Evidentiality in Uzbek and Kazakh

0.2 
Methodology and Conventions 
0.2.1 Data Sources 
For this work I have relied upon data from three very different sources: native speakers of Uzbek 
and Kazakh, data from literary publications and data from the Internet. I have avoided using data 
from grammars, as evidentiality and related meanings have been either ignored or inadequately 
described in most previous works. 
1
But see Erdal (1991, 383), (2004, 288, 320) for further discussion of the origin of this form. 

---

3 
Consultation with native speakers of Uzbek and Kazakh took place in and around 
Chicago and in Almaty, Kazakhstan. The data-gathering techniques conducted consisted of 
questionnaires, elicitation of examples, and open-ended conversation. Native speakers also 
participated in the elicitation of data via e-mail correspondence and online surveys. 
Data from literature comes from both original Uzbek and Kazakh literature and literature 
translated from English. This literature included novels (such as James Joyce’s A Portrait of the 
Artist as a Young Man), collections of folk tales, and newspapers. 
There are large communities of speakers of Uzbek and Kazakh on the Internet, and the 
various message boards and websites created by these speakers have been enormously beneficial 
for this work, owing to the breadth of available data provided by these sources. This data comes 
primarily from messageboards, but also comes from personal websites and news sites. To ensure 
that Internet data was not erroneous or produced by a non-native speaker, I made sure that 
unusual constructions were attested elsewhere and were deemed acceptable by native speakers.
In employing the sorts of spontaneous data found in native literature and on the Internet, I 
was able to find a great many constructions that speakers are unlikely to produce in isolation or 
in the course of ordinary consulation. The ability to search these corpora resulted in the location 
of a number of rare, yet grammatical, constructions and usages. Rhetorical questions and 
Kazakh -mIs, in particular, were two phenomena I would not have known about had I not been 
able to use these corpora. Internet data, in particular, provided a great many instances of the sorts 
of dubitative, rhetorical, ironic, and otherwise casual data that is otherwise not found in written 
literature; for this reason, it forms of the bulk of data in this work.
In all cases, I have been careful to employ data that is close to the standard dialects of 
Uzbek and Kazakh, although the similarities between Uzbek and Kazakh, at least in terms of 

---

4 
how evidential meaning is expressed, suggests that even highly divergent dialects in either 
language would not deviate from the patterns described here. Any mistakes in the analysis or 
presentation of this data in this work are those of the author.