“Imperfective Paradox” in Formal Model vs.

Cognitive
Force-theoretic Model

Chungmin Lee
Seoul National University
clee@snu.ac.kr

Abstract
This paper introduces a new proposal by Copley and Harley (2014,
2015) to overcome the ‘Imperfective Paradox’ by relying on the
notion of dynamic ‘forces’ by Talmy (2000). They treat dynamic
predicates as predicates of forces, represented by functions
from an initial situation to a final situation that occurs ceteris
paribus, that is, if other conditions are equal or if nothing external
intervenes. Thus, they eliminate causative entailments, avoiding
the Paradox, and explain the frustratives as well. However, we
view the Progressive aspect as incremental progress based on
the planned/scheduled culminating accomplishment. Another
important phenomenon is that there occurs an ambiguity between
progressive and resultative, as in Korean and Japanese wearing
verbs. A caused existence of the accomplishment Theme is not
entailed by the perfective but its plan still affects our grammar of
language and cognition. Imperfective and perfective, appearing
as the same in some languages, require a new syntax-semantics
interface analysis, and conative and frame idea, to solve the
paradox.

Receive 18 January 2023; Revised 29 January 2023; Accepted 30 January 2023

Journal of Cognitive Science 23(4): 467-492 December 2022
©2022 Institute for Cognitive Science, Seoul National University
 468 Chungmin Lee

Key words: Imperfective Paradox; Force dynamics; Progressive

aspect; Resultative; Accomplishment; Culmination; Theme; Efficacy;
Conative; Frame

1. Introduction
1. Introduction
1. Introduction
Thispaper
ThisThis
paper paper discusses
discusses
discusses the so-called
the so-called
the so-called “Imperfective
“Imperfective
“Imperfective
Paradox” Paradox”
Paradox”
appearing
appearingappearing
in the
inProgressive
the Progressive
formform
in the Progressive form in accomplishment telic sentences, encountered
in accomplishment
in accomplishment
telictelic
sentences,
sentences,
encountered
encountered
by formal
by formal
semantic
semantic
models,
models,
and and
the recently
the recently
by formal semantic models, and the recently proposed solution by a
proposed
proposed
solution
solution
by amodel
force-theoretic by
force-theoretic
a force-theoretic
by Copleymodelmodel
andbyHarley
Copley
by Copley
and and
Harley
(2015), Harley
(2015),
which is(2015),
which
still which
is still
formal is still
formal
formal
butcognitively-oriented.
cognitively-oriented.
but cognitively-oriented.
but In the Inprevious
the formal
In previous
the previous
formal formal
models,
models,models,
causal
causal
event causal
event event theories
decomposition
decomposition theories
decomposition theories show the insight that an accomplishment has the
showshow
the insight
the insight
that that
an accomplishment
an accomplishment
has the
hasAgent
the Agent
of only
of only
the first,
the first,
causing,
causing,
sub-event, e1; e1;
sub-event,
Agent of only the first, causing, sub-event, e1; this event then is ‘chained’
this this
eventevent
thenthen
is ‘chained’
is ‘chained’ e2, which
withwith e2, which
is itself
is itself
related
related
to the
to Theme
the Theme
(Pustejovsky
with e2, which is itself related to the Theme (Pustejovsky 1991). A(Pustejovsky
causal 1991).
1991).
A A

relation
causal
causal
relationof
ofthe
relation form
ofform
the forme1∃e1∃e2:
e2: e1
∃e1∃e2: e1CAUSE
CAUSE e2 isise2
e1 CAUSE
e2 generally
generally assumed
is generally
assumed (Dowty
assumed
(Dowty
(Dowty
1979,
1979,
Lewis
Lewis
1979, Lewis 1973) (non-neo-Davidsonian) but for e1 to cause e2 here both
1973)
1973)
(non-neo-Davidsonian)
(non-neo-Davidsonian) for e1
but but e1cause
for to e2 here
to cause e2 here
bothboth
events
events
mustmust
exist.
exist.
ThisThis
is is
events must exist. This is required in the existential binding of e1 and e2.
required
required
Butine2the in existential
the not
does existential
binding
binding
necessarily of e1ofand
occur, e1 e2. But
and
because e2.ceteris
e2 does
But e2non
does
notparibus
necessarily
not necessarily
occur,
‘all other occur,
because
because

things
ceteris
ceteris being
non non
paribus not
paribus ‘allequal’
‘all [ceteris
otherother
things
things
being not =equal’
paribus
being notallequal’
other conditions
[ceteris
[ceteris
paribus being equal,
= all= other
paribus all other
conditions
conditions
i.e., ‘inertia’]. Hence “Imperfective Paradox” arises.
being
being
equal,equal,
i.e., i.e.,
‘inertia’].
‘inertia’].
HenceHence
“Imperfective
“ImperfectiveParadox”
Paradox”
arises.
arises.
Examining the force-theoretic model (Copley and Harley 2015, adopting
Examining
Examining the force-theoretic
the force-theoreticmodelmodel
(Copley
(Copley
and and
HarleyHarley
2015,
2015,
adopting
adopting
cognitive
cognitive
force-
force-
cognitive force-dynamic theory including Talmy’s (2000)), we will
dynamic
dynamic theory
investigatetheory
including
someincluding
Talmy’s
verbs Talmy’s
such (2000)),
(2000)),
we will
as ‘wearing’ we will
investigate
investigate
in languages somesome
verbs
including verbs
suchsuch
Korean as
to ‘wearing’
as ‘wearing’

insee howincluding
in languages they
languages can Korean
be ambiguous
including Korean
to see between
to how
see how Progressive
theythey
can can and Resultative
be ambiguous
be ambiguous
between
betweenwith
Progressive
Progressive
and and
the same surface form. We can see intermediate transitions in Progressive
Resultative
Resultative
withwith
the same
the same
surface
surface
form.
form.
We can
We see
can intermediate
see intermediate
transitions
transitions
in Progressive
in Progressive
and and
and the final result state of wearing a dress. In English, typically the verb
the final
the
putfinal
result
result
on is statestate
of wearing
employed offor
wearing
thea Progressive,
dress.
a dress.
In English,
In whereas
English,
typically
typically
the the wear
verb verb
the verb
put putisonemployed
on
is employed is employed
for for
forProgressive,
the Perfective
the Progressive,
the whereasthough
whereas thewith
the verbverbthe
wear issame
wear is surface
employed
employedform
for for of
the V-ing.
thebePerfective
Perfective though
though
withwith
the same
the same
We will also examine destruction in progress or annihilation verbs. For
surface
surface
formform
of beofV-ing.
be V-ing.

We will
We will
alsoalso
examine
examine
destruction
destruction
in progress
in progress
or annihilation
or annihilation
verbs.
verbs.
For annihilation,
For annihilation,
we must
we must

destroy
destroy
something
something
completely
completely
so that
so that
nothing
nothing
is left.
is left.
 “Imperfective Paradox” in Formal Model vs. Cognitive Force-theoretic Model 469

annihilation, we must destroy something completely so that nothing is left.

2. The Paradox and the Force Mental Model

2.1 The Paradox and Affordance

It is fine to say that Mia was building a house without its entailing that
the result state of a building’s existence occurred. That’s how “Imperfective
Paradox” occurs. Compare it with the following no paradox sentences.
is not (1a) is a transitive
a culmination verb but
Theme) andthethe
object ‘a cart’
Agent’s is not created
exertion (it is by
of pressure not pushing
a the cart
culmination Theme) and the Agent’s exertion of pressure by pushing
is not a culmination Theme) and the Agent’s exertion of pressure by pushing the cart has n the
consequence
cart has noof consequence
reaching a of telic culmination
reaching Goal, without
a telic culmination Goal, without to the added. So, (1a
to the store
consequence of added.
store reaching So, acannot
telic
(1a) as culmination
an activity Goal, without
Progressive cannot to timeadded.
the store
take So, (1a) as a
activity Progressive take the time span adverbial ‘inthe
fifteen span
minutes’ (if That
adverbial ‘in
activity Progressive fifteentake
cannot minutes’
the (That’s why adverbial
time span (1a) entailment relation minutes’
‘in fifteen is justified (if That’s wh
(1a) entailment relation
just like (1b), whichisappears
justifiedas ajust like (1b),
classical which
example of appears as a classical exampl
no Imperfective
(1a) entailment relation
Paradox. (1a) mayis justified just like (1b), which
well be explained appearstheory
as a classical example of n
Imperfective Paradox. (1a) may wellbybeGibsonianexplained affordance
by Gibsonian(Gibsonaffordance theory (
1966, Steedman 2002, Pustejovsky 2022) by which we perceive (objects of)
Imperfective Paradox. (1a) may well be explained by Gibsonian affordance theory (Gibso
1966, the
Steedman
world for2002,action Pustejovsky 2022) by
such as buttons/carts for which
pushing,we perceive
wheels (objects of) the wo
for rolling,
1966, Steedman
knobs for2002, Pustejovsky
turning, handles for 2022) by which
pulling, wesliding,
levers for perceiveetc.(objects
However,of) the world fo
action such as buttons/carts for pushing, wheels for rolling, knobs for turning, hand
such actions have no telic culmination points in mind (see section 4). In
action such as buttons/carts for pushing, wheels for rolling, knobs for turning, handles fo
pulling, levers
(1a), for asliding,
‘pushing etc. just
cart’ means However,
exertingsuch actions
pressure with have no telic
no result. culmination
In (1b), no points i
pulling, levers for sliding,
distance etc. mile’
such as ‘one However, such
nor to the actions
store as telichave
Goalno telic culmination
is expressed. So, (1a points in min
(see section 4).
and 1b) involve no Imperfective Paradox.
(see section 4).

(1) a.(1)
Miaa. M
 was
ia was pushing
pushing a cart ⇒ Mia
a cart pushed
Mia a cart.
pushed a cart. (just exerting pressure with no
b. M
 ia was running Mia ran.
(1) a. Mia was pushing a cart ⇒ Mia pushed a cart. (just exerting pressure with no resul
b. Mia was running ⇒ Mia ran. (a and b: No Imperfective Paradox involved)
b. Mia
Onwasthe running Mia ran.
⇒ examine
other hand, the(afollowing
and b: No Imperfective
sentences Paradox involved)
with culmination
Goal-reaching or Theme-reaching. In (2), the Progressive shows Mia’s
On theposition changing
other hand, (to finally
examine be at the end),
the following which by
sentences notculmination
with being at the end
Goal-reaching or T
On the otherofhand,
the street, the Goal,
examine does not entail
the following the Perfective
sentences sentence,Goal-reaching
with culmination creating or Them
(2a) does not entail (2b), and (3a) shows the Progressive not entailing its culmination
(2a) does not entail (2b), and (3a) shows the Progressive not entailing its culmination Them
Perfective, both creating Imperfective Paradox. (3b) and (3b’), on the other hand,
Perfective, both creating Imperfective Paradox. (3b) and (3b’), on the other hand,
470 Chungmin Lee

Imperfective Paradox. (3a) also shows the Progressive not entailing its
culmination Theme Perfective. Thus, (2) and (3a) both create Imperfective
Paradox. (3b) and (3b’), on the other hand, using the adversary connective
but, a contradictory follow-up is possible or using the frustrative, as cited in
Copley and Harley’s (2014) O’odham one, or a similar use of -taka ‘changing
the course of action on the way’ in Korean, as in (3b’), the on-going action
in the progressive may be blocked or suspended and the culmination of
accomplishment is given up on the way. (2a) does not entail (2b), although
Mia’s position changing occurred, with the intention to finally be at the
end, and (3a) shows the Progressive not entailing its culmination Theme
Perfective, both creating Imperfective Paradox. (3b) and (3b’), on the other
hand, the actions in the Progressive stopped without culmination by means
of the adversary or frustrative connective. In (3c), however, a past sentence
(or perective) precedes such a frustration, resulting in oddmess.

(2) a. M
 ia was crossing the street.
b. M
 ia crossed the street.

(3) a. Mia was building a house => Mia built a house.

b. Mia was building a house, but she didn’t finish.
b.’ Mia-nun cip-ul cit-ko iss-taka cakum-i mocala cwungtanhay-ss-ta
 M-TOP house-ACC build-PROG-on the way funds-NOM lack
stopped
‘Mia was building a house but, on the way, she stopped because of
lack of funds.’
c. # Mia built a house, but she didn’t finish. (Here contradiction arises
because the first clause entails the event of a house existing as a
result of the causing sub-event, e1)

Copley and Harley (2015) give a Tagalog example to show how an

 “Imperfective Paradox” in Formal Model vs. Cognitive Force-theoretic Model 471

adversative/frustrative can block the completion of a telic accomplishment.

Their text cited here as (3) in our (4) below, however, seems to have an
error as follows in inalis, which is perfective, meaning ‘successfully
removed.’ It is not just ‘tried to remove.’ Therefore, as given in their text, it
creates a contradiction with the continuing adversative pero ‘but’ utterance.
The perfective inalis must be replaced by its imperfective form (like the
Progressive) inaalis (native Tagalog speaker Joanna Visaya, p.c. 2002).
The word mantas must be replaced by mantsa as well. ‘Quickly’ must be
replaced by ‘immediately’ in the translation and the entire translation must
be ‘I was removing the stain, but I ran out of soap immediately, so I couldn’t
remove it.’ Copley and Harley (2015) gives the different form inaalis in
their footnote 5 as a form of completion entailment, which seems to be
erroneous as the opposite. They further provide various other languages
such as Chinese (from J-P Koenig and L-C Chief 2008) to resort to “non-
result-entailing theories of causation” such as probabilistic (Reichenbach
1971) and force-dynamic (Copley and Wolff (2014)). But many other
Mandarin Chinese speakers (including Linmin Zhang, p.c. 2023) disagree
and say sha-le “kill-PF” does entail death. (4) in Tagalog also counters
Copley and Harley, as explained. All those perfective “non-result-entailing”
data cited from various languages seem to be either erroneous or marginal/
loose data, missing the maximal degree word ‘almost’ or something like
that in a special context pragmatically. They must involve telic predicates’
conative implications (Borer (2005)) in the exceptional speakers’ minds (as
we see ‘tried’ involved in the cited Tagalog translation).

(4) (3) Inalis ko ang mantas, pero naubusan ako kaagad

N-PF-remove GEN-I NOM stain, but run-out-of NOM-I rapidly
ng sabon, kaya hindi ko naalis.
GEN soap hence not GEN-I A-PF-remove
“I tried to remove (lit. ‘I removed’) the stain, but I ran out of soap,
472 Chungmin Lee

and couldn’t.” (Dell 1987, p. 186)

In face of the paradox, Copley and Harley (2015) try to account for non-
culmination by improving on the previous efforts, although it looks like
avoidance of the paradox, first, by Dowty (1979) of assuming a relation
between two sub-events and additionally requiring possible worlds to solve
non-culmination and the efforts, second, by Parsons (1989) of proposing a
non-culminated event as bearing some scalar relation to a culminated event,
though without clarity. Dowty’s possible worlds seem to be Copley and
Harley’s ‘situations’ for non-culmination, though how Parsons’s scalarity
idea is used is not clear in them either.
I will present the ambiguity of the Progressive and the Perfective in ‘wear’
verbs in Korean and Japanese, as in 3. The Puzzle of Ambiguity between
the Progressive and the Resultative in (13), to show how the situation
precedence relation by Copley and Harley’s (2015) reverse function
predec(s) = suc−1 (s) is not enough. The preceding clause verb ‘wear’ before
the connective -ko has its culmination interpretation, which is absent in the
Progressive.

2.2 Force and Situations

With such intuitions, Copley and Harley (2015) adopt “situations” to be
in line with interpretations of situation semantics/DRT approaches in which
situations/DRSs are representations of mental models.
To represent situations in a Montagovian semantics (Montague 1970),
they assign them the primitive type s. Let’s follow their formal steps (from
(5) to (11)):

(5) Type of situations: s

Type of forces: <s,s> Type f in abbreviation
 “Imperfective Paradox” in Formal Model vs. Cognitive Force-theoretic Model 473

Because we now have the force type f, let us take a look at their syntactic
structures first. They first adopt the current generative grammar syntax and
the distributive morphology (DM) theory, adopting conceptual meaning
decomposition, as shown in their syntactic trees, with model-theoretic types
assigned. Examine their (22) and (69), SC is a proposition <st> type, vP and
active S are <ft> type:

(22)

In the following diagram of causal chain, the vectors are depicted in the bubbles because the

In the followingconceptual
diagram of thecausal
forces arise from chain,
conceptual situations theby thevectors
represented bubbles. are depicted in the
bubbles because the conceptual forces arise from the conceptual situations
(6) Causal chain of situations with net forces

represented by the bubbles.

(6) Causal chain of situations with net forces

474 Chungmin Lee

(7)(7)
net(s)
net(s)=:=:the
the net forceofofs s
net force

The inverse of the f unction−1net is net−1. Given a (particular,

The inverse of the function net is net . Given a (particular, spatiotemporally bounded) force
spatiotemporally bounded) force f, we can refer to both the situation of
f, we can refer to both the situation of which it is a net force, and the situation that follows. The
which it is a net force, and the situation that follows. The initial situation of
initial
f is simplysituation of f is simply
the situation s ofthewhich
situation
it iss of which
a net it is aThe
force. net force.
finalThe final situation
situation is the is the
situation
situationthat
that results whenf takes
results when f takes
s assitsasargument.16
its argument.
The The functions
functions init andinit anddefined
fin are
finwith
are respect
definedto with respect
the inverse to theofinverse
function net. function of net.

(8) a. init(f) = net−1 (f)
(8) a. init(f) = net−1−1(f)
b. fin(f) = f(net (f))
b. fin(f) = f(net−1 (f))

Definition of a situation’s successor and predecessor situation:

Definition of a situation’s successor and predecessor situation:
(9) a. suc(s) = fin(net(s))
b. predec(s) = suc−1 (s)
(9) a. suc(s) = fin(net(s))

b. predec(s) = suc−1 (s)

Individuals with type e: variables x, y, z, ---; predicates p, q, etc. and (stage-
level) statives are predicates of situations, type <s,t>, and when they are
Individuals with type e: variables x, y, z, ---; predicates p, q, etc. and (stage-level) statives are
dynamic, they are predicates of forces, type <f, t>.
predicates of situations, type <s,t>, and when they are dynamic, they are predicates of forces,

2.3type
Efficacy
<f, t>.
A situation s0 is efficacious just in case its ceteris paribus (inertia)
successor situation occurs. Consider the diagram in (10) below. When
2.3 Efficacy
A situation s0 is efficacious just in case its ceteris paribus (inertia) successor situation occurs.
 “Imperfective Paradox” in Formal Model vs. Cognitive Force-theoretic Model 475

choosingConsider
an initial situation
the diagram s0,below.
in (10) the speaker chooses
When choosing (the)
an initial one s0,
situation that
theshe judges
speaker chooses
a
to be efficacious.
(the) one thatShe may tojudge
she judges that s0She
be efficacious. ismay
thejudge
efficacious situation;
that s0a is the efficaciousshe is
situation;
a
judging that
she is no forces
judging arising
that no at least
forces arising partially
at least partiallyfrom outside
from outside s0 perturb
s0a will will perturb
f0a , so that
f 0a, so that s0aindeed
s0a will will occur.
indeed occur.
Something Something
from from
outside s0a may
a
outsideinswhich
well intervene, 0 may casewell
perhaps
intervene,
thein which initial
efficacious casesituation
perhaps is s0the
b
, andefficacious initial
what will occur situation
is its successor s1b.is s0b, and
Force-theoretic
what will occur is itsThe
non-determinism. successor s1bpicked
causal chains . Force-theoretic non-determinism.
out by superscripts Thewell
a and b correspond rather
causal chains picked
to “histories” out by1970).
(Thomason superscripts a and b correspond rather well to
“histories” (Thomason 1970).
(10)

(10)

In accordance with such intuition but to avoid the paradox, Copley and
In accordance with such intuition but to avoid the paradox, Copley and Harley argue for forces
Harley argue for forces to model “efficacy” (as explained), removing
to model efficacy (as explained), removing causative entailments. They also argue for forces
causative entailments. They also argue for forces to model intervention,
to model intervention, which remains a long-standing puzzle in the event-structure literature
which remains a long-standing puzzle in the event-structure literature
concerning the status of verbs of maintaining like keep and stay (Jackendoff 1975, a.o.). They
concerning the status of verbs of maintaining like keep and stay (Jackendoff
are clearly dynamic, as diagnosed by the usual eventuality tests; for example, the progressive
1975, a.o.). They are clearly dynamic, as diagnosed by the usual eventuality
gets an ‘ongoing-now’ reading.1
tests; for example, the progressive gets an ‘ongoing-now’ reading.1

(11) a. The rock is keeping the door open.

b. The door
Similarly,
1 is staying
a stative verb such asopen.
‘stand’ can be in progressive: The statue is standing there for some time.

1
Similarly, a stative verb such as ‘stand’ can be in progressive: The statue is
standing there for some time.
 476 Chungmin Lee

They argue that their occurrence in the progressive shows that these verbs
can be dynamic. However, the fact that they are interpreted habitually in
sentences such as those in (12) with the simple present indicates that they
essive shows that these verbs can be dynamic.
must be dynamic. It is not easy to understand what distinguishes such
bitually in sentences such as those in (12) with
dynamic eventualities from stage-level statives such as The door is open.
e dynamic. It is not easy to understand
However, what adjective in English happens to be a passive
open, a stative
form
stage-level statives suchinasKorean,
The doorwith its action verb yel- ‘open’ being a base form but
is open.
functioning as a statve in passive. Talmy (2000) in cognitive semantics
happens to be a passive form in Korean, with its
points out that the verb stay in (11b) lexically encodes the presence of a
unctioning as a statve in passive. Talmy (2000)
force that intervenes to counteract the ceteris paribus result of an existing
n (11b) lexically encodes
force. the a forceextends the Talmian project to characterize verbal
presence of(2007)
Gaerdenfors
concepts
s result of an existing as patterns (2007)
force. Gaerdenfors of force.

bal concepts as patterns of force.

(12) a. The rock keeps the door open.
b. The door stays open.

3. The Puzzle of Ambiguity between the Progressive and the

Resultative
he Progressive and the Resultative
This postulation of a force that ‘intervenes to counteract the ceteris
s to counteract the paribus result result
ceteris paribus of an ofexisting
an force’ may give some clue to solving the
ambiguity
the ambiguity between betweenand
the Progressive thetheProgressive and the Resultative in Korean and
Japanese.
he following ambiguous Observe the following ambiguous sentence:
sentence:

(13) ai-ka say os-ul ip -ko iss -ta

o iss -ta child-NOM new dress-ACC wear-PROG/RESULTative DEC
G/RESULTative DEC a. ‘The child is putting on a new dress.’ (Progressive) henhi ‘slowly,’ as in a.’
ess.’ (Progressive) π(net(s)). In this Progressive reading, the verb ip- ‘put on’ can be modified by an adverb

Progressive reading may be captured by Copley and Harley’s denotation: ⟦progressive⟧ =

This Progressive reading may be captured by Copley and Harley’s
Copley and Harley’sdenotation: progressive == λπλs. π(net(s)). In this Progressive reading, the
denotation: ⟦progressive⟧

verb ip- ‘put on’ can be modified by an adverb a. ‘The child is putting on a new dress.’ (Progressive)
child-NOM new dress-ACC wear-PROG/RESULTative DEC
-ta iss ip -ko os-ul say ) ai-ka
 “Imperfective Paradox” in Formal Model vs. Cognitive Force-theoretic Model 477

verb ip- ‘put on’ can be modified by an adverb chenchenhi ‘slowly,’ as in a.’

a.’ ai-ka say os-ul chenchenhi ip -ko iss -ta

child-NOM new dress-ACC slowly wear-PROG DEC
‘The child is putting on a pretty dress slowly.’

The Resultative reading of (13), on the other hand, is as (b) and the verb iss-
‘exist’/‘remain’ can be modified by kamanhi ‘quietly’ as in b’:

b. ‘The child is wearing a new dress.’

b.’ ai-ka say os-ul ip -ko kamanhi iss -ta
child new dress-ACC wear-and quietly exist-DEC
‘The child remains quietly wearing a new dress.’

This kind of Progressive accomplishment verb sentence does not entail its
past form sentence in Korean (13), English (14), German (15), and Chinese
(16), still revealing the paradox (cf. (1a) of Copley and Harley 2015).

(14) The child is putting on a new dress. (Progressive)

As in (14), English shows no ambiguity of wearing verbs between the

Progressive and the Resultative because different verbs ‘put on’ vs. ‘wear’
are employed in the division of labor. ‘Wear’ refers to having clothes on
(accomplishment). When someone is wearing something, they are not doing
the action. They have it on them.

(14) =
 > The child put <PAST> on a new dress (--- wore a new dress in
the past).
Likewise, in Korean, the Progressive sentence cannot entail its past
sentence, as follows:
478 Chungmin Lee

(13a’) => ai-ka say os-ul chenchenhi ip -ess -ta (K)

child-NOM new dress-ACC slowly wear-PAST DEC
‘The child put on her new dress slowly.’
(15) a. Sie hat ihr Kleid angezogen. <present perfect but progressive
meaning> (German)
‘She is putting on her dress.’
b. Sie trägt ihr Kleid <present but stative meaning> (Frank Sode via
e-mail)
‘She is wearing her dress.’

(15a) => (15b)

(16) a. ta-men zai zao yi dong fang-zi (Ch)

they at/exist build one CL house
‘They are building a house.’ [CL= Classifier]
b. ta-men zao-le yi dong fang-zi
they build-PFV one CLAS SIF house
‘They(’ve) built a house.’

(16a) => (16b)

(17) a. ta (zheng)-zai chuan qun-zi

she right-at/exist put on dress
‘She’s putting on her dress.’
b. ta chuan -zhe qun-zi (Linmin Zhang, via e-mail)
she wear dress
‘She’s wearing her dress.’

(17a) => (17b)

 “Imperfective Paradox” in Formal Model vs. Cognitive Force-theoretic Model 479

(14) in English seems to well fit the denotation by Copley and Harley’s
(2015) progressive formula. Here the verb ‘putting on’ must be an
accomplishment verb but interestingly its telic result state is expressed by
‘wearing.'. For the telic consequence, a different verb ‘wear’ is typically
employed in English. Copley and Harley’s (69) [tree structure] above
may be an approximation. For the telic end existence, we may need
some frame in mind. Morphologically -ko iss in Korean here functions
as a grammaticalized progressive form as a unit and no adverbial can be
inserted between the segments, -ko and iss, as argued in Lee (1999). (13b),
on the other hand, has the Resultative reading, with -ko retaining its original
conjunction meaning ‘and (then).’ Here a temporal sequence marker -se
can be added to -ko ‘and’ to ascertain the temporal precedence sequence
and the completion of telicity in this reading, which requires a bi-sentential
construction, iss- in Korean retaining its original meaning of ‘exist’/’stay.’
Because they are separate morphemes and have different morphosyntactic
meanings, adverbials such as ‘quietly’ can freely be inserted between the
two morphemes, with an independent clause preceding the conjunctive
marker -ko ‘and’ in Korean. The second morpheme iss-, in its original
meaning of ‘exist’/’stay,’ in Korean, I argue, is an independent verb like
those in (11) ‘stay’ or ‘remain.’ (13b) above is repeated as (18) below. The
adverb kamanhi ‘quietly’ modifies the main clause verb iss- ‘exist’/’stay’
and the event time of the preceding subordinate clause precedes the present
time of the main clause. With the telic completion meaning solidified, the
main clause verb iss- ‘exist’ can be replaced by any non-contradicting verbs
such as ‘disappeared,’ as in (20b) [Japanese], and ‘danced.’ The crucial point
I raise here is that it is not a simple preceding situation by their predecessor
function, “predec(s),” as proposed by Copley and Harley (2015) to be
explained shortly. In this Resultative construction, the preceding clause
before the clause-linking conjunction -ko became perfective with the telic
culmination completed. Consequently, (18) entails (18’), which is in the past.
480 Chungmin Lee

No Imperfective Paradox is involved here. This crucial semantic renovation

has not been explored previously.

(18) = (13b’) aii-ka [[Φi say os-ul ip ] -ko] kamanhi iss -ta
child-NOM new dress-ACC wear -and quietly exist -DEC
‘The child remains quiet, wearing a new dress.’
(18’) aii-ka [[Φi say os-ul ip ] -ess -ta
child-NOM new dress-ACC wear PAST DEC
‘The child wore a new dress.’

Also, in English, a similar effect tends to arise in ‘is quietly wearing ---’
in (19) below. The verb ‘is’ here, I argue, is like ‘remain/stay’, as discussed
in (11), and is independent of the eventive verb ‘wear,’ tending to form
a bi-clausal construction.’ The verb ‘is’ here must be ‘exists,’ rather than
a copula, as in the Korean bi-clausal construction. Therefore, to have a
Progressive event, another verb ‘put on’ must be used and if that verb is
used, the movement (‘VP fronting’ a la John Whitman p.c.) is slightly odd
(Darcy Sperlich p.c.), as in (19b). The result state clause of wearing, just like
the Resultative ip-ko in Korean, freely appears as an adjunct with any non-
contradicting main clauses, as in (19c) (Darcy Sperlich p.c.). In this respect,
the three languages agree in Resultative constructions. One interesting trend
regarding the meaning of the progressive in English is that it can mean
‘wanna’ (Linmin Zhang providing (19d) pc, Darcy Sperlich agreeing pc
2003). Examine (19d). In this telic accomplishment, the progressive means
‘wanna’ regardless of the 2nd, 1st or 3rd person of the subject Agent, resulting
in .wanna wear -. This shows the conative nature of telicity (Borer (2005)).

(19) a. The child is wearing her dress.

a.’ Wearing her dress, the child is.
b. The child is putting on her dress.
 “Imperfective Paradox” in Formal Model vs. Cognitive Force-theoretic Model 481

b.’ #Putting on her dress, the child is.

c. The child disappeared wearing her dress.
d. S
 cenario: The kid is in her bed. She will get up soon and go to
kindergarten. Mom is preparing the kid's clothes and asking her
what she wants to wear.
Mom: What are you wearing today?
The kid: It looks sunny. I'm wearing my new dress.

In Japanese, as in (20a), -te iru has its resultative reading with the same
‘put on/wear’ verb, like in Korean and unlike in English. Native Japanese
speakers seem to accept the progressive meaning as well with the same
verb (Yukinori Takubo, via e-mail 2022). So, it is ambiguous. An emphatic
progressive form ki-te iru saicyu-da ‘in the middle of putting on ---’ is
preferably used by some. The main clause verb iru now can be replaced by
other verbs such as inakunatta ‘disappeared’ freely with the accomplishment
culmination intact, as in (20b). (20b’) in Korean is exactly equivalent to
(20b) in accomplishment culmination but (20c) is totally different with the
connective -myense ‘while,’ with no accomplishment culmination.

(20) a. kodomo-ga kireina fuku-o ki-te iru

child-NOM pretty dress-ACC wear-C exist
‘The child is putting on/wearing a pretty dress.’
b. kodomo-ga kireina fuku-o ki-te inakunatta
child-NOM pretty dress-ACC wear-C disappeared
‘The child was wearing a pretty dress and disappeared.’
b’. ai-ka yeppu-n os-ul ip-ko saraci-ess-ta
child-NOM pretty dress-ACC wear-and disappear-PAST-DEC
‘The child was wearing a pretty dress and disappeared.’
c. ai-ka yeppu-n os-ul ip-umyense saraci-ess-ta
child-NOM pretty dress-ACC wear-on the way disappear-PAST-DEC
 b. The door stays open.

3. The Puzzle of Ambiguity between the Progressive and the Resultative

482 This postulation of a force Chungmin Lee to counteract the ceteris paribus result of an
that ‘intervenes

existing force’ may give some clue to solving the ambiguity between the Progressive and the
‘Theinchild
Resultative wasand
Korean putting on a Observe
Japanese. pretty dress and disappeared.’
the following ambiguous sentence:

Surprisingly, we can see that the Resultative reading is commonly available

(13) ai-ka
for wearing say three languages
verbs in the os-ul ip -ko
we iss
examined. -ta

It wouldchild-NOM new dress-ACC

not be easy wear-PROG/RESULTative
for the proposed force model toDEC explain the
ambiguity of a. the
‘TheProgressive andonthe
child is putting Resultative
a new for wearing verbs in
dress.’ (Progressive)
chenchenhi ‘slowly,’ as in a.’
Korean and Japanese. The force model can explain the Progressive mono-
λπλs. π(net(s)). In this Progressive reading, the verb ip- ‘put on’ can be modified by an adverb

This Progressive reading may be captured by Copley and Harley’s denotation: ⟦progressive⟧ =
clausal construction
This Progressive by means
reading may be of the proposed
captured by Copleydenotational meaning: ⟦progressive⟧ =
and Harley’s denotation:
progressive = λπλs. π(net(s)) (11a’). But in the resultative reading, -ko
λπλs. π(net(s)). In this Progressive reading, the verb ip- ‘put on’ can be modified by an adverb
functions as a rather subordinating connective, and the same morpheme
iss-chenchenhi ‘slowly,’
‘be’ functions as as
aninindependent
a.’ main clause dynamic verb ‘exist,’
a. ‘The child is putting on a new dress.’ (Progressive)

‘stay,’ or ‘remain’ and can be modified by ‘quietly,’ etc, at least in Korean

child-NOM new dress-ACC wear-PROG/RESULTative DEC

and Japanese. See (13) above. Copley and Harley view the progressive
-ta iss ip -ko os-ul say (13) ai-ka
denotation as causal chains occurring in efficacy, not involving culmination
or destruction so that frustrative situations can easily be solved. Observe (21).
Resultative in Korean and Japanese. Observe the following ambiguous sentence:

(21) kutul-un cip-ul cit-ko iss-taka mwuneci-e cwungtanhay-ss-ta.

existing force’ may give some clue to solving the ambiguity between the Progressive and the

They-TOP house-ACC build-PROG-FR fall stop-PAST_DEC

This postulation of a force that ‘intervenes to counteract the ceteris paribus result of an
‘They were building a house and then it fell and they stopped.’
3. The Puzzle of Ambiguity between the Progressive and the Resultative

However, the activity involved is progressed based on the blueprint of the

house imagined for accomplishment; rate of construction being measured;
b. The door stays open.
even this expressed degree may function as a new culminating point. Upon
(12) a. The rock keeps the door open.
hearing (22), an immediate question may be: ‘Then, who are building the
other half?’ (Ken Turner, p.c.) Examine.
extends the Talmian project to characterize verbal concepts as patterns of force.

(22) They are building half of the house.

that intervenes to counteract the ceteris paribus result of an existing force. Gaerdenfors (2007)

in cognitive semantics points out that the verb in (11b) lexically encodes the presence of a force
Korean uses the same progressive form -ko iss for all Activity or
action verb yel- ‘open’ being a base form but functioning as a statve in passive. Talmy (2000)
unergative verbs such as ‘laugh,’ ‘run,’ etc., and uses the separate Resultative
However,open, a stative adjective in English happens to be a passive form in Korean, with its

distinguishes such dynamic eventualities from stage-level statives such as The door is open.

the simple present indicates that they must be dynamic. It is not easy to understand what

However, the fact that they are interpreted habitually in sentences such as those in (12) with
 “Imperfective Paradox” in Formal Model vs. Cognitive Force-theoretic Model 483

form -e/-a iss for Telic intransitive or inchoative achievement verbs such
as ‘sit,’ ‘stand,’ ‘lie,’ ‘die,’ ‘dry,’ etc. The latter's progressive forms in -ko iss-
can be used in their very slow motion and their causatitve forms with -i, hi,
ri/li, ki, etc. (24c) can regularly be derived.

(23) Mia-ka anc-a iss-ta

M-NOM sit-A ISS-DEC
‘Mia sits.’

(24) a. swuken-i maru-ko iss -ta2

towel-NOM dry-PROG DEC
‘The towel is drying.’
b. swuken-i maru-a iss -ta3
towel-NOM dry-RESULT- DEC
‘The towel dried and remains.’
c. Mia-ka swuken-ul mar-li-ko iss -ta
M-NOM towel-ACC dry-CAUS-PROG-DEC
‘Mia is drying the towel.’

In Japanese, Copley and Harley’s examples (45) are given below in (25):

(25) (45) a. Taroo-wa wara-te iru

T TOP laugh-TE be.PRES
‘Taroo is laughing.’
b. Taoru-wa kawai -te iru
towel -TOP dry.INTR TE be.PRES
‘The towel has dried.’

2
This is subtle but used often. A very slow change is presupposed. Intransitive
action such as ‘laughing’ and ‘crying’ are more naturally used in Korean as well.
3
A past tense form is used more often.
 - ‘The towel has dried.’

the Progressive and the Resultative in intransitive

484 verbs are distinct in Korean, as
Chungmin Lee

(24a) and (24b), respectively. (25b) = (45b). Copley and Harley attack Ogihara’s
Recall that the Progressive and the Resultative in intransitive verbs are
rtial events’ solution for not providing any conditions under which one event can be
distinct in Korean, as shown in (24a) and (24b), respectively. (25b) = (45b).
Copley
nother. Nishiyama’s and Harley
(2006) partialattack Ogihara’s
events (1998)worlds
plus inertia ‘partial as
events’ solution for(1998)
per Portner’s not
providing any conditions under which one event can be ‘part of’ another.
also attacked for the possible relationship between logical and cognitive systems.
Nishiyama’s (2006) partial events plus inertia worlds as per Portner’s (1998)
solution
ad argue that their is also
final attacked
situation of for theas
force possible relationship
the reference betweencan
situation logical
solveandthe
cognitive systems. They instead argue that their final situation of force as
of -te iru, putting
the reference predecessor
forth thesituation function,
can solve “predec(s),”
the denotation which
of -te iru, picks
putting outthethe
forth
predecessor
r situation s, as function,
a solution. They“predec(s),”
introduced which picks out the
“predec(s)” predecessor
earlier situationthe
to explain
s, as a solution. They introduced “predec(s)” earlier to explain the meaning
of again
of again in their (26a):in [[again]]
their (26a):=[[again]]
λpλs .=λp(s)
λpλs . λp(s) presupposed: ∃s’
presupposed: s’ prior
priorto to
(predec(s)): p(s’). Now their solution is (26) below = their (46):
): p(s’). Now their solution is (26) below = their (46):
(26) (46) [[-te iru]] = λπλs . λ(net(predec(s)))

For an. λ(net(predec(s)))

6) [[-te iru]] = λπλs Activity, they may use the Progressive denotation where the final
situation and the initial situation are not quite distinct. For a telic predicate,
however, the predecessor situation of s is picked out by (predec(s)) in
(26) = (46), thereby the final situation is claimed to be the result state,
which
ivity, they may use thebegins after the
Progressive initial situation.
denotation where theThis
finalissituation
claimed andto yield the
the initial
resultative interpretation. This treatment must have advantages over
re not quite distinct.
OgiharaForand
a telic predicate,
Nishiyama however,
but their the Hasegawa’s
own and predecessor(1996)
situation of s is
situation
by (predec(s)) precedence relation
in (26) = (46), is not fully
thereby specific
the final about semantic
situation is claimedconditions
to be theon telic
result
accomplishment and its resultant existence, as in ip- ‘wear.’
ch begins after Onethe reasonable
initial situation.
direction This is claimed
of solution seems totocome
yieldfrom
theHamm
resultative
and
Bott’s (2021) introduction of the “frame” in Artificial Intelligence (AI) (van
on. This treatment must have advantages over Ogihara and Nishiyama but their own
Lambalgen and Hamm 2005, Moens and Steedman 1988 a.o.), viewing
awa’s (1996) situation precedence
lexical meanings relation
as scenarios. is not fully specific
All accomplishments take theabout semantic
same form of
scenario, the differences being that the individual scenarios involve different
preparatory processes (run, build, put on/wear, etc.), incremental themes
 “Imperfective Paradox” in Formal Model vs. Cognitive Force-theoretic Model 485

(distance\((x)\, ‘run a mile’), house\((x)\), (worn) dress\((x)\),

etc.), and resultant states. The lexical meaning of the accomplishment run
a mile corresponds to scenario (27) = (54). Like the axioms, the scenario
takes the form of a logic program. (All variables are universally quantified.)
Following van Lambagen and Hamm 2005, the scenarios formalize the
event nucleus of Moens and Steedman (1988):

(27) = (54) (a) Initially(distance(a))

(b) Initiates(start, run, t)
(c) Releases(start, distance(x))
(d) Initiates(finish, distance(a+1), t)
(e) Terminates(finish, run, t)
(f) HoldsAt(run, t)→HoldsAt(distance(a+1), t) → Happens(finish, t)
(g) HoldsAt(distance(x), t) → Trajectory(run, t, distance(x+g(d)), d)

All accomplishment themes have incremental themes (distance(x), house(x),

dress(x), null(x) ‘destruction,’ etc.), and resultant states. If ‘Mary was
putting on the new dress when the fire alarm rang so she never got it’
(John Whitman, p.c), then with all other frustrative/adversary connections,
the resultant state of wearing it does not happen. There was an activity of
putting on the dress with the Agent’s intention of wearing it, but the result
state didn’t occur. Even a simple activity scenario ‘Ed was running’ can be
a proper preparatory part of (54). If we remove clauses (d)–(f) from (54),
the simpler activity scenario can be easily obtained. In this scenario, both a
past progressive activity sentence ‘Ed was running’ and a past progressive
accomplishment sentence ‘Ed was running a mile’ will trigger a discourse
update as stated in integrity constraint: ?HoldsAt(run,t)^t<now succeeds.
This suggested solution to the paradox is non-monotonic with no possible
worlds. The non-monotonic nature of the Event Calculus can be used to
model the incremental construction of a temporal model with inferences
486 Chungmin Lee

which hold locally at a discourse segment sn but which can be undone at

some later discourse unit sn+1 to solve the frustrative cases such as ‘Ed was
running a mile, when he was hit by a truck’ or ‘Mary was putting on the
new dress when the fire alarm rang so she never got it’ or ‘---she was never
wearing it.’ The Event Calculus will compute a minimal model by resorting
to the nob-monotonic reasoning, they claim.
For verbs of creation and destruction (annihilation), different languages
seem to show some different decompositional manifestation; in English
the adjective open is a root (see trees (22) and (69)) but the verb ‘open’
yel- is a root in Korean and its stative meaning must be in a passivized
form yel-li-(e, n) ‘opened.’ Korean has an interesting lexical negative
stative verb eps- ‘not existent,’ coming from the negation of iss- ‘existent’
and its causative form eps-ae- ‘annihilate’ derives from the same root iss-
‘existent’ and ‘all the creation verbs like ‘make,’ ‘create,’ ‘compose,’ etc.,
assert the Theme’s coming into being, but do not presuppose its existence.
Therefore, these verbs contain the elements of CAUSE and most deeply
embedded (BECOME(EXISTENT y)). The annihilation causative verbs
like ‘destroy,’ ‘annihilate,’ etc. have the same elements, this time with the
negative existential (BECOME(NOT(EXISTENT y))) (Lee 1973). All the
creation verbs, such as ‘build a house,’ if their telic culmination occurs,
as intended, (EXISTENT y) is the final end-point. If the expressed ‘half’
is the object in the Progressive sentence, it should be a new Theme out of
the originally planned ‘house.’ In the ‘wear’ verb situation, a fully-worn
state must be existent at the end. How the verb stem ip- ‘put on’/’wear’
plus the connective -ko, followed by the stative verb iss- ‘existent’
can come up with the intended fully-dressed state y becomes existent
(BECOME(EXISTENT y)). On the contrary, if destruction or annihilation
verbs take a Theme, the direction of change becomes the opposite of
creation verbs: the telic culmination end-point of Theme (concrete or
abstract) y (BECOME(NOT(EXISTENT y))).
 “Imperfective Paradox” in Formal Model vs. Cognitive Force-theoretic Model 487

.;

As we saw the variation between the root adjective open in English and the root verb Vi

yel- ‘open’ in Korean, different languages have different creation verbs. Creation verbs as
As we saw the variation
accomplishment verbs unlikebetween theno existence
change verbs have root adjective
presupposition, asopen in English and
in (27, 28).

the root verb Viis yel-

‘Coffee’ ‘open’
not presupposed in inKorean,
to exist different languages
(27). Their metonymically-derived versions follow have different

them, as in (27’) and (28’).

creation verbs. Creation verbs as accomplishment verbs unlike change verbs
(27) Mina-ka coffee-rul kkulh- i -ko iss-ta
have no existence presupposition,
M-NOM
as in (27, 28). ‘Coffee’ is not presupposed
-ACC boil-CAUS PROG-DEC
to exist in (27).‘Mina
Their metonymically-derived
is brewing coffee.’ (not ‘boiling’) versions follow them, as in
(27’) and (28’).
(27’) Mina-ka coffee-rul nayri -ko iss-ta

M-NOM -ACC brew PROG-DEC

‘Mina is brewing coffee.’

(27) Mina-ka coffee-rul

(28) Cathy is brewing (a cup kkulh- i -ko iss-ta
of) coffee.

M-NOM -ACC
(28’) Cathy is putting up (a potboil-CAUS
of) coffee. PROG-DEC
‘Mina4.isDiscussion:
brewing coffee.’
Telicity (not
as Conative and‘boiling’)
the Entailed Component as (CAUSE y to)

BECOME (∼) EXISTENT

(27’) Mina-ka coffee-rul nayri -ko iss-ta

M-NOM -ACC brew PROG-DEC
‘Mina is brewing coffee.’

(28) Cathy is brewing (a cup of) coffee.

(28’) Cathy is putting up (a pot of) coffee.

488 Chungmin Lee

4. Discussion: Telicity as Conative and the Entailed Component

as (CAUSE y to) BECOME (~) EXISTENT

We examined the complexity of the puzzle of the ambiguity residing

in the Progressive and the Resultative surrounding verbs of wearing.
The surface does not provide any clue to resolve the telic culmination
involved in the simple verb stem ip- ‘put on’/‘wear,’ followed by the simple
connective -ko ‘and (then),’ in the Resultative reading, which involves
complex meaning and constitutes an independent clause in a di-clausal
construction. A dress worn by y must [BECOME] EXISTENT at the end.
Gibsonian affordances we discussed do not require this endpoint of Theme’s
being existent.
Now let’s turn to the frustrative in the Progressive. Why are a frustrative
or adversative transitions required when the progression of the Progressive
is not smooth as planned or intended? The telic culmination is blocked on
the way. Even the traditional action theory requires DESIRE, BELIEF, and
ACTION for an action. That fits Borer’s (2005) telicity as conative. Copley
and Harley’s proposal must be linked to causal theory to be clear about
a situation to be a ceteribus paribus successor situation for some initial
situation and they don’t distinguish between Agents and simple (natural)
Causers (non-animate subjects are odd for action in many Asian languages).
Event theory together with ‘frame’ theory may be considered as an
alternative to solving the Imperfective Paradox. We can hardly see the
affordance, as discussed earlier, as "telic" (cf. Pustejovsky 2022). This
division must be rooted in the “frame” in AI (Moens and Steedman 1988
a.o.), viewing lexical meanings as scenarios. All accomplishments take the
same form of scenario, the differences being that the individual scenarios
involve different preparatory processes (run, build, etc.), incremental themes
(distance\((x)\, ‘run a mile’), house\((x)\), etc.), and resultant states. Mark
Steedman (e-mail Jan 2, 2023) states:
 “Imperfective Paradox” in Formal Model vs. Cognitive Force-theoretic Model 489

“The point that Copley and Harley (and other force-

theoreticians like Lambalgen and Hamm) miss is that telicity
is conative.4 Someone writing a poem or building a bridge
entails their intention to cause the poem or bridge to exist,
not their doing something that actually causes existence.
Intention can be frustrated, in a sense that causation cannot.
The progressive/imperfective latches onto the process of
writing with that intention. The perfective latches onto that
consequent state of existence. Affordance only comes into it
if I see some paper and a pen and think that it affords writing,
and use that fact to make a plan for the coming to existence of
my poem.”

We can hardly see the affordance as "telic" (cf. Pustejovsky 2022).

Let’s further consider such ‘perspective’ requiring examples, as cited
by Yeom (2003) from Landman (1992) ultimately borrowed from ter
Meulen (1987). Suppose that a plane just left for London, but that a team
of hijackers were aboard to take it to Havana. In this context, (29) and (30)
can be true; on (29) perspective, the event of flying to London starts when
the plane takes off and continues until it is hijacked. On (30) perspective,
the event of flying to Havana starts with the take-off of the plane with the
hijackers aboard and continues until it lands at Havana. This perspective
problem also seems to require ‘scenario’ and ‘frame’ solutions.

(29) Rebecca was flying to London when the plane was hijacked.

(30) R
 ebecca was flying to London; well, in fact, she wasn’t, she was
flying to Havana, but she didn’t know at that time.

4
connected with a wish, intention, or effort to do something’ in the dictionary.
490 Chungmin Lee

5. Concluding Remarks

We tried to understand Copley and Harley’s (2015) force-dynamic

theory of perfective and imperfective particularly in telic predicates. They
seem to have tried to simplify the theory involving ceteris (non) paribus
phenomena by force (and situation) in syntax and semantics but their
interface explanation seems to need further refinements. We presented the
case of the ambiguity of ip- ‘put on/wear’ between the Progressive and
the Resultative in Korean and Japanese to raise the problem and seek its
solution via ‘frame’ and scenarios as well.
We showed further points to consider --- telicity as conative and telic
culmination requiring the root of ((BECOME)) (~) EXISTENT), which
may be the ultimate conceptual root for the result state required by all
incremental Themes.

References

Borer, Hagit (2005) Structuring Sense: Volume II: The Normal Course of Events.
Oxford University Press.
Copley, Bridget, Harley Heidi (2015) A force-theoretic framework for event
structure. Linguistics and Philosophy, 38:103–158.
Copley, Bridget, Harley Heidi (2014) Eliminating causative entailments with the
force-theoretic framework: The case of the O’odham frustrative cem. In
Copley, Bridget, Fabienne Martin, eds. (2014).
Copley, Bridget, Fabienne Martin, eds. (2014) Causation in Grammatical
Structures. Oxford University Press.
Copley, Bridget and P. Wolff (2014) Theories of causation should inform
linguistic theory and vice versa. Copley and Martin (Eds.) Causation in
Grammatical Structures (pp 11-57).
Gardenfors, P. (2007) Representing actions and functional properties in
 “Imperfective Paradox” in Formal Model vs. Cognitive Force-theoretic Model 491

conceptual spaces. In Ziemke, Zlatev, & Frank (Eds.) Body, Language, and
Mind. Embodiment 1: 177-196.
Gibson, James, J. (1966) The senses considered as perceptual systems. Boston,
MA: Houghton Mifflin.
Koenig, J-P and L-C Chief (2008) Scalarity and state-changes in Mandarin
Chinese (and other languages). In Bonami and Hofherr (Eds.) Empirical
Issues in Syntax and Semantics 7 (241 -262). Paris: CSSP.
Hamm, Friedrich and Oliver Bott (2021) Tense and Aspect. In The Stanford
Encyclopedia of Philosophy, Edward N. Zalta (ed.), URL = <https://plato.
stanford.edu/archives/fall2021/entries/tense-aspect/>.
Hasegawa, Y. (1996) The (nonvacuous) semantics of TE-linkage in Japanese.
Journal of Pragmatics, 25: 6, 763–790.
Lee, Chungmin (1999) Aspects of aspect in Korean psych-predicates:
implications for psych-predicates in general in Tense-Aspect, Transitivity
and Causativity: Essays in honour of Vladimir Ned jalkov. Edited by
Werner Abraham and Leonid Kulikov [Studies in Language Companion
Series 50], Benjamins. 1999.
Lee, Chungmin (1973) Presupposition of existence of Theme for verbs of change
(In Korean and English). Foundations of Language 9: 3 (Jan., 1973) 384-
388.
Lewis, David K. (1973) Counterfactuals. Cambridge, MA: Blackwell.
Moens, M. and M. Steedman (1988) Temporal ontology and temporal reference.
Computational Linguistics 14: 2, 15-28.
Pustejovsky, J. (2022) ISO/TC 4 Language resource management. 4 N 1798 Doc
date 11-06. Semantic annotation framework (DemAF) – Part 10: Visual
Information (VoxML).
Reichenbach, H. (1971) The Direction of Time. Berkley: UC Press (originally
published in 1956)
Steedman, M. (2002) Plans, affordances, and combinatory grammar. Linguistics
and Philosophy, 25, No. 5/6 (Dec.), 723-753.
Talmy, Leonard (2000) Toward a Cognitive Semantics I and II. Cambridge, MA:
MIT Press.
Yeom, Jaeil (2003) The semantics of the English progressive and the imperfective
paradox. Language and Information, 7:2, 139-161.