Dienes & Perner (D&P) argue that nondeclarative knowledge can take multiple forms. We provide empirical support for this from two related lines of research about the development of mathematical reasoning. We then describe how different forms of procedural and declarative knowledge can be effectively modeled in Anderson's act-r theory, contrasting this computational approach with D&P's logical approach. The computational approach suggests that the commonly observed developmental progression from more implicit to more explicit knowledge can be viewed as a consequence of accumulating and strengthening mental representations.

The scenario used by Dienes & Perner to show that individual representation can be implicit when property representation is explicit can be adapted to show that property representation can be implicit when individual representation is explicit. So there is no hierarchy of explicitness, contrary to their claim. There is a reading of the “implicit/explicit” distinction that does appear to exhibit an asymmetry parallel to that alleged to hold between individual and property. But this is not a distinction Dienes & Perner mention, nor is it one that could be easily incorporated into their framework.

A comprehensive theory of implicit and explicit knowledge must explain phenomenal knowledge (e.g., knowledge regarding one's affective and motivational states), as well as propositional (i.e., “fact”-based) knowledge. Findings from several research areas (i.e., the subliminal mere exposure effect, artificial grammar learning, implicit and self-attributed dependency needs) are used to illustrate the importance of both phenomenal and propositional knowledge for a unified theory of implicit and explicit mental states.

Dienes & Perner argue that volitional control in artificial grammar learning is best understood in terms of the distinction between implicit and explicit knowledge representations. We maintain that direct, explicit access to knowledge organised in a hierarchy of implicitness/explicitness is neither necessary nor sufficient to explain volitional control. People can invoke volitional control when their knowledge is implicit, as in the case of artificial grammar learning, and they can invoke volitional control when any part of their knowledge representation is implicit, as can be seen by examining “feeling of knowing” phenomena.

Dienes & Perner's target article contains numerous but unsystematic references to the implicit or explicit knowledge of the temporal context of a known state of affairs such as may constitute the content of a propositional attitude. In this commentary, the forms of cognition that, according to D&P, require only implicit knowledge of time are contrasted with those for which explicit temporal knowledge is needed. It is suggested that the explicit representation of time may have been important in human evolution and that certain developmental disorders including autism may be (partly) caused by defective ability to represent time.

The visual system captures a unique contrast between implicit and explicit representation where the same event (location of a visible object) is coded in both ways in parallel. A method of differentiating the two representations is described using an illusion that affects only the explicit representation. Consistent with predictions, implicit information is available only from targets presently visible, but, surprisingly, a two-alternative decision does not disturb the implicit representation.

The notion of nonconceptual content in Dienes & Perner's theory is examined. A subject may be in a state with nonconceptual content without having the concepts that would be used to describe the state. Nonconceptual content does not seem to be a clear-cut case of either implicit or explicit knowledge. It underlies a kind of practical knowledge, which is not reducible to procedural knowledge, but is accessible to the subject and under voluntary control.

Dienes & Perner's target article constitutes a significant advance in thinking about implicit knowledge. However, it largely neglects processing details and thus the time scale of mental states realizing propositional attitudes. Considering real-time processing raises questions about the possible brevity of implicit representation, the nature of processes that generate explicit knowledge, and the points of view from which knowledge may be represented. Understanding the propositional attitude analysis in terms of momentary mental states points the way toward answering these questions.

Dienes & Perner argue that there is a hierarchy of forms of implicit knowledge. One level of their hierarchy involves factuality, where it may be merely implicit that the state of affairs is supposed to be a real one rather than something imagined or fictional. I argue that the factual or fictional status of a thought or utterance cannot be a matter of concept, implicit or explicit.

Dienes & Perner (D&P) link explicit knowledge of facts to predication. But predication is basically a linguistic notion. Their approach therefore makes it difficult to attribute knowledge of facts to non- language-users, such as animals. The explicit/implicit distinction, as D&P formulate it, is accordingly of little use for exploring the cognitive capacities of nonhuman primates – despite the increasing evidence for sophisticated social awareness among apes, implying mental representations of events in which participants are clearly distinguished. A revised formulation, less biased toward syntax as it happens to have evolved in humans, could avoid this drawback.

Dienes' & Perner's proposals are discussed in relation to the distinction between explicit and implicit systems of thinking. Evans and Over (1996) propose that explicit processing resources are required for hypothetical thinking, in which mental models of possible world states are constructed. Such thinking requires representations in which the individuals' propositional attitudes including relevant beliefs and goals are made fully explicit.

Dienes & Perner make three mistakes in their account of the “natural language meaning” of implicit-explicit knowledge: They fail to take the multiplicity of use of a concept seriously enough, they arbitrarily separate use of a concept and its conceptual structure, and they tend to tailor their analysis for use by the Representational Theory of Knowledge.

Dienes & Perner's (D&P's) target articles proposes an analysis of explicit knowledge based on a progressive transformation of implicit into explicit products, applying this gradient to different aspects of knowledge that can be represented. The goal is to integrate a philosophical concept of knowledge with relevant psychophysical and neuropsychological data. D&P seem to fill an impressive portion of the gap between these two areas. We focus on two examples where a full synthesis of theoretical and empirical data seems difficult to establish and would require further refinement of the model: action representation and the closely related consciousness of action, which is in turn related to self-consciousness.

Gesture does not have a fixed position in the Dienes & Perner framework. Its status depends on the way knowledge is expressed. Knowledge reflected in gesture can be fully implicit (neither factuality nor predication is explicit) if the goal is simply to move a pointing hand to a target. Knowledge reflected in gesture can be explicit (both factuality and predication are explicit) if the goal is to indicate an object. However, gesture is not restricted to these two extreme positions. When gestures are unconscious accompaniments to speech and represent information that is distinct from speech, the knowledge they convey is factuality-implicit but predication-explicit.

Dienes & Perner's theoretical framework should be applicable to two related areas: technological innovation and the psychology of scientific reasoning. For the former, this commentary focuses on the example of nuclear weapon design, and on the decision to launch the space shuttle Challenger. For the latter, this commentary focuses on Klayman and Ha's positive test heuristic and the invention of the telephone.

The propositional account of explicit and implicit knowledge interprets cognitive differences between direct and indirect test performance as emerging from the elements in different hierarchical levels of the propositional representation that have been made explicit. The hierarchical nature of explicitness is challenged, however, on the basis of neuropsychological dissociations between direct and indirect tests of memory, as well as the stochastic independence that has been observed between these two types of tests. Furthermore, format specificity on indirect test of memory challenges the basic notion of a propositional theory of implicit and explicit knowledge.

Three forms of implicit knowledge are presented (functional, structural, and procedural). These forms differ in the way they are made explicit and hence in how they are represented by the individual. We suggest that the framework presented by Dienes & Perner does not account for these differences.

In Dienes & Perner's analysis, implicitly represented knowledge differs from explicitly represented knowledge only in the attribution of properties to specific events and to self-awareness of the knower. This commentary questions whether implicit knowledge should be thought of as being represented in the same conceptual vocabulary; rather, it may involve a quite different form of representation.

Dienes & Perner's target article is not a satisfactory theory of implicit knowledge because in endorsing the representational theory of knowledge, the authors also inadvertently accept that only explicit knowledge can be causally efficacious, and hence that implicit knowledge is an inert category. This conflation between causal efficacy, knowledge, and explicitness is made clear through the authors' strategy, which consists of attributing any observable effect to the existence of representations that are as minimally explicit as needed to account for behavior. In contrast, we believe that causally efficacious and fully implicit knowledge exists, and is best embodied in frameworks that depart radically from classical assumptions.

Dienes & Perner's analysis provides a clear theoretical justification for using a demonstration of volitional control as a criterion for conscious awareness. However, in memory tasks, the converse does not hold: A phenomenological awareness of a memory episode can arise involuntarily, even when the task does not require retrieval of the episode. The varying amounts of volitional retrieval required by different memory tasks need to be recognized.