Reducibility of the holonomy of flat manifolds

Definitions and mathematical statement

Let $M$ be a connected closed flat Riemannian $n$-manifold. Its universal cover is Euclidean space and $M\cong{\mathbb{R}}^n/\Gamma$, where $\Gamma$ is a torsion-free discrete cocompact group of Euclidean isometries. Write each deck transformation as $x\mapsto A_\gamma x+b_\gamma$. The holonomy representation is the linear representation \[ \rho:\Gamma\longrightarrow O(n),\qquad \rho(\gamma)=A_\gamma. \] Its image is finite, and reducibility means existence of a real linear subspace $V$ with \[ 0<\dim V<n,\qquad A_\gamma V=V\quad\text{for every }\gamma\in\Gamma. \] This is reducibility of the full holonomy representation, not just triviality of restricted holonomy for a flat connection.

The source asks for an alternative geometric proof of this known reducibility statement. It does not ask to determine whether the theorem is true or to construct a new manifold. The phrase geometric proof has no formal criterion here; no particular allowed or forbidden algebraic tool is prescribed.

The dimension restriction is missing in the source. With the usual proper-nonzero-subspace definition, the one-dimensional trivial representation is irreducible, so the statement needs $n\geq2$ or a special convention for that case. The supplied summary also suggests a center/centralizer argument without hypotheses guaranteeing a nonzero common fixed subspace. Reducibility does not require an invariant vector, and no such proof is inferred from that sketch.

Short English statement

Give a more geometric proof of the established reducibility of full holonomy for closed flat manifolds in the intended dimensions. Clarify the dimension-one exception and the desired meaning of geometric proof.

Sources

• Local UnsolvedMath registry, permanent ID 6800003, exact source code catalogue entry, "Reducibility of the holonomy of flat manifolds"; proposed by Renato Bettiol, City University of New York (Lehman College), 2018. Registry status open, classification OPEN-TRIAGE; the supplied summary identifies the theorem itself as established and the requested proof style as the open task.
• Morgan and Pansu, A List of Open Problems in Differential Geometry (2018), question environment 3, https://www.imo.universite-paris-saclay.fr/~pansu/problems_MTDG.tex. Supplied source-TeX extraction accessed July 29, 2026.
• The supplied summary points to Auslander--Kuranishi and Charlap, Bieberbach Groups and Flat Manifolds, for Bieberbach-group treatments, without precise theorem, edition, or page citations. Its center/centralizer sketch does not provide a complete argument or its necessary hypotheses.
• Provenance flags current status and redistribution terms NEEDS_REVIEW; the source TeX is publicly available on a coauthor's publication page.