The perforator hypothesis: knots as small arteries held shut where they pierce the fascia

Position paper · preprint · not peer reviewed. Anyone with a foam roller can find them: small, tender places in the soft tissue that persist for years yet can let go within seconds, sometimes with a tingle across a patch of skin. No mechanism for them is established, and no release has been recorded as it happens. We propose one. A knot is a cutaneous perforator, a small artery travelling with its veins and a nerve through an opening in the fascia, held shut by the smooth muscle of its own wall. Each part rests on measured physiology. A toned small artery can snap shut and stay shut until its tone falls well below the level that shut it; held shut for hours, it relaxes only slowly and incompletely; the breath moves the skin's sympathetic drive within seconds; a squeeze and release widen a small artery; a nerve deprived of blood fires when blood returns; and a vessel held shut holds shut the branches it feeds. A knot, then, is a memory with an anatomy: written by the sympathetic drive, kept by the wall, read by the perforator's own nerve. We argue that it is now the strongest account in a specific sense: it explains more of what is reported, with less that is new, than six alternatives in their strongest forms. We state what it does not explain, weigh the claim that acupuncture points sit on perforators, consider the slower fascial changes reported after intensive practice, and their dangers, and give the measurements that would refute it. Caution. This paper describes; it does not prescribe. Nothing in it is a practice or a protocol. Authorship and AI use. Written by AI (Claude Opus 5.5, Anthropic) under the author's direction, and refereed by a second model (Claude Fable 5.1, Anthropic), which checked each claim against its source. Eugene Jhong conceived the hypothesis and directed the work. The author is an independent researcher. It supersedes, where they differ, the earlier preprint Knots as Pattern Memories (doi:10.5281/zenodo.22709006), which held the knot shut by a hyaluronan collar rather than by the vessel's wall. Web version, models and code: https://ejhong.github.io/knots/research/perforator-hypothesis/ and https://github.com/ejhong/knots. Correspondence and objections: https://github.com/ejhong/knots/issues