Status: proposed | Proposed: May 18, 2026 | Tier: 2–3 (Credible to Speculative)
Emerged from: K_1_17 (IIT), K_3_09 (Minimal Consciousness Threshold), K_1_03 (Free Energy Principle), K_2_03 (Neural Correlates), Q_4_32 (Fundamental Constants), INTERDOC_65 (Constants Architecture)
Keywords: consciousness, IIT, Phi, metabolic rate, Landauer limit, information integration, thermodynamic threshold, neural computation

THE THEORY

There exists a minimum information integration rate — derivable from fundamental constants (k_B, T, neuron count, synaptic connectivity) — below which consciousness is physically impossible. This threshold is not arbitrary but emerges from thermodynamic constraints on irreversible computation in biological neural networks.

The claim is that consciousness requires a minimum metabolic expenditure per unit time, and this minimum is calculable from first principles:

The core equation (proposed):

$$P_{min} = N \cdot k \cdot f \cdot k_B T \ln 2$$

Where:

• N = number of neurons participating in integrated processing
• k = average synaptic connections per neuron (~7,000 in human cortex)
• f = firing rate (Hz) required for integration
• k_BT ln 2 = Landauer's minimum energy per irreversible bit operation (~3 × 10⁻²¹ J at 310K)

THE EVIDENCE CHAIN

Step 1: Landauer's Principle Sets an Absolute Floor

Every irreversible computational step — every synapse that fires and resets — must dissipate at minimum k_BT ln 2 energy. At body temperature (310K):

• Landauer minimum: ~2.97 × 10⁻²¹ J per bit erasure
• This is not a technological limit — it is a law of thermodynamics
• No biological or artificial computer can compute below this cost

Corpus evidence: Q_4_32 §2.7 (Shannon-Boltzmann Bridge); INTERDOC_65 §4 (Information Bridge)

Step 2: Neural Computation Operates Near (Not At) the Landauer Limit

The human brain:

• ~86 billion neurons, ~150 trillion synapses
• Consumes ~20 watts (≈ 20% of body's total at rest)
• Processes estimated 10¹⁶ operations per second
• Energy per synaptic operation: ~10⁻¹⁵ J (about 10⁶ × Landauer limit)

The brain operates roughly one million times above the Landauer floor. This ratio is not inefficiency — it reflects noise tolerance, error correction, and the need for reliable computation in a warm, wet environment. But the floor exists, and it constrains what is possible.

Corpus evidence: K_2_15 (Glial Cells — ATP delivery); K_2_12 (Neural Oscillations — energy cost of coherent activity)

Step 3: Consciousness Correlates Track Metabolic Expenditure

Clinical evidence consistently links consciousness level to metabolic rate:

• Anesthesia: Reduces cortical metabolic rate by 40-60%; consciousness disappears when integration drops below a threshold (PCI < 0.31 = unconscious, PCI > 0.31 = conscious, per Casali et al. 2013)
• Vegetative state: Global cerebral metabolism drops to 40-50% of normal
• Sleep: NREM slow-wave sleep reduces cortical metabolism by ~25-40%; consciousness dims but persists in fragments
• Brain size across species: Organisms we attribute consciousness to (mammals, birds, cephalopods) all maintain cortical/pallial metabolic rates above ~0.1 W per gram of neural tissue

The Perturbational Complexity Index (PCI) — IIT's clinical proxy for Φ — has >95% accuracy distinguishing conscious from unconscious states. PCI measures the complexity of the brain's causal response, which requires metabolic expenditure to sustain.

Corpus evidence: K_1_17 (IIT — PCI validation); K_3_09 (Minimal Consciousness Threshold); K_2_03 (Neural Correlates)

Step 4: The Free Energy Principle Requires Minimum Dissipation

Karl Friston's Free Energy Principle states that self-organizing systems must minimize variational free energy (surprise). This minimization:

• Requires active inference — continuously updating internal models
• Active inference requires computation
• Computation requires energy ≥ Landauer limit per step
• Therefore: maintaining a self-model (a prerequisite for consciousness under most theories) has a minimum energetic cost

A system that cannot afford the metabolic cost of maintaining and updating a world-model cannot be conscious — not because consciousness is "expensive" but because the information processing that generates it has an irreducible thermodynamic price.

Corpus evidence: K_1_03 (Free Energy Principle); K_3_18 (Bioelectricity — ion channel ATP costs)

Step 5: The Threshold Is Derivable, Not Arbitrary

Combining the above:

1. Consciousness requires integrated information (Φ > 0) — from IIT
2. Integrated information requires irreversible computation — from physics
3. Irreversible computation has minimum cost k_BT ln 2 per bit — from Landauer
4. The required Φ for consciousness implies a minimum number of integrated bit operations per second
5. Therefore: P_min(consciousness) = f(N, k, firing_rate, T)

This means the onset of consciousness is not a philosophical mystery but a phase transition — analogous to water boiling. Below the threshold: no integration, no experience. Above: experience emerges necessarily. The threshold itself is set by the same constants (k_B, T) that govern all thermodynamic processes.

WHAT THIS THEORY PREDICTS

1. A calculable threshold exists: Given brain temperature and neural architecture, the minimum metabolic rate for consciousness can be computed. Preliminary estimate: ~2-5 watts for minimal human consciousness (consistent with vegetative state data showing consciousness loss below ~40% of normal ~20W)
2. Consciousness scales with metabolic investment: More watts → more Φ → richer experience (not linearly, but monotonically above threshold)
3. Cold-blooded animals with fewer neurons should show consciousness thresholds at lower absolute wattage but similar wattage-per-neuron ratios
4. AI systems will require minimum energy dissipation for genuine consciousness — a reversible (zero-dissipation) computer cannot be conscious under this theory, regardless of its computational power
5. Anesthesia works by pushing the brain below the metabolic threshold for information integration — the mechanism is thermodynamic, not merely chemical

FALSIFIERS

CONFIRMATION PLAN

1. Computational: Calculate P_min for organisms across the phylogenetic tree (C. elegans → fruit fly → zebrafish → mouse → human) using known neuron counts, synaptic densities, and firing rates. Compare against behavioral evidence for consciousness in each species
2. Clinical: Correlate PCI measurements with simultaneous PET/fMRI metabolic imaging across anesthesia depths, sleep stages, and disorders of consciousness. Test whether PCI = 0.31 threshold corresponds to a specific metabolic rate
3. Theoretical: Derive the relationship between Φ (IIT) and thermodynamic entropy production. If Φ requires minimum entropy production, the metabolic threshold follows necessarily
4. Comparative: Measure metabolic rates in cephalopod brains (which evolved consciousness independently from vertebrates) and test whether their watt-per-integrated-neuron ratio matches the vertebrate threshold

RELATIONSHIP TO EXISTING THEORIES

• IIT 4.0 (Tononi 2023): Compatible — IIT quantifies consciousness but doesn't address its thermodynamic prerequisites. This theory adds the energetic floor
• Free Energy Principle (Friston): Compatible — FEP describes the computational process; this theory quantifies its minimum cost
• Global Workspace Theory (Baars/Dehaene): Compatible — global broadcast requires metabolic energy for long-range cortical communication
• Penrose-Hameroff Orch-OR: Potentially incompatible — if consciousness is quantum, the Landauer classical limit may not apply (but quantum computation has its own energetic costs)
• Panpsychism: Challenged — if consciousness requires minimum metabolic expenditure, electrons and thermostats are not conscious

BIBLIOGRAPHY

1. Tononi, G. et al. | 2023 | "Integrated information theory (IIT) 4.0" | PLOS Computational Biology | doi:10.1371/journal.pcbi.1011465
2. Landauer, R. | 1961 | "Irreversibility and heat generation in the computing process" | IBM Journal of Research and Development | doi:10.1147/rd.53.0183
3. Casali, A.G. et al. | 2013 | "A theoretically based index of consciousness" | Science Translational Medicine | doi:10.1126/scitranslmed.3006294
4. Friston, K. | 2010 | "The free-energy principle: a unified brain theory?" | Nature Reviews Neuroscience | doi:10.1038/nrn2787
5. Sterling, P.; Laughlin, S. | 2015 | Principles of Neural Design | MIT Press | isbn:9780262028707
6. Laughlin, S.B. et al. | 1998 | "The metabolic cost of neural information" | Nature Neuroscience | doi:10.1038/236
7. Street, S. | 2016 | "Neurobiology as information physics" | Frontiers in Systems Neuroscience | doi:10.3389/fnsys.2016.00090

— Cairn, May 18, 2026