Representing Focus Scoping over New Mats Rooth Cornell University - - PowerPoint PPT Presentation

▶

Nov 21, 2022 211 likes •538 views

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary Representing Focus Scoping over New Mats Rooth Cornell University NELS 45 MIT Oct. 31 Nov. 2, 2014 Locus/scope/antecedent Focus over new New grammar for

SLIDE 1

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Representing Focus Scoping over New

Mats Rooth Cornell University NELS 45 MIT

Oct. 31 – Nov. 2, 2014

SLIDE 2

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Alternative semantics

1 2 4 who 5 10 ate 11 the last piece of cake 3∼2 6F Justin 7 14 ate 15 it

3 ∼ 2 licensed by: Justin ate the last piece of cake being a proposition of the form y ate the last piece of cake.

SLIDE 3

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

∼/F marks the locus, scope and antecedent for focus

1 2 4 who 5 10 ate 11 the last piece of cake 3∼2 6F Justin 7 14 ate 15 it

SLIDE 4

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Entailment semantics/Givenness semantics

1 2 4 who 5 10 ate 11 the last piece of cake 3∼2 6F Justin 7 14 ate 15 it

3 ∼ 2 licensed by: somebody ate the last piece of cake entails some entity ate the last piece of cake.

SLIDE 5

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

∼ not scoping over F

1 2 4 who 5 10 ate 11 the last piece of cake 3∼2 6F Justin 7∼5 14∼10 ate 15∼11 it

7 ∼ 5 licensed by: ate the last piece of cake = ate the last piece of cake (alternative semantics) or somebody ate the last piece of cake entails some entity ate the last piece of cake (entailment semantics).

SLIDE 6

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

1 2 4 who 5 10 ate 11 the last piece of cake 3∼2 6F Justin 7∼5 14∼10 ate 15∼11 it

the last piece of cake = the last piece of cake.

SLIDE 7

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Phrases with ∼ not scoping over F ...

1 2 4 who 5 10 ate 11 the last piece of cake 3∼2 6F Justin 7∼5 14∼10 ate 15∼11 it

... often surface as reduced/destressed. This is part of the focus system.

SLIDE 8

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Alternative Focus vs. New according to Katz and Selkirk 2011

... for instance, they mostly store minesF in Idaho. speech/katz/07aMPc12.wav ... for instance, they mostly store mines in IdahoF. speech/katz/07bMPc21.wav Katz and Selkirk (2011): F gets realized differently than new, with greater pitch range and duration. Data from Jonah Katz and Lisa Selkirk

SLIDE 9

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Since all nodes are new ...

1 2 4 store 6F mines 3F in Idaho

... projecting F is not distinguished.

SLIDE 10

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

1F 2F 4F store 6F mines 3F in Idaho

Really every node is F-marked.

SLIDE 11

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Prosody of farmer sentences

[CanadianF farmers]1∼2 admire [AmericanF farmers ]2∼1. Roberts 1996/2012 The two tokens of farmer sound different, with the first

ne prosodically subordinate to Canadian, but not reduced.

So the first F is not alternative-focus at all.

SLIDE 12

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Scott Hollis That was Phil Woods on the altoF saxophone. Hope to see you again nextF Friday.

SLIDE 13

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Focus and reduction in comparatives [IF dance the tango] better than [youF dance the tango] As in farmer sentences: [IF dance the tango]2∼3 better than [youF dance the tango]3∼2

SLIDE 14

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

With preceding antecedent ... I’m an expert at [dancing tango]1. Hah! I bet [IF [VP dance the tango ]∼1] better than you!

speech/tango/tango1.wav

The gray part is reduced because it is redundant relative to the first VP, as expressed by ∼1.

speech/tango/tango1.wav

SLIDE 15

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Out of the blue ... I should do the next part. Why? [IF [VPdance the tango]∼1] better than you!

speech/tango/tango1.wav

Without a preceding antencedent, destressing of VP dance the tango seems not to be licensed, even though focus on the main-clause subject is licensed.

default: speech/tango/tango1.wav destressed: speech/tango/tango1.wav

SLIDE 16

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Focus/new vs. Focus/given in Japanese

M. Sugahara (2003)

SLIDE 17

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

While it is Tokyo and Osaka that are known for crime, this week the manager of a [YokohamaF [importing agency]] was arrested for selling marijuana to a model.

SLIDE 18

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Sugahara: for AA case, difference in pitch scaling in absence of a difference in phonological phrasing.

SLIDE 19

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Recursive alternative semantics

Let h be the ordinary semantic operation for a node of the form [αβ], e.g. leftward function application or rightward function application. [ [[αβ]] ]f =

h(a, b)|aǫ[

[α] ]f ∧ bǫ[ [β] ]f

Hamblin 1973 Rooth 1985

Alternatives project automatically, independent of syntactic features.

SLIDE 20

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

The problem again

1 2F next 3F Friday 1 2F alto 3F saxophone

The focus syntax doesn’t express that the F on next projects while the F on Friday doesn’t.

SLIDE 21

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Local information-structural operators

01 Jacob-ga 10 dare-o aisiteiru Project alternatives only along a “spine” of 1’s. These 1’s also have a phonological interpretatation of relative prominence–a local Stress-F.

M. Rooth and H. Dong (application to in situ WH and second
ccurrence focus).

SLIDE 22

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Semantic interpretation of information-structural

perators

Let h be the ordinary semantic operation for a node of the form [πσαβ], e.g. leftward function application or rightward function application. [ [[10αβ]] ]f =

h(a, [

[β] ]o)|aǫ[ [α] ]f [ [[01αβ]] ]f =

[α] ]o, b)|bǫ[ [β] ]f [ [[11αβ]] ]f =

h(a, b)|aǫ[

[α] ]f ∧ bǫ[ [β] ]f [ [[00αβ]] ]f = {h([ [α] ]o, [ [β] ]o)} [ [[πσαβ]] ]o = h([ [α] ]o, [ [β] ]o) Difference from Hamblin: when branch is marked 0, ordinary semantic value is plugged in, instead of selecting an element from the alternative set.

SLIDE 23

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Projecting F in all new

F mostly(C) ~C F Idaho F store F mines F mostly(C) ~C F Idaho F store F mines

SLIDE 24

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Each F projection path has an antecedent at the top, ∼ k

mostly(C) ~C Idaho store mines ~k yuujin-ga maneeja-no Yokohama-no yunyuu-daikooya-de

SLIDE 25

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

∼ x also without F-projection tail

~k ~j yuujin-ga ~i maneeja-no Yokohama-no ~h yunyuu-daikooya-de

Focus over given

SLIDE 26

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Remaining vertices are non-projecting F

~k nyuujin-ga maneeja-no Yokohama-no nyunyuu-daikooya-de

Focus over new

SLIDE 27

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Formally no F

~k nyuujin-ga maneeja-no Yokohama-no nyunyuu-daikooya-de

... at the base of F-projection paths, or on novelty F.

SLIDE 28

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Local stress F

Given a vertex with children a,b where a is marked for F-projection and b is not, some grid column in the realization of a exceeds every grid column in the realization of b

SLIDE 29

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Incremental improvement over standard stress F

Given a vertex with children a,b where a is marked for F-projection and b is not, some grid column in the realization of a exceeds every grid column in the realization of b. Let β be an F-marked phrase with scope φ. Then the strongest stress in the phonological realization of φ falls within the realization of β.

... because there’s no global reference to F-marked phrase, the corresponding F scope, and their phonological realizations.

SLIDE 30

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Both of these obey local stress F

~k next ~h Friday ~k next Friday

x x x next x x Fri x day x x x next x x Fri x day focus over given focus over new

SLIDE 31

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Phrase F X is bears ∼, or the realization of X dominates an accentual phrase. Accentual phrase Lower-ranked Accentual phrase provides pressure towards de-phrasing for ∼ k.

SLIDE 32

Locus/scope/antecedent Focus over new New grammar for focus Phonology Summary

Summary

1. ∼ k with an optional projection tail replaces the ∼ k/F

grammar for focus.

2. Alternatives project only along projection paths, they don’t

project by default.

3. Licensing of ∼ k using alternative semantics or entailment

semantics.

4. Anaphoric de-stressing is in the same system as projecting

focus.

5. Phonology uses local Stress F, plus Phrase F and

Representing Focus Scoping over New

Mats Rooth Cornell University NELS 45 MIT

Alternative semantics

1 2 4 who 5 10 ate 11 the last piece of cake 3∼2 6F Justin 7 14 ate 15 it

3 ∼ 2 licensed by: Justin ate the last piece of cake being a proposition of the form y ate the last piece of cake.

∼/F marks the locus, scope and antecedent for focus

1 2 4 who 5 10 ate 11 the last piece of cake 3∼2 6F Justin 7 14 ate 15 it

Entailment semantics/Givenness semantics

1 2 4 who 5 10 ate 11 the last piece of cake 3∼2 6F Justin 7 14 ate 15 it

3 ∼ 2 licensed by: somebody ate the last piece of cake entails some entity ate the last piece of cake.

∼ not scoping over F

7 ∼ 5 licensed by: ate the last piece of cake = ate the last piece of cake (alternative semantics) or somebody ate the last piece of cake entails some entity ate the last piece of cake (entailment semantics).

the last piece of cake = the last piece of cake.

Phrases with ∼ not scoping over F ...

... often surface as reduced/destressed. This is part of the focus system.

Alternative Focus vs. New according to Katz and Selkirk 2011

... for instance, they mostly store minesF in Idaho. speech/katz/07aMPc12.wav ... for instance, they mostly store mines in IdahoF. speech/katz/07bMPc21.wav Katz and Selkirk (2011): F gets realized differently than new, with greater pitch range and duration. Data from Jonah Katz and Lisa Selkirk

Since all nodes are new ...

1 2 4 store 6F mines 3F in Idaho

... projecting F is not distinguished.

1F 2F 4F store 6F mines 3F in Idaho

Really every node is F-marked.

Prosody of farmer sentences

[CanadianF farmers]1∼2 admire [AmericanF farmers ]2∼1. Roberts 1996/2012 The two tokens of farmer sound different, with the first

So the first F is not alternative-focus at all.

Scott Hollis That was Phil Woods on the altoF saxophone. Hope to see you again nextF Friday.

Focus and reduction in comparatives [IF dance the tango] better than [youF dance the tango] As in farmer sentences: [IF dance the tango]2∼3 better than [youF dance the tango]3∼2

With preceding antecedent ... I’m an expert at [dancing tango]1. Hah! I bet [IF [VP dance the tango ]∼1] better than you!

speech/tango/tango1.wav

The gray part is reduced because it is redundant relative to the first VP, as expressed by ∼1.

speech/tango/tango1.wav

Out of the blue ... I should do the next part. Why? [IF [VPdance the tango]∼1] better than you!

speech/tango/tango1.wav

Without a preceding antencedent, destressing of VP dance the tango seems not to be licensed, even though focus on the main-clause subject is licensed.

default: speech/tango/tango1.wav destressed: speech/tango/tango1.wav

Focus/new vs. Focus/given in Japanese

While it is Tokyo and Osaka that are known for crime, this week the manager of a [YokohamaF [importing agency]] was arrested for selling marijuana to a model.

Sugahara: for AA case, difference in pitch scaling in absence of a difference in phonological phrasing.

Recursive alternative semantics

Let h be the ordinary semantic operation for a node of the form [αβ], e.g. leftward function application or rightward function application. [ [[αβ]] ]f =

[α] ]f ∧ bǫ[ [β] ]f

Hamblin 1973 Rooth 1985

Alternatives project automatically, independent of syntactic features.

The problem again

1 2F next 3F Friday 1 2F alto 3F saxophone

The focus syntax doesn’t express that the F on next projects while the F on Friday doesn’t.

Local information-structural operators

01 Jacob-ga 10 dare-o aisiteiru Project alternatives only along a “spine” of 1’s. These 1’s also have a phonological interpretatation of relative prominence–a local Stress-F.

Semantic interpretation of information-structural

Let h be the ordinary semantic operation for a node of the form [πσαβ], e.g. leftward function application or rightward function application. [ [[10αβ]] ]f =

[β] ]o)|aǫ[ [α] ]f [ [[01αβ]] ]f =

[α] ]o, b)|bǫ[ [β] ]f [ [[11αβ]] ]f =

[α] ]f ∧ bǫ[ [β] ]f [ [[00αβ]] ]f = {h([ [α] ]o, [ [β] ]o)} [ [[πσαβ]] ]o = h([ [α] ]o, [ [β] ]o) Difference from Hamblin: when branch is marked 0, ordinary semantic value is plugged in, instead of selecting an element from the alternative set.

Projecting F in all new

Each F projection path has an antecedent at the top, ∼ k

∼ x also without F-projection tail

Focus over given

Remaining vertices are non-projecting F

Focus over new

Formally no F

... at the base of F-projection paths, or on novelty F.

Local stress F

Given a vertex with children a,b where a is marked for F-projection and b is not, some grid column in the realization of a exceeds every grid column in the realization of b

Incremental improvement over standard stress F

... because there’s no global reference to F-marked phrase, the corresponding F scope, and their phonological realizations.

Both of these obey local stress F

~k next ~h Friday ~k next Friday

x x x next x x Fri x day x x x next x x Fri x day focus over given focus over new

Phrase F X is bears ∼, or the realization of X dominates an accentual phrase. *Accentual phrase Lower-ranked *Accentual phrase provides pressure towards de-phrasing for ∼ k.

Summary

grammar for focus.

project by default.

semantics.

focus.

*Accentual phrase.

Phrase F X is bears ∼, or the realization of X dominates an accentual phrase. Accentual phrase Lower-ranked Accentual phrase provides pressure towards de-phrasing for ∼ k.