Eyes do not lie but words do
Evidence from eye-movement monitoring during reading that misuse of evidentiality marking in Turkish is interpreted as
deceptive
Evidentiality encodes how a speaker has access to the information contained in his/her proposition. It has been
shown that some ‘evidential language’ speakers make a deliberate choice of evidentials while telling lies (
Aikhenvald 2004). In this study, we recruited 40 native speakers of Turkish, an ‘evidential language’, to
judge statements with evidentials using an eye-movement-monitoring-during-reading study with an end-of-sentence deception
detection task. The participants read sentences with four conditions, containing a direct or indirect evidential form either
compatible or incompatible with the given information source. Our results show that the indirect evidential condition was detected
as a lie more often than the direct evidential condition. Readers had the tendency to judge stimulus material with
source-evidentiality mismatch to be untruthful. These findings were mirrored in the eye-movement data, as we found gaze duration
to be longer at the critical verb region for indirect evidential and mismatch conditions.
Article outline
- 1.Introduction
- 2.Methods
- 2.1Participants
- 2.2Materials and norming
- 2.3Procedures
- 3.Results
- 4.Discussion
- Note
- Abbreviations in glosses
- Appendices
- Appendix 1.Stimulus characteristics: Mean, SDs, and 95%CIs for surface frequency (per mil), word length, and witnessability
- Appendix 2.Mean outcomes from eye movement monitoring tasks for the spillover region. FPRT = First Pass Reading Time (Gaze Duration, ms),
RPD = Regression Path Duration (ms); TFT = Total Fixation Time (ms). The standard error of each mean value is given in
parentheses
- Appendix 3.Statistical outputs from linear mixed effects regression models computed with the eye-movement measures for the spillover
region (FPRT = First Pass Reading Time [Gaze Duration, ms], RPD = Regression Path Duration [ms]; TFT = Total Fixation Time
[ms])
-
References