Neurolinguistic & psycholinguistic investigations on evidentiality in Turkish

participles (1.00 vs. 0.40; 
t
(18) = -1.567; 
p
= .135), or any of the other 
participle forms (–mIş
32
and –EcEk: both 
p
s > .11).
31
The infinitives produced by all participants were in embedded clauses. However, 
agrammatic participant A10 (see Appendix B4) produced most of the infinitives in 
isolation and only few in embeddings.
32
Not to be confused with the indirect evidential –mIş. The participle –mIş is used 
in relative clauses and is not a finite verb nor does it make reference to indirect 
information. The indirect evdiential and the participle –mIş were analyzed 
seperately.

---

72 
Table 3.3.
The means and standard deviations for the measures of verb 
inflection
3.3.4.
 
Verb tense and evidentiality 
In Table 3.4, the use of inflectional morphology for tense/aspect and 
evidentiality for both groups is given. The individual data are in Appendix 
B5. 
Table 3.4.
Means and standard deviations (sd) of the numbers and type 
token ratios (TTR) for the 3 most frequently used verb forms
There is no difference between the use of each type of inflection 
between the groups, although there is a trend for agrammatic speakers to 
produce more direct evidentials than the NBDs (direct evidentials: 
t
(18) = -
# finite 
verbs 
prop. finite 
verbs 
# non-finite verbs
Infinitives
Participles 
Gerunds 
agrammatic speakers
mean 
42.8 
0.87 
3.90 
1.10 
1.90 
sd 
9.78 
0.10 
4.88 
1.44 
1.85 
NBDs 
mean 
41.2 
0.78 
4.30 
4.60 
2.10 
sd 
9.53 
0.14 
3.46 
3.80 
1.37 
direct evidentials 
indirect evidentials 
present progressive 
number 
TTR 
number 
TTR 
number 
TTR 
agrammatic 
speakers
mean 
20.5 
0.72 
4.0 
0.68 
11.6 
0.71 
sd 
11.3 
0.14 
3.06 
0.32 
5.82 
0.19 
NBDs 
mean 
12.8 
0.86 
3.4 
0.86 
15.0 
0.83 
sd 
8.0 
0.13 
2.59 
0.31 
7.51 
0.16 

---

73 
1.758; 
p
= .096; indirect evidentials 
t
(18) = -0.473; 
p
= .642; present 
progressive: 
t
(18) = 1.132; 
p
= .273). 
Since the number of direct evidentials of the agrammatic speakers 
was surprisingly high, we inspected the raw data and looked at the diversity 
of the verbs that were used within each inflectional category.
The individual performances of the agrammatic speakers reveal a 
pattern similar to that of Bastiaanse and Jonkers (1998) and Anjarningsih 
and Bastiaanse (2011) with regard to the production of time reference. In