n the lurality of erbsastechow/lehre/ss06/events/kratzer05.k… · h ow do you pluralize a predicat...
TRANSCRIPT
![Page 1: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/1.jpg)
On the Plurality of VerbsAngelika Kratzer
Ventsislav ZhechevEvent Semantics, SS 2006
Sigrid Beck, Arnim von Stechow
![Page 2: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/2.jpg)
Agenda• Pluralization in Natural
Language
• Nominal number.Eliminating [singular] as a number feature
• On the plurality of verbs.Lexical Cumulativity
• How many readings?Phrasal Cumulativity
• Evidence forLexical Cumulativity
• The source ofPhrasal Cumulativity
• Conclusion
2
![Page 3: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/3.jpg)
Pluralization in Natural Language
![Page 4: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/4.jpg)
Are there plural verbs and how did they become that way?
• Fa%
• denotes a relation between individuals and events
• Verb meanings start out singular
• singular individuals are being linked to singular events
4
![Page 5: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/5.jpg)
• Kri&a (1992) and Landman (1996) suggest that verbs are born as plurals
• Then
• fa% could also link plural individuals to plural events
On the other hand…
5
![Page 6: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/6.jpg)
What about the VPs?
• VPs and bigger verbal proje)ions
• can be plural too
• but their plurality cannot always be inherited from their verbs
• so there must be another source of pluralization
6
![Page 7: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/7.jpg)
What about the VPs?
• Sternefeld (1998), Sauerland (1998), Beck (2000), and Beck and Sauerland (2000) have proposed that there is an optional and freely available operator in the syntax thatpluralizes predicates
7
![Page 8: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/8.jpg)
The claims of this paper
• There is a distin)ive theoretical place for lexical pluralization
• The pluralization of phrasal verbal proje)ions is constrained:
• it can only occur in the immediate neighborhood of a DP with plural agreement morphology
8
![Page 9: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/9.jpg)
How do you pluralize a predicate?
• The domain of entities De should contain both singular and plural individuals
• Following Link (1983), we construe plural individuals as sums and
• assume that De is cumulative:
• whenever x and y are in De, so is x+y
9
![Page 10: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/10.jpg)
How do you pluralize a predicate?
• In addition we need a domain of events Ds
• The sum operation is also defined for events and
• Ds can also be assumed to be cumulative
10
![Page 11: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/11.jpg)
How do you pluralize a predicate?
• Following Kri&a (1986) we extend the sum operation to ordered pairs built from members of De and Ds:
• <Mary, fall1>+<John, fall2> = <Mary+John, fall1+fall2>
• Now we can define pluralization as
• an operation ∗ that maps sets that come with a sum operation to their smallest cumulative superset
11
![Page 12: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/12.jpg)
Pluralizing fa%
• ⟦fall⟧ = {<Mary, fall1>, <John, fall2>}
• ⟦*fall⟧ = {<Mary, fall1>, <John, fall2>, <Mary+John, fall1+fall2>}
12
![Page 13: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/13.jpg)
Some questions
• Where do pluralization operators show up?
• Why do they show up where they do?
• How are they related to plural morphology on nouns and verbs?
• What is their semantic effe)?
• If there is pluralization, shouldn’t there be singularization too?
13
![Page 14: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/14.jpg)
Nominal number.Eliminating [singular] as a
number feature
![Page 15: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/15.jpg)
If there is pluralization, shouldn’t there be singularization too?
• There is no such thing as singular number
• Therefore, we do not expe) operators that singularize
15
![Page 16: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/16.jpg)
Cumulativity
• Manfred Kri&a’s universal:
• Simple predicat& in natural language typically are cumulative
• Apparent counterexamples:
• singular count nouns like child, chair, chin
16
![Page 17: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/17.jpg)
• Following Link, the extensions of singular count nouns are taken to be sets of singularities
• Hence, they could not be cumulative:
• If Josephine is a child, and Beatrice is too, the sum of Beatrice and Josephine is not a child
Cumulativity
17
![Page 18: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/18.jpg)
• However, child, chair, or chin are not necessarily simple predicates:
• they may already be complex by the time we get to see or hear them
• They each might consist of
• a root (√child, √chair, √chin) and
• a piece of nominal infle)ion
18
![Page 19: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/19.jpg)
• Müller (2000),Rullmann and You (2003)
• common noun roots have predicative, number-neutral (tra*numeral) denotations
• the root √child would denote the set consisting of all singular children and their sums
• Kri&a (1995) andYang (2001)building on Carlson (1977)
• noun roots are referential and refer to kinds
• it would now be part of the job of nominal infle)ion to turn those roots into predicates
Two views on the nature of roots
19
![Page 20: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/20.jpg)
Roots as kinds
• This proposal does not assume any particular mode of individuation or portioning for the denotations of noun roots
• Cross-linguistically, the fun)ion of individuating and portioning is often carried by classifiers
20
![Page 21: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/21.jpg)
Classifiers
• Following Kri&a (1995):
• English has a multiply ambiguousnon-overt classifier and
• the noun forms that are usually categorizedas ‘singular’ are in reality roots withan incorporated classifier
21
![Page 22: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/22.jpg)
Classifiers
a.⟦√zebra⟧ = zebra
b.⟦CLind⟧ = λx.λy.[kind(x) ∧ individual(y) ∧ y ≤ x]
⟦CLkind⟧ = λx.λy.[kind(x) ∧ kind(y) ∧ y ≤ x]
22
![Page 23: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/23.jpg)
• The word zebra is a ‘singular’ predicate by the time we see or hear it
• It was turned into a predicate by an incorporated ambiguous classifier, and is therefore ambiguous, too
23
a. ⟦√zebra⟧ = zebra
b. ⟦CLind⟧ = λx.λy.[kind(x) ∧ individual(y) ∧ y ≤ x]
c. ⟦CLkind⟧ = λx.λy.[kind(x) ∧ kind(y) ∧ y ≤ x]
![Page 24: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/24.jpg)
(3)
a. This zebra h, not been fed.
b.This zebra is almost extinct.
(4)
a. Those zebr, have not been fed.
b.Those two zebr, are almost extinct.
24
![Page 25: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/25.jpg)
(5)
a. This wine is for table 8.
b.You dropped two red win&.
(6)
a. This Pinot Noir is rare.
b.We t,ted five different Pinot Noirs.
25
![Page 26: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/26.jpg)
plural
classifier √zebra
• Either
• English has no such feature altogether, or
• it has an ambiguous classifier [singular]
[singular]
26
![Page 27: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/27.jpg)
Mass nouns
• In English they are predicative by the time we see them
• hence, should come with an obligatory classifier
• Following Chierchia (1998), that classifier should map a kind into the set of all of its singular or plural realizations
• thus, predicative mass nouns are already pluralized by their classifier and
• are not submitted to further pluralization
27
![Page 28: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/28.jpg)
• Unless mass nouns combine with an individual or a kind classifier, they cannot proje) [plural]
• Both mass nouns and non-plural count nouns proje) only a classifier
• therefore they both lack [plural] and
• therefore they both trigger singular agreement
28
![Page 29: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/29.jpg)
• Kratzer suggests that there is no feature [singular] in English
• What we call singular might bethe absence of plural
• in the morphology and
• in the semantics
• Non-overt classifiers are responsible for the presence of singular predicates
29
To sum up
![Page 30: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/30.jpg)
To sum up
• Agreement phenomena show that there is a tight conne)ion between nominal and verbal number
• if there is no nominal [singular]
• then there is no verbal [singular] either
• The conne)ion between nominal and verbal number should then be established via [plural] alone:
• when we see singular agreement, what we see is infle)ion that is there because of the absence of [plural]
30
![Page 31: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/31.jpg)
On the plurality of verbs.Lexical Cumulativity
![Page 32: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/32.jpg)
Kri&a’s cumulativity universal
• Simple predicat& in natural language typically are cumulative
• We posit referential denotations for noun roots
• Thus they are never predicative
• hence, they satisfy Kri&a’s universal trivially
• English nouns may become predicativein the course of a synta)ic derivation
• but when they do, they are no longer simple
32
![Page 33: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/33.jpg)
What about verb roots and verb stems?
• Verbs have the chara)eristic propertyof taking arguments
• Some of those arguments seem to be synta)ically acquired in the course of a derivation
• Marantz (1984) and Kratzer (1994) have argued that external arguments are always added in the syntax
• Pylkkänen (2001, 2002) makes the same pointfor applicative arguments
33
![Page 34: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/34.jpg)
What about verb roots and verb stems?
• Some dire) internal arguments seem to be introduced synta)ically, too, via secondary predicates or serialization, for example
• But there are also transitive and unaccusative verbs with inherently relational meanings:
• relate, connect, r&emble, surp,s, outdo,depend, hinder, ca1e
34
![Page 35: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/35.jpg)
What about verb roots and verb stems?
• With many transitive and unaccusative verbs, the kind of event described varies with the kind of dire) internal argument in sometimes erratic ways
• pick, pop, rise
• It would be hard to account for this dependency under the assumption thata verb’s argument stru)ure is always synta)ically constru)ed
35
![Page 36: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/36.jpg)
• Then there seems to be a large group of inherently relational verb roots
• and this suggests that as a class, verb roots might be predicative from the start
• then we would expe) them to fall under Kri&a’s generalization, and have cumulative denotations
• And if external and applicative arguments are not true arguments of their verbs,
• we need thematic role predicates like agent or goal to introduce them
• those predicates’ denotations should then be cumulative, too
36
![Page 37: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/37.jpg)
• If the denotations of verbs and thematic role predicates are cumulative from the start, we expe) effortless availability of a cumulative interpretation for sentences like
• Twenty children ate ten pizz,.
• Formal representation:
• ∃e∃x∃y [∗child(x) ∧ |x| = 20 ∧ ∗agent(x)(e) ∧ ∗pizza(y) ∧ |y| = 10 ∧ ∗eat(y)(e)]
• As in Landman (1996, 2000), the basic predicates of the metalanguage are singular predicates that are pluralized with the ∗-operator
37
![Page 38: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/38.jpg)
(8)Two children lifted two box&.• Suppose the two children are C,ey and Stacey, and
the two boxes are Red and Green. C,ey lifted Red on her own once, and Stacey did so twice. In addition, C,ey and Stacey jointly lifted Green.
• We have four events, e1, e2, e3, and e4:
38
Cumulativity and event semantics
Event Box liffted Box lifftere1 Red Caseye2 Red Staceye3 Red Staceye4 Green Casey+Stacey
![Page 39: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/39.jpg)
(9) Extension of lift
• {<e1, Red>, <e2, Red>, <e3, Red>, <e4, Green>, …}
(10) Extension of agent
• {<e1, C,ey>, <e2, Stacey>, <e3, Stacey>, <e4, C,ey + Stacey>, …}
39
Event Box liffted Box lifftere1 Red Caseye2 Red Staceye3 Red Staceye4 Green Casey + Stacey
![Page 40: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/40.jpg)
(11) a. Extension of ∗lift{<e1, Red>, <e2, Red>, <e3, Red>, <e4, Green>, <e1+e2, Red>,<e1+e3, Red>, <e1+e4, Red+Green>, <e2+e3, Red>,<e2+e4, Red+Green>, <e3+e4, Red+Green>, <e1+e2+e3, Red>,<e1+e2+e4, Red+Green>, <e1+e3+e4, Red+Green>,<e2+e3+e4, Red+Green>, <e1+e2+e3+e4, Red+Green>, …}
b.Extension of ∗agent{<e1, C,ey>, <e2, Stacey>, <e3, Stacey>, <e4, C,ey+Stacey>,<e1+e2, C,ey+Stacey>, <e1+e3, C,ey+Stacey>,<e1+e4, C,ey+Stacey>, <e2+e3, Stacey>, <e2+e4, C,ey+Stacey>, <e3+e4, C,ey+Stacey>, <e1+e2+e3, C,ey+Stacey>,<e1+e2+e4, C,ey+Stacey>, <e1+e3+e4, C,ey+Stacey>,<e2+e3+e4, C,ey+Stacey>, <e1+e2+e3+e4, C,ey+Stacey>, …}
40
![Page 41: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/41.jpg)
(12)
a.∃e∃x∃y [∗child(x) ∧ |x| = 2 ∧ ∗agent(x)(e) ∧ ∗box(y) ∧ |y| = 2 ∧ ∗lift(y)(e)]
b. [∗child(x) ∧ |x| = 2 ∧ ∗agent(x)(e) ∧ ∗box(y) ∧ |y| = 2 ∧ ∗lift(y)(e)]
c. ‘e’ → e1+e2+e3+e4
‘x’ → C,ey+Stacey‘y’ → Red+Green
41
![Page 42: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/42.jpg)
(13)
a. C,ey and Stacey lifted Red.
b.C,ey and Stacey lifted Green.
c. C,ey lifted Red (at le,t) once.
d.Stacey lifted Red (at le,t) twice.
42
![Page 43: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/43.jpg)
(11) a. Extension of ∗lift{<e1, Red>, <e2, Red>, <e3, Red>, <e4, Green>, <e1+e2, Red>,<e1+e3, Red>, <e1+e4, Red+Green>, <e2+e3, Red>,<e2+e4, Red+Green>, <e3+e4, Red+Green>, <e1+e2+e3, Red>,<e1+e2+e4, Red+Green>, <e1+e3+e4, Red+Green>,<e2+e3+e4, Red+Green>, <e1+e2+e3+e4, Red+Green>, …}
b.Extension of ∗agent{<e1, C,ey>, <e2, Stacey>, <e3, Stacey>, <e4, C,ey+Stacey>,<e1+e2, C,ey+Stacey>, <e1+e3, C,ey+Stacey>,<e1+e4, C,ey+Stacey>, <e2+e3, Stacey>, <e2+e4, C,ey+Stacey>, <e3+e4, C,ey+Stacey>, <e1+e2+e3, C,ey+Stacey>,<e1+e2+e4, C,ey+Stacey>, <e1+e3+e4, C,ey+Stacey>,<e2+e3+e4, C,ey+Stacey>, <e1+e2+e3+e4, C,ey+Stacey>, …}
43
![Page 44: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/44.jpg)
(13)
a. C,ey and Stacey lifted Red.
b.C,ey and Stacey lifted Green.
c. C,ey lifted Red (at le,t) once.
d.Stacey lifted Red (at le,t) twice.
44
(14)
a. Red w, lifted fourteen tim&.
b.C,ey and Stacey together did eleven liftings.
![Page 45: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/45.jpg)
• A basic principle of counting says that ifI count you as an entity, I can’t countyour head separately
• What really seems to count in countingis atomicity• The extension of ∗lift contains exa)ly three
atomic pairs that conne) Red to a lifting event
• The extension of ∗agent contains exa)ly one atomic pair that conne)s C,ey and Staceyto a lifting event.
45
(14) a. Red w, lifted fourteen tim&.b.C,ey and Stacey together did eleven liftings.
![Page 46: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/46.jpg)
• Cumulation preserves all information we want to extra) from a verb’s extension
• we have all the information we might need to get the semantics of adverbs like twice or three tim&, or individually, or together right
46
To sum up
![Page 47: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/47.jpg)
On the Plurality of VerbsAngelika Kratzer
Ventsislav ZhechevEvent Semantics, SS 2006
Sigrid Beck, Arnim von Stechow
PartII
![Page 48: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/48.jpg)
Agenda• Pluralization in Natural
Language
• Nominal number.Eliminating [singular] as a number feature
• On the plurality of verbs.Lexical Cumulativity
• How many readings?Phrasal Cumulativity
• Evidence forLexical Cumulativity
• The source ofPhrasal Cumulativity
• Conclusion
48
![Page 49: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/49.jpg)
How many readings?Phrasal Cumulativity
![Page 50: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/50.jpg)
How many readings?
(15)Two children lifted two box&.
• there is a reading of (15) that lumps together what are traditionally called collective and cumulative interpretations, and doesn’t distinguish between one-time and repetitive liftings
• Is it right to lump together all the interpretations that others have takenpains to distinguish?
50
![Page 51: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/51.jpg)
Evidence from VP-ellipsis• The ambiguity in the overt and in the silent
VP must be resolved in the same way
(16) I went to the bank, and you did, too.
(17) The two boys lifted the two box&, and the two girls did, too.
• Is (17) true in a situation in which the two boys jointly lifted each of the two boxes,but the two girls each lifted a different oneof the two boxes on her own?
51
![Page 52: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/52.jpg)
Distributivity
• In addition to the cumulative interpretation, (15) has two distributive interpretations
• Landman (1989) argued that when a plural DP produces distributive interpretations ofthis kind, they should be derived bypluralizing its sister predicate
52
(15)Two children lifted two box&.
![Page 53: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/53.jpg)
Distributivity
(18) a. (2 children) ∗ [lifted 2 boxes]b.(2 boxes) ∗ λ1 [2 children lifted t1]
(19) a. ∗ λxλe∃y [∗box(y) ∧ |y| = 2 ∧ ∗lift(y)(e) ∧ ∗agent(x)(e)]b.∗ λxλe∃y[∗child(y) ∧ |y| = 2 ∧ ∗lift(x)(e) ∧ ∗agent(y)(e)]
• Since starring a predicate always extends the original extension, both 18(a) and (b) still cover all the scenarios we discussed before
53
(15)Two children lifted two box&.
![Page 54: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/54.jpg)
Lumping the readings together• If pluralization of verbal predicates is the
corre) way of accounting for distributive interpretations, we are committed to lumping together interpretations in a particular way
54
1. Cumulative/Collective/Repetitive
2. Cumulative/Collective/RepetitiveSubject Distributive 3. Cumulative/Collective/RepetitiveObject Distributive
![Page 55: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/55.jpg)
How many readings?
• If starring of a plural DPs sister node is obligatory, as I will be suggested below, we only expe)two truly distin) readings for sentence (15),one of which is highly dispreferred
• The difference boils down to whether or notwe move the obje) over the subje)
55
(15)Two children lifted two box&.
![Page 56: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/56.jpg)
How many readings?
• As far as grammar goes, no distin)ion is made between subject distributive, cumulative, collective, and iterative interpretations
• All those different ways of understanding (15) correspond to a single reading that can be computed in a straightforward way from a single synta)ic representation
56
(15)Two children lifted two box&.
![Page 57: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/57.jpg)
Individuating the readings
• In ellipsis constru)ions, for example, it should be possible to mix distributive and cumulativeor co%ective interpretations
(20)The two chefs cooked a stew, and the two studentsdid, too. The chefs were very experienced, so they each prepared a Moroccan tagine. The two students worked together on a Bœuf Bourguignon.
57
![Page 58: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/58.jpg)
Individuating the readings
• Roger Schwarzschild has observed that separating distributive and co%ective/cumulative interpretations can have undesirable consequences in the scopeof negation
• Be,ly, better make sure those guys don’t win a car this week!
58
![Page 59: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/59.jpg)
To sum up
• We have seen what looked like initial support for lexical cumulativity,
• but we have also seen that lexical cumulativity alone is not enough
• Phrasal cumulativity is needed to account for certain cases of distributive interpretations
• thus, we need ∗-operators that can pluralize phrases
59
![Page 60: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/60.jpg)
To sum up
• Do we still need Lexical Cumulativity,if we already have the ∗-operators?
• Can’t those ∗-operators alone do the jobs we thought Lexical Cumulativity was responsible for?
60
![Page 61: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/61.jpg)
Evidence for Lexical Cumulativity
![Page 62: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/62.jpg)
Attacking Lexical Cumulativity
• Several authors have argued that the denotations of verbs can be rendered cumulative through the freely available optional presence of synta)ically represented ∗-operators that can pluralize any kind of verbal predicate:
• lexical, phr,al, b,ic or syntactica%y derived
62
![Page 63: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/63.jpg)
The goal of this se)ion
• To show that
• the proposal of Sternefeld, Sauerland, and Beck over-generates
• there is still a theoretically distinguished placefor Lexical Cumulativity andKri&a’s Cumulativity Universal
63
![Page 64: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/64.jpg)
(22)What does this intern do?a. She guards a parking lot.b.He cooks for an elderly lady.c. She waters a garden.d.He watch& a baby.e. She clea* an office building.
(23) a. I dialed a wrong phone number for 5 minut&.b.She bounced a ba% for 20 minut&.c. He kicked a wa% for a couple of hours.d.She opened and closed a drawer for half an hour.e. I petted a rabbit for two hours.
64
![Page 65: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/65.jpg)
• If ∗-operators could be inserted freely, they could immediately produce 24(b) from 24(a),for example, hence derive unattested interpretations for the sentences in (22) and (23):
(24)
a. λe∃x [ba%(x) ∧ ∗bounce(x)(e)]
b.∗ λe∃x [ba%(x) ∧ ∗bounce(x)(e)]
65
Restri) ∗-operators?(22) She guards a parking lot.(23) She bounced a ba% for 20 minut&.
![Page 66: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/66.jpg)
• λxλe ∗bounce(x)(e)
• λR<e<st>>λe∃x [ba%(x) ∧ R(x)(e)]
• λe∃x [ba%(x) ∧ ∗bounce(x)(e)]
• being a possibly plural event e such thatthere is a ba% x and e is an event of bouncing x
66
[bounce a ball]VP
![Page 67: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/67.jpg)
• In (22), habitual aspe) seems to be responsible for the necessarily iterative interpretation
• In (23), durativity plays a similar role
• Given lexical cumulativity, we still predi) the fa)s in (22) and (23), even if the habitual operator and durational adverbs take scope over the indefinite dire) obje)
67
Explaining (22) and (23)(22) She guards a parking lot.(23) She bounced a ba% for 20 minut&.
![Page 68: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/68.jpg)
(26)
a. Ich hab’ fünf Minuten langeine falsche Telefonnummer gewählt.
b.Ich hab’ eine falsche Telefonnummerfünf Minuten lang gewählt.
68
![Page 69: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/69.jpg)
• Suppose the denotation of durational adverbials like for 5 minut& is as in (27):
(27) λP<st>λe [P(e) ∧ e = σe′ [P(e′) ∧ e′ < e] ∧ fminute(e) = 5]
• The definition in (27) uses Link’s σ-operator.In our case, the operator maps the events in the set {e′| P(e′) ∧ e′ < e} to their supremum – if it exist
69
Durational adverbials
![Page 70: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/70.jpg)
• Following Morzycki’s Program of Modified Modification (Morzycki 2004) and the independently developed analysis of durational adverbs in van Geenhoven (2004), we would eventually want to split up the denotation of durational adverbials like for 5 minut& into at least two parts:
(28)
a. λP<st>λe [P(e) ∧ e = σe′ [P(e′) ∧ e′ < e]]
b.λe fminute(e) = 5
70
![Page 71: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/71.jpg)
• The denotation of dial a number for 5 minut&,for example, can now be computed by applying the denotation of for 5 minut& tothe denotation of dial a number
• the VP dial a number is thus clearly inthe scope of for 5 minut&
• The result is the denotation in (29):
(29) λes∃x [number(x) ∧ ∗dial(x)(e) ∧ e = σe′ [∃x [number(x) ∧ ∗dial(x)(e′)] ∧ e′ < e] ∧ fminute(e) = 5]
71
![Page 72: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/72.jpg)
• Assuming Lexical Cumulativity,
• iterative interpretations for verbs are possible from the very start
• iterativity without concurrent object distributivityis the automatic result of introducing an ordinary singular indefinite in the early stages ofa synta)ic derivation
72
To sum up
![Page 73: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/73.jpg)
To sum up• Habitual operators and durational adverbs
do no longer have to pluralize the predicatesthey operate over, or introduce quantification over sub-events• they merely have to make sure that those predicates
do not describe any singular events, but are properly plural in a lexically defined sense
• Thus, we no longer have to stipulate obligatory narrow scope for such operators• the desired interpretations can be derived,
even if the relevant aspe)ual operators aresitting above dire) obje)s
73
![Page 74: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/74.jpg)
To sum up
• ∗−operators cannot be inserted freely
• If they could, we wouldn’t expe) the ‘failure of distribution’ effe) illustrated in (22) and (23)
• What is the force that producesphrasal cumulativity,hence, many cases of distributivity?
74
(22) She guards a parking lot.(23) She bounced a ba% for 20 minut&.
![Page 75: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/75.jpg)
The source of Phrasal Cumulativity
![Page 76: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/76.jpg)
What makes phrasal cumulativity possible?
• It cannot be an accident that none of the sentences in (22) or (23) contained any plural DPs
• Maybe phrasal ∗-operators are necessarily tied to the presence of plural DPs in some way or other
• Schwarzschild (1993-94), for example, proposed that all plural VPs are obligatorily translatedwith the ∗-operator, hence always have cumulative denotations
76
(22) She guards a parking lot. (23) She bounced a ba% for 20 minut&.
![Page 77: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/77.jpg)
Reformulating Schwarzschild
• At the level where semantic interpretation takes place, sister constituents of plural DPs are pluralized, regardless of whether they are still in their base position or have moved away
77
![Page 78: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/78.jpg)
Plural DPs
• Following Sauerland (2005), when a plural DP is built from a determiner anda plural noun, for example, both the noun and the determiner come with their own number proje)ion
78
plural
Det
pluralclassifier N
![Page 79: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/79.jpg)
What is the role of the higher [plural]?
• It can’t seem to be interpretable within its DP
• Suppose that nominal [plural] is always interpretable, and it always carriesthe cross-categorial plural operator
• then the higher [plural] is forced to move out before semantic interpretation takes place
79
![Page 80: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/80.jpg)
What is the role of the higher [plural]?
• Moving as little as possible, it could becomea verbal infle)ional head right below its DP
• In this way, a DP could literally create itsown agreement proje)ion, possibly on topof another verbal proje)ion
80
![Page 81: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/81.jpg)
What is the role of the higher [plural]?
• When [plural] migrates out of its DP, we geta ∗-operator that pluralizes the DP’s sister node, possibly showing up as overt verbal agreement
81
DPplural= * …
Pluralized sister predicate
![Page 82: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/82.jpg)
• An immediate predi)ion of this proposal is that pluralization of phrasal verbal proje)ions should require the presence of DPs with [plural] agreement features in English
• But distributive/cumulative interpretations that can be produced by Lexical Cumulativity alone, should also be available for singular DPs
82
![Page 83: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/83.jpg)
(30)
a. She sent her off>ring to 5 different boarding schools.
b.She sent her off>ring to a boarding school.
(31)
a. She sent her children to 5 different boarding schools.
b.She sent her children to a boarding school.
83
Testing the predi)ion
![Page 84: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/84.jpg)
Testing the predi)ion
• Why is (32) bad?
(32) *Her off>ring each went to a boarding school.
• A possible explanation is that floated eachneeds to agree with [plural]
84
![Page 85: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/85.jpg)
Mass nouns
(33)
a. A% that furniture w, loaded onto 5 trucks.
b.A% that furniture w, loaded onto a truck.
(34)
a. Her off>ring inherited a% her jewelry.
b.Her off>ring inherited a vi%a in Tuscany.
85
![Page 86: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/86.jpg)
Mass nouns
(35) a. The moss on the rocks and the moss on the tre& is blighted.
b. Jane’s china and Alice’s china was stored in separate closets.
(36) a. The sugar for the coffee and the sugar for the cake was stored in a pl,tic jar.
b. Jane’s silverware and Patsy’s silverware was sent to a co1in.
86
![Page 87: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/87.jpg)
• It should not be possible to simultaneously cumulate two non-event arguments• we do not have lexical predicates with more than
two non-event arguments to begin with, and
• plausible assumptions about movement do not seem to allow us to derive any such predicatesin the syntax
• Thus, a DP’s sister constituent can only be of type <et> or <e<st>>• hence, the pluralization operation could not affe)
any other non-event argument position apart from the one that is about to be saturated bythe DP triggering the pluralization
87
![Page 88: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/88.jpg)
• Beck and Sauerland (2000):
(37) Th&e 5 teachers gave a bad mark to those 20 prot&ting students.
• Beck and Sauerland argue that the intended interpretation of (37) can only be derived bypluralizing the following 2-place relation:
• λxλy∃z [bad-mark(z) ∧ gave-to(y)(z)(x)]
88
![Page 89: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/89.jpg)
• The key for (37) is neo-Davidsonian association of the agent argument, coupled with movement of the indire) obje) toa position right above the dire) obje)
• The moved DP’s sister predicate would nowbe pluralized and would wind up withthe denotation in (38):
(38) ∗λyλe∃z [bad-mark(z) ∧ ∗gave(z)(e) ∧ ∗goal(y)(e)]
• This pluralized predicate is of type <e<st>>, hence only has one non-event argument
89
(37) Th&e 5 teachers gave a bad mark to those 20 prot&ting students.
![Page 90: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/90.jpg)
(39) ∗λyλe∃z [bad-mark(z) ∧ ∗gave(z)(e) ∧ ∗goal(y)(e)] those 20 prot&ting students
(40)λe [∗agent(those 5 teachers)(e) ∧((∗λyλe∃z [bad-mark(z) ∧ ∗gave(z)(e) ∧ ∗goal(y)(e)]) (those 20 prot&ting students))(e)]
• The interpretation captured in (40) says that those five teachers were the agents of an eventin which those 20 protesting students received one or more bad marks
90
(37) Th&e 5 teachers gave a bad mark to those 20 prot&ting students.
![Page 91: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/91.jpg)
(41) The children areholding a wheel.
• In situations of this kind, (41) is false orat least “highly strange”
91
Winter (2000)
31
Next, we add the agent argument and saturate it:
(32) !e ["agent(those five teachers)(e) & (" !y!e #z [bad-mark(z) & "gave-
to(y)(z)(e)])(those 20 protesting students)(e)]
The interpretation captured in (32) says that those five teachers were the agents of an event
in which those 20 protesting students received one or more bad marks. This is the desired
result.
A surprising consequence of the current analysis of phrasal verbal cumulativity relates to an
example presented in Winter (2000). Winter’s example is meant to show that theories that
account for distributive interpretations of plural DPs by pluralizing their sister predicates
overgenerate. They seem to predict interpretations that do not in fact exist. Interestingly,
Winter’s objection applies to event-less versions of the ‘nominal distributivity via verbal
cumulation’ idea, but not to the event-based account proposed here. Let us see why.
Winter asks us to judge the truth of sentence (33) in the scenario depicted in figure 5:
(33) The children are holding a wheel.
Figure 5
Winter observes correctly that in situations of this kind, (33) is false or at least “highly
strange”. On Winter’s own account, all non-lexical cases of distributivity are due to covert
monadic distributivity operators that enforce atomic distribution. (33) would then be true
just in case each child is holding a wheel. This condition is not satisfied in Winter’s scenario.
![Page 92: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/92.jpg)
Winter (2000)
• On Winter’s own account, (41) would be true just in case each child is holding a wheel
• Winter predi)s (41) to be false in his scenario, then
92
(41) The children are holding a wheel.
![Page 93: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/93.jpg)
Winter (2000)• Boy1 and Boy2 are holding a wheel,
and so do Boy2 and Boy3
• The denotation of the unstarred VP in (41)is therefore true of the two pluralitiesBoy1+Boy2 and Boy2+Boy3
• If the plural subje) the children induces starring of the VP, the denotation of that VP is trueof Boy1+Boy2+Boy3, and hence of the children
• (41) is thus predi)ed to be trueon Winter’s scenario
93
(41) The children are holding a wheel.
![Page 94: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/94.jpg)
Winter (2000)in an event-based approach
• On Kratzer’s account, the sister constituentof the plural subje) in (41) expresses a relation between individuals and events, and it isthat relation that is cumulated:
(42)
a. λxλe [∗agent(x)(e) ∧ ∃y [wheel(y) ∧ ∗hold(y)(e)]]
b.∗ λxλe [∗agent(x)(e) ∧ ∃y [wheel(y) ∧ ∗hold(y)(e)]]
94
(41) The children are holding a wheel.
![Page 95: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/95.jpg)
Winter (2000)in an event-based approach
• Does the pair consisting of the three boys and the event e represented in Winter’s scenario satisfy the starred relation in 42(b)?
• It could only do so if there are pairs <x1, e1> and <x2, e2> that satisfy the relation in 42(a), where x1+x2 = the children and e1+e2 = e
95
32
Winter predicts (33) to be false in his scenario, then. Winter’s point is that analyses where
plural DPs trigger cumulation of their sister make false predictions in this case. Let’s forget
about events for a moment, and reconstruct Winter’s argument. Boy1 and Boy2 are holding a
wheel, and so do Boy2 and Boy3. The denotation of the unstarred VP in (33) is therefore
true of the two pluralities Boy1+Boy2 and Boy2+Boy3. If the plural subject the children induces
starring of the VP, the denotation of that VP is true of Boy1+Boy2+Boy3, and hence of the
children. (33) is thus predicted to be true on Winter’s scenario.
Does our event-based scenario fare any better here? On our account, the sister constituent
of the plural subject in (33) expresses a relation between individuals and events, and it is that
relation that is cumulated.
(34) a. !x!e ["agent(x)(e) & #y [wheel(y) & "hold(y)(e)] ]
b. " !x !e ["agent(x)(e) & #y [wheel(y) & "hold(y)(e)] ]
Does the pair consisting of the three boys and the event e represented in Winter’s scenario
satisfy the starred relation in 34(b)? It could only do so if there are pairs <x1, e1> and
<x2, e2> that satisfy the relation in 34(a), where x1+x2 = the children and e1+e2 = e.
However, the event represented in figure 5 is most naturally conceptualized as a single event.
There are no natural, but only ‘strange’ or artificial ways of conceptualizing it as the sum of
two subevents. The subevents singled out in Figure 6, for example, do not seem to be
among the atoms in our domain of events:
Figure 6
(42) (∗) λxλe [∗agent(x)(e) ∧ ∃y [wheel(y) ∧ ∗hold(y)(e)]]
![Page 96: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/96.jpg)
Winter (2000)in an event-based approach
34
Figure 7
Sentence (33) is clearly true in the scenario of Figure (7). Rather than presenting a challenge
to our account, Winter’s example provides a surprising piece of support.
I conclude, then, that plural DPs are themselves sources of phrasal cumulativity - or more
concretely, their higher [plural] features are (in the sense of Sauerland (2005)). Pluralizing
their DP’s sister node seems to be the only way for those features to be put to semantic use.
Within an event semantics, a DP’s sister node often denotes a relation between individuals
and events. Consequently, judgments about the truth of sentences like (15) are bound to be
sensitive to the individuation of events.
7. Conclusion and questions for further research
It seems, then, that there are indeed at least two pluralization mechanisms at work in
languages like English. One is Lexical Cumulativity, which seems to be universal. The other
one is carried by the inflectional feature [plural]. I have argued that [plural] is always
interpretable, and always denotes the cross-categorial !-operator. Moreover, I have
suggested that [plural] always originates within a DP and pluralizes nominal or verbal
projections, depending on whether it occupies a high or a low position within its DP. The
low position provides access to a noun, the high position provides access to a verbal
projection. A single inflectional feature can thus create subtle variations in the availability of
distributive/cumulative interpretations, even within a single language. In the long run, the
96
•We expect (41) to be judged true in scenarios where the relevant subsituations are individuated more clearly
• and in the scenario above sentence (41) is clearly true
(41) The children are holding a wheel.
![Page 97: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/97.jpg)
To sum up• Plural DPs are themselves sources of
phrasal cumulativity• or more concretely, their higher [plural]
features are (in the sense of Sauerland (2005))
• pluralizing their DP’s sister node seems to be the only way for those features to be put to semantic use
• Within an event semantics, a DP’ssister node often denotes a relationbetween individuals and events• consequently, judgments about the truth
of sentences like (15) are bound to be sensitiveto the individuation of events
97
(15)Two children lifted two box&.
![Page 98: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/98.jpg)
Conclusion
![Page 99: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/99.jpg)
• It seems, that there are indeedat least two pluralization mechanismsat work in languages like English
• one is Lexical Cumulativity,which seems to be universal
• the other one is carried bythe infle)ional feature [plural]
99
![Page 100: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/100.jpg)
• is always interpretable, and
• always denotes the cross-categorial ∗-operator
• always originates within a DP and pluralizesnominal or verbal proje)ions, dependingon whether it occupies a high or a lowposition within its DP
• the low position provides access to a noun
• the high position provides access to a verbal proje)ion
100
[plural]
![Page 101: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/101.jpg)
(35)
• To get a distributive interpretation forChinese sentences like (35), the overt distributivity operator dou has to be used• in Chinese, then, dou is a carrier of
the ∗-operator (Lin 1998, Yang 2001)
Chinese
Tamen mai-le yi-bu chezi.They buy-Asp one-CL car‘They bought a car.’ Lin (1998), 201
101
![Page 102: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/102.jpg)
(36) Eine Kanne Milch hat jeweilsein Pfund Käse produziert.
• jeweils may be given the following interpretation:
(37) λP<st>λe [e = σe′ [P(e′) ∧ e′ < e]]
German
102
![Page 103: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/103.jpg)
• Phrasal plurality is not alwayslinked to nominal [plural]
• the feature [plural] does not have to bethe one and only possible source ofphrasal plurality, even in a languagethat also has [plural]
103
![Page 104: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/104.jpg)
• There is some indication thatsubje)-distributivity is hard to getwhen the subje) is left in a low position
(38) Am Nebentisch rauchten vier Männer eine Zigarre.
• Subje)s sitting in low positions are alsoknown to have different agreement properties in some languages
(39) Il &t arrivé d& enfants.‘There is arrived children’
Subje)-distributivity
104
![Page 105: n the lurality of erbsastechow/Lehre/SS06/Events/Kratzer05.k… · H ow do you pluralize a predicat e? ¥ F ollowing Kri&a '1986( w e ext end the sum operation t o ordered pairs built](https://reader030.vdocuments.us/reader030/viewer/2022013021/5f02a6bb7e708231d4055442/html5/thumbnails/105.jpg)
Thank you!