Antisingleton Indefinites in Persian · 2019-10-21 · I Persian is among these languages with Japanese, Quechua, and Turkish. I In this talk, I focus on singular definite and indefinite

INTRO NP YE-NP YE-NP-I NP-I References

Antisingleton Indefinites in Persian

Masoud Jasbi

Department of LinguisticsStanford University

California Universities

Semantics and Pragmatics (CUSP 8)


DEFINITENESS

I No definite but indefinite article: around 7% of the languages in WALS.I Persian is among these languages with Japanese, Quechua, and Turkish.I In this talk, I focus on singular definite and indefinite descriptions in Modern

Tehrani Colloquial Persian.

(Dryer, 2013; Dryer and Haspelmath, 2013)


DEFINITENESS IN PERSIAN

I No overt marker of definiteness.I There are two indefinite markers1:

i. The indefinite determiner “ye”.ii. The indefinite clitic “-i”.

I Four constructions:1. NP Definite2. ye-NP Indefinite3. ye-NP-i Antisingleton Indefinite4. NP-i Antisingleton Indefinite

1See Ghomeshi (2003); Toosarvandani and Nasser (2015) for some previousdiscussions of these markers



I No overt marker of definiteness.I There are two indefinite markers1:

i. The indefinite determiner “ye”.ii. The indefinite clitic “-i”.

I Four constructions:1. NP Definite2. ye-NP Indefinite3. ye-NP-i Antisingleton Indefinite4. NP-i Antisingleton Indefinite

1See Ghomeshi (2003); Toosarvandani and Nasser (2015) for some previousdiscussions of these markers


MAIN QUESTIONS

I What are the empirical facts on singular (in)definites in Persian?I What is the semantic contribution of each indefinite morpheme?I Can we have a compositional story when they appear together?

(ye-NP-i)

I Good News: Yes!


MAIN QUESTIONS

I What are the empirical facts on singular (in)definites in Persian?I What is the semantic contribution of each indefinite morpheme?I Can we have a compositional story when they appear together?

(ye-NP-i)I Good News: Yes!


PLAN FOR THE TALK

Construction Form Implication Projective CGDefinite NP

Indefinite ye-NPAntisingleton ye-NP-iIndefiniteAntisingleton NP-iIndefinite


PLAN FOR THE TALK

Construction Form Implication Projective CGDefinite NP

Indefinite ye-NPAntisingleton ye-NP-iIndefiniteAntisingleton NP-iIndefinite


BARE NOMINALS: IMPLICATIONS

I Bare nominals can carry EXISTENCE and UNIQUENESSimplications2.

(1) [The Falling Child Scenario | There is a family party. Suddenly there comes aloud thud. The host asks what happened. Someone says the followingsentence. How many children are at the party?]

[S bache]child

[V oftad]fall.PERF.3.SG

“The child fell.”

|CHILD|= 0 1 2+# X #

2They can also be interpreted as generic.


BARE NOMINALS: PROJECTIONI The EXISTENCE and UNIQUENESS implications of the bare

nominals are projective.

(2) a. Question:

bachechild

oftad?fall.PERF.3.SG

“Did the child fall?”b. Negation:

bachechild

na-yoftadNEG-fall.PERF.2.SG

“The child didn’t fall.”c. Antecedent of Conditionals:

ageif

bachechild

oftad,fall.PERF.2.SG,

be-mto-1.SG

be-guIMP-say

“Tell me if the child fell.”d. Possibility Modal:

shayadmaybe

bachechild

oftadfall.PERF.2.SG

“Maybe the child fell.”


BARE NOMINALS: CG STATUS

I They also need to be COMMON GROUND between the speakerand the addressee.

(3) a. [The Sleeping Child Scenario: Uniqueness | Amir goes to a family party.When he comes back home his wife asks if there were any children at theparty. Amir says yes and continues with . . . ]# bache

childhamashall.3.SG.CLC

xabsleep

budbe.PST.3.SG

“The child was constantly asleep.”b. [The Sleeping Child Scenario: Existence | In Amir’s family, only his sister

has a child but sometimes she leaves the child with a babysitter for familyparties. Amir comes back home. His wife asks how the party was. Hesays: “oh it was great . . . ]! bache

childhamashall.3.SG.CLC

xabsleep

budbe.PST.3.SG

“The child was constantly asleep.”


Construction Form Implication Projective CGDefinite NP Existence Yes Yes

Uniqueness Yes YesIndefinite ye-NPAntisingleton ye-NP-iIndefiniteAntisingleton NP-iIndefinite


BARE NOMINALS: FORMAL MODELING

FALL(ιx[CHILD(x)])t

λy[FALL(y)]et

oftad

ιx[CHILD(x)]e

λx[CHILD(x)]et

bache

IOTA



Uniqueness Yes YesIndefinite ye-NPAntisingleton ye-NP-iIndefiniteAntisingleton NP-iIndefinite


YE-NP: IMPLICATIONS

I NPs modified by the indefinite determiner “ye” carry anEXISTENCE implication but no UNIQUENESS implication.

(4) [The Falling Child Scenario ]

[S yeIndef.D

bache]child


“A child fell.”

|CHILD|= 0 1 2+# X X

I “ye-np” is compatible with a singleton indefinite reading.


YE-NP: IMPLICATIONS

I NPs modified by the indefinite determiner “ye” carry anEXISTENCE implication but no UNIQUENESS implication.


[S yeIndef.D

bache]child


“A child fell.”

|CHILD|= 0 1 2+# X X

I “ye-np” is compatible with a singleton indefinite reading.


YE-NP: PROJECTION

I The EXISTENCE implication of a “ye”-indefinite is not projective.

(5) a. Question:

yeIndef.D

bachechild


“Did a child fall?”b. Negation:

yeIndef.D

bachechild


“A child didn’t fall.” (∃ > ¬)c. Antecedent of Conditionals:

ageif

yeIndef.D

bachechild


be-mto-1.SG

be-guIMP-say

“Tell me if a child fell.”d. Possibility Modal:

shayadmaybe

yeIndef.D

bachechild

oftadfall.PERF.2.SG

“Maybe a child fell.”


YE-NP: CG STATUS

I The existence implication of the indefinite determiner need notbe common ground between the speaker and the addressee.

(6) [The Crying Baby Scenario | Amir comes back from a trip and his wife askshow the flight was. He says: ]

yeIndef.D

bachechild

hamashall.3.SG.CLC

geryecry

mi-kardIMPERF-be.PST.3.SG

“A child was constantly crying.”



Uniqueness Yes YesIndefinite ye-NP Existence No NoAntisingleton ye-NP-iIndefiniteAntisingleton NP-iIndefinite

I The crucial difference between definites and indefinites is theirstatus with respect to the CG.




I The crucial difference between definites and indefinites is theirstatus with respect to the CG.


YE-NP INDEFINITES: FORMAL MODELING

∃x[CHILD(x) ∧ FALL(x)]t

λy[FALL(y)]et

oftad

λQ[∃x[CHILD(x) ∧Q(x)]]〈et,t〉

λx[CHILD(x)]et

bache

λPλQ[∃x[P(x) ∧Q(x)]]〈et,〈et,t〉〉

ye






YE-NP-I: IMPLICATIONS

I NPs marked by both the indefinite determiner and the indefiniteclitic carry an ANTIUNIQUENESS implication.


[S yeIndef.D

bache-i ]child-Indef.C


“A child fell.”

|CHILD|= 0 1 2+# # X

I Here the domain is plural but the claim is not about a plurality ofindividuals. A single child falling verifies the statement.


YE-NP-I: IMPLICATIONS

I NPs marked by both the indefinite determiner and the indefiniteclitic carry an ANTIUNIQUENESS implication.


[S yeIndef.D

bache-i ]child-Indef.C


“A child fell.”

|CHILD|= 0 1 2+# # X

I Here the domain is plural but the claim is not about a plurality ofindividuals. A single child falling verifies the statement.


THE ANTIUNIQUENESS IMPLICATION

I Given |JOB| = 1:

(8) [You Had One Job! | . . . ]

a. [YEIndef.D

kar]work

be-tto-2.SG

sepord-amgive-1.SG

kethat

beto

un-amthat-even/too

gand-zad-istink-hit-2.SG

“You had ONE job and you even screwed that up!”

b. # [yeIndef.D

kar- i ]work-Indef.C

be-tto-2.SG

sepord-amgive-1.SG

kethat

beto

un-amthat-even/too

gand-zad-istink-hit-2.SG


YE-NP-I: PROJECTIONI I set this aside for now. As we will see, antiuniqueness in Persian

is the combination of two implications in my analysis and eachneeds to be checked independently.

(9) a. Question:

yeIndef.D

bache-ichild-Indef.C


“Did a child fall?”b. Negation:

yeIndef.D



“A child didn’t fall.” (∃ > ¬)c. Antecedent of Conditionals:

ageif

yeIndef.D



be-mto-1.SG

be-guIMP-say

“Tell me if a child fell.”d. Possibility Modal:

shayadmaybe

yeIndef.D


oftadfall.PERF.2.SG

“Maybe a child fell.”


YE-NP-I: CG STATUS

I The antiuniqueness implication does not have to be commonground between the speaker and the addressee.

(10) [The Dancing Party Scenario | There was a party last night and Ali dancedwith a girl. Reza is spreading the gossip the next morning: Guys, you knowwhat happened last night? . . .

AliAli

bawith

[yeIndef.D

doxtar- i ]girl-Indef.C

raqsiddance.PERF.3.SG

“Ali danced with a girl!”



Uniqueness Yes YesIndefinite ye-NP Existence No NoAntisingleton ye-NP-i Antiuniqueness NoIndefiniteAntisingleton NP-iIndefinite



Uniqueness Yes YesIndefinite ye-NP Existence No NoAntisingleton ye-NP-i Antiuniqueness NoIndefiniteAntisingleton NP-iIndefinite


NP-I

I The indefinite clitic cannot generate an indefinite reading byitself in positive episodic sentences.

(11) ∗ to2.SG

sib- iapple-Indef.C

xord-ieat.PREF-2.SG

“You ate an apple.”

I The acceptability of such sentences can be salvaged in two ways:1. Modification.2. Embedding.


NP-I: MODIFICATION

(12) Quantifiers:

a. to2.SG

yeIndef.D

sib- iapple-Indef.C

xord-ieat.PERF-2.SG

“You ate an apple.”

b. ? to2.SG

hareach

sib- iapple-Indef.C

xord-ieat.PERF-2.SG

“You ate every apple.”

c. to2.SG

hichno

sib- iapple-Indef.C

na-xord-iNEG-eat.PERF-2.SG

“You didn’t eat any apples.”

(13) Relative Clauses:

a. to2.SG

sib- iapple-Indef.C

kethat

avord-ibring.PERF-2.SG

xord-ieat.PERF-2.SG

“You ate the apple you brought.”


NP-I: EMBEDDING

(14) a. Question:

to2.SG

sib- iapple-Indef.C

xord-i?eat.PERF-2.SG

“Did you eat any apples?”b. Negation:

to2.SG

sib- iapple-Indef.C


“You didn’t eat any apples.” (¬ > ∃).c. Antecedent of Conditional:

ageif

sib- iapple-Indef.C

xord-i,eat.PERF-2.SG,

be-mto-1.SG

be-guIMP-say

“If you ate any apples, tell me.”d. Possibility Modal:

shayadmaybe

sib- i2.SG

xord-iapple-Indef.C eat.PERF-2.SG

“Maybe you ate an apple”


NP-I: IMPLICATION AND PROJECTION

(15) a. Negation:

to2.SG

sib- iapple-Indef.C


“You didn’t eat any apples.” (¬ > ∃).b. Antecedent of Conditional:

ageif

sib- iapple-Indef.C


be-mto-1.SG

be-guIMP-say

“If you ate any apples, tell me.”

I No existence implication.

I However, these sentences can never mean that a specific entity satisfiesthe claim (wide scope existential).

I Therefore, “NP-i” constructions are incompatible with singleton NPsbut unlike the Spanish algun (Alonso-Ovalle and Menendez-Benito,2010), they are compatible with situations where no entity satisfies thedescription.

I This ANTISINGLETON implication (|NP| 6= 1) is projective.



(15) a. Negation:

to2.SG

sib- iapple-Indef.C



ageif

sib- iapple-Indef.C


be-mto-1.SG

be-guIMP-say


I No existence implication.I However, these sentences can never mean that a specific entity satisfies

the claim (wide scope existential).

I Therefore, “NP-i” constructions are incompatible with singleton NPsbut unlike the Spanish algun (Alonso-Ovalle and Menendez-Benito,2010), they are compatible with situations where no entity satisfies thedescription.




(15) a. Negation:

to2.SG

sib- iapple-Indef.C



ageif

sib- iapple-Indef.C


be-mto-1.SG

be-guIMP-say



the claim (wide scope existential).I Therefore, “NP-i” constructions are incompatible with singleton NPs

but unlike the Spanish algun (Alonso-Ovalle and Menendez-Benito,2010), they are compatible with situations where no entity satisfies thedescription.




(15) a. Negation:

to2.SG

sib- iapple-Indef.C



ageif

sib- iapple-Indef.C


be-mto-1.SG

be-guIMP-say



the claim (wide scope existential).I Therefore, “NP-i” constructions are incompatible with singleton NPs

but unlike the Spanish algun (Alonso-Ovalle and Menendez-Benito,2010), they are compatible with situations where no entity satisfies thedescription.



NP-I: CG STATUS

I The antisingleton implication (constraint) does not have to becommon ground between the speaker and the addressee.

(16) a. ageif

sib- iapple-Indef.C


be-mto-1.SG

be-guIMP-say




Uniqueness Yes YesIndefinite ye-NP Existence No NoAntisingleton ye-NP-i Antiuniqueness NoIndefiniteAntisingleton NP-i Antisingleton Yes NoIndefinite




I “ye” EXISTENCE (|NP| ≥ 1)I “-i” ANTISINGLETON (|NP| 6= 1)I “ye” + “-i” EXISTENCE + ANTISINGLETON (|NP| > 1)

I ANTIUNIQUENESS




I “ye” EXISTENCE (|NP| ≥ 1)

I “-i” ANTISINGLETON (|NP| 6= 1)I “ye” + “-i” EXISTENCE + ANTISINGLETON (|NP| > 1)

I ANTIUNIQUENESS




I “ye” EXISTENCE (|NP| ≥ 1)I “-i” ANTISINGLETON (|NP| 6= 1)

I “ye” + “-i” EXISTENCE + ANTISINGLETON (|NP| > 1)I ANTIUNIQUENESS





I ANTIUNIQUENESS





I ANTIUNIQUENESS






Uniqueness Yes YesIndefinite ye-NP Existence No NoAntisingleton ye-NP-i ExistenceIndefinite AntisingletonAntisingleton NP-i Antisingleton Yes NoIndefinite



Uniqueness Yes YesIndefinite ye-NP Existence No NoAntisingleton ye-NP-i Existence No NoIndefinite AntisingletonAntisingleton NP-i Antisingleton Yes NoIndefinite



Uniqueness Yes YesIndefinite ye-NP Existence No NoAntisingleton ye-NP-i Existence No NoIndefinite Antisingleton Yes NoAntisingleton NP-i Antisingleton Yes NoIndefinite








YE-NP-I INDEFINITES: FORMAL MODELING

∃x[CHILD(x) ∧ FALL(x)] • |CHILD| 6= 1

λy[FALL(y)]et

oftad

λQ[∃x[CHILD(x) ∧Q(x)]] • |CHILD| 6= 1

CHILD • |CHILD| 6= 1

λP[|P| 6= 1]

-i

CHILDet

bache

λPλQ[∃x[P(x) ∧Q(x)]]〈et,〈et,t〉〉

ye


THANK YOU!

I An indefinite number of thanks to:I Cleo Condoravdi and James Collins for their continued help and

support with this project.I Chris Potts, Beth Levin, and the members of the Semprag group at

Stanford.I My patient and helpful informants: Elham Jasbi, Tahereh

Hajebrahimi, Amir Azad, Amire-Reza Saheb, Taha Hajitarkhani,Morteza Farazmand, Mohammad-Reza Shahidi, Hasan Ettefagh,and Ata Esmaili-Fazel.


REFERENCES

Alonso-Ovalle, L. and Menendez-Benito, P. (2010). Modal indefinites.Natural Language Semantics, 18(1):1–31.

Dryer, M. S. (2013). Definite Articles. Max Planck Institute forEvolutionary Anthropology, Leipzig.

Dryer, M. S. and Haspelmath, M., editors (2013). WALS Online. MaxPlanck Institute for Evolutionary Anthropology, Leipzig.

Ghomeshi, J. (2003). Plural marking, indefiniteness, and the nounphrase. Studia Linguistica, 57(2):47–74.

Toosarvandani, M. and Nasser, H. (2015). Quantification in persian.In Keenan, E. L. and Paperno, D., editors, Handbook of quantifiers innatural language. Springer, 2nd edition.

Documents

Antisingleton Indefinites in Persian · 2019-10-21 · I Persian is among these languages with Japanese, Quechua, and Turkish. I In this talk, I focus on singular definite and indefinite