Structuralist theories in consciousness science focus on the structural properties of phenomenal experience and seek to identify their neural encoding via similarities between quality spaces and neural state spaces. This paper argues that structuralism can be adopted as a methodological approach partly independent of philosophical assumptions about the nature of consciousness. The author situates methodological structuralism within questions of mental representation, psychophysical measurement, holism, and functional relevance of neural processes, and analyzes its relationship to the distinction between conscious and unconscious states. The goal is to make structuralist methodology more accessible to a broader scientific and philosophical audience.
Methodological structuralism aims to identify neural correlates of consciousness by mapping relationships among conscious experiences onto relationships among neural population activity. This paper argues that while structuralism effectively describes how content is encoded in the brain, it cannot fully explain why content is consciously experienced. Most current theories of consciousness propose an additional mechanism responsible for conscious experience. If structuralism is correct, progress requires studying interactions between neural mechanisms of consciousness and neural structures encoding content. This has implications for consciousness in artificial intelligence, and the paper discusses empirical approaches using advanced neuroscientific methods to investigate these interactions.
Distinguishing neural correlates that are genuinely tied to conscious experience from those that are merely coincidental requires a principled approach. The authors propose neurophenomenal structuralism, which relies on the fact that phenomenal experiences form structures of similarity and difference. By using these structures to individuate experiences in neuroscience and psychophysics, they derive a structural similarity constraint: a neural activation must preserve the structure of the associated phenomenal domain to qualify as a neural correlate of consciousness proper. An example from Brouwer and Heeger illustrates the constraint. This approach aims to refine the search for neural correlates by filtering out merely statistical associations.