Knowing before the psyche

PET is the author's theory of a semiotic and informational foundation for the sciences of mind. Its paper v1 is outlined under the working title Knowing Before the Psyche. The claim in its draft: knowing precedes the psyche; nature's lawfulness is language proper; knowing is that language taken up by self-maintaining systems; human formal language is a natural metalanguage that cannot fully state its own rule from inside.established

On 2026-11-01 the author worked the theory through in a spoken session. This report records the positions stated there, sets each against the literature that meets it, and lists what a referee will raise. The positions are the author's. The wording is a condensation, filed as positions 7 to 15 and claims C7 to C10 in the paper workspace.established

One pattern runs through everything. A rock, a cell and a mind are the same lawful process at different depths. There is no second way of being reserved for knowers. Physical boundaries are real, and epistemology sits inside the one ontology, not beside it.provisional

Nothing thinks; everything registers. Thinking, knowing and feeling name registration at greater depth. They are not separate acts and not separate ingredients. A rock registers what strikes it. A cell registers in order to rebuild itself. A mind registers its own registering. On this reading PET avoids panpsychism: feeling is not added to matter, it is what registration is at the deepest layer.provisional

The rules do not evolve; they unfold. What looks like a new law at a higher level is the same law seen from another layer of homeostasis. The layers are nested, and each depends on the one beneath: there is no psyche without geography, and no geography without the physics that makes planets. Structure adds no rule. It lets the same rules be seen at another layer.provisional

Determinism, and no free will. The world moves only as it was bound to move. Agents were never agents, so there is no randomness to accommodate and no special role for an observer. The author states that the thesis will not be softened because it is unwelcome.provisional

Maintenance and selection are bound, on different layers. Self-maintenance works now, inside one system. Selection works across generations and decides which patterns get to maintain themselves. Smolin's The Life of the Cosmos extends the same selection logic to universes that reproduce through black holes^†. PET keeps the selection and refuses his later claim that the laws themselves evolve.provisional

Autopoiesis. Varela, Maturana and Uribe (1974) define the living system as a network of processes that produces its own components and its own boundary^†. Their later slogan is that living is knowing. This is PET's thesis that knowing precedes the psyche, already worked out for the cell. The 1974 full text was not obtained: the publisher copy is paywalled, and the university repository holds the abstract only.provisional

Self-evidencing.Friston (2013) argues that any system with a Markov blanket will appear to perform active inference to preserve its integrity, which yields homoeostasis and “a simple form of autopoiesis.” Kirchhoff, Parr, Palacios, Friston and Kiverstein (2018) tie the same formalism to autopoietic enactivism, naming operational closure and sense making. Allen and Friston (2018) argue the continuity from cognitivism to autopoiesis.established

The beautiful loop.Laukkonen and Chandaria's preprint gives three conditions for consciousness: a reality model, inferential competition to enter it, and epistemic depth, the reflexive sharing of the model through the system. Through that loop “the world model contains the knowledge that it exists.” The conclusion allows the conditions in “possibly much simpler brains.” The preprint grounds its account in self-evidencing and never names autopoiesis, Maturana, Varela or Markov blankets.established

The gap.The pairs are in the literature. The chain is not. A search on 2026-11-01 found no source that runs from autopoiesis through Markov blankets to the beautiful loop as one mechanism from cell to consciousness. PET's section 2 can claim that chain: one self-evidencing loop at three depths, building itself, evidencing itself, registering its own registering.provisional

Nesting is already formal.Ramstead, Badcock and Friston (2018) describe the universe as “Markov blankets of Markov blankets.” Friston and colleagues (2023) sort particles into inert and active, dissipative and conservative, ordinary and strange, where strange particles can be read as inferring their own actions. The typology lines up with P2's ladder: an inert particle registers, an active one maintains itself, a strange one registers its own registering. The alignment is PET's reading, not theirs.converging

The hardest fight. Maturana denies that information passes into a living system. The environment only perturbs it, and the system's own structure determines the change^†. PET treats information as the substrate. The author's answer: disturbance and information are one event. Seen from outside, it is a perturbation. Seen from inside the receiving structure, it is information. What Maturana rejects is instruction, and receiver-relative information does not instruct.provisional

The criterion.Kolchinsky and Wolpert (2018) define semantic information as the part of a system's correlation with its environment that is causally necessary for the system to keep existing, tested by interventions that scramble the correlation. This separates registering from knowing without a second substance. It is PET's best reply to the charge that “nature is language” is a metaphor.converging

The match. A disturbance becomes information only when it matches a pattern the receiving structure already carries. Without the match it does nothing for the system. The author's word for this is the click. Candidate support: receptor and ligand recognition; Barbieri's code biology, where codes are mappings made by adaptor molecules^†; predictive processing, where input outside the generative model is noise.provisional

The economy.The match pays for itself: it lowers the system's expected surprise by more than it costs to make. This ties the claim to the economic point of view in Freud and to free energy in Friston, which is Module 7 of the PET curriculum.provisional

Many codes, one information. Because information lives in the mapping and not in the medium, different codes can carry it. The cell already shows this: the genetic code is degenerate, and several codons specify the same amino acid^†. Human languages show it at the top of the ladder.provisional

Computation is metalanguage. In PET's layers, computation is a human formalization that describes lawful registration. It cannot be what binds, because whether a physical system computes depends on the description chosen, the observer-relativity argument of Searle and Putnam^†. Pattee's epistemic cut between laws and symbols^† and Rosen's argument that organisms cannot be fully simulated^† are the nearest allies.provisional

What is provably beyond computation. Cubitt, Pérez-García and Wolf (2015) construct quantum lattice systems for which deciding whether the spectrum has a gap is undecidable. Wolpert (2008) proves impossibility results for any physical device that observes, predicts or recalls, results that hold whatever the particular laws. Both are limits on knowing from inside, not evidence of a non-physical process.established

The testability wall. Every experiment yields finite data, and finite data can always be computed. No observation can show that a process is beyond computation. PET should therefore not claim hypercomputation. It should claim the positional limit: the rule cannot be fully computed from inside, by self-reference and capacity, as Gödel, Turing and Wolpert show for systems embedded in what they describe.provisional

The quantum. Hossenfelder and Palmer (2020) define a superdeterministic theory as one that violates Statistical Independence between hidden variables and measurement settings. PET's claim that agents were never agents denies the premise that makes settings free, so it sits naturally with them. It may not need that much. Bohmian mechanics and the Everett interpretation are deterministic and give the observer no special role; in both, randomness is how the unfolding looks from inside^†. The recommended position: claim determinism, read randomness as positional, and hold superdeterminism in reserve for the case where locality must be kept.contested

Wolfram, the nearest prior art. A New Kind of Science (2002), the Physics Project (2020) and the ruliad (2021) hold that fixed simple rules unfold into everything, that the outcome cannot be shortcut, and that the laws an observer sees depend on what kind of observer it is^†. PET differs on two counts: its substrate is registration and sign, not computation, and it posits one rule where the ruliad holds all of them.contested

Sixteen works were added to the Musaeum library during the session, each with its text extracted beside the source for search. Five could not be obtained through open channels and are listed so they can be fetched through the university library.established

The positions were spoken and condensed; the author has not yet reviewed this wording. Sources marked with a dagger are cited from memory. The claim that no source joins the three depths rests on one session's searches and is an absence of evidence. The beautiful loop was read in its two-author preprint; the published version adds Friston and was not compared. The 1974 autopoiesis paper and the Markov blanket critique were not read, so they are cited at second hand. Several readings recorded here were proposed by the report's writer, among them the mapping in M5 and the recommendations in X3 and X4, and the reader should weigh them as proposals.