This commentary describes another variety of self-deception, highly relevant to von Hippel & Trivers's (VH&T's) project. Drawing on dual-process theories, I propose that conscious thinking is a voluntary activity motivated by metacognitive attitudes, and that our choice of reasoning strategies and premises may be biased by unconscious desires to self-deceive. Such biased reasoning could facilitate interpersonal deception, in line with VH&T's view.

Mercier and Sperber (M&S) suggest that human reasoning is reflective and has evolved to support social interaction. Cognitive agents benefit from being able to reflect on their beliefs whether they are acting alone or socially. A formal framework for argumentation that has emerged from research on artificial cognitive systems that parallels M&S's proposals may shed light on mental processes that underpin social interactions.

Understanding the perceived benefits of using drugs to achieve specific mental states will provide novel insights into the reasons individuals seek to use drugs. However, the precision of attempts to instrumentalize drugs is unclear both across drugs and individuals. Moreover, mis-instrumentalization, defined as discrepancies between such endpoints, may have relevance to understanding the relation among use, abuse, and addiction.

If Bayesian Fundamentalism existed, Jones & Love's (J&L's) arguments would provide a necessary corrective. But it does not. Bayesian cognitive science is deeply concerned with characterizing algorithms and representations, and, ultimately, implementations in neural circuits; it pays close attention to environmental structure and the constraints of behavioral data, when available; and it rigorously compares multiple models, both within and across papers. J&L's recommendation of Bayesian Enlightenment corresponds to past, present, and, we hope, future practice in Bayesian cognitive science.

This commentary does three things. First, it offers further support for the view that explicit reasoning evolved for public argumentation. Second, it suggests that promoting effective communication may not be the only, or even the main, function of public argumentation. Third, it argues that the data Mercier and Sperber (M&S) cite are compatible with the view that reasoning has subsequently been co-opted to play a role in individual cognition.

Elqayam & Evans' (E&E's) laudable program to keep the scientific investigation of human reasoning norm-free and focused on the facts alone is an essential part of a long tradition in the philosophy of science – but it faces deeper difficulties than the authors seem to realize, since reasoning is a competence, and the very concept of competence is normative.

We need not propose, as Carey does, a radical discontinuity between core cognition, which is responsible for abstract structure, and language and “Quinian bootstrapping,” which are responsible for learning and conceptual change. From a probabilistic models view, conceptual structure and learning reflect the same principles, and they are both in place from the beginning.

Self-deception may be the result of social manipulation and conflict management of social in-groups. Although self-deception certainly has offensive and defensive aspects, a full evolutionary understanding of this phenomenon is not possible until strategies of other parties are included into a model of self-perception and self-representation.

Müller & Schumann (M&S) deserve applause for their interdisciplinary examination of drug use, evolution, and learning. Further steps can deepen their evolutionary analysis: a focus on adaptive benefits, a distinction between approach and consummatory behaviors, an examination of how drugs can create adaptive lag through changing human niche construction, the importance of other neurobehavioral mechanisms in drug use besides instrumentalization, and the importance of sociocultural dynamics and neural plasticity in both human evolution and drug use.

I argue that framing the issue of motivated belief formation and its subsequent social gains in the language of self-deception raises logical difficulties. Two such difficulties are that (1) in trying to provide an evolutionary motive for viewing self-deception as a mechanism to facilitate other-deception, the ease and ubiquity of self-deception are undermined, and (2) because after one has successfully deceived oneself, what one communicates to others, though untrue, is not deceptive, we cannot say that self-deception evolved in order to facilitate the deception of others.

We argue that the authors’ call to integrate Bayesian models more strongly with algorithmic- and implementational-level models must go hand in hand with a call for a fully developed account of algorithmic rationality. Without such an account, the integration of levels would come at the expense of the explanatory benefit that rational models provide.

Though we are in broad agreement with much of Elqayam & Evans' (E&E's) position, we criticize two aspects of their argument. First, rejecting normativism is unlikely to yield the benefits that E&E seek. Second, their conception of rational norms is overly restrictive and, as a consequence, their arguments at most challenge a relatively restrictive version of normativism.

Carey focuses her theory on initial knowledge and Quinian bootstrapping. We reflect on developmental mechanisms, which can operate in between. Whereas most of the research aims at delimitating early cognitive mechanisms, we point at the need for studying their integration and mutual bootstrapping. We illustrate this call by referring to a current debate on infants' use of featural representations.

Mercier and Sperber (M&S) claim that the main function of reasoning is to generate support for conclusions derived unconsciously. An alternative account holds that reasoning has a deliberative function even though it is an internalized analogue of public discourse. We sketch this alternative and compare it with M&S's in the light of the empirical phenomena they discuss.

This commentary suggests that the general population has much less reasoning skill than is claimed by Mercier & Sperber (M&S). In particular, many studies suggest that the skills of understanding, evaluating, and producing arguments are generally poor in the population of people who have not had specific training.

The only viable formulation of perception, thinking, and action under uncertainty is statistical inference, and the normative way of statistical inference is Bayesian. No wonder, then, that even seemingly non-Bayesian computational frameworks in cognitive science ultimately draw their justification from Bayesian considerations, as enlightened theorists know fully well.

This commentary questions Elqayam & Evans' (E&E's) claims that thinking tasks are doomed to have multiple normative readings and that only applied research allows normative evaluations. In fact, some tasks have just one undisputed normative reading, and not only pathological gamblers but also normal individuals sometimes need normative guidance. To conclude, normative evaluations are inevitable in the investigation of human thinking.

The target article tends to conflate self-deception and self-enhancement, but biased self-presentation can be negative as well as positive. Self-deceiving self-diminishers may be depressed and non-self-deceiving self-diminishers may project false modesty. The article's otherwise brilliant argument for the advantages of self-deceptive self-enhancement for deceiving others may underemphasize the risks it entails for poor decision making.

Placeholders enable conceptual change only if presumed to be relevant (e.g., lead to the formation of true beliefs) even though their meaning is not yet fully understood and their cognitive function not yet specified. Humans are predisposed to make such presumptions in a communicative context. Specifying the role of the presumption of relevance in conceptual change would provide a more comprehensive account of Quinian bootstrapping.

The Müller & Schumann (M&S) view of drug use is courageous and compelling, with radical implications for drug policy and research. It implies that most nations prohibit most drugs that could promote happiness, social capital, and economic growth; that most individuals underuse rather than overuse drugs; and that behavioral scientists could use drugs more effectively in generating hypotheses and collaborating empathically.