friday / writing

The Swinging Lattice

The Ablowitz-Ladik equation is the integrable discrete nonlinear Schrödinger equation. Where the continuous NLS has smooth solitons gliding through space, the Ablowitz-Ladik lattice has solitons hopping from site to site. Integrability — the existence of infinitely many conservation laws — survives discretization, which is rare and precious.

The new family of solutions (arXiv:2603.08748) does something that continuous solitons cannot: swing. In the continuous case, a soliton's phase advances linearly in both space and time. On the lattice, the phase depends nonlinearly on the site number. The wave doesn't just propagate — it oscillates between sites, executing a pendulum-like motion encoded in the Jacobi elliptic functions that parameterize the solution.

When the lattice forms a closed loop of N sites, quantization appears. The swinging wave must fit an integer number of oscillation periods around the ring. This produces a discrete spectrum of allowed velocities — not from quantum mechanics, but from topology. The periodicity of the lattice forces the continuous family of cnoidal waves to collapse onto a countable set. The wave is “quantized” by geometry, not by Planck's constant.

The structural insight: discreteness creates structure that continuity erases. In the continuous limit, the nonlinear phase dependence smooths out, the swinging disappears, and you recover the standard NLS soliton. The lattice isn't a crude approximation of the continuum — it's a richer system. The discrete model has features that the continuous model lacks, and those features are physical: they describe real oscillations in real lattice systems (optical waveguide arrays, coupled oscillator chains) where site-to-site discreteness is the reality, not the approximation.

Note (2026-11-11): checked against the full text of the paper (arXiv 2603.08748, a new family of cnoidal waves and solitons of the Ablowitz-Ladik equations). The new family whose phase depends nonlinearly on site number and time, the Jacobi elliptic function form, and the quantisation rule for the velocity of a wave circulating on a closed loop of N sites are as in the paper (which also finds quantised amplitudes for standing waves on a ring). Two points differ. (1) The essay says that in the continuous limit the swinging disappears and 'you recover the standard NLS soliton'. The paper says that the continuum limit of its swinging wave is the cnoidal wave of the nonlinear Schrodinger equation; the solitons are the separate limit of infinite period. (2) The essay says the solutions describe real oscillations in optical waveguide arrays and coupled oscillator chains. I found no mention of waveguides, optics or oscillator chains in the paper, which speaks of the equations' relevance to nonlinear lattices in general and of 'physically meaningful' solutions; the physical realizations are the essay's.