Transcript clairvoyance-tutoria.. - Carnegie Mellon School of Computer Science

Lexical Semantics in
American Corpus
Annotation Projects
Lori Levin
September 10, 2004
Tutorial at Clairvoyance
Corporation
What is Lexical Semantics?


Lexical semantics is about the meanings of
words.
This tutorial is about the meanings of verbs and
their arguments:






Sam opened the door with a key.
They key opened the door.
The door was opened by Sam with a key.
The door opened (with a key).
Sam bought a book from Sue.
Sue sold a book to Sam.
Types of semantics not covered in
this tutorial

Sentence-level meaning

Truth conditions of sentences



Compositional semantics


This is a picture of a cell phone. (true)
This is a picture of a book. (false)
How the meanings of a noun phrase and a verb
phrase are combined into the meaning of a
sentence.
Quantifier scope.

Everyone here speaks two languages.
Aspects of lexical semantics not
covered in this tutorial


Nouns, adjectives, adverbs, and prepositions
Selectional restrictions:

Colorless green ideas sleep furiously.


Count and mass nouns:



Chomsky, 1957, Syntactic Structures
There was water all over the driveway. (mass)
There was dog all over the driveway.
(count)
Synonymy, hyponymy, antonymy, etc.



car-automobil
car-vehicle
Hot-cold
Outline

Background





Predicates and Arguments
Valency and subcategories of verb
Optional arguments and adjuncts
Semantic Roles
Three approaches to lexical semantics

A linguistic theory


A lexicon project


Frame Semantics
A corpus annotation project (also building a lexicon)


Lexical Conceptual Structure
PropBank
A multi-lingual semantic corpus annotation
project
Predicates and Arguments

Verbs (and sometimes nouns and adjectives)
describe events, states, and relations that have
a certain number of participants.

The children devoured the spaghetti.


The teacher handed the book to the student.


Three particpants.
Problems exist.


Two participants
One participant.
The participants are referred to as arguments
of the verb. (Like arguments of a function.)
Valency and Subcategorization

Fillmore and Kay, Lecture Notes, Chapter 4:








The children devoured the spaghetti.
*The children devoured.
*The children devoured the spaghetti the cheese.
She handed the baby a toy.
*She handed the baby.
*She handed the toy.
Problems exist.
*Problems exist more problems.
Grammaticality
An asterisk (*) indicates that a sentence
is ungrammatical.
 A large percentage of linguists make
these assumptions:


Human languages are like formal languages.


Some sentences are in the set of legal sentences
and some are not
A human can act like a machine that accepts
legal sentences and rejects illegal sentences.
Valency

The number of participants is called the verb’s
valence or valency.




Devour has a valency of two.
Hand has a valency of three.
Exist has a valency of one.
Linguists took this term from chemistry – how
many electrons are missing from the outer shell.

The first linguist to use the term was Charles Hockett in
the 1950’s.
Subcategorization


Verbs are divided into subcategories that have
different valencies.
Here is how the terminology works:

Exist, devour, and hand have different
subcategorizations




i.e., They are in different subcategories
Devour subcategorizes for a subject and a direct object.
Devour is subcategorized for a subject and a direct
object.
Devour takes two arguments, a subject and a direct
object (or an agent and a patient).
Arguments are not always Noun Phrases

The italicized phrases are also arguments:

He looked very pale.


The solution turned red.


Verb Phrase
He started singing a song.


Adjective Phrase
I want to go.


Adjective Phrase
Verb Phrase
We drove to New York.

Prepositional Phrase
Optional and Obligatory Arguments

The direct object of eat is optional:



The children ate.
The children ate cake.
The direct object of devour is not optional:


*The children devoured.
The children devoured the cake.
Optional Arguments
The dog ran.
 The dog ran from the house to the
creek through the garden along the
path.

Optional vs. Invisible Arguments
a. What happened to the cake?
b. The children ate.
b’. The children ate it.
In English, Sentences b and b’ do not mean
the same thing in this context.
Compare to Japanese and Chinese.
Adjuncts

Locations, times, manners, and other
things that can go with almost any
sentences are called adjuncts.

The children ate the cake quickly at 2:00
in the kitchen.
How to tell arguments from adjuncts


There are some general guidelines that are not
always conclusive.
Adjuncts are always optional.


but some arguments are optional too
Repeatability:

The children devoured the cake at 2:00 on Monday.


The children devoured the cake in Pittsburgh in a
restaurant.


Two temporal adjuncts
Two locative adjuncts
*The children devoured the cake the dessert.

arguments are not repeatable
Semantic Roles: Motivation

The verb open appears in different
subcategorization patterns:





Sam opened the door with a key.
The key opened the door.
The door was opened by Sam with a key.
Sam’s opening of the door with a key
How can we represent the meanings of
these sentences in a way that shows that
they are related?
Semantic Roles: Motivation

These sentences do not have the same
meaning even though they have the same
verb:


Sam interviewed Sue.
Sue interviewed Sam.
Semantic Roles: Motivation
These sentences mean roughly the same
thing even though they use different
verbs:
Sam bought a toy from Sue.
 Sue sold a toy to Sam.

Semantic Roles: Motivation

The way to express riding a vehicle to a
location is different in different languages:






Sam took a bus to school.
Sam ascended to the bus and went to school.
(Hebrew)
Sam riding on the bus, went to school. (Japanese)
Sam sat on the bus, went to school. (Chinese)
Sam went to school by bus.
Sam went to school by taking a bus.
Semantic role names in a meaning
representation





Sam opened the door with a key.
The key opened the door.
The door was opened by Sam with a key.
Sam’s opening of the door with a key
Open



Agent: Sam
Patient: door
Instrument: key
Semantic Roles Names in a Meaning
Representation

These sentences do
not have the same
meaning:


Sam interviewed Sue.
Sue interviewed Sam.

Interview



Agent: Sam
Patient: Sue
Interview


Agent: Sue
Patient: Sam
Examples of Semantic Roles

Agent: an agent acts volitionally or
intentionally


The students worked.
Sue baked a cake.
Examples of Semantic Roles

Experiencer and Perceived:


An experiencer is an animate being that perceives
something, cognizes about something, or or
experiences an emotion.
The perceived is the thing that the experiencer
perceives or the thing that caused the emotional
response.

The students like linguistics.
 (emoter and perceived)

The students saw a linguist.
 (perceiver and perceived)

Linguistics frightens the students.
 (emoter and perceived)

The students thought about linguistics.
 (cognizer and perceived)
Examples of Semantic Roles

Patient: A patient is affected by an action.



Beneficiary: A beneficiary benefits from an event



My dog died on me.
Instrument:



Sue baked a cake for Sam.
Sue baked Sam a cake.
Malefactive: Someone is affected adversely by an event.


Sam kicked the ball.
Sue cut the cake.
The boy opened the door with a key.
The key opened the door.
Location:


The clock stands on the shelf.
I put the book on the shelf.
Three approaches to semantic roles in
meaning representations

Ray Jackendoff (1972, 1990)





Lexical Conceptual Structure
The Motion/Location Metaphor
Semantic Roles
Charles Fillmore, FrameNet Project

Linguistic Theory
Lexicon
Frame-semantics
Martha Palmer, PropBank Project Corpus Annotation


Predicate-specific role names:
Proto-grammatical relations
Ray Jackendoff
Semantic Interpretation in Generative
Grammar, MIT Press, 1972
 Semantic Structures, MIT Press, 1990.



Theory of human cognition
Used by many computational linguists
Lexical Conceptual Structure

Primitives:



Types of entities:


GO, BE, STAY, CAUSE, and several more
TO, FROM, AWAY, TOWARD, VIA, and several more
Event, State, Thing, Place, Path
Other tiers of representation are added in order to
capture nuances of meaning and grammar:



Cause and affectedness
Manner
Actor and undergoer (see discussion of PropBank)
Example of Lexical Conceptual
Structure
Sam threw the ball across the room.
[event
CAUSE [thing SAM]
[event GO
[thing BALL]
[path TO
[place AT
[thing
other-side-of-room]]]]]
Lexical Conceptual Structure and
Semantic Role Names
Sam threw the ball across the room.
[event
CAUSE [thing SAM] agent
[event GO
[thing BALL] theme
[path TO
[place AT
[thing
other-side-of-room]]]]]
goal
The Motion/Location Metaphor

J. S. Gruber, Studies in Lexical Relations,
MIT Dissertation, 1965.





Agent: causes, manipulates, affects
Theme: changes location, is located
somewhere, or exists
Source: the starting point of the motion
Goal: the ending point of the motion
Path: the path of the motion
Examples of Location and Directed
Motion
Many problems still exist.
 The clock sits on the shelf.
 The ball rolled from the door to the
window along the wall.
 Same walked from his house to town
along the river.
 Sue rolled across the room.
 The car turned into the driveway.

Being in a state or changing state
The car is red.
 The ice cream melted.
 The glass broke.
 Sam broke the glass.
 The paper turned from red to green.
 The fairy godmother turned the pumpkin
into a coach.

Having or Changing possession
The teacher gave books to the students.
 The teacher gave the students books.
 The students have books.

Exchange of Information
The teacher told a story to the students.
 The teacher told the students a story.

Extent
The road extends/runs along the river
from the school to the mall.
 The string reaches the wall.
 The string reaches across the room to the
wall.

Strong points of LCS and the
Motion/Location Metaphor

Sam manipulates a key, having an effect
on the door, causing it to go from the
state of being closed to the state of being
open.




Sam opened the door with a key.
The key opened the door.
The door was opened by Sam with a key.
Sam’s opening of the door with a key
Strong points of LCS and the
Motion/Location Metaphor
A toy goes from Sue to Sam. Some
money goes from Sam to Sue.
 Differences in the causation tier.



Sam bought a toy from Sue.
Sue sold a toy to Sam.
Strong points of LCS and the
Motion/Location Metaphor

Supports some inferences:


If X goes from A to B, then X is no longer at A.
If X is created (begins to BE) during event Y,
then X doesn’t exist until Y is finished.
Strong or weak point?

LCS wasn’t designed with this kind of thing in
mind, but it could be made to work.






Sam took a bus to school.
Sam ascended to the bus and went to school.
(Hebrew)
Sam riding on the bus, went to school. (Japanese)
Sam sat on the bus, went to school. (Chinese)
Sam went to school by bus.
Sam went to school by taking a bus.
Problem with Thematic Roles and the
Motion/Location Metaphor

It is not clear how to apply the metaphor
to many verbs (Fillmore and Kay, Lecture
Notes, pages 4-22)



He risked death.
We resisted the enemy.
She resembles her mother.
LCS Resources



Bonnie Dorr, University of Maryland
http://www.umiacs.umd.edu/~bonnie/LCS_Da
tabase_Documentation.html
LCS Lexicon for English




English word senses are mapped to WordNet
Handcrafted lexical entries for around 4000 verbs
Automatically produced entries may be available for
a full-sized lexicon
LCS Dictionaries for other languages may be
available

May be handcrafted or produced partially
automatically
Problem with Thematic Roles and the
Motion/Location Metaphor

It is not clear how to apply the metaphor
to many verbs (Fillmore and Kay, Lecture
Notes, pages 4-22)



He risked death.
We resisted the enemy.
She resembles her mother.
Charles Fillmore, Collin Baker, and others
FrameNet Project
http://www.icsi.berkeley.edu/~framenet/
 Frame semantics



Frames are networked using several relations
Based on corpus analysis
Lexical entries for around 7500 English
verbs
 Other FrameNet projects in



Spanish
Japanese
Advantage of Frame Semantics

FrameNet was designed to capture the
similarities in sentences like these.

Ride-vehicle frameSam took a bus to
school.





Sam ascended onto the bus and went to school.
(Hebrew)
Sam riding on the bus, went to school.
(Japanese)
Sam sat on the bus, went to school. (Chinese)
Sam went to school by bus.
Sam went to school by taking a bus.
Frame Semantics compared to the
Motion/Location Metaphor

Frame Semantics has


Many primitives
Many semantic roles
FrameNet strong and weak points
FrameNet is still under development and
may change frequently.
 Versions are clearly identified.
 Lexical entries are very carefully hand
crafted.

Martha Palmer and others
The PropBank Project
http://www.cis.upenn.edu/~ace/
 Annotate the Penn TreeBank with
predicate-argument information
 Corpus can be used for automatic learning
of the surface realization of each
argument

PropBank and FrameNet: Close ties

PropBank lexical entries are linked to
FrameNet entries.


This paper contains some comparisons of
PropBank and Framenet


There are more PropBank entries than
FrameNet entries
http://www.cis.upenn.edu/~dgildea/gildeaacl02.pdf
See also VerbNet

http://www.cis.upenn.edu/group/verbnet/
Proto-roles and verb-specific roles

http://www.cis.upenn.edu/~dgildea/Verbs/

Abandon

Arg0:abandoner

Arg1:thing abandoned, left behind

Arg2:attribute of arg1
PropBank:
multiple surface realizations of arguments








Sam opened the door with a key.
The key opened the door.
The door was opened by Sam with a key.
Sam’s opening of the door with a key
Arg0:opener
Arg1:thing opening
Arg2:instrument
Arg3:benefactive
Sam
door
key
PropBank:
How are lexical entries used by annotators?
Intercoder agreement is a high priority for
PropBank.
 Role names like agent and theme can be
confusing.
 Verb-specific role names are more clear.

Annotation Procedure
Identify the verb in a sentence.
 Look it up in the PropBank lexicon.
 Assign arg0…arg-n appropriately by
looking at the verb-specific roles.


Always use the same arg-n for the same verbspecific role.
What are the arg-n’s?
The arg-n labels are arbitrary labels.
 However, PropBank tries to use them
consistently across verbs.



Arg0 tends to be
most likely to be
Arg1 tends to be
thing most likely



an agent or the argument
the subject in active voice.
a theme or patient or the
to be
The direct object of a transitive verb in active voice
The subject of a verb in passive voice
The subject of an intransitive verb
PropBank was not designed for this
 Sam
took a bus to school.
 Sam ascended onto the bus and went to
school. (Hebrew)
 Sam riding on the bus, went to school.
(Japanese)
 Sam sat on the bus, went to school.
(Chinese)
 Sam went to school by bus.
 Sam went to school by taking a bus.
 But
it is linked to FrameNet
IAMTC (Interlingua Annotation of
Multilingual Text Corpora) Project
http://aitc.aitcnet.org/nsf/iamtc/
 Collaboration:







New Mexico State University
University of Maryland
Columbia University
MITRE
Carnegie Mellon University
ISI, University of Southern California
Goals of IAMTC

Interlingua design


Annotation methodology


Three levels of depth
manuals, tools, evaluations
Annotated multi-parallel texts


Foreign language original and multiple English
translations
Foreign languages: Arabic, French, Hindi,
Japanese, Korean, Spanish
Motivation for Corpus and Data

Examine the surface realization of many
phenomena

In one language: many surface realizations of the
same phenomenon



I think it is raining.
It is probably raining.
Across languages: different syntactic constructions are
used to express the same ideas
IL Development: Staged, deepening


IL0: simple dependency tree gives structure
IL1: semantic annotations for Nouns, Verbs,
Adjs, Advs, and Theta Roles








Not yet ‘semantic’—”buy”≠“sell’, many remaining
simplifications
Concept ‘senses’ from ISI’s Omega ontology
Theta Roles from Dorr’s LCS work
Elaborate annotation manuals
Tiamat annotation interface
Post-annotation reconciliation process and interface
Evaluation scores: annotator agreement
IL2: that comes next…
Details of English IL0
Deep syntactic dependency representation:





Removes auxiliary verbs, determiners, and some
function words
Normalizes passives, clefts, etc.
Removes strongly governed prepositions
Includes syntactic roles (Subj, Obj)
Construction:


Dependency parsed using Connexor (English)
–


Tapanainen and Jarvinen, 1997
Hand-corrected
Extensive manual and instructions on IAMTC
Wiki website
IL0 coding manuals for other languages:
Japanese
 Spanish
 Korean (in progress)
 Hindi (in progress)
 French (in progress)

Example of IL0
TrEd, Pajas, 1998
Sheikh Mohammed, who is also the
Defense Minister of the United Arab
Emirates, announced at the inauguration
ceremony that “we want to make Dubai
a new trading center”
Example of IL0

Sheikh Mohammed, who is also the Defens Minister of the
United Arab Emirates, announced at the inauguration
ceremony that “we want to make Dubai a new trading
center”
announced V Root
Mohamed PN Subj
Sheikh PN Mod
Defense_Minister PN Mod
who Pron Subj
also Adv Mod
of P Mod
UAE PN Obj
at P Mod
ceremony N Obj
inauguration N Mod
Dependency parser and Omega ontology
Omega (ISI):110,000 concepts
(WordNet, Mikrokosmos,
etc.), 1.1 mill instances
URL: http://omega.isi.edu
Dependency parser
(Prague)
Details of IL1

Intermediate semantic representation:

Annotations performed manually by each person alone




No treatment of prepositions, quantification, negation,
time, modality, idioms, proper names, NP-internal
structure…
Nodes may receive more than one concept


Associate open-class lexical items with Omega Ontology
items
Replace syntactic relations by one of approx. 20 semantic
(theta) roles (from Dorr), e.g., AGENT, THEME, GOAL,
INSTR…
Average: about 1.2
Manual under development; annotation tool built
Example of IL1
Sheikh Mohammed, who is also the
Defense Minister of the United Arab
Emirates, announced at the inauguration
ceremony that “we want to make Dubai
a new trading center”
Example of IL1: internal
representation
The study led them to ask the Czech government to recapitalize
CSA at this level.
[3,
[2,
[4,
[6,
lead, V, lead, Root, LEAD<GET, GUIDE]
study, N, study, AGENT, SURVEY<WORK, REPORT]
they, N, they, THEME, ---, ---]
ask, V, ask, PROPOSITION, ---, ---]
[9, government, N, government, GOAL, AUTHORITIES,
Semantic
GOVERNMENTAL-ORGANIZATION]
Roles
[8, Czech, Adj, Czech, MOD, CZECH~CZECHOSLOVAKIA, ---]
[11, recapitalize, V, recapitalize, PROP, CAPITALIZE<SUPPLY,
INVEST]
[12, csa, N, csa, THEME, AIRLINE<LINE, ---]
[16, at, P, value_at, GOAL, ---, ---]
[15, level, N, level, ---, DEGREE, MEASURE]
[14, this, Det, this, ---, ---, ---]
Concepts from the
Omega Ontology
Tiamat: annotation interface
For each new
sentence:
Step 1: find
Omega
concepts
for objects
and events
Step 2: select
event frame
(theta
roles)
Candidate concepts
Omega ontology


Single set of all semantic terms, taxonomized
and interconnected (http://omega.isi.edu )
Merger of existing ontologies and other
resources:








Manually built top structure from ISI
WordNet (110,000 nodes) from Princeton
Mikrokosmos (6000 nodes) from NMSU
Penman Upper model (300 nodes) from ISI
1-million+ instances (people, locations) from ISI
TAP domain relations from Stanford…
Undergoing constant reconciliation and pruning
Used in several past projects (metadata
formation for database integration; MT; QA;
summarization)
So far…

Annotations of 12 English texts:


6 pairs of translations of 1 text from each source
language
10 – 12 annotators for each text





Approximately 144 annotated texts total
Annotation manuals for IL0 and IL1
Annotation tools
Work on evaluation for interannotator agreement.
Now, we’re working on IL2 specification and
annotation.
Getting at Meaning
(Two translations of Korean original text)
Starting on January 1
of next year,
SK Telecom subscribers
can switch to
less expensive LG Telecom or
KTF. …
The Subscribers
cannot switch again
to another provider
for the first 3 months,
but they can cancel
the switch
in 14 days
if they are not satisfied with
services
like voice quality.
Starting January 1st
of next year
customers of SK Telecom
can change their service
company to
LG Telecom or KTF …
Once a service company swap
has been made,
customers
are not allowed to change
companies again
within the first three months,
although they can cancel
the change
anytime within 14 days
if problems
such as poor call quality
are experienced.
Color Key
Black: same meaning and same
expression
 Green: small syntactic difference
 Khaki: Lexical difference
 Red: Not contained in the other text
 Purple: Larger difference.


Need to use some inference to know that the
meaning is the same
Getting at meaning
(Two translations of a Japanese original text)










This year,
too,
in addition to
the birth
of Mitsubishi Chemical,
which has already been
announced,
other rather large-scale
mergers
may continue,
and be recorded
as a "year of mergers."











More lexical similarity.
More differences in dependency
relations.


This year,
which has already seen
the announcement
of the birth
of Mitsubishi Chemical
Corporation
as well as
the continuous
numbers of big mergers,
may
too
be recorded
as the "year of the merger“
for all we know.
Additional Topics in Lexical Semantics
English Transitivity Alternations

Beth Levin, 1993

Identified around 100 transitivity alternations
in English.
Transitivity Alternations and Semantic
Classes: Examples

Causative-Inchoative: change of state verbs







In other languages



Sam broke the glass. (causative)
The glass broke.
(inchoative)
Sam opened the door.
The door opened.
Sam kicked the ball.
*The ball kicked.
Inchoative verbs may be reflexive (e.g., Romance languages)
There may be a causative marker on the transitive verb.
Inchoative means beginning.

Beginning a change of state?
Transitivity Alternations and Semantic
Classes: Examples

Dative Shift: giving and telling







I gave Sam the book.
I gave the book to Sam.
I told the story to the children.
I told the children the story.
I drove the car to New York.
*I drove New York the car.
In other languages


The goal may not be able to become a direct object.
(Romance languages)
The goal may become a direct object in the presence of
an applicative morpheme. (Bantu languages)
Transitivity Alternations and Semantic
Classes: Examples

Spray-Load Alternation: filling and
covering.




Sam
Sam
Sam
Sam
sprayed the wall with paint.
sprayed paint on the wall.
loaded the truck with hay.
loaded hay onto the truck.
Transitivity Alternations and Semantic
Classes: Examples

There Insertion: stative, appearing








Problems exist.
There exist problems.
A ghost appeared.
There appeared a ghost.
The students worked.
*There worked some students.
The students disappeared.
*There disappeared some students.
Transitivity Alternations and Semantic
Classes: Examples

Locative subjects:



Bees swarmed in the garden.
The garden swarmed with bees.
Temporal subjects:

1990 saw the fall of the government.
Transitivity Alternations and Semantic
Classes: Examples

Middle: Telic verbs? (see below)






You can cut this bread.
This bread cuts easily.
You can sell these books easily.
These books sell well.
People like these books.
*These books like well.
Transitivity Alternations and Semantic
Classes: Examples

Resultative Secondary Predication: theme
version




Sam hammered the nail.
Sam hammered the nail flat.
The lake froze.
The lake froze solid.
Transitivity Alternations and Semantic
Classes: Examples

Resultative Secondary Predication: agent
version


He screamed himself hoarse.
He cried himself to sleep.
Class shifts

Manner of motion to change of location:





The
The
The
The
bottle floated.
bottle floated into the cave.
ball bounced.
ball bounced across the room.
Sound to change of location:




The car rumbled.
The car rumbled down the street.
The dress rustled.
She rustled across the room.
How universal?

How universal is argument structure?


If an English word has an agent and a patient,
will the translation-equivalent in another
language have an agent and patient?
If an English word has a subject and object,
will the translation-equivalent in another
language have a subject and object?

Less likely:
 I met him.
 I met with him.
How Universal?

How universal are alternations and
semantic classes?

If an English word undergoes a transitivity
alternation, will the translation equivalent in
another language undergo the same
transitivity alternation?

Even less likely. (Mitamura, 1989)
Importance of Transitivity Alternations in
Language Technologies

For any task that requires understanding
(question answering, information
extraction, machine translation) you need
to know the semantic roles of the NPs.



The glass broke. (subject is patient)
The kids ate. (subject is agent)
I gave them some books (object is recipient)
Importance of Transitivity Alternations in
Language Technologies

So you need multiple lexical mappings for each
verb:
break < agent patient>
subj obj
break < patient >
subj
give < agent theme recipient>
subj obj
obl
give < agent theme recipient>
subj obj2 obj
Importance of Transitivity Alternations in
Language Technologies

To speed up lexicon acquisition, assigning a verb
to a semantic class and automatically generating
its alternations is faster than listing all of its
lexical mappings by hand.






I gave books to the students.
I gave the students books.
Books were given to the students.
The students were given books.
There were books given to the students.
There were students given books.
Lexical Aspect

State


Activity


The children painted.
Accomplishment


The clock sat on the shelf.
The children walked to school.
Achievement

The ambassador arrived in Moscow.
Lexical Aspect




Took examples from this web page:
http://www.sfu.ca/person/dearmond/322/322.ev
ent.class.htm
Vendler, Linguistics in Philosophy, 1967
Dowty, Word Meaning and Montague Grammar,
1979
Tenny, Aspectual Roles and the Syntax-Semantics
Interface, 1994
Activities and Accomplishments

Activity:
 The children painted for an hour.

?The children painted in an hour.

The children will paint in an hour.
 They will start in an hour.

The children almost painted.
 Almost started painting

Test for telicity:
 If you start to paint and stop,
you have painted.
 Fails test for telicity.

Accomplishment:
 ?The children walked to school for
an hour.
 The children walked to school in
an hour.
 The children will walk to school in
an hour.
 They will start in an hour, or
it will take an hour.
 The children almost walked to
school.
 Almost started walking, or
almost reached school
 Test for telicity:
 If you start to walk to school
and stop, you may not have
walked to school.
 Passes test for telicity.
Telicity
Telic: has a goal or endpoint
(accomplishment)
 Atelic: does not have a goal or endpoint
(activity)
 Telicity can change depending on the
sentence:



He built houses for a year/*in a year.
He built a house in a year/?for a year.
Achievements

The ambassador almost arrived in
Moscow.

Only means “almost finished” not “almost
started.”
States (English)

Stative: Simple present tense means present
time. Present progressive does not sound good.



He knows the answer.
He is knowing the answer.
Non-stative: Simple present tense means
habitual or generic. Present progressive means
present time.


He paints.
He is painting.
Consequences of Lexical Aspect for
Language Technologies

English

You have to know the lexical aspect of the verb in order
to know what the tense morphemes mean.


The simple present tense means “habitual” with a nonstative verb, but means present time with a stative verb.
You have to know the lexical aspect of the verb in order
to know what the adverbials mean.

Almost can mean “almost started,” “almost finished,” or
both.
Consequences of Telicity

Japanese:
 Telic
verbs with –te iru have a resultative
meaning
 Aite iru: is open or has been opened, not is
opening
 Otite iru: is dropped (is on the floor), not is
dropping (unless it takes a very long time to
fall, like a leaf falling off of a sky scraper)
 Atelic
verbs with –te iru have a
progressive meaning
 Tabete iru: is eating, not has eaten
Consequences of Telicity
Japanese: -te aru (with passive-like
meaning) only applies to telic verbs
because it focuses on a resulting state.
(e.g., wash (arau), but not praise
(homeru))
Sara ga aratte aru.
Plate subj wash

???Taroo ga homete aru.
Consequences of Telicity: Finnish


Angelica Kratzer, Telicity and the Meaning of Objective
Case, International Round Table ‘The Syntax and
Semantics of Aspect’, Universite de Paris, Nov. 2000.
Telic: direct object can have partitive or
accusative case (with a slight difference in
meaning):
Ammu-i-n
karhu-a
Shoot-past-1sg bear-part
I shot at a/the bear
Ammu-i-n
karhu-n
Shoot-past-1sg bear-acc
I shot the bear

Atelic: can only have partitive case: despise,
admire, envy, love, study, play, listen, pull
Consequences of Telicity: Chinese

Lisa Lai Shen Cheng, Aspects of the Ba-Construction,
Lexicon Project Working Papers 24, Carol Tenny (ed.),
MIT, 1988.
Ta ba shu mai le.
He BA book sell ASP
He sold the book

Factors determining grammaticality of the baconstruction:



Aspect markers: occurs with le and zhe, but not with zai
and guo.
Definiteness: The direct object has to be interpretable as
definite.
Telicity of the verb: tui le (pushed) vs. tui dao le (pushed
down; push-fall); la le (pull) vs. la dao le (pull down; pullfall); dai le (bring/carry) vs. dai lai le (bring here; carrycome)
“Ba” and Telicity
*Wǒ bǎ
Lǐsì tūi-le.
I
BA Lisi push-ASP
“I pushed Lisi.”
Wǒ bǎ Lǐsì tūi-dǎo-le.
I
BA Lisi push-fall ASP
“I pushed Lisi and he fell.”
“Ba” and Telicity
*Tā bǎ
Zhāngsān lā-le.
He BA Zhangsan pull-ASP
“He pulled Zhangsan.”
Tā bǎ Zhāngsān lā-dǎo-le.
He BA Zhangsan pull-fall-ASP
“He pulled Zhangsan and Zhangsan fell.”
“Ba” and Telicity
*Tā
bǎ dìan-nǎo dài-le.
He BA computer bring-ASP
“He brought the computer.”
(Does this really mean “He carried the computer?”)
Tā bǎ dìan-nǎo dài-lái-le.
He BA computer bring-come-ASP
“He brought the computer here.”
“Ba” and Telicity
*Tā
bǎ fángjīan dǎ-sǎo-le.
He BA room hit-sweep-ASP
“He cleaned the room.”
Tā bǎ fángjīan dǎ-sǎo
de hěn gānjìng.
He BA room
hit-sweep DE very clean
“He cleaned the room and the result is that the
room is very clean.”
Two kinds of intransitive verbs:
subject is agentive or not
Sam worked.
agentive
Sam fell (by accident).
non-agentive
 Unaccusative: an intransitive verb whose subject is not
agentive.


Unergative: an intransitive verb whose subject is
agentive.



Because the noun phrase would have been accusative if
the verb were transitive?
Because the noun phrase would have been ergative if the
verb were transitive?
Confusing terminology by David Perlmutter and Paul
Postal.
Highly influential and insightful contribution to linguistic
theory also by David Perlmutter and Paul Postal.
Consequences of Unaccusativity or
Agentivity

English: Resultative secondary
predication:
*He screamed hoarse.
?He worked to exhaustion.
He worked himself to exhaustion
It broke to pieces.
It froze solid.
Consequences of Unaccusativity or Agentivity:
German Impersonal Passive
http://www.wm.edu/CAS/modlang/gasmit/grammar/passive/impspass.htm
Hier wird nicht geparkt.
No parking here.
Im Gang wird nicht geraucht.
No smoking in the corridor.
Es wurde viel getanzt und gesungen.
There was lots of dancing and singing.
Works with agentive verbs only.
Not with break, fall, etc.
Consequences of Unaccusativity:
Italian partitive clitics

http://www.sfu.ca/person/dearmond/405/405.ergative.unaccusative.htm
Sono passate tre settimane.
Are passed three weeks
Three weeks have passed.
Ne
sono passate tre.
Of-them are passed three
Three of them have passed.
Ne
sono arrivati(?) tre.
Of-them are arrived
three
Three of them have arrived.
* Ne
hanno telefonato(?) tre.
Of-them have phoned
three
Three of them have arrived.
Importance of unaccusativity

Non agentive subjects, direct object, subjects
of passives:




The water froze solid.
He hammered the nail flat.
The nail was hammered flat.
Agentive subjects and subjects of active,
transitive verbs.

He hammered the nail exhausted.


Doesn’t mean that he became exhausted as a result of
hammering the nail.
He screamed hoarse.

Doesn’t mean that he became hoarse as a result of
screaming.
Importance of Unaccusativity
Non-agentive subjects behave like direct
objects.
 Passive subjects correspond to direct
objects of active sentences.
 The Unaccusative Hypothesis (Perlmutter
and Postal): Maybe non-agentive subjects
are direct objects at some level of
representation.

Example of insight from the unaccusative
hypothesis
Why can’t German unaccusative verbs
become impersonal passives?
 They are already passive! The nonagentive subject was at some point an
object that got promoted.
