Skip to main content

Altered States of Consciousness

Sleep, meditation, psychedelics, anaesthesia, hypnosis and lucid dreaming motivate different research protocols. Their identification with matrix trajectories is a bridge hypothesis [H/I], not a consequence of naming the axes of Γ\Gamma. The chapter separates the state model, observations, task capabilities and clinical outcomes.

Typed quantities

In a declared native frame, Γ∈D(C7)\Gamma\in\mathcal D(\mathbb C^7), P=TrΓ2P=\mathrm{Tr}\Gamma^2, Q=∑iγii2Q=\sum_i\gamma_{ii}^2, Φ=P/Q−1\Phi=P/Q-1 and R=1/(7P)R=1/(7P). For a specified numerical self-model M:D7→D7M:D_7\to D_7, RM=1−∥Γ−MΓ∥F2/PR_M=1-\|\Gamma-M\Gamma\|_F^2/P is a different score; it can be negative (orthogonal pure input and output give RM=−1R_M=-1). It is not the logical support reflector. For γij≠0\gamma_{ij}\ne0, phase Gap(i,j)=∣sin⁡arg⁡γij∣\mathrm{Gap}(i,j)=|\sin\arg\gamma_{ij}| ignores amplitude and oriented phase information; at zero coherence its phase is undefined. A mean over 21 pairs is defined only when all those phases are defined; otherwise the support/missing-value convention must be reported.

Capability and physical interpretation

The chosen gate Cap2\mathrm{Cap}_2 requires P>2/7P>2/7, R≥1/3R\ge1/3, Φ≥1\Phi\ge1 and Ddiff≥2D_{\mathrm{diff}}\ge2. The purity interval (2/7,3/7](2/7,3/7] checks only its first two clauses. L3 requires an independent nontrivial metamodel prediction certificate; L4 is the ideal compatible tower of such certificates. A high RMR_M, low Gap, fixed point or practice label supplies none of these missing tests. Identifying the gates with physical consciousness or clinical condition remains [H]. See the mathematical kernel and hierarchy.

Differentiation and extension

Dext=exp⁡SvN(ρE)D^{\mathrm{ext}}=\exp S_{vN}(\rho_E) needs a normalized state on a declared experiential space. A tensor factor permits a partial trace. A direct-sum E-sector instead uses its positive-weight compressed block divided by its trace. The 7D basis axis E is one-dimensional and cannot supply a multidimensional entropy by taking a partial trace over the other axes. The hierarchy's D7D=1+6CohED^{7D}=1+6\mathrm{Coh}_E is a separate stipulated proxy, not an approximation theorem for the extension spectrum. A 7D marginal does not uniquely determine an extension: Γ⊗ρE\Gamma\otimes\rho_E has the same marginal Γ\Gamma for every ρE\rho_E, with different experiential entropies. See 7D/42D typing.

Chapter roadmap​

Historical context; declared trajectories and observation models; research hypotheses for sleep and meditation; intervention outcomes; anaesthesia, hypnosis and dreaming; conditional transition dynamics.

1. Historical Perspective​

1.1 Charles Tart and the cartography of states​

In 1969, American psychologist Charles Tart published "Altered States of Consciousness", proposing the first systematic classification of altered states of consciousness (ASC). Tart viewed consciousness as a system possessing stable configurations — "discrete states of consciousness" (DSC). Each DSC is characterised by a set of "subsystems": input (perception), processing (thinking), output (behaviour), energy (attention), etc. Transition between DSCs is a destabilisation of one configuration and a transition to another.

Tart's key idea: Normal wakefulness is merely one of the possible configurations, not privileged from the standpoint of "truth". Sleep, meditation, the psychedelic state — these are equally legitimate configurations with their own regularities.

1.2 From Tart to UHM​

Tart's system perspective motivates a state-space model [I/H]. A measured configuration can be represented by Γ\Gamma only after defining a validated observation map. A stable category is an attractor only in a separately specified dynamics; changes between categories need not be bifurcations. The seven named axes, their numerical scores and their relation to Tart's subsystems are modeling choices, not an automatic translation theorem.

1.3 Predecessors and context​

Before Tart, altered states were studied in a fragmented way: William James (1902, "The Varieties of Religious Experience") described mystical states; Ludwig (1966) introduced the very term "altered states of consciousness"; Masters and Houston (1966) systematised psychedelic experience. But only Tart proposed a unified framework for all types of ASC.

In the 2000s, the neuroscience of ASC received a powerful impulse: fMRI studies of meditation (Lutz et al., 2004), neuroimaging of psychedelic states (Carhart-Harris et al., 2012), formalisation of the "entropic brain" (Carhart-Harris, 2014). The entropic brain hypothesis is the direct precursor of the parameter DdiffD_{\text{diff}} in UHM.


2. Altered States as Trajectories in Γ-Space​

Model definition [D]. A trajectory Γ(τ)\Gamma(\tau) and a specified numerical model MτM_\tau yield a feature trajectory s(τ)=(P,Φ,RM,D,G)s(\tau)=(P,\Phi,R_M,D,G) where each component is defined. Relative to a calibrated baseline, a declared criterion ∣sk(τ)−skbase∣>δk|s_k(\tau)-s_k^{\mathrm{base}}|>\delta_k marks a model feature deviation. The thresholds, uncertainty and timescale are part of the protocol. A deviation does not by itself establish an altered conscious state, clinical condition or unique experience [H].

The state space of 7×77\times7 density matrices has real dimension 4848 in its interior: six free diagonal populations and 21 complex off-diagonal entries. Five aggregates discard information and do not automatically distinguish all classes. Estimate the class from independently observed labels and validated distributions, not by defining it from the same features under evaluation.

2.1 State space: visualisation​

A (P,RM)(P,R_M) plot is a projection, not a complete classification. For a fixed MM, RMR_M need not be a function of PP; the canonical RR always is. The strip 2/7<P≤3/72/7<P\le3/7 is a scalar gate condition, not a boundary of life, a clinical norm or proof of experience. Estimated clusters may overlap and need not be attractors. Show uncertainty and independent labels; omit fabricated numerical coordinates for sleep, meditation or drug effects.

3. Sleep​

Sleep labels and reports must be obtained independently of the matrix score [Pr]. The following are candidate feature associations; they do not assign a universal capability level or infer absent experience from absent report.

3.1 NREM sleep (deep dreamless sleep)​

Candidate [H]. Compare independently identified NREM periods with self-model prediction accuracy, differentiation in a declared experiential model and task/report accessibility. The former assignments RM<1/3⇒L1R_M<1/3\Rightarrow\mathrm{L1} and “integration necessarily preserved above Φ=1\Phi=1” are not derived. A measured D<2D<2 would fail that chosen clause of Cap2\mathrm{Cap}_2, without proving the presence or absence of physical experience.

Vanishing γAE\gamma_{AE} does not imply GAE=1G_{AE}=1: its phase is undefined at zero. Failure to remember or report content cannot distinguish absent content from failed encoding or access without additional tests. No universal purity value for NREM is asserted.

3.2 REM sleep (dreaming)​

Candidate [H]. Relate independently identified REM periods and dream reports to features of a fixed encoding. Neither a sum of selected coherence moduli nor a low phase Gap is a sufficient condition for dreaming. Oriented phases and amplitudes may be investigated as predictors, but phase alone does not determine dream content. The relation between neural observations and entries named (S,E)(S,E) or (D,E)(D,E) requires an explicit observation model; matrix labels do not derive it.

A lucid or non-lucid report and a formal L-label are different outcomes. Report higher capabilities only after their full task certificates, and leave missing certificates unknown [Pr].

4. Meditation​

Practice labels suggest different task designs [H/I]. A voluntary instruction is an external protocol variable, not a demonstrated direct manipulation of matrix entries.

4.1 Shamatha (focusing attention)​

Candidate [H]. A focusing practice can be tested against an independently measured sustained-attention task and specified numerical self-model predictions. An increase of ∣γAE∣|\gamma_{AE}| does not force other moduli to fall: the trace-one family (1−t)I7/7+tuu†(1-t)I_7/7+tuu^\dagger raises all off-diagonal moduli simultaneously. The bound ∑X≠A∣γAX∣2≤γAA(1−γAA)\sum_{X\ne A}|\gamma_{AX}|^2\le\gamma_{AA}(1-\gamma_{AA}) is not a conserved budget. A tradeoff requires an additional controller law, as proved in Attention and memory.

A change in RMR_M or fitted generator parameters is a separate measurement. Neither monotone practice improvement nor a one-year or ten-year numerical trajectory follows from normalization or a two-timescale model. Such effects must be estimated from an actual protocol [Pr].

4.2 Vipassanā (insight)​

Candidate [H]. Compare reports of noticing previously unreported content with changes in a calibrated access task and separately estimated phases. A noticed discrepancy does not mathematically force Gap reduction or a bifurcation. A fixed score threshold has no universal order of bodily, emotional and logical “insights”.

Complete phase transparency is possible: the positive family in §4.1 has all 21 Gaps zero whenever t>0t>0. Hamming's code theorem does not require three of these pairs to remain opaque. Whether broader access improves a specified outcome must be tested independently; a phase statistic is not access itself.

4.3 Samādhi (deep absorption)​

A report of deep absorption does not certify L3 or L4 [H/I]. Self-model agreement and access to nontrivial higher-order predictions are distinct.

Counterexample to the former fixed-point proof [T]. For M=IdM=\mathrm{Id}, every state is a fixed point and RM=1R_M=1. At Γ=I7/7\Gamma=I_7/7, P=1/7P=1/7 and Φ=0\Phi=0, so Cap2\mathrm{Cap}_2 fails. Thus perfect self-agreement does not establish even the chosen base capability, much less a compatible metamodel tower or complete self-knowledge. The identity is a valid frozen CPTP map, so imposing CPTP does not fix the implication.

Feasibility check [T]. Since Q≥1/7Q\ge1/7, Φ≤7P−1\Phi\le7P-1. Within Cap2\mathrm{Cap}_2, P≤3/7P\le3/7 gives Φ≤2\Phi\le2. The old sample P=0.40P=0.40, Φ=3.5\Phi=3.5 cannot be a density-matrix certificate: its bound is 1.81.8. Maximizing Φ\Phi over all states yields 66 at a uniform pure state, whose R=1/7R=1/7 fails the access cut. Consequently “maximal integration” cannot replace the gate or the higher-order tests.

A study of absorption can test independent reports, tasks and metamodel predictions, including perturbations and held-out targets [Pr]. A stable state requires stability of a specified flow; a fixed point of MM alone supplies neither attraction nor a retention time.

4.4 Comparison of three practices​

Practice labelProposed independent target [H/Pr]What is not inferred
ShamathaSustained-attention performanceConserved coherence budget
VipassanāAccess/report discriminationMandatory phase-gap reduction
SamādhiAbsorption reports and higher-order task testsL4 from RM=1R_M=1 or a fixed point

The table compares research designs; it supplies no universal matrix profile or ordering of practice proficiency.

5. Psychedelics​

A matrix account of an intervention is a candidate model [H]. Its reconstructed features and independently assessed outcomes must remain separate. No dose, efficacy or safety recommendation is derived here.

5.1 General profile​

Changes in DD, RMR_M, PP or phase features can be studied under a fixed encoder, but their directions are not imposed by the matrix axioms. Reported “ego dissolution” is not equivalent to a fall in RMR_M, and a fall in Gap does not prove that suppressed records have become accessible. Rapid state motion does not by itself establish self-model prediction failure: that requires a task, target and an error readout. Approaching the chosen P=2/7P=2/7 cut does not by itself establish a physiological danger threshold.

5.2 Therapeutic window​

The former theorem of maximal therapeutic efficacy at 1/6<RM<1/31/6<R_M<1/3 is withdrawn [✗]. Neither bound follows from the definition of RMR_M, and purity alone does not certify the full gate or clinical safety.

Counterexample to the score implication. For a pure Γ=∣0⟩⟨0∣\Gamma=|0\rangle\langle0|, choose the valid replacement-channel output MΓ=(1−a)∣0⟩⟨0∣+a∣1⟩⟨1∣M\Gamma=(1-a)|0\rangle\langle0|+a|1\rangle\langle1|. Then RM=1−2a2R_M=1-2a^2; a=0.4a=\sqrt{0.4} gives RM=0.2R_M=0.2 inside the proposed window, while P=1P=1 gives canonical R=1/7R=1/7 outside Cap2\mathrm{Cap}_2. The window thus certifies neither base capability nor a clinical outcome. An independent outcome variable can differ while the same (Γ,M)(\Gamma,M) and score are retained; no outcome function is supplied by the score definition.

Research programme [H/Pr]. If an intervention hypothesis uses a model feature, preregister the observation model and an independently assessed benefit outcome, compare alternatives on held-out data, and assess clinical safety independently. Do not infer a treatment window, retained insight, dose or duration from RMR_M, Gap or the purity interval. The former timed numerical drug profile had no identified measurement basis and is withdrawn.

5.3 Connection to the entropic brain model​

The entropic-brain proposal is a motivating hypothesis, not an identification theorem [H/I]. Entropy of an observed neural time series, SvN(ρE)S_{vN}(\rho_E) and exp⁡SvN(ρE)\exp S_{vN}(\rho_E) are different quantities. Relating them requires an encoder, units and a validated statistical model. A “critical point” in a neural model is not automatically a phase boundary of the chosen Gap potential, and neither entropy measure alone establishes a phenomenal or therapeutic effect.

6. Anaesthesia​

A proposed model of anaesthesia must be linked to independent observations and outcome criteria [H/Pr]. A loss of selected matrix coherences is a mathematical process; equating it with a clinical state requires an additional validated bridge.

6.1 Sequence of coherence loss​

A declared dephasing channel can drive off-diagonal entries toward zero while preserving trace and populations [T]. Its rates and order of channel suppression depend on the actual generator; the universal sequence (E,U)→(L,E)→(A,E)→(S,E)(E,U)\to(L,E)\to(A,E)\to(S,E) does not follow from positivity or the seven-axis names. The former numerical induction table was an unvalidated assignment rather than reconstructed data.

For a nonzero entry approaching zero along different phases, Gap may stay at 00, 11, or another value until its phase becomes undefined. Hence vanishing coherence does not determine a phase-opacity progression or the instant at which physical experience ends.

6.2 Distinction from sleep​

A fitted model can compare independently identified sleep and anaesthesia conditions. It cannot define their distinction as “L1 versus L0” from RMR_M or Φ\Phi alone. Every valid density matrix satisfies the hierarchy's basic A0A_0 predicate; failing Cap2\mathrm{Cap}_2 does not determine which lower predicates hold or establish physical unconsciousness.

Counterexample to purity-based reversibility. A replacement channel sends every input record to a fixed pure state with P=1>2/7P=1>2/7 yet destroys distinguishability of those records at its output. A unitary channel preserves distinguishability at any input purity. Purity above the chosen cut therefore does not prove recoverability, survival or recovery of a person. Clinical support requirements are not deduced from this matrix inequality.

6.3 Awareness during anaesthesia (intraoperative awareness)​

The candidate relation between independently assessed awareness and selected features requires an outcome protocol [H/Pr]. A threshold ∣γAE∣>ε|\gamma_{AE}|>\varepsilon with GAE<1G_{AE}<1 does not prove felt experience, ability to report or a particular clinical complication. The encoder, observation noise and independent access/report measures must be specified. No rate or diagnosis is inferred from the matrix statistics here.

7. Hypnosis and Lucid Dreaming​

These labels can motivate distinct protocols; they do not supply a unique Γ-profile or capability certificate.

7.1 Hypnosis​

Candidate [H]. Compare independently measured response to suggestion, task attention and report with a fixed matrix model. Hypnosis is not defined by increased GLEG_{LE} or decreased RMR_M, and neither score establishes “logical control switched off”. A bypass from one named channel to another requires an actual causal model and intervention evidence; it is not inferred from index names or a static correlation.

7.2 Lucid dreaming​

Candidate [H/Pr]. Independently verified lucid-dream reports can be compared with task-dependent access and self-model prediction measures. A positive γAE\gamma_{AE} plus RM≥1/3R_M\ge1/3 is not a sufficient dreaming or L2 criterion. It omits Φ\Phi, canonical RR, differentiation and the observation bridge. Formal capability certification and a report of knowing that one is dreaming are recorded separately; no universal training trajectory or matrix values are asserted.

8. Summary Table: All ASC Classes​

Independently assigned conditionCandidate study target [H/Pr]Capability result
WakefulnessCalibrated baseline tasks and reportsRequires full certificate
NREM / REMIndependently identified sleep periods and access/report tasksNot fixed by sleep label
Meditation / absorptionPractice reports, attention and metamodel tasksL3/L4 not inferred from a fixed point
Psychedelic conditionSeparate feature estimates and intervention outcomesNo score-derived therapeutic window
AnaesthesiaIndependent observations and outcome criteriaPurity does not prove recovery
HypnosisIndependently measured response/task outcomesNot determined by phase Gap
Lucid dreamingVerified report and task-specific accessNot determined by RM≥1/3R_M\ge1/3

The former universal numerical profiles and L-assignments are withdrawn. Missing gate clauses or certificates remain unknown rather than imputed from the condition name.

9. Geometry of Transitions​

A trajectory may cross a feature threshold without any change in the stability of its underlying flow. A clinical transition, a score crossing and a mathematical bifurcation are three distinct claims.

9.1 Types of transitions​

Conditional saddle-node result [C]. For a declared one-dimensional smooth flow x˙=f(x,μ)\dot x=f(x,\mu), a saddle-node at (x0,μ0)(x_0,\mu_0) requires f=0f=0, fx=0f_x=0, fxx≠0f_{xx}\ne0 and fμ≠0f_\mu\ne0, with the usual local hypotheses. In higher dimensions a simple zero eigenvalue and the corresponding center-manifold nondegeneracy/transversality conditions are needed. A scalar readout RM=1/3R_M=1/3 supplies none of these derivatives. See Ghrist, bifurcation theory.

Counterexample to the former threshold proof [T]. The family x˙=−x+μ\dot x=-x+\mu has a unique globally stable equilibrium x∗=μx^*=\mu and eigenvalue −1-1 for every μ\mu. Its readout crosses 1/31/3 at μ=1/3\mu=1/3 without a saddle-node. Likewise a fixed point MΓ=ΓM\Gamma=\Gamma is not proof of a transition, attractor or cognitive level.

Falling asleep, entering a practice or changes in report may motivate a fitted model [H]. A Hopf, saddle-node or phase-transition interpretation requires the declared dynamics and independently estimated controls. No universal sequence, abruptness or clinical timescale is assigned here.

9.2 Transition diagram​

This diagram describes a research workflow [Pr], not a universal transition graph between clinical or phenomenal states. A dynamical model must explain observations beyond the score curve; successful prediction is assessed independently of its fitted features.

What we learned​

  1. Matrix features, task capabilities, reports and clinical outcomes have distinct types. Their correspondences are testable bridge hypotheses.
  2. The purity strip is only part of Cap2\mathrm{Cap}_2; a self-model score or phase Gap does not replace the remaining gate or higher-order prediction tests.
  3. An identity self-model has perfect agreement even at I7/7I_7/7, which fails the gate. A fixed point does not certify complete self-knowledge or L4.
  4. Full phase transparency is possible, while zero coherence has undefined phase. Hamming coding supplies no three-opaque-pair requirement.
  5. The proposed therapeutic score window and clinical numerical assignments are withdrawn; benefit, safety and recovery require independent outcome evidence.
  6. Crossing a scalar threshold is not a bifurcation. Conditional transition claims need a specified flow and verified nondegeneracy/transversality.

The Unconscious continues this distinction between phase, stored content and access; it does not identify every Gap =1=1 with a physically unconscious channel.