■ Research Note ■
Subjectivity and Objectivity in the
Grammar of Oriya
●
Junji Yamabe
1. Overview
In this article I examine two grammatical expressions in Oriya, the particle bi 'also, even' and the relative clause construction,l) as to how they are related to "subjective" and "objective" factors, namely, features of the depictor (or speaker) and the outer world, respectively. They are both basic grammatical devices in this language for rendering partici-pants of events (or things that are involved in happenings).
The Oriya expressions turn out to contrast with the comparable En-glish ones in one respect, despite they are practically equivalent in that the counterparts usually translate into each other. The Oriya expressions
treat subjective and objective factors on equal footings, while the con-cern of the English ones inclines toward objective factors. On this obser-vation I argue that the meanings inherent to the Oriya expressions are
concerned with the "appearances" of depicted entities to the depictor, whereas the meanings inherent to the English expressions are concerned
山部順 治 Junji Yamabe, Department of English Language and Literature, Notre Dame Seishin University.
Subject: Linguistics.
Recent publications: "What the Distribution of the Relative Pronoun in Oriya Tells us about an Anaphoric Rule" (in Japanese), in Proceedings of the 120th Meeting of the Linguistic Society of Japan, pp. 108-113, and Gengo Kenkyu 118, p. 144, 2000. "The
Meaning of the Japanese Auxiliary Verb oku" (in Japanese), Studies in Japanese Lan-guage and Literature, Notre Dame Seishin University 25 (1), pp. 53-78, 2001.
162 Journal of the Japanese Association for South Asian Studies, 13, 2001
with the depicted entities themselves and are accordingly susceptible to a description in truth-conditional terms.
I go on to propose that the contrasts point to the difference between the two languages in their overall grammatical design, in other words, in the way they represent the world, particularly, "things" (as opposed to happenings) constituting it. I also make suggestions as to how the way the Oriya grammar represents the world is mirrored in the way the cultural tradition of the region does the same.
Putting forth thus an outlook on subjectivity and objectivity from a linguistic standpoint, I would like to call for the insights of the research-ers of adjoining disciplines who investigate closely related topics.
The structure of this article is as follows. The data and their analysis come in sections 3 and 4: section 3 deals with the particle bi; section 4, with the relative clause. These two sections are sandwiched by two shorter sections: section 2 sketches the notion of subjectivity and its opposite, objectivity; section 5 gives a summary and prospects.
2.
The Notions of Subjectivity
and Objectivity
The notions of subjectivity and objectivity can be outlined as follows.
A sentence carries two semantic aspects with it by virtue of occurring
in speech events to make sense: inside it, it depicts a segment of the outer
world; outside it, it is constructed by the depictor (or speaker). Thus, it
can be phrased as having its object and subject: the outer world is the
object of depiction, and the depictor is the subject of depiction.
Expressions, of various kinds and sizes (for example, morphemes and
construction types), which constitute a sentence, each have a meaning in
inherent to them. Some have a meaning that is concrete enough to be
accessible to naive introspection; others, like the two that the present
article deals with, have a meaning that is so abstract as to call for
linguis-tic tricks. The expressions' meanings enable them to get employed in
actual sentences to impart features of either, or both, of the
above-men-tioned two aspects. Their concern with features of the outer world is
termed objectivity, and that with features of the depictor, subjectivity.2)
The general outline above now allows us to come to grips with the
topic of this article, and the examples in the following will provide
sub-stance.
3. The Particle bi 'also, even'
The particle bi has a variety of usages, and accordingly is translated in English and other languages familiar to us in a variety of ways.
3.1 The 'also' Sense
In its presumably basic usage, the particle bi is translated into 'also, too' (Japanese mo), as in (1).
(1) apekhyaa kar-a! mu•` bi jib-i. wait.OPT-2 me.NOM bi will.go-1s 'Wait .I will go,too.'
In this usage, bi brings with it the implication that, in addition to the referent of the overt noun phrase it is attached to, there is an unstated alternative instance determinable uniquely from context of which the event of clause holds true. For example, in the second sentence in (1), it is bringing in the implication that the event of 'going' holds true of 'you', in addition to mu•`'I'.
3.2 When bi Is Attached to a Noun Phrase
The particle also appears in the sequence hele bi.3) Such cases can be classified into, broadly, two usages. In the one, hele bi is translated into
'even' (Japanese demo), as in(2).
(2) a. choTa pilaa hele bi kari paarib-a. small child hele bi do-CP can.FUT-3s 'Even a small child will be able to do it.' b. baDa loka hele bi kari paarib-a ni. big man hele bi do-cP can.FUT-3s not 'Even an elder person will not be able to do it.'
c. 'This exam is tough.'
baapi hele bi phel hoi paar-e. Baapi hele bi fail become-CP can-3s 'Even Babi can fail in it.'
In (2), bi is bringing in the implication that the referent of the noun phrase it is attached to is the last instance to participate in the event of
164 Journal of the Japanese Association for South Asian Studies, 13, 2001
the clause as this event comes into reality ("low point" of Fauconnier 1975). As a consequence, the sentences convey that the event would hold true of any possible alternative instance, as well as of the referent of the overtly stated noun phrase. For example, in (2c), it is implied that Babi is the last person to participate in the realization (or occurrence in the real world) of the event of 'failing in the exam'. (Perhaps, he is a brilliant student.) The sentence conveys consequently that any possible student other than Babi would be likely to fail, as well as he is.
In the other usage, the hele bi sequence is translated into 'at least' or otherwise (Japanese sukunaku to mo, kurai (wa) , stressed wa).
(3) a. se krupaNa.
he.NOM miser
tathaapi alpa hele bi paisaa deb-a. even.then a.little hele bi money give-3s
'He is a
miser. Still,he will pay at least a bit.'
b. jaNe hele bi paas heb-a.
one (person) hele bi passed will.become-3s 'At least one
person will pass the exam.' c. alpa hele bi di-a!
a.bit hele bi give.OPT-2p 'Give me something
, a bit at least!' d. go Tie hele bi khaaib-a!
one hele bi will.eat-2s 'Eat at least one piece!'
e. nabina ki niru ghar-e na thaai paar-anti.
Nabin or Niru house-Loc not be-CP can-3p
kintu harita hele bi thib-a.
but Harit hele bi will.be-3s
'It may be that neither Nabin nor Niru are at home . But at least Harit will be there.'
f. taanka baahaaghara-ku mu•` khaali-haata-re their marriage-OBJ me .NOM empty-hand-LOC jaai na-thil-i.
choTa jinisa hele bi nei kari jaaithil-i. small thing hele bi take-cP had.gone-1s
'I didn't go to their marriage empty -handed. I did take something there,
though it was a small thing.'
In (3), bi is bringing in the implication that the referent of the noun phrase it is attached to is the last instance to participate in the event of the clause as it is conceptualized by the depictor. For example, in (3e), it is implied that Harit is the last person to participate in the conceptualization (or occurrence in the mind) of the event of 'staying in the house now'. This sentence was uttered when we were going to visit Nabin's family. The adult members of the family (Nabin and Niru) came into issue first and foremost, and the child (Harit) came under consideration only after them.
Note that, unlike the sentences in (2), those in (3) are not evoking an implied pragmatic scale according to which an instance is the least prob-able (or most probprob-able) to participate in the realization of the event.
The usages in (2) and (3) converge in that the referent of the noun phrase that hele bi is attached to is the last instance for the clause to hold of, in other words, the last instance to be acknowledged as appearing in the event of the clause. On the other hand, they diverge in the account
on which the noun phrase's referent is the last. In (2), it is so because it is most unlikely to participate in the realization of the event. In (3), it is so because it is most unlikely to participate in the conceptualization of the event by the depictor.
The preceding paragraph can be rephrased as follows. In (2) and (3) alike, the noun phrase's referent is situated barely inside the boundary (or threshold) between "the clause does and does not hold of the noun phrase". It is, briefly, a boundary case. On the other hand, (2) and (3) diverge in the dimension of the pragmatic scale on which the boundary lies: in (2), the boundary lies between the realization and non-realization of an event; in (3), it lies between the application and non-application of the attribute(s) characteristic of that event to a real-world phenomenon. In the former case, the boundary is due to objective factors, and in the latter, to subjective factors.
166 Journal of the Japanese Association for South Asian Studies, 13, 2001
are widely (and usually in popular terms) held to be governing human experience of the world. In order for an event to be acknowledged as occurring, both criteria must be satisfied: the one dictates that an event-like matter be out there; the other dictates that it be acknowledged by the cognizer as of the category of a particular event. If either fails, there is no acknowledgement of an event occurring.
The particle's inherent meaning, I contend, is concerned with the "appearance" (or look) of the noun phrase's referent as it occurs in the clause's event. Specifically, it means that the depictor's attention domain expands to obtain the intended appearance of the noun phrase's referent within the clause's event. This cognitive operation is carried out (in 3. 1 to 3. 3) in order to have the referent visible within the clause's event, that is, in order to have the referent recognized as being in the clause's event.
3.3 When bi Is Attached to a Clause
The particle can be attached to a dependent clause whose verb is in the conditional participle (henceforth, conditional clause).4) The way it exhibits a variety of senses in such cases parallels the way it does so as it is attached to a noun phrase. It brings about the implication that the event of the dependent clause is the last instance to hold together with the event of the main clause. Consequently, the dependent clause is understood as standing in the adversative relation to the main clause.
(4) baapaa manaa ka-le bi
father prohibit do-COND bi
se mada piibaa banda {karib-a/kal-aa} ni.
he.NOM liquor drink-GER will.stop-3s/stopped-3s not 'Even if father forbids, he/she will not stop drinking.'/ 'Although father forbade
, he/she did not stop drinking.' In some cases, as in (4), the sentence involves an implication of a cause-effect relation, a kind of relation pertaining to the arrangement of the outer world: normally, if the conditional clause's event ('a father's forbidding his son to drink') realizes, the main clause's event ('the son's not stopping drinking') does not realize; in other words, the real-world
occurrence of the former usually prevents the real-world occurrence of the latter.
an implication. This sentence does not point to such an expectation according to which, normally, if the conditional clause's event Ca girl's being pretty'), the main clause's event Cher not being humane') does not realize.
(5) a. mu•` kam maark-re paas hoith-ile bi me. NOM few mark-Loc passed have.become-COND bi
em e paas hoich-i. MA passed have.become-1s
'Although I passed with a low mark
, I did pass my MA.' b. se sundara hoith-ile bi dayaaLu nuh-e•`.
he/she beautiful have.become-COND bi humane is.not-3s 'Although she is beautiful
,she is not humane.'
Here, what is implied is not about the state of affairs of the outer world, but about the evaluation by the depictor of the notions of the conceived events: admitting the main-clause event makes admitting the conditional-clause event difficult for the depictor.
The usages illustrated respectively in (4) and (5) converge in that the conditional clause's event is the last instance to accompany the main clause's event. On this score, in order to have it in his/her view, the descriptor has to expand his/her attention domain. They diverge in the account on which the conditional clause's is the last: in (4), the condi-tional clause's event is the least probable to concur with the main's clause's event as they come into reality or occur in the real world; in (5), the conditional clause's event is the least probable to concur with the main clause's event as they are conceptualized or occur in the mind.
3.4 When bi is Attached to Indefinite Pronouns
The particle bi (and the sequence hele bi) can be attached to a variation of indefinite pronouns.5)
The particle bi (and hele bi) can be attached to the relative pronoun, the pronoun beginning with j-. The combination expresses either uni-versality ('any'), as in (6), or existence (`some...no matter who/what'), as in (7).
168 Journal of the Japanese Association for South Asian Studies, 13, 2001
(6)
jie
{bi/hele bi} paas
hu-e.
who(R).NOM bi / hele bi
passed become-3s
'Any body passes the exam .'
(7)
jie
{bi/hele bi} mo-ra
byaaga nei jaaich-i.
who(R).NOM bi/hele bi
me-GEN bag take-CP has.gone-3s
'Someone has taken away my bag
, no matter who.'
The universality in (6) is exceptionlessness. The existence in (7) is accompanied by the emphasized ignorance of, or indifference to, the identity.
(6) and (7) converge in that the particle bi indicates that the relative pronoun's referent is outside the current attention domain, that is, it is a referent that cannot be reached immediately. They diverge in the ac-count on which it is so. In (6), bi indicates that it is the last alternative to participate in the realization of the clause's event. Consequently, the clause holds true of any instance, including the least probable instance. In (7), the relative pronoun's referent is specific, but, as bi indicates, the notion of does not immediately become visible, hence the unaccessibility of the identity, which is typically (if not necessarily) due to the incurios-ity on the side of the depictor.
The particle can also be attached to pronouns beginning with k- (hence-forth, k-pronouns). The combination expresses existence ('some, there is one who/which...').
(8) a.
kie {bi
/hele bi} aasib-a.
who(Q).Nom
bi / hele bi
will.come-3s
'Somebody will come .'(= There is one who will come.)
b.
kehi
{bi/hele bi} mo-ra
byaag
somebody.NOM bi / hele bi
me-GEN bag
nei
jaaich-i.
take-CP has.gone-3s
'Somebody has taken away my bag .'
(= There is one who has taken away my bag.)
c. kichi bi thib-a.
something bi will.be-3s 'Th
ere will be something.'
would then express existence. The use of bi here is bringing in a subtle semantic effect, unrelated to the size of a set. (To be exact, it narrows the possible range of interpretations.) Thus, as is shown in (9), bi cannot go with a k-pronoun, if the latter is referring to an object directly perceived. The ƒÓ mark stands for "nothing added".
(9) A: 'There is a sitting place upstairs. We can sit there.' B: 'Let me check.' (Goes up to the place, and comes back.) A: 'It is occupied.'
kie {ƒÓ/*bi/*hele bi} basich-i. who(Q).NOM bi hele bi is.sitting-3s 'Somebody is sitting.'
Here, speaker A observes the sitting person to details. (The ignorance of the identity is expressed by the use of the interrogative pronoun.) The use of bi fails to be justified, because the referred-to object is not outside of the region presently in attention, in particular, it is not located beyond the current reach of the first-hand perception by the speaker.
A slight change in context makes the use of bi (and hele bi) possible.
(10) A: 'There is a sitting place upstairs. We can sit there.' B: kintu kie {ƒÓ/bi/hele bi} basithib-a.
but who(Q) .NOM bi hele bi will .be .sitting- 3s 'But somebody will be sitting.'
Here, speaker B does not observe the sitting person, but presumes the existence on the basis of circumstantial evidence. The sitting person is located beyond the current reach of the speaker's first-hand perception.
So in order to have the notion of it within the view of cognition, the speaker has to expand the domain of attention. The particle bi points to this cognitive operation.°
3. 4 When bi is Attached to Collective Pronouns
The particle bi can be attached to words that portray the set they refer to as a collection, in other words, words that profile the definite bound of the set referred to. bi in this case expresses the reconfirmation of the set's
collective feature. In (11) are words of collective universality (samaste 'all (people)', sabu 'all (things, peole)'). In (12) is a word that specifies the
170 Journal of the Japanese Association for South Asian Studies, 13, 2001
number of the members of a set (dui jaNa 'two CL (person)') as it is referring to a definite group ('the two people').
(11) a. samaste {ƒÓ/bi} paas hel-e. all.NOM bi passed became-3p
'All passed the exame .' b. sabu {ƒÓ/bi} jaNich-i
all bi has.known-1s
'I know e
verything.'
(12) se dui jaNa bi caali gal-e.
that two CL (person) bi walk-CP went-3p 'They both went away
.'
In (13) is a negative counterpart to (11) (and, less closely, also to (12)). adow only appears in contexts of negation (Japanese zenzen).
(13) mu•` adow {ƒÓ/bi} semiti kar-e ni. me.NOM at.all bi that.way do-1s not 'I never do that way .'
Incidentally, this usage is not possible with the sequence hele bi: in (11)-(13), hele bi in place of bi is not acceptable.
The usage of this section differs from those of the preceding sections on one score. In this usage, the particle bi presents the noun phrase's referent (samaste in (11 a)) in such an appearance as though the attention domain would have expanded to reach the endpoint so that there were no further course for it to expand along. Thus, in (12) for example, the particle expresses the reconfirmation that the set consisting of the two people at issue is considered collectively. Thus, in the usage of this section, illustrated in (11)-(13), bi indicates the type of the mental opera-tion carried out to put the referent of the word it is attached to in the intended appearance. In contrast to this usage, in the others, illustrated in sections 3.1-3.3, such mental operations are done to make the referent come into the purview of the depictor for the first time.
To capture this contrast, between the usage in 3.4 on the one hand, and the usages in 3.1-3.3 on the other hand, we can employ the analyti-cal term "end focus" of Lakoff [1987], which marks the former but not the latter.7 The particle bi inherently denotes a process that is the ex-tending of the attention domain. In the former usage, it is bringing into
focus the final part of the extending process, putting out of focus the process itself, the main body of the denoted concept. In contrast, in the other, unmarked usages, it is profiling the whole process of the extend-ing of the attention domain.
3.5 The Particles in Other Languages
Let us review the linguistic literature. The particles in other languages, in South Asia and elsewhere, which translate Oriya bi are termed "addi-tive", "totalizing" or "inclusive", all indicative of their meanings. Where these terms are accompanied by any substantial explanation, the claims are, in outline, as follows: (i) the particles are to bring in the implication that the clause holds true of an alternative or alternatives, in addition to the referent of the noun phrase they are attached to ("additive"); (ii) repeating such an additive operation leads to the event of the clause holding of all alternatives ("totalizing"). (The term "inclusive" seems to cover (i) and (ii).)
This view is argued for by Emeneau [1974] on the particles in lan-guages of India, and by Gil [1995] on those of Malayalam, Panjabi and other languages of the world. A variant of this view is that English any makes the statement of the clause logically stronger, by precluding the exceptions of marginal instances. Dayal [1995: 191-197] and Lahiri [1998] adopt this for a limited range of usages of Hindi bhii 'also, even'.
We can name this view, inclusive of the last-mentioned variant, as truth-conditional approach. It looks into the inclusion relation between the sets of instances fitting in with the descriptions of a pair of clauses, distinguished from each other only in having or lacking the relevant article.
However, the truth-conditional approach does not work for the de-scription of the Oriya facts. It will deal with one half of the variety of usages, but not the other half, that is, only with those usages involving an implication pertaining to the state of affairs of the outer world; those illustrated in (1) and (2), but not (3); (4), but not (5); (6), but not (7). It will not deal with (8, 9) and (11)-(13), either: in (8, 9), a single instance is first introduced into a discourse with the utterance of the sentence, so it does not have any smaller set in consideration that could be compared with; in (11)-(13), the size of the referred-to set is fixed by a pronoun, and is not altered by the use of bi or anything else.
172 Journal of the Japanese Association for South AsianStudies, 13, 2001
4. The Relative Clause Construction8)
Relative clauses can be classified on a semantic ground into restrictive relative clauses and non-restrictive relative clauses. A restrictive relative clause can be defined as in (14). A relative clause that is not restrictive relative clause is a non-restrictive relative clause.
(14) A restrictive relative clause limits the set of the entities which is referred to by the noun phrase in the main clause ("the head noun") to the subset consisting of the entities of which it holds.9)
The notions characteristic of this definition are "to limit" and "sub-set". Note that these are predicates of truth-conditional or set-inclusion relation. In English, these notions play a central role: the restrictive and non-restrictive relative clauses they define diverge in a number of syn-tactic respects."
The relative clause in Oriya, in contrast, is not concerned with truth-conditional notions such as these, as we shall see below. It is concerned with the appearance of something in some setting to the depictor, exactly as the particle bi was. Specifically, it means that the referent of the rela-tive pronoun shows the depictor contrasting guises in the settings of the relative and main clauses. [Yamabe 1998a]
4. 1 When the Relative Clause is Restrictive
The relative clauses in (15) do fit in with the semantic specification in (14), and therefore are restrictive relative clauses. For example, in (15a), the set of people in general is limited to the subset consisting of the people who study well.
(15) a. jie bhala paDh-e,
who(R).NOM well study-3s
se paas hu-e.
he.NOM passed become-3s
'The one who studies well passes the exam .' b. jeu∼chaatra bhala paDh-e,
sei chaatra paas hu-e. that student passed become-3s
'The student who studies well passes the e xam.'
c. jeu•`jhia seThi basich-i, which(R) girl there has.sat-3s
sei jhia mo bhowNi. that girl my sister
'The girl who is sitting there is my sister .'
The referent of the relative pronoun in some cases is indefinite, as in (15a,b), and in other cases is definite, as in (15c). (In the following, for brevity's sake, the phrase "relative pronoun" subsumes the reference to the sequence of 'relative pronoun + noun', e.g. jeu•`jhia `which(R) girl' in (15c).)
4.2 When the Relative Clause is Non-restrictive
(16) is an example of non-restrictive clause: the relative clause does not accord with the semantic specification in (14). The intention here is not that a particular person named Nabin is singled out from the set of people with the same name. The sentence in (16) is unacceptable.
(16) a. * jeu•`nabina paakha-re paisaa na thil-aa, which(R) Nabin side-Loc money not was-3s se nabina kouThi jaa-i paaril-aa naahi•` that Nabin anywhere go-cP could-3s not Intended: `Nabin, who had no money, couldn't
go anywhere.'
(16b, c) shows that the unacceptability persists so long as the same intention remains, irrespective of whether the name nabina `Nabin' is repeated, as in (16a), or not, as in (b, c).
(16) b. * jeu•`nabina paakha-re paisaa na thil-aa, se kouThi jaa-i paaril-aa naahi•`. c. * jaahaa paakha-re paisaa na thil-aa,
se nabina kouThi jaa-i paaril-aa naahi•`.
174 Journal of the Japanese Association for South Asian Studies, 13, 2001
relative clause, the type which marks the relativization site with a gap and has the verb of the relative clause in the gerund form (GERUND), as in (17).
(17) paisaa na th-ibaa nabina money not be-GERUND Nabin
kouThi jaa-i paaril-aa naahi•` . anywhere go-CP could-3s not 'Nabin
,
whohadno money, couldn't go anywhere.'
This construction type is free of the restriction under discussion. Thus, it can rephrase not only the non-restrictive (16), but also the restrictive (15).
The observations thus far might suggest that the relative clause in Oriya depends for its possibility on whether it is restrictive or not. (Such a view is adopted with respect to the relative clause in Hindi by Dayal [1996: 155].) However, a further observation reveals that it does not. The relative clauses in (18) do not comply with the specification in (14), and thus are non-restrictive. Still, the sentences are all acceptable.
(18) a. jeu•` maami kaanduthil-aa, which(R) Mami was.weeping-3s
sei maami ebe hasuch-i. that Mami now is.laughing-3s 'Mami
,
who was weeping, is now laughing.'
b. jeu•` maami ebe hasuch-i, which(R) Mami now is.laughing-3s
sei maami kaanduthil-aa. that Mami was.weeping-3s
'Mami
, who is now laughing, was weeping.'
c. jeu•` nabina paakha-re gata barsa-re paisaa na thil-aa, what(R) Nabin side-Loc last year-Loc money not was-3s
sei nabina paakha-re ebe bahuta paisaa ach-i. that Nabin side-Loc now much money is-3s 'Nabin
, who had no money last year, has much
money now.'
on whether or not the referent of the relative pronoun looks contrastively in the settings depicted in the relative and main clauses . The relative clause's situation renders the referent an appearance already presup-posed in discourse, and the main clause's situation, an appearance worth introducing into it.
For example, in the acceptable (18c), two appearances of Nabin are presented contrastively, the one in the situation in which he had no money last year, and the one in the situation in which he has much money now. His having no money last year is part of the presupposition of discourse, and his having much money now is newly reported. Unlike (18c), the unacceptable (15) presents two appearances of Nabin as of the same quality, the one in the situation in which he had no money, and the one in the situation in which he could not go anywhere.
4. 3 How Restriction is Effected
Let us come back to the restrictive relative clause as in (15). In Oriya, narrowing a set to a subset as for the domain of reference is just one of the effects that can derive from contrasting two appearances of one and the same referent. For example, the sentence in (15b) contrasts two appearances of a student belonging to a certain subset of students , the one in the situation in which he studies well and the one in the situation in which he passes the exam. The situation depicted in the relative clause is already part of the presupposition of discourse, and by this virtue it allows one to single out a certain subset of students from the set of all students. Thus, the understanding of restriction or limiting the domain of reference is obtained as an effect.
5. A Summary and Prospects
In sections 3 and 4, we have seen how English and Oriya contrast as for the meanings inherently associated with grammatical expressions .
In English, even, also and the restrictive relative clause are all to indi-cate what the outer world, the object of depiction, is like. By this virtue, they can be quite accurately described in truth-conditional terms.
In contrast, in Oriya, the translationally equivalent expressions are to indicate what the world looks like. The appearance of the world is af-fected, of course, by the features of the outer world, the object of
depic-176 Journal of the Japanese Association for South Asian Studies, 13, 2001
tion. But it is also affected by the features of the depictor, the subject of depiction. Accordingly, these expressions can impart two apparently very different things. They can, as in (1, 2, 4, 6) and (15), relate to features of the outer world, which are captured in truth-conditional terms such as "additive" and "restrictive" . In other times, they can, as in (3, 5, 7) and
(10), relate to features of the depictor, namely, his/her attitude toward the depicted contents.
(The usage of bi in (9, 10) could, if any, be classified as imparting a feature of the outer world, that is, the fact that the referent of a noun phrase is located beyond the current reach of direct perception, and the usage in (11)-(13), as imparting a feature of the depictor, that is, the reconfirming of the collectivity sense inherent to the word it is attached to. Anyway, this uncertainty should not bother us, since our contention is that Oriya bi is not concerned with the partition between subjectivity and objectivity.)
I suppose this pattern of the contrast between Oriya and English ex-tends farther than the two cases taken up in this article, covering a number of devices for rendering the participants of events. I am thus supposing that the characteristics described in the preceding paragraphs constitute the grammatical "predispositions" of these languages, which concern the dependents of the verbs.
I would like to word the predispositions as follows: English (E-i) takes for granted the division of the subject and object of cognition, and (E-ii) pays more attention to the object: Oriya (0-i) is indifferent to the divi-sion of the subject and object of depiction/cognition, but (0-ii) directs attention to the appearance of the experienced world, which is emerging somewhere between the subject and the object.")
The above characterization has assimilated into its (E-ii) part an in-sight found in the literature: English is characterized as allotting much of its attention to the objective aspects, in comparison to Japanese [for instance, Watanabe 1996: 125]. We might expect that this character of English is mirrored in the architecture of Western semantic theories. Actually, truth-conditional or formal semantics is a case in point, which sees meaning as the constraint on referring to things out in a world, real or possible.
The explanation of this article, in its turn, presents us with new puzzles, two of which I mention below. (I should admit that what follows
pre-sumes a degree of validity of what is said above .)
The first is the typology of languages. The above characterization of English and Oriya points out the following parameters (or scales): (i) whether or not, and to what degree, a language respects the subjectivity-objectivity division (as it refers to things); and (ii) which region of the subjectivity-objectivity expanses it inclines to pay attention to . These parameters together place the two languages far apart. Applied to lan-guages of the world, what classification will they yield?
As to this question, the literature supplies data for the inquiry into parameter (ii) with regard to Japanese: the language has been character-ized as having a variety of content words that involve subjective features , and also function words that express differing shades of the depictor's attitude to the depicted state of affairs. It remains to be seen what posi-tion it is assigned on parameter (i).
While this article sets about typology with the expressions of things in events (approximately corresponding to the grammar of nominals) , a fuller theory would have to comprise other domains of grammar , nota-bly, the expressions of events (approximately corresponding to the gram-mar of verbs). Specifically, it would have to give explicit answers to questions like the following. What is the "appearance" of an event?: Is there any kind or degree of correlation between the inclination toward the appearance in the case of things (or nominals) on the one hand , and that in the case of events (or verbs)?
The second puzzle is to understand grammars within broader con-texts, cultural and social and otherwise. To be more specific , one can investigate how the basic design of a grammar is mirrored in the features of the images of the experienced world of the speakers [Whorf 1956 , Foley 1997: part vi, Nuyts and Pederson 1997]. Such images assume explicit forms in variety of cultural phenomena, and also tacitly regulate the courses people take in everyday behavior .
The images of the world of Indians , as they get manifested in their cultural tradition, show us some parallelisms to the design of the gram-mar of Oriya, (0-i) and (0-ii), mentioned above . I would like to point out the following two cases that strike me as notable , although my lim-ited knowledge on relevant matters precludes technically detailed expli-cations and scholarly convincing arguments .")
opin-178 Journal of the Japanese Association for South Asian Studies, 13, 2001
ions, has hardly (or never?) taken for granted the division of the self and the world, continuously disputing whether such a thing is essentially there at all, or where it lies. Second, the Indian mythology depicts one and the same god in dramatically different appearances in different situ-ations in which he/her occurs. The various appearances of a god together receive devotion of people, being felt to be no less substantial than the "essence" or "substance" that would characterize that god as an indi-vidual. This course of imaging gods is also manifest in linguistic
conven-tions: a god assumes different names in different situations for him/her to occur in; such different names are not just improvisations of relaters, but form part of the permanent lexicon of languages so that they can be constituents of compound names (for example, `. . . Temple', `. . . River'). Having spoken out this much, I would like to call for insights of the researchers of the adjoining disciplines.
Notes
* This is the English version of the paper read at the 13th national meeting of the Japa-nese Association for South Asian Studies, October 7th, 2000, University of Tokyo. I would like to thank Shingo Einoo, Nozomi Kodama, Tomio Mizokami, Yoshifumi Mizuno and Teiji Sakata for their helpful comments on the earlier paper. Special thanks are due to an anonymous JJASAS reviewer for hisfher advices that led me to revision at a number of crucial points.
1) For a more thorough description of these expressions, see Yamabe [1998a, 1999] and unpublished [1998b: Chapter 4].
2) The streams of semantic studies which have the notion of subjectivity play central roles in description and explanation are generally called cognitive semantics, in distinction to formal semantics [Lakoff 1987, Lakoff and Johnson 1980, Langacker 2000, Talmy 2000]. The notion has also been respected in traditional Japanese linguistics (kokugogaku) [for instance, Watanabe 1996].
The description this article is going to give is cognitive on this account, making crucial references to subjectivity. The facts do not just admit, but require, a cogni-tive viewpoint.
Note also that, while sharing the core tenet, the streams vary in details. Variation is found, for one thing, in the kind of facts they deal with in terms of subjectivity and objectivity, and in correlation, in the specific formulation of the notions. In this article, I will not go into the discussion of such variation but remain somewhat vague on that matter, since my claims are independent of it.
3) What hele bi means is the result of the synthesis of what hele itself means and what bi itself means. Etymologically, hele is the conditional participle of the verb he-`become' . In this article, I am arguing that the meaning of bi itself contributes its meaning in the sequence hele bi. For the meaning of hele itself, and its work in hele
bi, see Yamabe [1998b: 4.2.2].
4) For the meaning of the conditional participle ending, see Yamabe [1998b: 4.2.1]. 5) For the meanings of indefinite pronouns and their work as they are accompanied by
a particle, see Yamabe [1997, 1998b: 4.1, 4.3].
6) Concerning the particle bhii and the infinite pronoun in Hindi, which correspond, respectively, to the particle bi and a type of k-pronoun in Oriya, Davison [1978] and Haspelmath [1997] note that attaching of bhii to the indefinite pronoun gives the "emphasis" . A more precise description of this "emphasis" is, I claim, the implica-tion that relevant entities are expected to be scarce. The original works translate the indefinite pronoun without vs. with bhii, respectively, as 'somebody' vs. 'anybody', as in (i), or 'anybody' vs. 'anybody at all', as in (ii). (In gloss, ERG = ergative case, OBJ = objective case, Q = question marker.)
(i) kyaa aap ne {kisii ko / kisii ko bhii} dekhaa. Q you ERG someone OBJ / someone OBJ even saw `Did you see {somebody / anybody} ?'
(ii) agar koii (bhii) fon kare, mujhe bataanaa. if someone even phone does me.oBj tell `If anybody (at all) calls
, tell me.'
((i) = Davison, pp. 27-28, (ii) = Haspelmath, p. 285, notation and gloss adapted)
As for English some and any, where they stand in opposition, as in the above trans-lation of (i), the former imparts the less remote possibility of the existence of an entity satisfying the description of the clause, than the latter. Similarly, anybody at all, as in (ii), implies the scarce possibility of the existence of an entity satisfying the description of the clause, more strongly than any solely does.
We can extend the claim of this article on the meaning of Oriya bi, to Hindi bhii, to explain this finding. Assuming that, like Oriya bi, Hindi bhii directs the depictor's attention to a non-immediate domain, we can say that, in the case of (i) and (ii), bhii is indicating that search must be more extensive than usual in order to come across an instance of the concept of the indefinite pronoun. Hence, the implication of the remote possibility of existence.
7) To give another instance, the notion of end focus distinguishes one usage from another of the English preposition over [Lakoff 1987]: a bridge over the river is ambiguous so that the bridge can be either situated on the other side of the river ("end focus") interpretation or overarching the river. The two interpretations share the imagery trajectory extending from this side of the river to the other side, which over the river denotes. In the end-focus interpretation, the final part of the trajectory is in focus.
8) Oriya has a number (three, in a moderate count) of types of relative clause construc-tions. In the discussion of Oriya in this article, by "relative clause (construction)", I mean the type commonly called correlative (construction). (Another type is in (17).) It is characterized, and distinguished from the others, by the following features: (i) it employs the relative pronoun, rather than a gap, for the marking of the relativization site; (ii) the word order is usually 'relative clause - main clause', rather than the reverse. The reasons for choosing this type rather thanothers in the comparison
180 Journal of the Japanese Association for South Asian Studies, 13, 2001
with the English relative clause construction are the following. First, it is structur-ally closely parallels to the English construction, containing the relative pronoun in it. Second, it can be regarded as the most "robust" in Oriya: in use, it occurs very frequently; in grammar, it serves as a mother scheme producing several derived constructions.
9) The term "entities" is used rather non-technically in (14) and the following. It subsumes the reference not only to individuals, but also to stages of an individual. The consideration of this causes a slight complication, but after all does not affect the points made in this article. See Yamabe [1998b: 4.1.2.2].
10) See Fabb [1990], for how the restrictive and non-restrictive relative clauses in En-glish differ syntactically. Notable in this connection is that the syntactic differences are all reflections of the semantic difference of restrictive and non-restive, the fact that any serious analysis must be able to explain.
11) I reason that Oriya and English are giving comparable information in the following manners. In Oriya, a majority of basic and simple expressions concern the appear-ances of things, so information concerning the things themselves are being given, either somewhat indirectly by way of the basic simple linguistic devices, or directly by way of marginal and more or less complicated devices whose primary denotations are keyed to that information. As for English, the reverse to Oriya holds.
I am thus adopting the logic of the reasoning that is wildly done in the literature with respect to the choice between aspect and tense as the major category of lan-guages for rendering the time-related properties of events. Some lanlan-guages (for instance, Slavic languages) give information about the temporal locations of events by way of the category of aspect, while others (for instance, languages of West Europe) give information about the views of events by way of the category of tense. For a recent development on this topic, see Bhat [1999].
12) Nakamura [1988] is a work that takes up this same task and does give technically detailed explications. While it diverges from the present paper in linguistic facts discussed and scholarly background, it looks into the same realms of cultural tradi-tion that the present paper wishes to look into. (However, the comparison of the two works remains a desideratum in this article. The languages they deal with differ (Sanskrit vs. Oriya), but in my view this point does not bear on the matters that would come to issue in the comparison of them.)
I thank Yoshifumi Mizuno for bringing Nakamura's book to my notice.
Abbreviations
CL = classifier, CP = conjunctive participle, GEN = genitive, FUT = future, Loc = locative, NOM = nominative, OBJ = objective, OPT = optative, p = plural, (Q) = interrogative pronoun, (R) = relative pronoun, s = singular,
1/2/3 = 1st/2nd/3rd person, * = unacceptable sentence.
Pronunciation
a = [a], as = [a], D/L/N/T = retroflexes, consonants + h = aspirates,
References
Bach, Emmon, Eloise Jelinek, Angelika Kratzer, and Barbara H. Partee (eds.), 1995, Quantification in Natural Languages. Dordrecht: Kluwer.
Bhat, D. N. S., 1999, The Prominence of Tense, Aspect and Mood. Amsterdam: John Benjamins.
Davison, Alice, 1978, "Negative Scope and Rules of Conversation: Evidence from an OV Language", Syntax and Semantics 9, pp. 23-45.
Dayal, Veneeta, 1995, "Quantification in Correlatives", in Bach et al. (eds.), pp. 179-205. •\•\ , 1996, Locality in Wh Quantification. Dordrecht: Kluwer.
Emeneau, Murray B., 1974, "The Indian Linguistic Area Revisited", International Jour-nal of Dravidian Linguistics 3, pp. 92-134. Reprinted in: 1980, Language and Linguistic Area: Essays by Murray B. Emeneau, ed. by Anwal S. Dil. Stanford: Stanford Univer-sity Press, pp. 197-249.
Fabb, Nigel, 1990, "The Difference between English Restrictive and Nonrestrictive Relative Clauses", Journal of Linguistics 26, pp. 57-78.
Fauconnier, Gilles, 1975, "Pragmatic Scales and Logical Structures", Linguistic Inquiry 6, pp. 353-375.
Foley, William A., 1997, Anthropological Linguistics. London: Blackwell.
Gil, David, 1995, "Universal Quantifiers and Distributivity", in Bach et al. (eds.), pp. 321-362.
Gonda, J., 1968, "The Sanskrit Particle api" , Lingua 21, pp. 183-196. Haspelmath, Martin, 1997, Indefinite Pronouns. Oxford: Clarendon Press.
Lahiri, Utpal, 1998, "Focus and Negative Polarity in Hindi", Natural Language Seman-tics 6, pp. 57-123.
Lakoff, George, 1987, Women, Fire, and Dangerous Things: What Categories Reveal about the Mind. Chicago: Chicago University Press.
Lakoff, George, and Mark Johnson, 1980, Metaphors we Live by. Chicago: University of Chicago Press.
Langacker, Ronald W., 2000, Grammar and Conceptualization. Berlin: Mouton de Gruyter. Nakamura, Hajime, 1988, Indozin no sii hoohoo (Ways of thinking of the Indians), definitive edition. Tokyo: Shunjusha. [English translation of an earlier edition: Hajime Nakamura, 1964, Ways of Thinking of Eastern Peoples. Honolulu: East-West Center Press.] Nuyts, Jan, and Eric Pederson (eds.), 1997, Language and Conceptualization. Cambridge:
Cambridge University Press.
Talmy, Leonard, 2000, Toward a Cognitive Semantics, twovolumes. Cambridge, Mass.: MIT Press.
Watanabe, Minoru, 1996, Nihongo gaisetu (An introduction to Japanese linguistics). To-kyo: Iwanami.
Whorf, Benjamin, 1956, Language, Thought and Reality: Selected Writing of Benjamin Lee Whorf, ed. by John B. Carroll. Cambridge, Mass.: MIT Press.
Yamabe, Junji, 1997, "The Meanings of the Relative and Interrogative Pronouns in Oriya" (in Japanese), in Proceedings of the 114th Meeting of the Linguistic Society of Japan, pp. 222-227, and Gengo Kenkyu 112, pp. 192-193.
•\•\
, 1998a, "The Meanings and Anaphoric Behavior of the Relative and Reflexive Pronouns in Oriya" (in Japanese), in Proceedings of the 117th Meeting of the Linguistic
182 Journal of the Japanese Association for South Asian Studies, 13, 2001 Society of Japan, pp. 204-209, and Gengo Kenkyu 115, p. 182. •\•\
, 1998b, The Relative and Interrogative Pronouns in Oriya. Doctoral dissertation, University of Tokyo. •\•\
, 1999, "The Meaning of the Particle bi in Oriya" (in Japanese), in Proceedings of the 119th Meeting of the Linguistic Society of Japan, pp. 165-170, and Gengo Kenkyu 117, p. 167.
