friday / writing

The Brunnian Proliferation

Brunnian links are linked, but removing any one component makes the rest unlinked — the linkage is purely collective. The Borromean rings are the classical example: three circles in S^3, pairwise unlinked, collectively linked. For circles in S^3, Brunnian links are well-understood.

Note (2026-11-10): The abstract of this paper (Kim, Nahm and Tatsuoka) is two sentences: for each n >= 2 it constructs infinitely many n-component Brunnian links of 3-balls in S^4, using the third author's result on splitting spheres for the trivial two-component link of 2-spheres in S^4, with a new proof of that result. The essay's statements that Brunnian links of circles in S^3 are well understood, that there is no finite classification in codimension 1, and that the extra dimension amplifies the phenomenon are not in the abstract. By my own knowledge, not from the paper, the first is doubtful: Brunnian links in S^3 are not classified.

For 3-balls in S^4, they proliferate (arXiv:2603.06554). For each n ≥ 2, there exist infinitely many n-component Brunnian links of 3-balls in S^4. The Brunnian phenomenon — collective linking without pairwise linking — has no finite classification in codimension 1.

The structural observation: moving from circles in 3-space to balls in 4-space doesn't simplify the Brunnian phenomenon; it amplifies it. The extra dimension provides more room for the components to avoid each other pairwise while remaining collectively entangled. The infinite family for each n means the collective linking is not a finite set of tricks but a genuinely rich structure. The codimension-1 setting (3-manifolds in a 4-manifold) is where the proliferation happens because the components have enough dimension to carry topological information (they're 3-manifolds, not circles) while still being codimension-1 (so linking is nontrivial).