WEAK NOUN PHRASES: SEMANTICS AND SYNTAX Barbara H. Partee University of Massachusetts, Amherst.

Slides:



Advertisements
Similar presentations
Problems of syntax-semantics interface ESSLLI 02 Trento.
Advertisements

1 Yoad Winter – Technion/Utrecht (Joint work with Sela Mador-Haim – Technion/UPenn) Spatial Meaning and Quantification SALT paper downloadable at:
First-Order Logic.
Semantics (Representing Meaning)
Ambiguous contents? Arvid Båve, Higher seminar in Theoretical Philosophy, FLoV, Gothenburg University, 8 May 2013.
Syntactic analysis using Context Free Grammars. Analysis of language Morphological analysis – Chairs, Part Of Speech (POS) tagging – The/DT man/NN left/VBD.
Albert Gatt LIN3021 Formal Semantics Lecture 9. In this lecture Noun phrases as generalised quantifiers: some further concepts.
© by Kenneth H. Rosen, Discrete Mathematics & its Applications, Sixth Edition, Mc Graw-Hill, 2007 Chapter 1: (Part 2): The Foundations: Logic and Proofs.
CAS LX 502 8a. Formal semantics Truth and meaning The basis of formal semantics: knowing the meaning of a sentence is knowing under what conditions.
Albert Gatt LIN3021 Formal Semantics Lecture 5. In this lecture Modification: How adjectives modify nouns The problem of vagueness Different types of.
2009/91 Predicates and Quantifiers (§1.3) In the sentence “x is greater than 3”: The phrase “x” denotes the subject - the object or entity that the sentence.
CAS LX b. Questions. Seeking truth Much of what we’ve done this semester has to do with characterizing (our knowledge of) the conditions under which.
Language and Logic: tools for the semantic study of natural language Henriëtte de Swart Barcelona, May 2005.
Sag et al., Chapter 4 Complex Feature Values 10/7/04 Michael Mulyar.
Type-shifting and beyond Henriëtte de Swart Barcelona, May 2005.
CAS LX a. A notational holiday. Sets A set is a collection of entities of any kind. They can be finite: {√2, John Saeed, 1984}. They can be infinite:
Syntax and Grammar John Goldsmith Cognitive Neuroscience May 1999.
1 CSC 594 Topics in AI – Applied Natural Language Processing Fall 2009/ Outline of English Syntax.
Type shifting and coercion Henriëtte de Swart November 2010.
Meaning and Language Part 1.
Models of Generative Grammar Smriti Singh. Generative Grammar  A Generative Grammar is a set of formal rules that can generate an infinite set of sentences.
Linguistic Theory Lecture 3 Movement. A brief history of movement Movements as ‘special rules’ proposed to capture facts that phrase structure rules cannot.
February 2009Introduction to Semantics1 Logic, Representation and Inference Introduction to Semantics What is semantics for? Role of FOL Montague Approach.
Debbie Mueller Mathematical Logic Spring English sentences take the form Q A B Q is a determiner expression  the, every, some, more than, at least,
Albert Gatt LIN 3098 Corpus Linguistics. In this lecture Some more on corpora and grammar Construction Grammar as a theoretical framework Collostructional.
Context Free Grammars Reading: Chap 12-13, Jurafsky & Martin This slide set was adapted from J. Martin, U. Colorado Instructor: Paul Tarau, based on Rada.
CAS LX 502 Semantics 3a. A formalism for meaning (cont ’ d) 3.2, 3.6.
LIN1180/LIN5082 Semantics Lecture 3
Logical Equivalence & Predicate Logic
Discrete Mathematics and Its Applications
Syntax Lecture 8: Verb Types 1. Introduction We have seen: – The subject starts off close to the verb, but moves to specifier of IP – The verb starts.
Relative clauses Chapter 11.
Lecture 9: The Gerund.  The English gerund is an intriguing structure which causes a particular problem for X-bar theory  [His constantly complaining.
A small semantics quiz. 2 Guess the determiner P Q  x(P(x)&Q(x)) 2. P Q  x(Plural(x)&P(x)&Q(x)) 3. P Q  x(P(x)  Q(x)) 4. P Q  x(P(x)&  y(P(y)
Continuous Discontinuity in It-Clefts Introduction Tension between the two approaches Our proposal: TAG analysis Equative it-cleft: It was Ohno who won.
November 2003CSA4050: Semantics I1 CSA4050: Advanced Topics in NLP Semantics I What is semantics for? Role of FOL Montague Approach.
Diathesis Alternations and NP Semantics Barbara H. Partee University of Massachusetts, Amherst.
October 2004CSA4050: Semantics III1 CSA4050: Advanced Topics in NLP Semantics III Quantified Sentences.
Lecture 7 Natural Language Determiners Ling 442. exercises 1. (a) is ambiguous. Explain the two interpretations. (a)Bill might have been killed. 2. Do.
1 Predicate (Relational) Logic 1. Introduction The propositional logic is not powerful enough to express certain types of relationship between propositions.
Entity Theories of Meaning. Meaning Talk Theory should make sense of meaning talk Theory should make sense of meaning talk What sorts of things do we.
Semantic Construction lecture 2. Semantic Construction Is there a systematic way of constructing semantic representation from a sentence of English? This.
Interpreting Language (with Logic)
Albert Gatt LIN3021 Formal Semantics Lecture 4. In this lecture Compositionality in Natural Langauge revisited: The role of types The typed lambda calculus.
Rules, Movement, Ambiguity
Chapter 2 Logic 2.1 Statements 2.2 The Negation of a Statement 2.3 The Disjunction and Conjunction of Statements 2.4 The Implication 2.5 More on Implications.
1 Introduction to Abstract Mathematics Predicate Logic Instructor: Hayk Melikya Purpose of Section: To introduce predicate logic (or.
Lectures 8-9 Ling 442. Exercises (1) Reconstruct the original English sentence for each: 1.|birds  fly| > ½ |birds| 2.dog  bite  {} 3.student  study_hard.
For Wednesday Read chapter 9, sections 1-3 Homework: –Chapter 7, exercises 8 and 9.
CS 285- Discrete Mathematics Lecture 4. Section 1.3 Predicate logic Predicate logic is an extension of propositional logic that permits concisely reasoning.
For Friday Read chapter 8 Homework: –Chapter 7, exercises 2 and 10 Program 1, Milestone 2 due.
Lecture 2 (Chapter 2) Introduction to Semantics and Pragmatics.
Albert Gatt LIN3021 Formal Semantics Lecture 8. In this lecture Noun phrases as generalised quantifiers.
LTAG Semantics for Questions Aleksandar Savkov. Contents Introduction Hamblin’s idea Karttunen’s upgrade Goals of the paper Scope properties of wh-phrases.
CAS LX 502 9b. Formal semantics Pronouns and quantifiers.
First-Order Logic Semantics Reading: Chapter 8, , FOL Syntax and Semantics read: FOL Knowledge Engineering read: FOL.
A small semantics quiz. 2 Guess the determiner P Q  x(P(x)&Q(x)) 2. P Q  x(Plural(x)&P(x)&Q(x)) 3. P Q  x(P(x)  Q(x)) 4. P Q  x(P(x)&  y(P(y)
1 Lecture 3 The Languages of K, T, B and S4. 2 Last time we extended the language PC to the language S5 by adding two new symbols ‘□’ (for ‘It is necessary.
Section 1.4. Propositional Functions Propositional functions become propositions (and have truth values) when their variables are each replaced by a value.
PHILOSOPHY OF LANGUAGE Some topics and historical issues of the 20 th century.
Propositional Logic. Assignment Write any five rules each from two games which you like by using propositional logic notations.
Chapter Eight Predicate Logic Semantics. 1. Interpretations in Predicate Logic An argument is valid in predicate logic iff there is no valuation on which.
English Syntax Week 12. NP movement Text 9.2 & 9.3.
Lecture 3: Functional Phrases
Statistical NLP: Lecture 3
Semantics (Representing Meaning)
Chapter 1 The Foundations: Logic and Proofs
CS201: Data Structures and Discrete Mathematics I
Week #2 – 4/6 September 2002 Prof. Marie desJardins
CS201: Data Structures and Discrete Mathematics I
Presentation transcript:

WEAK NOUN PHRASES: SEMANTICS AND SYNTAX Barbara H. Partee University of Massachusetts, Amherst

2 Acknowledgements Thanks to many students in classes at RGGU and MGU for data, suggestions, and ideas about weak NPs in Russian. Thanks to Vladimir Borschev, Elena Paducheva, Ekaterina Rakhilina, and Yakov Testelets for ongoing discussion of the Russian Genitive of Negation. This material is based upon work supported in part by the National Science Foundation under Grant No. BCS to B.H. Partee and V. Borschev.

3 Outline NPs as Generalized Quantifers,t> Determiners as functions Weak NPs and existential sentences Property-type interpretations of NPs Intensional contexts Genitive of negation hypothesis, conjecture for future research.

4 Introduction: NPs as Generalized Quantifiers Montague : Noun Phrases denote sets of properties. Semantic type for NPs: (e  t)  t

5 Some NP interpretations John λP[P(j)]type (e  t)  t (the set of all of John’s properties) John walks λP[P(j)] (walk)  walk (j) (function-argument application) j: type eP, walk: type e  t

6 NP interpretations, continued every student type (e  t)  t λP  x[student(x)  P(x)] (the set of properties that every student has) every student walks λP  x[student(x)  P(x)] (walk) (function-argument application)   x[student(x)  walk (x)]

7 NP interpretations, continued a student: λP  x[student(x) & P(x)] (the set of properties at least one student has) the king: λP [  x[king(x) &  y ( king(y)  y = x) & P(x))] (the set of properties the one and only king has)

8 Syntactic structure S 2 NP VP 2 walk DET CN every student Semantics: CN(P) (Common Noun (phrase)): type e  t VP: type e  t  Note: It is more common now to have DP where I have NP, and NP where I have CN(P).

9 Semantics of DET DET: interpreted as a function of type (e  t)  ((e  t)  t) it applies to CN meaning, type (e  t), to give a a generalized quantifier, a function of type (e  t)  t, which in turn applies to a VP meaning to give truth value. NP: type (e  t)  t

10 Semantic structure truth-value 2 function(arg1) (arg2) 2 walks function (arg1) every student || every|| ( || student|| ) ( || walks|| )

11 Determiners as functions ||Every||(A) = {B|  x ( x  A  x  B)}. Equivalently: ||Every|| = Q[ P[  x ( Q(x)  P(x) )]]. Some, a: takes as argument a set A and gives as result {B| A  B   }. || a || = Q[ P[  x ( Q(x) & P(x) )]]

12 So Determiner Properties Project through Whole Sentence  2  (arg2) 2 function (arg1) DET Determiners can license Negative Polarity Items inside NP and/or in “arg2”, the “rest of the sentence”. Weak vs. strong determiners: crucial for “existential sentences”

13 “Weak” determiners and existential sentences. Data: OK, normal:  There is a new problem.  There are three semantics textbooks.  There are many unstable governments. Anomalous:  #There is every linguistics student.  #There are most democratic governments.  #There is the solution. (# with “existential” there)

14 Semantic explanation – Milsark, Barwise and Cooper, Keenan Definition (Keenan 1987): A determiner D is a basic existential determiner if for all models M and all A,B  E, D(A)(B) = D(A  B)(E). English test: “Det CN VP” is true iff “Det CN which VP exist(s)” is true.

15 Examples (i) Three is an existential determiner: Three cats are in the tree iff three cats which are in the tree exist. (ii) Every is not existential: Suppose there are 5 cats, and 3 are in the tree. Then: “Every cat is in the tree” is false but “Every cat which is in the tree exists” is true.

16 Existential = Symmetric Basic existential determiners = symmetric determiners.  One can prove, given that all determiners are conservative (Barwise and Cooper 1981), that Keenan’s basic existential determiners are exactly the symmetric determiners. Symmetry: A determiner D is symmetric iff for all A, B, D(A)(B) ≡ D(B)(A).

17 Testing symmetry Weak (symmetric): Three cats are in the kitchen ≡ Three things in the kitchen are cats. More than 5 students are women ≡ More than 5 women are students. Strong (non-symmetric): Every Zhiguli is a Russian car  Every Russian car is a Zhiguli.

18 Test symmetry in Russian Три черные кошки на кухне ≡ Три вещи на кухне черные кошки три is weak Все черные кошки на кухне  Все вещи на кухне черные кошки все is strong See abstract for more discussion of Russian

19 Further related topics Partee (1991) suggests a systematic connection between weak-strong, Heimian tripartite structures, and topic-focus structure, which is further explored in Hajicová, Partee and Sgall (1998). See also Partee (1989) on the weak-strong ambiguity of English many, few and Babko-Malaya (1998) on the focus-sensitivity of English many and the distinction between weak много and strong многие in Russian.

20 CNP and NP types Prototypical case:  NPs are type e (proper names, pronouns, referring terms) or type,t> (quantifier phrases).  CNPs are type (predicates). Sometimes NPs shift to type (predicate nominals: John is a student.) Sometimes bare CNPs shift to e type (Russian singular count nouns in e-type argument positions: Молодой лингвист кончил свой доклад во- время.)

21 Property-type NP interpretations NP types: e: entity type : (extensional) predicate type >: (intensional) property type,t>: generalized quantifiers: Montague’s NP type, and the agreed-on type for essentially quantificational NPs

22 Property-type NP interpretations, continued Where do predicate-type and property-type NPs appear? Predicate-type : (i) predicate nominals: John is a student. (ii) Some say also: There is a cat on the mat. (McNally, Landman, Kamp) Property-type: recent proposals, to be discussed next.

23 Property-type NP interpretations, continued Zimmermann 1993: argues against Montague’s analysis of “intensional transitive verbs” like seek Montague: object is intensional generalized quantifier, type >,t>. Zimmermann: object is property-type, type >.

24 Fundamental properties of intensional contexts (11)Caroline found a unicorn. (extensional, unambiguous) (12)Caroline sought a unicorn. (intensional, ambiguous) Sentences with seek are ambiguous between a specific and a non-specific reading (or transparent vs. opaque reading). (11) is unambiguous, (12) is ambiguous. On the opaque reading of (12), the existence of a unicorn is not entailed.

25 Fundamental properties of intensional contexts, continued Substitution of extensionally equivalent expressions in an intensional context does not always preserve truth-value. Caroline is looking for a unicorn The set of unicorns = the set of 13-leaf clovers Not entailed: Caroline is looking for a 13- leaf clover

26 The classical analysis Everyone agrees since Frege: the complement of seek must be intensional, not extensional. Quine (1960) argued that seek should be decomposed into try to find. He argued that intensionality is (in general) the result of embedding a proposition under an intensional operator, such as the verb try. Within Caroline try [Caroline find x], there are then two places a quantifier phrase could take its scope: the higher clause, giving the transparent reading the lower clause, giving the opaque reading.

27 The classical analysis, continued Montague (1973) argued that the same semantic effect can be achieved with a simpler syntax: seek + NP, if NPs like a unicorn express Generalized Quantifiers. The argument of an intensional verb gets an intensional operator “^” applied to it. So Montague treats a verb like seek 1 as denoting a relation between an individual and an intensional generalized quantifier. The transparent reading results from “quantifying in” to an e-type argument position of seek 2, a relation between two individuals.

28 The classical analysis, continued For Montague, the relation between seek and try to find is captured not by decomposition but by a meaning postulate. Meaning postulate: seek’ (x, ^Q)  try’ (x, ^[Q( y find’ (x,y))]).

29 Problems with the classical analysis But there are problems with Quine’s and Montague’s classical analyses. Among other problems, (Zimmermann 1993) points out an overgeneration problem: True quantifier phrases like every doctor are normally unambiguously “transparent” after intensional transitive verbs like compare, seek, although they are ambiguous in constructions like try to find, so Montague and Quine predict ambiguity.

30 Problems with the classical analysis, continued. Simple indefinites with a, on the other hand, are indeed ambiguous with intensional verbs. Compare: (a) Alain is seeking a comic book. (ambiguous) (b) Alain is seeking each comic book. (unambiguous; lacks ambiguity of (c)) (c) Alain is trying to find each comic book. (ambiguous)

31 Zimmermann’s alternative account Zimmermann: we can capture the relevant generalizations if we treat definite and indefinite arguments of intensional verbs, (but not generalized quantifiers) as properties, type >. Zimmermann’s proposal is that a verb like seek 1 denotes a relation between an individual and a property.

32 Zimmermann’s alternative account, continued Zimmermann: seek a unicorn: seek’(^unicorn’) ( ^ is Montague’s ‘intension operator’) This is a case of NP type-shifting by coercion: seek demands a property-type argument. We know that indefinite NPs easily shift into > readings, as was shown for predicate nominals in (Partee 1986). transparent, or de re, reading: “quantify in” to e-type argument position of seek 2.

33 Russian Genitive of Negation Hypothesis: Wherever we see Nom/Gen and Acc/Gen alternation (both under negation and under intensional verbs), Nom or Acc represents an ordinary e-type argument position (‘referential’; and may be quantified), whereas a Gen NP is always interpreted as property-type:, or >.

34 Russian Genitive of Negation, continued. In the case of intensional verbs like ждать, this agrees with Zimmermann’s analysis. There is a similar connection to the work of van Geenhoven, who treats ‘weak’ object NPs in West Greenlandic as “incorporated to the verb”: they are not fully independent objects, but get an existential quantifier from the verb.

35 Russian Genitive of Negation, continued. In the case of Genitive of Negation, the construction is not intensional. But Russian linguists from Jakobson to Paducheva have argued that Genitive- marked NPs have reduced “referential status”, and Western linguists have generally claimed that they must be “indefinite”.

36 Russian Genitive of Negation, continued. A shift of NP-meanings to property-type under Negation could capture those insights and intuitions. But negation is not really intensional; there seem to be different kinds of ‘reduced referentiality’. We have a few facts in favor, but also some doubts.

37 Russian Genitive of Negation, continued. Evidence in favor: (a)Петя нашел ответ. (b)Петя не нашел ответ. (c) Петя не нашел ответа. Competing analyses of case (c):  Standard analysis: definite vs. indefinite -- (c) has an indefinite (weak) NP under the scope of Neg.  Suggested analysis: (c) has a property-type NP under the scope of Neg.

38 Russian Genitive of Negation, continued. Evidence casting doubt on property analysis: (a)Я не видела Машу. (b)Я не видела Маши. The (b) case causes problems for all “quantificational” approaches to the Genitive of Negation, unless we suggest a meaning like “any trace of Masha”. (c)Ваня не решил все задачи. (d)Ваня не решил всех задач. Exs. (c-d) may differ in scope, but not in intensionality.

39 Russian Genitive of Negation, continued. For examples with negated indefinites, the property-type analysis for Gen Neg examples looks good. For examples with proper names or strong quantifiers, the property-type analysis does not look good. But no uniform semantic approach looks good for all cases (yet). This issue is still under exploration – more coming in future years.

40 Спасибо за внимание.