Reading focus
Syllabus: How Can We Study Conscious Experience?
Compare explanatory targets as well as mechanisms, especially experience, access, and particular contents.
Why you’re reading it. Make the alternatives intelligible before evaluating the adversarial study; the two tested theories do not exhaust the field.
A question to consider. Choose one prediction shared by two theories and explain why confirming it would leave their disagreement unresolved.
Why this reading
Theories can disagree because they propose different mechanisms or because they explain different aspects of consciousness. Seth and Bayne compare four broad families and ask what would make their claims more precise and testable. The review offers a way to compare explanatory aims before counting supporting findings.
Before you read
An explanandum is the phenomenon a theory aims to explain. Higher-order theories concern representations of one's own mental states; global workspace theories concern information becoming widely available. Integrated information theory links experience to a system's intrinsic causal organization. Re-entry and predictive-processing approaches emphasize recurrent interactions, with substantial variation within each family.
Read for
- Separate predictions about the presence of experience, its particular contents, and cognitive access. Are competing accounts addressing the same question?
- Follow the connection from theoretical commitments to measures and anatomical predictions. Which observations are shared predictions rather than discriminating tests?
Consider afterwards
- Two theories both predict activity in a posterior region. What additional prediction could make their comparison informative?
- How could a useful consciousness measure remain compatible with several theories?
Editorial context
This 2022 review represents theories as developing families, not fixed exhaustive alternatives. Some are more computationally specified than others. Perturbational complexity is not identical to IIT's integrated-information quantity, and its usefulness does not uniquely validate IIT. Later adversarial testing provides an example of translating disputed frameworks into agreed predictions, while leaving their interpretation contested.