Zero Pronouns and Conditionals 
in Japanese Instruction Manuals 
Tatsunori Mori and Hiroshi Nakagawa 
Division of Electrical and Computer Engineering, l"acnlty of Engineering, Yokohama National University, 
Tokiwa-dai 156, llodogztya-kn, Yokoh,~ma 240, JAPAN 
E-maih mori@forest.dnj.ynu.ac.jp, nakagawa@naklab.dnj.ynu.ac.jp 
Abstract 
This paper proposes a method of the 
zero pronoun resohition, which is one 
of the essential processes in understand- 
ing systems for Japanese manual sen- 
tences. It is based on pragmatic proper- 
ties of Japanese conditionals. We exam- 
ined a uumber of sentences appearing in 
Japanese manuals according to the clas- 
sillcation based on the types of agent and 
the types of verb phrase. As ~ result, we 
obtained the following pattern of usage 
in matrix clauses: 1) The connective par- 
ticles TO and REBA have tt, e same distri- 
bution of usage. TARA and NARA have 
the same distribution of usage. 2) '\['he 
distribution of usage of TO and REBA, 
and that of TARA and NARA are com- 
plementary to each other. We show that 
these distributions of usage can be used 
for resolution of zero subjects. 
1 Introduction 
From simple electrical appliances to complex com- 
puter systems, almost all machines are accom- 
panied by instruction manuals. Since recently 
there are many machines whose operating pro- 
cedures are complicated, we have much trouble 
in many cases including translating their manuals 
into other languages, maintaining consistency be- 
tween the description in manuals and the actual 
behavior of the machines. To solve these prob- 
lems, we have to have a computer assisted system 
tbr processing Japanese manual sentences, espe- 
cially tbr understanding manual sentences. 
A large number of researchers have gotten to 
grip with the method of understanding some types 
of text inehlding instruction nlanuals(Abe et al., 
1988; Nomura, 1992; Eugenio, 1992). One of the 
most important matters of concern in tliese types 
of system is how we can fix ambiguities in seman- 
tic representations and fill uuderspecified parts of 
them. Generally speaking, almost all systems de- 
scribed above take the following scheme, l"irstly, 
each sentence in'a text is translated into a seman- 
tic representation, hi this process, the system uses 
only non-defeasible syntactic and semantic coll- 
straints. Most of pragmatic information and coln- 
rnousense knowledge are not used here, because 
the result of these knowledge would be overrid- 
den by some other information such as contex- 
tual intbrmation. Therefore the semantic repre- 
sentation would include some undetermined parts 
which would be fixed by other kind of information 
including context. This way of analysis is known 
as the Noudcfeasibility Thesis(Kameyama, 1995). 
Secondly, all of undetermined parts of the seman- 
tic representation are filled or settled by some kind 
of inferences based on ttie donlain knowledge. 
This type of method, which uses a, large ~%lnollut 
of domain knowledge, seems to be dominant froni 
the viewpoint of disambiguation. Moreover it 
scarcely depends on the language in use becmlsc 
the way of disambiguation is based oil the infer- 
ence with a certain knowledge base. On the otlLer 
hand, ill order to use this method, we have to pre- 
pa.re the amount of knowledge being large euough 
to cope with various type of described objects. 
Unfortunately, so far we have not had such a com- 
monsense knowledge base. 
One of ways to get rid of this situation is to 
adopt some knowledge which hardly depends on 
some particular domain. As such a kind of knowl- 
edge, we pay our attention to pragmatic con- 
straints, which haw.' not been used sufficiently in 
the former methods. We expect that by prag- 
matic constraints the ambiguity in manual sen- 
tences would be resolw;d to some extent not in 
the process of inference but in the process of the 
translation of manual sentences into semantic rep- 
resentations. 
We do not commit ourselves to the domain spe- 
citlc knowledge, but use some ontological knowl- 
edge in general manuals. For example, the col 
respondence of objects in the mamlal sentences 
to the objects in linguistic coustra.ints, like the 
speaker, the hearer, and so on. Note tha.t tile on- 
tology in this paper does not refer to all of objects 
in the world described by manuals, like a certain 
part of machine. Aiming at iridependence from 
the doniain knowledge of objects, we adopt erie of 
general ontologies which is applicable to almost all 
manuals. In short, our scheme consists of tile fol- 
lowing three parts: 1) a parser based on tim non- 
defeasiblity thesis, 2) pragmatic constraints spe- 
cific to linguistic expressions, and 3) the general 
ontology of the worhl described by tnanuals. 
In the rest of this paper, we will focns on 
the zero pronoun resolution. In Jal)anese , zero 
pronouns frequently make a sentence ambiguous. 
Zero pronouns are ellipsis of obligatory ca.ses, 
which very frequently appear in Japanese sen- 
782 
tences. F, specially, subjects are omitted very o1: 
ten. It is called "zero subject." In some sense, the 
resolution of zero pronouns' referents, especially 
the resolution of "zero subject", is the essential 
part of the knowledge extraction fi'om JaI)atmse 
illanuals~ becanse once referents of zero prOllOtlllS 
are identified, we can use w~rious methods already 
been l)roposed to recognize the structure of sen- 
tence and to map it into the suitable knowledge 
representation. To capture pragmatic constraints, 
we have paid our attention to conditionals, which 
occur very frequently in instruction manuals. In 
this paper, we will show that in instruction man- 
uals, the constraint of conditionals can I)e used to 
identify the referents of zero subjects. Although, 
of course, not all the zero pronomls can t)e solved 
with the constraints shown in the paper, our ex- 
amination for a lot of manual sentences shows 
that the constraints work very effectively and ac- 
curately in sentences with conditionals. 
Now we have to deline the term 'subject' we 
used in this paper. Generally, the term 'subject' 
d('notes a nominative from the grammatical point 
of view. In this paper, however, we will use the 
term SUBJECT to denote a main participant of the 
sentence. Roughly speaking, in tile active voice, 
the SUBJECT is the nominative., on the other hand, 
in the passive voice, the SUBJECT is the nomina- 
tive of the corresponding sentence in the active 
voice. 
2 Zero pronouns in manual 
sentences 
Let's consider the following Japanese sentence, 
which shows a certain instrnction. 
(1) ¢~, kono-botan-o osu -to, 
(~a-N()M this-button-ACe push -TO 
¢b der -are -mas -u. 
qSb-NOM go ollt -car| -I'Ol,-NONPAST. 
If lb, push(es) this button,then ¢b can go oul.. 
Native speakers of Japanese have the following in- 
tuitive interpretation for (1) without any special 
context. 
(2) ¢. : = the hearer (= the u, er) 
Here, 'TO' is a Japanese conjunctive particle 
which represents a causal relation. 'MASU' shows 
politeness, Milch is expressed by cot, iu(l). The 
'ARE' shows ability or permission. 
On the other hand, the following sentence, 
which does uot have the verbal suffix of possi- 
bility 'ARE' in the matrix clause, has a different 
interpretation. 
(3) ¢c kono-botan-o osu -to, 
(~c-N()M this-button-Ace push -TO, 
Ca de -mas -u. 
q~d-NOM conic out -POL -NONPAST. 1 
If ¢¢ push(es) this button, then Ca will come 
out. 
The zero pronoun Ca does not refer to tim 
hearer(the user), even though qS~ refers to the user 
as well a.s (1). The intuition of native sl)eal~ers of 
Japanesc for (3) is that Cd refers to a machine or a 
certain part of the machine. Note that when only 
the matrix clause of (3) is use(t as shown in (4), 
¢,. can lie interpreted as either the hearer or the 
machine ~, 
(4) gb e de -lnas-u. 
(/)e-NOM go out -POl, -NONPAST. 
¢~ will go out. 
These examples show that the expressions TO and 
ARE impose some constraints on the referents of 
SUBJECTS of the sentences. As described above, 
there are many cases that linguistic expressions 
give us a key information to resolve some tyl)e of 
ambiguity like the a.nal)hora of a zero pronoun. In 
the rest of this paper, we will show several prag- 
matic constraints, which can aceonllt for the in 
terpretations of these sentences described above. 
I)ohsaka(l)ohsaka, 1994) l)roposes a similar ap- 
proach, in which pragmatic constraints are used 
to determine rethrents of zero pronouns. While 
his apl)roach treats dialogue, our targets are man- 
ual sentences. His approach utilizes honorific ex- 
pressions and the speaker's point of view. Since 
the constraints are efl'ective in the (lifferent target 
from ours, the accuracy of identifying the referents 
of zero pronouns would be improved much more 
by using both of his constraints and the constraint 
we proposed. As for the identifying method 
available in general discourses, the centering the- 
ory(Brennan et al., 1987; Walker et al., 1990) and 
the property sharing theory(Kameyama, 1988) are 
proposed. Although this kind of theory has a 
good point that it is independent of the type o17 
discourse, the linguistic constraints specitic to ex- 
pressions like the pragmatic constraints l/roposed 
by Dohsaka or us are more accurate than theirs 
when the speeitlc constraints are applicable. 
3 General ontology in manuals and 
prinmry constraints 
In this section, we consider the general ontology 
which can be used in ,dl types of manuals. 
We shouhl consider two types of information as 
the parts of ontology: the properties of the ob- 
jects in manuals and the discourse situation that 
is characterized by linguistic roles like a writer and 
a reader. 
Constraint 1 (Objects) 
User has intention. 
Manutheturer has intention. 
Machine has no intenlion. 
1The English translation of 'DERU' ill (3) is dif- 
ferent from the translation ill (1). It is due to the 
ditference of tile viewl)oint be.tweei, Japanese and Eli- 
glish. The difference has no effect on the selection of 
zero prOllOllll~S refercllt. 
2It seems to be more natural that Ce is interl)reted 
as the hearer. 
783 
Constraint 2 (Discourse Situation) 
Speaker(Writer) = Manufacturer 
Hearer(Reader) = User 
From these constraints of the ontology, we can 
obtain the constraint of persons as follows. 
Constraint 3 (Persons) 
First Person = Manufacturer 
Second Person -- User 
Third Person = Machine 
Before considering the constraints of Japanese 
conditionals, we had better mention the more ba- 
sic expressions in manuals. In Japanese, simple 
operation procedures, like those which do not in- 
elude some conditions, are often described as sim- 
ple sentences with no subjects whose verbs arc of 
one of the following types: the RU form, the re- 
quest form or the solicitation form. The RU form is 
the basic form of verbs and it denotes the non-past 
tense. Since the RO form has a neutral meaning, it 
does not impose any restriction on the SUBJECT. 
On the other hand, the request form and the 
solicitation form have some constraints. The 
speaker uses the sentences to prompt hearers to 
do an action described by the sentence. There- 
fore, we have the following constraint. 
Constraint 4 (SUBJECT of sentence in the 
request form) 
The SUBJECT of a sentence in either the requesl 
form or the solicitation form is the hearer, 
Manual sentences may have a kind of modal- 
ity expressing the permission, the possibility, the 
obligation, and so on. Sentences which have the 
expressions of ability or permission mean not. only 
that it is possible for the SUBJECT of the sentence 
to do the action, but also thai; the SUBJECT has 
their choice of whether to do the action or not to 
do it. Therefore, we have the following. 
Constraint 5 (SUBJECT of sentence with 
ability expressions) 
A SUBJECT of a sentence with the expressions of 
ability or permission must have his~her intention 
to make a choice about the action described by the 
.sentence. 
4 Semantics of Japanese 
Conditionals 
Japanese has four conditional particles, TO, REBA, 
TARA and NARA, which are attached to the end of 
subordinate clauses as described in (1). The sub- 
ordinate clause aud the matrix clause conjoined 
by one of these particles correspond to the an- 
tecedent and the consequence, respectively. Each 
expression has its own meaning as shown in Ta- 
ble l(Masuoka, 1993). TARA and NARA are very 
rarely used in manual sentences as far as we exam- 
ined. For example, the rates of use of each condi- 
tional in over a dozen of instruction manuals are as 
follows3: TO is 77.6 % (385 sentences) of all condi- 
tionals, REBA is 19.4 % (96 sentences), TARA is 2.6 
3As described later, we have examined several 
other manuals especially for the consideration of the 
% (13 sentences) and NARA is 0.4 % (2 sentences). 
Roughly speaking, TO and REBA show causality 
relations, namely some general rules, and TARA 
and NARA are used in the case that the the an- 
tecedent is an assumption. The fact that not as- 
sumptions but general rules are usually described 
in the context of instruction is the reason why 
TARA and NARA are used less fi'equently than TO 
and REBA. 
The difference of constraints of these expres- 
sions are shown in the following sentences, which 
are the variants of the sentence (3). 
(5) ¢i kono-botan-o use -ba, 
(~i-NOM this-button-ace push -REBA, 
ej de -mas -u. 
(fij-NOM come out -POL -NONPAST. 
If ¢i push(es) this button, then ej will come 
out. 
(6) ¢k kono-botan-o osi -tara, 
ek-NOM this-button-Ace push -TARA, 
et de -mas -u. 
~/-NOM come out/go out -POL -NONPAST. 
If ek push(es) this button, then et will come 
out/go out. 
(7) em kom>botan-o osu -nara, 
em-NOM this-button-Ace push -NARA, 
en de -mas -u. 
en-NOM come out/go out -POL -NONPAST. 
If em push(es) this button, then ¢,~ will come 
out/go out. 
As well as the sentence (3), for Japanese native 
speakers, the SUBJECT of the matrix clause of (5) 
should be a machine. On the other hand, in the 
case of the sentences (6) and (7), the SUBJECTS 
of the matrix clauses can be either users or ma- 
(:hines. These phenomena probably due to the na- 
ture of each conditionals. Since a causal relation, 
which is shown by TO or REBA, expresses a general 
rule, the consequence cannot include speaker's at- 
titude, like volition and request. Therefore, the 
SUBJECT of the matrix clause should be a ma- 
chine. In contrast, in the case of assumptions, 
that is TARA and NARA , there are no such re- 
strictions on the SUBJECT . 
It depends on the volitionality of the verb 
whether a sentence shows a speaker's attitude, or 
not. Therefore, we consider each Japanese condi- 
tionals in terms of volitionality of the verb. Note 
that the electronic dictinary IPAL provides the in- 
formation of volitionality for each Japanese verb 
entry(IPA Te.chn01ogy center, 1987). We can use 
it to analyze sentences based on our proposal. 
4.1 SUBJECTS of complex sentences with 
the eonditional TO 
A matrix clause of the sentence with TO expresses 
a consequence of a causal relation. Consequently, 
conditionals REBA , TARA and NARA, since they oc- 
cur less frequently than TO in manuals and we have 
to collect more examples to estimate their property in 
nralluals. 
784 
TO 
REBA 
TARA 
NARA 
Table h (\]haracteristics of Japanese Conditionals 
shows successi{;eness of two"e;eents observed in a real situation. 
shows a universal causal relation. ' ..... 
It becomes more assurnlStive when the subordinate clause shows a state. 
shows l) two individual events occur with ~ie l)assing of tile time, or 
2) an event which is expected to occur on the uncertain assumption expressed 
in the subordinate clause. 
shows that the antecedent of the senl.e.nce is an ~ssumpi, i0n ..... 
and tire consequence holds on that ;~ssmnption. 
in matrix clauses, we can use either the mood of 
the description of fa.cts or the mood of evidentials 
like conjectures, judgment and so on. In contrast, 
we may not use the expressions of volition, re- 
quests and so on. We consider only the mood 
of the description of facts, because manual sen- 
tences should describe only facts and must not 
in('ludc sl)eaker's at, titude. The sentences having 
the mood are classified into two types: tile (le- 
scription of an action and the description of a state 
like an expression for the ability of some action. 
The former type is problematic, because the RU- 
l'orm~ which is the normal inflection form of verbs 
and describes an action, is ambiguous in its mean- 
ing. The RU-fonn can show one of tit(', followings: 
speaker's volition, speaker's request t,o hearers, or 
the action done hy a third party. 
In the analysis of the description of an action, it, 
is important to examine whether the verb phrase 
expresses a volitional action or not. According 
to the classitlcation by IPA(IPA Technology een- 
ter, 1987), all of Japanese verbs are classitied 
into two types, volitioual verbs, which usually ex- 
press int, entional actions, and non-volitional verbs, 
which express non-intentional actions. Although 
non-volitional verbs only exl)ress non-volitional 
actions, volitional verbs are cb~ssitied into two 
kind of verbs. One is the type of verbs which 
can be used tbr not only volitional actions but 
also non-volitional actiorts. The other is the type 
of verbs which are used only for volitional ac- 
tions. For example, ITAMO(havc a pain) is a 
non-volitional verb, OTOSO(drop/loose) is a voli- 
tional verb which has also the non-volitional use, 
SAGASU(scarch) is a volitional wM) which has only 
the volitional use. 
Let us consider the interpretations of the ma-- 
trix clauses of the sentences with 1-O. The first 
case is that verbs in the matrix clauses are in voli- 
tional use. If the SUBJECT is the speaker, the verb 
in volitional use expresses speaker's volition. If 
the SUBJECT is the hearer, the st)eaker expresses 
his/her expectation that the hearer makes a vo- 
litional action shown by the sentence. This is 
the case of requests. Consequently, the SUBJECT 
should be neither the speaker nor the hearer due 
to the constraint that we cannot express sortie vo- 
lition o1: request in a matrix clause of the TO sen- 
tence. On the other hand, a third paJ'ty can 1)e 
the SUBJECT, because a sentence whose SUBJECT 
is a third party does not express any volition, in- 
vitations, requests or injunctions. Since tile man- 
ufactm'e is the speaker and the user is the hearer 
according to the constraint of the discourse sit- 
uation, the mauufacture Ltll(l tile user cannot be 
the SUBJECT of the matrix clause. Therefore, the 
only possible interpretation is that the SUBJECT 
of the matrix clause is the machine. 
The second case is thal, verbs in the nlatrix 
clauses are in nell-volitional use. If a verb of the 
matrix chmse has a non-volitional use, thai, is, it' 
it is possible for the action of the clause to be 
done unconsciously, the constraint is not applied, 
because the w:rb in non-volitional use does not 
express any volition, invitations, requests and im 
junctions. For example, the SUBJECT of the ma- 
trix clause of the following sentence refers to the 
users. 
(8) ~b\] O.q fllreru-l,o, 
qSj-NOM q59-ACC touch-TO, 
'/'h kandenshi-mas-u. 
(~h-NOM get_an_eh~ct ric~shocl~-p OL-N O N I'AST. 
\[1'@ touch(es) qSq, then 4h will get an eh'.etric 
shock. 
To examine the accuracy of interpretations 
bused on our estinmtion we have collected about 
400 sentences, which include TO and some of 
which also inch.Me possibility expressions, from 
several types of inanuals, l/y these sentence.s, 
we check Constraint 5 and our estimati(m of TO. 
Then, it is contirmed that there are no excel)tion 
to them, at least in the collected sentences. 
4.2 SUBJECTS of comI)h~.x s(mten('es with 
the conditionals REBA,TARA and NARA 
Because of the characteristics of each conditionals 
descril)e(t in Table 1, we expect that a) the con- 
junctive REBA, which shows a causal relation, has 
the same constraint as TO has, which also express 
causality, b) since both of TARA and NARA ex- 
press an assumption, they have the same type of 
constraint, which is difl'erent fi'om the constraint 
of TO and REBA. As the first step to confirm this 
expectation, let us examine whether the matrix 
clause may have a request form, or not, in the 
cruses of REBA,TARA and NARA. At lirst, note that 
the hearer, namely the use.r, is the agent of Cite re- 
quested action if the matrix clause is a. re(luest 
form. In the case that the conjunctive shows 
causality, the matrix chmse should show some 
inevitable result of tile event expressed by the 
785 
subordinate clause. Therefore, tile matrix clause 
should not express the judgement and attitude of 
the speaker. As for the conjunctive REBA, the 
fact that tile conjunctive represents some causal- 
ity means that the matrix clause does not have 
a request form. Note that the exception is tile 
case that the subordinate clause is stative, or a 
non-volitional action. As described in 'Fable 1, 
in those eases, the subordinate clause shows an 
assumption rather than a cause, and the matrix 
clause may be a request as shown in the following 
example. 
(9) hitsuyou-ga nake -reba, 
Necessity-NON there-is-no-REBA, 
¢o Cp sutete -kudasa -i. 
Co-NON Cp-ACC discard -REQPOL -NONPAST. 
If there is no need of Cp, please discard Cp. 
The usages of the conjunctives TARA and NARA, 
which express assumptions, are explained as fol- 
lows. Since the assumptions are introduced by the 
speaker, the matrix clause is to describe speaker's 
expectation or desire. Therefore, it is quite prob- 
able that not only the normal form but also some 
request form, which is considered as a kind of 
wish, appears in the matrix clause. 
In order to ascertain our estimation, we have 
examined a bunch of real sentences, which appear 
in real instruction manuals. First of all, in about 
400 TO sentences, all of tile matrix clauses have no 
request form. In the REBA case, few request form 
appear in the matrix clauses. The exceptions are 
the same type of sentences as (9). 
Next, we consider the usage of TARA and NARA. 
Even if the conjunctive REBA in the sentence (9) 
is changed for TARA or NARA, the sentences are 
still acceptable. As we exepcted, it shows that the 
matrix clause of the sentence with TARA or NARA 
may have a request form, that is , the SUBJECT 
of the matrix clause may be a user. Then, can 
the SUBJECT of the matrix chmse be a machine? 
We expect that there are few cases that the SUB- 
JECT of the matrix clause is a machine, because 
the highly context specific assumption, which is 
expressed by TARA or NARA, is not suitable for 
tile description of general rules. Moreover, fi'om 
the fact that the matrix clause of TO and REBA 
cannot express the speaker's attitude, we prag- 
matically infer that TARA and NARA are expected 
to be used only for expressing the speaker's atti- 
tude. Our expectation is summarized in Table 2. 
Note that a SUBJECT should be either a user or a 
machine because manufacturers have finished all 
the actions appeared in the context of instruction 
before shipment. 
Our estimation about TO has been already con- 
firmed in Section 4.1. In order to confirm our 
estimation about REBA,TARA and NARA, let us 
examine real examples. Since tile constraints we 
pursue here are those which restrict the types of 
SUBJECTS, we examined the correlation among 
the types of conjunctives, the types of verbs and 
the SUBJECT. As for the types of SUBJECTS, a 
SUBJECT should be either a user or a machine. 
Table 2: Our estimate of the usage of tile matrix 
clause 
Speaker's attitude = 
User's Volitional Act. 
TO,REBA Not available 
TARA,NARA Available 
Others 
Available 
Not available 
As for tile types of verbs, each clause is classified 
into two classes according to volitionality of verb. 
One of them is tile cl;~ss of verbs in volitional use, 
the other is the class of other non-volitional predi- 
cates. Therefore each clause belongs to one of the 
followings: 
• SUBJECT = user and Predicate = verb in vo- 
litional use (U/V, hereafter) 
• SUBJECT = user and Predicate = others (u/o) 
• SUBJECT = machine and Predicate = verb in 
volitional use (M/V) 
• SUBJECT = machine and Predicate = others (M/O) 
Table 3, 4 and 5 show the distribution of usage 
of each conjunctive. Each number shows the fre- 
quency of use in the examples we examined. Note 
that to create 'Fable 3,4 and 5, several Japanese 
native speakers determine referents of zero SUB- 
JECTS according to contexts. 
Table 3: Distribution of use of REBA 
Matrix Clause 
U/V U/O M/V M/O Total 
U/V 1 65 52 14 132 
0.4% 26.9% 23.1% 6.2% 58.7% 
U/O 4 12 1 0 17 
1.8% 5.3% 0.4% 0.0% 7.6% 
M/V 0 0 1 4 5 0.0% 0.0% 0.4% 1.8% 2.2% 
M/O 6 20 38 7 71 
2.7% 9.0% 16.9% 3,1% 31,9% 
11 97 92 25 225 
rotal 4.9% 43.1% 40.9% 11.1% 100% 
First of all, as we expected before, the distri- 
bution of the use of REBA is different from those 
of TARA and NARA. While we call see several dif- 
ferences of use, the most remarkable one is the 
difference of use of the matrix clause. The matrix 
clauses of REBA are hardly any user's volitional 
action. The exceptions are only about 5% of all 
examples. The distribution of use of the matrix 
clauses of TARA and NARA is complementary to 
the distribution of REBA, that is, the majority of 
the matrix clause of TARA(about 90% of all exam- 
ples) and NARA(100% of all examples) are user's 
volitional actions, although the number of the to- 
tal examples of NARA is not so numerous. The 
empirical result supports the our estimation. 
786 
Table 4: l)istribution of use of TARA 
M atrix Clause 
U/V U/O M/V M/O Total' 
25 0 0 0 25 
U/V 42.4% 0.0% 0.0% 0.0% 42.4% 
7 1 0 0 8 .... 
U/O 11.9% 1.7% 0.0% 0.0% 13.6% 
8 2 O 1 11 
M/V 13.6% 3.4% 0.0% 1.7% 18.6% 
13 0 2 0 15 
M/O 22.0% 0.0% 3.4% 0.0% 25.4% 
53 3 2 1 59 
\]'ot a\] 89.8% 5.1% 3.4% 1.7% 100% 
S 
U 
I) 
TMJe 5: Distrilmtion of use of NARA 
Matrix Clause 
U/V U/O M/V M/O Total 
0 0 0 0 0 
U/V 0.0% 0.o% 0.0% o.o% 0.0% 
8 0 0 0 8 
U/O 88.9% 0.0% 0.(l% 0.0% 88.9% 
0 0 0 0 0 
M/V 0.0% 0.0% 0.0% 0.0% 0.0% 
I 0 0 0 1 
M/O 11.1% 0.0% 0.0% o.0% 11.1% 
9 0 0 0 9 Tot;ill 100% 0.0% 0.0% 0.0% 100% 
4.3 Default rules of usage of REBA, TARA 
and NARA 
The tendency of use of the conjunctives gives us a 
couple of strong defimlts to resolve the zero pro- 
noun in tit(; matrix clauses. We propose the fol- 
lowing defaults. 
Default 1 (SUBJECT of sentem:e with TO or 
REBA) 
In a complex sentence with the connectzve parti- 
cle TO orREBA, the matrix clause does not empress 
user's volitional action. Therefore, the SUBJECT 
of the matriz clause should be a machine, if the 
verb of the matrix clause does not have the non- 
volitional use. 
Default 2 (SUBJECT of sentence with TARA 
or NARA) 
In a complex sentence with. the connective par- 
ticle TARA or NARA, the matrix clause e~:presses 
only user's volitional action. 7'h.erefore, the SUB- 
JECT of the malrix clause should be a user. 
The accuracy of the default rules of 
-I-O,REBA,TARA and NARA is 100%, 95.1%, 89.8% 
and 100%, respectively, as far as we examined. 
5 Conclusion 
In l;his pat)er , we proposed a scheme which closely 
depends not on domain knowledge of objects de- 
scribed in manual but on pragmatic constraints 
which linguistic expressions innately have. This 
method uses only the linguistic constraints and 
the general ontology of the world described by 
manuals. Especially, we have shown that we can 
deternfine the referents of zero pronouns to some 
extent with our linguistic constraints, like the con 
straint of the Japanese Conditionals. llowew~'r, 
we do not have enough knowledge about the fob 
lowing l)oints. They are important portions of our 
future work. 
• Utilization of discourse, structure. 
• Analysis for the other types of manual sen- 
tences, like definitions. 
References 
Norihiro Abe, Tomohiro lshikawa, and Tsuji 
Salmro. 1988. Generating an asscnfl)ly pro- 
cedure from asselnbly instruction. Journal 
of Japanese Society for Artificial intelligence, 
3(5):590-598, September. (in Japanese). 
Susan E. F/rennan, Marilyn Walker Friedman, and 
Carl 3. Pollard. 1987. A centering approach 
to pronouns, In Proceedings of th.e 25th Annual 
Meeting of the American. Association of Compu- 
tational Linguistics (ACL 87), pages 155:162. 
Kohji Dohsaka. 1994. hlentifying the referents 
of Japanese zero pronouns based on pra.gmatic 
condition interpretation. 7Yansactions of In for 
malion Processing Society o\] Japan, 35(5):768 
778. (in Japanese). 
Barbara Di Eugenio. 1992. Understanding natu- 
ral language instructions: The case of purl)use 
clause_ In Proceedings of 30th Annual Meeting 
of A CL, pages 120 127. 
II'A Technology center, 1987. Th.e lexz- 
eon of the Japanese basic verbs for Com- 
lrulers, hfformation-teehnology I'romotion 
Agency(IPA), ,lapan, March. (in :lapanesc). 
Megumi KarneyaIna. 1988. Japanese zero 
pronominal binding: Where syntax and dis- 
course meet. In W. Poser, editor, Japanese 5'yn- 
tax, pages 351-369. CSIA,Stantbrd. 
Megumi Kameyama. 1995. The syntax and se- 
nm.utics of the .lal)anese language engine. In 
R,. Mazuka and N. Nagai, editors, Japanese 
Syntactic Processing. Lawrence l,\]rll)aum Asso- 
ciates. 
'Fakashi Masuoka, editor. 1993. Nihongo- 
no Joken-llyogen (Conditional t'2~:pr'essions in 
,/apanese). Kurosio Publishers, Tokyo. (in 
Japanese). 
llirosato Nomura. 1992. l,inguistic analysis of law 
sentences. In Proceedings of the 6th Interna- 
tional Symposium: Legal l(nowledge and Legal 
Reasoning Systems, pages 10 18. l,egal Exl)ert 
System Association, October. 
Marilyn Walker, Masayo 1ida, a.nd Sharon Cote. 
1990. Centering in Japanese discourse. In Pro- 
ceedings of COLING 90. 
787 
