Evidentiality in Uzbek and Kazakh

6.2.2 Theoretical Implications 
Many of the claims made in this dissertation are similar to those made for similar phenomena in 
other languages. In many ways, the expression of evidentiality and related meanings in Uzbek 
and Kazakh is similar to the expression of the same in Turkish, Azerbaijani, Georgian, Lak, 
Albanian, and Macedonian (see Friedman 1978; 1988). Perhaps the most important conclusion 

---

173 
that may be drawn in comparing all of these languages is that 
EVIDENTIALITY
is not a category, 
per se, but rather a possible interpretation of 
NON
-
CONFIRMATIVITY
. 
In typological works on evidentiality (such as Aikhenvald and Dixon 2003;Aikhenvald 
2004) the Turkic languages, Macedonian, Bulgarian, Georgian, and other Eurasian languages are 
considered to possess grammatical evidentiality. Works on Balkan and Caucasian languages 
(Darden 1977; Friedman 1977; 1978; 1988) have shown that an analysis involving (non-) 
confirmativity better accounts for the wide range of meanings expressed by so-called 
“evidential” morphemes. The analysis of Uzbek and Kazakh presented here also supports the 
non-confirmative analysis, and the brief examination of other Central Asian languages in the 
previous sections suggests that a non-confirmative analysis may apply to them as well. 
Aronson (1991), among others, has claimed that he knows of no language that possesses 
evidentiality as a grammatical category, and he only tentatively preserves it in his restructuring 
of Jakobson’s (1957/1971) verbal categories. The languages discussed in this work, and many of 
the other languages referred to, exhibit evidential meaning only as a result of the expression of 
non-confirmativity, which falls into Jakobson’s category of 
STATUS
. It is, in fact, the expression 
of evidential meaning via non-confirmativity that is one of the hallmarks of the Eurasian 
evidentiality belt. Outside of this belt, many languages have been claimed to possess an 
evidential category, yet it is still somewhat rare for these languages to be examined with a non-
confirmative analysis in mind, so it is still unknown how widespread the association between 
non-confirmativity and evidential meaning is. 
In further examining evidentiality in other languages, we expect that one of three possible 
conditions will hold true: that the expression of evidentiality is dependent upon the expression of 
non-confirmativity or some related sub-category; that evidentiality is epiphenomenal, and may 

---

174 
be expressed via several different media; or that evidentiality does, in fact, exist in as a category 
in some, but not all of the languages for which an evidential category has been claimed. 
Outside of certain schools of linguistics, there exists little work on (non-)confirmativity 
as a subtype of either 
STATUS
or 
MODALITY
. There does exist, however, a fairly recent body of 
work concerning subjective modality, particularly as it is expressed by English auxiliary verbs 
(Lyons 1977, Verstraete 2001, Papafragou 2006). Although, as explained in the preface, there 
are reasons to make distinctions between 
STATUS
and 
MODALITY
, treating (non-)confirmativity as 
a form of subjective 
MODALITY
may provide better insight into how this phenomenon functions.
As noted by Verstraete (2001), it is quite difficult to distinguish the subjective and objective 
interpretations of English modal verbs. If we treat the various (non-)confirmative morphemes in 
Uzbek and Kazakh (and other languages) as expressions of subjective 
MODALITY
, we have access 
to canonical forms with unequivocally subjective interpretations. 
In Chapter 5, I discuss the emotive function of language as described by Jakobson (1960) 
in reference to admiratives and rhetorical questions. Other functions of language are associated 
with certain types of utterances, namely, the 
REFERENTIAL 
function with declaratives, and the 
CONATIVE
with imperatives. What I have proposed is that admiratives and rhetorical questions 
be seen as examples of language being employed in its emotive function. I further include 
exclamatives, which function quite similarly to admiratives, in emotive class of utterances (see 
Andueza and Gutiérrez-Rexach 2010). The properties of exclamatives are well-known, yet it is 
still somewhat rare for utterances not exhibiting canonical exclamative morphosyntax to be 
included in a broader class of exclamative-like constructions. Following Zanuttini and Portner 
(2003), I propose that there exists a broad class of emotive utterances that includes exclamatives, 
admiratives, and rhetorical questions. Each of these types of utterances possesses unique 

---

175 
properties that vary language to language, yet by examining them as a class, we may gain insight 
into the emotive function of language. 

---

176