Mathematics has a new object for describing one of democracy's most stubborn properties: a bottle with no inside.

Three researchers have shown that Condorcet's paradox — the 240-year-old observation that a majority can vote itself into a logical loop — has the shape of a Klein bottle, the surface with no distinct inside or outside. Quanta Magazine reports the result, from Ori Livson, a doctoral student at the University of Sydney, his adviser Mikhail Prokopenko, and Siddharth Pritam of the Chennai Mathematical Institute. Livson began in April 2025 by looking at the two ends of a model that Yuliy Baryshnikov, now a professor at the University of Illinois, Urbana-Champaign, had proposed in 1993 and deliberately left open.

The problem itself is old. In 1785 Nicolas de Condorcet noticed that with three or more candidates, every individual ballot can be perfectly coherent while the tally runs in circles. One third of voters rank Alice over Bob over Carol; one third rank Bob over Carol over Alice; one third rank Carol over Alice over Bob. Compared pairwise, Alice beats Bob, Bob beats Carol, and Carol beats Alice. Nobody beats everyone.

Kenneth Arrow showed in 1951 how deep this runs. Any method for combining individual rankings into a collective order must sacrifice at least one reasonable democratic requirement — no dictator, independence from irrelevant alternatives, unanimity. No choice of voting system escapes it; the choice is only which condition to fail. In 1980 the mathematician Graciela Chichilnisky mused in a footnote that Condorcet's case in particular might be a topological problem. It turns out she was right.

The construction is visual. A single ballot in a three-candidate race is one of six orderings. Each can be drawn as a triangle whose corners are the three pairwise comparisons the ballot implies. Two orderings that agree on two comparisons produce triangles sharing an edge, so they glue together. Glue all six and you get a strip; its two ends also match, so bend it round into a tube with an opening at each end.

Baryshnikov used that tube to give a new topological proof of Arrow's theorem. He ignored the rims. But the rims are precisely the Condorcet cycles — around one, the comparisons read Alice over Bob, Bob over Carol, Carol over Alice; around the other, the cycle runs in reverse. Livson wanted a model that distinguishes cases where such cycles are permitted, as in rock-paper-scissors, from cases like elections where they are not.

Both rims say the same impossible thing: every candidate ranks above every other. In topology, once you know two things are the same, you may glue them. The obvious move — bend the tube until the openings meet, producing a torus — fails, because it pairs edges carrying incompatible claims: an edge meaning "Bob over Alice over Carol" ends up glued to one meaning "Bob over Carol over Alice." The seam would manufacture the contradiction the model exists to avoid.

The fix is a twist. Rather than matching the rims point for point in the same order, the researchers reversed the order of preferences around one rim, making the two cycles equivalent. "Gluing with the twist is the only valid option," Livson said. Twist a strip and join its ends and you get a Möbius strip. Twist a tube and you get something stranger: a Klein bottle, one-sided, edgeless, with no distinguished inside. In three dimensions it crosses through itself; mathematicians build it in four, where it closes cleanly.

The analogy is the point. Each voter supplies a coherent ranking, yet the electorate's preferences refuse to fit into a single order. Each small patch of a Klein bottle appears to have two sides, yet those sides cannot be distinguished consistently across the whole surface. Something that makes sense piece by piece breaks down when the pieces are assembled.

That raises the larger question the trio is chasing: whether other fundamental limits in mathematics, when given a shape, turn out to share a deeper geometric explanation.

For anyone actually designing elections, nothing changes. Choosing a voting system still means choosing which reasonable requirement to sacrifice; topology makes the trade-off tangible without resolving it. "It's not really a mathematical debate," said Iain McLean of the University of Oxford. "It's a debate about values." The bottle shows why the loop cannot be defined away.