Handout - Genericity: Interpretation and Uses

Transcription

GENIUS II
Workshop on Dispositions, Abilities and States
ENS, Institut Jean Nicod
Paris, June 24-25, 2010
Genericity and stativity inheritance in “agent” deverbal nominals
R-nominals:
[DP/NP N [-ation] [XP->VP form ] ]
If deverbal nominals of participants in the event(uality) (agents, patients, holders, etc.) are
built syntactically in a comparable fashion (cf., Alexiadou & Schäfer 2010, Schäfer 2010),
they rise an interesting question:
o since they nominalize an argument, they can only be built from a structure akin
to the one in (5) (as opposed to (6)); thus involving full AS
o if built from a structure akin to (5), they should always give rise to an eventive
interpretation (correlating with full AS)
•
Deverbal nominalizations and clauses
(2)
Grimshaw (1990):
AS-nominals
event reading
obligatory arguments
compatible with aspectual
modifiers like in three hours
constant, frequent with
the singular
by-phrase is an argument
v.
(6)
[email protected]
[email protected]
a. the destruction of the city by the enemy
b. the enemy destroyed the city
iv.
AS-nominals:
[DP/NP N [AspP [AspE [-ation] [XP->VP form ] ] ] ] ]
!
(1)
i.
ii.
iii.
(5)
Paris, June 24-25, 2010
Isabelle Roy & Elena Soare
UMR 7023 Paris 8 / CNRS
1 Introduction
•
GENIUS II
Naming participants in the event(uality)
This is not the case – e.g., English –er nominals can have an eventive reading, along with a
non-eventive or dispositional reading:
(7)
R-nominals
no event
arguments not obligatory
not compatible with
aspectual modifiers
constant, frequent possible
only with the plural
by-phrase is not an argument
the saver of lives
;
the mower of the lawn ;
the seller of the horse ;
the life-saver
the lawn-mower
the seller
Alternatively, there exist AS-nominals and R-nominals of participants – the latter do not
involve any event (cf., above).
The crucial issue is what we believe dispositional nominals to be:
are they eventive? Or non eventive?
do they involve an AS? Or not?
•
(3) AS-nominals
a. the examination *(of the students) in three hours
b. the examination *(of the students) by the professor
c. the frequent examination *(of the students)
Goals
Show on the basis of French that “agent” nominals can be integrated in a unified theory of
deverbal nominalizations once we accept:
i) that nominalizations may inherit stativity and genericity from the base eventuality
provided by the underlying verbal phrase
ii) dispositional deverbal nominals are eventive (although not episodic) nominals.
(4) R-nominals
a. the examination / exam was on the table
b. the exam (*in three hours)
c. the frequent exams
This work is compatible with recent findings at the interface between syntax and semantics in
the domain of derived nominals that strongly supports the idea that derived nominals inherit
fine aspectual distinctions from their verbal base. This is also true of the generic character of
the underlying eventuality, we argue.
Syntactic approach to word formation
The internal structure of (complex) words is build syntactically, i.e. using the mechanisms
also found at the level of phrase structure; cf. Halle and Marantz (1993); Marantz (1997);
Alexiadou (2001); Borer (2003, 2005).
Correlation between obligatory presence of an argument structure and event interpretation: cf.
Borer (2001, 2003), see also Alexiadou (2001):
Roy & Soare
1
Roy & Soare
2
GENIUS II
Paris, June 24-25, 2010
2 –Eur nominals
•
! The external argument generalization
correct generalization: –eur/-er nominals denote participants with quite different theta-roles,
as long as they are assigned to the external argument (Fabb 1984, Keyser & Roeper 1984,
Rappaport-Hovav & Levin 1992, Van Hout & Roeper 1998)
(8)
a. … he is a teacher
b. … is a great defuser of pent-up emotions
c. ... a holder of a Visa or Master cart
d. ...as a dazzled admirer of Washington
e. this is a grinder
f. he is a receiver of compliments
(9)
a. … un mangeur de pommes
b. … un arracheur de maïs
c. … un détenteur de ressources pétrolières
d. … un admirateur de Napoléon
e. … un compteur de visiteurs gratuit pour votre site
(agent)
(causer)
(holder)
(experiencer)
(instrument)
(beneficiary)
•
Instrumental –eur/-er nominals are the only class on non-eventive nominals of
participants. In particular, they differ from dispositional nominals (e.g., life-saver,
sauveteur) in not accepting, in addition to (14), modification by adjectives like e.g., gros
‘big’, vieux ‘old’ and heureux ‘happy’ in their event-oriented reading (Larson 1998):
Combining the two tests (frequent-modifiers and gros-modifiers), a third class of
deverbal –eur nominals must be distinguished, that are compatible with both modifiers:
(17) Episodic Ns :
a. un consommateur fréquent de plusieurs drogues douces
a frequent user of many soft drugs
b. un heureux/gros consommateur de plusieurs drogues douces
a happy/big user of many soft drugs
•
2.1 Event and non-event reading
As commonly accepted in the literature –er nominals fall into two classes [± event]
depending on whether or not they have an eventive interpretation (Rappaport & Levin
1992, Van Hout & Roeper 1998 among many others).
the life-saver
the lawn-mower
the seller
Rappaport & Levin 1992 correlation: an –er nominal has an AS iff it has an eventive
interpretation [+event]:
(12) a. the constant defender *(of the government's policies)
b. a frequent consumer *(of tobacco)
Note: Difference between complex-event nominals and [+event] –er nominals: the latter do not allow adverbial
modification, for instance (cf., Greek, from Alexiadou 2001):
a. *i damastes ton fotonion mesa se /gia enan eona
the tamers the-GEN photons within / for a century
b. *o katharistis tu ktiriu
epi ena mina telika apolithike
the cleaner the-GEN building for a month finally got fired
Roy & Soare
•
(16) Dispositional N:
a. # un consommateur fréquent ; #un vendeur fréquent de voitures
a fréquent consummer
a frequent seller of cars
b. Les gros consommateurs/vendeurs font tourner la machine économique.
The big consummers / sellers run the economy
(agent)
(cause)
(holder)
(experiencer)
(instrument)
(10) a. un mangeur 'an eater',
un coureur 'a runner'
b. *un alleur 'a goer',
*un veneur 'comer'
*un arriveur 'an arriver'
c. *a faller ; *an arriver
(13)
For Rappaport & Levin, the instrumental (hence inanimate) –er nominals lack an
eventive reading, and an AS.
(15) Instrument N:
a. # un broyeur frequent
a frequent grinder
b. # un gros broyeur (with the event-oriented reading, meaning ‘which grinds much’)
a big grinder
Accordingly, -eur/-er never take unaccusative bases:
(11) the saver of lives
;
the mower of the lawn ;
the seller of the horse;
Paris, June 24-25, 2010
(14) *This machine continues to be our only frequent transmitter
Traditionally French –eur and English -er nominalizations are called “agentive nominals”.
•
GENIUS II
3
The event-oriented reading of gros ‘big’, vieux ‘old’, heureux ‘happy’ does not require
that the arguments be overtly realized:
(18) a. un gros mangeur
a big eater
‘someone who is big’/ ‘someone who eats much’
b. un vieil admirateur
an old admirer
‘someone who is old’ / ‘someone who has admired for a long time’
(19) #un heureux tableau; #un gros aimant
a happy painting
a big magnet
(20) a. #un vieux portier
an old door.man (lit. door.ierN)
b. #un gros exemplier
a big handout (lit. example.ierN)
2.2 Episodic and dispositional Ns
! We accept that two eventive readings for -eur nominals must be distinguished,
namely episodic and dispositional (cf., Alexiadou & Schäfer 2010). However,
contrar Alexiadou & Schäfer, dispositional Ns and instrument Ns do not fall into
the same class.
Roy & Soare
4
GENIUS II
•
Paris, June 24-25, 2010
We call episodic nominals, those that involve a particular event; and dispositional
nominals those that do not.
(21) Le dresseur de Simba
the trainer of Simba (= the one who trains Simba)
a particular event of training must have taken place.
>
>
> non specific DP
cette / la voiture
des voitures
américaines
des voitures
voiture
cars
car
>
> generic event
particular episodic
event >
Similar contrast is found with non-durative verbs:
(23) Le vendeur de ce tapis aux puces
the seller of this carpet at the flea market
(24) Le vendeur de tapis
the seller of carpets
frequent-modifiers mark an underlying episodic event, as opposed to a disposition (and
not an underlying event altogether).
2.3 The source of the episodic vs. dispositional readings
! The correlation between AS/event (Grimshaw 1990; Rappaport Levin 1992) must be
restated for –eur nominals in terms of a correlation between ‘specific arguments’ and
‘episodic’ reading.
! In the light of (21)-(22), it is not the presence/absence of AS altogether but the nature
of the arguments that matters in determining the episodic/dispositional reading (contra
Alexiadou & Schäfer).
Episodic and dispositional –eur nominals are both AS-nominals (as they are eventive).
Their respective AS is of a different nature, however.
" Specific arguments (definite, demonstrative, etc.) correlate with the episodic
reading for the derived nominal;
" Non-specific arguments (bare singulars, indefinite plurals, etc.) correlate with the
dispositional reading.
•
specific DP >
american cars
no event of lion training must have taken place.
•
Paris, June 24-25, 2010
(25)
this/the car
(22) Le dresseur de lion(s)
the trainer of lion(s) (= the one who has the disposition to train lions.)
•
GENIUS II
The degree of (non)-specificity in the AS matches the degree of episodicity vs. genericity
in the underlying eventuality. This, in turn leads to a gradation in episodicity for the
related nominal.
>
[vendre cette voiture / [vendre des voitures [vendre des voitures]
la voiture]
américaines]
sell cars
sell this /the car
sell american cars
[vendre de la voiture]
episodic nominals >
dispositional
nominals
>
>
le vendeur de cette le vendeur de voitures le vendeur de voitures
voiture / de la voiture américaines
the seller of cars
rouge
the seller of american
the seller of this car / cars
of the red car
•
le vendeur de voiture
the seller of car
In English, only non-specific arguments can undergo incorporation:
(26) a. the seller of this car
b. the seller of cars
c. the seller of big red cars
d. the seller of american cars
e. *the seller of car
•
sell car
*the this-car seller
*the cars-seller
*the big red cars seller
the american-car seller
the car-seller
frequent-modifiers and specificity of the arguments:
(27) a. # un vendeur fréquent de cette voiture
a frequent seller of this car
b. un vendeur fréquent de voitures américaines
a frequent seller of american cars
c. ?? un vendeur fréquent de voitures
a frequent seller of cars
d. *un vendeur fréquent de voiture
a frequent car-seller
(28) a. un consommateur fréquent de cette drogue
a frequent consumer of this drug
b. un consommateur fréquent de drogues douces
a frequent consumer of light drugs
c. ? un consommateur fréquent de drogues
d. *un consommateur fréquent de drogue
a frequent consumer of drug(s)
Roy & Soare
5
Roy & Soare
6
GENIUS II
•
Paris, June 24-25, 2010
GENIUS II
3.1 Agents
In the absence of overt AS, and therefore of any source of specificity, we expect the
eventive nominals to receive a dispositional reading:
(29) a. le vendeur; le sauveteur, le chercheur
! A second strong generalization: –ant nominal are unable to express agents.
(36) a. *un enseignant délibéré, *un accompagnant volontaire, *un attaquant volontaire
'a deliberate teacher', 'an voluntary accompanying person', 'a voluntary attacker'
b. the seller, the teacher, the driver
However, this is a property that -ant nominals share with dispositional –eur nominals.
This is so, we believe, because dispositionals inherit the stative nature of the underlying
generic eventuality. As a result, they are interpreted as holders rather than agents.
3 –ANT NOMINALS
(37) a. un défenseur délibéré /volontaire *(des opprimés)
a defender deliberate / voluntary of.the oppressed
b. l'accusateur obstiné *(de cette jeune femme)
the accuser stubborn of this young lady
! Some tricky pairs
(30) a. attaquant 'attacker' vs. défenseur 'defender'
b. enseignant 'teacher' vs. chercheur 'researcher'
c. accompagnant 'accompanying person' vs. accompagnateur 'guide'
d. soignant ‘carer’ vs. guérisseur ‘healer’
(31) a. accompagnateur ‘guide’
b. serveur / serviteur ‘waiter / maid’
c. suiveur / successeur ‘follower / successor’
d. exécuteur ‘executor’
e. débiteur ‘debtor’
Only episodic French –eur nominals verify tests identifying agentivity: adjectival
modification with delibéré, volontaire, intentionnel, obstiné
(38) *l’attaquant délibéré vs. l’agresseur délibéré de la jeune femme
accompagnant ‘accompanying person’
servant ‘servant’
suivant ‘following/follower’
exécutant testamentaire ‘executing’
vs. débitant (de tabac) ‘tobacconist’
the accuser deliberate vs. the agressor deliberate of the young lady
! Generalization of the subject – not the external argument : unaccusative bases
available
(32) a. un habitant ‘an inhabitant’ vs. *un habiteur ‘an inhabiter’
b. un arrivant ‘an arrivant’ vs. *un arriveur ‘an arriver’
(33) a. un soignant, un remplaçant, un servant
(Holder)
‘a carer’, ‘a substitute’, ‘a servant’
b. un mourant, un accompagnant, un soupirant, plaignant
(Experiencer)
a dying person, an accompanying person, a suitor, a complainant
c. un arrivant, un accédant
(Theme)
a comer, a reacher
•
If the lack of agentivity is correlated to the dispositional reading. The question is now:
since–ant nominals are non-agentive, are they always dispositional?
•
The question arises for eventive –ant nominals only, to the exclusion of
instruments/’product’ nominals. These are referential (inanimate) nominals, which have
no underlying eventuality (cf. also instrument –eur nominals).
(39) a. #un vieux tranquillisant ‘an old tranquillizer’
b.#un gros amincissant ‘a big slimming cream’
c. #le petit détachant ‘the little stain remover’
d. #un heureux cicatrisant ‘a happy healing cream’
•
Frequent-constant modification (cf (40)) is not available in –ant nominals, while eventoriented adjectives are possible (cf. (41)). With respect to these two tests, eventive -ant
nominals pattern with dispositional –eur nominals.
(40) a. *un militant fréquent contre la peine de mort
‘a frequent campaigner against the death sentence’
b. *le gagnant fréquent du gros lot
‘the frequent winner of the jackpot’
c. *un aidant constant des malades de l’aile A
‘a constant helper of the patients from wing A’
(34) stative bases, perception verbs :
un voyant 'a psychic' vs. #voyeur (idiosyncratic only)
un apprenant 'learner' vs *appreneur
un savant 'scientist' vs. *saveur
(35) may be derived from verbs with middle reflexive se:
a. se défendre > défendant
vs. défendre > défenseur
‘to defend oneself > defendant’ ‘to defend > defender’
b. se plaindre > plaignant vs. *plaigneur
‘to complain > complainant’ vs. ‘complainer’
Roy & Soare
Paris, June 24-25, 2010
(41) a. les (bien)heureux militants pour la paix
‘the happy campaigners for the peace = who are happy to fight for peace’
b. un vieux gagnant du loto
‘an old winner of the jackpot = someone who won the jackpot for a long time’
c. les petits publiants
‘the small publishing fellows = the ones who publish little work’
7
Roy & Soare
8
GENIUS II
Paris, June 24-25, 2010
GENIUS II
Paris, June 24-25, 2010
b. Parlant d’une voix forte, il a réveillé le bébé.
speaking with a loud voice, he woke up the baby
3.2 AS and the type of event
•
Productive new creations of –ant nominals are argument-less :
(42) les écoutants ‘the listeners’, le discutant ‘the commentator’, un aidant ‘a helper’, les
publiants ‘the ones who publish’
•
Nevertheless, AS can be inherited, and when it is, an episodic reading is obtained in the
presence of specific arguments.
(43) a. le gagnant du gros lot
‘the winner of the jackpot’
b. les militants contre les mesures de Sarkozy
‘the campaigners against Sarkozy’s policies’
•
And a dispositional meaning is obtained in the presence of non-specific arguments:
•
We conclude that–ant nominals are not always dispositional. When formed from a true
activity verb, what sort of participant they denote needs to be explained, since the agent
from the underlying eventuality seems to be lost.
3.3 –ant nominals as stative
•
–ant nominals involve a null nominalizing head realizing stative Aspect, assuming as in
Borer (2005) that nominalizing suffixes realize Aspect.
(51)
(44) a. l’accédant à la propriété
‘the home-buyer’
b. un militant contre la peine de mort
‘a campaigner against the death sentence’
•
(48) a. Enseignant le français, il en connaît bien les subtilités.
teaching French, he knows all its subtleties
b. Enseignant mon cours de syntaxe cet après-midi, je ne pourrai pas venir à la réunion.
teaching my syntax class this afternoon, I will not be able to come to the meeting
(49) *la personne habitante ‘the inhabiting person’
(50) a. le bien aimé ‘the best-beloved’, le très miséricordieux ‘the all-merciful’
b. *le très-publiant ‘someone who publishes much’, *le bien-aimant ‘someone who
loves much’
We take the absence of agentivity in –ant nominals to be connected to their underlying
stativity, cf. the imperative test, adapted from Lakoff (1966)
(45) a. kiss Lynn!
[ n [ Asp-S ØN [ VP perdant]]]
[ n [ Asp-S ØN [ VP gagnant du gros lot]]]
'looser'
'winner of the jackpot'
For –eur nominals, the situation is rather different. We argue that there is only one –eur
suffix, which realizes eventive Aspect (highlighted by the agentivity of –eur nominals).
Here, arguments are realized outside of the VP, and for that reason can contribute to the
interpretation of the entire construal (episodic or generic).
(52) le conducteur du train de 19h30 'the driver of the 7:30 train '
[ n [ Asp-Ev -eurN [ Asp-Q le train de 19h30 [ VP conduct- ]]]
EPIS
(53) un chercheur 'a researcher'
[ n [ Asp-Ev -eurN [ Asp-Q
[ VP cherche- ]]]
DISP
[ VP
DISP
(54) un éleveur de chiens 'a dog-breeder'
[ n [ Asp-Ev -eurN [ Asp-Q chiens
4
Ø
élev- ]]]
Internal arguments
b. *love Lynn!
(46) a. sois l’accompagnateur des enfants!
'be
the guide
of the children '
b. *sois l’accompagnant (des enfants) !
'be the accompanying person (for the children)!'
•
(55) le marié, un immigré, l’arrondi, un insoumis, un imprévu, un malvenu, un pendu
•
The hypothesis of –ant nominals as stative allows for a better understanding their
properties: the type of Th-role they are linked to (cf 33), and the specific bases they allow
for. The event-oriented adjectives (cf. 41) point to the presence of an eventuality, which in
this case is a state. We speculate that the incompatibility with frequent and constant
modifiers (cf. 40) is also a result of the inherent stativity of –ant nominals.
! Remaining problem: the episodic reading with specific arguments.
•
–ant nominals are derived from present participles (as opposed to the alternative,
conceivable in French, of an adjectival source, cf. (49-50)). Present participle’s default
value: imperfective – durative aspect (with an active or stative meaning).
(47) a. Parlant d’une voix forte, il n’a pas besoin de micro.
speaking with a loud voice, he needs no microphone
Roy & Soare
9
é/u nominals by default realize AS: they nominalize the internal argument
(56) le détenu, un invité
(57) les nouveaux venus, un insoumis
(Patient)
(Thème)
Since arguments can be specific and non-specific, and in our system this leads to episodic and
dispositional reading respectively, we expect that internal argument nominals equally allow
those readings.
• Generalization: é/u nominals systematically lack a dispositional reading
(58) un blessé = a wounded man and not someone that has the disposition to be wounded
une mariée = the bride and not a woman who has the disposition to be married
trois inconnus = three individuals that I didn’t know at the moment, etc.
Roy & Soare
10
GENIUS II
Paris, June 24-25, 2010
This can be explained if we accept that only specific arguments can be nominalized,
leading necessarily to an episodic reading.
(59) [ n [ Asp-Ev [ Asp-Q –é/uN [ VP
mari- ]]]
•
5 Conclusions
•
There is a tight connection between argument structure and event interpretation, which
goes beyond what has been suggested for event nominals (e.g., Grimshaw’s 1990 wellknown correlation): the nature (specific or non-specific) of the argument structure is
important in building the episodic vs. dispositional meaning in names of participants.
•
A three-way classification of names of participants, in terms of episodic, dispositional and
referential, is more accurate than a bi-partite one, either in terms of [animate-eventive] as
in Rappaport Hovav and Levin 1992 or in terms of [episodic-dispositional] as in
Alexiadou & Schäfer 2010.
•
On the basis of differences between –eur and –ant nominals in French, we have shown
that genericity can be inherited in nominalizations, leading to a stative-like
interpretation.
•
This inner genericity is to be distinguished from an outer genericity brought in by the
Determiner system. However, the interaction between the two types of genericity needs to
be better understood.
•
So must the interaction between inner genericity and the semantics of the underlying
predicate.
GENIUS II
Paris, June 24-25, 2010
22, 2001), Booij G., J. de Cesaris, S. Scalise, A. Ralli (eds). 177-196. Barcelona: IULAUniversitat Pompeu Fabra
Rappaport Hovav, M. and B. Levin. 1992. ``--er Nominals: Implications for a Theory of
Argument Structure'', in T. Stowell and E. Wehrli, eds., Syntax and Semantics 26:
Syntax and the Lexicon, Academic Press, New York, NY, 127-153.
Van Hout A. and Roeper, T. 1998. “Events and Aspectual Structure in Derivational
Morphology”. In Heidi Harley, Papers from the Upenn/MIT Roundtable on Argument
Structure and Aspect, vol. 32. MIT Papers in Linguistics, Cambridge, Mass, pages 175200.
Larson, R.K. 1998. Events and modification in nominals. In D. Strolovitch and A. Lawson
(eds.) Proceedings from Semantics and Linguistic Theory (SALT) VIII. Cornell
University, Ithaca, NY
McIntyre, A., 2009, "Argument inheritance and category-changing morphology', talk given at
Generative Grammatik des Südens, University of Leipzig.
Kratzer, A. 1998. Aspect in Adjectival Passives, Paper presented at WCCFL 17,
February 20, 1998. University of British Columbia at Vancouver.
Lakoff, G. 1966 Stative adjectives and verbs in English. Mathematical linguistics and
automatic translation; report to the National Science Foundation 17, Computational
Laboratory, Harvard University.
References
Alexiadou, A. 2001. Functional structure in nominals: nominalization and ergativity. John
Benjamins.
Alexiadou, A. & Schäffer. 2010. On the syntax of episodical vs. dispositional -er nominals. In
A. Alexiadou & M. Rathert (eds.) Nominalizations across languages and frameworks.
Berlin: Mouton de Gruyter.
Anscombre J-C. 2001. A propos des mécanismes sémantiques de formation des noms d'agent
en français et en espagnol. Langages 143, pp. 28-48.
Borer, H. 1999. The form, the forming and the formation of nominals. paper presented at the
2nd Mediterranean Morphology Meeting, September 1999.
Borer, H. 2003. Exo-skeletal vs. endo-skeletal explanations. In Moore, J. and M. Polinsky
(eds). The Nature of Explanation in Linguistic Theory. Chicago: CSLI and University of
Chicago Press.
Borer, H. 2005. The Normal Course of Events. Structuring Sense, Volume II. Oxford: Oxford
University Press.
Borer, H. and I. Roy. 2007. The name of the adjective. To Appear in P. Hoffher Cabredo and
O. Matushansky, eds., Adjectives. John Benjamins.
Embick, D. 2003. Locality, listedness and morphological identity. Studia Linguistica 57, 143169.
Fradin, B. & F. Kerleroux. 2003. Troubles with Lexemes. In Topics in Morphology. Selected
papers from the Third Mediterranean Morphology Meeting (Barcelona, September 20-
Roy & Soare
11
Roy & Soare
12

Handout - Genericity: Interpretation and Uses

Transcription

Documents pareils

How to buy a used car

Introduction - Presses Universitaires de Vincennes

washington county - Wisconsin Forty and Eight

Drain Nut Wrench

Domaine Viticole du Haut Bourg www.hautbourg.fr Gros Plant du

Guitare BASSE " Lag Jacobacci 400 Series"

Fact sheet of purchase

Family Transfer of Designated Property

Fact sheet of purchase