Prism Hypothesis in Topological Cosmology

• The prism hypothesis is a topological model where prism-shaped spherical manifolds truncate Laplacian modes to suppress large-scale CMB correlations.
• Its mathematical foundation employs binary dihedral groups to construct homogeneous prism spaces and double-action manifolds with strict eigenmode selection rules.
• Observationally, these models predict distinctive circle-in-the-sky signatures and quantitative CMB suppression linked to the injectivity radius and cell geometry.

The prism hypothesis in topological cosmology pertains to the impact of multi-connected spherical 3-manifolds with prism-shaped fundamental domains (specifically quotients by binary dihedral groups and their extensions) on the suppression of large-scale temperature correlations observed in the cosmic microwave background (CMB). Central to this hypothesis is the notion that specific topological identifications—particularly those yielding "well-proportioned" or polyhedral-like cell geometries—truncate the Laplacian mode spectrum, naturally suppressing the power at large angular scales in the CMB, and potentially offering an explanation for the observed low amplitude of CMB correlations at large angles. The group-theoretical underpinnings, spectral properties, and their physical manifestations in observational cosmology have been rigorously studied in the context of single-action and double-action prism spaces, especially in comparison with lens and Platonic topologies.

1. Mathematical Foundation of Prism Spaces

Compact spherical 3-manifolds can be represented as quotients $S^3/\Gamma$, where $S^3 \simeq \mathrm{SU}(2)$ and $\Gamma$ is a discrete, fixed-point-free subgroup of $$\mathrm{SO}(4) \simeq (\mathrm{SU}(2) \times \mathrm{SU}(2))/\{\pm1\}$$. Prism spaces, denoted $D_p$, arise when $\Gamma$ is the binary dihedral group $D_p^*$ of order $4p$.

In matrix coordinates, an element of $S^3$ is given by $$u = \begin{pmatrix} z_1 & i z_2 \ i \bar{z}_2 & \bar{z}_1 \end{pmatrix} \in \mathrm{SU}(2)$$, with $z_1 = x_0 + i x_3$ and $z_2 = x_1 + i x_2$. $D_p^*$ is generated by two left-actions:

• $$g_{a1} = \text{diag}(e^{-i\Psi_{az}}, e^{i\Psi_{az}}),\ \Psi_{az} = 2\pi \cdot (2/p)$$ (rotation about $z$-axis)
• $g_{a2} =$ rotation by $2\pi \cdot (1/4)$ about the $y$-axis

The deck group is $$\Gamma = \langle (g_{a1},1),(g_{a2},1) \rangle \simeq D_p^*$$, making pure prism spaces homogeneous.

Prism double-action manifolds generalize this construction to $M_{p,n} = S^3/(D_p^* \times Z_n)$, where $Z_n$ is a cyclic group acting on the right. The requirement $gcd(p,n) = 1$ ensures a freely acting group. The order is $|D_p^* \times Z_n| = 4pn$ (Aurich et al., 2012, Aurich et al., 2012).

2. Fundamental Domain and Well-Proportioned Conjecture

For any observer at $x_0 \in S^3$, the fundamental domain (Voronoi cell) $F$ is defined as

$$F = \{ x \in S^3\:|\:d(x_0, x) \le d(x_0, g \cdot x)\ \forall g \in \Gamma \}$$

with geodesic distance $d(x,y) = \arccos(\langle x, y \rangle)$. For prism spaces, $F$ takes the form of a "spherical prism"—bounded by $p$ vertical planes and two p-gonal "caps." In double-action cases, the Dirichlet domain can interpolate between various polyhedral shapes depending on the observer position (Aurich et al., 2012, Aurich et al., 2012).

The well-proportioned conjecture (Weeks et al.) posits that spaces whose cells are equally extended in all directions (minimizing the variance $\sigma_\tau^2$ of the cell radius $\tau(\hat n)$) would yield maximal suppression of large-scale CMB power. However, evidence demonstrates that identical domain shapes can correspond to distinct CMB statistics, directly challenging this conjecture and showing that cell geometry alone is insufficient to predict CMB properties (Aurich et al., 2012).

3. CMB Spectra and Mode Suppression in Prism Spaces

The Laplace–Beltrami eigenmodes on $S^3$ are labeled by $j \in \mathbb{N}_0$ with eigenvalues $E_j = 4j(j+1) = \beta^2 - 1$. Topological identification projects out non-invariant modes, setting selection rules:

• $m_a \equiv 0~(\text{mod}~p/4)$ from dihedral action,
• $2m_b \equiv 0~(\text{mod}~n)$ from the cyclic action.

Only those eigenmodes satisfying these conditions survive in $\mathcal{H}_{S^3/\Gamma}$ (Aurich et al., 2012, Aurich et al., 2012). The suppression of long-wavelength (low-$\beta$) modes leads to a distinctive cutoff in the angular two-point correlation,

$$C(\theta) = \sum_\ell \frac{2\ell + 1}{4\pi} C_\ell P_\ell(\cos\theta)$$

with multipoles

$$C_\ell = \sum_{\beta} P(\beta) T_\ell^2(\beta) \frac{r^M(\beta)}{\beta^2}$$

where $r^M(\beta)$ is the multiplicity and $P(\beta) \propto 1/(E_\beta \beta^{2-n_s})$ with $n_s \simeq 0.96$. Studies reveal that for large $p$, the ratio $S_{D_p}(60^\circ)/S_{P^3}(60^\circ) \sim 1/p^2$ (Aurich et al., 2012).

4. Quantitative Results and Comparative Suppression

Extensive numerical surveys of prism double-action spaces ($4pn \leq 180$) show that some configurations, particularly $DZ(8,3)$, $DZ(16,3)$, and $DZ(20,3)$, yield strong suppression of $S(60^\circ)/S_{S^3}$, reaching as low as $0.29$ at $\Omega_{\text{tot}}\approx1.036$ for $DZ(16,3)$ and $0.33$ at $\Omega_{\text{tot}}\approx1.03$ for $DZ(8,3)$. These results are competitive with, but do not surpass, the suppression achieved in regular polyhedral (Platonic) spaces such as the Poincaré dodecahedral space ($I^*$) (Aurich et al., 2012, Aurich et al., 2012).

Notably, the physical scale of suppression is set by the injectivity radius $r_{\text{inj}} = \min(\pi/p, 2\pi/n)$, restricting observable correlations at large angles ($\theta \gtrsim 60^\circ$).

Comparison with lens spaces $L(p,q)$ reveals counter-examples: even when the Voronoi cells coincide, the CMB statistics can differ, due to differences in the multiplicities and structure of cyclic Clifford subgroups, underscoring the need to consider not just cell geometry but also face-identification rules (Aurich et al., 2012).

5. Observational Implications and Circle Searches

The prism hypothesis naturally predicts the suppression of long-wavelength correlations via topological mode truncation, with the injectivity radius determining the cutoff scale. For the best-case prism spaces, the fundamental cell fits multiple times across the last scattering surface, accounting for the deficit of correlations at $\theta > 60^\circ$.

Such spaces generically predict a small ($\sim$2–4) number of nearly antipodal "circle-in-the-sky" pairs with radii $20^\circ$–$40^\circ$, a key observational signature at the threshold of current search sensitivity. Non-detection of such circles would require $\Omega_{\text{tot}}$ to be closer to unity or would exclude particular $(p,n)$ values (Aurich et al., 2012).

6. Critical Evaluation and Limitations

Analysis across multiple prism and lens spaces demonstrates that the well-proportioned criterion ($\sigma_\tau$ minimization) is neither necessary nor sufficient to guarantee maximal large-angle CMB suppression. The full eigenmode structure, the action of Clifford translation subgroups, and the multiplicities of these subgroups play decisive roles. Group-theoretic insights, particularly the classification of cyclic subgroups and their action, offer a deeper explanatory framework for the observed suppression.

Viability of specific candidate topologies is sensitive to current measurement bounds on curvature ($\Omega_{\text{tot}} \approx 1.002 \pm 0.02$). The sweet-spot for the strongest prism-suppression ($\Omega_{\text{tot}} \approx 1.036$) is marginal given cosmological bounds, with more realistic candidates ($DZ(20,3)$ at $\Omega_{\text{tot}}\approx1.02$) remaining viable.

No known double-action prism topologies produce strictly greater large-angle CMB suppression than the best Platonic spaces (notably the Poincaré dodecahedral space), though further classification and phenomenological analysis is ongoing (Aurich et al., 2012, Aurich et al., 2012).

7. Summary Table: Notable Prism Double-Action Manifolds and CMB Suppression

Manifold $DZ(p,n)$	Group Order $4pn$	$S_{\min}/S_{S^3}$	$\Omega_{\text{tot}}$ at Minimum
(8,3)	96	$\approx 0.33$	$\approx 1.03$
(12,3)	144	$\approx 0.40$	$\approx 1.02$
(10,3)	120	$\approx 0.45$	$\approx 1.015$
(16,3)	192	$\approx 0.29$	$\approx 1.036$
(20,3)	240	$\approx 0.42$	$\approx 1.02$

These examples demonstrate that prism double-action manifolds provide viable multi-connected candidates for explaining the CMB large-angle anomaly, subject to geometric, topological, and observational constraints (Aurich et al., 2012, Aurich et al., 2012).

---

References

• Aurich & Lustig, "How well-proportioned are lens and prism spaces?" (Aurich et al., 2012)
• Aurich & Lustig, "Cosmic Topology of Prism Double-Action Manifolds" (Aurich et al., 2012)
• Aurich & Lustig, "Cosmic Topology of Polyhedral Double-Action Manifolds" (Aurich et al., 2012)