Abstract

Is experience stored in the brain? The universal assumption is that it is, yet the answer to this question is critical to the theory of consciousness. If “yes,” it must be understood that this answer absolutely constrains all theories of the origin and nature of consciousness. Memory images, dreams, even perceptual images and perceptual experience must somehow be generated from stored elements within the neural substrate. If the answer is “no,” Searle’s principle of “neurobiological sufficiency,” as one example, carries no force, while on the other hand, a theory of direct perception can be construed to actually require this “no” answer, but then also requires a theory of memory to support this. Yet a model of memory which does not require storage of our experience within the brain has not been attempted. An initial review is undertaken of Gibson’s (1969, 1979) ecological perception, his notion of the “direct specification” of the external perceptual image, and subsequent findings of the ecological research paradigm. Central to this model is the notion of invariance laws defining external events, for example, the mathematical laws defining texture gradients, velocity flow fields, tau ratios on these fields, inertia tensors, etc. This invariance law information drives the brain’s “specification” of the environmental image/event. Yet, problematically, the specification is necessarily always to the past, i.e., to a past form of the transformation of the environmental field. Gibson’s model has a high affinity with a model developed by Henri Bergson (1896). The model presciently assumed holographic properties in the matter-field and broke from the classical model of space and time, arguing that time-evolution or motion of the field must be treated as indivisible or, as physics is now arguing, as non-differentiable. In this framework, the brain can be viewed as supporting a modulated reconstructive wave passing through a holographic matter-field and specifying an image of the past motion of the field. In such a model, experience cannot be stored within the brain for it is not occurring solely within the brain. Yet the model contains a basic mechanism for redintegration or retrieval of experienced events. The research results in cued recall are reviewed in several areas, to include verbal learning experiments, the effect of imagery in paired-associate tasks, priming phenomena, and the very ecologically oriented subject performed tasks (SPTs). The mathematical description of events via invariance allows a precise definition of the similarity of the cue-event and allows for parametric variation of the power of the cue, thus biasing recall or recognition performance. The model is predictive, and at the minimum, holds its own with current memory models. A short survey of current models reveals that most are very vague as to what events are stored, and what “elements” of these events are stored, and even vaguer on how an experience is reconstructed. Since these models ignore the actual structure of events via invariance laws, no current models actually explain the retrieval of events per se. The proposed model, however, does not require the storage of experience within the brain, in fact holding that time-extended events, for example, velocity flow fields, cannot be so stored. Some larger implications are explored: the in principle possibility of retrieval of all experience, the operation of free recall, the nature of cognition in the model, and potential distinguishing experiments relative to current models.