Evidentiality in Uzbek and Kazakh
|
| ekan/eken occurs in interrogatives. The following examples (223) and (224) are illustrative of
this issue: 151 (223) Peter Moskva-ni yaxshi ko’r-gan ekan (Uz) Peter Mäskew-di žaqsï kör-gen eken (Kaz) Peter Moscow- ACC good see- PRF EVID ‘Peter liked Moscow.’ (224) Kim Moskva-ni yaxshi ko’r-gan ekan (Uz) Kim Mäskew-di žaqsï kör-gen eken (Kaz) who Moscow- ACC good see- PRF EVID ‘Who liked Moscow?’ The two possible interpretations of (223) can be clearly derivable from a non-confirmative treatment of ekan/eken, as seen in (225). Drawing from Lyons’ (1977) observation that subjectivity may be paralleled by performative verbs, we are able to make the connection between non-confirmativity and the resulting meanings explicit: (225) a. Peter (apparently) liked Moscow ≈ I will not confirm that Peter liked Moscow (because I have no firsthand evidence for that fact) b. Wow! Peter liked Moscow! ≈ I will not confirm that Peter liked Moscow (because I am surprised that he would) It is less easy to derive the two possible interpretations of (224), the interrogative utterance, by appealing to non-confirmativity. (226) a. Who (do you think) liked Moscow? Who will I not confirm liked Moscow (due to a lack of firsthand evidence for that fact) b. Who (the hell) likes Moscow?! Who will I not confirm liked Moscow (because I am surprised by that fact) The problem with allowing forms that indicate subjective evaluation in questions is that the question asked becomes a question about the speaker’s knowledge. In most languages, this sort of question is either highly constrained or simply not allowed. Futher evidence against a non-confirmative analysis of ekan/eken in questions is their ability to co-occur with markers of confirmativity, i.e. the confirmative past tense in -di/-DI. In Uzbek, at least, ekan may not follow any form of the past -di when in a declarative clause, and in Kazakh, the co-occurrence of eken and -DI is mostly restricted to certain types of cause-and- 152 effect constructions demarcated by the complentizer dep (227), suggesting that in embedded contexts, eken may bear a reportative function similar to -mIs. It is otherwise very rare for eken and -DI to co-occur. (227) De-gen-men, dağdarïs kel-di eken dep, qol quwsïr-ïp qara-p otïr-uw-ğa say- PRF -1 SG crisis come- PST EVID COMP arm cross- CVB look- CVB sit- INF - DAT bol-ma-y-dï. (Kaz) be- NEG - PRES -3 ‘As I’ve said, because a crisis has apparently come, we can’t sit around waiting with our arms crossed.’ 17 In questions, however, neither Uzbek nor Kazakh has any restrictions on the co-occurrence of ekan/eken and -di/-DI (228-229). (228) U ayt-di-mi-kan? (Uz) He say- PST - Q - EVID ‘Did he say that?’ (229) Bar-dï eken be? ~ Bar-dï ma eken? (Kaz) go- PST EVID Q ~ go- PST Q EVID ‘Did she go?’ This co-occurrence is unexpected because ekan/eken have been analyzed up to this point as non- confirmative, whereas -di/-DI have been analyzed as confirmative; their co-occurrence should, then, be semantically impossible. In addition to the semantic arguments outlined previously, this distributional evidence supports the idea that ekan/eken lose their non-confirmativity when they appear in questions. Uzbek and Kazakh appear to have developed rhetorical questions and evidential questions as a strategy for dealing with the incompatibility of non-confirmativity and interrogativity. Rather than completely disallowing these combination of interrogativity and non-confirmativity the secondary meanings expressed by non-confirmative forms (non-firsthand 17 Arnur, Asqar. 2009. “Ulïqpan Žoldasov: Almatïnï aralamağanïma üš žïl boldï.” Dastan Studiyasï, 11 Sep. Accessed 6 Oct 2010. http://www.dastan-studio.kz/?p=1904 153 information source and emotivity) become primary. The combination of interrogativity with non-firsthand information source, or evidential meaning, results in questions about the hearer’s knowledge or source of information, and the combination of emotivity and interrogativity results in rhetorical questions. These facts, combined with the inability of the non-confirmative meanings of ekan/eken to surface in the presence of a conditional, provide strong support for the claim that the primary meanings of these forms are not truly evidential, but are instead ones of a non-confirmative type of STATUS / MODALITY . In Uzbek and Kazakh, ekan/eken have identical morphosyntactic properties whether the intended reading is one of emotivity or of non-firsthand information source. However, in Nogay, a Kipchak language related to Kazakh, the non-firsthand information source meanings and emotive meanings of the cognate eken can differ in their morphosyntactic realizations. This distinction crucially occurs only in questions (Johanson 2003, data from Karakoç 2005): (230) a. Ne-ge kel-gen eken-ler? what- DAT come- PRF EVID - PL ‘Why have they (reportedly) come?’ b. Ne-ge kel-gen-ler eken? what- DAT come- PRF - PL EMOT ‘I wonder why they have come.’ When the question asked concerns source of information (as in 230a), person and number marking is placed on eken. When a rhetorical question is intended (230b), person and number marking is placed on the main part of the predicate. In declarative clauses, the placement of agreement markers is variable in Nogay, just as in Uzbek and Kazakh, and no distinction is made between the evidential or emotive types of non-confirmative meaning. The differences between these two forms are, I propose, the result of the grammaticalization of evidential and rhetorical questions. Whereas the various meanings expressed by ekan/eken in declarative clauses are 154 clearly related to non-confirmativity, in questions, the incompatibility of non-confirmativity has forced evidential and emotive meanings to become primary. In Uzbek and Kazakh, this appears to be a pragmatic effect, with speakers and hearers agreeing upon conventionalized interpretations of these semantically incompatible combinations. In Nogay, this pragmatic effect has grammaticalized to the point that these conventionalized interpretations are marked differently in the grammar. By appealing to non-confirmativity as the primary feature of ekan/eken, we are able to unify the various uses of these forms to express non-firsthand information source and admirativity. Because non-confirmativity is related to the subjective, or speaker-internal evaluation of the state of affairs, it follows that when non-confirmative forms are employed, the resultant meanings have to do with the speaker’s knowledge (as expressed by the speaker’s indication of non-firsthand information source) or a contradiction between the speaker’s knowledge and the state of affairs (as expressed by the speaker’s indication of admirativity). Although interrogativity is generally incompatible with the expression of non-confirmativity, there exist ways of reconciling this incompatibility. Because the secondary meanings of non- confirmative forms are compatible with interrogativity, these meanings, evidentiality and emotivity, become primary in these cases. 155 Yüklə Dostları ilə paylaş:
|