·ankfurt-Bern-New-York: Lang. !ern. Zur Gestaltung von Ansagetexten heim, unpublished master thesis. : eine Kommunikationsform der bes " xhergruppe, Kommunikationsbereich. :;.=.

chiques des messages laisses sur :'!":'==-• ws de semantique et de relations efiP'£ NEFRI-Strasbourg. Strasbourg, 427--+.!.:

:tures. Pragmatics, vol7, No.2, 125-1-­~~ais contemporain. Berne: Hermes. s. Semiotica 8/4: 289-327. mplest Systematics for the Organisarioc

II openings. In : Gumperz & D. H~ . ~

~lefonische Kommunikationstechisch no:· =1!!:11111: 1 & R. Fiehler, Technisierte

hrung. Berlin: Schmidt. M. Metzeltin & C. Schmitt (eds.), Lexi· _ ·188. c:tion? The case of the telephone answe 1g versus speaking Language, Text, Dis.-

onische Kommunikation, Bern, PTT M >ragmatica. Napoli: Liguori. The Telephone Answering Machine and -

n und Strukturen . Berlin: Akademie Veri2Z orterkommunikation zwischen Selbstd~~ 1iskurs, Heidelberg, Lang, (to appear). .nn (ed.), Arbeiten zur Konversations

icriptOral ia 40. TUbingen: Gunter Narr. Analysis of the structure and traditional 1dies II. 81-1 03.

TENSE-ASPECT CONTROVERSY REVISITED: THE-TA AND -RU FORMS IN JAPANESE

Y oko Hasegawa

troduction

of fhe verbal sufftxes -TA and -RU l.n Japanese are used to refer to a situation 'in past, t, or future time. Nevertheless, they are frequently glossed as past and nonpast , respectively, mainly because they refer only to past or nonpast time in mono­

sentences in isolation. The competing analysis claims that -TA and -RU are · y aspect markers, and that tense interpretation emerges as a derivative from the

- . .._...., .... 1 meanings. Despite ample work involving in-depth probes into the nature of markers, it is surprisingly rare to fmd research that compares and evaluates those ting analyses of -TA and -RU. The present study attempts such a critical inquiry.

After listing in section 2 typical uses of -TA and -RU, the present paper summarizes in n 3 six hitherto proposed analyses. Their strengths and weaknesses, based on

_."Sis of 19 Japanese novels taken from various periods and styles, are then discussed .sa:tion 4. Conclusions follow in section 5.

· o (1967:56-68) lists the uses of the -TA and -RU forms as follows (unless · se specified, examples are supplied by the present author; the verbals in English

~ ond-ing to those marked by -TA and -RU are in small capitals):

-T A form is used to express: 1) A situation completed in the past or that lasted for a certain length of time in

the past: haha ni tegami o kai-TA 'I WROTE a letter to my mother.' 2) Habitual repetition in the past: ano koro wa yoku undoo si-TA ' I USED TO

EXERCISE frequently in those days. ' (3) A situation that occurred under a certain circumstance in the past: kaeru to

sugu ni te o arat- TA 'I WOULD WASH my hands immediately after I got home. ' (4) A situation that materialized in the past and still exists: ookiku nat-TA ne

'You'VE GROWN, haven't you?'

226 Yoko Hasegawa

(5) Sudden discovery that a certain state has continuously existed (the so­TA form of discovery; a special case of (4): a, soko ni i-TA no 'Oh, ARE!'

(6) Sudden recalling of a future event or plan that the speaker has definite: a, asita siken ga at-TA 'Oh, I HAVE an exam tomorrow!'

(7) A request for hearer's confirmation of a fact (normally only in q anata wa donata desi-TA ka 'Who ARE you?' (Mikarni 1953).

(8) Proclamation or assertion of the realization of a situation which has realized: yosi kat-TA 'All right, (I'm sold) I'LL BUY it!' (Suzuki 1965).

(9) A command: doi-TA, doi-TA 'STEP back! STEP back!'

-Ru (-u) form is used to express (-u = allomorph of //Ru/f) 1

(10) A present state: kono isu wa sakki kara koko ni ar-u 'This chair HAS here for a while' (Mikarni 1953).

(II) A present psychological state: hara ga tat-u 'I'M ANGRY.' (12) A speech act (hatsugen-genzai 'the utterance present'): yakusoku su­

PROMISE it.' (13) An event occurring in front of one's eyes: a, teppan ga oti-RU 'Oh. a

plate HAS FALLEN down!' (Suzuki 1965). (14) A situation that is certain to occur in the future: asita wa yasumi da

COPULA-RU) 'Tomorrow IS a holiday.' (15) The speaker's present intention or plan: asita amerika ni tat-u 'I'LL LEA=.

America tomorrow.' (16) A past event: kikizute naranai koto o i-u ne ' You'VE SAID something I

ignore.' (17) Habitual repetition: itumo 12-zi ni ne-RU 'I always GO TO BED at 12.' (18) A step by step procedure (e.g. in recipes): tugi ni sio o ire-RU 'Next.

some salt.' (19) A situation regularly occurring under a certain condition: tabi o suru to

manab-u 'We LEARN a lot by travelling.' (20) A characteristic or general truth: ana hito wa yoku syaber-u 'He TALKS a (21) A command: sugu ni tabe-RU 'EAT it now!'

Although it is questionable whether those items are all distinctive, they show that the­form can be used to refer to a situation in the past (1-3 ), in the present ( 4-7), or iL future (8-9), and the -RU form to a situation in the past (16), in the present (10-13), iL future (14-15), or to an atemporal one (17-21 ).

3. Previous Analyses 3.1. Tense Analysis

Those who advocate that -TA and -RU are primarily tense markers contend that, wh complications occur in multi-clausal sentences and/or in discourse, they refer only to or nonpast time in mono-clausal sentences in isolation.

Tense-aspect contl

2-) kinoo wa itinitizyuu uti ni i-T,~o ' I STAYED home all day yesterc

23) asita kaimono ni *it-TA!ik-u. 'I'LL GO shopping tomorrow.'

- those markers are rarely used ·- can deal with only very limited c

· 6) as exceptions. Furthermore, bee =ounted cases, this analysis is the lea :he fact that -TA and -RU frequently

another problem to the tense an; ·on should be regarded as a rhet it is widely agreed, has a tense S)

for a past situation - the so-c; ·ve present is considered by many ·olfson 1979, Schiffrin 1981, Fleiscl

~ J.th the exception of the uses of -rz. :x>ssible for an author to use only the ~ound ' or 'background.'

2 Likewise,

0.: possible to use nothing but the nc events described. In the former case fact way, while in the latter it will b ;:vents himself. In this sense, there constitute an element of the effective

wing Hopper (1979), Soga conte ·ve are frequently expressed with -1

· g the discourse elusive. Analyzing Soseki Natsume's Sans

, the forgroundlbackground distinc - call them tense markers). She points

be ended with -TA, as shown in (24-2 :.cpnent are Terakura's.)

(24) CD hirota-san wa hige no sita o motte-i-RUI??motte-i-TA.

'CD Mr. Hirota SMILED showir good-looking teeth. '

(25) CD "odeni narimasen ka" to no mama kubi o yoko ni hut-r.

'CD "Wouldn't you like to go faintly and SHOOK his head ~ gesture.'

:e has continuously existed (the so of (4): a, soko ni i-TA no 'Oh, there

or plan that the speaker has kno~­I HAVE an exam tomorrow!' 1 of a fact (normally only in questi tE you?' (Mikami 1953). lization of a situation which has not -;old) I'LL BUY it!' (Suzuki 1965). ck! STEP back!'

orph of //Ru/!)1

kara koko ni ar-u 'This chair HAS

·a tat-U ' I'M ANGRY.'

utterance present'): yakusoku su-R~

s eyes: a, teppan ga oti-Ru 'Oh, a 55). li1 the future: asita wa yasumi da .

m: as ita amerika ni tat-u 'I'LL LEA\~

J i-u ne 'You'VE SAID somethling I

·RU ' IalwaysGOTOBEDat 12.' ~cipes): tugi ni sio o ire-Ru 'Next -

a certain condition: tabi o suru to tg.' hito wa yoku syaber-u 'He TALKS a' now!'

rre all distinctive, they show that the -ast (1-3), in the present (4-7), or C. ! past (16), in the present (10-13), ir:

ily tense markers contend that, wh .dJor in discourse, they refer only to. 'Ifion.

Tense-aspect controversy revisited: The -ta and -ru f orms in Japanese 227

(22) kinoo wa itinitizyuu uti ni i-TAI*i-RU. ' I STAYED home all day yesterday.'

(23) asita kaimono ni *it-TAiik-u 'I'LL GO shopping tomorrow.'

3ecause those markers are rarely used in isolated mono-clausal sentences, the tense l:Ja!ysis can deal with only very limited data and usage: it needs to treat the uses in ( 4-9)

d (16) as exceptions. Furthermore, because it inevitably relies on auxiliary theories for ccounted cases, thlis analysis is the least authentic. The fact that -TA and -RU frequently alternate in written as well as spoken discourse

ses another problem to the tense analysis. Advocates commonly argue that their 'ternation should be regarded as a rhetorical phenomenon, as in the case of English,

which, it is widely agreed, has a tense system, although its nonpast/present tense can be '!Sed for a past situation - the so-called historical/narrative present. Historical! _ ·arrative present is considered by many to be a stylistic device with limited distribution :f. Wolfson 1979, Schiffrin 1981 , Fleischman 1990). Soga (1983:219) states:

With the exception of the uses of -ru required by grammatical restrictions, it is quite possible for an author to use only the -ta form regardless of whether an event is ' fore­

ground' or 'background.'2

Likewise, although it may not be very common, it should be possible to use nothing but the non-past tense form regardless of the types of the events described. In the former case, the story will be perceived only in a matter-of­fact way, while in the latter it will be perceived as if the reader is experiencing the events hlimself. In thlis sense, therefore, it seems that proper uses of tense forms constitute an element of the effective specific style of an author or of a story.

7 ollowing Hopper (1979), Soga contends that background statements in Japanese =.mative are frequently expressed with -RU, but it is possible to change it to -TA without ::laking the discourse elusive.

Analyzing Soseki Natsume's Sanshiro, Takakura (1990) acknowledges that in ~era!, the forgroundlbackground distinction accurately predicts the choice of -TA or -RU

she call them tense markers). She points out, however, that not all sentences in narrative :an be ended with -TA, as shown in (24-25) ®. (The translation and the appropriateness _ "Jdgment are Terakura's.)

(24) CD hirota-san wa hige no sita kara ha o dasite warat-TA. ® wariai kireina ha o motte-i-RUI? ?motte-i-TA. 'CD Mr. Hirota SMILED showing his teeth below his moustache. ® He'D GOT good-looking teeth. '

(25) CD "odeni narimasen ka " to kiku to, sensei wa sukosi warai nagara, mugon no mama kubi o yoko ni hut-TA. ® kodomo no yoona syosa o su-Rul??si-TA. 'CD "Wouldn't you like to go out?" asked Sanshiro, and then Sensei smiled

faintly and SHOOK his head without saying a word. ® That WAS a childlike gesture.'

228 Yoko Hasegawa

Takakura confirms that sentences representing perception, exemplified by (24) (Z. normally occur with -RU, but they permit the TA-RU alternation. When -TA is selected, she explains, the percipient disappears from the narrative, and the described situation is conceived as an objective fact in the voice of the narrator. Significantly, she informs thai in sentences in the free indirect style -representing a character's consciousness -exemplified by (25) ®, -RU resists replacement with -TA. This fact poses anothe;

problem for the tense analysis because, as demonstrated in (25) ®, the represented ide2. entered the character's consciousness in the narrative past, and thus the use of the "past tense" must be allowed, at least as a possible variation. This is not the case, however, as Terakura points out.

Still another problem with the tense analysis is that, as Soga acknowledges, there are grammatical restrictions on the usage of -TA and -RU. For example, in certain subordinate clauses, they do not contrast, i.e. either -TA or -RU can appear, but not both, regardless of the matrix time or mood of the sentence, as shown in (26-27).

(26) mado o ake-TAI*ake-RU ra, kanarazu simete kudasai ne. ' Please make sure to close it if you OPEN the window. '

(27) mado o *ake-TAiake-RU to, tori ga haitte ki-TA. ' When I OPENED the window, a bird FLEW in.'

Unless an explanation of why some connectives require a specific "tense" is provided, the claim of -TA and -RU as primarily tense markers cannot be upheld in a straightforware manner, and, consequently, a more elaborate analysis is needed.

3.2. Time-Relationship Analysis

Ota (1972) argues for two tense types- (a) a deictic notion and (b) 'time-relationship·

indicating an alignment of event and reference times (a Ia Reichenbach)? He contends that, while English has both types, Japanese encodes only time-relationships: -T;,

indicates event time being prior to reference time, and -RU otherwise. If no reference time is specified, speech time serves as reference time, making -TA and -RU resemble deictic tenses. In a dependent construction, e.g. subordinate clause, reference time is supplied b~

the construction on which it depends. Miura (1974) points out that, in dependent constructions, not only the matrix-clause

time, but also speech time, can serve as reference time. For example:

(28) kanai no ike-TA hana ga asita no tenrankai ni de-RU. ' Flowers that my wife HAS ARRANGED/WlLL ARRANGE will be displayed in tomorrow's exhibition. '

In one reading, the flowers were arranged before speech time, and in the other, they wii: be arranged before the exhibition, but not before speech time. In either case, -TA indicates anteriority.

It is frequently commented that the usage of -TA and -RU in Japanese discourse appears to be bewildering or even illogical to non-native speakers. Isshiki (1965), for

Tense-aspect con

example, claims that in a long discourst cumbersome once the passage is establi: have very little sense of time. Argueing -TA or -RU is not arbitrary in the time-1 example taken from Yasunari Kawabata'

(29) CD "osida," to singo wa tub1

'CD "This one' s mute," Shin1 that had sung so loudly.'

Miura explains that in (29) CD, the refere

in the past established by CD. He write~ past (or .. . in Shingo's place). . .. This : Japanese and frequently occurs in narrati is whether it is necessary to assume that because ® is in the free indirect stylt

Terakura. Had ® appeared with -TA, t narrator, not of Shingo (which is the case

The problem with the time-relations marked events must be understood to ha time, e.g. (27) and (30).

(30) naite tanom-U kara kane o ka 'Because you IMPLORED me, I

In both (27) and (30), the -RU event has c clause -TA event. The time-relationship a before the-TA event - which is not the c<

3.3. Aspect Analysis

Miyoshi (1974:24) claims, "Japanese has forms for past and present [nonpast] confusion for the reader. " However, if th to what do the "forms for past and present do, that -TA and -RU are primarily tense discourse, it may not be appropriate to suggested by Soga.

Responding to Miyoshi, Miller (1975

markers of perfective and imperfective, res

'Tense' is a grammatical term that languages in which the different fa concerned with identifying, markir

perception, exemplified by (24) ®, J alternation. When -TA is selected, she rative, and the described situation is tarrator. Significantly, she informs that :nting a character's consciousness -t with -TA. This fact poses another strated in (25) ®, the represented idea tive past, and thus the use of the "past 1tion. This is not the case, however, as

s that, as Soga acknowledges, there are ~u. For example, in certain subordinate can appear, but not both, regardless of in (26-27).

imete kudasai ne. N the window.' te ki-TA. oW in.'

quire a specific "tense" is provided, the cannot be upheld in a straightforward sis is needed.

:ictic notion and (b) 'time-relationship'

nes (a Ia Reichenbach).3 He contends encodes only time-relationships: -TA and -RU otherwise. If no reference time :, making -TA and -RU resemble deictic ate clause, reference time is supplied by

)nstructions, not only the matrix-clause time. For example:

rnkai ni de-RU. ~D/WILL ARRANGE will be displayed in

speech time, and in the other, they will peech time. In either case, -TA indicates

of -TA and -RU in Japanese discourse on-native speakers. Isshiki (1965), for

Tense-aspect controversy revisited: The-ta and -ruforms in Japanese 229

example, claims that in a long discourse, "rigid sequencing of tenses" makes the passage cumbersome once the passage is established to refer to past events, and that the Japanese have very little sense of time. Argueing against Isshiki, Miura asserts that the selection of -TA or -RU is not arbitrary in the time-relationship analysis. He provides the following example taken from Yasunari Kawabata's Yama no oto.

(29) CD "osida," to singo wa tubuyai-TA. ® gyatto itta semi to wa tiga-u. 'CD "This one's mute," Shingo MUTTERED. ® It WAS different from the one that had sung so loudly.'

Miura explains that in (29) CD, the reference time is speech time, but in ®, it is the point in the past established by CD. He writes, "The author is suddenly putting himself in the past (or ... in Shingo's place). . .. This shifting in point of reference is characteristic of Japanese and frequently occurs in narratives" (98). The crucial question to be raised here is whether it is necessary to assume that the reference point is shifted. It turns out that, because ® is in the free indirect style, the use of -RU is predictable, as argued by Terakura. Had ® appeared with -TA, the statement would have been the voice of the narrator, not of Shingo (which is the case in the English translation).

The problem with the time-relationship analysis is that, as Ota keenly observes, -RU

marked events must be understood to have occurred and completed before the reference time, e.g. (27) and (30).

(30) naite tanom-u kara kane o kasite yat-TA no da. (Onoe 1995) 'Because you IMPLORED me, I LENT you the money.'

In both (27) and (30), the -RU event has clearly occurred and completed before the matrix clause -TA event. The time-relationship analysis predicts that the -RU event has occurred before the-TA event- which is not the case.

3.3. Aspect Analysis

Miyoshi (1974:24) claims, "Japanese has no clearly established grammatical tense, and forms for past and present [nonpast] are often interchanged without creating any confusion for the reader." However, if there is no clearly established grammatical tense, to what do the "forms for past and present" refer? If one considers, as Miyoshi appears to do, that -TA and -RU are primarily tense markers but used interchangeably in Japanese discourse, it may not be appropriate to investigate them in the realm of grammar, as suggested by Soga.

Responding to Miyoshi, Miller (1975) claims that -TA and -RU are aspect, not tense,

markers of perfective and imperfective, respectively.4 He defines these notions as:

'Tense' is a grammatical term that is generally reserved for the verb systems of languages in which the different forms of the verbs are essentially, or mainly, concerned with identifying, marking, or ... 'expressing' differences in time,

230 Yolw Hasegawa

indicating when the action or event to which the verb has reference took place. . .. 'Aspect' is a grammatical term that is generally reserved for the verb systems of languages in which the different forms of the verb are concerned not with when something was done, but how, particularly and typically (e.g., the Slavic languages) with whether or not a given action was or is completed (over and done with) or was or is still continuing (not over and done with, uncompleted or incomplete). (3-4)5

Using Kusamakura by Soseki Natsume, Miller provides numerous convincing examples. (The free translation is from Turney ( 1965).)

(31) CD ... rokuzyoo hodo no tiisana zasiki ni ire rare-TA. @ mukasi kit a toki to wa maru de kentoo ga tiga-u.

'CD I WAS eventually shown into a small room about twelve feet by nine. @ This WAS not at all as I had remembered the place from the last time I was there. '

(32) CD ... mooretu naru undoo o kaisi si-TA. @ kono tume ga, ... sippuu no sokudo de tooru gotoku ni oorai su-RU.

'CD (The barber) BEGAN a violent exercise. @ His fingernails ... WENT back and forth at whirlwind speed.'

In (31), the verb CD irerare- TA ' be shown into+ TA ' is perfective, whereas @ tiga- u ' be different+ RU' is imperfective. In (32), the verb CD kaisi si-TA 'begin+ TA' (perfective) contrasts with @ oorai su-RU 'go back and forth+ RU' (imperfective).

The aspect analysis is in accordance with the idea that more salient concepts than tense per se are (i) whether or not a certain change has occurred, and (ii) whether or not the speaker is certain about its occurrence. If the speaker is confident about a given situation having been completed (the notion of perfective), the addressee naturally interprets it as having occurred in the past. Tense thereby emerges as something secondary. Tense, which grasps time linearly as a past-present-future continuum, is the

more tangible one, for it can be analyzed independently of the speaker's mental state. 6

Therefore, researchers tend to consider the notion of tense as most basic. Diachronically, however, tense is secondary to aspect in Indo-European languages (Kuryiowicz 1964, Izui 1967, Bybee 1985), and Pidgin and Creole systems are inherently aspectual (Giv6n 1982, Kotsinas 1989).

One technical difficulty that should be addressed here is that, as Bybee et al. ( 1994) point out, if a given language has both perfective and imperfective aspects, they must contrast and should be mutually exclusive. Japanese has the -TE 1- form, a well-agreed­upon imperfective construction, and it freely co-occurs with either -TA or -RU, as shown in (33) @ .

(33) CD akiko ga ki-TA. @ aikawarazu kuroi huku o ki-TEI-TAiki-TEI-RU.

'CD Akiko CAME. @ As usual, she WAS WEARING a black dress.'

This implies either -TA or -RU are not genuine (or typical) aspect markers, or that they mark a different kind of aspect from the -TE 1- form.

Tense-aspect conn

What Miller has in mind in fact der analysis. He claims that the perfective ar by the meaning of the predicate. (The tra

(34) CD "sensei, sensei " to hutako e kao o das- u. @ boosi wa A 'CD "Sensei, sensei," she CA

RAISED my head above the qt

In (34), CD kake-TA ' call + TA ' and @ oti

@ das- 'put out one's head ('raise' in tl perfective as well, for it is a punctual v Miller explains that das- is marked as iJ his head out over the quince bush and following passage" (Miller 1975: 13).

The significance of this claim is tl reflect not the completion or incompleti• rhetorical intention of the statement. Th; because they are worth mentioning, but above the bush enabled the observatic because Kusamakura is in first-person n; without event ® , it is impossible to narr:

This usage resembles the use of · introduces a precondition for the matri conjoined, the connective TO would be t predicts that such a use should be abse1 omniscient. Of the 19 novels from va narratives. There were several problema

(35) CD siawase to wa sono t1 @ tetubin no yu o yutanpo

@) netuku made no aida, kai de keisan o site miru no dat­

'CD She COULDN'T RECALL

thing. @ She POURED hotw a cold bed. @) Scrunching fell asleep.' (Tatsuzo Ishika

Miller' s analysis can account for the D

occurrence of @ . However, as shown 1: @ ,it is natural to consider that ® and (

sentences. (In fact , deviding @ and @ ; Another problem with the aspect

sentences like (36) cannot be regarded a

1e verb has reference took place. . .. ly reserved for the verb systems of verb are concerned not with when

typically (e.g., the Slavic languages) npleted (over and done with) or was :ompleted or incomplete). (3-4i

des numerous convincing examples.

rerare-TA. @ mukasi kita told to wa

-oom about twelve feet by nine. @ . the place from the last time I was

@ kono fume ga, . . . sippuu no

·· @ His fingernails ... WENT back

: perfective, whereas@ tiga- u 'be kaisi si-TA 'begin+ TA' (perfective) (inlperfective ).

ea that more salient concepts than LS occurred, and (ii) whether or not :peaker is confident about a given ~rfective), the addressee naturally >e thereby emerges as something st-present-future continuum, is the 1tly of the speaker's mental state.6

:nse as most basic. Diachronically, 1 languages (Kurylowicz I964 Izui : inherently aspectual (Giv6n I982,

here is that, as Bybee et al. (1994) .d inlperfective aspects, they must has the -TE 1- form, a well-agreed­with either -TA or -RU, as shown in

ru 0 ki-TE 1-TA/ki-TE 1-RU. IRING a black dress.'

>ical) aspect markers, or that they

Tense-aspect controversy revisited: The-ta and -ruforms in Japanese 231

What Miller has in mind in fact departs from the traditional perfective-imperfective analysis. He clainls that the perfective and inlperfective aspects are not determined solely by the meaning of the predicate. (The translation in (34) is mine.)

(34) CD "sensei, sensei" to hutakoe kake-TA . ... @ "nan desu" to yo wa bake no ue e kao o das-u @ boosi wa kusahara e oti-TA.

'CD "Sensei, sensei," she CALLED twice .... @ "Yes, what do you want?" I RAISED my head above the quince bush. @ My hat FELL on the grass.'

In (34), CD kake-TA 'call + TA' and @ oti-TA 'fall + TA' are perfective and so marked, but @ das- 'put out one's head ('raise' in this particular context)' is naturally interpreted as perfective as well, for it is a punctual verb and the event occurred between CD and @. Miller explains that das- is marked as inlperfective here because the protagonist "sticks his head out over the quince bush and leaves it visible there continuously during the following passage" (Miller I975: 13).

The significance of this clainl is that the Japanese aspect system is considered to reflect not the completion or incompletion of a given situation, but, rather, the encoder's rhetorical intention of the statement. That is, in (34), three events are presented not solely because they are worth mentioning, but also because the protagonist's raising his head above the bush enabled the observation of event @. This explanation is plausible because Kusamakura is in first-person narrative, i.e. the narrator is not omniscient. Thus, without event@, it is inlpossible to narrate event@.

This usage resembles the use of -RU in TO subordinate clause, which frequently introduces a precondition for the matrix event. In fact, if (34) @ and @ were to be conjoined, the connective TO would be the only natural choice. In addition, this analysis predicts that such a use should be absent in third person narrative where the narrator is omniscient. Of the I9 novels from various periods I investigated, II are third-person narratives. There were several problematic cases.

(35) CD siawase to wa sana teido no mono da to sika, kangaerarenakat-TA. @ tetubin no yu o yutanpo ni ire-RU @ sosite tumetai nedoko ni hai-RU ® netuku made no aida, kanozyo wa karada o tizimete, mata atama no naka de keisan o site miru no dat-TA.

'CD She COULDN'T RECALL happiness as anything more than such a trivial thing. @ She POURED hotwater into a foot warmer. @ And she SLIPPED into a cold bed. ® Scrunching herself up, she THOUGHT about it again until she fell asleep.' (Tatsuzo Ishikawa, Seishun no satetsu)

Miller's analysis can account for the relationship between (35) @ and ®, but not the occurrence of@. However, as shown by the conjunctive sosite 'and' at the beginning of @, it is natural to consider that @ and @ form a unit, so that the author used -RU in both sentences. (In fact, deviding@ and@ and marking both with -RU sounds deviant to me.)

Another problem with the aspect analysis is that, as Suzuki (I976) points out, sentences like (36) cannot be regarded as perfective.

232 Yoko Hasegawa

(36) ldnoo tornatta ryokan ni wa niwa ni ooldna rnatu no ld ga at-TA. 'The inn I stayed at yesterday HAD a big pine tree in the garden.'

The aspect analysis predicts that the garden no longer exists, which is improbable. And the narrator is likely to continue talking about events in which the garden plays some significant role. Thus the narrator's rhetorical intention should select -RU, but (36) sounds perfectly natural.

3.4. Dual-Function Analysis

This may be the most widely-accepted analysis. Teramura (1971) points out that -TA is ambiguous in such sentences as (37), and that this ambiguity becomes overt when the sentence is negated. Here, Teramura contrasts imperfective with perfect, rather than

perfective.7

To answer question (37), (38) with -TA is selected to negate the past-tense reading, or (39) with -TE 1-RU to negate the perfect reading.

(37) hirurnesi o tabe-TA ka. 'DID you eat lunch?' or 'HAVE you eaten lunch? '

(38) iya, tabenakat-TA. 'No, I DIDN'T EAT lunch. '

(39) iya, tabete-1-NA -1. (NA-J= NEGATIVE-RU) 'No, I HAVEN'T EATEN lunch.'

It is, however, worth questioning whether or not the existence of two possible answers proves that (37) is indeed ambiguous. It may be the case that it is unambiguous, but one may respond to the inferred intention of the questioner: it is possible to respond to the question "Did you have lunch?" with "No, i haven't eaten anything today."

Like Teramura, Nakau (1980) considers that -TA and -RU primarily indicate tense but they also signify aspects. His argument is that, in an adverbial clause, they designate either tense or aspect, depending on the nature of the clause connective. For example, with noni ' although' , they mark tense, whereas with told 'when', they mark aspect. In ( 40) with noni, the connected clauses can carry different time adverbials, but in ( 41) with told, they cannot.

(40) ldnoo arne ga hut-TA noni, kyoo wa atu-1. (-1= allomorph of //Ru/1) 'Although it RAINED yesterday, it's hot today. '

(41) *kinoo arne ga hut-TA toki (wa), kyoo wa atu-1. ' When it RAINED yesterday, it's hot today.' (Intended; this is impossible in both Japanese and English.)

Nakau explains that in (40), both -TA and -RU function as tense markers, i.e. the two clauses have different tenses, and, consequently, two time adverbials are permitted. In ( 41 ), by contrast, because -TA is an aspect marker, the entire sentence has only one tense encoded by the matrix clause -RU, and thus the sentence can take only one adverbial .

Kunihiro (1980) disagrees with Nakau, however, by suggesting that recognition of the aspect-marking function alone is sufficient to account for the difference between (40)

Tense-aspec

and (41). The reason that -TA is i1 nothing to do with the tense-aspe• concessive noni, which requires the

The dual-function analysis is tt for the usage of -TA and -RU, it is no

3.5. Perspective Analysis

Like Miller, Kunihiro (1967) cons Instead, he argues that -TA indicates of the situation having been materi< an unspecified person' s subjective . viz. the person is certain about the has already materialized. I will call

The perspective analysis is no• analysis. Kunihiro cites Mikarni ( and (1 0) in section 2 with -RU, re situation, the former is understood the latter as a direct expression ofh

( 42) kono isu wa sakld kara 'This chair HAS BEEN he

(43) kono isu wa sakld kara

Kunihiro explains that if a person asserts it as such, the situation is m consequently, -TA appears to be a p< encode tense, it can occur in a ser conditional clause, e.g. (44).

(44) arne ga hut-TAra siai w ' If it RAINS, the game Wi

Although he can explain why -T,

provide an explanation as to why -1

Regarding the usage of -RU, K naturally associated with the speec (10-13) above. When the verb is p the realization of the situation 1, Kunihiro contends that this fact ex1 future, and (15), i.e. the speaker' s the usage in ( 16), i.e. a past event the statement to be about a past without assertion - is compatibl1

(17-21) - atemporal situations -

matu no ki ga at-TA. ~ tree in the garden.'

exists, which is improbable. And ; in which the garden plays some 1tion should select -RU, but (36)

mura (1971) points out that -TA is nbiguity becomes overt when the :rfective with perfect, rather than ; selected to negate the past-tense ng.

tch?'

:xistence of two possible answers tSe that it is unambiguous, but one :r: it is possible to respond to the en anything today." td -RU primarily indicate tense but 1 adverbial clause, they designate clause connective. For example, oki 'when' , they mark aspect. In tt time adverbials, but in ( 41) with

1 = allomorph of I /Ru/ I) '

1-1.

e and English.)

Jn as tense markers, i.e. the two ime adverbials are permitted. In !ntire sentence has only one tense can take only one adverbial. by suggesting that recognition of nt for the difference between ( 40)

Tense-aspect controversy revisited: The -ta and -ru f orms in Japanese 233

and (41). The reason that -TA is interpreted as referring to a past situation in (40) has nothing to do with the tense-aspect distinction; rather, it is due to the nature of the concessive noni, which requires the presupposition of the referent of the noni clause.

The dual-function analysis is the last resort. If a uni-functional analysis can account for the usage of -TA and -RU, it is not necessary to posit two equally-weighted functions .

3.5. Perspective Analysis

Like Miller, Kunihiro (1967) considers that neither -TA nor -RU encodes tense proper. Instead, he argues that -TA indicates an objective judgment made by an unspecified person of the situation having been materialized at a certain point in time, and that -RU indicates an unspecified person's subjective judgment regarding the realization of the situation -viz. the person is certain about the realization of the situation, but does not assert that it has already materialized. I will call this the perspective analysis.

The perspective analysis is not subject to the problem of (36) raised for the aspect analysis. Kunihiro cites Mikarni (1953), who observes that while (42) below with -TA

and (10) in section 2 with -RU, restated here as (43), can be used to depict the same situation, the former is understood as an indirect report of the speaker's experience, and the latter as a direct expression of his/her perception.

(42) kono isu wa sakki kara koko ni at-TA. 'This chair HAS BEEN here for a while.'

(43) kono isu wa sakki kara koko ni ar- u.

Kunihiro explains that if a person is confident about the realization of a situation and asserts it as such, the situation is naturally understood to have taken place in the past, and consequently, -TA appears to be a past-tense marker. However, because -TA itself does not encode tense, it can occur in a sentence used as an imperative, e.g. (9), as well as in a conditional clause, e.g. ( 44).

(44) arne ga hut-TA ra siai wa tyuusi desu. ' If it RAINS, the game will be cancelled.'

Although he can explain why -TA is not incompatible with conditional, he does not provide an explanation as to why -TA should be selected.

Regarding the usage of -RU, Kunihiro explains that temporally neutral formatives are naturally associated with the speech time, and that this principle accounts for the uses of (10-13) above. When the verb is punctual, speaker' s certainty about, but not assertion of, the realization of the situation logically implies its occurrence to be in the future . Kunihiro contends that this fact explains (14), i.e. a situation that is certain to occur in the future, and (15), i.e. the speaker's present intention or plan. Furthermore, he states that the usage in (16), i.e. a past event, depends heavily on a context that clearly guarantees the statement to be about a past situation, and that the meaning of -RU - certainty without assertion - is compatible with this function. Finally, Kunihiro states that for

(17-21)- atemporal situations- -RU is the most natural selection.8

234 Yoko Hasegawa

One problem with this perspective analysis of objective -TA and subjective -RU is that, when both -TA and the -TE I- plus -RU refer to a past situation, a salient difference between them is that -TA implies subjectivity, and -TE I-RU objectivity - contrary to Kunihiro. Inoue (1978) argues that the -TE 1-RU construction marks evidentiality; viz. the described situation is not part of the speaker's direct experience. She refers to this function as reportative, since it is frequently found in journalistic writing conveying objectivity and availability of evidence. Comparing the sentences in (45-46), she concludes that -TA lacks this reportative sense. (Grarnmaticality judgments are Inoue's.)

(45) zyon wa 10-nen mae ni zyeneraru mootaazu o yamesaserare-TE 1-RUI ?yamesaserare-TA.

'John was laid off by General Motors ten years ago.' (46) watasi wa onazi tosi ni kuraisuraa o *yamesaserare-TE 1-RU!yamesaserare-TA.

'I was laid off by Chrysler in the same year.'

She explains that the use of -TE I-RU in (46) is anomalous because it implies that the speaker has forgotten about the lay-off s/he personally suffered, and hence must check some document regarding his/her own past employment - an unlikely situation. While some native speakers may disagree with Inoue's grarnmaticality/appropriateness judgments, it is true that -TE 1-RU frequently conveys the idea that the asserted situation is not part of the speaker's direct experience.9

3.6. Grounding Analysis

Onoe (1995) claims that Japanese has neither tense nor aspect as a morphosyntactic category, and that the contrast between -TA and -RU cannot be expressed in terms of tense or aspect. Rather, he advocates that provision of the characteristics of -TA, -RU, -TE I-RU,

and -TE I-TA as predicative forms and clarification of their mutual relationships are essential.

He argues that the contrast between -TA and -RU does not involve a tense opposition because, as we have already seen, -TA need not refer to past time, and -RU need not refer to nonpast time. Between -RU and -TE I-RU, the opposition cannot be perfective vs. imperfective, respectively, because the latter can also be perfective, as shown in (47).

(47) sono koro made ni wa kono koozi wa owat-TEI-RU. 'The construction WILL BE FINISHED by that time.'

He characterizes -TA as the form to describe a situation in relation to the point in time at which the speaker recognizes it to have already happened. When -TA appears in the matrix clause, this time point is normally speech time. -TE I-RU marks either the perfect or the progressive. Because, in terms of aspect, the notion of perfect corresponds to the perfective, and the progressive to the imperfective, and because -TE I-RU covers both, Onoe concludes that Japanese does not have an aspectual system. He characterizes -TE I­

TA as an indicator of the pluperfect, past progressive, present perfect, or the discovery present, as exemplified in (5). Because -TE I-TA as a whole can indicate a situation

Tense-a;

connected with speech time, i.e. 1

cannot be derived compositional uses and comments that its typic: encodes present or future time. ( a situation as a type. He cone temporality,' but -RU is not.

Onoe's analysis is insightful decomposed into -TE I- and -TAl­

a past specific situation, -TA en what the perspective claims. If' objection becomes irrelevant.

The problem with Onoe's ar investigation of narratives. In situations are narrated: CD th spectacular (-TA), ®she poured bed (-RU), and ® she thought al clear in what sense ( 48) CD and (

( 48) CD siawase to wa ® tetubin no yu o . ® netuku made no 1

de keisan o site min 'CD She COULDN'T 1

thing. (V She POUR

a cold bed. ® Sen fell asleep.' (Tatsuz

4. Discussion

Section 3 summarized six anal: aspect, (iv) perspective, (v) dua organized into three groups ac temporal sequentiality (i-ii), CB: iv), (C) a combination of bot! because it assumes that -TA is t -RU is free from such temporal assessment of such analyses anc

I do not have much to sa problems by positing dual fu satisfactory characterization of·

In type A, the tense an~

applicability; (ii) some past eve (iii) it cannot explain why cei1 event time. The time-relationsl

:ive -TA and subjective -RU is : situation, a salient difference RU objectivity - contrary to m marks evidentiality; viz. the {perience. She refers to this ournalistic writing conveying b.e sentences in (45-46), she :1lity judgments are Inoue's.)

>amesaserare-TE 1-RUI

tgo.' ~rare-TE 1-Rulyamesaserare-TA.

us because it implies that the tffered, and hence must check - an unlikely situation. While ~ammaticality/appropriateness

ea that the asserted situation is

aspect as a morphosyntactic be expressed in terms of tense cteristics of -TA, -RU, -TE 1-RU, :heir mutual relationships are

not involve a tense opposition st time, and -RU need not refer ition cannot be perfective vs. rfective, as shown in ( 4 7).

-RU.

in relation to the point in time 1ed. When -TA appears in the E 1-RU marks either the perfect 1 of perfect corresponds to the because -TE I-RU covers both, :ystem. He characterizes -TE !­

:sent perfect, or the discovery vhole can indicate a situation

Tense-aspect controversy revisited: The -ta and -ruforms in Japanese 235

connected with speech time, i.e. the present perfect, he argues that the meaning of -TE I-TA cannot be derived compositionally from -TE 1- and -TA. Regarding -RU, Onoe posits 18 uses and comments that its typical uses do not involve any temporal concept, i.e. it rarely encodes present or future time. Consequently, he characterizes -RU as the form to present a situation as a type. He concludes that -TA, -TE 1-RU, and -TE I-TA are grounded in temporality,' but -RU is not.

Onoe's analysis is insightful, and his proposal that -TE I-TA and -TE 1-RU should not be decomposed into -TE 1- and -TAI-RU saves the perspective analysis. Recall that to refer to a past specific situation, -TA encodes subjectivity, and -TE 1-RU objectivity, contrary to what the perspective claims. If we do not posit the existence of-RUin -TE 1-RU, then my objection becomes irrelevant.

The problem with Onoe's analysis is that it is difficult to apply it straightforwardly to investigation of narratives. In the passage in (35), restated as (48), for example, four situations are narrated: CD the character could not recall happiness as something

spectacular (-TA), ® she poured hotwater into a foot warmer (-RU), @ she slipped into a

bed (-RU), and @ she thought about happiness again until she fell asleep (-TA). It is not

clear in what sense ( 48) CD and @ are grounded, but ® and @ are not.

(48) CD siawase to wa sono teido no mono da to sika, kangaerarenakat-TA. ® tetubin no yu o yutanpo ni ire-RU. @ sosite tumetai nedoko ni hai-RU. @ netuku made no aida, kanozyo wa karada o tizimete, mata atama no naka de keisan o site miru no dat- TA. 'CD She COULDN'T RECALL happiness as anything more than such a trivial

thing. ® She POURED hotwater into a foot warmer. @ And she SLIPPED into a cold bed. @ Scrunching herself up, she THOUGHT about it again until she fell asleep.' (Tatsuzo Ishikawa, Seishun no satetsu)

4. Discussion

Section 3 summarized six analyses of -TA and -RU: (i) tense, (ii) time-relationship, (iii) aspect, (iv) perspective, (v) dual function, and (vi) grounding. These six analyses can be organized into three groups according to the concept presumed to be essential: (A) temporal sequentiality (i-ii), (B) the speaker's attitude toward the encoded situations (iii­iv), (C) a combination of both (v-vi). The grounding analysis is categorized as (C) because it assumes that -TA is temporally grounded with respect to some other time, but -RU is free from such temporal sequentiality like aspect. This section provides further assessment of such analyses and their relative validity.

I do not have much to say about the type C analyses, for they attempt to solve problems by positing dual functions . If a type A or type B analysis can provide satisfactory characterization of -TA and -RU, positing both will become superfluous.

In type A, the tense analysis presents three problems: (i) it has very limited applicability; (ii) some past events cannot be marked the alleged past-tense marker -TA; (iii) it cannot explain why certain connectives require a particular ' tense ' regardless of event time. The time-relationship analysis has wider validity, and it gives an explanation

236 Yoko Hasegawa

for (ii), by shifting reference time. However, (iii) remains problematic for the time­relationship analysis because it cannot explain why such clauses as TO require the -RU

form to express anteriority. In type B, the aspect analysis needs to posit two types of aspect because the alleged

perfective -TA can co-occur with the typical imperfective marker -TE 1- . In Miller's proposal, what -TA and -RU encode is not the traditional notions of perfective and imperfective, but, rather, more general judgments of how a given situation occurs and how the speaker presents it, a step closer to the notion of modus. As mentioned earlier, the most serious challenge to the aspect analysis is how to account for the use of -TA in sentences like (36), restated here as (49).

(49) kinoo tomatta ryokan ni wa niwa ni ookina matu no ki ga at-TA. 'The inn I stayed at yesterday HAD a big pine tree in the garden.'

This use of -TA can be explained by the perspective analysis. As discussed in section 3.5, the function of -TA in this case is to express an indirect report of the speaker's experience. If -RU were used instead, the sentence would be a direct expression of the speaker's perception.

Although integration of the aspect and perspective analyses is beyond the scope of the present paper, analyses based on the speaker's involvement, vis-a-vis the sequentiality of situations, can more accurately account for the data. Nevertheless, many (myself included) are inclined to consider that the prototypical uses of -TA and -RU are to mark tense. Soga (1990: 1 04), who nonetheless insists that -TA and -RU are primarily tense markers, states:

If the past tense form is used, it indicates that the speaker perceives the situation to be remote, but if the non-past tense is used, it means that the past situation is perceived as if it were directly before the speaker. Put another way, the use of the past tense suggests a detached and objective attitude on the part of the speaker toward the situation, but the use of the non-past tense suggests the speaker's subjective and psychological involvement with the situation . ... the difference in use between past and non-past tense markers reflects a difference in the speaker's or writer's perception of and psychological attitude toward the situation being described, and that is a modality difference.

This statement resembles the perspective analysis. The crucial difference, however, is its direction of semantic extension. Soga attempts to derive other usage from deictic tense, whereas the perspective analysis claims that tense interpretation is derived from the subjectivity and objectivity opposition.

The time-relationship analysis (A type) also resembles the aspect analysis (B type). In both analyses, the -TA situation is understood to be completed before some other situation. The crucial difference between the two is, again, that whereas the former maintains that -TA and -RU primarily indicate some notions of temporal alignment, the latter affirms that such notions are secondary, for the term aspect is generally defmed as the assessment or characterization of the denoted situation 'as it progresses or as it is

Tense-asp .

distributed in time, but irrespectiv action, mentioned or implied' (Ma

This discrepancy in the directi priori mental permanence of abst mentioned earlier, in many langua: than tense (and modus is more p1 and objective; aspect is definite, iJ (1989) also comments:

Time .. . does not present a happen, but rather time is sequentiality of events rather time is primary only to the e: in scientific events ... This p consequences for the linguist' that 'tense' has been singled c modality.

The usage of -TA and -RU can be d take tense as necessity.

5. Conclusion

The present paper discussed the 1

Japanese. Having examined six speaker's judgments about given rather than the sequentiality of the

Admittedly, it is extremely d notion of time and to interpret : subjective dimensions. This is UJ

for most people in modern times. and they inevitably reflect cone However, such essential componc the case of Japanese, the aspectl although new usage has emergec firmly encoded in the future.

NOTES

1 //Ru// stands for the morpheme

2 Hopper (1979:213) divides stat signaling the major thread of the He calls the former, foregrounc

ins problematic for the time­clauses as TO require the -RU

; of aspect because the alleged ·e marker -TE 1- . In Miller's 1al notions of perfective and r a given situation occurs and modus. As mentioned earlier, ) account for the use of -TA in

tu no ki ga at-TA. :e in the garden.'

is. As discussed in section 3.5, ort of the speaker's experience. ct expression of the speaker's

nalyses is beyond the scope of nent, vis-a-vis the sequentiality

Nevertheless, many (myself ;es of -TA and -RU are to mark A. and -RU are primarily tense

eaker perceives the situation to eans that the past situation is :>ut another way, the use of the de on the part of the speaker

tense suggests the speaker's tuation .... the difference in use difference in the speaker's or

: toward the situation being

11cial difference, however, is its other usage from deictic tense,

:rpretation is derived from the

.es the aspect analysis (B type). : completed before some other again, that whereas the former .ons of temporal alignment, the n aspect is generally defined as ion 'as it progresses or as it is

Tense-aspect controversy revisited: The -ta and -ro forms in Japanese 23 7

distributed in time, but irrespective of the moment of speech or ... of the time of another action, mentioned or implied' (Maslov 1988:63).

This discrepancy in the direction of semantic extension reflects the conjecture of an a priori mental permanence of abstract linear time. Izui, however, argues otherwise. As mentioned earlier, in many languages, including Indo-European, aspect is more prominent than tense (and modus is more prominent than aspect). "Tense is abstract, intellectual, and objective; aspect is definite, impressionistic, and subjective" (Izui 1967:85). Hopper (1989) also comments:

Time .. . does not present a preexistent fj;amework, a flux against which things happen, but rather time is secondary to sequencing. Time is defined by the sequentiality of events rather than the other way around ... It follows that sequential time is primary only to the extent that eventive discourse is privileged, for example in scientific events ... This privileging of scientific sequentiality has had important consequences for the linguist' s approach to tense, aspect, and modality. It has meant that 'tense' has been singled out as a primary category, and separated from aspect and modality.

The usage of -TA and -RU can be described, and explained to a certain extent, if we do not take tense as necessity.

5. Conclusion

The present paper discussed the tense-aspect controversy associated with -TA and -RU in Japanese. Having exanlined six analyses, it is concluded that -TA and -RU encode the speaker's judgments about given situations and his/her intention of their presentation, rather than the sequentiality of those situations.

Admittedly, it is extremely difficult, if not impossible, for me to abandon the linear notion of time and to interpret sentences solely in terms of aspect and objective and subjective dimensions. This is understandable because the concept of linear time is real for most people in modern times. Languages have developed over a long course of time, and they inevitably reflect conceptualizations different from those of modern users. However, such essential components as verbal morphology resist abrupt alternations. In the case of Japanese, the aspectual meaning in its broadest sense is still predominant, although new usage has emerged, and the notion of tense will certainly become more firmly encoded in the future.

NOTES

1 //Ru// stands for the morpheme of which -RU is an allomorph.

2 Hopper (1979:213) divides statements in narrative discourse into two categories: those signaling the major thread of the narrative, and those providing supportive information. He calls the former, foreground, and the latter, background. Foreground statements

238 Yoko Hasegawa

express the events in chronological order, making the completion of one event a necessary condition for the next; verbs that appear in foreground tend to be punctual or perfective. Background statements need not be sequential with respect to foreground events, tend to amplify or comment on the main narrative events, and are concurrent with the main events; they tend to be durative, stative, iterative, or imperfective. 3 Ota calls these two tense types the primary and secondary tenses. Because his primary tense is identical with the conventional notion of tense, I call the former, ( deictic) tense. Following Miura (1974), I use the term time-relationship to refer to the latter. 4

Because the objective of Miller's article is to scrutinize Miyoshi (1974), who makes the claim based on his analysis of modern Japanese fiction, Miller also restricts his claim to only the written language of modern Japanese fiction. 5

Smith (1986: 101) defines perfective as indicating "a situation as a whole (that is, complete with initial and final endpoints)." 6

A totally opposite view has been expressed by Hopper (1982:4): "Aspectual structures are highly abstract. Whereas even tense morphemes have a concrete relationship to the observer - the observer's own time-line - aspect depends on an absolute, observer­independent shaping of a state or action." It is, however, difficult to envisage aspect as more objective than tense, because, as Michaelis (1993: 17) argues, aspectual meaning involves such notions as evidentiality, resultativity, modality, individuation, perspective, transitivity, and distribution of focus over sentential constituents. 7

The present perfect differs from perfective and imperfective (the aspects proper), because it tells us "nothing directly about the situation in itself, but rather relates some state to a preceding situation" (Comrie 1976:52). If a language has both present perfect and past-tense forms, the semantics of the two invariably overlap. The difference is traditionally stated in such a way that, while both terms refer to past situations, the present perfect also signals their 'present (current) relevance,' although this characterization is manifestly inadequate. In normal discourse, the very act of mentioning a situation guarantees its present relevance, and the use of the present perfect does not transform an irrelevant piece of information into a relevant one. 8

This line of analysis has recently become popular among literary critics (Karatani 1993, Noguchi 1994). It is said that the modern usage of -TA was developed by Japanese intellectuals who had become acquainted with European languages, which are equipped with a tense system. They contends that the newly invented use of -TA enabled authors to efface the narrator, and that it gave the text a single third-person point of view. 9 Shinzato (1993:51-52) proposes that -TE I-RU conveys the notion of temporal distance , which can be viewed as reflecting the degree of immediacy of information processing. In keeping with traditional cognitive psychology, she posits two kinds of cognitive processes - perception and cognition. Perception is a cognitive process in which 'perceived stimuli are translated and encoded in a mandatory and instantaneous manner. Cognition ... involves further processing of the information using all resources available to the speaker.' The evaluative and objective views reflect the speaker's deliberation on the acquired information, and this implies non-instantaneous information processing.

Tense-asp

REFERENCES

Bybee, Joan (1985) Morphology Amsterdam: John Benjarnins

Bybee, Joan, Perkins, Revere, I Chicago: University of Chica

Comrie, Bernard (1976) Aspect. C

Fleischman, Suzanne (1990) Tens

Givan, Talmy (1982) Tense-asp( Hopper (ed.), Tense-aspect: Benjamins, 115-63.

Hopper, Paul (1979) Aspect and f semantics, vol. 12, discourse

Hopper, Paul (1982) Aspect betv the volume. In P. Hopper 1

Amsterdam: John Benjarnins

Hopper, Paul (1989) Time of th( linguistics, Humanities Instit

Inoue, Kyoko (1978) Speaker' s Michigan Papers in Linguist

Isshiki, Masako (1965) Niti-eigo 8.

Izui, Hisanosuke (1967) Gengo n

Karatani, Kojin (1993) Origins q_ deBary. Durham, N.C.: Duk

Kotsinas, Ulla-Britt (1989) Com second Scandinavian symbc International, 33-48.

Kunihiro, Tetsuya (1967) Koozoc

Kunihiro, Tetsuya (1980) Hensh vol2, bunpoo. Tokyo: Taishu

Kuryiowicz, Jerzy (1964) The il Winter.

Maslov, Jurij (1988) Resultative, resultative constructions. Jo

Michaelis, Laura (1993) Towarc California, Berkeley.

Mikarni, Akira (1953) Gendai gc

Miller, Roy A. (1975) Do the Association of Teachers of J

completion of one event a ;round tend to be punctual or 1 with respect to foreground rents, and are concurrent with r imperfective.

' tenses. Because his primary ill the former, ( deictic) tense. refer to the latter.

iyoshi (1974), who makes the ller also restricts his claim to

ituation as a whole (that is,

982:4): "Aspectual structures a concrete relationship to the tds on an absolute, observer­jifficult to envisage aspect as 7) argues, aspectual meaning ty, individuation, perspective, 1ents.

fective (the aspects proper), itself, but rather relates some ~age has both present perfect y overlap. The difference is ; refer to past situations, the

relevance,' although this rse, the very act of mentioning f the present perfect does not ~:me.

literary critics (Karatani 1993, , was developed by Japanese mguages, which are equipped l use of -TA enabled authors to rson point of view.

~ notion of temporal distance, of information processing. In

ro kinds of cognitive processes process in which 'perceived and instantaneous manner.

n using all resources available t the speaker's deliberation on IS information processing.

Tense-aspect controversy revisited: The-ta and -rufonns in Japanese 239

REFERENCES

Bybee, Joan (1985) Morphology: A study of the relation between meaning and form. Amsterdam: John Benjarnins.

Bybee, Joan, Perkins, Revere, Pagliuca, William (1994) The evolution of grammar. Chicago: University of Chicago Press.

Comrie, Bernard (1976) Aspect. Cambridge: Cambridge University Press.

Fleischman, Suzanne (1990) Tense and narrativity. Austin: University of Texas Press.

Giv6n, Talmy (1982) Tense-aspect-modality: The creole proto-type and beyond. In P. Hopper (ed.), Tense-aspect: Between semantics & pragmatics. Amsterdam: John Benjamins, 115-63.

Hopper, Paul (1979) Aspect and foregrounding in discourse. InT. Givon (ed.), Syntax and semantics, val. 12, discourse and syntax. New York: Academic Press, 213-41.

Hopper, Paul (1982) Aspect between discourse and grammar: An introductory essay for the volume. In P. Hopper (ed.), Tense-aspect: Between semantics & pragmatics. Amsterdam: John Benjarnins, 3-18.

Hopper, Paul (1989) Time of the sign. Lectures on discourse and temporality in recent linguistics, Humanities Institute, Santa Barbara, October 16-21, 1989.

Inoue, Kyoko (1978) Speaker's perspectives and temporal expressions. University of Michigan Papers in Linguistics 2:105-15.

Isshiki, Masako (1965) Niti-eigo tensu no hikaku nooto kara. The Rising Generation 3:6-8.

Izui, Hisanosuke (1967) Gengo no koozoo. Tokyo: Kinokuniya Shoten.

Karatani, Kojin (1993) Origins of modern Japanese literature. Translation edited by Brett deBary. Durham, N.C.: Duke University Press.

Kotsinas, Ulla-Britt (1989) Come, stay, finish. In L.-G. Larsson (ed.), Proceedings of the second Scandinavian symbosium on aspectology. Stockholm: Almqvist & Wiksell International, 33-48.

Kunihiro, Tetsuya (1967) Koozooteld imiron. Tokyo: Sanseido.

Kunihiro, Tetsuya (1980) Hensha hosetu. In T. Kunihiro (ed.), Niti-eigo hikaku kooza, vol2, bunpoo. Tokyo: Taishukan, 267-76.

Kurylowicz, Jerzy (1964) The inflectional categories of Indo-European. Heidelberg: C. Winter.

Maslov, Jurij (1988) Resultative, perfect, and aspect. In V. P. Nedjalkov (ed.), Typology of resultative constructions. John Benjamins, 63-85.

Michaelis, Laura (1993) Toward a grammar of aspect. Ph.D. dissertation, University of California, Berkeley.

Mikarni, Akira (1953) Gendai gohoo zyosetu. Tokyo: Toko Shoin.

Miller, Roy A. (1975) Do the Japanese know how to tell time? The Journal of the Association of Teachers of Japanese 10:1-18.

240 Yoko Hasegawa

Miura, Akira (1974) The V-u form vs. the V-ta form. Papers in Japanese Linguistics 3:95-121.

Miyoshi, Masao (1974) Accomplices of silence. Berkeley: University of California Press.

Nakau, Minoru (1980) Tensu asupekuto no hikaku. InT. Kunihiro (ed.), Niti-eigo hikaku kooza, vol2, bunpoo. Tokyo: Taishukan, 101-55.

Noguchi, Takehiko (1994) Sanninsyoo no hakken made. Tokyo: Chikurna Shobo.

Onoe, Keisuke (1995) Grounding keishiki to shite no SIDTA, SIDTEIRU. Handout distributed at the CLC Gengogaku Syuutyuu Koogi, Kansai, July 2, 1995. Tokyo: CLC Nihongo Gakuin.

Ota, Akira (1972) Tense correlations in English and Japanese. In J.E. Alatis (ed.) Studies in honor of Albert H Marckwardt. Washington, D.C.: Teachers of English to Speakers of Other Languages, 121-34.

Schiffrin, Deborah (1981) Tense variation in narrative. Language 57:45-62.

Shinzato, Rurniko (1993) A unified analysis of Japanese aspect marker te iru. Language Quarterly 31:41-57.

Smith, Carlota (1986) A speaker-based approach to aspect. Linguistics and Philosophy 9:97-115.

Soga, Matsuo (1983) Tense and aspect in modern colloquial Japanese. Vancouver: University of British Columbia Press.

Soga, Matuo (1990) The modal nature of tense markers in Japanese. In 0. Kamada and W. Jacobsen (eds.), On Japanese and how to teach it. Tokyo: The Japan Times, 100-10.

Suzuki, Shigeyuki (1965) Gendai nihongo no doosi no tensu. Tokyo: National Language Research Institute.

Suzuki, Shigeyuki (1976) Nihongo no doosi no toki ni tuite. Gekkan Gengo 5(12):50-58.

Takakura, Hiroko (1990) Tense variation in Japanese narratives: Why do some sentences resist the RU/TA alternation? Papers from the Middlebury conference on Japanese linguistics and Japanese language teaching. Middlebury, VT: Middlebury College, 166-84.

Teramura, Hideo (1971) "Ta" no imi to kinoo. In Gengogaku to nihongo mandai. Tokyo: Kuroshio Shuppan.

Turney, Alan (1965) The three-cornered world. London: Peter Owen.

Wolfson, Nessa (1979) The conversational historical present alternation. Language 55:168-82.

TEACHERS' IN T(

0 Introduction An important difference interaction consists in thE subjectively shared meani1 common knowledge is achi understanding of one ano Every next turn displays tl: and, unless it gets repairec shared meaning. Conversat in various ways to construe

As Streeck (1979) and ) (intersubjektiv geteilter K1 classrooms emerges in quit' the teacher becomes part limited possibilities on the of these meanings. Meanin not be considered as shared teacher. In other words: discourse with a higher e pupils.

Crucial to the process , classroom interaction is tl sequence observed for c Coulthard 1975): initiah time, the initiation is a que sequences (Mercer 1995) -pupils' knowledge, rather 1

question / answer cycle is lecturing) precisely becam turn. It offers a teacher tl

PRAGMATICS in 1998

Selected Papers from the 6th International Pragmatics Conference

Vol. 2.

edited by

JefVerschueren

Fund for Scientific Research - Flanders & University of Antwerp

~ ~

International Pragmatics Association (IPrA), v.z.w. Antwerp, Belgium

1999