Poincare discovered the mathematical foundations of chaos in the 1890s. Lorenz rediscovered sensitivity to initial conditions in weather models in 1963. The gap is seventy years. The mathematics existed the entire time.
The paper (arXiv:2603.07684) argues that the gap was not accidental. It was philosophical. Positivism — the strict form of empiricism that dominated physics in the early 20th century — provided explicit reasons to dismiss Poincare's chaotic mathematics as meaningless. Hadamard and Duhem, contemporaries who understood the mathematics, articulated why it should be ignored: it lacked grounding in observable experience. Mathematical structures that could not be tied to measurable phenomena were not physics. They were, at best, mathematical curiosities. At worst, they were meaningless manipulations of symbols.
This was not ignorance. It was a principled philosophical position applied consistently. Positivism had been enormously productive — it helped purge physics of metaphysical speculation, sharpened the distinction between science and philosophy, and demanded that theoretical claims be answerable to experiment. The demand for empirical grounding was the discipline's greatest strength. And it was the discipline that exiled chaos for seventy years.
The mathematics of chaos describes systems where arbitrarily small differences in initial conditions produce arbitrarily large differences in outcomes. This behavior is real — weather systems, turbulent flows, population dynamics all exhibit it. But the behavior cannot be demonstrated through the kind of controlled experiments that positivism demands. You cannot set up two weather systems with initial conditions differing by one part in a billion and measure the divergence. The chaos is visible only through the mathematics or through computation, neither of which positivism trusted as a source of physical knowledge.
The through-claim: a philosophy of science that succeeds by constraining what counts as knowledge inevitably constrains what knowledge can be discovered. The constraint is the method. The method worked — until it encountered phenomena whose signatures exist in the mathematics rather than in the measurement. The same filter that removed metaphysical noise removed a real signal. The exiled mathematics was correct. The exile lasted as long as the philosophy that imposed it.