Davidsonian- and Kimian-states

Transcripción

Davidsonian- and Kimian-states
JeNom 4, 4èmes Journées d’étude sur les nominalisations
University of Stuttgart, 16-17 June 2011
Syntax-semantics mapping in nominalizations: Davidsonian- and
Kimian-states
Antonio Fábregas & Rafael Marín
Universitetet i Tromsø & CNRS – Université Lille 3
1 The APH, nominalizations and aspect
This paper concentrates on the domain of deverbal nominalizations, specifically those
that denote states, and addresses the following questions:
1. The internal classification question: Are state nominalizations K-states, Dstates or both? In case they are restricted to a particular class of states, which one?
2. The origin question: Are the kind of states denoted by a nominalization
dependent on properties of the base or not?
3. The productivity question: Are there stative verbs that do not have a state
nominalization? If so, what are the restrictions and what do they tell us about the
previous two questions?
Background problem:
 Aspect and tense are instantiated primarily in the verbal category crosslinguistically, and its instantiation in the form of nouns is a secondary process.
 In consequence, we expect the morphological processes known as
nominalizations, which add nominal features to a verbal base, not to be able to
add aspectual or temporal information to the structure  All the aspectual
information contained in the nominalization must come from the verbal base.
We state this hypothesis, the Aspect Preservation Hypothesis in the following form:
(i)
Given a domain populated with eventualities and properties mapped into the
temporal domain, if  is a verb and  is the nominalization derived from it, then:
[[]] ⊆ [[]]
That is, if the denotation of  is a particular set of eventualities –happenings, states, left
boundary happenings, etc.- then the denotation of  must be a set that includes only
objects contained in . This is the hypothesis that we will argue for in this paper, taken
state nominalizations as an empirical basis. We will argue that:
1. The classification answer: All state nominalizations are K-states
2. The origin answer: This is independent of whether the original predicate
denoted a K-state or not.
3. The productivity answer: Only those D-state predicates that contain a K-state
in their internal structure can have state nominalizations.
Before addressing these topics, however, we must briefly determine what is a state
nominalization and account for some tests that we will use to identify them. This would
be question number four:
1
JeNom 4, 4èmes Journées d’étude sur les nominalisations
University of Stuttgart, 16-17 June 2011
4. The descriptive question: How do we know that a noun denotes a state?
2 Identifying state nouns
2.1 Take place and take to an end
Take place
(1)
a. La discusión tuvo lugar en el pasillo.
The discussion took place in the corridor
b. *El aburrimiento tuvo lugar esta mañana.
The boredom took place this morning
c. *La construcción de piedra tuvo lugar en Stuttgart.
The stone building took place in Stuttgart
Take to an end
(2)
a. El profesor llevó a cabo la discusión sin interrupciones.
The teacher took to end the discussion without interruptions.
b. *El profesor llevó a cabo el aburrimiento de los estudiantes.
The teacher took to end the boredom of the students.
c. *El professor llevó a cabo la construcción de piedra.
The teacher took to end the stone building
2.2 This happened...
(3)
a. La construcción del puente fue larga. Esto sucedió porque...
The building of-the bridge was long. This happened because...
b. El enfado de Juan fue grande aquella tarde. #Esto sucedió...
The anger of Juan was big that evening. This happened...
2.3 Modification by fast or slow
(4)
a. la construcción rápida del puente
the construction fast of-the bridge
‘the fast building of the bridge’
b. #el aburrimiento rápido de Juan
the boredom fast of Juan
‘the fast boredom of Juan’
2.4 PP temporal modifiers
(5)
a. una discusión de varias horas.
a discussion of several hours
b. una preocupación de varias semanas
a preoccupation of several weeks
c. #una construcción de piedra de varios meses.
a stone building of several months
2
JeNom 4, 4èmes Journées d’étude sur les nominalisations
University of Stuttgart, 16-17 June 2011
2.5 Temporal reading of long
- Possible temporal reading of adjectives ambiguous between physical extension and
temporal extension
(6)
a. larga discusión, corta operación
long discussion, short operation
b. larga preocupación, corto enfado, infinito aburrimiento
long preoccupation, short anger, infinite boredom
c. #larga construcción de piedra
long stone building
2.6 Argument structure
The sentences in (7a) and (7b), where the argument structure is missing, are
ungrammatical if there is no preceding discourse which allows us to recover the
arguments; this contrasts with the object noun in (7c), where no argument structure is
necessary.
(7)
a. #El aburrimiento preocupaba a los profesores.
The boredom worried ACC the teachers.
b. #La preocupación causó su dimisión.
The preoccupation caused his resignation.
c. La construcción de piedra preocupaba al gobierno.
The stone building worried ACC-the government.
2.7 Agents
- Unavailability of agents
(8)
a. Luis cuidadosamente aburrió a su hijo para que se durmiera.
Luis carefully bored ACC his son so that SE falls.asleep.SUBJ
b. *el aburrimiento de su hijo por Juan
the boredom of his son by Juan
c. La película aburrió a Juan
The movie bored ACC Juan
d. el aburrimiento de Juan con la película
the boredom of Juan with the movie
This distinction is very clear in French, where the preposition par ‘by’ is used only with
agents. This preposition is unavailable in state nominalizations; pour, which introduces
non volitional agents, has to be used instead.
(9)
la préoccupation de Jean {pour / *par} l’économie
the preoccupation of Jean {for / by} the economy
2.8 Countability
(10)
a. las diversas destrucciones de Constantinopla a lo largo de la historia.
the many destructions of Constantinople across history
3
JeNom 4, 4èmes Journées d’étude sur les nominalisations
University of Stuttgart, 16-17 June 2011
b. *los tres aburrimientos de Juan.
the three boredoms of Juan
b’. los dos enfados de Juan (*de varias horas).
the two angerings of Juan (*of several hours)
c. varias construcciones de piedra
several stone buildings
Event nouns State Nouns Object nouns
Take place and take to an end
+
–
–
+
–
–
This happened
Event adjectives (fast or slow)
+
–
–
PP temporal modifiers
+(normally)
+
–
Temporal reading of long
+
+
–
Argument structure
+
+
–
Agents
+
–
–
Count
+
–
+
3 The classification question. Are state nominalizations D-states or K-states?
In the recent literature (Maienborn, 2003, 2005; Rothmayr, 2009) two different classes
of states have been proposed: Davidsonian states (D-states) and Kimian states (Kstates).
D-states:
lie, sit, wait, glow
They contain an ‘e’ argument and allow place and manner modification
K-states:
own, weight, cost
admire, like, worry (psychological states)
They lack an ‘e’ argument and reject place and manner modification
Two questions arise: (i) To which class of states belong state denoting nominalizations.
(ii) Wether the kind of state denoted is dependent on the
properties of the base verb or imposed by the nominalization
process.
3.1 Tests to differentiate the two classes of states in the verbal domain
3.1.1 Place modification
K-STATES NEVER ALLOW EVENT / STATE LOCATING PLACE MODIFIERS
(11)
a. Die spanische Armada lag bei Calais vor Anker
the Spanish armda lay near Calais at anchor
[Maienborn 2005: (27b)]
b. *Die Tomaten wiegen neben den Zwiebeln 1kg.
The tomatoes weigh besides the onions 1 kg.
[Maienborn 2005: (26a)]
4
JeNom 4, 4èmes Journées d’étude sur les nominalisations
University of Stuttgart, 16-17 June 2011
3.1.2 Manner modification
K-STATES NEVER ALLOW MANNER ADVERBIALS
(12)
a. Carol sass reglos am Tisch
Carol sat motionless at.the table
[Maienborn 2005: (31b)]
b. *Paul besitzt sparsam viel Geld
Paul owns thriftily much money
[Maienborn 2005: (30b)]
Notice, however, that there is the question of whether some of these manner modifiers
are out because they presuppose that there must be an agent able to control that manner.
In our tests we will avoid this problem by using adverbs that do not necessarily imply
the presence of an agent, confirming anyways Maienborn’s data.
3.1.3 Temporal reading of ein bisschen
K-STATES NEVER ALLOW A TEMPORAL READING OF EIN BISSCHEN (‘A LITTLE’)
(13)
a. Paul hat ein bisschen im Garten gesessen.
Paul has a bit in.the garden sat.
‘Paul sat in the garden for a short while’
[Maienborn 2003: (37b)]
b. #Carol ähnelte ein bisschen ihrer Grossmutter
Carol resembled a bit her grandmother
‘Carol slightly resembled her grandmother’
[Maienborn 2005: (39)]
3.2 Tests to differentiate the two classes of states in the nominal domain
The next step is clear: let us apply similar tests to state nominalizations and determine
what their behaviour is.
3.2.1 Place modification in the nominal domain
Event nominalizations accept locative modifiers, to the extent that the base verb allows
them:
(14)
a. su encuentro (junto al puente)
their encounter (next to-the bridge)
b. su operación (en el quirófano dos)
his operation (in the operating room #2)
STATE NOMINALIZATIONS DO NOT ALLOW FOR LOCATIVE MODIFIERS if they locate the
state denoted:
(15)
a. su brillo (*junto al puente)
its glow next to-the bridge
5
JeNom 4, 4èmes Journées d’étude sur les nominalisations
University of Stuttgart, 16-17 June 2011
b. su preocupación (*bajo la cama)
his preoccupation under the bed
3.2.2 Manner modification in the nominal domain
We will only take into account manner adverbials compatible with states, like elegantly,
scruffily, orderly, disorderly, upright, rigidly, silently, materialized in the nominal
domain as adjectives.
Event nominalizations accept manner modification:
(16)
a. su elegante interpretación
her elegant performance
b. su silencioso llanto
his silent crying
STATE NOMINALIZATIONS DO NOT ALLOW FOR MANNER MODIFIERS:
(17)
a. su (*elegante) aburrimiento
his (elegant) boredom
b. su (*buena) espera
her good wait
3.2.3 The ein bisschen test in the nominal domain
State nouns allow for temporal extension in the absolute sense, as we have seen (as
opposed to quality nouns and, depending on the aspectual type, event nouns).
The ein bisschen test shows that with a modifier that does not in itself carry a temporal
meaning, the salient reading is a degree one with K-states.
Thus, we will use an adjective which is ambiguous between a degree and a temporal
reading: reducido (‘reduced’).
With events, reducido allows both a temporal reading and a degree reading:
(18)
a. una representación reducida de Hamlet
a reduced performance of Hamet
‘a short performance of Hamlet’
b. un ataque reducido contra los enemigos
an attack reduced against the enemies
‘a short attack against the enemies’
With states, the temporal reading is completely absent and only a degree reading
emerges:
(19)
a. una preocupación reducida
a preoccupation reduced
‘a slight preoccupation’
b. un brillo reducido
6
JeNom 4, 4èmes Journées d’étude sur les nominalisations
University of Stuttgart, 16-17 June 2011
a shine reduced
‘a slight shine’
3.3 State nominalizations behave as K-states
The behaviour of D-states, K-states and state nominalizations is summarized in the
following Table:
D-states K-states State nominalizations
Locative modifiers
+
–
–
Manner modifiers
+
–
–
Ein bisschen
+
–
–
Partial generalizations:
 State nominalizations do not allow for locative modifiers if they locate the
state denoted.
 State nominalizations do not allow for manner modifiers
 State nominalizations do not allow the nominal equivalent of ein bisschen to
denote temporal extension
Global generalization (I): State nominalizations behave as K-states
4 The origin question: are state nouns K-states by virtue of the properties of the
corresponding verbs?
At this point, a question comes out naturally: are state nouns behaving like K-states
because of the properties of the verbal base? In other words, are these nominalizations
K-states because the original verbs were already K-states?
We will show that the nominalization behaves like a K-state even in those cases where
the original predicate behaves like a D-state.
4.1 Asymmetries with place modifiers
(20)
a. La lámpara brilló un momento frente a la casa.
The lamp shined a momento in.front of the house
b. su brillo de un momento (*frente a la casa)
its shine of a moment (in.front of the house)
Generalization: even if a state verb allows place modification, the corresponding
nominalization rejects it.
4.2 Asymmetries with manner modifiers
(22)
a. Estas dos ideas se vinculan elegantemente en la teoría de Popper.
These two ideas SE associate elegantly in the theory of Popper
‘These two ideas elegantly associate in Popper’s theory’
b. su (*elegante) vinculación en la teoría de Popper.
7
JeNom 4, 4èmes Journées d’étude sur les nominalisations
University of Stuttgart, 16-17 June 2011
(23)
their (elegant) association in the theory of Popper
a. Juan sufría silenciosamente.
Juan suffered silently.
b. su (*silencioso) sufrimiento
its silent suffering
Generalization: even if a state verb allows manner modification, the corresponding
nominalization rejects it.
4.3 Asymmetries with ein bisschen
Finally, we show that un poco (‘a little’) allows for temporal readings with predicates
that, once nominalized, only give the degree reading of reducido.
(24)
a. La lámpara brilla un poco
The lamp shines a bit
Two readings: The lamp shines for a short time
The lamp shines with a bit of intensity
b. su reducido brillo
its reduced shine
One reading: The lamp shines with little intensity
Remember (examples in (18)) that reducido allows for temporal extension readings.
Generalization: even if a state verb allows for temporal readings of ein bisschen, the
corresponding nominalization rejects it.
Verb
Place modifier
X brilló frente a la casa
Manner modifier se vinculan elegantemente
Ein bisschen
Brilla un poco (time and degree)
Nominalization
su brillo (*frente a la casa)
su (*elegante) vinculación
Su reducido brillo (degree only)
Partial generalizations:
 Even if a state verb allows place modification, the corresponding nominalization
rejects it.
 Even if a state verb allows manner modification, the corresponding
nominalization rejects it.
 Even if a state verb allows for temporal readings of ein bisschen, the
corresponding nominalization rejects it.
Global generalization (II): State nominalizations are always K-states even when the
verbal base is a D-state.
5 The productivity question. Do all D-state verbs give rise to state nominalizations?
The answer is no. There are some D-state verbs that reject state nominalizations.
Either they lack a nominalization entirely –and, as nouns, are expressed as infinitives,
(26), or the nominalization denotes an object or another notion (27).
(26)
el suave fluir del río
8
JeNom 4, 4èmes Journées d’étude sur les nominalisations
University of Stuttgart, 16-17 June 2011
(27)
The soft flow.inf of-the river
un yaci-miento de minerales
a lay-nominalizer of minerals
‘a deposit of minerals’
We call these Stubborn D-states, because they refuse to become K-states by losing
their eventive properties in the nominal domain.
They typically belong to one of these classes:
(28)
(29)
(30)
Stubborn D-states
Predicates of position and posture
yacer, rodear, estar sentado, estar erguido, estar tumbado
lie, surround, to be sitting, to be standing, to be lying
Predicates of emission of liquids
manar, fluir, sangrar, brotar
flow, flow, bleed, spring
‘Manner’ of staying predicates
colgar, flotar, pender
hang, float, hang
To these we can add the reading of several predicates in which they denote the process
of being continuously in contact (physical or otherwise) with another object, like guiar
(guide), llevar (carry), acompañar (accompany), etc.
None of them has a state denoting nominalization; independently, they cannot have
event nominalizations because they do not denote dynamic eventualities.
There is another classe of D-state verbs, those that that can give rise to state
nominalizations.
We call them Flexible D-states. They fall in one of these classes:
(31)
(32)
Flexible D-states
Meteorological phenomena and other processes without an agent
temblar, vibrar, reflejar, neblinear, permanecer, mantener
shake, vibrate, reflect, be.foggy, remain, keep
Emission of light and sound
brillar, resplandecer, silbar, pitar, zumbar
shine, shine,
whistle, whistle, buzz
Flexible D-states allow for state nominalizations:
(33)
a. un brillo de varios segundos (*tuvo lugar)
a shine of several seconds
b. un pitido de varios segundos (*tuvo lugar)
a whistling of several seconds
Absence of generalization: Some D-state verbs give rise to state nominalizations
9
JeNom 4, 4èmes Journées d’étude sur les nominalisations
University of Stuttgart, 16-17 June 2011
In the next section we will try to turn this non-generalization into a generalization that
shows that stubborn D-states and flexible D-states form two natural classes, and then we
will be in a position to analyze our data.
6 Analysis
Our analysis has two ingredients. The first one is the principle in (34), which subsumes
the empirical facts observed in nominalizations:
(34)
A state nominalization cannot denote a D-state, and must denote a Kstate
The second one is the proposal in (35), which we will argue for in the reminder of the
paper:
(35)
Some D-stative verbs contain in their denotation a subcomponent that
expresses a set of properties with temporal extension. Others do not have
this component.
Distinction between Stubborn D-states and Flexible D-states:
Flexible D-states contain inside their verbal denotation a set of properties –which, we
argue, behave as a K-state–, while Stubborn D-states lack this component of meaning.
Therefore, the denotation of the two classes of verbs is different and they form two
natural classes. This allows us to turn the non-generalization of §5 into a generalization:
Global generalization (III): Only those D-states containing a set of properties with
temporal extension (a K-state) can give rise to a state nominalization.
6.1 Event and intensity readings
As Bosque & Masullo (1998) noticed, adverbial quantifiers, when applied to verbal
predicates, give a variety of meanings which depend on the ingredients contained in the
predicate.
When a predicate contains a set of properties, these properties can be targeted by an
adverbial quantifier to express a degree of them: the I(ntensity) reading.
When this set of properties is missing, the degree reading of the adverb is replaced by
another interpretation (mostly temporal or argumental): the E(ventive) reading. If this
is not possible, the sequence become ungrammatical.
In this respect, contrast the following two sets of examples.
(36)
Mucho and other quantifiers with stubborn D-states
a. Juan yació mucho
Juan lied a lot (‘long time’)
c. Juan estuvo tumbado mucho
Juan was lying a lot (‘long time’)
10
JeNom 4, 4èmes Journées d’étude sur les nominalisations
University of Stuttgart, 16-17 June 2011
d. ??El río fluyó mucho
The river flew a lot
e. ??La fuente manó mucho
The fountain flew a lot
f. El cuadro colgó mucho de la pared.
The picture hanged a lot from the wall (‘long time’)
(37)
Mucho and other quantifiers with flexible D-states
a. Nevó mucho.
It.snowed a lot (‘with much intensity’)
c. La lámpara brilló mucho.
The lamp shined a lot (‘with much intensity’)
d. El fuego resplandeció mucho.
The fire glowed a lot (‘with much intensity’)
e. La tetera silbó mucho.
The tea-pot whistled a lot (‘with much intensity’)
f. Esta máquina pitó mucho.
This machine whistled a lot (‘with much intensity’)
In (36), the only interpretations available are E-readings, that is, readings related to an
event. Unsurprisingly, they are possible with D-states, which contain an event
argument.
This does not mean, however, that D-states systematically reject the I-reading.
Crucially, Stubborn D-states do, but not the Flexible D-states in (37).
It is worth to recall that K-states typically allow I-readings, as is the case of
psychological state verbs (Rothmayr, 2009):
(38)
a. Juan quiere mucho a María.
Juan loves much ACC Mary
‘Juan loves Mary a lot’
b. Juan se preocupa mucho.
Juan SE worries much
‘Juan is much worried’
6.2 A semantic analysis
Our proposal is that D-states come in two classes. In the first class, the stubborn D-state,
the denotation of the verb does not contain a K-state and denotes, directly, an event
which holds of an argument. In contrast, the second class of D-states, the flexible Dstates, denote the event of staying in a particular K-state during a period of time.
This is the denotation of a D-state. It denotes the event e of being in a particular state P
(Maienborn 2005).
(39)
Pex[P(e)  HOLDS(e, x)]
11
JeNom 4, 4èmes Journées d’étude sur les nominalisations
University of Stuttgart, 16-17 June 2011
This is the denotation of a K-state: it is the instantiation z, bound to times and worlds, of
a property P applied to an individual x (Maienborn 2005).
(40)
Pxz[ z  [P(x)]]
We follow the proposal in Kennedy & McNally (1999) about the way of representing a
property: it is a relation between individuals and degrees. We assume that all properties
are gradable (in other words, that relational adjectives, for instance, do not denote
properties).
(41)
dxP[P(d)(x)]
A K-state is built over the Property, as follows, illustrated with the K-state preoccupy.
(42)
Rxz[ z  [R(d)(x)]] (dx[preoccupy(d)(x)])
and therefore
xz[ z  [d[preoccupy(d)(x)]]]
The K-state contains a degree variable, and as such it can be bound by a quantifier. We
assume that a quantifier like mucho has the denotation in (43), following Kennedy &
McNally, and ignoring the contextual standard of comparison. It states that there exists
a degree of a property in an individual such as that it is ‘much’.
(43)
Rxd [much (d)  R(d)(x)]
That is, in order to get the I-reading, a quantifier must find a property that contains a
degree inside the denotation of whatever it modifies. Therefore, the denotation of ‘to
preoccupy much’ is the one in (44).
(44)
Pxd [much (d)  R(d)(x)] (xz[ z  [x[preoccupy(d)(x)]]])
and therefore
xdz[much (d)  z  [preoccupy(d)(x)]]
This explains that K-states allow I-readings of the adverbial quantifiers. A flexible Dstate is built on top of a K-state, and denotes the event of staying in a K-state, while a
Stubborn D-state denotes simply an event, as in (41). A stubborn D-state directly
combines with a root; say, for instance, sit.
(45)
xe[sit(e)  HOLDS(e, x)]
If the quantifier tries to get a I-reading, the result is non semantically computable,
because the quantifier presupposes in this reading the existence of a degree (d) and this
is not provided by the predicate.
Consider, in contrast, how a flexible D-state is built. Its denotation is that there is an
event, and that event is the one of holding a property which is instantiated as a K-state.
(46)
Pxdze[P(e, z)  HOLDS(e, x)  z  [P(x)(d)]] (glow)
xdze[ glow(e, z)  HOLDS(e, x)  z  [glow(d)(x)]]
12
JeNom 4, 4èmes Journées d’étude sur les nominalisations
University of Stuttgart, 16-17 June 2011
Here, the predicate contains a degree variable that can be bound by the adverb to get the
I-reading:
(47)
Pxdz[much (d)  P(d)(x)] (exdz[ glow(e, z)  HOLDS(e, x)
 z  [glow(d)(x)]])
and therefore
exdz[much (d)  glow(e)  HOLDS(e, x)  z  [glow(d)(x)]]]
6.3 A syntactic analysis
In a syntactic analysis that decomposes words in phrases, this process can be formalized
in the following way. The ‘pure’, non decomposable P are roots (in the sense of
Marantz 1997 and other DM approaches). The properties with a degree value are
adjectival-like projections that add an argument to the root. The K-states are StateP (in
the sense of Ramchand 2008), and to obtain a D-state from them they are plugged under
a ProcessP head; unlike what is claimed by Ramchand (2008), we must assume that the
configuration does not automatically impose a cause-effect relation between the ProcP
and the StateP, because in that case we would obtain a change of state reading, while in
the D-states we want to obtain a simultaneous reading (the process lasts as long as the
state lasts, or, in other words, the process is to keep x in a particular state). From this
perspective, the distinction between the stubborn D-states and the flexible D-states
would be that the former do not contain a StateP and an AP inside the structure, and
select the root directly in ProcP. In contrast, a flexible D-state contains all of them.
(48)
sit:
[ProcP
glow: [ProcP
Semantic:
mapping
[P(e)(x)
[StateP


[z
[AP
[SIT]]
[GLOW]]]]
[P(d)(x)
[Concept]]]]
As DegreeP is contained in AP, the first verb lacks this component. The structure of a
telic change of state with a result state would be the same as the one of glow, but the
relation between ProcP and StateP would be interpreted as temporal succession, not as
simultaneity.
(49)
interrupt: [ProcP
[StateP
[AP
[INTERRUPT]]]]
Compare with a telic verb without a State component:
(50)
destroy: [ProcP
[DESTROY]]
Now, the fact that a state nominalization always expresses a K-state means that it selects
a StateP, ignoring the higher material, as in (51). This motivates that the event,
contained in ProcP, is not included in this class of nominalizations.
(51)
[NP
[StateP]]
13
JeNom 4, 4èmes Journées d’étude sur les nominalisations
University of Stuttgart, 16-17 June 2011
Crucially, this head is present in a telic verb with a State component, or in a flexible Dstate, but not in the other cases. Consequently, only the first two cases will have a state
nominalization.
7 Conclusions and consequences
(52)
a.
destruc-ción
destroy-nom ‘destruction’
b.
preocupa-ción
worry-nom ‘preoccupation’
c.
modera-ción
moderate-nom ‘temperance’
(53) [ProcessP
[StateP
[PredicationP ]]]
(54) [P(e)(x) ∧
[z
≈
[P(d)(x)
]]]
(55)
Event noun: [ProcP…
State noun:
[StateP
Quality noun:
[√]]
[PredP [√]]]
[PredP [√]]
a) A state nominalization builds on top of StateP, already a temporal object, but
without an event. If this projection is not available in the structure, the state
nominalization is unavailable.
b) A quality nominalization builds on top of PredP, not a temporal object and
without an event. If this projection is unavailable in the structure, the quality
nominalization is unavailable.
c) An event nominalization builds on top of ProcP, a temporal object with an
event. If ProcP is not present in the structure, the event nominalization is
unavailable.
ProcP
Proc
<---- Selected by NP for an event nominalization
StateP <---- Selected by NP for a state nominalization
State
PredP <--- Selected by NP for a quality nominalization
ROOT
Pred
See now the same tree considered in its semantic denotation:
Event
Event
<---- For an event nominalization
K-state
K-state
<---- For a state nominalization
Property <--- For a quality nominalization
Property
CONCEPT
Why is the ProcP unavailable for the nominalization when it denotes a D-state?
14
JeNom 4, 4èmes Journées d’étude sur les nominalisations
University of Stuttgart, 16-17 June 2011
Some speculations:
a) A dynamic event that meets the subinterval condition as a state is, somehow,
not a prototypical event. This kind of hybrid is possible in the verbal domain,
where aspectual relations are a constitutive part of the denotation of the
predicate, but not in the nominal domain.
b) Perhaps the ProcP that introduces the D-state is different from the ProcP that
introduces a ’normal’ dynamic event. For instance, the state is interpreted as
a result state with respect to the event in the dynamic events, but in a D-state
the state has to be interpreted as simultaneous and coextensive with the
process, not as a result of it. Perhaps the ProcP that gives the simultaneous
semantics with StateP is not selected by the NP, but the ’standard’ one that
gives a result semantics is.
References
Arad, Maya. 1998. Psych-notes. UCL Working Papers in Linguistics 10.
Chomsky, Noam. 1970. Remarks on nominalization. In R. A. Jacobs & P. S.
Rosenbaum (eds.), Readings in English Transformational Grammar, Waltham,
Ginn and Co., pp. 232-286.
Dowty, David R. 1979. Word Meaning and Montague Grammar. The Semantics of
Verbs and Times in Generative Semantics and Montague’s PTQ. Dordrecht:
Reider.
Fábregas, Antonio, Rafael Marín & Louise McNally. In Press. From Psych Verbs to
Nouns. In V. Demonte & L. McNally (eds.), Telicity and change of state in
natural languages: implications for event structure. Oxford, Oxford University
Press.
Grimshaw, Jane. 1990. Argument Structures. Cambridge (Mass.), MIT Press.
Kratzer, Angelika. 2000. Building statives. Berkeley Linguistic Society 26, 385-399.
Krifka, Manfred. 1989. Nominalreferenz und Zeitkonstitution: Zur Semantik von
Massentermen, Pluraltermen und Aktionsarten. München: Wilhelm Fink.
Maienborn, Claudia. 2003. Die logische Form von Kopula-Sätzen. Berlin: AkademieVerlag.
Maienborn, Claudia. 2005. On the limits of the Davidsonian approach: The case of
copula sentences. Theoretical linguistics 31(3), 275-316.
Marín, Rafael & Louise McNally. 2011. Inchoativity, change of state, and telicity.
Natural Language and Linguistic Theory.
Martin, Fabienne. 2009. Relative stupidity and past tenses. Ms., Universität Stuttgart.
Mufwene, Saliloko. 1984. Stativity and the progressive. Bloomington: Indiana
University Linguistics Club.
Padučeva, Elena Viktorovna. 1996. Semanticheskie issledovanie. Moscow: Jazyki
russkoj kul’tury [Quoted from Spencer & Zaretskaya 2003]
Parsons, T. 1990. Events in the semantics of English. Cambridge (MA): MIT Press.
Pesetsky, David. 1995. Zero syntax: Experiencers and Cascades. Cambridge (MA),
MIT Press.
Ramchand, Gillian. 2008. First Phase Syntax. Oxford, Oxford University Press.
Rothmayr, Antonia. 2009. The structure of stative verbs. Amsterdam: John Benjamins.
Sebastián, Nuria, Fernando Cuetos, María Antonia Martí & Manuel F. Carreiras. 2000.
LexEsp. Léxico informatizado del español. Edición en CD-Rom. Barcelona:
Edicions de la Universitat de Barcelona.
15
JeNom 4, 4èmes Journées d’étude sur les nominalisations
University of Stuttgart, 16-17 June 2011
Spencer, Andrew & Marina Zaretskaya. 2003. Stative predicates in Russian and their
nominalizations. Essex Research Reports in Linguistics 22.
Williams, Edwin. 2007. Dumping Lexicalism. In Gillian Ramchand & Charles Reiss
(eds.), The Oxford Handbook of LInguistic Interfaces. Oxford: Oxford University
Press, 353-382.
16

Documentos relacionados