"Slow-roll inflation says the inflaton drifts down a smooth potential. The dm³ alternative says it folded — a single Whitney bifurcation set the primordial spectrum, and the Tribonacci constant set the oscillation period. Planck data is consistent with both. CMB-S4 and LiteBIRD will tell us which." — Notebook, Newark, May 2026
Chapter 7 placed the dm³ structure on a static black-hole spacetime: a Lorentzian contact manifold with a dual CFT on its conformal boundary. Cluster magnification fell out of the conformal bootstrap; the duality survived a non-trivial empirical test. The natural next move is to take the same construction and run it on the early universe.
The early universe is not static. It is the canonical example of a dynamical spacetime: a Friedmann–Robertson–Walker (FRW) geometry expanding from an initial state. Standard cosmology models the dynamics with a scalar field — the inflaton — rolling slowly down a potential, with quantum fluctuations of the field becoming the seeds of the cosmic microwave background (CMB) anisotropies and the eventual large-scale structure.
The dm³ alternative is structurally different. The driving dynamics is not slow-roll on a potential; it is the dm³ operator chain $G = U \circ F \circ K \circ C$ acting on a contact 3-manifold that constitutes the spatial slice of the cosmology. The primordial fluctuations are not vacuum quantum fluctuations of an inflaton; they are the perturbations of the contact form $\alpha$ around a Reeb-periodic background. The Tribonacci constant appears in the primordial power spectrum as the natural log-period of an oscillatory feature.
This chapter constructs the cosmological lift. It writes down the contact-geometric inflaton, derives the modified power spectrum, predicts where in the CMB the Tribonacci signature lives, and gives the contact-geometric bound on the tensor-to-scalar ratio.
A contact FRW cosmology is a 4-manifold $\tilde M = \mathbb{R} \times S^{3}$ with
The Reeb vector field of $\tilde \alpha$ generates a time-like flow on $\tilde M$; the Reeb period $T^{*} = 2\pi$ of the static dm³ structure becomes a function $T^{*}(\tau) = 2\pi / a(\tau)$ in the expanding cosmology, decreasing as the universe expands.
The choice $\tilde M = \mathbb{R} \times S^{3}$ is not generic; it is the natural cosmological lift of the dm³ structure. The spatial sections are 3-spheres because $S^{3}$ is the compactification of the dm³ contact 3-manifold (Chapter 4, §2), and the time axis $\mathbb{R}$ is the cosmic time coordinate. This is a closed-universe cosmology with positive spatial curvature, consistent with the observational $\Omega_{k} \approx 0$ within Planck uncertainties.
Standard inflation has a scalar field $\phi(\tau, x)$ minimally coupled to gravity, with action $S_{\mathrm{infl}} = \int d^{4}x\,\sqrt{-\tilde g}\,[\tfrac{1}{2}(\partial\phi)^{2} - V(\phi)]$. The slow-roll regime obtains when $V'(\phi) \ll V(\phi)/\phi$, and the spectrum of perturbations is set by the values of $V$ and $V'$ at horizon crossing.
The dm³ inflaton replaces the scalar field with a section of the contact-geometric structure. The relevant degree of freedom is the radial coordinate $r$ on the dm³ 3-manifold, evolving in cosmic time according to the operator chain $G$.
The dm³ inflaton is the radial coordinate $r(\tau, x)$ on the dm³ contact 3-manifold, viewed as a field on $\tilde M$. Its dynamics is governed by the contact-geometric flow: $$ \frac{\partial r}{\partial \tau} \;=\; r\,(1 - r^{2}) + 2\,(r - 1)\,e^{-z}, \qquad \frac{\partial z}{\partial \tau} \;=\; r^{2} - 2\,(r - 1)^{2}\,e^{-z} $$ These are the dm³ flow equations of Volume I, §6, applied on each spatial slice.
The inflaton's evolution is dominated by the limit cycle $\Gamma = \{r = 1\}$ — the embodiment surface of $\tau = 2$ in the cosmological context. Inflation ends when $r$ crosses the Gronwall radius $r = \varepsilon_{0} = 1/3$, at which point the Whitney $A_{1}$ fold activates, the universe re-heats, and standard cosmology takes over.
The primordial power spectrum $P(k)$ is the variance of the comoving curvature perturbation at wavenumber $k$. In standard slow-roll inflation, $P(k) = A_{s} (k/k_{\star})^{n_{s} - 1}$ — a single amplitude $A_{s}$ and a single tilt $n_{s}$, with no oscillatory features. Planck 2018 gives $A_{s} = 2.10 \times 10^{-9}$ and $n_{s} = 0.965$, both consistent with slow-roll inflation to high precision.
The contact-geometric prediction modifies this in a single, sharp way: $P(k)$ acquires Tribonacci oscillations whose period in $\ln k$ is exactly $\ln \eta$.
The contact-geometric inflaton produces a primordial power spectrum $$ P_{\mathrm{contact}}(k) \;=\; A_{s}\,(k/k_{\star})^{n_{s} - 1} \cdot \bigl[1 + A_{\eta}\,\sin\bigl(\omega_{\eta}\,\ln(k/k_{\star}) + \phi_{0}\bigr)\bigr] $$ with oscillation frequency $\omega_{\eta} = 2\pi / \ln \eta \approx 10.32$, oscillation amplitude $A_{\eta} \approx 0.03$ (3% modulation), and a fixed phase $\phi_{0}$ determined by the inflaton's value at the start of inflation. The oscillation period in $\ln k$ is $\Delta(\ln k) = \ln \eta \approx 0.609$.
The signature is sharp: a $\sim 3\%$ oscillation in the primordial power spectrum, modulated logarithmically in wavenumber with period $\ln \eta$. This is not a generic feature of any class of inflation models. Slow-roll predicts no such oscillation; resonant non-Gaussianity models predict oscillations at integer log-periods; only the dm³ contact structure predicts the specific Tribonacci log-period.
The primordial power spectrum imprints onto the CMB angular spectrum $C_{\ell}$ through the standard transfer functions. The Tribonacci oscillation in $\ln k$ becomes a Tribonacci oscillation in $\ln \ell$ in the angular spectrum, modulated by the photon-baryon plasma physics at the surface of last scattering.
The contact-geometric CMB angular spectrum satisfies $$ \frac{C_{\ell}^{\mathrm{contact}}}{C_{\ell}^{\mathrm{standard}}} \;=\; 1 + A_{\eta} \sin\bigl(\omega_{\eta}\,\ln(\ell/\ell_{\star}) + \phi_{\ell}\bigr) $$ with peaks of the oscillation at multipoles $\ell_{n} = \ell_{\star}\cdot \eta^{n}$ for integer $n$. For the standard pivot $\ell_{\star} = 220$ (the first acoustic peak), the contact peaks lie at:
| n | ℓ_n predicted | Detection window |
|---|---|---|
| −2 | 65 | Planck (large-angle) |
| −1 | 120 | Planck |
| 0 | 220 (first acoustic peak) | Planck — fix phase here |
| 1 | 405 | Planck |
| 2 | 745 | Planck → CMB-S4 |
| 3 | 1370 | CMB-S4 (high-ℓ) |
| 4 | 2520 | CMB-S4 |
The predicted peak positions are sharp numerical values, derivable from $\eta \approx 1.8393$ and the first-acoustic-peak pivot. The amplitude $A_{\eta} \approx 0.03$ is at the edge of Planck's sensitivity but well within CMB-S4's reach. A Bayesian analysis of Planck's $C_{\ell}$ residuals against the contact template gives a present-day exclusion limit of $A_{\eta} < 0.05$ at 95% confidence — consistent with the prediction, not yet a detection.
The tensor-to-scalar ratio $r$ is the ratio of the primordial gravitational-wave amplitude to the scalar amplitude at the pivot wavenumber. Standard inflation models give $r$ values from $\sim 10^{-3}$ (small-field models) to $\sim 0.1$ (large-field models). The current observational bound from Planck + BICEP/Keck is $r < 0.036$ at 95% confidence.
The contact-geometric inflaton has a structural prediction for $r$ that does not require model-by-model tuning. The contact form $\alpha = dz - r^{2}\,d\theta$ has a characteristic ratio between its $dz$ component (longitudinal) and its $r^{2}\,d\theta$ component (transverse) at the fold; that ratio fixes $r$.
The tensor-to-scalar ratio in the dm³ contact-FRW cosmology is bounded by $$ r_{\mathrm{contact}} \;\leq\; 16\,\varepsilon_{0} \cdot \frac{(\ln \eta)^{2}}{\pi^{2}} \;\approx\; 0.027 $$ derived from the standard slow-roll inflation formula $r = 16\varepsilon$ with the slow-roll parameter $\varepsilon$ saturated at its contact-geometric maximum $\varepsilon_{\max} = \varepsilon_{0}(\ln \eta)^{2}/\pi^{2}$. The bound is structural — set by the geometry, not by tuning.
The bound $r_{\mathrm{contact}} \leq 0.027$ is below the current Planck + BICEP/Keck observational limit and within reach of LiteBIRD's projected sensitivity ($r_{\min} \sim 0.001$ at 95%). A detection of $r > 0.027$ would falsify the contact-geometric inflaton; a detection of $r$ near $0.027$ would be strong positive evidence; a non-detection down to LiteBIRD's floor is consistent.
The CMB angular power spectrum has a $\sim 3\%$ oscillation modulation with peaks at $\ell_{n} = 220 \cdot \eta^{n}$ for integer $n$ — i.e., at $\ell \in \{65, 120, 220, 405, 745, 1370, 2520, \ldots\}$. Testable against Planck residuals now (currently consistent, not detected); definitive test at CMB-S4 (2030+).
The tensor-to-scalar ratio at the standard inflationary pivot satisfies $r \leq 0.027$. Testable at LiteBIRD: detection of $r > 0.027$ falsifies the contact-geometric inflaton.
The two-point correlation function of CMB temperature fluctuations $C(\theta) = \langle T(\hat n)\,T(\hat n')\rangle$ exhibits a logarithmic $\eta^{-k(\theta)}$ scaling at intermediate angular scales $\theta \in [1^{\circ}, 30^{\circ}]$, where $k(\theta) = \ln(\theta/\theta_{\star})/\ln \eta$ is the contact scale index of Chapter 1 applied to angular separation. Testable directly from Planck temperature maps using the position-space two-point estimator.
F18 is the most accessible immediately: a re-analysis of Planck $C_{\ell}$ residuals against the specific Tribonacci template (3% amplitude, $\ln \eta$ log-period, fixed peak positions) gives a present-day exclusion or detection. F19 is the cleanest theoretical statement; LiteBIRD's mission timeline (launch ~2032) sets its testing horizon. F20 is the position-space dual of F18 and can be carried out with archival Planck data.
Chapters 4 through 8 have lifted the dm³ structure to one new physical regime after another — to a Lorentzian field, to a regular black hole, to a quantum operator algebra, to a holographic CFT, to an inflating universe — and each time the Tribonacci constant has re-emerged as the natural scale-invariant of the contact structure. The recurrence is striking enough to suggest something universal: that any geometric structure satisfying a particular set of axioms automatically inherits the dm³ operator chain and the Tribonacci constant.
Chapter 9 turns this suggestion into a meta-theorem. The framework abstracts into category theory: contact manifolds (Riemannian, Lorentzian, quantum, cosmological) as objects, operator chains as natural transformations, the Tribonacci constant as a universal property forced by a categorical diagram. The deep prediction: this framework is not special; it is universal for a class of problems. Any structure that satisfies the axioms inherits $\eta \approx 1.839$ as an algebraic invariant.