VorticesWhirlpools in a field of phases — the defects the whole transition is about

Fill a plane with oscillators, one at every point, each carrying a phase — an arrow on its own little clock face. If the coupling is doing its job, neighbors mostly agree, and the field of arrows looks like a well-combed lawn. But there is one kind of disturbance that combing cannot fix: a point the arrows circulate around. Walk a small loop around such a point and the phase you carry turns through one full 2π — the loop has a winding number. You drew exactly this integer with your own hand in the topology module; here it stops being a game and becomes a physical object. A point the field winds around is called a vortex, and every claim in this module — and the wall the whole library is built on — is about what a population of these costs and does.

1. Place them and look

The figure below is a field of phase arrows (hue underneath repeats the same information, since phase is a color wheel as naturally as it is an angle). Click to drop a vortex with winding +1; toggle the button to drop −1 instead (an anti-vortex — the arrows circulate the other way). The readout measures the winding of the phase carried around the border of the whole box: it always equals the sum of the charges you placed inside. Individual arrows change everywhere with every vortex you add — but that one integer is doing bookkeeping no local rearrangement can cheat.

Click anywhere to place a vortex. Arrows show the phase field; the background hue is the same field as color. The readout integrates the phase around the box border — it reads exactly the total charge inside, however the field between rearranges. border winding = 0 · charges placed: none

2. Try to comb it away

Why call this a defect rather than just a messy spot? Because of what happens when you let the field relax. Relaxation is the coupling doing what coupling does: every arrow turns, a little at a time, toward the average of its neighbors. Smooth ripples iron out and their energy drains toward zero. A vortex does not. The arrows near it can shuffle, the core can shift, but the winding is an integer counted around any loop — and an integer cannot change by a sequence of tiny moves. To remove it you would have to tear the field somewhere, which coupling will not do. This is topological protection: the same reason your drawn loop in the topology module could not unwind without crossing the hole.

Left: a smooth random ripple. Right: one vortex. The same relaxation runs on both. The ripple's gradient energy drains toward zero — combed flat. The vortex's energy falls to a floor and stops, and its winding stays exactly 1: there is no continuous path from "winds once" to "winds zero times". press relax

3. The one legal way out: meet your opposite

There is exactly one way a vortex can die, and it is the way that matters for everything in the next section: find an anti-vortex. A +1 and a −1 close together look, from far away, like nothing at all — walk a big loop around the pair and the windings cancel; the far field barely knows they exist. Drag them together and they annihilate outright, leaving a combed lawn. Drag them apart and you must pay for every step: the field between them stays torqued. That cost of separation is the entire subject of section 02.

A vortex (orange core) and an anti-vortex (purple core). Drag either. The meter reads the disturbance the pair leaves on the far border: tight pair — the outside barely ripples; pulled apart — the disturbance grows. Bring them within touching distance and they annihilate. drag a core
What's real here These fields are ideal textbook constructions — phase patterns written directly by formula on a small grid, so you can play with the topology bare-handed. Nothing is simulated thermally yet; that begins in section 02, where a live finite-temperature lattice takes over. Vortices themselves are entirely real physics: they are the defects of the 2-D XY model (superfluid films, arrays of coupled oscillators, the site's lattices), and the winding-number bookkeeping shown here is exact, not illustrative. In the essays these same objects appear as the defects that set the wall (The Physics of Mind §05) and as the topologically protected structures the Coherent Learning Rule must not starve (Foundations §03's Fiedler channel).