handout

Transcript

handout
Workshop on Weak Referentiality
Utrecht, 29 March
Comparative Issues in the Syntax and Semantics of Bare Predicate Nominals:
Towards a Principled Account
Denis Delfitto (joint work with Gaetano Fiorin)
1a.
1b.
2a.
2b.
3a.
3b.
Questa è acqua
this is water
Ieri ho bevuto acqua
yesterday I have drunk water
Questo è *(un) tavolo
this is (a) table
Ieri ho acquistato *(un) tavolo
yesterday I have bought (a) table
Questi sono tavoli
these are tables
Ieri ho acquistato tavoli
yesterday I have bought tables
4.
Questions:
(A) how does it come that bare mass nouns and bare plurals are acceptable both as arguments and
as predicates?
(B) How does it come that bare singulars are excluded from both positions?
(C) How does it come that indefinite noun-phrases are compulsory as predicates?
5. The domain of interpretation is a Complete Atomic Join Semilattice
5a.
{a, b, c}
…
{a, c} {a, b} {a, b}
…
a b c
…
|
|
|
PL(F)
F
5b.
i. {a, b} ≤ {a, b, c}
ii. a ≤ {a, b}
5c.
PL(F) = λx [¬F(x) ∧ ∀y[y ≤ x ∧At(y) → F(y)]]
5d.
5e.
=
If you have children, you should behave responsibly
Everyone who has children should behave responsibly
6.
Residual Questions:
(A) how does it come that bare mass nouns and bare plurals are acceptable as arguments?
(B) How does it come that indefinite noun-phrases are compulsory as predicates?
7a.
NbrP
=
λxPL table(xPL)
Nbr
nP
=
Bare
plural
predicate
[+plural]
n
N
[ucount]
7b.
NbrP
=
λxAT
table(xAT)
Nbr
nP
=
Bare
singular
predicate
[+sing]
n
N
[ucount]
8a.
DP
=
xPL
D
=
xPL
NbrP
λxPL : table(xPL). xPL
=
Bare
plural
argument
[+plural]
Nbr
nP
[+plural]
8b.
DP
=
xAT
D
=
xAT
NbrP
λxAT : table(xAT). xAT
=
Bare
singular
argument
[+sing]
Nbr
nP
[+sing]
9. A preliminary answer to the Residual Questions in 6.:
(A) Bare mass nouns and bare plurals are DPs when found in argument position. Under a
presuppositional treatment of the predicative restriction nP, these DPs simply introduce free
variables (type <e>: xAT and xPL) ranging over atoms and pluralities, respectively. Existential closure
or unselective binding follows (exemplified in 10.), triggering type-lifting of the usual sort (from <e>
to <et, t>)
10a.
10b.
I met students the whole day
Syntactic puzzles are always difficult to solve
(existential closure of xPL)
(unselective binding of xPL by ∀)
(B) Indefinite noun-phrases are compulsory (also) as predicates because the indefinite article is not a
Determiner! It is simply a valued occurrence of “sing” (say, in Nbr), necessary to value the unvalued
count feature on n!
(C) There is an important consequence: Singular indefinite noun-phrases may have the syntax in 7b
(they are then directly interpreted as properties (type: <e,t>) or the syntax in 8b (they are then free
variables ranging over singularities): in this case we may assume that predication structures of the
sort This is a table provide a dedicated functional head (say, Pred) encoding λ-astraction over the free
variable. As we will see, this possibility leads to a well-defined range of interpretive variation
11. Prediction: Bare Singular Predicates should be ruled out!
12a.
12b.
12c.
12d.
12e.
12f.
13.
*Lui/*Mario è uomo
he/Mario is man
*Lui/*Fido è cane
he/ Fido is dog
*Questo è tavolo / libro
this is table / book
Lui/Mario è un uomo
he/Mario is a man
Lui/Fido è un cane
he/Fido is a dog
Questo è un tavolo / un libro
this is a table / a book
Lui/Mario è medico/insegnante/cattolico (roles, professions, religious qualifications, etc.)
he/Mario is doctor/teacher/Catholic
14-16. Some more fine-grained empirical properties of Bare Singular Predicates:
14a.
14b.
15b.
15b.
16a.
16b.
*Questo è medico/insegnante/cattolico
this is doctor/teacher/Catholic
Questo è un medico/un insegnante/un cattolico
this is a doctor/a teacher/a Catholic
(sensitivity to the nature of the subject)
Lui/Mario è uomo di assoluta onestà
he/Mario is man of extraordinary probity
Lui/Fido è cane di grandi qualità
he/Fido is dog of great qualities
(sensitivity to modification)
La Queen Elizabeth è *nave
(extension to inanimate subjects)
the Queen Elizabeth is ship
La Queen Elizabeth è (?) nave di grande mole / nave di linea
the Queen Elizabeth is ship of great bulk / liner
17.
Deriving the distribution of Bare Singular Predicates
17a.
The set of data in (13): valuation of ucount in n depends on subject-predicate agreement: pronouns
and proper names may be endowed with a valued +count feature, due to the conceptual
features they may express
17b.
The set of data in (14a): the demonstrative is not endowed with a valued +count feature
(i)
*Questo è cane / ok questo è oro
this is dog / this is gold
(ii)
*Questa è sedia / ok questa è acqua
17c.
The set of data in (14b): valuation takes place by means of the indefinite article, we need not
care about the subject
17d.
The set of data in (12a-b): Why is subject-predicate agreement not enough?
Proposal: The property ascribed to the subject must be distinct from the conceptual feature
presupposed by the subject
(i) Suppose I am asked to guess – at a masked party – who of my masked friends is a man
and who is a woman. I can then say: Lui è uomo / lei è donna, etc.
(ii) I can easily say: Lui è uomo to express that he is a male, or to express that he is a macho or
extremely brave…
Similarly, when we ascribe a role, profession or religious qualification to someone, we are
normally presupposing that she is a human being! Ascribed property and presupposed
property are thus clearly distinct
17e.
The set of data in (15)-(16): the modified property is clearly distinct from the original – i.e.
non-modified – property
18.
Primitive properties and derived properties
We predict that Bare Singular Predicates (whenever syntactically licensed through subjectpredicate agreement) are nPs: they should then express primitive properties, i.e. wellestablished properties. Indefinite noun-phrases are derived properties: they are turned back
into properties by applying λ-astraction to the free variable expressed at the DP-level (cf. 9C)
18a.
18b.
19a.
19b.
?Mario è medico competente
Mario is competent physician
Mario è medico della mutua
Mario is general practitioner
??Mario è buon cattolico
Mario is good Catholic
Mario è cattolico praticante
Mario is Practising Catholic
20.
Only derived properties (= Indefinite noun-phrases) provide discourse antecedents: the reason
is that they provide an object-referring expression (type <e>) at the DP-level: a free variable
ranging over atoms. Primitive properties (= bare singular predicates) do not provide doscourse
antecedents, since they are not object-referring at any level of their derivation
20a.
Mario è induista. In precedenza non *ne avevo mai incontrati
Mario is induist. In the past (I) not CL-gen had never met
‘Mario is induist. In the past I had never met any of them’
Mario è un induista. In precedenza non (?)ne avevo mai incontrati
Mario is a induist. In the past (I) not CL-gen had never met
‘Mario is induist. In the past I had never met any of them’
20b.
21-22. Semantic shift with indefinite noun-phrases
21a.
21b.
Mario è un avvocato
Mario is a lawyer
*Mario è avvocato
Mario is lawyer
(meaning: he is a good debater)
(meaning: he is a good debater)
22a.
22b.
22c.
22d.
23.
Il marito di Maria è un dittatore
(meaning: he manifests authoritarian behavior)
Maria’s husband is a dictator
*Il marito di Maria è dittatore
(meaning: he manifests authoritarian behavior)
Maria’s husband is dictator
Berlusconi è stato Primo Ministro per 8 anni senza esser mai stato un Primo Ministro
Berlusconi has been PM for 8 years without ever being a PM
Berlusconi è stato un Primo Ministro per 8 anni senza esser mai stato Primo Ministro
Berlusconi has been a PM for 8 years without ever being PM
Why should semantic shift be limited to Indefinite noun-phrases? In principle, λ-astraction
through Pred should simply restore the original property!
Proposal: Covert operations come with a cost and are thus bound to produce novel
interpretations (D. Fox).
Corollary: λ-abstraction as encoded by Pred, whenever applied to object-referring indefinite
nouns, should deliver a distinct interpretation with respect to the interpretation delivered by
the bare predicates. This is exactly what we find!
24.
Given a predicative constant F, we have thus:
(i) Primitive property = λx.F(x)
(ii) Derived property = λx [∃P∀w (P w ⊇ Fw) ∧ P(x)],
where P is a variable ranging over properties
Conclusion: Semantic shift with Indefinite Singulars is derived from general properties of the
language design!
25-26. Another fine-grained prediction: Semantic shift is uniformly ruled out with Bare Singular
Predicates (that are nPs) but should be allowed with Bare Plural Predicates (that may be DPs)
25a.
25b.
26a.
26b.
In famiglia, Mario è un vero dittatore
in his family life, Mario is a true dictator
*In famiglia, Mario è vero dittatore
in his family life, Mario is true dictator
In famiglia, Mario e Gianni sono dei veri dittatori
in their family life, Mario and Gianni are (part.art.) true dictators
In famiglia, Mario e Gianni sono veri dittatori
in their family life, Mario and Gianni are true dictators
27.
Bare Singular predicates in English
27a.
27b.
Bill is *(a) doctor / *(a) teacher / *(an) employee
Obama is president of the US / head of the department / employee of the week
28.
Three questions:
(i) Why are Bare Singular Predicates generally excluded?
(ii) What is the role of uniquely instantiated properties (as in 27b)?
(iii) What is the relationship between (27b) and (28a)?
28a.
Obama is the president of the US / the head of the department / the employee of the week
29.
Here are some tentative answers to the questions in 28:
(iii) In (28a), the Definite Article in D is a ι-operator. We adopt the following version of the
Frege/Russell definition of the ι-operator (Chierchia 1998):
29a.
ι X = the largest member of X if there is one (else, undefined)
the dogs = ι DOGS = the largest plurality of dogs
the dog = ι DOG = the only dog (if there is one)
Given the semantics in (29a), the Definite Article expresses a valued +sing feature. In (28a), it
is this feature in D that values the unvalued ucount feature on n. This is why (28a) is formally
licensed as a DP
(ii) In (27b), world knowledge supplies us with the information that the properties expressed
by nP are uniquely instantiated. We propose that a Nbr-position is projected and that this
position is contextually identified as +sing on the grounds of the above-mentioned piece of
world knowledge. It is the +sing feature in Nbr that values ucount in n. And this is why (27b)
is formally licensed as a NbrP
(i) The reason why Bare Singular Predicates are generally ruled out in English is that ucount
cannot be valued by means of subject-predicate agreement. Subject-predicate agreement is
not robust enough in English:
29b.
29c.
30.
Bill is brave / Mary is brave
Mario è coraggioso / Maria è coraggiosa
Bill is an employee / Mary is an employee
Mario è un impiegato / Maria è un’impiegata
(adjectival predicates)
(nominal predicates)
Three noticeable consequences of the proposed analysis:
(i) The interpretation of the Definite Singular Predicate in (28a) is as in (30a):
30a.
λx. x = ιzPz
(ii) However, since (30a) is indistinguishable from λx. P(x) (where P is uniquely instantiated),
semantic shift effects are predicted for (28a). This prediction is borne out:
In a situation where the mayor of Verona has just achieved sweeping electoral success, (30b)
is acceptable but (30c) is not:
30b.
30c.
The mayor is the king of Verona
??The mayor is king of Verona
(iii) The behavior of Definite Singular Predicates in inverse copular constructions in Italian is
clearly reminiscent of their behavior in English (the property must be uniquely instantiated!).
By hypothesis, the English strategy is the only way to achieve valuation of ucount in these
structures. The relevant contrast is shown in (30d-e):
30d.
30e.
Un difensore dei diritti umani è Chomsky
a champion of human rights is Chomsky
Difensore dei diritti umani è Chomsky
the champion of human rights is Chomsky
Some concluding remarks: If the line of analysis put forward in this talk is on the right track, inquiry
into the syntax and semantics of Bare Singular Predicates shows that matters of syntactic
structure and feature-endowment are crucial in determining into which kind of logical
elements (variables, properties, etc.) bare nouns translate within a given language. More
controversially, perhaps, this inquiry reveals the potential richness of the interface
phenomena by means of which some selected aspects of conceptual stucture and world
knowledge interfere with the result of the syntactic computation.
Selected References
Baker, M. (2008). The syntax of agreement and concord. Cambridge, Cambridge University Press.
Beyssade, C. and C. Dobrovie-Sorin. “Predication and identity in copular sentences”, ms., Institut
Jean Nicod and LLF, Paris 7.
Blutner, R. (2004). Pragmatics and the lexicon. In L. Horn and G. Ward, eds.,
Handbook of Pragmatics. Oxford, Blackwell, 488-514.
Carlson, G. (1977). Reference to Kinds in English. Doctoral Dissertation. University of
Massachussetts, Amherst.
Cheng, L. and R. Sybesma (1999). “Bare and Not-So-Bare Nouns and the Structure of
NP”. Linguistic Inquiry 4, 509-542.
Chierchia, G. (1998). ‘Reference to Kinds Across Languages’. Natural Language Semantics
6, 339-405.
Dayal, V. (2004). ‘Number Marking and (In)definiteness in Kind Terms’. Linguistics
and Philosophy 27, 393-450.
Delfitto, D. (2002). Genericity in Language, Alessandria, Italy, Edizioni Dell’Orso.
Delfitto, D. (2003). “Results in semantics”. Lingue e linguaggio 2:261-275
Delfitto, D. (2005), “Bare Plurals”. In M. Everaert and H. van Riemsdijk (eds.), Blackwell
Companion to Syntax, Malden, Mass., Blackwell Publishing, vol. 1, 214-259
Delfitto, D. (2009). “Universals and semantics”. In S. Scalise, E. Magni and A. Bisetto
(eds.), Universals of language today. Berlin, Springer, 209-224
Delfitto, D. and G. Fiorin (2011). Person features and pronominal anaphora.
Linguistic Inquiry 42.2
Fox, D. (1999). Economy and semantic interpretation. MIT Press, Cambridge MA.
Heim, Irene (1992). The semantics of definite and indefinite noun phrases in English. PhD
thesis, University of Massachusetts at Amherst.
Heim, I., and A. Kratzer (1998). Semantics in Generative Grammar. Oxford, Blackwell.
Heycock, C. and R. Zamparelli (2003). Coordinated bare definites. Linguistic
Inquiry 34.3, 443-469.
Higginbotham, J. (1985). On semantics, Linguistic Inquiry 16, 547-593.
Horn, L. (1984). Towards a new taxonomy of pragmatic inference: Q-based and Rbased implicature’. In D. Schiffrin (ed.), Meaning, Form, and Use in Context: Linguistic
Applications. GURT84, 11-42, Washington; Georgetown University Press.
Kamp, H. and U. Reyle (1993). From discourse to logic. Dordrecht: Kluwer Academic
Press.
Kratzer, A. (2009). “Making a pronoun: Fake-indexicals as a window into the properties
of pronouns”. Linguistic Inquiry 40.2, 187-237.
Longobardi, G. (1994). Reference and proper names: A theory of N movement in
syntax and logical form. Linguistic Inquiry 25(4), 609-665.
Longobardi, G. (2001). How comparative is semantics? A unified parametric theory of
bare nouns and proper names, Natural Language Semantics, 9: 335-369.
Munn, A. and C. Schmitt (2005). Number and indefinites. Lingua 115, 821-855.
Partee, B. H. (1986). “Noun Phrase Interpretation and Type-Shifting principles”. In J.
Groenendijk, D. de Jong, and M. Stokhof (eds.), Studies in Discourse Representation Theories and
the Theory of Generalized Quantifiers. Dordrecht, Foris.
Pérez-Leroux, A. T. and T. Roeper (1999). Scope and the structure of bare nominals:
evidence from child language. Linguistics 37, 927-960.
Rizzi, L. (1994). “Early null subjects and root null subjects”, in T. Hoekstra and
B. D. Schwartz (eds.), Language acquisition studies in generative grammar,151–176.
Amsterdam, John Benjamins.
Swart, H. de and J. Zwarts (2009). “Less form – more meaning: Why bare singular
nouns are special”. Lingua 119.2, 280-295.
Zamparelli, R. (2003). “Bare predicate nominals in Romance languages”. In Høeg
Müller, Henrik and Alex Klinge (eds.), Essays on nominal determination: From morphology to
discourse management. 369 pp. (pp. 101–130).