Neurolinguistic & psycholinguistic investigations on evidentiality in Turkish

Participant
: 
İzledim!
“I watched it” (or points to the appropriate choice) 
Notice that the participants were provided with two clues: they could 
answer on the basis of their memory on the previous task and they could use 
the information provided by the verb inflection. The practice items from the 
production task, one for each condition, were used as practice items for the 
second task as well. These items were repeated until it was clear that the 
participants understood the task. The scores were analyzed quantitatively 
(accuracy for source identification) and qualitatively (error type analysis). A 
response was counted as correct when the participant mentioned the correct 
source. When the participants did not remember the source, it was counted 
as no-recognition. When they misattributed a source type to another, it was 
counted as a source misattribution. There were three types of source 
misattributions: (a) ‘thought to have heard’ is when the participant 
substituted a source by the reportative condition, (b) ‘thought to have 
inferred’ is when a participant substituted a source by the inferential 

---

42 
condition, (c) ‘thought to have seen’ is when a participant substituted a 
source by the direct perception condition.
21
2.3.
 
Results 
2.3.1.
 
Sentence completion
The NBDs performed at ceiling in all conditions for production. Therefore, 
their data were not further analyzed. Since the sample size was rather small 
and not normally distributed, non-parametric tests (
Wilcoxon
) were used for 
statistical analysis. The mean numbers and ranges of correctly completed 
sentences of agrammatic speakers in the three conditions are given in Table 
2.3 (see Appendix A3 for individual scores).