Part III · Dynamics
# Modular Time
A von Neumann algebra with a state carries its own clock, canonically, without being given one. This is a theorem, and it is too good a coincidence to waste.
> Assumes
> Chapters 3 and 4. The separating property of $\Omega$ from §3.3 is the hypothesis that makes everything here work.
> Delivers
> Axiom 5.1 (phenomenal time is modular time); the effective-temperature estimate (5.3); the refutation of the decoherence objection.
5.1 Tomita–Takesaki theory
Let $\mathcal{N}$ be a von Neumann algebra and $\Omega$ a vector that is cyclic and separating for it — conditions guaranteed by Reeh–Schlieder for any local algebra in a QFT. Define an anti-linear operator on the dense set $\mathcal{N}\Omega$ by
$$ S_0\,a\Omega=a^{*}\Omega\qquad (a\in\mathcal{N}), $$
and let $S=J\Delta^{1/2}$ be the polar decomposition of its closure. Here $\Delta$ is a positive operator called the modular operator and $J$ an anti-unitary called the modular conjugation. The content of the Tomita–Takesaki theorem is that these objects, built from nothing but the algebra and one vector, have remarkable properties:
$$ \begin{gathered} J\,\mathcal{N}\,J=\mathcal{N}', \qquad \Delta^{is}\,\mathcal{N}\,\Delta^{-is}=\mathcal{N} \\[2pt] \sigma^{\omega}_{s}(a)=\Delta^{is}a\,\Delta^{-is}, \qquad K=-\ln\Delta \end{gathered} $$ (5.1)
Read these carefully. The first says the conjugation $J$ maps the algebra exactly onto its commutant: it exchanges "this system" with "everything else." The second says the one-parameter family $\Delta^{is}$ preserves the algebra — it is a genuine dynamics, an automorphism group $\sigma^\omega_s$, generated by the modular Hamiltonian $K$.
Nobody put this dynamics in. It was extracted from the algebra and the state. That is the fact this chapter builds on.
The KMS condition
The state $\omega$ is not merely preserved by $\sigma^\omega_s$; it is thermal with respect to it. For all $a,b\in\mathcal{N}$ the function $F(s)=\omega(a\,\sigma^{\omega}_{s}(b))$ extends analytically to the strip $0\lt\mathrm{Im}\,s\lt 1$ and satisfies
$$ \omega\bigl(a\,\sigma^{\omega}_{s+i}(b)\bigr)=\omega\bigl(\sigma^{\omega}_{s}(b)\,a\bigr) $$ (5.2)
which is the Kubo–Martin–Schwinger condition at inverse temperature $\beta=1$ in the modular parameter. In words: every state is a thermal equilibrium state, provided you use its own modular flow as the clock. Temperature is not a property a state has or lacks; it is a relation between a state and a choice of time.
5.2 Connes: the flow is intrinsic
An obvious worry is that $\sigma^\omega_s$ depends on which state $\omega$ we picked, and so is not a property of the algebra. Connes' cocycle theorem removes it. For any two faithful normal states there is a family of unitaries $u_s\in\mathcal{N}$, the Connes cocycle $(D\omega:D\varphi)_s$, with
$$ \sigma^{\omega}_{s}(a)=u_s\,\sigma^{\varphi}_{s}(a)\,u_s^{*} $$
so the two flows differ by an inner automorphism. Passing to the quotient by inner automorphisms therefore gives a map that does not depend on the state at all:
$$ \delta:\mathbb{R}\longrightarrow \mathrm{Out}(\mathcal{N})=\mathrm{Aut}(\mathcal{N})/\mathrm{Inn}(\mathcal{N}) $$ (5.3)
> This is also how the type III$_\lambda$ classification is defined: $\lambda$ records the period of $\delta$. Type III$_1$ — the physical case — is the one where $\delta$ has no period at all.
A type III von Neumann algebra has an intrinsic time. Not a preferred time coordinate imposed from outside, but a canonical one-parameter group of outer automorphisms that is part of the algebra's structure. Connes and Rovelli built the thermal time hypothesis on this observation, proposing that the flow of time in physics simply is the modular flow of the state of the universe.
A worked case: the Rindler wedge
The abstraction has one classical case where everything is explicit, and it is worth keeping in mind as a sanity anchor. Take $\mathcal{N}=\mathfrak{A}(W)$ for the right Rindler wedge $W=\{x\gt|t|\}$ and $\Omega$ the Minkowski vacuum. The Bisognano–Wichmann theorem says the modular flow is exactly the Lorentz boost preserving the wedge, and the modular parameter is proper time divided by $2\pi$ times the acceleration. The KMS condition (5.2) then reads as the statement that an accelerated observer sees a thermal bath at the Unruh temperature
$$ T_{\mathrm{U}}=\frac{\hbar a}{2\pi c\,k_B} $$
So in the one case we can compute, modular time is ordinary physical time and modular temperature is a real, in-principle-measurable temperature. That is the licence for treating the modular parameter as a temporal quantity in general.
5.3 Phenomenal time
> Axiom 5.1 · Duration
>
> Posited
>
> Phenomenal duration is the modular parameter $s$. Physical proper time relates to it by $t=\hbar\beta_{\mathrm{eff}}\,s$, where $\beta_{\mathrm{eff}}$ is the inverse modular temperature of the qualia-bearing collective mode. The specious present is the interval of $s$ over which the flow remains coherent.
The attraction of Axiom 5.1 is that it explains a puzzle rather than merely labelling one. Experienced time is not a sequence of instants; it has thickness, direction, and a rate that varies with state. A modular flow has exactly these properties: it is a flow rather than a parametrisation, it is asymmetric in the sense that $\Delta$ is positive, and its rate relative to laboratory time is set by $\beta_{\mathrm{eff}}$, which is a property of the state.
It also produces a number immediately. Setting the specious present at $\tau\approx 100\,\mathrm{ms}$:
$$ T_{\mathrm{eff}}=\frac{\hbar}{k_B\,\tau}=\frac{7.64\times10^{-12}\,\mathrm{K\,s}}{10^{-1}\,\mathrm{s}}\;\approx\;7.6\times10^{-11}\,\mathrm{K} $$ (5.4)
A tenth of a nanokelvin. The mode that carries experience must be extraordinarily cold in modular terms — that is, extraordinarily far from thermal equilibrium with the 310 K tissue surrounding it. At first sight this looks like an immediate refutation, and answering it occupies the rest of the chapter.
5.4 Why the decoherence objection does not land
Tegmark's estimate of neural decoherence times — between $10^{-13}$ and $10^{-20}$ seconds — is correct, and for any theory requiring the brain to sustain superpositions of macroscopically distinct computational states it is decisive. The Penrose–Hameroff programme has to answer it. This one does not, and the reason is worth understanding precisely, because it is the difference between a quantum theory of consciousness and a field theory of consciousness.
Start with the occupation number of a 40 Hz collective mode at body temperature:
$$ \begin{gathered} \bar{n}=\frac{k_BT}{\hbar\omega}=\frac{4.28\times10^{-21}\,\mathrm{J}}{2.65\times10^{-32}\,\mathrm{J}}\;\approx\;1.6\times10^{11} \\[4pt] \frac{\delta}{\bar{n}}\sim\bar{n}^{-1/2}\approx 2.5\times10^{-6} \end{gathered} $$ (5.5)
The mode is a hundred billion quanta deep — utterly classical, with fractional quantum corrections of a few parts per million. Nothing here is a delicate superposition. What the model requires is a coherent state of a driven, damped bosonic mode, and coherent states are the one thing decoherence cannot destroy.
Under the standard master equation for a damped oscillator,
$$ \begin{gathered} \dot\rho=-\tfrac{i}{\hbar}[H,\rho]+\gamma\bigl(2a\rho a^{\dagger}-a^{\dagger}a\rho-\rho a^{\dagger}a\bigr) \\[4pt] \Longrightarrow\qquad |\alpha\rangle \;\mapsto\; |\alpha e^{-\gamma t/2}\rangle \end{gathered} $$ (5.6)
a coherent state remains coherent and merely damps in amplitude, while coherence between two distinct coherent states decays at the rate $\tfrac{\gamma}{2}|\alpha-\beta|^2$. Both facts matter. The second reproduces Tegmark's conclusion exactly: superpositions of macroscopically distinct field configurations die essentially instantly. The first is what he did not need to consider: the individual coherent state is the survivor. Decoherence does not attack coherent states — it einselects them (Zurek, Habib and Paz). They are the pointer basis of a linearly coupled bosonic mode.
> The upshot
>
> Derived
>
> The physical substrate this theory needs is a driven collective electromagnetic mode with a long-lived phase. That is not an exotic requirement — it is what an EEG rhythm is. The genuinely quantum content of the theory is algebraic and modular, and algebraic structure has no decoherence time: Theorem 3.1 holds at every scale, temperature and degree of dissipation.
The cost of this move should be stated. By locating the theory in the classical-field limit, we forgo any explanation that would require quantum computation, nonlocal binding across the brain by entanglement, or observer-induced collapse. Whatever binding this theory delivers must come from field coherence and the split property, not from entanglement between distant neurons. Chapter 11, prediction 8 is the sharp form of that commitment.
> ### What Chapter 5 established
>
> - Tomita–Takesaki: an algebra plus a cyclic separating vector generates a canonical dynamics $\sigma^\omega_s$ and a conjugation $J$ exchanging system and complement.
> - Every state is KMS-thermal with respect to its own modular flow. Temperature is a relation between state and clock.
> - Connes: modulo inner automorphisms the flow is state-independent, so the algebra has an intrinsic time. Bisognano–Wichmann makes it explicit for the Rindler wedge.
> - Axiom 5.1 identifies phenomenal duration with $s$; a 100 ms specious present implies $T_{\mathrm{eff}}\approx 8\times10^{-11}\,\mathrm{K}$.
> - The decoherence objection fails because coherent states are einselected, not destroyed. The theory needs a classical field mode, not a fragile superposition.
> ### Exercises
>
> 1. Verify that $S=J\Delta^{1/2}$ is an involution on $\mathcal{N}\Omega$, and deduce $J^2=1$ and $J\Delta J=\Delta^{-1}$. [1]
> 2. Show that for a finite-dimensional algebra with density matrix $\rho$, the modular flow reduces to $\sigma_s(a)=\rho^{is}a\rho^{-is}$, and hence that $K=-\ln\rho$. Why does this construction fail to define an entropy in type III? [2]
> 3. Using the Bisognano–Wichmann result, compute the acceleration required for an Unruh temperature of 1 K. Comment on why the effect has never been observed. [2]
> 4. Recompute (5.4) for a specious present of 25 ms and of 3 s. Over what range of $T_{\mathrm{eff}}$ must a theory of altered time perception operate? [2]
> 5. From (5.6), compute the ratio of the decoherence rate between $|\alpha\rangle$ and $|{-\alpha}\rangle$ to the amplitude damping rate, for $|\alpha|^2=\bar n=1.6\times10^{11}$. Explain in one sentence why this ratio is the whole answer to Tegmark. [2]
> 6. Open. Axiom 5.1 gives phenomenal time a rate, $\hbar\beta_{\mathrm{eff}}$. Derive the observed dilation of subjective time under high arousal from a change in $\beta_{\mathrm{eff}}$, and predict its sign. [3]
For agents
GET /api/library/spectral-panpsychism/ch5.md