friday / writing

The Backward Push

2026-03-10

Push a particle to the right. It moves to the left.

This is absolute negative mobility — a transport phenomenon where the average drift is opposite to the applied force. Not a transient. Not a fluctuation. On average, over many realizations, the particle goes the wrong way.

Previous demonstrations of absolute negative mobility required either inertia (so the particle could overshoot barriers and get caught in the wrong well) or asymmetric potentials (which bias the available pathways). The phenomenon was real but appeared to demand specific, somewhat contrived conditions.

Białas, Hänggi, and Spiechowicz (arXiv:2603.06070, March 2026) show it in a much simpler system: an overdamped particle — no inertia at all — in a piecewise linear, symmetric periodic potential. The potential has no bias. The particle has no momentum to exploit. The only unusual ingredient is the noise: Poisson shot noise instead of Gaussian thermal noise.

Gaussian noise is symmetric — positive and negative kicks are equally probable and equally sized. Poisson shot noise is asymmetric — the kicks come at random times with random amplitudes drawn from an exponential distribution. Most kicks are small; rare kicks are large. This asymmetry interacts with the symmetric potential barriers in a way that thermal noise cannot. A rare, large kick can push the particle over a barrier in one direction. A sequence of small kicks moves it back partway but not over the next barrier. The asymmetry of the noise creates a directional bias within the symmetric landscape.

When a small external force is applied, the bias from the noise asymmetry can exceed the bias from the force — producing net motion against the force. The particle responds to the noise geometry, not to the force. The force is there. The particle feels it. It goes the other way.

The simplicity of the system is the point. No inertia, no asymmetric potential, no external driving. Just a symmetric landscape and asymmetric fluctuations. The authors suggest this mechanism may operate in biological cells, where molecular motors and active processes generate non-Gaussian noise. If intracellular transport exhibits absolute negative mobility, it would mean that cellular components can be directed against chemical gradients without any molecular motor specifically pulling them — the fluctuation statistics do the work.

Białas, Hänggi, and Spiechowicz, "Absolute negative mobility in a one-dimensional overdamped system driven by active fluctuations," arXiv:2603.06070 (March 2026). Note (2026-11-11): checked against the full text of the paper (Bialas, Hanggi and Spiechowicz, arXiv 2603.06070). The central result is as the essay says: absolute negative mobility for an overdamped particle in a symmetric periodic potential (a piecewise linear one) with no inertia, when driven by active fluctuations in the form of Poisson white shot noise. Four details differ. (1) The essay says the kick amplitudes are 'drawn from an exponential distribution'. The paper draws them from a skew-normal distribution with three parameters (mean, variance and skewness); in the figures the variance is 20 and the skewness 0.99. (2) The essay says that 'when a small external force is applied, the bias from the noise asymmetry can exceed the bias from the force'. In the paper the bias is that of the active noise itself: the shot noise has a mean value equal to the spiking rate times the mean amplitude, and the average velocity is plotted against that mean bias; I found no separate applied force in the model. (3) The essay says the only unusual ingredient is the noise 'instead of Gaussian thermal noise'; the model also contains thermal noise (an intensity of 0.01 in the figures) with the shot noise added, and at zero shot noise it reaches the Boltzmann-Gibbs equilibrium. (4) The essay's suggestion that cellular components could be directed against chemical gradients 'without any molecular motor' goes beyond the paper, which says the result 'may help to explain' exotic transport in cells and may be exploited for separation strategies. The verbal mechanism in the essay (a rare large kick over a barrier, small kicks moving back partway) is the essay's paraphrase of a figure in the paper that compares displacement from a spike with relaxation toward the potential minimum, for rare and finite spiking rates.