Automatic Extraction of Aspectual Information 
from a Monolingual Corpus 
Akira Oishi Yuji Matsumoto 
Graduate School of Information Science 
Nara Institute of Science and TechnologT 
8916-5 Takayama, Ikoma, 
Nara 630-01 Japan 
{ryo-o, matsu}~is, aist-nara, ac. j p 
Abstract 
This paper describes an approach to ex- 
tract the aspectual information of Japanese 
verb phrases from a monolingual corpus. 
We classify Verbs into six categories by 
means of the aspectual features which are 
defined on the basis of the possibility of 
co-occurrence with aspectual forms and ad- 
verbs. A unique category could be identi- 
fied for 96% of the target verbs. To evalu- 
ate the result of the experiment, we exam- 
ined the meaning of -leiru which is one of 
the most fundamental aspectual markers in 
Japanese, and obtained the correct recog- 
nition score of 71% for the 200 sentences. 
1 Introduction 
Aspect refers to the internal temporal structure of 
events and is distinguished from tense, which has a 
deictic element in it, of reference to a point of time 
anchored by the speaker's utterance. There is a vo- 
luminous literature on aspect within linguistics and 
philosophy. Recently, computational linguists also 
have joined in the act within the context of machine 
translation or text understanding etc. For example, 
consider the following Japanese sentences (quoted 
from (Gunji, 1992)). 
(a). Ken-wa ima tonarino heya-de kimono-wo ki-te-i-ru. 
Ken-TOP now next room-LOC kimono-ACC put- 
on-PRES 
'Ken is now putting on kimono in the next room.' 
(b). Ken-wa kesa-kara zutto axto kimono-wo ki-te-i-ru. 
Ken-TOP this morning-since always that kimono- 
ACC weax-PRES 
'Ken has been wearing that kimono since this morn- 
ing.' 
(e). Ken-wa ano kimono-wo san-nen maeni ki-te-i-ru. 
Ken-TOP that kimono-ACC three-year before 
wear-PRES 
'Ken has the experience of wearing that kimono 
three years ago.' 
Notice that English translations use separate lex- 
ical items (put on for (a) and wear for (b), (c)) 
and different aspectual configurations (the progres- 
sive for (a), the perfect progressive for (b), and an- 
other for (c)), while all Japanese sentences contain 
the same verbal form ki-te-i-ru. Thus. when the sys- 
tem tries to translate these sentences, it must be 
aware of the difference among them. 
This paper describes an approach to extract the 
aspectual information of Japanese verb phrases from 
a monolingual corpus. In the next section, we will 
classify Japanese verbs into six categories by means 
of aspectual features following the framework of 
(Bennett et al., 1990). The aspectual forms land 
adverbs are defined as the functions which operate 
on verbs' aspectual features and changes their val- 
ues. By using the constraints of the applicability of 
the functions, we can identify a unique category for 
each verb automatically. If one can acquire aspec- 
tual properties of verbs properly and know how the 
other constituents in a sentence operate on them, 
then the aspectual meaning of the whole sentence 
will be determined monotonically. To evaluate the 
result of the experiment, we will examine the mean- 
ing of -teiru which is one of the most fundamental 
aspectual forms, since the classification itself is dif- 
ficult to evaluate objectively. 
2 Realization Process of Aspectual 
Meaning 
We consider that the whole aspectual meaning of 
verb phrases is determined in the following order: 
verbs ---, arguments ~ adverbs ~ aspectual 
forms, Adverbs and aspectual forms are defined 
as indicators of such cognitive processes as "zoom- 
ing" and "focusing" which operate on the time-line 
representation. They are sinfilar to the notions "as- 
pectual coercion" (Moens and Steedman, 1988) or 
I The term "form" refers to grammatical morphemes 
which axe defined in terms of derivation. In this paper, 
we refer to the aspectual morphemes which follow verbs 
as "aspectual forms", including compound verbs such as 
.hazimevu(begin), suffixes with epenthetic -re such as - 
teiru, and aspectual nominals such as -bakaviOust now) 
etc. 
352 
"views" (Gunji, 1992). We explain each in turn. 
2.1 Aspectual Categories of Verbs 
A number of aspectually oriented lexical-semantic 
representations have been proposed. ~Ve adopt 
and extend the feature-based framework proposed 
by (Bennett et al., 1990) in the spirit of (Moens 
and Steedman, 1988). They uses three features: 
±dynamic, ±telic, and ±atomic. We add two more 
features: ±process and ±gradual. 
The feature dynamicity distinguishes between 
states(-d) and events(+d), and atomicity dis- 
tinguishes between point events(+a) and extended 
events(-a). The duration described by verbs is 
twofold: an ongoing process and a consequent state. 
The feature process concerns an ongoing process 
and distinguishes whether events described by verbs 
have the duration for which some actions unfold. 
The feature telicity distinguishes between culmi- 
native events(+t) and nonculminative events(-t). 
It presupposes a process. The feature graduality 
characterizes events in which some kind of change is 
included and the change gradually develops. 
We can classify verbs by means of different com- 
binations of the five features. Since there are depen- 
dences between features, only subsets of the com- 
binatorially possible configurations of features are 
defined as shown in the Table 1. 
In the Table 1, 1.stative verbs are those that 
are not dynamic. 2.atomic verbs are those that 
express an atomic event. 3.resultative verbs ex- 
press a punctual event followed by a new state which 
holds over some interval of time. 4.process+result 
verbs are those that express a complex situation 
consisting of a process which culminates in a new 
state. 5.non-gradual process verbs are those 
that express only processes and not changes of state. 
6.gradual process verbs are those that have grad- 
uality. Although the verbs of the categories 5 and 
6 don't contain telicity, the arguments of the verbs 
or some kinds of adverbs can set up the endpoint of 
the process as discussed later. In Vendlerian classi- 
fication, states correspond to 1, achievements to 2 
and 3, accomplishments to 4 and 6, activities to 5, 
respectively (Vendler, 1957). 
Table 1: Aspectual categories of verbs 
2.2 Arguments 
Tenny points out that internal argument of a verb 
can be defined as that which temporally delimits or 
measures out the event (Tenny, 1994). 
The direct internal argument can aspectually 
• 'measure out the event" to which the verb refers. To 
clarify what is meant by "'mesuring-out", she gives 
examples of three kinds of measuring-out: incremen- 
tal theme verbs (eat an apple, build a house etc.), 
change-of-state verbs (ripen the fruit etc.) and path 
objects of route verbs (climbed the ladder, play a 
sonata etc.). 
On the other hand, the indirect internal argument 
can provide a temporal terminus for the event de- 
scribed by the verb. The terminus causes the event 
to be delimited as in push the car to a gas station. 
There is only one kind of internal argument, in terms 
of thematic roles, that does provide an event termi- 
nus, and that is a goal. 
In terms of the current framework, both of them 
add the telicity to the verb which does not inherently 
contain the telicity. They play a role of framing the 
interval on which the focus should be brought. 
2.3 Adverbs 
In general, adverbs focus on the subpart of the event 
described by a verb and give a more detailed de- 
scription. According to the discussion in (Moriyama, 
1988), adverbs can be classified as follows in terms 
of the subpart on which they focus. 
Processes modifiers modify verbs which have 
process (+p). This class includes reduplicative ono- 
matopoeia such as gasagasa, batabata, suisui, ses- 
seto, butubutu, etc., which are expressing sound or 
manner of directed motion, and rate adverbs such 
as yukkuri(slowly), tebayaku(quickly), etc., which ex- 
press the speed of motions. They focus on the on- 
going process of events described by verbs. 
Gradual change indicators express the 
progress of change of state, such as dandan (grad- 
ually), sukosizutu (little by little), jojom (gradually), 
dondon (constantly). sidaini (by degrees), etc.. which 
modify gradual process verbs (+g) and focus on the 
process. 
Continuous adverbs are those that can mod- 
ify both states verbs (-d) and process verbs (+p), 
such as zutto(for a long time), itumademo(forever), 
etc. They express a continuance of an event or a 
maintenance of a state. 
categories features examples 
1.stative 
2.atomic 
3.resultative 
4.process+result 
5.non-gradual process 
6.gradual process 
\[-d\] \[+d,+a\] 
\[+d,-a,-p\] 
\[+d,-a,+p,+t\] 
\[+d,-a,+p,-t,-g\] 
\[+d,-a,+p,-t,+g\] 
aru(be), sobieru(  se), sonzaisuru( e=isO 
hirameku(flash), mikakeru(notice) 
suwaru(sit down), tatu(stand up) korosu(kill), Urn(put on~wear), ake,' (open) 
aruku(walk), in(say), utau(sing) 
kusaru(turn sour), takamaru(become high) 
353 
Atomic adverbs make any events instantaneous, 
such as satto, ponto, gatatto, potarito, syunkan, etc., 
which express instantaneous sound emission or an 
instant. When these adverbs co-occur with verbs, 
the events are understood as instantaneous. This 
doesn't necessarily imply that the verb itself is in- 
stantaneous. 
Quantity regulators measure out events, such 
as gokiro aruku(walk 5kin). gojikan seizasita(sit 
straight for 5 hours), etc. These include time, dis- 
tance, and any quantity of contents. 
End state modifiers express the consequent 
state of events, such as mapputatuni(into two ex- 
act halves), konagonani(into pieces), pechankoni(be 
fiat), barabarani(come apart), etc. They focus on 
the resultant state. 
So far we have described adverbs which concern a 
single event, but some adverbs regulate the multiple 
events which involves iteration of a single event. By 
iteration, the whole process of a collective event can 
be taken up regardless of the inherent features of 
verbs. 
There are two kinds of Repetition adverbs: 
one regulates the whole quantity of the iteration 
of events such as san-kai(three times) or nan- 
domo(many times) etc., and the other describes the 
habitual repetition of events such as itumo(always) 
or syottyuu(very often) etc. Both describe many 
events each of which involves one person's act. 
Finally, we shall mention Time in the past ad- 
verbs. There are cases where the form -teiru, which 
marks the present tense, can co-occur with tempo- 
ral adverbs describing the past. (See the exan~ple 
(lc) in the introduction.) It describes the experien- 
tial fact of an event. Such adverbs as katute(once), 
mukasi(in the past) and izen(be/ore) determine the 
temporal structure of the event related with tense. 
2.4 Aspectual Forms 
The ability of aspectual forms to follow verbs is con- 
strained by the inherent features of verbs. We briefly 
describe some of aspectual forms used in the exper- 
iment. 
The forms -you-to-suru(be going to) and 
kakeru(be about to) take up the occurrence of events. 
They can follow the verbs which are dynamic(+d). 
The form -tuzukeru(continue)can follow the verbs 
which have duration(-a). It can take up either the 
ongoing process or the resultant state. The form - 
hajimeru(begin) can follow the verbs which have pro- 
cess(-bp) and takes up the start time of the process. 
On the other hand, the forms -owaru(cease) and - 
oeru(finish) can follow the verbs which are telic(+t) 
and takes up the end point of the process. However, 
these constraints on the inherent features of verbs 
are only concerned with a single event. By itera- 
tion, the whole process of a collective event can be 
taken up regardless of the inherent features of verbs, 
as mentioned above. 
The forms -tutuaru(be in progress), -tekuru(come 
into state) and -teiku(go into state) focus on the 
gradual process of change. -Tutuaru(be in progress) 
takes up it as a kind of state, -tekuru(come into 
state) views it from the end state of change while 
-teiku(go into state) from the initial state of change. 
Both of -tekuru and -teiku have usages other than 
aspect, as in mot-tekuru(bring) or mot-teiku(take). 
3 Experiment 
We carried out an experiment to classify Japanese 
verbs into six categories in the Table 1 by means of 
corpus data. 
As shown in the Figure 1, each category is defined 
in terms of the ability to co-occur with aspectual 
forms. However, the discrimination of the categories 
needs negative evidence which we cannot use by def- 
inition. A corpus only provides positive evidence. 
Furthermore, some forms can be used regardless of 
the features and have usages other than aspect as 
discussed in the previous section. ~Ve must establish 
a method which takes into these facts into account. 
 \oo -- 
2. atomic verbs +~p --/ \ 
+t //~ . t 3. resultative verb= =°/ \ 
process+result /~ 4" verbs +g/ ~-g 
4¢i1~ 
6. gr=dual process 5. non-gradual process verbs verbs 
Figure 1: The relation between categories of verbs 
and features 
3.1 Algorithm 
We used the EDR Japanese Corpus and the EDR 
Japanese Co-occurrence Dictionary (EDR, 1995) as 
material to extract syntactic clues in the experi- 
ment. The corpus contains 220,000 sentences from 
various genres of text. The results of the parsing 
analysis of these sentences indicates that the con- 
stituents of the sentence have a dependency struc- 
354 
STEP:I Pick out the items of which the governing and 
dependent words are a verb and an adverb from 
the EDR Co-occurrence Dictionary and store them 
with the frequency in an array called PAIRS (cf. 
Table 2). 
STEP:2 For each adverb in PAIRS, give an adverb class 
label (the initial letter of the class name) on the ba- 
sis of the discussion in sec. 2.3 and store them in an 
array called ADVERBS (cf. Table3 and Table4). 
STEP:3 For each verb in PAIRS, add up the frequency 
of the co-occurrence with the adverbs contained in 
the array ADVERBS. If the sum is greater than 4, 
store the verb in a list called VERBS. 
STEP:4 For each sentence in the corpus, find a verb 
and if it is contained in VERBS, then: 
STEP:4°I If the form following the verb is con- 
tained in the predefined list (Table5), make 
an array FORMS\[/,j\] positive (where i is the 
position of the verb in the list VERBS and j 
is the position of the form in the Table 5, see 
Table6), provided that the verb is not modi- 
fied by repetition adverbs(R). When the form 
is -tekuru or -teiku, put it on record only if 
the verb is modified by gradual change indica- 
tors(G). 
STEP:4-2 If the verb is modified by the adverbs 
contained in the array ADVERBS, refer to 
the adverb class label and add 1 to an array 
MODIFIED\[i, k\] (where i is the position of the 
verb in the list VERBS and k is the position of 
the adverb class label in the Table4. When the 
adverb is continuous one(C), distinguish the 
cases where the verb is followed by -teiru(C1) 
from the other eases(C2), see Table7), pro- 
vided that the verb is not followed by negative 
forms such as -nai or -zu, nor the forms which 
change the voice such as -reru(the passivizer) 
or -seru(the causativizer), since they affect the 
aspeetual properties of the attached verb. 
STEP:5 For each verb in VERBS: 
STEP:5-1 Narrow down the candidates by means 
of the array FORMS (on the basis of possible 
categories shown in Table 5). 
STEP:5-2 In the ease where the category of the 
verb cannot be uniquely identified in STEP:5- 
1, i.e., other than the category 6, determine it 
by means of the array MODIFIED as follows: 
the category 6 
the category 5 
the category 4 
the category 3 
the category 2 
the category 1 
ambiguous 
ture. That is. the constituents have a governing- 
dependent relation. It is these constituents that 
form the head phrases of the Japanese Co-occurrence 
Dictionary which describes collocational information 
in the form of binary relations. Each item in the 
Japanese Co-occurrence Dictionary consists of a gov- 
erning word. a dependent word, the relator between 
the words, and supplementary co-occurrence item 
information which is composed of the frequency of 
the co-occurrence relation and a portion of the ac- 
tual example sentence from which the co-occurrence 
relation was taken. 
The algorithm used for classifying verbs is shown 
in Figure 2. 
Table 2: A part of the array PAIRS 
I "d~b I verb I r~q • I 
an(like that) tu(say) 1 
an(like that) suru( do ) 1 ai (mutually) au(rneet ) 1 
a~kawarazu(as usual) iru(be) 1 aikawarazu(as usual) otituku(settle) 1 
aituide(one after another) sannyuusuru(join) 3 
aituide(one after another) seturitusuru(establish) 4 
Table 3: A part of the array ADVERBS 
adverb I label \] 
aikawarazu(as usual) C 
aegiaegi(gasping) P 
akaakato(brightly) P 
akuseku(busily} P 
atafuta(in a hurry} P 
atafutato(in a hurry} P 
attoiuma(in an instance) A 
ikiiki(vividly) P 
(if the verb is modified by gradual change indicators(G)) 
(if modified by process modifiers(P) and not by end state 
modifiers(E)) 
(if modified by both process modifiers(P) and end state 
modifiers(E)) 
(if modified by end state modifiers(E) and not byprocess 
modifiers(P)) 
(if modified by only atomic adverbs(A)) 
(if modified by continuous adverbs without being followed 
by .teiru(C2) and not modified by process modifiers(P) nor 
gradual change indicators(G) nor end state modifiers(E)) 
(otherwise) 
Figure 2: The algorithm for classifying verbs 
355 
Table 4: Results of the classification of adverbs adverb class(label) 
process modifiers (P) 
gradual change indicators (G) 
continuous adverbs (C) 
atomic adverbs (A) 
quantity regulators (Q) 
end state modifiers (E) 
repetition adverbs (R) 
' | U\[|ll I| I\[| .~l~ I'\]~i .Ill| 
\] total I examples 
470 yukkuri(slowly), gasagasa, batabata, sui~ui, sesseto, butubutu,... 
52 sidaini(gradually), masumasu(increasingly), jojoni(gradually)... 
78 sonomama(as it is), zutto(for a long time), itumademo(forever)... 
294 satto, ponto, gatatto, potarito, syunkan(instantaneously)... 
12 180-do(180 degree), ippo(a step), ippai(a cup)... 
86 mapputatuni(into two exact halves), konagonani(into powder)... 
122 nandomo(many times), itumo(always), syottyuu(very often)... 
Table 5: The aspectual forms used in the experiment 
forms followable verb categories 
-you-to-suru(be going to),-kakeru(be about to) 
- tuzukeru(continue) 
-hajimeru(begin) 
-owaru(end) and -oeru(finish) 
-tutuaru(be in progress), -tekuru(come into state), 
-teiku(go into state) 
2, 3, 4, 5, 6 
3, 4, 5, 6 
4,5,6 
5,6 
verb 
akkasuru(become worse) 
nigiru(catch) 
anteisuru(become stable) 
isikisuru(become conscious) 
kotonaru(differ) 
idousuru(move) 
ijisuru(maintain) 
tigau( differ) 
sodatu(grow) 
sodateru(bring up) 
ittisuru(agree) 
Tabl 6: A part of the array FORMS 
-kakeru 
+ 
- + 
+ 
- + 
- + 
- + 
+ 
- + 
forms 
I -tuzu er  l -hajime   l -owa,  I -tutuaru 
- + 
- + 
- + 
- + 
- + 
m 
Table 7: A part of the array MODIFIED 
verb adverb class labels PIGICl C2\]AIQIE 
akkasuru(become worse) 0 5 0 0 1 0 0 
nigiru(eatch) 0 1 0 1 0 0 1 
anteisuru(become stable) 0 1 1 1 0 0 1 
isikisuru(become conscious) 0 1 0 1 0 0 0 
kotonaru( differ) 0 1 0 0 0 0 1 
idousuru(move) 1 1 0 1 1 0 0 
ijisuru(maintain) 0 0 0 4 0 0 0 
tigau(differ) 0 1 0 0 1 0 0 
sodatu(grow) 5 3 0 0 0 1 1 
sodateru(bring up) 3 1 0 1 0 0 0 
ittisuru(agree) 0 0 0 0 3 0 2 
356 
The steps 1, 2 find 3 are the processes to deter- 
mine the target verbs. There are 431 verbs modified 
by the classified adverbs more than 4 times. In step 
2, we classify adverbs on the basis of the discussion 
in the previous section. Although the classification 
has been done by hand, it is much easier than that 
of verbs, since adverbs are fewer than verbs in num- 
ber (2,563 vs. 12,766 in the corpus) and have higher 
"iconicity" -- the isomorphism between form and 
meaning -- than verbs. This classification of ad- 
verbs is used not only for determining the aspectual 
categories of verbs but also for examining the mean- 
ing of -teiru as mentioned later. 
The step 4 is a process to register the co-occurring 
forms and adverbs for each verb. By using these 
data, we identify the aspectual categories of verbs in 
the step 5. Since the categories cannot be uniquely 
identified by aspectual forms only, we use adverbs 
which can modify the only restricted set of verbs as 
shown in Table 8. 
Table 8: 
categories 
adverb cl ass (Ta-b'e--i~ 
Adverb classes and their modifiable verb 
verb cate~ 
process modifiers (P) 
gradual change indicators (G) 
continuous adverbs (C) 
atomic adverbs (A) 
quantity regulators (Q) 
end state modifiers (E) 
4,5,6 
6 
1,3,4,5,6 
2,3,4,5,6 
1,3,4,5,6 
3, 4, 6 
3.2 Evaluation and Discussion 
Out of 431 target verbs, we could uniquely identify 
categories for 375 verbs. As for the rest 56 verbs, 37 
verbs were identified in the step 5-2 as the category 
which was not included in the set of categories out- 
putted by the step 5-1. This seems to be due to the 
failure to detect the expression of repetition, there- 
fore, we chose the category determined in the step 
5-2. Table 9 shows the results. 
We confirmed that more than 80% of verbs are 
correctly classified. However, this is a subjective 
judgement. To evaluate the results of the classifi- 
cation more objectively, we focus on one evaluation 
metric; namely the automatic examination of the 
meaning of -teiru which can represent several dis- 
tinct senses as described in the introduction. 
The form -teiru indicates "zoom in" operation: it 
is a function that takes an event as its input and 
returns a type of states, which refers to unbounded 
regions i.e., a part of the time-line with no distinct 
boundaries. Figure 3 shows the time-line representa- 
tion for each aspectual category of verbs. Aspectual 
distinctions correspond to how parts of the time-line 
are delineated. 
1. staUvo verbs 
t ) t l (1) (2) 
2. atomic verbs 
......................... --£) .......................... ; ........ ; 
(3) 3. resultatlve verbs 
............................. 
t ~' (4) t I, (5) 
4. process+result verbs 
............. © © 
t J ~__I 4 J 
(s) (7) (e) 
5. non-gradual procese verbs 
..... ---(3 t J ................... ; .... -i 
(9) (10) 
gradual process verbs 
t # 
t ) (12) (11) 
Figure 3: The time-line representation for aspectual 
categories of verbs 
In Figure3, thick line segments signify regions, 
dashed line segments signify unbounded ends of re- 
gions and large open dots signify points in time 
boundaries or punctate events). 
Table 9: The verb classification obtained by the ex- 
periment 
\[ verb catel~ory \[ no. I examples 
1.stative 30 
2.atomic 19 
3.resultative 29 
4.process+result 30 
5.non-gradual process 94 
6.gradual process 210 
ambiguous 19 
mitumeru(stare) ijisuru(maintatn) sumu(live) sonzaisuru(ezist) 
nagameru(view, damaru(be silent) kumkaesu(repeat) tukaeru(can be used) ... 
nageru(throw) haneagaru(leap up) kizuku(notiee) mikakeru(happen to see) 
gouisuru(arrive at an agreement) kireru(snap) furnikiru(launch out) ... 
nureru(become wet) turnaru(become packed) tunagaru(make a connection) au(meet) 
suwaru(sit down) tatamu(fold) kureru(get dark) atehamaru(fit) ... 
tateru(build) nobasu(lengthen) rnatomeru(put together) narabu(form a line) 
tutumu(wrap) majiwaru(associate) tiru(fall) torikakomu(surround) ... 
nomu(drink) hakobu(carry) tanosimu(enjoy) kansatusuru(observe) furueru(shake) 
hibiku(ring) tobimawaru(fly about) taberu(eat) sugosu(spend)... 
akkasuru(get worse) tuyornaru(get strong) takarnaru(become raised) 
sinkoukasuru(get more acute) seityousuru(grow up) kappatukasuru(make active) ... 
kuwawaru(join) tutomeru(be employed) tomonau(accompany) tazuneru(visit) 
rainitisuru(eome to Japan) uwamawaru(be more than) hokoru(boast) ... 
357 
Since -teiru cannot include a time instant at which 
a state is drastically changed, it must denote one of 
the intervals depicted below the lines. The interval 
(1) in Figure3 designates a state which is a part 
of the state described by a lex_ical stative verb. It 
means a state holding before a speaker's eyes. 
It has been stated from (Kindaichi, 1976) that the 
form -teiru has three distinct senses: "a simple state', 
'a progressive state' and 'a consequent state'. (1) 
corresponds to a simple state. (4) and (7) to a con- 
sequent state, (6), (9) and (11) to a progressive state. 
respectively. Though not represented in Figure 3, a 
consequent state can be taken up with the verbs of 
categories 5 and 6 if the endpoints of the processes 
are set up by explicit expressions. 
Kudo (Kudo, 1982) has pointed out that there are 
inherent meaning and derivative meaning for both 
progressive and consequent states and has sorted out 
them as follows. 
(i) inherent meaning of 'a progressive state': an 
ongoing process 
(ii) derivative meaning of 'a progressive state': an 
iteration 
(iii) inherent meaning of 'a consequent state': a re- 
sultative state 
(iv) derivative meaning of 'a consequent state': an 
experiential state 
(v) otherwise: a simple state 
(ii) is the above-mentioned process of a collective 
event; "a line as a set of points", so to speak. (iv) 
is a state where a speaker has an experience of the 
event described by a verb and corresponds to the 
intervals (2), (3), (5), (8), (10), (12) in Figure3. 
These derivative meanings are conditioned syntacti- 
cally or contextually, that is, they are stipulated as 
derivative by explicit linguistic expressions such as 
adverbials etc., while not concerned with the inher- 
ent features of verbs -- they can appear with most 
of verbs regardless of their aspectual categories. 
We carried out an experiment to examine the 
meaning of -teiru automatically by means of the clas- 
sifications of verbs and adverbs obtained in the pre- 
vious experiment. Table 10 shows the determination 
process of the meaning of -teiru. We checked the 
cases in Table 10 downward from the top. 
Table 11 shows the results obtained from running 
the process of Table 10 on 200 sentences containing 
-teiru which are randomly selected from the EDR 
Japanese Corpus. 
The precision on the whole is 71%. Note that 
the sense (i) 'an ongoing process' has high recall but 
low precision, while (iii) 'a resultative state' and (iv) 
'an experiential state' show the opposite. This is 
due to the fact that the test sentences contain many 
"speech-act" verbs such as syuchousuru(insist), se- 
tumeisuru(explain), hyoumeisuru( declare) etc. They 
are classified as 5.non-gradual process verbs, and by 
Table 10: The determination process of the meaning 
of -teiru 
case output 
(1).the verb is modified by repetition (ii) an iteration 
adverbs( R} 
(2).the verb is modified by time in 
the past adverbs(P) or its category 
is 2. atomic verbs 
(3).the category of the verb is 
1. stative verbs 
(4).the category of the verb is 
3. resultative verbs 
(5).the verb is modified by process 
modifiers(P} or gradual change 
indicators (G} 
(6).the endpoint of the process is 
explicitly set up (the verb is modified 
by end state modifiers(E) ot quantity 
regulators(Q) or it takes a goal 
arsument i.e., ni(~o)-case etc. 
(7).the process cannot be taken up 
(the verb is modified by atomic 
adverbs(A) or sudeni(already), etc.) 
(iv) an experiential 
state 
(v) a simple state 
(iii) a resultative state 
(i) an ongoing process 
(iii} a resultative state 
(iii) a resultative state 
(8).the category of the verb is (i} an ongoing process 
5. non-gradual process or 6. gradual 
process verbs 
(9).the category of the verb is ambiguous: (i) or (iii) 
4. process+result verbs 
the case 8 in Table 10, the senses of -teiru follow- 
ing them are determined as (i) 'an ongoing process'. 
However, they takes a quotative to-case that marks 
the content of the statement and this measures out 
the event described by verbs. Therefore the resulta- 
tive or experiential readings are preferred. 
The other errors are caused by polysemous 
verbs such as kakaru (hangflie//all...) or ataru 
(hit/strike~be exposed/shine...). Their aspectual 
properties are changed by the complements they 
take. The analysis of how complements influence 
the aspectual properties of their governing verbs is 
beyond the scope of this paper. It seems to be a mat- 
ter of pragmatic world knowledge rather than sense- 
semantics (but see (Verkuyl, 1993) for English). 
4 Related Work 
The approach proposed here is similar to that of 
Dorr's (Dorr, 1992: Dorr. 1993), but different from 
it in scale and determinability of the categories. She 
adopts the four-way classification system following 
Vendler (Vendler, 1957) and utilizes Dowty's test 
(Dowty, 1991) for deternfining aspectual categories 
of English verbs. She reports the results obtained 
from running the program on 219 sentences of the 
LOB corpus. Although we cannot know how many 
verbs she tested because she has shown only a subset 
of the verbs, the program was not able to pare down 
the aspectual category to one in 18 cases out of 27 
verbs. 
Brent (Brent, 1991) discusses an implemented 
program that automatically classifies verbs into two 
groups, stative vs. non-stative, on the basis of their 
syntactic contexts. He uses the progressive and rate- 
358 
Table 11: The restdts of the evaluation ex)eriment 
the sense of 
-te=ru 
judgement 
by human(a) 
output of 
program(b) 
number of 
agreements(c) 
recall(%) 
c/a x 100 
precision(%) 
c/b x 100 
(i) an ongoing process 95 137 88 93 64 
(ii) an iteration 4 2 2 50 100 
(iii) a resultative state 
(iv) an experiential state 
(v) a simple state 
29 48 93 
39 15 14 36 93 
19 19 15 79 79 
ambiguous 14 12 9 64 75 
total 200 200 142 71 71 
adverbs constructions in combination with some sort 
of statistical smoothing technique. He identified 
eleven verbs as purely stative, of the 204 distinct 
verbs occurring at least 100 times in the LOB cor- 
pus. 
We think that the extraction of aspectual infor- 
mation must be based on principles that are well- 
grounded in linguistic theory. However, some sort 
of noise reduction technique such as the confidence 
intervals used by Brent may be needed to detect the 
cue more accurately. 
5 Conclusion 
In this paper, we have proposed a method for classi- 
fying Japanese verbs on the basis of surface evidence 
from a monolingual corpus, and examined the mean- 
ing of the form -teiru by means of the classifications 
of verbs and adverbs. 
The aspect of verb phrases provides not only the 
temporal configuration within a single event but also 
the information needed for processing temporal re- 
lation between multiple events (Dowty, 1986; Pas- 
sonneau, 1988; Webber, 1988). 
Furthermore, the lexical aspect of verbs is closely 
related with their deep complement structures which 
may not be directly reflected on the surface argu- 
ment structures. Therefore, by combining the aspec- 
tual categories of verbs and those that are defined in 
terms of their surface argument structures, we can 
obtain an elaborate classification based on seman- 
tic types of verbs. (Preliminary experiments on this 
issue can be seen in (Oishi and Matsumoto, 1996).) 
Thus, the information obtained here can be used 
for various applications. 

References 
S. W. Bennett, T. Herlick, K. Hoyt, J. Lifo, and A. Santisteban. 
1990. A computational model of aspect and verb semantics. 
Mashine 7~ranslation, 4(4):247-280. 
M. It. Brent. 1991. Automatic semantic classification of verbs 
from their syntactic contexts: An implemented classifier for 
stativity. In Proceedings of the 5th ACL European Chapter, 
pages 222-226. 
B. J. Dorr. 1992. A parameterized approach to integrating aspect 
with lexical-semantic for machine translation. In Proceedings 
of the 30th Annual Meeting of ACL. pages 257-264. 
B. J. Dorr. 1993. Machine Translation -- A View from the 
Lezicon. The MIT Press. 
D. R. Dowty. 1986. The effects of aspectual class on the temporal 
structure of discourse. Linguistics and Philosophy, 9(1):37- 
61. 
D. R. Dowty. 1991. Word Meaning and Montague Grammar : 
The Semantics of Verbs and Times in Generative Semantics 
and in Montague's PTQ, volume 7 of Studies in Linguistics 
and Philosophy(SLAP). Kluwer Academic Publishers. 
Japan Electronic Dictionary Research Institute Ltd. EDIt. 1995. 
the EDR Electronic Dictionary Technical Guide. (in 
Japanese). 
T. Gunji. 1992. A proto-lexical analysis of temporal properties of 
japanese verbs. In Linguistics Studies on Natural Language, 
Kyung Hee Language Institute Monograph One, pages 197- 
217. Hanshin Publishing. 
H. Kindaichi. 1976. Nihongo Dousi-no Asupekuto ('Aspect of 
Japanese Verbs'). Mugi Shobo. {in Japanese). 
M. Kudo. 1982. Siteiru-kei-no imi-kijutu ('the description of 
the meaning of the form -teiru'). Muzashi University Jinbun 
Gakkai Zasshi, 13(4). 
M. Moens and M. Steedman. 1988. Temporal ontology and tem- 
poral reference. Computational Linguistics, 14(2):15-28. 
T. Moriyama. 1988. Nihongo Doushi Jutsugobun no Kenkyuu 
('A Study of Japanese Verb-pradicate Sentences'). Meiji 
Shoin. (in Japanese). 
A. Oishi and Y. Matsumoto. 1996. Detecting the organization 
of semantic subclasses of Japanese verbs. Technical Iteport 
NAIST-IS-TIt96019, Nara Institute of Science and Technology. 
It. J. Passonneau. 1988. A computational model of the semantics 
of tense and aspect. Computational Linguistics, 14(2}:44-60. 
C. L. Tenny. 1994. Aspectual Roles and the Syntaz-Semantics 
Interface, volume 52 of Studies in Linguistics and Philoso- 
phy(SLAP). Kluwer Academic. 
Z. Vendler. 1957. Verbs and times. Philosophical Review, 
66:143-160. 
H. Verkuyl. 1993. A Theory of Aspectuality. Cambridge Uni- 
versity Press. 
B. L. Webber. 1988. Tense as discourse anaphor. Computational 
Linguistics, 14(2):61-73. 
