invited {every boy/#the boys} but John. b. I invited almost {every boy/#the boys}. (8) Mei-NPs without dou are compatible with Q-sensitive expressions a. Chule Except Lisi, Lisi, wo I qing.le invite.asp mei.ge every san.nianji third.grade xuesheng. student “I invited every third-grade student but Lisi.” b. Lisi Lisi jihu almost qing.le invite.asp mei.ge every san.nianji third.grade xuesheng. student “Lisi invited almost every third-grade student.”

9

Mei-NPs without dou: Q-sensitive expressions

(7) Q-sensitive expressions a. I invited {every boy/#the boys} but John. b. I invited almost {every boy/#the boys}. (8) Mei-NPs without dou are compatible with Q-sensitive expressions a. Chule Except Lisi, Lisi, wo I qing.le invite.asp mei.ge every san.nianji third.grade xuesheng. student “I invited every third-grade student but Lisi.” b. Lisi Lisi jihu almost qing.le invite.asp mei.ge every san.nianji third.grade xuesheng. student “Lisi invited almost every third-grade student.” (9) Plural definites without dou are incompatible with Q-sensitive expressions a. #Chule Except Lisi, Lisi, wo I qing.le invite.asp zhe.xie these san.nianji third.grade xuesheng. student “#I invited these third-grade students but Lisi.” b. #Lisi Lisi jihu almost qing.le invite.asp zhe.xie these san.nianji third.grade xuesheng. student “#Lisi invited almost these third-grade student(s).”

9

Mei-NPs without dou: Q-sensitive expressions

(7) Q-sensitive expressions a. I invited {every boy/#the boys} but John. b. I invited almost {every boy/#the boys}. (8) Mei-NPs without dou are compatible with Q-sensitive expressions a. Chule Except Lisi, Lisi, wo I qing.le invite.asp mei.ge every san.nianji third.grade xuesheng. student “I invited every third-grade student but Lisi.” b. Lisi Lisi jihu almost qing.le invite.asp mei.ge every san.nianji third.grade xuesheng. student “Lisi invited almost every third-grade student.” (9) Plural definites without dou are incompatible with Q-sensitive expressions a. #Chule Except Lisi, Lisi, wo I qing.le invite.asp zhe.xie these san.nianji third.grade xuesheng. student “#I invited these third-grade students but Lisi.” b. #Lisi Lisi jihu almost qing.le invite.asp zhe.xie these san.nianji third.grade xuesheng. student “#Lisi invited almost these third-grade student(s).”

◮ Mei-NP without dou still behaves like English every (quantificational).

9

Mei-NPs with dou: no partitives

(10) a. Many of the boxes were stolen. b. *Many of every box were stolen. (11) definites with dou are compatible with partitives while mei-NPs not a. {Zhexie.xuesheng/Tamen} {these.students/they} {daduo/henduo} {most/many} dou dou xihuan like Jin.Yong. Jin.Yong “Most/Many of these students/them like Jin Yong.” b. *Mei.ge every xuesheng student {daduo/henduo} {most/many} dou dou xihuan like Jin.Yong. Jin.Yong “∗Most/many of these students like Jin Yong.”

◮ The contrast shows plural definites but not mei-NPs are referential

sum-denoting expressions, suggesting the later are in fact quantificaitonal elements.

10

Mei-NPs with dou: Pair-list phenomena

Only true distributive universals allow for certain pair-list phenomena (Bumford 2015). (12) Only distributive universals license singular internal different a. Every boy read a different book. Beghelli & Stowell 1997 b. #The boys read a different book. Beghelli & Stowell, Moltmann 1992 c. The boys read different books.

11

Mei-NPs with dou: Pair-list phenomena

Only true distributive universals allow for certain pair-list phenomena (Bumford 2015). (12) Only distributive universals license singular internal different a. Every boy read a different book. Beghelli & Stowell 1997 b. #The boys read a different book. Beghelli & Stowell, Moltmann 1992 c. The boys read different books. (13) Only mei-NPs with dou license singular internal different a. Mei.ge Every xuesheng student dou dou mai.le buy.asp yi.ben

• ne

butong different de de shu. book “Every student bought a different book.” b. #{Zhe.xie.xuesheng these.student /tamen} /they dou dou mai.le buy.asp yi.ben

• ne

butong different de de shu. book “#{These students/they} bought a different book.” (cf. (12b)) c. {Zhe.xie.xuesheng these /tamen} student dou dou mai.le buy.asp butong differnt de de shu. book “These students/they bought different books.”

11

Mei-NPs with dou: Pair-list phenomena (cont.)

Qestions (with singular wh’s) that contain distributive universals allow for real pair-list answers (Krifka, 1992; Dayal, 1992). (14) Which movie did every boy rent least night? a. (Every boy rented) Z. b. Al rented A, Bill rented B, and Carl rented C. (15) Which movie did the boys rent least night? a. (Every boy rented) Z. b. #Al rented A, Bill rented B, and Carl rented C.

12

Mei-NPs with dou: Pair-list phenomena (cont.)

Qestions (with singular wh’s) that contain distributive universals allow for real pair-list answers (Krifka, 1992; Dayal, 1992). (14) Which movie did every boy rent least night? a. (Every boy rented) Z. b. Al rented A, Bill rented B, and Carl rented C. (15) Which movie did the boys rent least night? a. (Every boy rented) Z. b. #Al rented A, Bill rented B, and Carl rented C. (16) a. Mei.ge every xuesheng student dou dou mai.le buy.asp yi.ben

• ne.cl

shenme what shu? book “Which book did every student buy?” b. (Every boy bought) Z. c. Al bought A, Bill bought B, and Carl bought C. (17) a. {Zhe.xie.xuesheng these.students /tamen} /they dou dou mai.le buy.asp yi.ben

• ne.cl

shenme what shu? book “Which book did {these students/them} buy?” b. (Every boy bought) Z. c. #Al bought A, Bill bought B, and Carl bought C.

12

Summary

(18) Evidence for the quantificational status of mei-NPs a. Even without dou, mei-NPs still lack homogeneity and non-maximality — two properties exhibited by plural definites without dou, and retain their maximal universal force in both positive and negative contexts, similar to English every-NPs. b. Even without dou, mei-NPs are still compatible with Q-sensitive expressions, similar to English every-NPs, but different from their plural definite counterparts. c. Mei-NPs are incompatible with partitive constructions, similar to English every-NPs, but different from the corresponding plural definites (even with dou). d. Unlike plural definites with dou, mei-NPs license pair-list phenomena, a property that only true distributive universals have.

13

Where we are

The puzzle and a solution Mei-NPs are quantificational Non-quantificational dou Obligatory presuppositions A more nuanced characterization of obligatory-dou

14

dou as a strongest-prejacent operator

◮ Dou is truth-conditionally vacuous, but presupposes its prejacent to be the

strongest among its alternatives (Liao 2011, Liu 2017, cf. Xiang 2020).

◮ Different ‘uses’ of dou are analyzed by conceptualizing strength differently.

(19) a. Lisi Lisi dou dou mai.le buy.asp yi

• ne

liang cl Tesla. Tesla ‘Even Lisi bought a Tesla.’ Even-dou ←Likelihood b. Mali Mali he and Lisi Lisi dou dou mai.le buy.asp yi

• ne

liang cl Tesla. Tesla ‘Mary and Lisi each bought a Tesla.’ Distributive-dou←Entailment

15

dou as a strongest-prejacent operator

◮ Dou is truth-conditionally vacuous, but presupposes its prejacent to be the

strongest among its alternatives (Liao 2011, Liu 2017, cf. Xiang 2020).

◮ Different ‘uses’ of dou are analyzed by conceptualizing strength differently.

(19) a. Lisi Lisi dou dou mai.le buy.asp yi

• ne

liang cl Tesla. Tesla ‘Even Lisi bought a Tesla.’ Even-dou ←Likelihood b. Mali Mali he and Lisi Lisi dou dou mai.le buy.asp yi

• ne

liang cl Tesla. Tesla ‘Mary and Lisi each bought a Tesla.’ Distributive-dou←Entailment The analysis brings together two prominent accounts of dou: dist (Lin, 1998) vs. max (Giannakidou & Cheng, 2006; Xiang, 2008). (20) San.ge three.cl xuesheng student dou dou mai.le buy.asp wu.ben five.cl shu. book ‘The three students each bought five books.’ (21) Alt=3 :        3 students each bought five books (= π), 2 students each bought five books, 1 students (each) bought a books,        Each is required to ensure entailment among alternatives while the is needed so that the prejacent could entail all the other alternatives:

◮ strongest (in terms of entailment) = distributivity + maximality

15

The puzzle

◮ Why is dou required for mei-NP, which as we have demonstrated is a true

universal quantifier?

16

Where we are

The puzzle and a solution Mei-NPs are quantificational Non-quantificational dou Obligatory presuppositions A more nuanced characterization of obligatory-dou

17

A parallel: obligatory additives and obligatory dou

Additive particles sometimes are obligatory (Kaplan, Krifka, Zeevat, Sæboø, ...). (22) a. John went to the party. Bill went to the party #(, too). b. Mary went swimming yesterday. She went swimming #(again) today. c. Sam was in New York yesterday. He is #(still) there today. (23) Chule In.addition.to Lisi, Lisi Zhangsan Zhangsan #(ye) also lai.le pass.asp ‘In addition to Lisi, Zhangsan also passed.’

18

A parallel: obligatory additives and obligatory dou

Additive particles sometimes are obligatory (Kaplan, Krifka, Zeevat, Sæboø, ...). (22) a. John went to the party. Bill went to the party #(, too). b. Mary went swimming yesterday. She went swimming #(again) today. c. Sam was in New York yesterday. He is #(still) there today. (23) Chule In.addition.to Lisi, Lisi Zhangsan Zhangsan #(ye) also lai.le pass.asp ‘In addition to Lisi, Zhangsan also passed.’ cf. (24) Obligatory-dou Mei.ge every xuesheng student ∗(dou) dou lai.le. come “Every student came.”

18

Obligatory additives and obligatory presupposition

Obligatory presupposition: presupposition triggers gives rise to obligatory presence when their presuppositions are satisfied (Amsili & Beyssade 2009). (25) a. Mary went swimming yesterday. She went swimming again today. b. #Mary went swimming yesterday. She went swimming today. In (25), again presupposes the swimming happened before, and the requirement is locally satisfied by the first clause in (25), and hence again is obligatory.

19

Obligatory additives and obligatory presupposition

Obligatory presupposition: presupposition triggers gives rise to obligatory presence when their presuppositions are satisfied (Amsili & Beyssade 2009). (25) a. Mary went swimming yesterday. She went swimming again today. b. #Mary went swimming yesterday. She went swimming today. In (25), again presupposes the swimming happened before, and the requirement is locally satisfied by the first clause in (25), and hence again is obligatory. (26) {The/#A} sun is shining. (27) I washed {both/#All} of my hands. (28) Sam {knows/#thinks} that Paris is in France.

19

Obligatory additives and obligatory presupposition

Obligatory presupposition: presupposition triggers gives rise to obligatory presence when their presuppositions are satisfied (Amsili & Beyssade 2009). (25) a. Mary went swimming yesterday. She went swimming again today. b. #Mary went swimming yesterday. She went swimming today. In (25), again presupposes the swimming happened before, and the requirement is locally satisfied by the first clause in (25), and hence again is obligatory. (26) {The/#A} sun is shining. (27) I washed {both/#All} of my hands. (28) Sam {knows/#thinks} that Paris is in France. Obligatory presupposition can be explained by the pragmatic principle Maximize Presupposition in (29), proposed in Heim (1991). (29) Maximize Presupposition Make your contribution presuppose as much as possible. Maximize Presupposition mandates that a speaker choose among sentences with identical assertive information the one that has more/stronger presuppositions, when the presuppositions are satisfied.

19

Obligatory dou as obligatory presupposition

(30) Explaining obligatory dou via obligatory presupposition a. Obligatory-dou Mei.ge every xuesheng student ∗(dou) dou lai.le. come “Every third-grade student came.” b. ∀x[student(x) → came(x)] Prejacent of dou c. douC S is defined only if ∀q ∈ ALT(S)∩C[S q → S ⊂ q] if defined, dou S = S Semantics of dou d.              student a came, student b came, student c came, ...              Alternatives e. Dou’s prejacent entails all the alternatives and its presupposition satisfied and thus[mei.ge xuesheng dou lai.le] blocks #[mei.ge xuesheng lai.le] via MP

20

Obligatory dou as obligatory presupposition

(30) Explaining obligatory dou via obligatory presupposition a. Obligatory-dou Mei.ge every xuesheng student ∗(dou) dou lai.le. come “Every third-grade student came.” b. ∀x[student(x) → came(x)] Prejacent of dou c. douC S is defined only if ∀q ∈ ALT(S)∩C[S q → S ⊂ q] if defined, dou S = S Semantics of dou d.              student a came, student b came, student c came, ...              Alternatives e. Dou’s prejacent entails all the alternatives and its presupposition satisfied and thus[mei.ge xuesheng dou lai.le] blocks #[mei.ge xuesheng lai.le] via MP

• cf. Zeijlstra (2017), based on some of the positive polarity properties of Dutch

iedereen ‘everybody’, argues that it is a universal quantifier that triggers similar alternatives.

20

Where we are

The puzzle and a solution Mei-NPs are quantificational Non-quantificational dou Obligatory presuppositions A more nuanced characterization of obligatory-dou

21

mei-NPs without dou: irrelevance of individual alternatives

(31) [At a secondhand bookstore] The owner: We are now on sale! Mei.ben sells.for $10 . a John: What about this comic book? It seems brand new! b The owner: Mei.ben dou sells.for $10 . c

22

mei-NPs without dou: irrelevance of individual alternatives

(31) [At a secondhand bookstore] The owner: We are now on sale! Mei.ben sells.for $10 . a John: What about this comic book? It seems brand new! b The owner: Mei.ben dou sells.for $10 . c

◮ In (31a), the owner’s focus was on $10 (stress on shi) and it is naturally

understood that she (as the owner) was assuming that every book was sold at the same price and the QUD is how much IS a book?.

◮ In such a context, individual books are intuitively not relevant to the

Qestion under Discussion, and thus are not in the alternative set that is needed for the interpretation of dou.

◮ As a result, the set of contextually relevant alternatives associated with dou

is the empty set. Assuming that dou, as other alternative sensitive operators, needs to be associated with a non-empty set of alternatives, the absence of dou is correctly predicted.

22

mei-NPs without dou: irrelevance of individual alternatives

(31) [At a secondhand bookstore] The owner: We are now on sale! Mei.ben sells.for $10 . a John: What about this comic book? It seems brand new! b The owner: Mei.ben dou sells.for $10 . c

◮ In (31a), the owner’s focus was on $10 (stress on shi) and it is naturally

understood that she (as the owner) was assuming that every book was sold at the same price and the QUD is how much IS a book?.

◮ In such a context, individual books are intuitively not relevant to the

Qestion under Discussion, and thus are not in the alternative set that is needed for the interpretation of dou.

◮ As a result, the set of contextually relevant alternatives associated with dou

is the empty set. Assuming that dou, as other alternative sensitive operators, needs to be associated with a non-empty set of alternatives, the absence of dou is correctly predicted.

◮ By asking about a particular comic book, John shifed the QUD to which

books are $10?.

◮ In this new context, individual books are clearly relevant and they get into

the alternative set of dou.

◮ Since all the alternatives in this set are entailed by the universal prejacent,

dou’s presupposition is satisfied and its obligatory presence is required by Maximize Presupposition.

22

Alternatives evaluated by other focus sensitive operators

Another type of examples where dou is absent from mei-NPs are cases where there is another focus sensitive operator in the sentence that evaluates alternatives triggered by the mei-NP. (32) a. Mei.ge every zuo.le do.perf zuoye homework de de xuesheng student ∗(dou) dou de.le get.perf gao.fen. high.score ‘Every student who did the homework got a high score (in the exam).’ b. Zhiyou Only mei.ge every zuo.le do.perf Zuoye homework de de xuesheng student de.le get.perf gao.fen. high.score ‘Only every student [who did the homework]F got a high score (in the exam).’ (33) a. Zuotian yesterday mei.ge every lingdao leader ∗(dou) dou ma.le scold.perf Lisi. Lisi ‘Every leader scolded Lisi yesterday.’ b. Zuotian yesterday shi shi mei.ge every Lingdao leader mai.le scold.perf lisi, Lisi, bushi not mei.ge every kuaiji. account ‘It was every leaderF that scolded Lisi yesterday, not every accountant.’

23

Mei-NPs without dou: post-verbal mei-NPs

(34) a. Wo I qing.le invite.perf mei.yi.ge every san.nianji third.grade xuesheng. student “I invited every third-grade student.” b. *Wo I dou dou qing.le invite.perf mei.yi.ge every san.nianji third.grade xuesheng. student

◮ For syntactic reasons, dou only associates with items to its lef and thus (34b)

is ungrammatical.

◮ Consequently, (34b) cannot block (34a) via maximize presupposition

(competition and blocking only happens between grammatical sentences), even if the every-NP in (34a) could trigger individual alternatives.

24

Obligatory dou with conjunction

◮ The proposal predicts that obligatory dou is not limited to mei-NPs.

(35) a. (Question with two alternatives) Zhangsan Zhangsan he and Lisi Lisi shei who lai.le? come.asp ‘Who among Zhangsan and Lisi came?’ b. (Infelicitous without dou) #Zhangsan Zhangsan he and Lisi Lisi lai.le. come.asp ‘#Zhangsan and Lisi came.’ c. (Felicitous with obligatory dou) Zhangsan Zhangsan he and Lisi Lisi dou dou lai.le. come.asp ‘Both Zhangsan and Lisi came’

◮ Interestingly, if the question in (35) is changed into who among Zhangsan, Lisi

and Wangwu came? with three relevant individuals, then (35b) becomes an felicitous answer.

◮ This is expected under our proposal, since in the new context with three

alternatives, Zhangsan and Lisi came does not entail ALL the alternatives, dou’s presupposition not satisfied, and hence Maximize Presupposition does not apply and the blocking effect not observed

25

Conclusions

◮ I have defended the view that mei-NPs are true universal quantifiers while

dou is not, based on a large array of novel empirical facts.

◮ Dou is truth-conditionally vacuous but carries a presupposition that its

prejacent is the strongest among its alternatives.

◮ A pragmatic explanation of the mei-dou co-occurrence has been offered: in

most contexts where a mei-NP is used, its individual alternatives are relevant, and thus the universal prejacent entails all the other alternatives and dou’s strongest-prejacent-presupposition is satisfied;

◮ Maximize Presupposition then mandates that a speaker choose mei-dou

instead of mei without dou, for the former carries more presuppositions.

◮ As we have seen, the proposal predicts a more nuanced distribution of

• bligatory-dou, sensitive to discourse contexts.

26

References

Amsili, Pascal & Claire Beyssade. 2009. Obligatory presuppositions in discourse. In

• P. Kuehnlein, A. Benz & C. Sidner (eds.), Constraints in discourse 2: Pragmatics and beyond,

John Benjamins. Beghelli, Filippo & Tim Stowell. 1997. Distributivity and negation: The syntax of each and

• every. In Anna Szabolcsi (ed.), Ways of scope taking, vol. 65, 71–107. Dordrecht,

Netherlands: Kluwer. Bumford, Dylan. 2015. Incremental quantification and the dynamics of pair-list phenomena. Semantics and Pragmatics 8(9). 1–70. Chen, Zhenyu & Chengfeng Liu. 2019. The functional evolution of ‘mei’ and its co-occurrence with ‘dou’: An interpretation based on rhetoric pragmatics and grammaticalization. Contemporary Rhetoric (2). 56–71. Dayal, Veneeta. 1992. Two types of universal terms in questions. In Nels 22, 443–457. Giannakidou, Anastasia & Lisa L.-S. Cheng. 2006. (in)definiteness, polarity, and the role of wh-morphology in free choice. Journal of Semantics 23(2). 135–183. Heim, Irene. 1991. Artikel und definitheit. In Semantik: Ein internationales Handbuch der zeitgenssischen Forschung, 487–535. Berlin: de Gruyter. Huang, Shizhe. 1996. Qantification and predication in Mandarin Chinese: A case study of dou: University of Pennsylvania dissertation. Krifka, Manfred. 1992. Definite nps aren’t quantifiers. Linguistic Inquiry 23. 156–163.

27

References (cont.)

Liao, Hsiu-Chen. 2011. Alternatives and Exhaustification: Non-Interrogative Uses of Chinese Wh-words. Cambridge: Harvard University dissertation. Lin, Jo-Wang. 1998. Distributivity in Chinese and its implications. Natural Language Semantics 6. 201–243. Link, Godehard. 1987. Generalized quantifiers and plurals. In Peter Gärdenfors (ed.), Generalized quantifiers, 151–180. Berlin, Germany: Springer. Liu, Lin. 2019. On the co-ocurence of ‘mei p’ and dou. Chinese Teaching in the World 33(4). 468–480. Liu, Mingming. 2017. Varieties of alternatives: Mandarin focus particles. Linguistics and Philosophy 40(1). 61–95. Moltmann, Friederike. 1992. Reciprocals and same/different: Towards a semantic analysis. Linguistics and Philosophy 15(4). 411–462. Xiang, Ming. 2008. Plurality, maximality and scalar inferences: a case study of mandarin dou. Journal of East Asian Linguistics 17. 227–245. Xiang, Yimei. 2020. Function alternations of the mandarin particle dou: Distributor, free choice licensor, and ‘even’. Journal of Semantics 37(2). 171–217. Zeijlstra, Hedde. 2017. Universal quantifier PPIs. Glossa: a journal of general linguistics 2(1). Zhang, Lei & Haihua Pan. 2019. A reanalysis of the semantics of mei ‘every’. Contemporary Linguistics 21(4). 492–514.

28