In their thoughtful and stimulating essay, von Hippel & Trivers (VH&T) assert a number of intriguing proposals, none more thought-provoking than their central one that self-deception evolved in order to facilitate the deception of others.

My overall reaction to this central assertion is favorable. It is a well-formed hypothesis that readers easily grasp and that resonates with their intuition. The hypothesis, however, lacks one characteristic I wish it had more of – data. That is, the hypothesis is not completely new, having been forwarded, in some form or another, over that last quarter-century (Trivers Reference Trivers1985; Reference Trivers, Robinson and Tiger1991), and so it could profit now from direct data that potentially support it rather than from any additional weaving of indirect arguments and findings such as those the authors have spun here. It should be relatively easy to construct empirical studies to see if people engage in self-deception more eagerly when they must persuade another person of some proposition. Similarly, it should be easy to create experiments to see if people are more persuasive to others to the extent they have persuaded themselves of some untruth first.

Such empirical evidence, ultimately, is essential for two reasons. First, although intuitively compelling, VH&T's hypothesis already faces an empirical headwind. Many research findings oppose it. For example, VH&T suggest that people need self-deception to become convincing liars because others would be so good at catching their lies otherwise. However, one can reasonably read the literature on lie-detection to suggest that people are not very good at detecting lies (e.g., Bond & DePaulo Reference Bond and DePaulo2006; Ekman Reference Ekman1996), so this pressure does not really exist. Moreover, people do not seem to be especially skilled at lie detection under circumstances in which they arguably should be, given the authors' assertions. For example, people are not much better at detecting lies among those they know well (Anderson et al. Reference Anderson, DePaulo and Ansfield2002) or among those they are interacting with directly rather than merely overhearing (Bond & DePaulo Reference Bond and DePaulo2006).

In addition, one could argue from extant data that evolution would not have selected for self-deception, and the lying it supports, in the social setting associated with most of human evolution – one in which humans huddled together in small, interdependent groups, banded together against the potentially fatal dangers of nature and other competitive social factions (Brewer & Caporael Reference Brewer, Caporael, Schaller, Simpson and Kenrick2006). Recent arguments suggest that this small group setting boosted the survival value of groups whose members cooperated and cared for one another over groups whose members were more egoistic and selfish (e.g., Boyd & Richerson Reference Boyd and Richerson2009). One can presume that truth telling would be one of those behaviors so selected – and that self-deceptive lying within the small group would place any of its practitioners at a disadvantage.

Consider the fate of people in small groups who actively pursue one form of self-deception that VH&T discuss at length – people who engage in self-deceptive self-enhancement that allows them to display more confidence about themselves than their reality actually warrants. According to VH&T, this self-deceptive confidence is associated with many social advantages, but modern-day empirical data suggest the opposite – that this self-deception does not work well in small-group settings. At least two studies have found that people who boast with confidence about their talents and character are initially well-liked in small groups. However, over time, these individuals become the most disliked and least valued within those small social groups (John & Robins Reference John and Robins1994; Paulhus Reference Paulhus1998). Presumably, their self-deceptions are eventually found out, and whatever advantages they obtain initially are ones the group increasingly withholds as time goes on. This leads to a paradox. Perhaps self-deception in the service of deceiving others may plausibly work in contemporary social life, which is marked by a rather anonymous, ever socially shifting world. In the modern day, one can deceive and then move on to deceive other strangers. But what about a human evolutionary past in which people did not move on, but rather woke up each morning to deal with the same small group of individuals for most of their mortal lives?

Finally, VH&T's intuitive central assertion must be put to empirical test because there is an equally intuitive alternative. People may become more persuasive not because they deceive themselves of some illegitimate fact, but they instead decide that the facts just do not matter. They lay aside the truth and are unconcerned about it, making whatever claims they think will be the most convincing to the other person. This technique of simply not caring about the truth has been labeled by the philosopher Harry Frankfurt as bullshitting (Frankfurt Reference Frankfurt2005), and it potentially is a strategy that would make self-deception unnecessary under the authors' current framework. Thus, if evolution crafted the best liars among humans, perhaps it did so not via the route of self-deception but rather by creating individuals (or a species) who could dismiss any worry about the truth-value of what they were saying as an active consideration as they said it. Laboratory work could examine this. Are people better persuaders when they self-deceive themselves into some untruth? Or are they better persuaders when they strike any consideration of truth or falsity from their minds? Current social cognitive techniques seem tailor-made to tackle this question.

All these observations lead me to ask of the authors – or any interested reader: Get thee to a laboratory. The central contention guiding this essay is too broad and deep in its implications not to deserve direct and extensive empirical study.