Lying and truth-telling in social interaction: Basic mechanisms

Although deception mostly occurs in the spoken or written language, verbal deception--lying-- has received comparably little attention from both deception and language production research. Here we present three pre-registered online experiments (total N = 118, aged 18-35) using a novel paradigm to investigate lying and truth-telling in a socially interactive setting. Participants engaged in a picture-naming card game with a simulated task partner, sometimes choosing to lie strategically by producing the name of a card associated with their current card. Across experiments, they were slower to initiate lying than truth-telling, as shown by linear mixed-effects regression models. Interestingly though, response times (RT) and typing durations (analyzed across experiments) of truth-responses were relatively more affected by the social lying context than lie-responses, as shown by a comparison to a control task. In Experiments 2 and 3, we additionally manipulated memory load by displaying the potential lie underneath the current card (Experiment 2) and by comparing semantically related and unrelated truths and lies (e.g., skirt-pants vs. skirt-fork; Experiment 3). Here, lying was more strongly affected by task-associated memory load than truth-telling, but to a similar degree as non-deceptive responses in the control task. Taken together, while underscoring the robustness of RT-slowing in lying, these results hint at a special role of the truth and limited specificity for the effects observed in lying. Indicating influences of planning, control, and monitoring processes, the results can inform and integrate theories of deception and language production. Studying (verbal) lying offers novel perspectives to both research fields.