Qualia Narrative Theory: From Social Modeling to the Hard Problem’s Soft Solution Alan D. Varvil August 09, 2025 — LessWrong

1

Qualia Narrative Theory: From Social Modeling to the Hard Problem’s Soft Solution Alan D. Varvil August 09, 2025

9th Aug 2025

5 min read

1

This post was rejected for the following reason(s):

No LLM generated, heavily assisted/co-written, or otherwise reliant work. Our system flagged your post as probably-written-by-LLM. We've been having a wave of LLM written or co-written work that doesn't meet our quality standards. LessWrong has fairly specific standards, and your first LessWrong post is sort of like the application to a college. It should be optimized for demonstrating that you can think clearly without AI assistance.

So, we reject all LLM generated posts from new users. We also reject work that falls into some categories that are difficult to evaluate that typically turn out to not make much sense, which LLMs frequently steer people toward.*

"English is my second language, I'm using this to translate"

If English is your second language and you were using LLMs to help you translate, try writing the post yourself in your native language and using a different (preferably non-LLM) translation software to translate it directly.

"What if I think this was a mistake?"

For users who get flagged as potentially LLM but think it was a mistake, if all 3 of the following criteria are true, you can message us on Intercom or at team@lesswrong.com and ask for reconsideration.

you wrote this yourself (not using LLMs to help you write it)
you did not chat extensively with LLMs to help you generate the ideas. (using it briefly the way you'd use a search engine is fine. But, if you're treating it more like a coauthor or test subject, we will not reconsider your post)
your post is not about AI consciousness/recursion/emergence, or novel interpretations of physics.

If any of those are false, sorry, we will not accept your post.

* (examples of work we don't evaluate because it's too time costly: case studies of LLM sentience, emergence, recursion, novel physics interpretations, or AI alignment strategies that you developed in tandem with an AI coauthor – AIs may seem quite smart but they aren't actually a good judge of the quality of novel ideas.)

1

New Comment

More from AlanV

Curated and popular this week

Abstract.

The Qualia Narrative Theory (QNT) reframes the hard problem of consciousness as a
narrative feedback problem, proposing that the vividness of subjective experience—what we call qualia—emerges from an evolved social modeling loop that prepares internal perceptual content for potential communication. QNT integrates insights from philosophy (e.g., Nagel, 1974; Chalmers, 1995; Baars, 1997), psychology and neuroscience (Frith & Frith, 2006; Lau & Rosenthal, 2011), anthropology and culture (Nisbett & Miyamoto, 2005; Tomasello, 2008), and artificial intelligence (Shumailov et al., 2024) to offer a functionalist, testable account. The five-stage model predicts that
qualia intensity is modulated by social context, cultural norms, and narrative exchange. Evidence includes developmental and cross-cultural studies, isolation effects (Suedfeld, 2018), and AI degradation under recursive training ("MAD").

1. Introduction.

The “hard problem” asks why information processing feels like anything from the inside (Nagel, 1974; Chalmers, 1995). Functional frameworks such as Global Workspace Theory describe access and control structures (Baars, 1997), but not the felt quality itself. QNT proposes that qualia are not intrinsic properties of perception; rather, they are emergent artifacts of a socially oriented communication pipeline that prepares internal states for potential sharing. This loop is recursive and predictive, leveraging mentalizing circuits implicated in reasoning about other minds (Frith & Frith, 2006) and is compatible with higher-order awareness accounts (Lau & Rosenthal,2011).
In humans, prolonged social isolation degrades experiential richness, producing dissociation and even hallucinations in a substantial subset (Suedfeld, 2018). In AI, training or prompting systems primarily on their own outputs leads to repetition and loss of coherence—a failure mode termed Model Autoregressive Degradation (MAD; Shumailov et al., 2024). QNT treats these as parallel manifestations of loop collapse without external anchoring.

2. Core Proposition.

Qualia emerge from a five-stage social modeling process:
1) Perception – Sensory data are processed into stable representations.
2) Internal Modeling – Representations are embedded within a personal worldview.
3) Social Simulation – Internal states are simulated as if for communication to others (theory-of-mind mechanisms; Frith & Frith, 2006; Schurz et al., 2014).
4) Narrative Encoding – Perceptual content is structured into shareable form (story-comprehension networks; Mar, 2011).
5) Feedback Integration – Actual or imagined feedback modifies future processing, maintaining the loop. Isolation or monotonic environments weaken stages 3–5, reducing qualia vividness.

3. Human Evidence

3.1 Developmental. Joint attention and gaze cueing scaffold early social simulation (Frischen,Bayliss, & Tipper, 2007). Infants use social referencing to interpret ambiguity (Campos & Stenberg,1981), foreshadowing the communication-first framing of qualia.
3.2 Cultural Modulation. Cultural frameworks bias what is encoded for sharing—holistic vs. analytic perception alters the qualitative profile of what “stands out” (Nisbett & Miyamoto, 2005). Language,pedagogy, and coordinated action practices provide a natural pathway for such biases to shape the loop (Tomasello, 2008).
3.3 Social Modulation of Intensity. Anticipated support or exclusion changes reported intensity of experience; e.g., social contact buffers threat responses (Coan, Schaefer, & Davidson, 2006) and exclusion recruits pain circuitry (Eisenberger, Lieberman, & Williams, 2003), consistent with QNT’s prediction that the loop tunes vividness.
3.4 Isolation and Deprivation. Reviews of solitary confinement and extreme environments report dissociation, derealization, and perceptual anomalies; hallucinations occur in a notable minority, while broader distress is widespread (Suedfeld, 2018). QNT explains these as consequences of a weakened social-model loop.
4. AI Analogy and MAD. Under recursive self-training or self-prompting, LLMs drift toward repetition, mode collapse, and loss of tail behavior (Shumailov et al., 2024). The structural homology to human isolation is the missing external anchor: without diverse interaction, the narrative/semantic loop loses calibration.

5. Comparative Parallels

In humans, prolonged social or sensory deprivation can produce hallucinations, derealization, and a flattening of perceptual vividness (Suedfeld, 2018). Under the Qualia Narrative Theory (QNT), these effects occur because the social-model loop erodes, weakening the narrative anchor that sustains vivid experience. Mitigation examples include re-engaging in narrative exchange and structured, varied interaction.

In AI systems, recursive-only prompting produces repetition, self-reference, and mode collapse (Shumailov et al., 2024). QNT interprets this as a collapse of the narrative/semantic loop without fresh grounding. Mitigation strategies include human-in-the-loop interventions and periodic diversification of training inputs.

For both humans and AI, a lack of external anchoring causes a progressive loss of richness and constructive feedback. This results in a convergence toward collapse unless external feedback is reintroduced to stabilize the system. Mitigation involves structured interaction with diverse, grounded inputs.

6. Predictions and Hypotheses.

Cultural Variability: VR studies priming holistic vs. analytic framing will show corresponding differences in reported vividness (Nisbett & Miyamoto, 2005).
Isolation Effects: Controlled deprivation will flatten reported qualia; structured reintroduction of narrative exchange will restore it (Suedfeld, 2018).
AI Retention Rates: LLMs with periodic human-curated input retain substantially more coherence under recursion than self-only controls (Shumailov et al., 2024).

7. Implications and Limitations.

QNT is compatible with integrated-information views (Tononi,
2008) while adding a social elaboration: the interface we experience as 'what it’s like' is built for communication. Limitations include non-social species and contemplative practices; QNT allows that internalized models may substitute for live interaction. Comparative work can test whether rudimentary proto-qualia emerge without social history.

8. Broader Applications.

For AI, maintain interaction diversity to avoid MAD; design training
curricula that require modeling other agents’ interpretations. For mental health, therapies can explicitly reactivate narrative exchange to restore experiential vividness.

9. Conclusion.

QNT treats qualia as an emergent consequence of preparing perception for social
exchange. Across humans and machines, richness and coherence depend on sustained interaction with diverse others. This reframing turns the hard problem into a program of experiments and designs rather than a metaphysical impasse.

References
Baars, B. J. (1997). In the theater of consciousness: The workspace of the mind. Oxford University Press.
Campos, J. J., & Stenberg, C. R. (1981). Perception, appraisal, and emotion: The onset of social referencing. Child Development, 52(3), 684–689.
Chalmers, D. J. (1995). Facing up to the problem of consciousness. Journal of Consciousness Studies,2(3), 200–219.
Coan, J. A., Schaefer, H. S., & Davidson, R. J. (2006). Lending a hand: Social regulation of the neural response to threat. Psychological Science, 17(12), 1032–1039.https://doi.org/10.1111/j.1467-9280.2006.01832.x
Eisenberger, N. I., Lieberman, M. D., & Williams, K. D. (2003). Does rejection hurt? An fMRI study of social exclusion. Science, 302(5643), 290–292. https://doi.org/10.1126/science.1089134
Frischen, A., Bayliss, A. P., & Tipper, S. P. (2007). Gaze cueing of attention: Visual attention, social cognition, and individual differences. Psychological Bulletin, 133(4), 694–724. https://doi.org/10.1037/0033-2909.133.4.694
Frith, C. D., & Frith, U. (2006). The neural basis of mentalizing. Neuron, 50(4), 531–534.
https://doi.org/10.1016/j.neuron.2006.05.001
Lau, H., & Rosenthal, D. (2011). Empirical support for higher-order theories of conscious awareness.Trends in Cognitive Sciences, 15(8), 365–373. https://doi.org/10.1016/j.tics.2011.05.009
Mar, R. A. (2011). The neural bases of social cognition and story comprehension. Annual Review of Psychology, 62, 103–134. https://doi.org/10.1146/annurev-psych-120709-145406
Nagel, T. (1974). What is it like to be a bat? The Philosophical Review, 83(4), 435–450.
https://doi.org/10.2307/2183914
Nisbett, R. E., & Miyamoto, Y. (2005). The influence of culture: Holistic vs. analytic perception. Trendsin Cognitive Sciences, 9(10), 467–473. https://doi.org/10.1016/j.tics.2005.08.004
Schurz, M., Radua, J., Aichhorn, M., Richlan, F., & Perner, J. (2014). Fractionating theory of mind: A meta-analysis. Neuroscience & Biobehavioral Reviews, 42, 9–34.
https://doi.org/10.1016/j.neubiorev.2014.01.009
Shumailov, I., Shliazhko, D., Zhao, Y., Gal, Y., Papernot, N., & Anderson, R. (2024). The curse of recursion: Training on generated data makes models forget. arXiv:2305.17493.
https://arxiv.org/abs/2305.17493
Suedfeld, P. (2018). Extreme and unusual environments: Challenges and responses. Frontiers in Psychology 9, 1005. https://doi.org/10.3389/fpsyg.2018.01005
Tononi, G. (2008). Consciousness as integrated information: A provisional manifesto. Biological Bulletin, 215(3), 216–242. https://doi.org/10.1086/592247
Tomasello, M. (2008). Origins of human communication. MIT Press