Evidentiality in Uzbek and Kazakh

1.2 
Evidentiality as an Areal Feature 
At about the same time that Jakobson (1957/1971) first employed the term evidential, 
other scholars began to examine how evidential meaning was expressed in similar ways 
throughout much of Eurasia. The expression of evidentiality in the Balkans, the Middle East, the 
Caucasus, Central Asia, and Siberia is quite different from that observed in the Americas, East 
Asia, and New Guinea. This has led to the suggestion that the expression of evidential and 
related meanings is an areal feature in Eurasia. 
Languages bearing this areal feature may be said to belong to an “evidentiality belt” 
(Aikhenvald 2004). Situated at the heart of this belt are the Turkic languages, which has led at 
least Aikhenvald (2004) to claim that these languages are responsible for the spread of the 
grammatical marking for evidentiality as an areal feature. Such claims make the study of Uzbek 
and Kazakh especially important, as they are situated at the heart of the proposed belt and 
represent two separate branches of Turkic. 

---

22 
1.2.1 The Eurasian Evidentiality Belt 
The evidentiality expressed in central Eurasia is characterized by three main features: 
i. 
Expression of non-firsthand information source by a certain morpheme, with no 
counterpart that expresses firsthand information source 
ii. 
Expression of admirativity by this same morpheme, as well as other non-confirmative 
meanings 
iii. 
Close association between the expression of evidential meaning and past tense.
It is quite common in these languages that morphemes expressing non-firsthand information 
source and admirativity are derived from the perfect, whereas simple past tenses express 
confirmativity. Also common is the association between highly marked non-confirmativity and 
either the double marking of the perfect (pluperfects) or copular forms of the perfect (Friedman 
1979). When this is the case, the simple (i.e. non-copular or non-doubly marked) perfect is 
merely unmarked for confirmativity and may therefore express any range of confirmative or non-
confirmative meaning. 
Among the earliest accounts of evidentiality as an areal feature in Eurasia was that of 
Conev, who focused on the Balkans (1910/1911). Because the Balkans have been the starting 
point for many studies of evidentiality, it is worthwhile to examine briefly how evidentiality and 
related meanings are expressed in the languages of that region. Keep in mind that it was in 
reference to Balkan languages that confirmativity was first proposed as a feature relevant to the 
expression of evidential meaning; the use of this feature accounts for the distinction of the past 
and perfect forms, as well as the polysemy of so-called evidential morphemes, which express not 
only non-firsthand information source, but also admirativity and, sometimes, dubitativity. 

---

23 
In Macedonian and Bulgarian, the non-resultative and non-taxic past tenses are divided 
into paradigms that are traditionally labeled the definite past and the indefinite past. These 
correspond quite well to the (rather inappropriately named) görülen geçmiş zaman (seen past 
tense) and duyulan geçmiş zaman (heard/perceived/reported past tense) in Turkish. Under a 
theory that employs confirmativity as the relevant feature, these forms are more appropriately 
described as a simple past, which is positively marked for confirmativity, and a perfect, which is 
unmarked for confirmativity. Examples of the 3rd singular past tense of the verb ‘do’ in three 
languages of the Balkans are given in the Table 7: 
Table 7: Past and Perfect in Balkan Languages 
past [+
CONFIRMATIVE
] 
perfect [Ø 
CONFIRMATIVE
] 
Bulgarian