What syntax doesnt feed semantics Fake indexicals as indexicals - - PowerPoint PPT Presentation

What syntax doesn’t feed semantics Fake indexicals as indexicals

Emar Maier

ILLC/University of Amsterdam www.ru.nl/ncs/~emar

ESSLLI Workshop ‘What syntax feeds semantics?’, Hamburg August 11, 2008

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance

Introduction: I

• Kaplan (1977,1989):
• I is indexical, like today
• 1. context-dependent
• 2. directly referential
• 2D semantics

I am speaking ≡ the speaker is speaking

• Heim (1991,2008), Kratzer (1998,2008), Jacobson (2008):
• I is a pronoun, like he
• pronouns have bound and referential readings

Only I did my homework sloppy others didn’t do theirs: ∀x[x = i → ¬do.hw(x, x)]

• today: defend Kaplan

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance

Introduction: I

• Kaplan (1977,1989):
• I is indexical, like today
• 1. context-dependent
• 2. directly referential
• 2D semantics

I am speaking ≡ the speaker is speaking

• Heim (1991,2008), Kratzer (1998,2008), Jacobson (2008):
• I is a pronoun, like he
• pronouns have bound and referential readings

Only I did my homework sloppy others didn’t do theirs: ∀x[x = i → ¬do.hw(x, x)]

• today: defend Kaplan

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance

Introduction: I

• Kaplan (1977,1989):
• I is indexical, like today
• 1. context-dependent
• 2. directly referential
• 2D semantics

I am speaking ≡ the speaker is speaking

• Heim (1991,2008), Kratzer (1998,2008), Jacobson (2008):
• I is a pronoun, like he
• pronouns have bound and referential readings

Only I did my homework sloppy others didn’t do theirs: ∀x[x = i → ¬do.hw(x, x)]

• today: defend Kaplan

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance

Introduction: I

• Kaplan (1977,1989):
• I is indexical, like today
• 1. context-dependent
• 2. directly referential
• 2D semantics

I am speaking ≡ the speaker is speaking

• Heim (1991,2008), Kratzer (1998,2008), Jacobson (2008):
• I is a pronoun, like he
• pronouns have bound and referential readings

Only I did my homework sloppy others didn’t do theirs: ∀x[x = i → ¬do.hw(x, x)]

• today: defend Kaplan

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance

Outline

1 Introduction 2 Fake indexicals

Pronouns in Generative Linguistics VP ellipsis Only

3 . . . as indexicals

Ellipsis resolution by unification Only by unification Avoiding sloppy names

4 De se binding and de re acquaintance

Avoiding de se names

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Outline

1 Introduction 2 Fake indexicals

Pronouns in Generative Linguistics VP ellipsis Only

3 . . . as indexicals

Ellipsis resolution by unification Only by unification Avoiding sloppy names

4 De se binding and de re acquaintance

Avoiding de se names

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

The syntax-semantics interface

DS SS

wh,...

• QR
• LF
• PF

L

• m

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

The syntax-semantics interface

DS SS

wh,...

• QR
• LF
• PF

L

• m

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

The syntax-semantics interface

DS SS

wh,...

• QR
• LF
• PF

L

• m

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Binding and coreference

SS: John did his homework ւ ց

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Binding and coreference

SS: John did his homework ւ ց LF: [John did his homework]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Binding and coreference

SS: John did his homework ւ ց LF: [John did his homework] ↓ L: do.homework.of(j, x)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Binding and coreference

SS: John did his homework ւ ց LF: [John did his homework] ↓ L: do.homework.of(j, x) ↓ m: do.homework.of(j, x)f

w = 1

iff John did homework of f (x)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Binding and coreference

SS: John did his homework ւ ց LF: [John did his homework] ↓ L: do.homework.of(j, x) ↓ m: do.homework.of(j, x)f

w = 1

iff John did homework of f (x) context: his = f (x) = John

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Binding and coreference

SS: John did his homework ւ ց LF: [John did his homework] ↓ L: do.homework.of(j, x) ↓ m: do.homework.of(j, x)f

w = 1

iff John did homework of John context: his = f (x) = John John1 [ t1 did his1 homework]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Binding and coreference

SS: John did his homework ւ ց LF: [John did his homework] ↓ L: do.homework.of(j, x) ↓ m: do.homework.of(j, x)f

w = 1

iff John did homework of John context: his = f (x) = John John1 [ t1 did his1 homework] ↓ λx[do.homework.of(x, x)](j)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Binding and coreference

SS: John did his homework ւ ց LF: [John did his homework] ↓ L: do.homework.of(j, x) ↓ m: do.homework.of(j, x)f

w = 1

iff John did homework of John context: his = f (x) = John John1 [ t1 did his1 homework] ↓ do.homework.of(j, j)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Binding and coreference

SS: John did his homework ւ ց LF: [John did his homework] ↓ L: do.homework.of(j, x) ↓ m: do.homework.of(j, x)f

w = 1

iff John did homework of John context: his = f (x) = John John1 [ t1 did his1 homework] ↓ do.homework.of(j, j) ↓ do.homework.of(j, j)f

w = 1 iff

John did homework of John

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

VP ellipsis

PF: John likes his dad but Peter doesn’t

• ambiguous:

strict: Peter doesn’t like John’s sloppy: Peter doesn’t like his own

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

VP ellipsis

PF: John likes his dad but Peter doesn’t

• ambiguous:

strict: Peter doesn’t like John’s sloppy: Peter doesn’t like his own

• Sag/Williams: reduce to referential–bound ambiguity
• delete an LF constituent at PF if it’s semantically equivalent to

an earlier constituent at LF

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

VP ellipsis

PF: John likes his dad but Peter doesn’t

• ambiguous:

strict: Peter doesn’t like John’s sloppy: Peter doesn’t like his own

• Sag/Williams: reduce to referential–bound ambiguity
• delete an LF constituent at PF if it’s semantically equivalent to

an earlier constituent at LF

LF: John likes his dad but Peter doesn’t like his dad (context: his = his = John)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

VP ellipsis

PF: John likes his dad but Peter doesn’t

• ambiguous:

strict: Peter doesn’t like John’s sloppy: Peter doesn’t like his own

• Sag/Williams: reduce to referential–bound ambiguity
• delete an LF constituent at PF if it’s semantically equivalent to

an earlier constituent at LF

LF: John likes his dad but Peter doesn’t like his dad (context: his = his = John)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

VP ellipsis

PF: John likes his dad but Peter doesn’t

• ambiguous:

strict: Peter doesn’t like John’s sloppy: Peter doesn’t like his own

• Sag/Williams: reduce to referential–bound ambiguity
• delete an LF constituent at PF if it’s semantically equivalent to

an earlier constituent at LF

LF: John likes his dad but Peter doesn’t like his dad (context: his = his = John) like.dad(j, x) ∧ like.dad(p, y)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

VP ellipsis

PF: John likes his dad but Peter doesn’t

• ambiguous:

strict: Peter doesn’t like John’s sloppy: Peter doesn’t like his own

• Sag/Williams: reduce to referential–bound ambiguity
• delete an LF constituent at PF if it’s semantically equivalent to

an earlier constituent at LF

LF: John likes his dad but Peter doesn’t like his dad (context: f(x)=f(y)=j) like.dad(j, x) ∧ like.dad(p, y)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

VP ellipsis

PF: John likes his dad but Peter doesn’t

• ambiguous:

strict: Peter doesn’t like John’s sloppy: Peter doesn’t like his own

• Sag/Williams: reduce to referential–bound ambiguity
• delete an LF constituent at PF if it’s semantically equivalent to

an earlier constituent at LF

LF: John likes his dad but Peter doesn’t like his dad (context: f(x)=f(y)=j) like.dad(j, x) ∧ like.dad(p, y) John1 [t1 likes his1 dad] but Peter1 [t1 doesn’t like his1 dad]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

VP ellipsis

PF: John likes his dad but Peter doesn’t

• ambiguous:

strict: Peter doesn’t like John’s sloppy: Peter doesn’t like his own

• Sag/Williams: reduce to referential–bound ambiguity
• delete an LF constituent at PF if it’s semantically equivalent to

an earlier constituent at LF

LF: John likes his dad but Peter doesn’t like his dad (context: f(x)=f(y)=j) like.dad(j, x) ∧ like.dad(p, y) John1 [t1 likes his1 dad] but Peter1 [t1 doesn’t like his1 dad]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

VP ellipsis

PF: John likes his dad but Peter doesn’t

• ambiguous:

strict: Peter doesn’t like John’s sloppy: Peter doesn’t like his own

• Sag/Williams: reduce to referential–bound ambiguity
• delete an LF constituent at PF if it’s semantically equivalent to

an earlier constituent at LF

LF: John likes his dad but Peter doesn’t like his dad (context: f(x)=f(y)=j) like.dad(j, x) ∧ like.dad(p, y) John1 [t1 likes his1 dad] but Peter1 [t1 doesn’t like his1 dad] λx[like.dad(x, x)](j)∧ ¬λx[like.dad(x, x)](p)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

VP ellipsis

PF: John likes his dad but Peter doesn’t

• ambiguous:

strict: Peter doesn’t like John’s sloppy: Peter doesn’t like his own

• Sag/Williams: reduce to referential–bound ambiguity
• delete an LF constituent at PF if it’s semantically equivalent to

an earlier constituent at LF

LF: John likes his dad but Peter doesn’t like his dad (context: f(x)=f(y)=j) like.dad(j, x) ∧ like.dad(p, y) John1 [t1 likes his1 dad] but Peter1 [t1 doesn’t like his1 dad] like.dad(j, j)∧ ¬like.dad(p, p)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Binding an indexical?

PF: I like my job, but Sue doesn’t ւ ց

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Binding an indexical?

PF: I like my job, but Sue doesn’t ւ ց LF: I like my job but Sue doesn’t like my job

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Binding an indexical?

PF: I like my job, but Sue doesn’t ւ ց LF: I like my job but Sue doesn’t like my job like.job(i, i) ∧ ¬like.job(s, i)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Binding an indexical?

PF: I like my job, but Sue doesn’t ւ ց LF: I like my job but Sue doesn’t like my job like.job(i, i) ∧ ¬like.job(s, i) I1 [t1 like my1 job] but Sue1 [t1 doesn’t like my1 job]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Binding an indexical?

PF: I like my job, but Sue doesn’t ւ ց LF: I like my job but Sue doesn’t like my job like.job(i, i) ∧ ¬like.job(s, i) I1 [t1 like my1 job] but Sue1 [t1 doesn’t like my1 job] λx[like.job(x, x)](i)∧ ¬λx[like.job(x, x)](s)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Binding an indexical?

PF: I like my job, but Sue doesn’t ւ ց LF: I like my job but Sue doesn’t like my job like.job(i, i) ∧ ¬like.job(s, i) I1 [t1 like my1 job] but Sue1 [t1 doesn’t like my1 job] like.job(i, i)∧ ¬like.job(s, s)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Only

PF: Only I did my homework ւ ց

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Only

PF: Only I did my homework ւ ց [Only I]1 [t1 did my homework]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Only

PF: Only I did my homework ւ ց [Only I]1 [t1 did my homework]

• nly(i)(λx[do.hw(x, i)])

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Only

PF: Only I did my homework ւ ց [Only I]1 [t1 did my homework]

• nly(i)(λx[do.hw(x, i)])
• nly(x)(P)≡∀y[y = x → ¬P(y)]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Only

PF: Only I did my homework ւ ց [Only I]1 [t1 did my homework]

• nly(i)(λx[do.hw(x, i)])

[Only I]1 [t1 did my1 homework]

• nly(x)(P)≡∀y[y = x → ¬P(y)]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Pronouns in Generative Linguistics VP ellipsis Only

Only

PF: Only I did my homework ւ ց [Only I]1 [t1 did my homework]

• nly(i)(λx[do.hw(x, i)])

[Only I]1 [t1 did my1 homework]

• nly(i)(λx[do.hw(x, x)])
• nly(x)(P)≡∀y[y = x → ¬P(y)]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Outline

1 Introduction 2 Fake indexicals

Pronouns in Generative Linguistics VP ellipsis Only

3 . . . as indexicals

Ellipsis resolution by unification Only by unification Avoiding sloppy names

4 De se binding and de re acquaintance

Avoiding de se names

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Ellipsis by unification

SS: I like my job, but Sue doesn’t

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Ellipsis by unification

SS: I like my job, but Sue doesn’t like.job(i, i) ∧ ¬P(s)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Ellipsis by unification

SS: I like my job, but Sue doesn’t like.job(i, i) ∧ ¬P(s)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Ellipsis by unification

SS: I like my job, but Sue doesn’t like.job(i, i) ∧ ¬P(s) P(i) . = like.job(i, i)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Ellipsis by unification

SS: I like my job, but Sue doesn’t like.job(i, i) ∧ ¬P(s) P(i) . = like.job(i, i) ւ ց P→λx[like.job(x, i)] P→λx[like.job(x, x)]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Ellipsis by unification

SS: I like my job, but Sue doesn’t like.job(i, i) ∧ ¬P(s) P(i) . = like.job(i, i) ւ ց P→λx[like.job(x, i)] like.job(i, i) ∧ ¬like.job(s, i) P→λx[like.job(x, x)]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Ellipsis by unification

SS: I like my job, but Sue doesn’t like.job(i, i) ∧ ¬P(s) P(i) . = like.job(i, i) ւ ց P→λx[like.job(x, i)] like.job(i, i) ∧ ¬like.job(s, i) P→λx[like.job(x, x)] like.job(i, i) ∧ ¬like.job(s, s)

• cf. Dalrymple et al. 1991

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Only by unification

SS: Only [I]F did my homework

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Only by unification

SS: Only [I]F did my homework ∀x[x = i → ¬B(x)]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Only by unification

SS: Only [I]F did my homework ∀x[x = i → ¬B(x)] B(i) . = do.hw(i, i)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Only by unification

SS: Only [I]F did my homework ∀x[x = i → ¬B(x)] B(i) . = do.hw(i, i) ւ ց B→λx[do.hw(x, i)] B→λx[do.hw(x, x)]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Only by unification

SS: Only [I]F did my homework ∀x[x = i → ¬B(x)] B(i) . = do.hw(i, i) ւ ց B→λx[do.hw(x, i)] ∀x[x = i → ¬do.hw(x, i)] B→λx[do.hw(x, x)] ∀x[x = i → ¬do.hw(x, x)]

• cf. Pulman (1997)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Conclusions

• semantic/pragmatic alternative:
• minimized syntactic levels
• I is true indexical, interpreted in situ
• derive strict/sloppy by HOU

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Conclusions

• semantic/pragmatic alternative:
• minimized syntactic levels
• I is true indexical, interpreted in situ
• derive strict/sloppy by HOU
• Kaplan saved?

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

No sloppy names

John likes John’s job but Sue doesn’t

• strict: Sue doesn’t like John’s job

Only Mary likes Mary’s job

• strict: others don’t like Mary’s

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Predictions

• generative:
• names = pronouns
• Principle C prohibits bound names

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Predictions

• generative:
• names = pronouns
• Principle C prohibits bound names
• prediction: only reference, only strict

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Predictions

• generative:
• names = pronouns
• Principle C prohibits bound names
• prediction: only reference, only strict
• pragmatic:
• names ≈ indexicals: directly referential

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Predictions

• generative:
• names = pronouns
• Principle C prohibits bound names
• prediction: only reference, only strict
• pragmatic:
• names ≈ indexicals: directly referential
• prediction: strict + sloppy (by HOU)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Pragmatic blocking

• competing alternatives:

(1) Only Mary likes Mary’s job (2) Only Mary likes her job

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Pragmatic blocking

• competing alternatives:

(1) Only Mary likes Mary’s job (2) Only Mary likes her job

• (1) violates Principle C
• (1) more marked by referential hierarchy:
• definite descriptions > names > pronouns

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Pragmatic blocking

• competing alternatives:

(1) Only Mary likes Mary’s job (2) Only Mary likes her job

• (1) violates Principle C
• (1) more marked by referential hierarchy:
• definite descriptions > names > pronouns
• ulterior pragmatic motive for using (1)?

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Pragmatic blocking

• competing alternatives:

(1) Only Mary likes Mary’s job (2) Only Mary likes her job

• (1) violates Principle C
• (1) more marked by referential hierarchy:
• definite descriptions > names > pronouns
• ulterior pragmatic motive for using (1)?
• topicalizes/presupposes/makes salient Mary’s job
• prioritize background containing Mary’s job

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Pragmatic blocking

• competing alternatives:

(1) Only Mary likes Mary’s job (2) Only Mary likes her job

• (1) violates Principle C
• (1) more marked by referential hierarchy:
• definite descriptions > names > pronouns
• ulterior pragmatic motive for using (1)?
• topicalizes/presupposes/makes salient Mary’s job
• prioritize background containing Mary’s job
• B→{λx[like.job(x, m)], λx[like.job(x, x)]}

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Conclusions

• generative
• syntax/semantics: PF, LF, SS, L, m
• pronouns vs names
• binding/reference ambiguity: he, she, they, I, you,. . .
• reference: John, Sue,. . .
• in ellipsis, focus, only:
• reference → strict
• binding → sloppy

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Ellipsis resolution by unification Only by unification Avoiding sloppy names

Conclusions

• generative
• syntax/semantics: PF, LF, SS, L, m
• pronouns vs names
• binding/reference ambiguity: he, she, they, I, you,. . .
• reference: John, Sue,. . .
• in ellipsis, focus, only:
• reference → strict
• binding → sloppy
• pragmatic
• semantics/pragmatics: SS, L, m
• anaphoric vs directly referential
• anaphoric: he, she, they,. . .
• referential: I, you, John, today, . . .
• HOU pragmatically derives strict/sloppy
• sloppy names pragmatically blocked by anaphoric alternative

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Avoiding de se names

Outline

1 Introduction 2 Fake indexicals

Pronouns in Generative Linguistics VP ellipsis Only

3 . . . as indexicals

Ellipsis resolution by unification Only by unification Avoiding sloppy names

4 De se binding and de re acquaintance

Avoiding de se names

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Avoiding de se names

De se and de re

Kaplan is telling the story of the time he didn’t realize his pants were

• n fire while seeing himself on fire on live TV

I thought I was at a safe distance from the fire

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Avoiding de se names

De se and de re

Kaplan is telling the story of the time he didn’t realize his pants were

• n fire while seeing himself on fire on live TV

I thought I was at a safe distance from the fire “I’m at safe distance from fire”

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Avoiding de se names

De se and de re

Kaplan is telling the story of the time he didn’t realize his pants were

• n fire while seeing himself on fire on live TV

I thought I was at a safe distance from the fire “I’m at safe distance from fire” beli[safe(i)]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Avoiding de se names

De se and de re

Kaplan is telling the story of the time he didn’t realize his pants were

• n fire while seeing himself on fire on live TV

I thought I was at a safe distance from the fire “I’m at safe distance from fire” beli[safe(i)]

?I thought that I was remarkably

calm “that guy is remarkably calm”

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Avoiding de se names

De se and de re

Kaplan is telling the story of the time he didn’t realize his pants were

• n fire while seeing himself on fire on live TV

I thought I was at a safe distance from the fire “I’m at safe distance from fire” beli[safe(i)]

?I thought that I was remarkably

calm “that guy is remarkably calm” beli[remarkably.calm(i)]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Avoiding de se names

De se and de re

Kaplan is telling the story of the time he didn’t realize his pants were

• n fire while seeing himself on fire on live TV

I thought I was at a safe distance from the fire “I’m at safe distance from fire” bel∗

iλx[safe(x)]

?I thought that I was remarkably

calm “that guy is remarkably calm” beli[remarkably.calm(i)]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Avoiding de se names

De se and de re

Kaplan is telling the story of the time he didn’t realize his pants were

• n fire while seeing himself on fire on live TV

I thought I was at a safe distance from the fire “I’m at safe distance from fire” bel∗

iλx[safe(x)]

?I thought that I was remarkably

calm “that guy is remarkably calm” bele×et

i

i, λx[r.calm(x)]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Avoiding de se names

De se and de re

Kaplan is telling the story of the time he didn’t realize his pants were

• n fire while seeing himself on fire on live TV

I thought I was at a safe distance from the fire “I’m at safe distance from fire” bel∗

iλx[safe(x)]

?I thought that I was remarkably

calm “that guy is remarkably calm” R(i, i) ∧ bel∗

iλx[r.calm(

ι y[R(x, y)])]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Avoiding de se names

De se and de re

Kaplan is telling the story of the time he didn’t realize his pants were

• n fire while seeing himself on fire on live TV

I thought I was at a safe distance from the fire “I’m at safe distance from fire” bel∗

iλx[safe(x)]

?I thought that I was remarkably

calm “that guy is remarkably calm” R(i, i) ∧ bel∗

iλx[r.calm(

ι y[R(x, y)])] R = λxλy[see.on.tv(x, y)]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Avoiding de se names

De se and de re

Kaplan is telling the story of the time he didn’t realize his pants were

• n fire while seeing himself on fire on live TV

I thought I was at a safe distance from the fire “I’m at safe distance from fire” R(i, i) ∧ bel∗

iλx[safe(

ι y[R(x, y)])] R = λxλy[x = y]

?I thought that I was remarkably

calm “that guy is remarkably calm” R(i, i) ∧ bel∗

iλx[r.calm(

ι y[R(x, y)])] R = λxλy[see.on.tv(x, y)]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Avoiding de se names

De se and de re

Kaplan is telling the story of the time he didn’t realize his pants were

• n fire while seeing himself on fire on live TV

I thought I was at a safe distance from the fire “I’m at safe distance from fire” R(i, i) ∧ bel∗

iλx[safe(

ι y[R(x, y)])] R = λxλy[x = y]

?I thought that I was remarkably

calm “that guy is remarkably calm” R(i, i) ∧ bel∗

iλx[r.calm(

ι y[R(x, y)])] R = λxλy[see.on.tv(x, y)]

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Avoiding de se names

De se names

1 # Kaplan thought Kaplan was at a safe distance from the fire

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Avoiding de se names

De se names

1 # Kaplan thought Kaplan was at a safe distance from the fire

• Chierchia’89: Principle C blocks binding but¡2-¿ coreference ⇒

de re (non-de se)

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Avoiding de se names

De se names

1 # Kaplan thought Kaplan was at a safe distance from the fire

• Chierchia’89: Principle C blocks binding but¡2-¿ coreference ⇒

de re (non-de se)

• cheaper alternative:

2 Kaplan thought he was at a safe distance from the fire

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Avoiding de se names

De se names

1 # Kaplan thought Kaplan was at a safe distance from the fire

• Chierchia’89: Principle C blocks binding but¡2-¿ coreference ⇒

de re (non-de se)

• cheaper alternative:

2 Kaplan thought he was at a safe distance from the fire

• ulterior pragmatic motive for using (1)?

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Avoiding de se names

De se names

1 # Kaplan thought Kaplan was at a safe distance from the fire

• Chierchia’89: Principle C blocks binding but¡2-¿ coreference ⇒

de re (non-de se)

• cheaper alternative:

2 Kaplan thought he was at a safe distance from the fire

• ulterior pragmatic motive for using (1)?
• Kaplan ∈ reported thought

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals

Introduction Fake indexicals . . . as indexicals De se binding and de re acquaintance Avoiding de se names

De se names

1 # Kaplan thought Kaplan was at a safe distance from the fire

• Chierchia’89: Principle C blocks binding but¡2-¿ coreference ⇒

de re (non-de se)

• cheaper alternative:

2 Kaplan thought he was at a safe distance from the fire

• ulterior pragmatic motive for using (1)?
• Kaplan ∈ reported thought
• generalization: use marked coref res X only if X matches the

reported thought character # Kaplan thought Kaplan was remarkably calm Kaplan thought the guy on TV was remarkably calm

Emar Maier What syntax doesn’t feed semantics Fake indexicals as indexicals