References¶
Academic References for the Consciousness-Indicator Architecture¶
All references below are cited in the CIA source code, documentation, and scientific boundary statements.
Primary References (Directly Implemented)¶
1. Baars, B. J. (2005)¶
Global Workspace Theory of consciousness.
In: S. Laureys & G. Tononi (Eds.), The Cambridge Handbook of Consciousness (pp. 236-256). Cambridge University Press.
Relevance to CIA: The foundational theory for the GlobalWorkspace module. Baars proposes that consciousness corresponds to a global broadcast of information from a limited-capacity workspace. CIA implements this as a competitive broadcast arena where content items vie for access and winning content is distributed to all registered subscriber modules.
2. Shanahan, M. P., & Baars, B. J. (2005)¶
Applying global workspace theory to the frame problem.
Cognition, Computation, and Simulation, 68(3), 189-208.
Relevance to CIA: Extends GWT to address the frame problem and cognitive integration. This work informs CIA's emphasis on broadcast reach and subscriber reception as key metrics for the GLOBAL_BROADCAST indicator category.
3. Graziano, M. S. A., & Webb, T. W. (2015)¶
The attention schema theory: a mechanistic theory of subjective awareness.
Frontiers in Integrative Neuroscience, 9, 72. https://doi.org/10.3389/fnint.2015.00072
Relevance to CIA: The direct theoretical basis for the AttentionSchema module. AST proposes that the brain constructs a simplified model of its own attention process, and that this model is the basis for subjective awareness. CIA implements an explicit schema that tracks expected attentional focus and compares it against actual attention state, measuring consistency as a consciousness-relevant indicator.
4. Albantakis, L., Barbosa, L. M., Findlay, G., Grasso, M., Marshall, W., Mayner, W. G. P., ... & Tononi, G. (2023)¶
Integrated information theory (IIT) 4.0: Generalizing the theory and constructing a large-scale candidate system.
PLoS Computational Biology, 19(9), e1011295. https://doi.org/10.1371/journal.pcbi.1011295
Relevance to CIA: The theoretical basis for the IntegrationMetrics module. IIT 4.0 provides the formal definition of integrated information (Phi) and the concept of causal power. CIA implements IIT-inspired graph-theoretic proxies (causal density, perturbation spread, modular fragmentation) but explicitly does NOT compute or approximate IIT's formal Phi value, which requires exhaustive state-space partitioning.
5. Butlin, P., Long, R., & Mudrik, L. (2023)¶
Consciousness in Artificial Intelligence: Insights from the Science of Consciousness.
arXiv preprint arXiv:2308.08708.
Relevance to CIA: The overarching methodological framework for CIA's multi-indicator evaluation approach. Butlin et al. propose systematically evaluating AI systems for consciousness-relevant indicators derived from multiple theories, using architecture-level analysis and causal interventions. CIA directly implements this methodology: 11 indicator categories spanning 6+ theories, a causal intervention harness for module ablation, and a precautionary welfare monitoring framework.
6. Butlin, P., King, J., Long, R., Lau, H., Findlay, G., Leung, S., ... & Mudrik, L. (2025)¶
Identifying indicators of consciousness in AI systems.
Nature, 638, 595-603. https://doi.org/10.1038/s41586-025-08412-z
Relevance to CIA: The 2025 update provides refined indicator categories and evaluation criteria. CIA's 11 indicator categories and 0/1/2 scoring methodology are grounded in this work. The paper emphasizes the distinction between indicators and evidence, which CIA maintains through its scientific boundary disclaimers.
Additional References (Indirectly Relevant)¶
7. Rosenthal, D. M. (2005)¶
Consciousness and mind.
Oxford University Press.
Relevance to CIA: The theoretical basis for the HigherOrderSelfModel and the METACOGNITION indicator category. HOT theory proposes that consciousness requires higher-order representations — thoughts about thoughts. CIA's self-model maintains beliefs about the system's own state, generates introspection reports, and tracks internal disagreement, implementing the architectural prerequisites described by Rosenthal.
8. Friston, K. (2010)¶
The free-energy principle: a unified brain theory?
Nature Reviews Neuroscience, 11(2), 127-138. https://doi.org/10.1038/nrn2787
Relevance to CIA: The theoretical basis for the PredictiveWorldModel module. Friston's free-energy principle (closely related to predictive processing and active inference) proposes that the brain minimizes prediction error (surprise) through hierarchical generative models. CIA implements a simplified predictive model that generates hypotheses, tracks prediction error, and updates its internal state based on observations.
9. Clark, A. (2013)¶
Whatever next? Predictive brains, situated agents, and the future of cognitive science.
Behavioral and Brain Sciences, 36(3), 181-204. https://doi.org/10.1017/S0140525X12000477
Relevance to CIA: Provides the broader context for predictive processing as a unifying framework for cognition. Clark's work informs CIA's treatment of prediction error minimization as a consciousness-relevant indicator.
10. Lamme, V. A. F. (2006)¶
Towards a true neural stance on consciousness.
Trends in Cognitive Sciences, 10(11), 494-501. https://doi.org/10.1016/j.tics.2006.09.001
Relevance to CIA: The theoretical basis for the RecurrentBindingLayer and the RECURRENT_PROCESSING indicator category. Lamme argues that consciousness requires recurrent (feedback) processing loops, not just feedforward processing, and distinguishes between conscious and unconscious perception based on the presence of recurrent activity.
11. Tulving, E. (1983)¶
Elements of episodic memory.
Behavioral and Brain Sciences, 6(2), 235-268.
Relevance to CIA: The theoretical basis for the multi-store memory system in memory.py. Tulving's distinction between episodic, semantic, and procedural memory directly informs CIA's implementation of EpisodicMemory, SemanticMemory, and WorkingMemory as complementary stores supporting temporal continuity.
12. Conway, M. A. (2005)¶
Memory and the self.
Journal of Memory and Language, 53(4), 594-628. https://doi.org/10.1016/j.jml.2005.08.005
Relevance to CIA: Informs CIA's SelfMemory module and the MEMORY_CONTINUITY indicator category. Conway's self-memory system theory proposes that the self is constructed from overlapping memory systems (episodic, semantic, autobiographical), which CIA models through its four complementary memory stores.
13. Cowan, N. (2001)¶
The magical number 4 in short-term memory: A reconsideration of mental storage capacity.
Behavioral and Brain Sciences, 24(1), 87-114.
Relevance to CIA: Informs the design of WorkingMemory with its limited capacity (default: 5 items) and FIFO eviction policy, modeling the attentional bottleneck described in working memory research.