Completive aspect, emotion, and the dynamic eventive: the case of Korean V-a/e pelita, Japanese V-te shimau, and Spanish se (1).

Strauss, Susan

Abstract

This paper examines the auxiliary verb constructions V-a/e pelita in Korean and V-te shimau in Japanese as markers of both completive aspect and speaker stance. These constructions are, for the most part, grammatically optional in their respective languages. When they occur, they generally mark finality or completion of some event or action in addition to some aspect of the speaker's judgment. The two constructions, while not at all related morphologically or semantically to the reflexive in either language, pattern in a manner strikingly similar to some occurrences of the middle marker se in Spanish. That is, in all three languages, the constructions are generally optional, they tend to cooccur with certain types of verbs, they tend to increase or decrease the sense of volitionality/control expressed by the main verbs with which they cooccur, and they all express aspects of speaker stance such as emphatic perspective, accidentality/uncontrollability, regret, relief pride, counterexpectations, and strong resolve. In view of these parallels with certain functions of Spanish se, this paper will posit that Korean V-a/e pelita, Japanese V-te shimau, and the "energetic middle" are instances of the dynamic eventive.

1. Introduction

Both Korean and Japanese possess a rich system of auxiliary constructions (2) that serve to express event-focused relationships such as temporal aspect (e.g. progressive, resultative, perfective) and participant-focused relationships, such as causatives and benefactives. These constructions consist of a main verb marked by a nonfinite ending (-ko or -a/e in Korean and -te in Japanese), and an additional verbal element that comes from a closed set of approximately fifteen auxiliary verbs in each language. In the great majority of cases, these auxiliary verbs also function as main verbs in each language. (3)

The focus of this paper is on three grammatical constructions: two auxiliary constructions, V-a/e pelita for Korean and V-te shimau for Japanese, and the middle marker se in Spanish. The two auxiliary constructions are parallel in the following ways: (1) they derive from lexical sources that project somewhat similar cognitive motivations for their grammaticization into auxiliary constructions with similar functions, (2) they mark temporal aspect (typically inchoative, perfective, or completive), (3) they tend to affect the sense of volition or control expressed by the main verb with which they cooccur, (4) they are grammatically optional, and (5) utterances without the auxiliary tend to be more neutral while utterances in which the auxiliary appears tend to mark some type of emotion or evaluative stance on the part of the speaker.

In the same vein, certain instances of Spanish se will also be shown to mark temporal aspect, to influence the sense of control or volitionality expressed by the main verb with which it occurs, to be grammatically optional, and to indicate some type of speaker evaluative stance in the cases where its optional occurrence appears. I will be presenting a number of examples that demonstrate clear parallels in use and pragmatic function between the two auxiliaries under investigation and certain uses of the middle marker se in Spanish. All three constructions will be analyzed from the point of view of the "dynamic eventive"--a term developed to capture the pragmatic similarities of these forms across the three languages.

2. Korean pelita and Japanese shimau as lexical verbs

In this section we will briefly examine the sources of the Korean and Japanese auxiliaries. In both languages, these auxiliaries derive from lexical verbs that still exist and are naturally used.

2.1. Korean pelita

According to Lee (1993) the meaning of pelita as a lexical verb can be captured as 'to throw away' or 'to spoil'. Examples (1) and (2) below, adapted from Lee (1993: 238-239), illustrate these two meanings:

(1) Changswu-ka ssuleyki-lul pelie-ss-ta.
 Changswu TM garbage-ACC throw away-PLN-PST
 'Changswu threw away the garbage.'
(2) Changswu-nun sin-ul peli-ess-ta.
 Changswu TM shoe-ACC throw away-PLN-PST
 'Changswu ruined the shoes.'

More generally, then, rather than expressing two seemingly distinct senses of 'throwing something away' or 'spoiling something,' what pelita means, according to Lee, is that the subject of the sentence 'throws something out of the domain of his possession or influence' (1993: 238), whether intentionally or unintentionally. He represents this relationship schematically according to the diagram in Figure 1.

In this schema, x represents the object of pelita, or the thing that is thrown away, and Y (i.e. the larger circle) represents the source location where x once was. Thus for (1), x would represent ssuleyki 'the garbage' and Y, the source location where the garbage was before it was thrown out. In (2), the relationship between x and Y is less clear. In fact, according to Lee, the interpretation of pelita is ambiguous: it could mean either that Changswu ruined his shoes or that he threw them away. In any case, the interpretation of the verb 'ruin' or 'spoil' seems to also signal the notion of 'rendering something completely unusable or useless'. Invented examples (3) and (4) below illustrate this sense of the verb:

(3) silswulo sokum-taysin selthang-ul neh-e-se kwuk-ul
 accidentally salt-instead sugar ACC put-RSN soup-ACC
 peli-ess-ta.
 spoil-PST-PLN
 'I accidentally put sugar into the soup instead of salt and ruined
 it.'
(4) nemwu olay kyeysok koham-ul cilly-e-se
 too much long time continuously shout-ACC shout-RSN
 moksoli-lul peli-ess-ta.
 voice-ACC spoil-PST-PLN
 'I kept yelling for a long time and ruined my voice.'

In both cases, pelita implies an irreversibility of the action, such that in (3), the soup was completely beyond usability; it was ruined to the point of actually having to be discarded. Fortunately, the human body is a bit more forgiving, and the notion of irreversibility in (4) is not quite as extreme as in (3). What is underscored, however, is that the speaker needed her voice for some special purpose, for example to sing, to give a lecture, to tell a story, but it was rendered unusable by virtue of her earlier behavior.

By extension, pelita can also be used to express less concrete and less physical relationships, as in (5) and (6), also adapted from Lee (1993: 239):

(5) Changswu-nun casik-ul peli-ess-ta.
 Changswu-TM children-ACC throw away-PST-PLN
 'Changswu abandoned his children.'
(6) Changswu-nun huimang-ul peli-ess-ta.
 Changswu-TM hope-ACC throw away-PST-PLN
 'Changswu gave up hope.'

In both examples, x, the object of pelita is not actually 'thrown away' or 'ruined', but rather cast out beyond the sphere of the subject, Y. These two sentences also evoke a sense of irreversibility, such that in (5), chances may be that Changswu's relationship with his children is irreparable; in (6), what is implied is that it would be difficult, if not impossible, for Changswu to regain his hope.

2.2. Japanese shimau

Alfonso (1966: 463) lists the meaning of shimau as 'to finish' or 'to put an end to'; Martin (1975: 533) indicates that shimau means 'puts (stores) away' or 'shuts up (completely)'. The Koojien (1991) dictionary provides two examples that essentially illustrate these same meanings, as noted in (7) and (8) below (both from Koojien 1991: 1176):

(7) shigoto o shimau [meaning: 'finish']
 work ACC finish-PLN
 'to finish (or stop) working'
(8) omoide o mune ni shimau [meaning: 'store away']
 memory ACC heart LOC store away-PLN
 'to store a fond memory in one's heart'

Interestingly, however, most native speakers of Japanese with whom I have consulted do not accept (7) as a "natural"-sounding sentence, and thus do not accept the general meaning of 'finish' proposed by such linguists as Alfonso and as indicated in the Koojien (1991) as the notion of 'finishing or completing one's work'. In order to capture this sense of shirnau, my informants suggest instead a scenario like that in (9):

(9) watashi wa taitei shichiji ni mise o shimaimasu.
 I TM usually 7 o'clock at shop ACC close up-NPST-POL
 'I usually close my shop at seven o'clock.'

Here, the sentence still captures the sense of someone 'finishing' work, but refers to the actual closing up of the shop (for the day), rather than the termination or completion of a particular project or task.

Moreover, while the Koojien sentence in (8) creates a somewhat poetic and even abstract image (storing a memory in one's heart), shimau can also be used in a similar sense with perfectly mundane objects and can evoke concrete images as in (10) and (11) below:

(10) Kenji wa saifu o kaban ni shimatta.
 Kenji TM wallet ACC briefcase LOC put away-PST-PLN
 'Kenji put his wallet into his briefcase.'
(11) Kaori wa sentakumono o tatande tansu ni
 Kaori TM laundry ACC fold-TE dresser LOC
 shimatta.
 put away-PST-PLN
 'Kaori folded the laundry and put it away in the dresser.'

Semantically, then, what these example sentences have in common is the fact that they all express the meaning of 'closing something up' or 'putting something away'. As a result of this action, the object in question is no longer visible and no longer accessible, having been closed up or stored away in a particular place--both out of sight and out of mind.

Thus, in examining the meanings of both pelita and shimau as lexical verbs, we note that they share the common semantic feature of rendering something invisible and/or beyond the domain of control or influence of the sentential subject. The action of these two verbs as lexical verbs removes the object from sight and/or from accessibility.

3. Korean V-a/e pelita and Japanese V-te shiimau: auxiliary constructions

3.1. Korean V-a/e pelita

Korean pelita as an auxiliary occurs following the nonfinite -a/e form of the main verb. H. Sohn (1994: 333) refers to the V-a/e pelita construction as a marker of "terminative aspect," expressing the meaning of 'finish up'. Sohn adds that this construction performs a function similar to the so-called "terminative adverbials" such as motwu 'all', machimnay 'at last', and tutie 'finally'.

Example (12), from Sohn, will illustrate:

(12) (from Sohn 1994: 333, example [190a]; emphasis added)
 Yongho-nun apeci swul-ul masi-e peli-ess-ta
 -TC father wine-ACC drink-A/E PELITA-PST-DC
 'Yongho drank up his father's wine.'

Note that the counterpart without the auxiliary as in (12') simply indicates that Yongho drank his father's wine. Whether he drank it 'up' (i.e. completely finishing it or not) is not relevant:

(12') (variation of Sohn 1994: 333, example [190a])
 Yongho-nun apeci swul-ul masi-ess-ta
 -TC father wine-ACC drink-PST-DC
 'Yongho drank his father's wine.'

The alternation in meaning between (12) and (12') demonstrates precisely why Sohn mentions a similarity in meaning between the auxiliary itself and the adverbials meaning 'all', 'completely', or 'finally'. That is, in cases where this construction occurs, it tends to underscore the total completion of some action or emphasize the fact that the speaker has waited for or anticipated the completion of an action that has finally occurred.

In addition to terminative aspect, the use of Korean V-a/e pelita is also associated with the speaker's evaluative stance. Lee (1993: 239-243), points out that such evaluative meaning conveyed by this auxiliary stems from an overall interpretation of the construction by virtue of the semantics of pelita as a main verb. In other words, it indicates the notion of 'spoiling the speaker's expectation', hence expressing such feelings as disappointment or regret, and 'removing psychological hindrance', thus expressing relief. Examples of each are shown in (13) and (14) respectively:

(13) (adapted from Lee 1993: 240, example [4])
 [Clause 1]
 onul ku-eykey il-ul com sikhilye-ko hay-ss-nuntey.
 today him-DAT thing-OM a little ask-CONN do-PST-CONN
 'I was going to ask him to do something for me today, but ...'
 [Clause 2a]
 ku-nun ecey ka-ss-ta.
 he-TM yesterday go-PST-DC
 'He went yesterday' (to an implied destination).
 [Clause 2b]
 ku-nun ecey ka-pelie-sst-a.
 he-TM yesterday go-A/E-PELITA PST DC
 '(To my regret) he went yesterday' (i.e. he left).

In the above example, what is notable first is that the sentence in clause 2b marked with the V-a/e pelita auxiliary conveys a stance of regret and/or disappointment (negative affect) as pointed out by Lee, while the utterance in clause 2a simply states that the person in question had already left (and is now gone). Both sentences are well formed and grammatical; they differ only in terms of the auxiliary construction as a means of expressing terminative aspect and speaker stance.

Example (14), also from Lee, further illustrates the use of this construction to express positive or negative affect.

(14) (adapted from Lee 1993: 243, example [10a]; emphasis added)
 ku-tul-I nay kwaca-lul meke-peli-e-ss-ta
 he-PL-SM my cookie(s)-OM eat-A/E-PELITA-PST DC
 '(To my relief) they ate (up) my cookies.'
 '(To my regret) they ate (up) my cookies.'

Note that the speaker's attitude with respect to regret or relief is ambiguous; the utterance could also very well express the speaker's stance of surprise. Whether it is relief, regret, or surprise that is being expressed would depend upon the context of the utterance. A neutral and objective expression of the same event would generally appear without V-a/e pelita, as in (14').

(14') ku-tul-I nay kwaca-lul meke-ss-ta
 he-PL-SM my cookie(s)-OM eat-PST DC
 'They ate my cookies' [neutral].

3.2. Japanese V-te shimau

Syntactically and semantically, Japanese V-te shimau behaves much in the same way as Korean V-a/e pelita. (4) That is, like Korean V-a/e pelita, V-te shimau is formed by combining the nonfinite form of the main verb, that is, the -te form, with the auxiliary verb shimau. Its use has been accounted for in reference grammars and linguistic treatments as a marker of "emotive terminative aspect" (Soga 1983). It has also been discussed as a grammatical means of signaling irreversibility, the lack of control over a particular situation, the automatic or spontaneous occurrence of an event or situation, the speaker's reaction to some unexpected happen-stance, and the speaker's strong resolve to accomplish some action (e.g. Alfonso 1966; Makino and Tsutsui 1987; Martin 1975; Soga 1983; Ono 1992; Ono and Suzuki 1992; Strauss 1994, 1996, 2002; Teramura 1982), as well as a means of conveying the notions of completion and totality (Martin 1975, Soga 1983, Teramura 1982, Strauss 2002).

Soga (1983) presents an extensive discussion of the construction, providing a number of invented sample sentences that capture the range of functions and meanings noted above. As for "emotive terminative aspect," Soga points out that the construction may imply either regret or pride, depending upon whether or not the outcome expressed is an undesirable one (in the case of regret) or a desirable one (in the case of pride). Examples (15) and (16) below, from Soga (1983: 167), illustrate how these two emotions may be associated with the V-te shimau construction.

(15) (from Soga 1983: 167, example [70b])
 Okane ga nakunatte simatta.
 money SM disappear
 'The money has disappeared (and I regret it).'
(16) (from Soga 1983: 167, example [70d])
 Sono mondai wa toite simana.
 that problem TM solve
 '(I) have solved that problem (and I am proud of it).'

According to Soga, the sense of regret or pride conveyed through the use of the V-te shimau construction derives from the fact that "Vote s[h]imau often implies that the terminated action is unrecoverable" (1983: 166). Neutral readings for each sentence appear below as (17') and (18') which do not include the V-te shimau auxiliary.

(17') (variations of Soga 1983: 167, example [70b])
 Okane ga naku natta.
 money SM disappear
 'The money has disappeared' [neutral as to affective stance
 marking].
(18') (from Soga 1983: 167, example [70d])
 Sono mondai wa toita.
 that problem TM solve
 '(I) have solved that problem' [neutral as to affective stance
 marking].

Makino and Tsutsui (1986: 404) note that the V-te shimau auxiliary expresses the completion of an action and add, like Sohn (1994) for Korean, that this construction might frequently occur with such adverbs as kanzenni 'completely', zenbu 'all', and sukkari 'completely', citing the following example:

(19) maiku wa sukkari nihongo o wasurete
 Mike TM completely Japanese ACC forgt-TE
 shimatta.
 shimauAUX-PST-PLN
 'Mike has completely forgotten Japanese.'

In order to illustrate the completive function of this auxiliary, Makino and Tsutsui provide this example:

(20) koko ni oite oku to Jimu ga tabete
 here LOC put-TE oku-AUX if Jim SUBJ eat-TE
 shimau yo.
 shimauAUX-POL-NPST PRT
 'If you leave it here, Jim will eat it (up).'

The sentence expresses the idea that if the food item in question is left unattended, Jim will eat it all and nothing will be left of it--an outcome that would likely engender regret, disappointment, or even relief, depending upon the context.

Ono (1992) treats the V-te shimau auxiliary from the point of view of grammaticization, designating it as a marker of "frustrative," nonvolitionality, and evidentiality. Ono's analysis of V-te shimau is based largely on its opposition to another auxiliary V-te oku, which he designates as a marker of "preparative/purpose" and volitionality; both auxiliary constructions are analyzed as having grammaticized as markers of the perfect. Ono's analysis is predicated on the notion of a "minimal pair" contrast between the two target constructions, and he establishes this primarily on the basis of the contrasting aspectual differences between V-te shimau and V-te oku.

Ono and Suzuki (1992) centers on V-te shimau only and is also based on the concept of grammaticization. Here, the authors posit that the meaning of the lexical verb 'put away' or 'finish' engenders a progression of meaning in the V-te shimau construction such that it expresses pragmatic meanings of 'inability to undo' and 'automatic'. According to the authors, the pragmatic meaning of 'inability to undo' further leads to other pragmatic meanings, the first of which is 'speaker's negative attitude', followed by 'speaker's guiltily positive attitude'. (5)

Thus far, it has been shown from the majority of previous work cited above that when used as auxiliaries both V-te shimau and V-a/e pelita occur with the nonfinite form of a main verb and function primarily as aspectual markers, specifically as markers of completive aspect. Because of the sense of finality and totality expressed through completive aspect, both constructions also extend to imply such notions as the irreversibility of an event, the totality of some occurrence or change of state, a counter-expectation, the accidental or spontaneous occurrence of an event, an event depicted as beyond the speaker's control, in addition to a stance denoting determination to achieve some desirable outcome. In all cases, in addition to the notion of totality and/or completion, what seems to influence this range of emotive stances is the degree of volition or control that the speaker or sentential subject has over the outcome of the event being described.

4. Spanish se

Much in the same way as Korean V-a/e pelita and Japanese V-te shimau, some uses of the Spanish middle marker se also affect the degree of control by the speaker or subject and express accidentality, noncontrollability, strong resolve, emphatic involvement in an action, and counter-expectations, in addition to completion and totality. In this sense, Spanish se also serves to mark the speaker's emotive stance toward the occurrence of an action or an event.

Examples (21) and (21') will illustrate. These sentences represent translations into Spanish of the Korean sentence pair originally appearing as (14) and (14'), denoting the affectively marked and affectively neutral stance, respectively.

(21) Spanish translation of Korean (14)
 (ellos) se comieron mis galletas.
 '(To my regret, surprise, relief etc.) they ate (up) my cookies.'
(21') Spanish translation of Korean (14')
 (ellos) comieron mis galletas.
 they eat-PST my-PL cookie-PL
 'They ate my cookies' [neutral as to affective stance marking].

We find similar parallels between Japanese V-te shimau and Spanish middle marker se, as in examples (22) and (23), which are translations of the Japanese examples appearing as (19) and (20).

(22) Spanish translation of Japanese (19)
 Miguel se ha olvidado completamente del japones.
 'Mike has forgotten all of his Japanese.'
(23) Spanish translation of Japanese (20)
 Si lo dejas aca, Jaime se lo comera.
 'If you leave it here, Jim will eat it (up).'

And once again, examples (19) and (20) without the V-te shimau auxiliary and examples (22) and (23) without the middle marker se would be perfectly grammatical and would represent the neutral counterparts for each. Just as in the cases where the two target auxiliaries appear, what creates the sense of total completion of an action (i.e. eating something up or completely forgetting a language) is the reflexive marker se. In all four examples, as in their Korean and Japanese counterparts, what is underscored is the total occurrence of some event--and through the expression of this totality or finality comes the marking of the speaker's affective stance.

Many other semantic and pragmatic parallels exist across the three languages involving the three morphemes in question, which are evident even at the lexical level. The word triplets in (24) below are just a few examples.

(24) Semantic variations through the addition of V-a/e pelita, V-te
 shimau, and se

 a. 'to go' 'to go away'
 Korean: kata ka-pelita
 Japanese: iku it-te shimau
 Spanish: ir irse
 b. 'to sleep' 'to fall asleep'
 Korean: cata ca-pelita
 Japanese: neru ne-te shimau
 Spanish: dormir dormirse
 c. 'to take' 'to take away'
 Korean: kacyekata kacyeka-pelita
 Japanese: toru tot-te shimau
 Spanish: llevar llevarse

Note that in the (a) through (c) examples, the addition of the morpheme in question is responsible for the meaning change from 'go' to 'go away', from 'sleep' to 'fall asleep', and from 'take' to 'take away'. These represent simply word-level oppositions, and hence whether or not their use is grammatically optional would, of course, depend upon the intention of the speaker.

In the (a) triplets, which mark the distinction between 'go' and 'go away' in all three languages, the use of the verb 'go' (i.e. kata in Korean, iku in Japanese, and ir in Spanish) would require some kind of implied or inferable destination, if not an explicit one. Thus, to say simply that someone or something went away or has quit some location without knowing the goal or destination of the subject would constitute an ungrammatical utterance. This verb is perhaps the only case where absolute grammatical optionality might be called into question, since the added morphemes would actually be obligatory if no destination is stated, known, or implied. In the (b) and (c) triplets, what comes into play both semantically and pragmatically would be the degree of control the subject has over his/her own actions. In the (b) examples, what is underscored by the added morpheme is a kind of lack of control or volitionality over the activity of sleeping, with the verbs (ca-pelita [K], ne-te shimau [J], and dormirse [S]) indicating a much lower degree of control than their unmarked counterparts; in the (c) examples, the added morphemes create instead a sense of increased volitionality, such that the subject in question not only takes something, s/he 'takes it away' (completely).

We will return to these and other types of semantic/pragmatic nuances effected by the target morphemes in the next section.

4.1. Spanish se as a middle marker

Kemmer (1993) is an extensive discussion of middle voice for some 80 languages, from the standpoint of typology, semantics, and diachronic development. In the majority of cases, middle marking has been accounted for on the basis of the reflexive, either as a core source for middle development or as a related concept, such as the reciprocal. That is, for the majority of languages analyzed, Kemmer notes the reflexive as "the most frequent diachronic source" for middle marking (1993:151).

According to Kemmer (1993: 267-270), middle marking tends to occur in a variety of expressive domains, from the emphatic middles ('he did it himself'); indirect middles expressing the coming into possession of something ('eat', 'take', 'seize', 'receive'); verbs of translational motion ('go', 'fly', 'fall'); verbs of cognition ('realize', 'forget', 'be surprised', 'decide'); verbs expressing spontaneous events ('die', 'swell up', 'boil', 'dry', 'freeze', 'disappear', 'break', 'explode', 'fall apart', 'split'); verbs expressing naturally reciprocal events ('meet', 'greet', 'flock together'), and so forth.

What is interesting is that V-te shimau and V-a/e pelita pattern in much the same way as a good number of the middle examples noted by Kemmer, and neither auxiliary bears any morphological, semantic, or syntactic relationship to the reflexive or the reciprocal in either language. That is, in Japanese and Korean, the respective V-te shimau and V-a/e pelita auxiliary quite frequently and quite naturally cooccurs with such main verbs as 'take', 'eat', 'go', 'fall', 'realize', 'forget', 'die', 'disappear', 'break', 'become', 'explode', and so on. These verbs also tend to occur in Spanish with middle marker se.

Maldonado (1992) provides an extensive and impressive treatment of se as the Spanish middle voice marker. Since the middle in the majority of languages described by Kemmer (1993), including Spanish, derives from the reflexive, a central focus of the discussion in Maldonado (1992) centers on functional and grammatical distinctions between true reflexives and true middles. For the purposes of our discussion including Japanese and Korean, since this proposed middle derives from an altogether different source, we will only focus on Maldonado's treatment of true middles in Spanish.

According to Maldonado (1992), the middle construction in Spanish functions to decrease transitivity by emphasizing a lack of control by the agent, to increase transitivity by underscoring volitional control of the subject or agent of the verb, to designate a change of state; to express some special feature of temporal aspect (e.g. inchoative, punctual) of an event, and to express unexpected occurrences. Essentially, according to Maldonado, what the middle construction does in Spanish is to grammatically express the conceptualization of an event in a way that differs from an objective, neutral, and unmarked conceptualization of that same event. "A FUNDAMENTAL SEMANTIC IMPORT OF SE IS TO ELIMINATE THE CIRCUMSTANCES THAT ACCOMPANY THE NORMAL DEVELOPMENT OF AN EVENT IN SUCH A WAY THAT ONLY THE CHANGE OF STATE IS PROFILED" (1992: 318-319; emphasis in original). In other words, events described as processes or normal occurrences of something would tend to resist the middle. These categorizations are true for Spanish se as well as for Korean V-a/e pelita and Japanese V-te shimau.

Most linguists would hesitate labeling V-a/e pelita and V-te shimau as middle markers per se, especially since a typical middle in Spanish, for example, would read as something like (25):

(25) Alli se venden libros.
 'Books are sold there.'

and the target forms in Korean and Japanese do not follow the identical properties of the middle in every case. There are, however, a good number of similarities in function among Spanish se, Korean V-a/e pelita, and Japanese V-te shimau, and because of these overlapping functions I will henceforth refer to the three forms in question as instances of the "dynamic eventive."

4.2. Spanish se, Korean V-a/e pelita, Japanese V-te shimau as instances of the dynamic eventive

In the following sections, we will illustrate in detail the pragmatic parallels of the three target constructions as instances of the dynamic eventive from the point of view of their functions of decreasing transitivity, increasing transitivity, expressing special features of temporal aspect, and expressing events conceptualized as "unexpected" and "out of the ordinary."

4.2.1. The dynamic eventive and decreased transitivity. With respect to the function of decreasing transitivity, Maldonado (1992) provides a number of sample sentences for Spanish se where volitional control by the subject, agent, or speaker is reduced by virtue of the addition of the middle marker se, rendering an interpretation of an "accidental" or "noncontrollable" outcome of an event. The following example is from Maldonado, citing Garcia (1975), and illustrates how the transitivity of the verb 'forget' is decreased by the middle marker.

(26) (from Maldonado 1992: 62, example [44])
 Me olivide las llaves (from Garcia 1975)
 'I forgot the keys [accidentally].'

The verb 'forget' is a mental verb for which the subject generally does not have the freedom to exert varying degrees of control. That is, one could 'try to forget' something in the sense of not thinking about it consciously any longer, but the actual act of forgetting is something that one usually has little control over. In any case, what the addition of se does here is to reduce the already low degree of volitionality inherent in the verb olvidar such that the action is now characterized as an accident. We find an identical parallel in Korean and Japanese, as indicated in (27) and (28), respectively. Here, the V-a/e pelita and V-te shimau constructions also imply that the event is beyond the speaker's control and that the outcome was completely unintentional.

(27) yelsoi-lul ice-peli-ess-ta.
 key-OM forget-A/E PELITA-PST-PLN
 'I forgot the keys [accidentally].'
(28) kagi-o wasure-te shimatta
 key-OM forget-TE SHIMAU-PST-PLN
 'I forgot the keys [accidentally].'

4.2.2. The dynamic eventive and increased transitivity. In contrast with olvidar and the process through which its inherent transitivity is reduced through the middle marker se, Maldonado provides a number of other examples of mental verbs whose inherent semantics do allow for variations in degree of volitionality and control. In these cases, the addition of se serves to increase the transitivity of the verb by increasing the degree of control that the subject or agent of that verb has over the object. Example (29) below will illustrate:

(29) (adapted from examples [24a] and [24b] in Maldonado 1992: 18)

 a. Me pregunto cuando aprenderan algo de espanol estos
 gringos.
 'I wonder when these gringos will learn some Spanish.'

 b. Me pregunto cuando se aprederan la leccion los alumnos.
 'I wonder when the students will learn the lesson by heart.'

Maldonado explains the transitivity increase engendered by the middle marker se in the (b) sentence mainly on the basis of the fact that the "object of the complement clause is specific" (1992: 18). In other words, in the (a) sentence, the object algo de espanol 'some Spanish' is indefinite, nonfinite, and not subject to precise measure or evaluation. In contrast, the object in (b), la leccion 'the lesson' constitutes a finite, measurable entity; it is something that can be learned in its entirety. Additionally, Maldonado points out that the meaning distinctions between aprender on the one hand and aprenderse on the other rest on the fact that the latter involves the expression of "complete mental control over the object of cognition" (1992: 19). In the (b) example, by virtue of la leccion being a finite entity that one can learn in its entirety, the middle marker se fits well here to underscore that very act of learning the lesson by heart. Along these same lines, we note parallel expressions in Korean and Japanese.

In Korean, the verb oyta means 'to learn' with a specific sense of 'to memorize something', as opposed to the more general mental action of studying or being taught something. The sentence pair in (30) indicating the contrast between an utterance without V-a/e pelita and an utterance with V-a/e pelita will illustrate a shift in transitivity very similar to the one we observed in the Spanish examples above:

(30) Korean

 a. ecey sihom-i iss-e-se yenge-tane-lul
 yesterday exam-SM exist-RSN English-vocabulary-OM
 oywe-e-ss-e
 learn-PST-INFML
 'Because I had a test yesterday, I learned/memorized (some)
 English vocabulary words' [neutral, objective].

 b. ecey sihom-i iss-e-se yenge-tane-lul ta
 yesterday exam-SM exist-RSN English-vocabulary-OM all
 oyw-e-peli-ess-e
 learn-A/E PELITA-PST-INFL
 'Because I had a test yesterday, I learned/memorized all of the
 English vocabulary words' [thoroughness/intensity of the
 action; emphatic speaker stance].

In (30a), the object complement of the sentence is a nonfinite set of English vocabulary words that the speaker has memorized; in (30b), on the other hand, with the addition of the adverb ta 'all' and the V-a/e pelita auxiliary, we can clearly note a higher degree of transitivity in the verb oyta, in the sense that the object now involves the full and finite set of vocabulary words and the action of the verb has been intensified to express the action of 'thoroughly committing something to memory'. The (a) utterance can be considered as a more neutral and objective utterance; the (b) version clearly indicates an emphatic stance on the part of the speaker expressing some kind of feeling of pride, relief, complaint, and so forth.

The Japanese verb oboeru 'to learn/memorize' is similar to Korean oyta from the point of view of its inherent semantics and its ability to cooccur with V-te shimau. When it does occur with the auxiliary, what is underscored is the totality of the object being studied in addition to the intensity and thoroughness of the action itself. This is shown in the sentence pair in (31) below:

(31) (Japanese)

 a. kono shi o ashita made ni oboenai to sensei
 this poem OM tomorrow by learn-NEG COND teacher
 ni shikarareru.
 by scold-PSV-PLN
 'If I don't learn this poem by tomorrow, I'll be scolded by my
 teacher' [neutral, objective].

 b. kono shi o ashita made ni oboete
 this poem OM tomorrow by learn-TE
 shimawanai-to sensei ni shikarareru.
 SHIMAU NEG-COND teacher by scold-PSV-PLN
 'If I don't learn this poem (thoroughly and completely) by
 tomorrow, I'll be scolded by my teacher' [thoroughness/
 intensity of the action; emphatic speaker stance].

And, again, in this contrastive pair we find a similar distinction between the neutral expression in (a) and the emphatic stance in (b), achieved in large part because of the absence or presence of the V-te shimau auxiliary.

It must be noted that while the examples given in this section involve verbs of mental activity, the increase or decrease of transitivity effected by the dynamic eventive is not at all limited to this type of verb. For example, the transitivity of a verb like 'eat' clearly increases when a dynamic eventive marker is added, as we saw in some of the previous examples in all three languages.

Other instances of increased transitivity through the addition of Spanish se, Korean V-a/e pelita, and Japanese V-te shimau can be observed in the following set of sentence pairs:

(32) Spanish: 'I solved the problem'

 a. Resolvi el problema.
 'I solved the problem' [neutral, objective].

 b. Me resolvi el problema.
 'I solved the problem (after a struggle, finally)' [stance of
 pride, relief, confidence etc.].

(33) Korean: 'I solved the problem'

 a. na-nun ku mwuncey-lul haykyel-hay-ss-ta
 I-TM that problem-OM solve-do-PST-PLN
 'I solved the problem' [neutral, objective].

 b. na-nun ku mwuncey-lul haykyel-hay-peli-ess-ta
 I-TM that problem-OM solve-do-A/E PELITA PST-PLN
 'I solved the problem (after a struggle, finally)' [stance of
 pride, relief, confidence etc.].

(34) Japanese: 'I solved the problem'

 a. watashi wa sono mondai o toita.
 I-TM that problem-OM solved-PST-PLN
 'I solved the problem' [neutral, objective]

 b. watashi wa sono mondai o toite shimatta.
 I-TM that problem-OM solve-TE SHIMAU PST-PLN
 'I solved the problem (after a struggle, finally)' [stance of
 pride, relief, confidence etc.].

In (32), (33), and (34) we find a nearly parallel pragmatic phenomenon in Spanish on the one hand with the reflexive/middle marker se and in Korean and Japanese on the other, with the two auxiliary verbs in question. What is striking is that these constructions are completely unrelated grammatically yet their pragmatic function is nearly identical.

Thus, we have observed that both auxiliaries in Korean and Japanese function in much the same way as Spanish se in terms of their ability to both decrease and increase the degree of transitivity in the main verb, to express completive aspect, and to underscore control or lack of control on the part of the speaker or subject, as well as to express the speaker's emphatic or subjective stance.

4.2.3. Dynamic eventives and special features of temporal aspect. With respect to event conceptualization and middle voice marking, Maldonado (1992) and others have pointed out that middle marking often serves to capture certain features of temporal aspect, specifically punctual and inchoative aspects. Arce-Arenales et al. (1984: 7) indicate that "in general, middle diathesis marking will have associated aspectual functions, such that the middle diathesis marker is associated with punctual, change of state events." They provide the following sentence pair in Spanish in order to illustrate:

(35) (from Arce-Arenales et al. 1984: 6, original example [8])
 Durmio en el auto
 slept in the car
 'S/he slept in the car.'

(36) (from Arce-Arenales et al. 1984: 6, original example [9])
 se durmio en el auto
 slept in the car
 'S/he fell asleep in the car.'

Here, according to Arce-Arenales et al. (1984), with dormir 'to sleep' being an intransitive verb, the reading in (35) is aspectually durative, while it is punctual in (36). They claim that "[i]n both intransitives and hypertransitives [as in the case of comer 'to eat' vs. comerse 'to eat up'], se indicates increased perfectivity--a more completed and/or more puncutal action" (Arce-Arenales 1989: 286-365; Arce-Arenales et al. 1984: 7). In other words, what is being underscored by se, according to these authors, is the relationship between middle marking and completive aspect.

Note that the example pairs in (37) for Korean and (38) for Japanese, indicating the contrast with and without the V-a/e pelita and V-te shimau, express the identical semantic and aspectual distinction as the sentences in (35) and (36), which contrast the absence and presence of Spanish se.

(37) Korean translation of (35) and (36):

 a. Ku nun chatongcha an eyse cassta
 he TM car in LOC sleep-PST-PLN
 'He slept in the car.'

b. Ku nun chatongcha an eyse ca-peli-ess-taa
 he TM car in LOC sleep-A/E PELITA PST-PLN
 'He fell asleep in the car.'

(38) Japanese translation of (35) and (36):

 a. kare wa jidousha de neta.
 he TM car LOC sleep-PST-PLN
 'He slept in the car.'

 b. kare wa jidousha de nete-shimatta.
 he TM car LOC sleep-TE SHIMAU-PST-PLN
 'He fell asleep in the car.'

Thus, with the addition of the specific target markers for all three languages, we see that the action of falling asleep in the car is at once punctual and completive. Additionally, the action could easily be construed as being beyond the subject's control, unplanned, unintentional, and so forth.

The emphatic nature of the dynamic eventive and its effect on event conceptualizations and aspect marking can be further developed when we examine Maldonado's discussion of the "dynamic" values of se in Spanish. This becomes salient with verbs of motion, and especially so when those verbs also denote some type of directionality, such as 'to climb up', 'to fall down', 'to turn around', and so forth. Aspectually, what occurs with the combination of these types of verb and the middle marker se is that the action denoted by the verb plus middle construction becomes "compressed" and takes place within "a condensed amount of time" (Maldonado 1992: 336). Examples (39) and (40) will illustrate:

(39) (from Maldonado 1992: 335, original examples [25a] and [25b])

 a. Se fue de prisa, volando, corriendo, etc.
 'He left in a hurry, flying, running, etc.'

 b. ?? Se fue lentamente
 'He left slowly.'

(40) (adapted from Maldonado 1992: 336)

 a. Martina se subio a la mesa de un salto.
 'Martina got on the table in one jump' [lit. 'in one shot';
 compressed time].

 b. ??Martina se subio la escalera en diez minutos.
 'Martina went up the ladder in ten minutes' [extended
 duration > resists se].

In (39), the use of the middle marker with "dynamic" adverbials such as 'in a hurry', 'flying', 'running', is perfectly natural in the (a) sentence, whereas it sounds odd in the (b) sentence with an adverbial of extended duration and of low energy, lentamente 'slowly'. Similarly, in (40), the use of se in the (a) sentence is perfectly compatible with the adverbial 'in one jump', expressing both punctuality and compressed time. Its use in the (b) sentence is unnatural since the meaning indicates that the process of climbing the ladder was extended to take a full ten minutes.

Interestingly, the use of V-a/e pelita and V-te shimau patterns in a nearly identical fashion with the same sentence pair contrasts. Examples (41) and (42) illustrate this for Korean and (43) and (44) for Japanese.

(41) (Korean)

 a. ku nun ppalli ka-peli-ess-ta
 he-TM fast go-A/E PELITA-PST-PLN
 'He left quickly' [punctual, compressed time].

 b. ??ku nun chen chen-hi ka-peli-ess-ta
 he-TM slowly-ADV go-A/E PELITA-PST-PLN
 'He left slowly' [nonpunctual, extended time > resists V-a/e
 pelita].

(42) (Korean)

 a. Seongcheol-i cheksang ui-lul han-pen-ey
 Seongeheol SM desk on-OM one-time-in
 ttui-e-olla-ka-peli-ess-ta.
 climb-up-go-A/E PELITA-PST PLN
 'Seongcheol got on the table in one jump' [in one shot;
 compressed time].

 b. ??Seongcheol-i satari-lul sip-pwun-ey
 Seongcheol SM -ladder-OM ten minutes-in
 ttui-e-olla-ka-peli-ess-ta.
 climb-up-go-A/E PELITA-PST PLN
 'Seongcheol went up the ladder in ten minutes' [extended
 duration > resists V-a/e pelita].

(43) (Japanese)

 a. kare wa hayaku itte-shimatta
 he TM fast-ADV go-TE SHIMAU-PST-PLN
 'He left quickly' [punctual, compressed time].

 b. ??kare wa yukkuri itte-shimatta
 he TM slowly go-TE SHIMAU-PST-PLN
 'He left slowly' [nonpunctual, extended time > resists V-te
 shimau].

(44) (Japanese)

 a. Taro wa tsukue no ue ni ippen ni nobot-te
 Taro TM table GEN on LOC once in go up-TE
 shimatta
 SHIMAU-PST-PLN
 'Taro got on the table in one jump' [in one shot; compressed
 time].

 b. ??Taro wa hashigo o ichijikan de nobot-te
 Taro TM ladder-OM one hour in go up-TE
 shimatta
 SHIMAU-PST-PLN
 'Taro went up the ladder in one hour' [extended
 duration > resists V-te shimau].

The readings in the (b) sentences from (39) through (44) sound unnatural for all three languages. This is due, in large part, to the aspectual coloring that is associated with dynamic eventive marking, in the sense that it conveys punctual, inchoative, perfective, and completive aspect. Because the scenarios in the (b) sentences involve durative processes or at least processes that are clearly not punctual nor inchoative nor perfective nor completive, the added morphemes in question render a reading that is odd at best.

However, if we were to add background or context to each (b) sentence from the set in (39) through (44) it could indeed be conceivable that the dynamic eventive would work. That is, if we set the scene to portray the event as somehow unexpected or beyond the scope of a normal occurrence or that the outcome was long anticipated and finally it happened, then the readings in (b) would sound much more natural in each language. The fact that dynamic eventives tend to occur in situations where occurrences are portrayed as events that are out of the ordinary or unexpected will be examined in detail in the final section.

Essentially, then, dynamic eventives do in fact express special features of temporal aspect, which we have observed across the three particular morphemes or morpheme groups under investigation.

4.2.4. The dynamic eventive expressing events conceptualized as "unexpected," "out of the ordinary," etc. We have shown thus far that the dynamic eventive in Spanish, Korean, and Japanese is associated with issues of transitivity, speaker/subject control over a situation, and temporal aspect, as well as speaker stance. With respect to the latter, we have noted that dynamic eventive-marked utterances tend to carry such emotive tones as relief, pride, anxiety, determination, disappointment, and so forth. In relation to these issues, we will also examine this phenomenon as a grammatical and cognitive means of expressing an event conceptualization that deviates from what would be considered an ordinary or expected occurrence.

Example (45) from Maldonado (1992) will introduce this notion. The sentence pairs below are constructed such that the (a) version represents what is expected, that is, that leaves fall from trees in autumn. In this utterance, the use of se in Spanish would render the reading odd. In the (b) example, on the other hand, the occurrence is clearly beyond what one would expect from nature, that is, leaves fell from the tree in the springtime, and the utterance here naturally takes Spanish se.

(45) (adapted from Maldonado 1992: 345-346)

 a. En el otono, las hojas (*se)caen de los arboles.
 in the fall the leaves (*MM) fall-3PL from the trees
 'In the fall, the leaves fall from the trees' [no middle;
 expected occurrence] [neutral reading].

 b. En la primavera, las hojas (se)cayeron de los
 in the spring the leaves (MM) fall-3PL-PST from the
 arboles.
 trees
 'In the spring, the leaves fell from the trees' [middle
 marking, unexpected occurrence] [emphatic reading].

The Korean and Japanese counterparts in (46) and (47), respectively, express the identical expected/unexpected contrast, with the auxiliaries in question affecting the (a) and (b) utterances much in the same way as Spanish se in (45).

(46) (Korean translation of [45])

 a. kaul ey-nun namwu eyse iph-i ttele-cinta
 autumn in-TM tree from leaf-SM fall-PLN
 'In autumn, leaves fall from the trees' [no middle, expected
 occurrence] [neutral reading].

 b. pom ey-nun namwu eyse iph-i
 spring in-TM tree from leaf-SM
 ttele-cye-peli-ess-ta
 fall-A/E PELITA-PST-PLN
 'In spring, leaves fell from the trees' [middle marking,
 unexpected occurrence] [emphatic reading].

(47) (Japanese translation of [45])

 a. aki ni wa konoha ga ki kara ochiru
 autumn in TM leaves SM tree from fall-NPST
 'In autumn, leaves fall from the trees' [no middle, expected
 occurrence] [neutral reading].

 b. haru ni wa konoha ga ki kara ochi-te
 spring in TM leaves SM tree from fall-TE
 shimaimashita
 SHIMAU-PST POL
 'In spring, leaves fell from the trees' [middle marking,
 unexpected occurrence] [emphatic reading].

The sentence pairs in (45)-(47) provide a nice contrast between expected and unexpected occurrences in natural environments, and thus between canonical event conceptualizations and those which the dynamic eventive marks as noncanonical and out of the ordinary. This type of contrast is also observed in all three languages in (48)-(50), where the (a) version of the sentence pairs expresses a neutral statement of an event, and the (b) version, a stance of sadness, disappointment, anxiety, etc.

(48) Spanish:

 a. Llevo al nino a la guarderia
 'S/he took the child to kindergarten' [neutral, expected,
 ordinary].

 b. Se llevo al nino(al campo)
 'S/he took the child to the countryside' [emphatic stance;
 occurrence = out of the ordinary].

Here, by virtue of the use of se in the (b) sentence, the event is clearly portrayed as out of the ordinary. The speaker's stance could be one of disappointment ('unfortunately, s/he took the child to the countryside'), the child could have been kidnapped, and so forth. From the point of view of temporal aspect, this event could have occurred abruptly, it could have occurred only once, etc. If we were to add the phrase como siempre 'as usual', the reading with se would then become odd, as in (48').

(48') ([48b] portrayed as a normal, usual occurrence)
 ??Se llevo al nino(al campo) COMO SIEMPRE
 'S/he took the child to the countryside, as usual.'

Let's now examine the Korean and Japanese counterparts. (49) below expresses the identical sentence pair as (48):

(49) Korean

 a. ku salam-un ku ai-lul yuchiwen-ey
 the person-TM the child-OM kindergarten-DAT
 teyli-ko-kass-ta
 take-and-go-PST-PLN
 'S/he took the child to kindergarten' [neutral, expected,
 ordinary].

 b. ku salam-un ku ai-lul -sikol-lo
 the person-TM the child-GM -countryside-LOC
 teyli-ko-ka-peli-ess-ta
 take-and-go-A/E PELITA PST-PLN
 'S/he took the child to the countryside' [emphatic stance;
 occurrence = out of the ordinary].

Once again, the (a) sentence above denotes an expected, ordinary occurrence, whereas the (b) utterance with V-a/e pelita portrays the event as somehow beyond the speaker's expectation. This utterance carries with it a tone of 'Unfortunately, the person took the child to the countryside.' And once again, if we were to add the adverbial phrase denoting the sense that the action was a usual one, the reading with V-a/e pelita would become odd, as in (49'):

(49') ([49b] portrayed as a usual, ordinary occurrence)
 ??ku salam-un ku ai-lul -sikol-lo
 the person-TM the child-OM -countryside-LOC
 YENU TTAY-WA KATCH-I
 AS USUAL
 teyli-ko-ka-peli-ess-ta
 take-and-go-A/E PELITA PST-PLN
 'S/he took the child to the countryside.'

Japanese example (50) is identically parallel, both semantically and pragmatically, to (48) and (49):

(50) Japanese

 a. kare wa kodomo o youchien ni tsurete itta.
 he-TM child-OM kindergarten-DAT take-go-PST-PLN
 'He took the child to kindergarten' [neutral, expected,
 ordinary].

 b. kare wa kodomo o inaka ni tsurete itte
 he-TM child-OM countryside-DAT take-goTE
 shimatta.
 SHIMAU-PST-PLN
 'He took the child to the countryside' [emphatic stance;
 occurrence = out of the ordinary].

Here, in the (b) version, the speaker's stance again could be one of disappointment or anxiety; the subject 'he' could have taken the child back to his hometown without his mother's consent, it could have been an abrupt occurrence, etc. Clearly, though, the reading in (b) is not at all neutral.

In (50') we have added the adverbial phrase itsumo no you ni 'as usual' to the (b) example, and its reading, too, now becomes odd and unnatural:

(50') ([46b] portrayed as a usual, ordinary occurrence)
 ??kare wa kodomo o ITSUMO NO YOU NI inaka ni
 he-TM child-OM AS USUAL countryside-DAT
 tsurete itte shimatta.
 take-go TE SHIMAU-PST-PLN
 'He took the child to the countryside, as usual.'

Thus, the relationship between dynamic eventive marking and canonical event conceptualizations seems rather clear as it affects all three languages under investigation. Where events are portrayed as expected, routine, or otherwise within the domain of natural occurrences, we generally do not find evidence of the dynamic eventive. Conversely, where events are portrayed as out of the ordinary and beyond the realm of natural expectation, they can indeed be marked grammatically with a dynamic eventive morpheme or morpheme group.

5. Conclusion

This study has attempted to demonstrate some striking similarities between Spanish se and the two auxiliary constructions in Korean and Japanese, respectively, all of which can be designated as instances of the dynamic eventive. While these two auxiliaries derive from a completely different source than the Spanish reflexive, their grammatical and pragmatic patternings actually overlap in powerful ways.

More importantly, what this paper has attempted to show is the grammatical systematicity that exists in human language with respect to how we perceive, conceptualize, and report events. That is, what is meaningful in event conceptualizations such as transitivity valences, aspectual characteristics, degree of control over an event occurrence, the expectedness vs. the unexpectedness of an occurrence, etc., seems to pattern in highly systematic ways across languages that are totally unrelated to each other. Grammar is ostensibly much more than sentence structure and the meaningful stringing together of words it is a systematic reflection of human cognition, event perception, and emotion.

Received 23 July 2001

Revised version received

13 June 2003

Pennsylvania State University

Notes

(1.) The author would like to thank Noriko Akatsuka and Hanae Katayama for their help with the Japanese examples, Jong Oh Eun, Bong Jo Kang, and Eunju Kim, with the Korean examples, and Jacqueline Toribio, Jaime Gelabert, and Eduardo Negueruela, with the Spanish examples. The term "dynamic eventive" is credited with gratitude to Philip Baldi. The author is also indebted to the two anonymous reviewers for their invaluable comments and suggestions. The following abbreviations are used in the glosses:

ADV adverb MOD modal
ATTR attributive NPST nonpast
COM comitative OM object marker
COP copula PLN plain form
DAT dative POL polite
DEF deferential PRT particle
DIR directional PST past
FML formal PURP purposive
GEN genitive RSN reason
GOAL goal QT quotative
IMPER imperative SE sentence order
INSTR instrumental SG singular
INTERR interrogative SM subject marker
INTRPT interruptive SUSP suspective marker
LOC locative TM topic marker
MM middle marker

Correspondence address: Department of Linguistics and Applied Language Studies, The Pennsylvania State University, University Park Campus, State College, PA 16802, USA. E-mail: [email protected].

(2.) Alfonso (1966) refers to this general class of constructions as verb clusters; H. Sohn (1994: 423-426), as compound verbs; Falsgraf and Park (1994) as complex predicate constructions, and so on. The current paper will simply use the term "auxiliary verb" to express the compound nature of these constructions.

(3.) One exception to this is the Korean auxiliary V-ko malta, which is formed with the main verb + ko and the auxiliary malta. While malta on its own means 'to end' or 'to stop', it generally does not exist on its own as a lexical verb. (See Strauss 2002 for further discussion).

(4.) However, the patternings of Japanese V-te shimau and Korean -a/e pelita in actual discourse (both written and spoken) are not so straightforwardly similar, as shown in Strauss (1994).

(5.) In both Ono(1992) and Ono and Suzuki (1992), while some element of speaker emotive stance is mentioned with respect to the V-te shimau construction, that stance is strictly a negative one. However, Strauss (1994, 1996, 2002) has shown that these constructions also cooccur with expressions of positive evaluations.

References

Alfonso, Anthony (1966). Japanese Language Patterns. A Structural Approach. Tokyo: Sophia University Center of Applied Linguistics.

Arce-Arenales, Manuel (1989). Semantic structure and syntactic function: the case of Spanish se. Unpublished Ph.D. dissertation, University of Colorado, Boulder.

--; Axelrod, Melissa; and Fox, Barbara A. (1984). In Voice: Form and Function, Barbara Fox and Paul Hopper (eds.), 1-21. Amsterdam and Philadelphia: Benjamins.

Falsgraf, Carl; and Park, Insun (1994). Synchronic and diachronic aspects of complex predicates in Korean and Japanese. In Japanese/Korean Linguistics, vol. 4, Noriko Akatsuka (ed.), 221-237. Stanford: CSLI.

Garcia, Erica (1975). The Role of Theory in Linguistic Analysis. Amsterdam and Oxford: North Holland.

Kemmer, Suzanne (1993). The Middle Voice. Philadelphia: Benjamins.

Koojien (1991). Koojien Dictionary. Tokyo: Iwanami Shoten.

Lee, Keedong (1993) A Korean Grammar on Semantic-Pragmatic Principles. Seoul: Hangkuk Moon Hwa-sa.

Makino, Seiichi; and Tsutsui, Michio (1987). A Dictionary of Basic Japanese Grammar. Tokyo: Japan Times.

Maldonado, Ricardo Soto (1992). Middle voice: the case of Spanish se. Unpublished Ph.D. dissertation, University of California, San Diego.

Martin, Samuel (1975). A Reference Grammar of Japanese. New Haven and London: Yale University Press.

Ono, Tsuyoshi (1992). The grammaticization of the Japanese verbs oku and shimau. Cognitive Linguistics 3-4, 367-390.

--; and Suzuki, Ryoko (1992). The development of a marker of speaker's attitude: the pragmatic use of the Japanese grammaticized verb shimau in conversation. BLS 18, 204-213.

Soga, Matsuo (1983). Tense and Aspect in Modern Colloquial Japanese. Vancouver: University of British Columbia Press.

Sohn, Ho-Min (1994). Korean. London: Routledge.

--; and Sohn, Sung-Ock (1996). Grammaticalization in Korean. Paper presented at the International Conference on Korean Linguistics, Sydney, Australia.

Strauss, Susan (1994). A cross-linguistic analysis of Japanese, Korean, and Spanish: -te shimau, -a/e pelita, and the Romance reflexive se. In Japanese/Korean Linguistics, vol. 4, N. Akatsuka (ed.). Stanford: CSLI.

--(1996). Metaphors of "total enclosure" grammaticizing into middle voice markers. In The Construal of Space in Language and Thought, Rene Dirven and Martin Piitz (eds.), 395-416. Berlin: Mouton de Gruyter.

--(2002). Distinctions in completives: Korean V-a/e pelita, Korean V-ko malta, and Japanese V-te shimau. Journal of Pragmatics 34, 143-166.

--; and Sohn, Sung-Ock (1998). Grammaticalization, aspect, and emotion: the case of Japanese -te shimau and Korean -a/e pelita. In Japanese/Korean Linguistics, vol. 8, D. Silva (ed.). Stanford: CSLI.

Teramura, Hideo (1982). Nihongo no shintaxu to imi [Japanese Syntax and Meaning], Part II. Tokyo: Kuroshio.