Evidentiality in Uzbek and Kazakh

CHAPTER 6 
CONCLUSIONS 
6.1 
Evidentiality in Uzbek and Kazakh: A Summary 
Although 
EVIDENTIAL
meaning in Uzbek and Kazakh is expressed by a number of different 
forms, I have attempted in the previous chapters to explain how those forms are related, their 
origins, and their precise meanings, particularly in regard to their expression of 
(non-)confirmativity. The following sections attempt to condense the findings of the previous 
chapters into brief summaries. 
6.1.1 Major Findings 
In beginning to address the way that evidential meanings are expressed in Uzbek and 
Kazakh, it is necessary first to understand how ekan/eken, the major bearers of evidential 
meaning, are distributed. Although there is some dispute over the diachronic basis of these 
forms (Erdal 1991; 2004), synchronically, they behave as though they were copular forms of the 
perfect -gan/-GAn. Many grammars of Uzbek and Kazakh, and indeed, of other Turkic 
languages, take an approach to the verbal systems of these languages that does not account for 
the fact that most inflectional morphemes exhibit predictable distribution. My approach instead 
treats every possible configuration of the various markers of mood, negation, tense, etc. as an 
independent paradigm. 
Because these previous approaches made could not predict the distribution of ekan/eken, 
in Chapter 2, I proposed a basic distinction between finite and non-finite forms of the verb.
There are three finite verbal paradigms in Uzbek and Kazakh: the imperative-voluntative-
optative series, the simple past tense in -di/-DI, and the conditional in -sa/-sA. I call these 

---

156 
paradigms finite because they occur only when the verb they are attached to is acting as a 
predicate and because they take different types of agreement than other verbal paradigms (see 
Tables 10, 11, and 16 in Chapter 2). Non-finite series include gerunds, infinitives, agentives,