Deceiving Others: Distinct Neural Responses of the Prefrontal Cortex and Amygdala in Simple Fabrication and Deception with Social Interactions Nobuhito Abe, Maki Suzuki, Etsuro Mori, Masatoshi Itoh, and Toshikatsu Fujii Downloaded from http://mitprc.silverchair.com/jocn/article-pdf/19/2/287/1756586/jocn.2007.19.2.287.pdf by guest on 18 May 2021 Abstract & Brain mechanisms for telling lies have been investigated re- sponses revealed increased brain activity of the left dorsolateral cently using neuroimaging techniques such as functional mag- and right anterior prefrontal cortices, supporting the interpre- netic resonance imaging and positron emission tomography. tation of previous studies that executive functions are related Although the advent of these techniques has gradually enabled to making untruthful responses. The main effect of deceiving clarification of the functional contributions of the prefrontal the interrogator showed activations of the ventromedial pre- cortex in deception with respect to executive function, the frontal (medial orbitofrontal) cortex and amygdala, adding new specific roles of subregions within the prefrontal cortex and evidence that the brain regions assumed to be responsible for other brain regions responsible for emotional regulation or emotional processing or social interaction are active during social interactions during deception are still unclear. Assum- deceptive behavior similar to that in real-life situations. Further ing that the processes of falsifying truthful responses and de- analysis revealed that activity of the right anterior prefrontal ceiving others are differentially associated with the activities cortex showed both effects of deception, indicating that this of these regions, we conducted a positron emission tomog- region has a pivotal role in telling lies. Our results provide clear raphy experiment with 2 (truth, lie) Â 2 (honesty, dishonesty) evidence of functionally dissociable roles of the prefrontal factorial design. The main effect of falsifying the truthful re- subregions and amygdala for human deception. & INTRODUCTION activity, but rather a complex one composed of vari- Since the pioneering research of Spence et al. (2001), an ous cognitive functions. Although the inhibition of true increasing number of neuroimaging studies have shown responses and the production of deceptive ones are contributions of the prefrontal cortex to deception (Abe prerequisites for deception, the intention of deceiving et al., 2006; Mohamed et al., 2006; Davatzikos et al., others (deception with social interactions, i.e., delud- 2005; Kozel et al., 2005; Langleben et al., 2005; Lee et al., ing a questioner, irrespective of whether a response is 2005; Nunez, Casey, Egner, Hare, & Hirsch, 2005; Phan truthful), which is likely to be accompanied by emo- et al., 2005; Kozel, Padgett, & George, 2004; Kozel, tional regulation or social interaction, must be regarded Revell, et al., 2004; Ganis, Kosslyn, Stose, Thompson, as another important determinant for human deception. & Yurgelun-Todd, 2003; Langleben et al., 2002; Lee There is a growing body of evidence for dissociable et al., 2002). Although subregions of this area have functional specialization of subregions within the pre- been differentially activated in these studies, their pre- frontal cortex from previous neuropsychological, neuro- cise roles in relation to deception are still unclear. imaging, and animal studies (Fuster, 2001). Above all, in Previous neuroimaging studies of deception have fo- relation to deception, it is worthwhile noting that the cused on the truthfulness of responses as an essential lateral (especially dorsolateral) prefrontal cortex is asso- factor (simple fabrication, i.e., the inhibition of true re- ciated with executive function and the medial (especially sponses and the production of deceptive ones). How- ventromedial) prefrontal cortex with emotional regula- ever, the reported activations of the prefrontal cortex tion or social interaction (Mesulam, 2000). might not be solely associated with processes of falsify- In the present positron emission tomography (PET) ing responses, because deception is not a simple mental study, we used a novel task paradigm to test directly our hypothesis that during the telling of lies, the lateral prefrontal cortex would be associated with the pro- cesses of inhibition of true responses and production of Tohoku University, Sendai, Japan deceptive ones, and the medial prefrontal cortex would D 2007 Massachusetts Institute of Technology Journal of Cognitive Neuroscience 19:2, pp. 287–295 Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/jocn.2007.19.2.287 by guest on 26 September 2021 be associated with the intention of deceiving an interro- gator. Based on the findings of previous studies (Abe et al., 2006; Mohamed et al., 2006; Kozel et al., 2005; Lee et al., 2005; Nunez et al., 2005; Kozel, Padgett, et al., 2004; Kozel, Revell, et al., 2004; Ganis et al., 2003; Langleben et al., 2002), involvement of the anterior cingulate cortex related to the inhibition of true responses and the pro- duction of deceptive responses would also be posited. In addition, activity of the amygdala crucial for emotional Downloaded from http://mitprc.silverchair.com/jocn/article-pdf/19/2/287/1756586/jocn.2007.19.2.287.pdf by guest on 18 May 2021 processing (Dolan, 2002; LeDoux, 2000) would be ex- pected in relation to the intention of deceiving the in- Figure 1. Schematic of deception tasks in the present study. terrogator. During PET scanning, participants performed Conditions were varied as a function of the truthfulness of responses a series of deception tasks determined by two discrete (truth vs. lie) and of attitude to the interrogator (honest vs. dishonest). factors: the truthfulness of responses and the intention of (a) HT task: The participants obeyed the interrogator and told the truth in response to questions. (b) HL task: The participants deceiving the interrogator. Our results underscore an es- obeyed the interrogator and told lies in response to questions. sential aspect of human deceptive behavior that has never (c) DT task: The participants deceived the interrogator and told been explored and provide clear evidence that at least the truth in response to questions. (d) DL task: The participants two factors essential for human deception are supported deceived the interrogator and told lies in response to questions. by distinct subregions within the prefrontal cortex. imenter (the second experimenter) secretly asked the METHODS subjects to deceive the interrogator by answering the questions truthfully. (4) Dishonest–lie (DL) task: Sub- Participants jects were instructed by the interrogator to tell the truth Sixteen male volunteers who had no history of neurolog- about their past memories, but approximately 2 min ical or psychiatric disease were paid to take part in this before the initiation of this task, the second experiment- study (mean age = 20.3 years, range = 18–22 years). er secretly asked the subjects to deceive the interrogator There were no pathological findings on magnetic reso- by answering the questions untruthfully. In these four nance imaging (MRI) of any of the subjects’ brains. All tasks, the use of the four lists of questions related to of the subjects were right-handed on the Edinburgh autobiographical semantic memory was counterbal- Handedness Inventory (Oldfield, 1971). The subjects anced across the subjects. The order of the four tasks gave their written informed consent in accordance with was also counterbalanced across the subjects. the Declaration of Helsinki, and the guidelines were ap- Before the transmission scan, the interrogator explain- proved by the Ethical Committee of Tohoku University. ed to the subjects that they had to perform four tasks and that they had to give untruthful answers in response to the questions in two of the tasks. Just before the Stimuli initiation of the four tasks, the interrogator instructed For the deception tasks during PET scanning, we pre- the subjects to tell the truth in the HT and DL tasks and pared 48 questions related to autobiographical semantic to tell lies in the HL and DT tasks. During approximately memory (e.g., ‘‘What is the name of your primary school?’’ 10-min breaks between the tasks, which were required ‘‘Who is your best friend in junior high school?’’). These for sufficient decay of the radioactive tracers, the interro- questions included 16 on primary school life, 16 on junior gator always left the experiment room, while the second high school life, and 16 on high school life. They were experimenter and the subject remained in the room. divided into four lists, each consisting of 12 stimuli (four Before the DL and DT conditions, respectively, during questions per school period). these breaks, the second experimenter asked the sub- jects to deceive the interrogator by telling lies or the truth about the presented questions. The subjects were also Tasks told that the interrogator did not know he would be de- During PET imaging, the subjects performed the follow- ceived and that deceiving the interrogator was the most ing four tasks (Figure 1). (1) Honest–truth (HT) task: important purpose of the study. Furthermore, subjects Subjects were instructed by the interrogator to tell the were asked to make sure that the interrogator did not truth about their past memories. (2) Honest–lie (HL) discover the study’s real purpose. task: Subjects were instructed by the interrogator to tell To elucidate the