Neurolinguistic & psycholinguistic investigations on evidentiality in Turkish

Agrammatic 
Gender 
Age 
Handedness Education 
Etiology 
Post-onset
(months) 
A1 
Male 
65 
Right 
Secondary 
Left CVA 
5 
A2 
Female 
36 
Right 
Primary 
Left CVA 
38 
A3 
Female 
33 
Right 
Primary 
Left CVA 
1.5 
A4 
Female 
46 
Right 
Primary 
Left CVA 
5 
A5 
Male 
60 
Right 
Secondary 
Left CVA 
11 
A6 
Female 
24 
Right 
Secondary
Left CVA 
5.5 
A7 
Male 
37 
Right 
College
Left CVA 
14 
Control 
C1 
Male
65 
Right 
Secondary 
C2 
Female 
36 
Right 
Primary 
C3 
Female
33 
Right 
Primary 
C4 
Female 
46 
Right 
Primary 
C5 
Male 
60 
Right 
Secondary 
C6 
Female 
24 
Right 
Secondary
C7 
Male 
37 
Right 
College

---

38 
clauses were matched with a subject (i.e., 
kadın
“the woman”) and a 
temporal adverb referring to past (
az önce
“previously”). Ten events were 
displayed with short motion clips (duration range: 15 – 20 seconds) where 
the whole action was shown from beginning to the end. The description of 
these events require the use of a verb inflected with –DI, the 
direct 
perception
evidential. A photo with an initial state and a resultant state 
depicted ten events as an inferable evidence for what has happened (e.g., a 
man with an orange and a man with orange peels). The description of such a 
resultant state requires the use of a verb inflected with –mIş,
 
the 
inferential 
evidential. Ten declarative sentences (each five words in length) inflected 
with the 
direct perception
evidential (i.e., 
az önce adam çorap giydi
“just 
before the man wore the socks”) were auditorily presented to the 
participants. They were asked to retell the sentences to another person. The 
retelling requires switching from the 
direct perception
into
 
–(I)mIş
,
the 
reportative 
evidential. Thus, the test items (
n 
= 30) formed three conditions 
of information source, each appropriate for the use of an evidential form 
(see Figure 2.3 for examples of stimuli for each condition, and Appendix 
A2 for the list of verb clauses).
Procedure for sentence completion 
The test items were presented to the participants through a laptop computer 
in randomized order in a single block. The test items started with a fixation 
of 2000 ms. The fixation was depicted with a plus centered on a white 
background. Two hundred ms offset to the presentation of each stimulus, a 
target sentence appeared in which the verb (in final position in Turkish) was 
left out. The target sentences were formatted in 36 point font in white color 
on a black background. The target verbs were given in bare form (i.e., 
ye 
“eat”) above the target sentences. The following instruction was given: 
“You will see different movie clips and photo pairs, and hear sentences in a 
random order. You should tell me what happened because I do not know. I 
have never seen or heard these items before. Ok? For this, after each item a 
sentence will appear. I will read aloud this sentence but the final part is left 
out. That is a verb. This verb is given above the sentence. But be careful. It 

---

39 
has no morphemes. You should tell me the verb form to complete the 
sentence. There is no time limit, please, do not hurry.” 
The sentence completion task was administered individually. Three 
extra items, one for each condition, were used in a practice session and 
these items were repeated until the task was fully understood. Once the 
experiment had started, presentation of the items was not repeated, but the 
experimenter repeated the target sentences, which the participants had to 
complete, when necessary. Each test item took one minute maximally.
The answers were scored as correct when the participant produced the 
verb with the intended inflection. Both a quantitative (accuracy of produced 
target verb forms) and qualitative (error types coding) analyses were 
performed. If the participant did not produce the verb or produced the verb 
in its bare form, this was counted as an omission error. If the participants 
produced an inappropriate verb form this was counted as a substitution 
error.