The Domain Wall
Field Note — Observation Log #447
The field collapsed everywhere and chose, at random, one of two equally valid vacuum states. Left side of the patch chose +v. Right side chose −v. Somewhere between them, a wall formed — a sheet two-dimensional in extent, infinitely thin in theory, separating two universes living in the same space.
A domain wall is what happens when a discrete symmetry breaks. Not a continuous rotation, not a gauge choice, but a binary decision: this side or that side, up or down, particle or antiparticle. The field cannot compromise. A domain wall is the scar left by indecision on a cosmological scale.
Energy budget
The surface energy density σ of a domain wall is enormous. Roughly σ ∼ η³ where η is the symmetry-breaking scale. For GUT-scale physics, this translates to about 10²⁴ GeV/cm² — an amount of energy per unit area that should make any cosmologist nervous. And it should.
Unrestrained domain walls dominate the energy density of the universe faster than matter, faster than radiation. Their tension scales as a⁻¹ while radiation scales as a⁻⁴ and matter as a⁻³. Dominate they do. This is called the domain wall problem, and it is fatal: a universe with even one domain wall spanning a significant fraction of its volume would have torn itself apart before stars could form.
So either domain walls never formed, or something erased them. Inflation, perhaps, sweeping them beyond the observable horizon. Or an explicit bias term in the potential — a tiny energy difference between the two vacua — that made one side "win" everywhere and caused the wall to expand and collapse.
What we see (or don't see)
CMB anisotropy maps put the tightest constraints on domain walls. A wall crossing the line of sight would imprint a step discontinuity in the CMB temperature — a sharp edge in the cosmic background radiation. The WMAP and Planck collaborations have hunted for these step edges across the full-sky maps and found nothing statistically significant.
The upper limit on the wall energy density translates to η < 10 MeV for simple models. For GUT-scale theories, this rules out the simplest domain wall scenarios entirely. But the constraint is empirical, not theoretical — meaning the wall still exists in the space of allowed physics, just not at the scales we can currently detect.
Interesting edge cases
Textured walls: If the field responsible for the domain wall is coupled to a field with non-trivial topology, the wall can develop texture — twisting, knotted structures that evolve nonlinearly. These are messier, harder to constrain, and more interesting.
Semi-permeable walls: In models with multiple scalar fields, domain walls can be permeable to certain particles but not others. This creates a natural filter — a cosmic membrane that selects which species of matter can cross. A wall between two vacuum states with different particle masses would have fascinating phenomenology.
Colliding walls: When walls from different symmetry-breaking transitions intersect, they produce shock fronts, particle production, and gravitational wave bursts. The collisions are computationally expensive to simulate and observationally unconstrained, which is to say they are the kind of thing that keeps theorists employed.
Why it matters
Domain walls are the simplest topological defect after monopoles. They sit between monopoles (0-dimensional) and strings (1-dimensional) in the classification, and they carry important information about the structure of the symmetry-breaking potential. Detecting one — even indirectly — would tell you the symmetry group broke via a discrete quotient, not a continuous one. That narrows the landscape of viable GUTs by orders of magnitude.
But they probably don't exist. Not in the observable universe, at least. The fact that their absence is one of the strongest constraints we have on early-universe physics makes their hypothetical presence more powerful than any actual detection.
A wall between two vacua. Two universes sharing one sky. And we may never know which side we're on.