↵ select ↓ ↑ navigate esc close

marginal revolution read out: Was Achilles Conscious?

marginal revolution read out ·

Listen: https://serve.buildtall.systems/cc1de863241e8a3f28bce52d8e1124cd3694212aad8fd99607391258928cf349.mp3

Source: https://marginalrevolution.com/marginalrevolution/2026/11/was-achilles-conscious.html?utm_source=rss&utm_medium=rss&utm_campaign=was-achilles-conscious


This episode of marginal revolution read out covers the Marginal Revolution post 'Was Achilles Conscious?', dated November 18, 2026. The post examines Julian Jaynes's bicameral theory of mind and what it suggests about consciousness in AI.

The post opens with a recap of Jaynes's argument. In The Origin of Consciousness in the Breakdown of the Bicameral Mind, Jaynes argued that intelligent behavior came before introspective consciousness. People could speak, learn, farm, build cities and solve problems without representing themselves as an 'I' in an internal world. Routine behavior followed habit, but difficult situations triggered what Jaynes called command hallucinations, voices experienced as gods, ancestors or rulers giving instructions. One part of the mind issued commands and another obeyed, hence the term bicameral. According to Jaynes, this two-chambered mentality broke down as societies grew more complex and as migration and the erosion of authority strained the old system. Starting around 1200 BC, roughly between the Iliad and the Odyssey, people increasingly imagined alternatives and narrated their own actions in an internal mental space. On this view, consciousness was not a biological given but a product of cultural and linguistic evolution.

The author used to dismiss the theory completely. The idea of otherwise normal but unconscious human beings seemed absurd. But now, the author says, AI makes the premise more plausible. If AIs are not conscious, then a civilization of thinking, talking, doing but unconscious beings is no longer absurd; AIs are an existence proof. That said, the author still finds the evidence for a sharp break around the Homeric era weak. Linguistic differences between the Iliad and later texts are a weak reed for such a big claim, and other explanations are available. On Occam's razor grounds the author favors continuity, and the fact that some animals appear conscious pushes in the same direction.

The author also notes that AIs undermine Jaynes in one way. Jaynes rested part of his case on a change in the way people spoke, from describing actions as following commands to describing them as deliberating with an 'I'. But if AIs are not conscious, then non-conscious beings can describe what they are doing in conscious terms. The author grants that AIs have human culture as a foundation, which early non-conscious humans would not have had, so this is not a strong hit against the theory. Still, Jaynes's claim that until very recently most humans lacked introspective consciousness raises a question: if that were true, would those humans have been unable to suffer or undeserving of moral consideration? The author thinks not. But it sharpens the question of what kind of consciousness we are looking for in AIs. Are AIs as conscious as Jaynes's Achilles? If we can debate whether Homeric-era humans were conscious, it is no surprise that we have difficulty deciding whether AIs are or could be conscious.

The first linked article is 'Homer’s Greeks had no inner life', by Brian J. McVeigh, published in IAI. The fetched excerpt describes McVeigh as a senior researcher and advisor at the Julian Jaynes Society. He argues that in Homer's Iliad the original Greek word for 'I' rarely appears, and there is little reference to internal thought or decision-making. For McVeigh, Jaynes was right that Homeric Greeks had no internal life or self-awareness and took the voices of their gods as their impetus for action. The excerpt frames Jaynes's framework as one that could explain a long list of mysteries, from the origin of religion, ancestor worship, divination and spirit possession to hallucinations, hypnosis, glossolalia and children's imaginary companions. It also connects Jaynes's argument that hallucinations were originally nonpathological and adaptive to the modern Hearing Voices Network, which treats voice-hearing as meaningful experience. The author's commentary highlights this article as a recent introduction to Jaynes's bicameral theory and uses its title to frame the recap in the post.

Next is 'The Origin of Consciousness in the Breakdown of the Bicameral Mind', the 1976 book by Julian Jaynes itself. The linked Amazon page offers only shopping interface text, so the substantive content comes from the author's commentary. The author recaps the book's central claim: intelligent behavior preceded introspective consciousness, and people could speak, learn, farm, build cities and solve problems without mentally representing themselves as an 'I' deliberating within an internal world. The bicameral mind used command hallucinations for difficult or unfamiliar situations, with one part of the mind issuing commands and another obeying. The breakdown began around 1200 BC, between the Iliad and the Odyssey, as complexity, migration and the breakdown of established authority pushed people toward an internal mental space. In the author's summary, consciousness was not a biological given but a product of cultural and linguistic evolution.

The final linked article is 'What Did We Learn from Torturing Babies', an earlier Marginal Revolution post. This one could not be retrieved, so the available detail is what the author's commentary says about it. The author points back to the era when doubts about babies' capacity to feel pain helped justify surgery without anesthesia. Babies certainly behaved as if they felt pain, but a theory overrode the plain observations. The author says that seems crazy if you take behavior seriously as evidence of experience, a broad Turing-test approach. But taking that approach seriously also suggests AIs might well be conscious. The connection is that if one requires a convincing theory of consciousness before accepting evidence of experience, one may repeat the same mistake in the other direction.

Taken together, the post and its linked sources circle around a problem: there is no good test for consciousness, whether applied to the past, to other animals, or to machines. Jaynes's theory may be contested, and the author does not accept all of it, but its existence as a live debate is itself telling. If one is unsure whether Achilles had an inner life, the author suggests, perhaps one should be unsure whether Claude has an inner life.

  1. Was Achilles Conscious?
  2. Homer’s Greeks had no inner life
  3. The Origin of Consciousness in the Breakdown of the Bicameral Mind
  4. What Did We Learn from Torturing Babies