Evidentiality in Uzbek and Kazakh

1.1 
A History of the Study of Evidentiality 
Although evidential meaning was indentified by Al-Kāšğarī in the 11
th
century, it has only been 
studied in depth from 1957, when Jakobson formulated a calculus of verbal categories and 
included evidentiality in it. We can therefore divide the history of evidentiality studies into a 
pre-1957 and a post-1957 stages. It was in Jakobson’s (1957/1971) work that two major 
landmarks in evidentiality studies occurred: the creation of the term evidentiality and its 
formulation as a distinct verbal category. While definitions of evidential meaning have been 
based around how Jakobson characterizes it, the notion of evidentiality as a distinct verbal 
category has been the subject of much debate. 
1.1.1 Early Accounts of Evidentiality 
The first known mention of anything resembling evidentiality was formulated by Maḥmoud Al-
Kāšğarī in his 11
th
Century Dīvān Luğāt at-Turk, a comparative grammar of Turkic dialects (Al-
Kāšğarī 1982). In describing the distinction between two past tenses, he stated that “the 
difference between these two forms is that the dāl yā’ [-DI] on preterite verbs indicate that the 
action occurred in the presence of the speaker. The action is verified by its occurrence in his 
presence.” “Mīm šīn [-mIš], on the other hand, indicate that the action occurred in the absence of 
the speaker” (Al-Kāšğarī 1982, 297, cited in Friedman 2003, 189). Cognates of these forms are 
present in most Turkic languages, e.g. Turkish -DI and -mIş, Uzbek -di and emish; and the 
distinction described by Al-Kāšğarī is broadly applicable to these modern forms. While Al-
Kāšğarī’s description includes elements of information source (firsthand vs. non-firsthand), it 
also covers an important distinction that is the primary one for these morphemes: verificational 
(confirmative) vs. non-confirmative (Friedman 1978). This early description of the distinction 

---

14 
between the simple past -DI and the perfect -mIš is remarkably similar to Jakobson’s 
(1957/1971) definition of evidentiality nine centuries later. 
A somewhat more modern formulation of evidentiality is found in Franz Boas’ grammar 
of Kwakiutl (Kwak’wala), a Wakashan language of Vancouver Island. He states: “To the 
suffixes expressing subjective relation belong those expressing the source of subjective 
knowledge -- as by hearsay, or by a dream” (1911, 443). Boas’ description also anticipated 
modern descriptions of evidentiality, in as much as it references hearsay and includes the key 
phrase source of subjective knowledge. This reference to subjective knowledge is important, as 
this implies reference to 
STATUS
or 
MODALITY
, categories to which evidential meaning is 
inherently tied. 
Drawing of the work of Boas, Jakobson defined evidentiality as a speaker’s report of an 
event “on the basis of someone else’s report (quotative, i.e. hearsay evidence), of a dream 
(revelative evidence), of a guess (presumptive evidence), or of his own previous experience 
(memory evidence)”. In Jakobson’s calculus this is expressed as E
n
E
ns
/E
s
- the characterization 
of a narrated event and a narrated speech event (the source of information) with respect to a 
speech event (1957/1971, 135). This definition is the basic starting point for virtually all 
subsequent studies.