Timothy R. Levine

Distinguished Professor & Chair of Communication Studies, University of Alabama at Birmingham

My expertise involves topics of deception detection, interpersonal communication skills, credibility assessment & enhancement, interrogation, and persuasion/influence. 

Student, media, and consulting inquiries are welcome.

Truth Default Theory (TDT)

I call my theory Truth Default Theory (TDT for short). It offers an alternative view of deception and deception detection.

The basic idea of TDT is that when we communicate with other people, we not only tend to believe them, but the thought that maybe we shouldn’t does not even come to mind. This is a good thing for two reasons. First, and most important, the truth-default is needed for communication to function. Second, most people are mostly honest most of the time. But, the truth-default makes us vulnerable to deception. Fortunately, there are “triggers” that can break us out of our default-to-honest mindset and enable lie detection. TDT covers how this works and why.

One big difference between between TDT and prior views of deception is that TDT rejects the idea that the best way to detect deception is pay attention to nonverbal and verbal “cues.” Using cues make people poor lie detectors. My research shows that how people come off can be misleading and that there are much better ways to not getting fooled.

Outlines of the theory have been presented at The National Communication Association in November 2012 Orlando, CogSci 2013 (Cognitive Science Society) in Berlin, and International Conference on Language and Social Psychology in Honolulu in 2014. An article length summary has been published in a special issue of the Journal of Language and Social Psychology. The article length summary is [here]. The full theory is presented in book-length format to be published November, 2019 here: http://www.uapress.ua.edu/product/Duped,7305.aspx

The idea behind the theory first began to take shape when I saw the results of my first deception detection experiment (McCornack & Levine, 1990) when I was still a graduate student. We manipulated suspicion but found that even in the high suspicion condition, subjects were still truth-biased. To this day, all my results are consistent with truth-bias and I have never seen lie-bias in my own data. My main take away from that was the powerful and robust nature of truth-bias. Initially, I saw truth-bias as the result of flawed or lazy cognitive processing and as the proximate cause of poor deception detection accuracy. Over time, however, I came to a radically different conclusion. I now see truth-bias as functional, adaptive, and conducive to correct inference. Originally I had thought to call the theory “Truth Bias Theory,” but I changed “bias” to “default” to better capture the basic idea and because bias tends to have a negative connotation. Thanks to Professor Torsten Reimer for the suggestion to change bias to default.

The theory represents an integration of ideas by Dan Gilbert (Spinozian model of belief states), Gerg Gigerenzer (bounded rationality, adaptive toolbox, the simple heuristics that make us smart), Paul Grice (conversational implicature), and Erving Goffman (demeanor, presentation of self in everyday life) all applied to deception, and used abductively to explain and predict deception. Hee Sun Park, Pete Blair, Steve McCornack, and Torsten Reimer also contributed substantially to the core ideas. Rachel Kim and David Clare contributed substantially to the empirical testing. The National Science Foundation provided funding.

The theory predicts when people are truth-biased and when they abandon the truth default. It explains why accuracy in most experiments is better than chance, but only slightly better than chance. It also points to ways that accuracy can be improved. All the major claims of the theory have now been tested empirically and replicated.

TDT’s logical structure is reflected in 12 propositions and it is comprised of 13 modules, mini-theories, and effects. The propositional structure, the modules, and key definitions are provided in the [Levine (2014) JLSP article]. Thanks to Howie Giles for supporting publication. Duped provides the full story and the evidence to date.