Human belief comes in two forms, occupying different levels of mental organization. The basic mind consists of specialized subsystems shaped by evolution and learning, generating non-conscious partial beliefs that guide action directly. The supermind is a workspace for conscious, deliberative reasoning, operating with flat-out belief states that are voluntarily adopted, inferentially integrated, and typically language-involving.
Building on ideas from Ronald de Sousa, Daniel Dennett, and Jonathan Cohen, I argue that the folk-psychological term ‘belief’ refers to two distinct types of mental state with different properties and explanatory roles. This two-level account resolves puzzles that have troubled philosophers. It explains how we can simultaneously hold contradictory beliefs (different levels, different systems), why beliefs seem both passive (basic mind) and active (supermind), and clarifies the relationship between belief and action while vindicating folk-psychological commitments without risky claims about neural structure.
Basic beliefs and superbeliefs
Basic beliefs are non-conscious attitudes generated by specialized cognitive subsystems. They’re partial (representing features as likely to some degree rather than true or false), passive, and action-guiding. The basic mind is the biological, non-verbal level of mentality we share with other animals.
Superbeliefs (also called virtual beliefs) are conscious, flat-out commitments actively formed through what I call premising policies—resolutions to use certain propositions as premises in conscious reasoning. When you make up your mind about something, you’re not passively acquiring a new belief. You’re committing yourself to treat a proposition as a premise, work out what it entails, and regulate your thoughts accordingly.
These superbeliefs are not biologically given but actively constructed. They’re realized in basic-level attitudes—beliefs about one’s own cognitive commitments together with desires to honour those commitments. The supermind is thus a virtual system, dependent on the basic mind for its implementation but functionally distinct from it.
Premising policies and mental agency
This framework reveals belief as sometimes active rather than wholly passive. You can’t believe at will if that means forming a belief without evidence (strong voluntarism is necessarily false). But you can directly adopt beliefs as premises for pragmatic reasons when you have sufficient epistemic support (weak voluntarism). When the evidence leaves the matter open, practical considerations can legitimately tip the balance. This also helps explain non-doxastic acceptance—adopting propositions as premises for pragmatic or emotional reasons rather than epistemic ones, as may occur in cases of religious faith or ideological commitment.
The distinction between basic beliefs and superbeliefs also illuminates the nature of conscious thought and deliberation. Conscious reasoning involves actively manipulating superbeliefs—premises we’ve committed to—using language-involving processes. This connects directly to dual-process theories and the role of language in thought.
Applications: akrasia, self-deception, and delusion
The two-level framework helps explain puzzling phenomena. Akrasia (weakness of will) involves conflict between what you’ve decided to do (a supermind commitment) and what your basic mind is disposed to do. Self-deception involves accepting a proposition for comfort while non-consciously manipulating evidence to avoid acknowledging what you’ve done—believing something at one level while disbelieving it at another.
Delusions may be non-doxastic acceptances—premising commitments adopted for emotional reasons rather than epistemic ones, analogous to religious faith. This explains why delusional subjects can simultaneously maintain bizarre beliefs while acting normally in most contexts. Their explicit commitments (level 2) diverge from their implicit attitudes (level 1).
The framework also addresses implicit bias. We can explicitly endorse egalitarian principles while implicitly acting on prejudice because these attitudes occupy different levels with different formation mechanisms. Overriding implicit bias requires strong metacognitive motivation to monitor and correct automatic responses. See Self-knowledge for more on this.
Connections to consciousness
The layered picture fits naturally with illusionism about consciousness. If the supermind is a cognitive construction rather than a neural given, then the appearance of a unified phenomenal perspective may itself be part of that construction—a useful illusion rather than a metaphysical fact.
Key publications
Natural language and virtual belief (1998) — I argue that some beliefs—virtual beliefs—are actively formed by adopting premising policies, and that forming them typically requires language. Virtual beliefs are realized in basic beliefs about one’s cognitive commitments. Language provides the scaffolding for articulating, manipulating, and evaluating these commitments. In P. Carruthers and J. Boucher (eds.), Language and Thought. Cambridge University Press.
Mind and Supermind (2004) — Develops the two-level framework comprehensively, applying it to akrasia, self-deception, and first-person authority. Argues that avowals are performatives that make premising commitments, explaining first-person authority as performative self-guarantee. Shows how the framework resolves puzzles about weakness of will and self-deceit without partitioning strategies or abandoning rationality constraints. Cambridge University Press.
Deciding to believe again (2007) — I distinguish weak and strong voluntarism about belief. Strong voluntarism—forming beliefs at will without epistemic support—is necessarily false. But weak voluntarism—forming beliefs directly and for pragmatic reasons, provided one has epistemic support—remains defensible. I argue that actively formed beliefs are premising policies: resolutions to use certain propositions as premises in conscious reasoning. Mind 116: 523–47.
Partial belief and flat-out belief (2009) — Addresses the duality in folk psychology between partial beliefs (degrees of confidence) and flat-out beliefs (categorical beliefs). I review various attempts to reduce one to the other and argue none are attractive. Instead, I defend the premising view: flat-out beliefs are complex behavioural dispositions realized in the agent’s partial beliefs and desires. Flat-out belief is a disposition to take the believed proposition as a premise in conscious explicit reasoning and decision-making. This view resolves the Bayesian Challenge—showing how both partial and flat-out beliefs can guide action without conflict, since flat-out states are effective in virtue of the underlying partial ones. In F. Huber and C. Schmidt-Petri (eds.), Degrees of Belief. Springer, pp. 75–93.
Delusions, levels of belief, and non-doxastic acceptances (2011) — Applies the two-level framework to delusions. I argue that delusions are level 2 beliefs—conscious premising commitments—and suggest some may be non-doxastic acceptances adopted for emotional reasons rather than epistemic ones, analogous to religious faith. Explores how rationality constraints operate differently at each level. Neuroethics 4: 153–57.
Dual systems and dual attitudes (2012) — Defends the compatibility of the action-based view of System 2 with dual-attitude theory. System 2 attitudes are virtual attitudes, partially realized in System 1 attitudes. When we make up our minds or change them, we form higher-order beliefs about our cognitive commitments, which then regulate our explicit reasoning. Mind & Society 11: 41–51.
Playing double: Implicit bias, dual levels, and self-control (2015) — Develops the dual-level framework to explain implicit bias as involving implicit beliefs that conflict with explicit commitments. Argues that override of implicit bias requires strong metacognitive motivation. In M. Brownstein and J. Saul (eds.), Implicit Bias and Philosophy, Volume I. Oxford University Press.
Related topics
Dual-process theories—The relation between System 1/System 2 and basic mind/supermind
Language and thought—How language enables the supermind
Self-knowledge—Performative authority vs interpretive access
Illusionism—How the two-level view connects to consciousness
