Mysteries of Visual Experience

Abstract: Science is a crowning glory of the human spirit and its applications remain our best hope for social progress. But there are limitations to current science and perhaps to any science. The general mind-body problem is known to be intractable and currently mysterious. This is one of many deep problems that are universally agreed to be beyond the current purview of Science, including quantum phenomena, etc. But all of these famous unsolved problems are either remote from everyday experience (entanglement, dark matter) or are hard to even define sharply (phenomenology, consciousness, etc.). An updated summary of this work has been published as: Feldman, J. (2022). Computation, perception, and mind. Behavioral and Brain Sciences, 45, E48. doi:10.1017/S0140525X21001886 A more readable, open access, version is: https://escholarship.org/uc/item/6cs78450

• The paper challenges prevailing visual perception models by revealing that neural architectures lack sufficient computational resources to produce our rich, detailed experience.
• It quantitatively demonstrates that the brain’s limited neural capacity cannot support continuous, high-resolution vision, particularly in the peripheral field.
• The analysis urges a reassessment of mind-brain theories and encourages innovative research frameworks in neuroscience and AI.

Analysis of "Mysteries of Visual Experience" by Jerome Feldman

Jerome Feldman's paper, "Mysteries of Visual Experience," presents a critical examination of the current understanding of visual perception and its significant inconsistencies with established theories in cognitive neuroscience. Contrary to the prevalent notion that the brain fully accounts for our subjective experience of a detailed and stable visual world through well-understood neural mechanisms, Feldman posits that current scientific explanations fall short, underscoring potential gaps in our comprehension of the mind-brain relationship.

Feldman addresses two major phenomena that highlight these inconsistencies: the neural binding problem and the experience of perceiving a stable visual world. He argues that the detailed, continuous perception of our visual environment is not supported by current neural processing models, given the limited number of neurons available for complex visual computation. Feldman critiques the prevalent theories, underscoring the limitations in neural computation capability to account for continuous, high-resolution vision across the visual field.

Key Insights and Results

1. Experience vs. Neural Capability: Feldman provides a quantitative argument regarding the brain's perceptual accuracy, particularly around the fovea and peripheral vision. He highlights that the neural resources required to support the high level of detail seen in peripheral vision do not exist, suggesting that the brain's architecture, as currently understood, cannot fully account for the richness of visual experience.
2. Visual Stability and Binding: The process by which the brain maintains a stable and cohesive visual experience despite frequent saccades remains unexplained. Feldman argues that the phenomena of binding separate neural computations into a unified perceptual experience challenges current neural theory.
3. Implications on Mind-Brain theories: Feldman challenges the dominant view that mental processes are reducible to neural activity. He calls into question the sufficiency of neural spike-based communication for explaining consciousness, advocating for the consideration of alternative theoretical frameworks.
4. Computational Constraints: By examining the spatial and temporal capabilities of neural processes, Feldman effectively demonstrates that the computational requirements for subjective visual detail surpass what is theoretically feasible within known brain structures.

Theoretical and Practical Implications

Feldman's analysis calls for a reassessment of the theories correlating neural activity and subjective experience, potentially guiding future research toward unexplored areas within neurobiology, such as re-evaluating traditional models like the binding problem and considering new frameworks that integrate subjective experiences. The paradox of a detailed stable visual experience, juxtaposed with a lack of corresponding neural representation, suggests that alternative models, perhaps involving broader systems of information processing or other elements like the glia, might be required.

Future Directions in AI and Cognitive Science

The unresolved issues outlined by Feldman open avenues for innovation in both AI and neuroscientific research. Understanding the limitations of current models in explaining visual perception can inspire advancements in artificial vision systems, aiming to mimic or even surpass biological capabilities. Furthermore, Feldman's work encourages a more integrated approach between cognitive science and AI, leveraging insights from computational models to hypothesize potential biological mechanisms that could resolve the identified inconsistencies.

In conclusion, Jerome Feldman's paper serves as a catalyst for further exploration into the enigmatic areas of visual perception, emphasizing the need to reevaluate existing models and potentially redefine the boundaries of cognitive neuroscience. While the paper does not resolve the mystery of subjective visual experiences, it significantly contributes to the ongoing discourse by identifying crucial gaps and paving the way for deeper inquiry into the neural correlates of consciousness.