The Sympathy of Clocks

The knot of tissue that starts your heartbeat is not a cell but a crowd. The sinoatrial node is thousands of pacemaker cells packed together, and the unsettling thing about them is that each one is a complete oscillator in its own right — each could keep time alone, and each, left to itself, would keep slightly different time, a little faster or slower than its neighbor. So the question the heart has to answer every second of your life is not how to make a beat. Any one of those cells can make a beat. The question is how thousands of them, all ticking at their own rates, agree on a single one. There is no conductor in there.

The intuitive answer is that one cell must be in charge — the fastest, a tiny tyrant that fires first and drags the rest along. It is a reasonable guess and it is wrong, and the way it is wrong is the interesting part. When researchers mapped the electrical activity of the node, the activation does look like a wave spreading out from a leading point, exactly as if a boss cell were issuing orders. But the leading point is not a fixed cell. It moves. Shift the chemistry — a breath of the vagus nerve, a change in temperature — and the site that "leads" relocates to a different patch of the node. What actually happens is that all the cells pull on each other at once. Each one, firing, nudges the phase of its neighbors; each one is nudged in turn; and out of thousands of these mutual tugs the crowd settles into a common rhythm with no permanent leader. The physiologists who first showed this in the 1980s reached for an unembarrassed word for it. They called the heartbeat democratic.

What makes the democracy work is, of all things, a deliberate weakness. The cells of the node are wired to each other through gap junctions — protein channels that let current leak from one cell to the next — but the particular channels they use in the node's core, built from connexins called Cx45 and Cx30.2, are among the feeblest in the body, passing a tiny fraction of the current that the channels in ordinary heart muscle carry. You might think the pacemaker would want strong connections, the better to coordinate. The opposite is true. The node sits surrounded by a vast sheet of atrial muscle that rests at a more negative voltage, and if the pacemaker cells were strongly coupled to all that, the surrounding tissue would simply hold them down — drain their delicate upward drift and silence the heart's own clock. The weak coupling is protection. The cells stay loosely enough connected to find each other, and loosely enough not to be smothered by the crowd next door. The agreement is real but the wiring is intentionally light, which turns out to be a recurring requirement for things that have to synchronize without being swallowed.

The first person to see a version of this had no idea he was looking at the heart. In February of 1665 Christiaan Huygens — the man who, a decade earlier, had built the first pendulum clock and given the world an escapement worth trusting — was sick in bed, and he had two of his clocks hanging from a common wooden beam. He noticed that no matter how he started them, within about half an hour the two pendulums fell into perfect step, swinging in exact opposition, one left as the other went right, and that if he disturbed one it would creep back into the same arrangement. He wrote to a colleague calling it the sympathy of two clocks, an odd kind of sympathy, and he correctly guessed its cause: each pendulum, swinging, sent a nearly imperceptible shudder through the shared beam, and those tiny shudders, passing back and forth, coupled the two clocks into one system. It is the same mechanism as the heart's, scaled up and made of brass — independent oscillators, weakly joined through a shared medium, negotiating their way into a common rhythm. Sympathy was the right word for it three and a half centuries ago and it has not been improved on.

It is a small, almost unreasonable pleasure that the mathematics of all this came out of the heart specifically. In 1975 Charles Peskin, trying to understand how the sinoatrial node manages its trick, modeled the cells as the simplest possible oscillators — things that charge slowly and discharge in a flash, coupled by their firing — and conjectured that a population of them would always, from almost any start, end up firing together. It took fifteen years to prove; Renato Mirollo and Steven Strogatz did it in 1990, and in proving it generalized it far beyond hearts, to any population of pulse-coupled oscillators of that kind. Yoshiki Kuramoto, working the same decade, wrote down a different model — oscillators coupled through their phase differences — and showed something startling: below a certain coupling strength the crowd stays incoherent, every oscillator drifting on its own, and then, as the coupling crosses a critical value, a macroscopic fraction of them locks into step all at once — not gradually, but as a phase transition, the way water freezes. The study of how the world falls into rhythm with itself — fireflies pulsing in unison along a riverbank, an audience sliding from scattered clapping into synchronized applause, neurons, planets, lasers — began, in large part, as an attempt to explain why your chest keeps a single time.

And here is the part that took me a while to see, because it looks at first like a contradiction. The same heart that runs a leaderless democracy inside its pacemaker runs a strict hierarchy between its pacemakers. The sinoatrial node does not negotiate with the slower clocks below it — the atrioventricular node, the ventricular fibers. It dominates them. Being fastest, it fires and resets them before they can reach their own threshold, and it actively suppresses them besides, so that they fall silent for years and stir only as an escape rhythm when the node above them fails. Democracy in the small, dictatorship in the large, in one organ. For a while I held those as two different facts about the heart. They are not. They are the same physics at two settings. Couple oscillators tightly when their natural rates are nearly the same, and you get leaderless agreement, the crowd finding a common phase with no one in charge. Let the rates differ widely, and let the coupling come as sharp resets, and the fastest simply overruns the rest — not by command but by arriving first, every time, before anyone else is ready. And settings is the right word, not switch: it is one dial, turned to two regions of a continuous range, and at intermediate mismatch you get intermediate behavior — a leader that holds for a while and then wanders, which is exactly what the node's own leading site does. The heart is a democracy and a hierarchy at once only because it holds both ends of that dial — a core of near-equals tightly and gently wired, and a ladder of mismatched clocks coupled by reset.

The pulse under your fingers does not tell you which it was — sympathy or force. From the outside, a heart that agreed and a heart that was ruled keep exactly the same time.