Evidentiality in Uzbek and Kazakh

4.1.5.2 On -mIs in Kazakh 
Kazakh -mIs expresses similar meaing to that of Uzbek emish, in that it expresses 
reportativity. It has, however, taken a somewhat different path in its development, becoming a 
clitic and moving in the direction of becoming a quotative marker. 
The term clitic is used here in reference to -mIs for both phonological and morphological 
reasons. In Kazakh, most affixes beginning with /m/ alternate with /b/ and /p/ depending on the 
surrounding environment, but the /m/ in -mIs never changes. Moreover, some speakers employ 
disharmonic forms such as žoq-mis or keledi-mïs, which indicate that -mIs occupies a status 
somewhere between a phonologically independent word and a bound morpheme (see Zwicky 
1977). In terms of morphology, -mIs is extremely promiscuous and may affix itself to essentially 
any word that may occur sentence finally. This includes eken, which is typically always 
sentence-final in Kazakh: 
(158) Ol šetel-de eken-mis. (Kaz) 
he foreign.country-
LOC EVID
-
REPORT
'He's in a foreign country.' 
 
-mIs may also affix itself to other non-confirmative forms, namely, to the converbial past in -(I)p.
Note that person marking occurs between the converbial past marker and -mIs: 
(159) Olar-dïŋ ušağ-ï aspan-da bir žer-de žoğal-ïp ket-ip-ti-mis. (Kaz) 
they-
GEN
airplane sky-
LOC
a place-
LOC
disappear-
CVB
go-
CPST
-3-
REPORT
'Their plane (reportedly) became lost somewhere in the sky.' 
18
 
As previously noted, -mIs may not attach to verb stems, likely because it is derived from 
an earlier copular form *emis.  
18
Yerkimbay, Askhat. 2008. “Äwežaydağï ‘orïstilifobiya sindromï.’” Neweurasia, 7 Feb. 
Accessed 8 Feb 2011. http://neweurasia.net/kazakhstan/саясат/əәуежайдағы-орыстіліфобия-
синдромы/ 

---

111 
(160) a. 
*kel-mis (Kaz) 
come-
REPORT
b. 
kel-ip-ti-mis 
 
come-
CPST
-3-
REPORT
‘She (reportedly) came.’ 
 
Like copular forms, however, -mIs may attach to non-verbal predicates, including nouns, 
existentials, adjectives, and deontics: 
(161) Qïz bol-uw kerek-mis (Kaz) 
 
girl be-
INF
necessary-
REPORT
‘They say you have to be a girl.’
19
 
 
Aside from the third person marker -DI, which is found only on the present in -A/y and the 
converbial past in -(I)p, forms bearing -mIs are incompatible with person or number marking.
This restriction was likely due to the rarity of employing reportative forms to discuss events that 
happened to either the speaker or the hearer; this rarity was later encoded as a lacuna in the 
distribution of -mIs. 
Many of the contexts in which -mIs is found are quotative, and in these quotative contexts 
-mIs appears to emphasize the quotative or reportative nature of what is being repeated: 
(162) Sibir-ge žür-giz-e-di-mis degen xabar Kenesarï-ğa da kel-e-di (Kaz) 
Siberia-
DAT
go-
CAUS
-
PRES
-3-
REPORT COMP
news Kenesarï-
DAT
too come-
PRES
-3 
'The news is also coming to Kenesarï that they are being forced to go to Siberia.'
20
 
 
When in quotative contexts, -mIs is often followed by a complementizer such as dep or degen.
These forms are derived from the verb meaning 'say', and have historically been used to indicate 
quotations or reported speech. Recently, however, the scope of these forms has expanded 
beyond the introduction of speech complement clauses to include complements of thought and 
knowledge, as well as to introduce clauses indicating reason or purpose. Cognate forms in 
19
2009. “Algan žarïm qïz bolmay šïqtï ne isteymin.” Kazakh.ru, 25 Jun. Accessed 7 Feb 2011. 
http://www.kazakh.ru/talk/mmess.phtml?idt=102685&page=4 
20
Esenberlin, Iliyas. 2009. “Qošpendiler.” Ädebiy Älem. Accessed 8 Feb 2011. http://adebiet.kz/ 

---

112 
Uzbek have undergone similar changes, and it appears that when -mIs is present, its purpose is to 
indicate that the preceding material is reported speech, and not thought or knowledge or an 
indication of purpose. In (162), for example, without -mIs, there is an ambiguity between the 
gloss provided above, and one meaning 'The news is coming to Kenesarï in order to make them 
go to Siberia.' 
Given the unusual behavior of -mIs, it is perhaps best to analyze it as a purely reportative 
form with no non-confirmative meaning. The fact that -mIs is often affixed to non-confirmative 
forms indicates that it cannot express that meaning on its own, as does the fact that it does not 
express admirativity, a meaning which is associated not only with evidentiality, but also with 
non-confirmativity. Moreover, its use as a marker of reported speech often occurs in explicitly 
confirmative contexts, such as the one below in (163). In this example, as in (162), the speaker 
is merely repeating the contents of what has been previously uttered, as is indicated by the word 
söz 'word, news', and the use of the confirmative past on the verb esit- 'hear': 
(163) astana-ğa qayt-ïp ket-e-di-mis degen söz-di esti-di-k (Kaz) 
capital-
DAT
return-
CVB
go-
PRES
-3-
REPORT COMP
word-
ACC
hear-
PST
-1
PL
'We heard the news that he was going back to the capital.' 
21
 
 
Although the clitic -mIs is potentially useful in differentiating reported information from 
other kinds of non-firsthand information and in distinguishing between the various uses of the 
Kazakh say-complementizers, it is still somewhat rare. Not all speakers accept forms with -mIs, 
and of those who do, many identify these forms as archaic or literary.  
21
Dosžan, Dükenbay. 2009. “Tört patšanï körgen keywana.” Ädebiy Älem. Accessed 8 Feb 
2011. http://adebiet.kz/ 

---

113