Menger's Theorem states that the maximum number of edge-disjoint paths between two vertices equals the minimum number of edges whose removal disconnects them. This theorem reveals that connectivity is dual to separability; it follows as a consequence of max-flow min-cut by translating path-disjointness into flow problems.
From max-flow min-cut, you know the deep duality between flow and cuts: the most you can send equals the least you can block. Menger's Theorem expresses this same duality in purely combinatorial language, stripping away the numerical flow values and asking a yes/no structural question: how well-connected are two vertices, really?
Two paths from u to v are edge-disjoint if they share no edge — they may share intermediate vertices, but every edge is used by at most one path. The maximum number of edge-disjoint paths from u to v is a measure of how "many independent routes" connect them. An edge cut separating u from v is a set of edges whose removal destroys every path from u to v. Menger's Theorem says these two quantities are equal: the max number of edge-disjoint paths = the min size of an edge cut. This is not obvious! It says that whatever structural bottleneck limits the number of independent paths (the cut) is exactly as tight as the number of paths you can actually find.
The proof is essentially max-flow min-cut in disguise. Assign capacity 1 to every edge. An integer flow of value k from u to v corresponds to k edge-disjoint paths (flow conservation forces the flow to split into disjoint path-flows when capacities are 1). A cut of capacity k corresponds to k edges whose removal disconnects u from v. Max-flow min-cut then gives max-paths = min-cut directly. This reduction — translating a combinatorial statement into a flow problem — is a template for much of combinatorial optimization: bipartite matching, Hall's theorem, and many other results all follow from the same translation.
There is also a vertex form of Menger's theorem: the maximum number of internally vertex-disjoint paths (sharing no intermediate vertex) equals the minimum number of vertices whose removal disconnects u from v. The vertex version follows by a splitting construction: replace each internal vertex v with two vertices v_in and v_out connected by a single edge of capacity 1, forcing every path through v to "use up" that edge. The edge form of the theorem then applies to the split graph. Together, the two forms give a precise characterization of connectivity: a graph is k-connected (remains connected after removing any k−1 vertices) if and only if any two vertices are connected by k internally vertex-disjoint paths. This is a structural fingerprint of resilient networks — whether in communications infrastructure, transportation grids, or biological systems.