Evidentiality in Uzbek and Kazakh

1.1.2 Evidentiality and Verbal Categories 
In formulating evidentiality as E
n
E
ns
/E
s
, Jakobson (1957/1971) stated that he considered 
evidentiality to be a high-level verbal category on par with tense, aspect, or mood. All of these 
categories are composed of the following four pieces: 
i. 
E
n
- narrated events 

---

15 
ii. 
E
s
- speech events 
iii. 
P
n
- participants of narrated events 
iv. 
P
s
- participants of speech events 
These may be combined in a variety of ways, resulting in the ten categories shown in Table 4. 
Table 4: Jakobson's Verbal Categories 
P involved 
P not involved 
Designator 
Connector 
Designator 
Connector 
Qualifier 
G
ENDER
S
TATUS
Quantifier 
P
n 
N
UMBER
P
n
P
e
V
OICE
E
n 
A
SPECT
E
n
E
n
T
AXIS
Shifter 
P
n
/P
s
P
ERSON
P
n
E
n
/P
s
M
OOD
E
n
/E
s
T
ENSE
E
n
E
ns
/E
s 
E
VIDENTIALITY
The most important changes in Aronson’s (1991) new calculus of verbal categories 
involved the redefinition of 
STATUS
and the elimination of evidentiality as a verbal category.
Whereas Jakobson had defined 
STATUS
as that “which characterizes [qualifies] the narrated event 
without involving its participants and without reference to the speech event” (E
n
) (Jakobson 
1957/1971, 134), Aronson defines 
STATUS
as the “subjective evaluation of the narrated event by 
the speaker, i.e., E
n
/P
s
” (1991, 114 - emphasis added) and redefines 
MOOD
using the formula 
previously assigned to 
STATUS
: E
n
. Evidentiality is only tentatively included in Aronson’s 
calculus, as the quantification of a narrated event in relation to the participant in a speech event 
(E
n
/P
s
). Aronson states that evidentiality should be regarded as “closely related to, or, better, a 
subvariety of 
STATUS
(E
n
/P
s
)” (1991, 116) as he knows of “no language that has a grammatical 
category that has evidential as its invariant meaning” (1991, 130). Aronson’s verbal categories 
are outlined in Table 5. 

---

16 
Table 5: Aronson's Verbal Categories 
Characterizing the participants of a narrated 
event 
Characterizing the narrated event itself 
With reference to the 
narrated event 
Without reference 
to the narrated 
event 
With reference to 
another narrated 
event 
Without reference to another 
narrated event 
Qualita-
tively 
Quantita-
tively 
Qualita-
tively 
Quantita-
tively 
Qualita-
tively 
Quantita-
tively 
Qualita-
tively 
Quantitatively 
E
s 
E
n
/E
s
T
ENSE
P
s 
P
n
P
s
P
ERSON
E
n
/P
s
S
TATUS
E
n
/P
s
E
VIDENTIAL
? 
-
E
s
, 
P
s 
P
n
/E
n
V
ERSION
P
n
/E
n
O
BLIQUE 
V
ALENCE
P
n
C
LASS
P
n
N
UMBER
E
n
/E
n
T
AXIS
E
n
M
OOD
E
n
A
SPECT
,
R
ESULTATIVITY
Aronson’s formulation of 
STATUS
as involving P
s
reorients this category back toward the 
speaker: the locus of evaluation and of knowledge. A number of recent theories relating 
STATUS
or 
MODALITY
and evidentiality relate these two ranges of meanings via the sorts of truth-value 
judgments a speaker makes based on his or her knowledge. That is, when a speaker’s knowledge 
is firsthand or shared with others, the speaker is more likely to be able to make an objective 
evaluation of the truth of an utterance, whereas when a speaker’s knowledge is non-firsthand or 
not shared with others, the speaker is less likely to be able to make an objective evaluation of this 
truth (see Nuyts 2001). 
This reformulation reflects Aronson’s previous work in which he claimed that 
evidentiality should be viewed as a subtype of 
CONFIRMATIVITY
, a subvariety of 
STATUS
. On the 
basis of data from Bulgarian, he proposed that when 
NON
-
CONFIRMATIVITY
is expressed, 
evidential meaning is a possible interpretation (1967). This analysis was expanded by Friedman 
(1977; 1978, etc.) to account for phenomena in Macedonian, and later, Turkish, Albanian, Lak, 
Avar, Georgian, and other languages of Eurasia. 
A consequence of this non-confirmative analysis of so-called evidential markers is that it 
allows a unified account of the apparent polysemy of these morphemes in many languages.

---

17 
Morphemes expressing evidential meaning (and in particular, non-firsthand information source) 
frequently express other non-confirmative meanings, such as dubitativity, non-volitionality, and, 
most notably, 
ADMIRATIVITY
. 
Throughout much of Eurasia, evidential morphemes express a set of meanings termed 
ADMIRATIVITY
(sometimes called “mirativity”), which is the linguistic expression of surprise at 
newly discovered information. Admirativity was first identified as a phenomenon by Dozon 
(1879) in his grammar of Albanian. Like evidentiality, its status as an independent category has 
been a point of debate. Some scholars, such as DeLancey (1997; 2001) have claimed that it 
merits the status of a category while others, such as Friedman (1988) and Lazard (1999) place it 
in the category of 
STATUS
or 
MODALITY
. 
Darden (1977) proposed that admirativity (in Bulgarian) could be accounted for by the 
same means as evidentiality. Given different contexts, a morpheme bearing non-confirmative 
meaning could bear either evidential or admirative meaning. In contexts where it is clear that the 
speaker has not witnessed an event, the use of 
NON
-
CONFIRMATIVITY
may result in the marking 
of non-firsthand information source. When the speaker has clearly just witnessed an event, the 
use of a non-confirmative form results in admirative meanings of irony or surprise. 
The connection between evidentiality and non-confirmativity, or, at least, the speaker’s 
evaluation of the truth of an utterance, is implicit in a number of terms employed by other 
authors. The term médiatif, employed in a number of French-language works, refers to the 
mediation of knowledge, and relates to the notion of the use speaker’s knowledge to evaluate 
statements (Guentcheva 1996; 2007; Lazard 1996; 2000). This term is frequently employed in 
discussions of evidential meaning in the languages of Eurasia, many of which employ non-
confirmative forms to express non-firsthand information source, admirativity, doubt, and the 

---

18 
like. In Turkic studies, the term indirective has been popularized by Johanson (2000; 2003) to 
refer to this same range of meanings. He differentiates indirectivity from standard definitions of 
evidentiality as follows: 
The source of knowledge--the way in which the event is acknowledged by the conscious 
subject in question--is not criterial; it is unessential whether the reception is realized 
through hearsay, logical conclusion or direct perception. The receiver P is not necessarily 
the speaker, P
s
, but may also be a participant of the narrated event, P
n
. Consequently, 
markers of indirectivity do not fit into narrow evidential schemes primarily distinguishing 
between the speaker’s non-first-hand and first-hand information (Johanson 2000, 61). 
Because markers of non-confirmativity in Turkic express both non-firsthand information source 
and admirativity, this distinction between pure information source and indirectivity is important.
While the claim of this work is that non-confirmativity is the most elegant way of accounting for 
the morphemes that can express evidential meaning, the term indirective is a succinct way of 
discussing the various consequences of the expression of non-confirmativity. 
A number of recent typological works make the implicit claim that evidentiality is an 
independent category. Among the broadest are those of Aikhenvald (2003; 2004). In these 
works, she claims that information source may be broken down into six types (Aikhenvald 2004, 
64):