Notwithstanding many similarities between Thelen et al.'s and Piaget's accounts of the A-not-B error, we argue that, in contrast to Piaget, they do not explicitly address the issue of objectivity. We suggest that this omission is partly due to the fact that Thelen et al. and Piaget's accounts are pitched at different levels of explanation.

In this commentary, we question (1) how embodied Thelen et al.'s model is relative to their aims, and (2) how embodied the behavior of children is in particular response systems, relative to how much dynamic systems theory emphasizes this idea. We close with corrections to mischaracterizations of an alternative, neural network perspective on infant behavior.

Thelen et al. criticize “spatial coding” approaches to the A-not-B error. However, newer thinking about spatial coding provides more precise analytic categories and recognizes that different spatial coding systems normally coexist. Theorizing about spatial coding is largely compatible with dynamic-systems theory, augmenting it with an analysis of what one means when discussing “location at A” (or B).

By delineating the parametric variations that affect infant performance in the standard A-not-B search task, the Thelen et al. model provides an important contribution to the field of infant development. We discuss several broad issues pertinent to interpreting the model. We note that the phenomenon modeled by Thelen et al. is not necessarily the one originally described by Piaget. We describe data on infant self-correction that are not addressed by the Thelen et al. model. Finally, we suggest that psychological constructs such as representation and knowledge structures are valuable to our understanding of the A-not-B phenomenon in particular and psychological development in general.

Thelen et al. present a convincing explanation of the A-not-B error, but contrary to their own claims, their explanation essentially involves mental representations. As is too common among cognitive scientists, they equate mental representations with representations of external physical objects. They clearly show, however, that representations of bodily actions on physical objects are central to the dynamical system producing the error.

The field theoretic approach of the target article is simplified by setting the parameters of the dynamical field equation so that the system is near the critical point between cooperative and non-cooperative dynamics. However, embodiment of cognitive development would require a closer connection between the dynamical field interactions and the physiology of the cerebral cortex.

In this commentary, I raise several points. First, I argue that non-search tasks show that the A-not-B task is about object representation, even if perseveration can occur without objects. Second, I provide an alternative interpretation for the finding that changing body posture reduces A-not-B errors. Third, I provide an alternative interpretation for the finding of convergence in reaching behavior in two-target tasks. Fourth, I suggest attention deficits can explain the A-not-B error on their own with no necessity for motor memories.

Thelen et al.'s account of cognition as the dynamic interaction of processes of perceiving, reaching, and remembering within a movement planning field is a useful articulation of the Piagetian concept of sensorimotor cognition. The claim that the same kind of analysis applies to all kinds of cognition at all ages, however, is questioned in light of the distinction between sensorimotor and symbolic cognition.

Psychologists have long debated the underlying cause of infants' perseverative reaching. Thelen et al. explain the error in terms of general processes that make goal-directed actions to remembered locations. The context- and experience-dependent nature of their model implies that there is no single cause of the A-not-B error, and, more generally, no core essence to cognitive development.

The “A-not-B” error is consistent with an old memory principle, Jost's Law. Quantitative properties of the effect can be explained by a dynamic model for habituation that is also consistent with Jost. Piaget was well aware of the resemblance between adult memory errors and the “A-not-B” effect and, contrary to their assertions, Thelen et al.'s analysis of the object concept is much the same as his, though couched in different language.

I present my observations of combinatory manipulations by three infant chimpanzees in a series of test tasks. Common characteristics of motor patterns were observed across the tasks between both infant chimpanzees and 1-year-old infants. Based on the results, I point out that comparative approach can illuminate Thelen et al.'s arguments.

Thelen et al. offer an inspiring behavior-based theory of a long-standing cognitive problem. They demonstrate how joining traditions, old (the Gestaltist field theory) and new (dynamical systems theory) may open up the path towards embodied cognition. We discuss possible next steps. Self-organization theory (synergetics) could be used to address the formation of gaze/reach attractors and their optimality, given environmental control parameters. Finally, some clinical applications of the field model are advocated.

Reaching in the A-not-B situation is not the product of a single person, but rather of a person-person system. We argue that models of embodied cognition distributed over persons may be necessary to capture the essential qualities of evolving behaviors, even those as simple asperseverative reaching.

We discuss the assumptions put forward by Thelen et al. about motor planning processes. We examine the results of an experiment inspired by the authors' contention that the motor plans of both infants and adults are continuous and graded. We wondered whether adults, in an adapted version of the A-not-B error paradigm, would point between the A and B targets, as in some degraded conditions of pointing (Ghez et al. 1990), or would make A-not-B errors. Unexpectedly, we observed that adults tended to shift the direction of their pointings to B away from A, and did not make any A-not-B errors.

We argue that mentalistic constructs like the “object concept” are not substitutes for process explanations of cognition, and that it is impossible to prove the existence of such constructs with behavioral tasks. We defend the field theory as an appropriate level for modeling embodiment. Finally, we discuss the model's biological plausibility and its extensions to other tasks and other species.

Miller (1956) summarized evidence that people can remember about seven chunks in short-term memory (STM) tasks. However, that number was meant more as a rough estimate and a rhetorical device than as a real capacity limit. Others have since suggested that there is a more precise capacity limit, but that it is only three to five chunks. The present target article brings together a wide variety of data on capacity limits suggesting that the smaller capacity limit is real. Capacity limits will be useful in analyses of information processing only if the boundary conditions for observing them can be carefully described. Four basic conditions in which chunks can be identified and capacity limits can accordingly be observed are: (1) when information overload limits chunks to individual stimulus items, (2) when other steps are taken specifically to block the recoding of stimulus items into larger chunks, (3) in performance discontinuities caused by the capacity limit, and (4) in various indirect effects of the capacity limit. Under these conditions, rehearsal and long-term memory cannot be used to combine stimulus items into chunks of an unknown size; nor can storage mechanisms that are not capacity-limited, such as sensory memory, allow the capacity-limited storage mechanism to be refilled during recall. A single, central capacity limit averaging about four chunks is implicated along with other, noncapacity-limited sources. The pure STM capacity limit expressed in chunks is distinguished from compound STM limits obtained when the number of separately held chunks is unclear. Reasons why pure capacity estimates fall within a narrow range are discussed and a capacity limit for the focus of attention is proposed.

The empirical data do not unequivocally support a consistent fixed capacity of four chunks. We propose an alternative account whereby capacity is limited by the precision of specifying the temporal and spatial context in which items appear, that similar psychophysical constraints limit number estimation, and that short term memory (STM) is continuous with long term memory (LTM).

The limited capacity of immediate memory “rides” on the even more limited capacity of consciousness, which reflects the dynamic activity of the thalamocortical core of the brain. Recent views of the conscious narrow-capacity component of the brain are explored with reference to global workspace theory (Baars 1988; 1993; 1998). The radical limits of immediate memory must be explained in terms of biocognitive brain architecture.

According to span theory, a behavioral theory of the magical numbers, Cowan's 4 and Miller's 7 are simply two different points on the same ogive describing the relation between performance and span load, a fundamental task characteristic. Span theory explains the magical numbers in terms of a unitary limited span ability, a mathematical abstraction from that ogive.

The data presented in the target article make a persuasive case for the notion that there is a fundamental limit on short term memory (STM) of about four items. Two possible means of further testing this claim are suggested and data regarding scene coherence and memory capacity for ordered information are reviewed.