How does the scalar potential's functional form affect which shape functions become dynamically accessible?

This is a sophisticated technical question about how the mathematical form of a scalar potential function constrains which dynamical solutions (called "shape functions" or equilibrium configurations) become physically realizable. Let me work through what the archive says.

The core mechanism

The key insight is that different potential shapes have different *symmetries and asymptotic behaviors*, which determine which critical points (equilibria) can exist in the dynamical system, and which trajectories can actually connect them [AG-2026.04-427].

When you formulate cosmological equations with a scalar field as a dynamical system (using techniques like expansion-normalised variables), the form of V(φ) directly shapes the geometry of the phase space. Crucially, [AG-2026.04-427] shows that viable cosmological histories are often confined to invariant planes or manifolds—special lower-dimensional subsets of the full phase space. For hybrid power-law-exponential potentials, trajectories connecting a matter-dominated era to late-time acceleration live on a two-dimensional invariant plane containing both the matter point and the accelerated attractor point.

Potential shape determines accessible attractors

Quadratic potentials behave very differently from hybrid or power-law forms [AG-2025.11-338]. For quadratic V = Λ ± ½m²φ², the dynamical system admits both slow-roll and hilltop quintessence solutions, each with distinct global structures. The different signs and functional forms yield different sets of accessible equilibrium points.

More generally, [AG-2025.10-345] shows that power-law potentials V ∝ φⁿ act as "shape-shifters": depending on the exponent n and the epoch, the same field mimics different matter types (radiation, dust, domain walls, phantom matter), each with distinct dynamical evolution. The potential's exponent directly selects which barotropic equations of state become dynamically accessible.

Near universality at special limits

Interestingly, [AG-2025.12-109] reveals that *near the massless limit*, the detailed potential shape becomes less important. At a codimension-two organizing centre, the dynamics reduces to a two-parameter unfolding controlled by geometric parameters (tilt and curvature) rather than the full potential. This means that for near-quadratic potentials, standard slow-roll parameters emerge as *universal functions* of just two variables, independent of the specific V(φ) choice—but only in that narrow regime.

The relevance: understanding how potential form gates which solutions are accessible is essential for both inflationary model-building and dark-energy cosmology, because observations can only see trajectories that the dynamics actually permits.

---

The preprints focus most directly on cosmological scalar fields; they don't address whether the question refers to field configurations in other contexts (e.g., solitons, topological defects, or quantum mechanics), so coverage here is specialized to gravity and cosmology.