A point in the domain of a multivalued complex function where analytic continuation along closed loops around the point produces different values (nontrivial monodromy), so that a single-valued branch cannot be defined on any neighborhood that contains a loop encircling the point.

Principle

A branch point is where monodromy is nontrivial: following analytic continuation around the point permutes function values or sheets of the associated Riemann surface, demanding branch choices or passage to a covering surface.

Demonstration

The function f(z) = sqrt(z) has a branch point at z = 0 because analytic continuation of a chosen square-root value around a circuit encircling 0 changes the sign; only after two circuits does the value return to its start.

Misapplication

Treating a branch point as a removable singularity or as an ordinary point and attempting to define a single analytic value on a punctured neighborhood without specifying a branch leads to contradictions under continuation.

Consequence

Presence of a branch point forces either the removal of closed loops from the domain (branch cuts) or the use of a multi-sheeted Riemann surface; global definitions and integrals must account for monodromy and branch choices.

Reversal

If analytic continuation around the point leaves values invariant, the point is ordinary (non-branching); the absence of monodromy is the inverse situation to a branch point.

Boundary

Branch points concern multivaluedness arising from algebraic or logarithmic dependencies in single-variable complex functions; they are distinct from poles, essential singularities, and from accumulation phenomena in several variables.

Semantic Tension

Distinguish branch points from branch cuts: the branch point is the intrinsic obstruction at a location, while a branch cut is an extrinsic choice in the domain to avoid loops; also separate from isolated singularities that change magnitude but not sheet structure.

Synthesis

A branch point is a location where analytic continuation produces nontrivial monodromy of a multivalued function, requiring branches or a covering surface to restore single-valuedness and altering how integrals and continuations are carried out.