Evidentiality in Uzbek and Kazakh

Mandarin 
‘How tall he is!/How tall is he?’ 
(197) Nan to yuu uso o tsuku no deshoo! Japanese 
‘What lies he tells!’ 
The existence of these constructions in Mandarin and Japanese does not appear to have been 
substantiated by any further studies, and there have been contradictory claims that a completely 
distinct Japanese construction is equivalent to the European exclamative. This Japanese 
construction does not possess any 
WH
-element (Yamato 2010). 
In Uzbek and Kazakh, it is possible to form exclamative-like constructions that employ 
WH
-elements: 

---

138 
(198) Mïnaw qanday dämdi et! (Kaz) 
 
this how tasty meat 
‘What tasty meat this is!’ 
(199) Qanday yaxshi! (Uz) 
how good 
‘How good (it is)!’ 
In both Uzbek and Kazakh, this 
WH
-element is always ‘how’ (Uzbek: qanday, Kazakh: qanday, 
qalay). 
There is some overlap between the two types of emotive utterances in Uzbek and 
Kazakh, as it is common that these exclamative-like constructions formed by the presence of 
how will also be marked with ekan.
(200) Qanday yaxshi odam ekan! (Uz) 
how good man
EMOT
‘What a good man he is!’ 
Zanuttini and Portner (2003) argue that the emotive aspect of 
WH
-marked exclamatives derives 
from the presence of the 
WH
-element, which results in a type of pragmatic inference that they call 
widening. The result of this widening is an interpretation of surprise, unexpectedness, or 
extreme degree. While this approach clearly cannot apply to utterances marked only with 
ekan/eken, it also is insufficient to account for utterances marked with how. The approach 
offered by Castroviejo Miró (2006; 2010) treats the 
WH
-elements in exclamatives as degree 
constructions. This approach appears to be the best way to reconcile those situations in which 
both how and ekan/eken are present, as, I claim, how in these contexts is a degree operator (and 
specifically an intensifier), which leaves ekan/eken as the primary markers of emotivity. 
The first piece of evidence in support of Castroviejo Miró’s degree analysis (and, 
specifically, the intensifier analysis proposed here) comes from the inability of how-type 
exclamatives in Uzbek and Kazakh co-occur with any other intensifiers: 

---

139 
(201) Qanday *juda yaxshi odam! (Uz) 
 
Qanday *öte žaqsï adam! (Kaz) 
how *very good man 
‘What a very good man!’ 
In neither Uzbek nor Kazakh may both how and another degree operator or intensifier (such as 
Uzbek juda or Kazakh öte ‘very’) co-occur; this is in contrast with English, where (as seen in the 
gloss), this is entirely possible. As the double marking of degree is not allowed in these 
languages, it is reasonable to presume that qanday/qalay is a degree marker as well.
A second piece of evidence in support of the intensifier analysis comes from the inability 
of qanday or qalay to directly modify nouns, at least in their exclamative/intensifying capacity: 
(202) #Qanday ayol!
(Uz) 
 
#Qanday äyel!
(Kaz) 
 
how woman 
‘What a woman!’ 
The only possible analysis for utterances such as (202) is that of a question (What kind of 
woman?). Under the analysis of qanday/qalay as intensifiers, this sort of behavior is expected, 
as other intensifiers such as juda and öte cannot modify nouns. 
Because only qanday/qalay occurs in these constructions, as opposed to other 
WH
-
elements (such as what), it is likely that qanday/qalay in these cases has been reanalyzed as an 
intensifier. The emotive properties of degree operators, and intensifiers in particular, are well-
known (Athanasiadou 2007, among others), so the treatment of utterances marked by 
qanday/qalay as expressing emotivity is not at all problematic. It appears to be the case that the 
choice of the marked qanday/qalay over a non-
WH
-intensifier such as juda or öte has the 
pragmatic effect of specifically indicating emotivity. 
In terms of semantics, both admiratives and the true 
WH
-exclamatives found in English 
and many other languages have much in common. Zanuttini and Portner’s impressionistic 

---

140 
analysis of exclamatives as expressing “‘a sense of surprise,’ ‘unexpectedness,’ ‘extreme 
degree,’ and the like” (2003, 40) is nearly identical to the range of meanings described for the 
admiratives found in the languages of the Balkans (Friedman 1980; 1981). Zannutini and 
Portner propose a number of criteria for defining exclamatives, and while most of these criteria 
are syntactic, and therefore not applicable to non-
WH
-type utterances, one criterion in particular, 
that of scalar implicature, can apply to both 
WH
-exclamatives and admiratives. 
The scalar implicature criterion proposed by Zannutini and Portner (2003) requires that 
some aspect of the propositional content of the exclamative must fall outside the realm of normal 
expectation. In the example How well he plays the violin, it is implied that the violin playing in 
question should exceed ordinary expectations for violin playing, or, at least, for the violin 
playing of the subject. In Uzbek and Kazakh, this criterion is only upheld if the predicate of the 
admirative utterance is gradable: 
(203) Aqtöbe-de bala-lar-diŋ žağday-ï žaqsï eken! (Kaz) 
Aqtöbe-
LOC
child-
PL
-
GEN
condition-3 good 
EMOT
‘The conditions for children in Aqtöbe are good!’
7
 
 
(204) Yor, ko’z-lar-ing qop~qora ekan! 
Friend, eye-
PL
-2
SG INTENS
-black 
EMOT
‘How very dark your eyes are, friend!’
9
 
In (203), the interpretation is that the good-ness of the conditions in Aqtöbe is somehow 
exceptional, and in (204), the combination of reduplication (qop~qora) and the presence of ekan 
indicates that the darkness of the addressee’s eyes is similarly exceptional. In cases where the 
7
Žalmağanbetova, Aynur. 2009. “Aqtöbede balalardïŋ žağdayï žaqsï eken!” Aqtöbe Oblïstïq 
Qoğamdïq Sayasiy Gazet 28 May. Accessed 10 Mar 2011. http://www.aktobegazeti.kz/?p=1853. 
9
2007. Forum TFI: Stixi o Ljubvi, 9 Feb. Accessed 8 Feb 2011. 
forum.tfi.uz/forum_posts.asp?TID=389&PID=78150 

---

141 
predicate is non-gradable (such as an existential or a non-stative verb), this extreme degree 
reading is not found: 
(205) Aqša-m žoq eken! (Kaz) 
Money-1
SG NEG
.
EXIST EMOT
‘My money’s gone!’ 
(206) Ah, qo’l chiq-ib ket-gan ekan-da 
Ah arm dislocate-
CVB PFV
-
PRF EMOT
-
EXCL
‘Ah! His arm has become dislocated!’
11
Without the scalar implicature interpretation available, the above utterances express surprise or 
other strong emotions. 
On the basis of Turkish data, Zanuttini and Portner (2003) suggested that admiratives 
might be considered a type of exclamative. While admiratives and exclamatives do share much 
in common, at least in terms of semantics, such an analysis is bound to fail. First, admiratives 
and exclamatives differ greatly in terms of their morphosyntax. Admiratives (at least those of 
the Eurasian evidentiality belt) are marked with forms derived from perfects, and exclamatives 
are marked by 
WH
-elements. Secondly, the expression of admirativity is inextricably tied to the 
expression of non-confirmativity. The use of a non-confirmative form in a context where the 
speaker ought to be able to confirm the propositional content of an utterance results in meanings 
of surprise or irony or some other strong emotion. These resulting meanings must be pragmatic 
in nature, as the same morphemes that express admirativity also express non-firsthand 
information source and other non-confirmative meanings. Although both admiratives and 
exclamatives both indicate an emotive use of language, they should not be considered the same.
11
Xolmirzayev, Shukur. 1990. “Qadimda Bo’lgan Ekan.” Ziyouz. 
http://www.ziyouz.com/index.php?option=com_content&task=view&id=3077&Itemid=228 

---

142 
The meanings common to both types of utterance are likely due to the typological and pragmatic 
implications of employing language to express emotion.