Linguistic analysis of deception cues from a convicted murderer

A paper authored by Gina Villar, Joanne Arciuli and myself has been accepted for publication in an upcoming issue of Applied Psycholinguistics. The paper reports the research that Gina conducted for her Honours dissertation, and which she presented at the Australian Psychological Society’s 2009 conference in Darwin. Gina was awarded Charles Sturt University’s APS Prize for 2009 and is now a doctoral candidate at the University of Sydney.

This project relates to our earlier Applied Psycholinguistics paper. In the previous project, we examined linguistic cues to lying in a laboratory experiment, in which we instructed participants to lie or tell the truth about certain things. The advantage of the experimental approach is that we knew with absolute certainty when the participants were lying and when they were telling the truth. But the downside is that it was an artificial situation and the stakes involved in lying were low. We tried to motivate the participants to try their hardest to fool the interviewer, but there weren’t any consequences outside of the experiment if they failed to do so.

In doing research that looks at real-world, high-stakes interviews, the problem is that it is more difficult to know with certainty what is true and what is false. But this paper identified a single case in which a reasonable determination of truths and lies was possible — in the statements of Scott Peterson, who was convicted of murdering his wife and unborn child. By comparing statements that Peterson made during media interviews and in recorded conversations with independent evidence that was introduced during his murder trial, it was possible to identify specific statements that were clearly corroborated (truths) or contradicted (lies) and compare their linguistic features.

Here is the abstract, followed by the citation details:

Previous studies have demonstrated a link between language behaviours and deception; however, questions remain about the role of specific linguistic cues, especially in real-life high-stakes lies. This study investigated use of the so-called filler, ‘um’, in externally verifiable truthful versus deceptive speech of a convicted murderer. The data revealed significantly fewer instances of ‘um’ in deceptive speech. These results are in line with our recent study of ‘um’ in laboratory elicited low-stakes lies (Arciuli, Mallard, & Villar, 2010). Rather than constituting a filled pause or speech disfluency, ‘um’ may have a lexical status similar to other English words and may be under the strategic control of the speaker. In an attempt to successfully deceive, humans may alter their speech, perhaps in order to avoid certain language behaviours that they think might give them away.

Villar, G., Arciuli, J., & Mallard, D. (in press). Use of ‘um’ in the deceptive speech of a convicted murderer. Applied Psycholinguistics.