Rumiko Shinzato
2. Subjectivity vs. intersubjectivity
2.3 Unidirectionality in grammaticalization
The previous sections discussed the heirarchical and implicational relationship between subjectivity and intersubjectivity in a synchronic frame. This section points out its additional relevance to diachronic change, especially to the unidirec-tionality in grammaticalization. Not surprisingly, the subjectivity/intersubjectivity dichotomy finds its counterpart in grammaticalization theory as subjectification/i ntersubjectification, which Traugott (2003: 128) defines as follows:
(13) “...while subjectification is a mechanism whereby meanings become more deeply centered on the speaker, intersubjectification is a mechanism whereby meanings become more centered on the addressee....The hypoth-esis is that, for any lexeme L, intersubjectification is historically later than and arises out of subjectification.”
A perfect example of the shift from subjectification to intersubjectification comes from Japanese as in the development of mental verbs into speech act verbs. Traugott and Dasher (1987: 570) note the following development (MV and SAV stand for mental verb and speech act verb respectively).
(14) kotowaru: attested as MV meaning <<discern, discriminate>> from the 8th C, and a SAV with the meanings <<give reasons, apologize>> from the late 12th C, <<announce judgement>> from the late 13th C. The present meaning <<refuse>> is attested from the mid 19th C.
mitomeru: mi-miru <<see>> + tomeru <<stop>> attested as a MV mean-ing <<recognize>> from the 13th C, and as a SAV from the 17th C.
A similar path is also observed in the English verb, find, as below:
(15) find: appears from OE on as a MV; does not appear as SAV until 1400, and then only in legal contexts (e.g. find guilty means <<determine and declare guilty>>)
Subjectivity, intersubjectivity and Japanese grammar
Referring to such a unidirectional developmental path, they (ibid: 570) state:
(16) “Being in a certain state of mind is a prerequisite for a speech act... claim involves the speaker’s belief in the proposition, order involves the speaker’s desire for the addressee to do whatever is named in the proposition...”
Note that their remark is strikingly similar to Nakau’s implicational relationship (cf. example 4), and in line with the Japanese predicative order (cf. example 1).
In this junction, it should be mentioned that Romaine and Lange (1991: 265) follows Traugott and Dasher (1987) in hypothesizing that the expression, be like first developed as a marker of thought as in (17), and then as a marker of speech, as in (18).
(17) My mother said, “David where are you,” and he just came right out. I was like, I thought I lost him. I really thought I lost him. (ibid: 265)
(18) I was like, “Mom?” She was like “What!” (ibid: 253)
Romaine and Lange (ibid: 243) found that in their data, like tends to be used for self-representation, say for the speech of others, and go for both. Regarding this statistical difference, they state, “Insofar as only the speaker can have access to his/
her own thoughts, and like (rather than say or go) is more likely to be used for the representation of thought, this trend is not surprising.” The higher rate of the oc-currence of be like with the 1st person subject is also noted in Tagliamonte and Hudson (1999). After all, we know what others are thinking from what they say, not by our omniscient power.
Another relevant mental vs. speech act contrast can be seen in the develop-mental history of I see – you see. According to OED, I see as in the sense of ‘under-stand’, that is, the next stage of development from its concrete visual perception sense, was attested around 15C as in (19), while you see as an interpersonal dis-course marker appeared around 17C as in (20). This again conforms to the direc-tionality of subjectification > intersubjectification.
(19) Now I see and vnderstande that myn old synne hyndereth me and shameth
me. (1470–85 MALORY Arthur XIII. xix. 639)
(20) Because, you see, the present Government has 1,900,000 l.
(1657 CROMWELL Sp. 21 Apr. in Carlyle Lett. & Sp.) Some Japanese modal auxiliaries such as rashii ‘it appears’, yooda ‘it seems’ and sooda ‘I hear’ are polysemous between evidential meaning and mitigating mean-ing. In Nitta’s (1992) example (21), yooda (yoo desu) has an evidential function, indicating the proposition it appends to was yielded through the speaker’s inferen-tial process based on the available information (in this case what is expressed in the previous sentence). In contrast, the same auxiliary in (22), does not express
Rumiko Shinzato
any evidential meaning, but rather it makes the tone softer and indirect. Thus, it concerns more with politeness than evidentiality. Needless to say, the former aligns with mental verbs, while the latter does so with speech act verbs.
(21) Me-o-korasu to rei-no-futari ga butchouzura-de dete kita.
gaze when those two sbj serious-look came out Doomo senka wa sappari datta yooda
Somehow results top not great was seem
‘When I was gazing, those two came out with a serious-look (on their faces). It seems that the results were not that great.’
(22) (After checking his watch to make sure the time has already passed) Jikoku ni natta yoodesu. Honjitsu no kaigi wa kore de
time has come seem today ’s meeting top this with ohiraki-ni shi tai to omoimasu.
ajourn make would like to comp think
‘It seems the time is up. I would like to adjourn today’s meeting now.’
Nitta (1992: 7) asserts that what he calls ‘communication modality’ as in (22) was developed later than ‘judgement modality’ as in (21). Though his analysis is syn-chronic in essence, nonetheless, here again, the same directionality from mental to speech act verbs, or subjectification → intersubjectification is confirmed.15
Though not reflecting the contrast between mental and speech act verbs per se, Japanese examples such as (23) are also consistent with the unidirectionality (subjectification > intersubjectification) under discussion. As illustrated below, this unidirectionality is not limited to a particular part of speech, but applies to items across the board: (a) verb; (b) formal noun; (c) clausal connective; (d) quota-tive conditional; and (e) sentence-final particle. In the following examples, the ear-lier stages depict the happenings in the speaker’s inner world (affect, inference, judgement, exclamation), while the later stages exemplify other-oriented speech acts (camaraderie, invited hearer’s inference, vocative, summon). The earlier stag-es corrstag-espond to subjectification, as opposed to the later onstag-es which exhibit inter-subjectification:
(23) a. te shimau (affective marker > social dialect/camaraderie)
(Strauss & Sohn 1998)
15. In the case of yooda/yoodesu, the subjective/intersubjective distinction does not seem to be
‘dichotomous’, but rather ‘a matter of degree’. In fact, Traugott (2003: 134) also states, ‘...although it may in some instances be difficult to determine whether a new meaning is strictly subjective before it becomes intersubjective, nevertheless, nonsubjective > intersubjective > subjective is hypothesized not to be likely.’
Subjectivity, intersubjectivity and Japanese grammar
b. wake (speaker’s inference > invited hearer’s inference) (Suzuki 1998);
mono (subjective judgement > amae, dependency on the addressee
(Fujii 2000)
c. demo (subjective adversative connective > discourse marker of claim-ing floor, and changclaim-ing topic) (Onodera 2004) d. ttara (metalinguistic and subjective judgement > vocative)
(Shinzato 2007)
e. na (exclamation > summon) (Onodera 2004) In summing up this section, it discussed the diachronic counterpart of the syn-chronic hierarchical relation between subjectivity and intersubjectivity. Drawing on existing studies, it demonstrated that the diachronic shift from mental verbs >
speech act verbs, and also the shift from mental acts > speech acts are prevalent in Japanese across grammatical categories.