Evidentiality in Uzbek and Kazakh
|
| -di/-DI Past
-gan/-GAn Perfect -(i)b/-(I)p Converb Past [+Confirmative] [Ø Confirmative] [-Confirmative] [+Definite] [-Definite] [Ø Definite] [-Distant] 1 [+Distant] [Ø Distant] While the converbial past -(i)b/-(I)p may be used to express non-confirmativity, it is limited in its uses as it must also express past tense. The copular forms ekan/eken are also marked as non-confirmative, but only sometimes express past tense, and may therefore be used in a wider variety of contexts. These forms are, however, restricted in the types of non- confirmativity that they may express, indicating non-firsthand information source and admirativity. This may be because they are copular forms, and therefore more formally marked; this formal markedness then corresponds to functional markedness. Uzbek and Kazakh differ from Balkan Slavic, Turkish, and other better-studied languages of the Eurasian evidentiality belt in that they possess forms like ekan/eken. The presence of these forms shifts the locus of evidential meaning away from the past tense and, instead, refocuses it on morphemes that more independently express non-confirmativity. 1 Temporal distance is only a feature in Uzbek – see 3.1. 158 When ekan/eken express non-firsthand information source, they are not limited to a certain type of non-firsthand information source. Aikenvald’s (2004) typology of evidential meaning breaks non-firsthand information source into four types: INFERENCE , ASSUMPTION , HEARSAY and QUOTATION . As shown in Chapter 5, ekan/eken express all of these meanings, except for quotations, which are expressed by means of a separate, complementation construction. In Uzbek, if a speaker wishes to make explicit that HEARSAY is the intended meaning, the form emish may be used; this form is never employed to indicate other types of non-firsthand information source. In Kazakh, the cognate form -mIs may be employed, but as described in Chapter 5, -mIs appears to have evolved into a marker of pure reportativity, with no meanings of non-confirmativity. Most descriptions of Kazakh either omit -mIs or state that it may attach to a bare verb stem (Johanson 2000; 2003), which it may not. When ekan/eken is employed in questions, one possible meaning is what I refer to as the evidential question. Evidential questions are employed to indicate the expectation that the hearer will base the response on the best possible grounds, or to express expectation that the hearer will have non-firsthand evidence for the answer (see Faller 2002). Evidential questions are typically questions into the knowledge of the speaker; they are therefore often used to indicate politeness or hedging, or, they may occur as part of an exchange in which the speaker and hearer discuss events of which neither party has firsthand knowledge. Ekan/eken, as well as Uzbek emish can also be employed to express admirativity, which is traditionally described as the linguistic expression of unexpected information. While many analyses treat non-firsthand information source (i.e. evidential meaning) and admirativity as separate phenomena (see DeLancey 1997, 2001), or consider admirative meaning to be an extension of evidential meaning (e.g. Johanson 2000, 2003), we are able to account for both 159 meanings by looking to the non-confirmative analysis employed by Darden (1977) and Friedman (1978, 1980). What distinguishes admirativity from evidential meaning is context. When the speaker has clearly witnessed the event being described, yet chooses to employ a non- confirmative form, the combination of non-confirmativity and clear, first-hand information produces the ironic, surprised, or unexpected meanings ascribed to admirative utterances (see Darden 1977 for a similar analysis of Bulgarian, and Friedman 1981 for Macedonian and Albanian). Perhaps the most surprising use of ekan/eken is the formation of rhetorical questions. Most of better-studied languages of the Eurasian evidentiality belt (e.g. Turkish, Macedonian) do not employ non-confirmative morphemes to create rhetorical questions, so there has been no previous attempt to incorporate rhetorical questions in the aforementioned non-confirmative analyses. This analysis readily accounts for admirativity and non-firsthand information source, but rhetorical questions pose a problem for this analysis. In Chapter 6, I proposed that admirativity and rhetorical questions be grouped together as EMOTIVE uses of language, following Jakobson’s (1960) account of the functions of language. The use of language in its emotive capacity serves to express the inner state of the speaker and is speaker-oriented language; admiratives, much like the exclamatives found in European languages, express a strong emotion on the part of the speaker, while rhetorical questions are questions the speaker asks of his or herself or questions that are intended to express emotional state, rather than to prompt a response. By combining the two possible interpretations of ekan/eken - non-firsthand information source and emotivity - with the two clause types that these forms may occur in, we produce the matrix found in Table 32. 160 Table 32: The Four Functions of Ekan/Eken [+E VIDENTIAL ] [+E MOTIVE ] [+D ECLARATIVE ] Evidential Statement Admirative [+I NTERROGATIVE ] Evidential Question Rhetorical Question In order to account for the fact that neither evidential questions nor rhetorical questions can easily be accounted for with the non-confirmativity analysis, I propose that these meanings exist as strategies for providing meaning to utterances that are morphosyntactically possible, but semantically incompatible. A number of scholars have noted that subjective types of modality are incompatible with certain types of questions (Lyons 1977; Papafragou 2006). While the categories of STATUS and MODALITY are somewhat different (as described in the Preface), it useful to compare the two concepts, as analyses of subjective MODALITY would appear to encompass non-confirmativity, which was originally proposed as a subvariety of STATUS . If we interpret non-confirmativity as a subjective type of meaning, we are able to account for the incompatibility of non-confirmative meanings and questions. Because the combination of ekan/eken and question forms is otherwise allowed, the semantic incompatibility of these two phenomena is remedied by the production of utterances bearing only the secondary meanings expressed via non-confirmativity: (non-firsthand) evidentiality and emotivity. The notion that non-confirmativity is, indeed, a subjective type of meaning is further reinforced by the meanings that are produced when ekan/eken combine with conditionals. A number of semantic works (Lyons 1977, etc.) have postulated that subjective types of MODALITY are incompatible not only with certain types of questions (as evidenced by the strong evidential and emotive meanings described above), but also with conditionals. In Uzbek and Kazakh, the combination of conditional forms and ekan/eken is morphologically possible, but the non- 161 confirmative semantics of ekan/eken should result in an ill-formed utterance. To remedy this incompatibility, a new meaning, one of obligation or desiderativity has resulted (231). (231) Men kim bilan bor-sa-m ekan? (Uz) I who with go- COND -1 SG EVID ‘Who should I go with?/Who would I like to go with?’ These incompatibilities, and the strategies that exist to repair these incompatibilities, indicate that non-confirmativity, which is a type of STATUS , should be connected to the concept of subjective MODALITY . Yüklə Dostları ilə paylaş:
|