Seifert Surfaces and Genus - Main Theorem

The fundamental theorems relating Seifert surfaces to knot invariants establish genus as a central quantity in knot theory.

Theorem

Seifert's Existence Theorem (1934): Every oriented knot $K \subset S^3$ bounds a compact, connected, orientable surface embedded in $S^3$ . Moreover, such a surface can be constructed algorithmically from any knot diagram.

Proof

Given an oriented knot diagram $D$ :

Smooth all crossings following orientation, producing $s$ disjoint oriented circles (Seifert circles)
Each circle bounds a disk in $\mathbb{R}^3$ (by Jordan curve theorem)
Position disks at different heights so they don't intersect
At each original crossing, the two smoothed arcs lie on different disks
Connect these disks with a half-twisted band at each crossing location
The bands are oriented to match the original knot orientation

The result is a compact, connected, orientable surface $F$ with $\partial F = K$ . Connectivity follows from the diagram being connected; orientability from consistent orientation choices.

■

Theorem

Fox-Milnor Genus Bound: For any knot $K$ with Alexander polynomial $\Delta_K(t)$ , $\deg(\Delta_K) \leq 2g(K)$

For fibered knots, equality holds: $\deg(\Delta_K) = 2g(K)$ .

This provides a computable lower bound on genus using the easily-computed Alexander polynomial. The bound is sharp for important knot families including torus knots and many alternating knots.

Theorem

Genus Formula for Torus Knots: For the $(p,q)$ -torus knot with $\gcd(p,q) = 1$ and $p, q > 0$ : $g(T(p,q)) = \frac{(p-1)(q-1)}{2}$

This exact formula shows genus grows quadratically with torus knot parameters.

Proof

The torus knot $T(p,q)$ can be represented as a closed braid with $q$ strands and $p(q-1)$ crossings. Applying Seifert's algorithm:

Number of Seifert circles: $s = q$
Number of crossings: $c = pq - p$ (from braid representation)
Genus formula: $g = \frac{c - s + 1}{2} = \frac{pq - p - q + 1}{2} = \frac{(p-1)(q-1)}{2}$

To show this is minimal, use Alexander polynomial: $\Delta_{T(p,q)}(t) = \frac{(t^{pq}-1)(t-1)}{(t^p-1)(t^q-1)}$ has degree $pq - p - q + 1 = 2g$ , matching the Fox-Milnor bound. Since torus knots are fibered, the bound is tight.

■

Example

For $T(3,5)$ : $g(T(3,5)) = \frac{(3-1)(5-1)}{2} = \frac{2 \cdot 4}{2} = 4$

Verification: $\deg(\Delta_{T(3,5)}) = 15 - 3 - 5 + 1 = 8 = 2 \cdot 4$ . ✓

Theorem

Slice-Ribbon Conjecture: A knot $K$ is slice (bounds a disk in $D^4$ ) if and only if it is ribbon (admits an immersed disk in $S^3$ with only ribbon singularities).

Direction $(\Leftarrow)$ is known: ribbon implies slice. The converse $(\Rightarrow)$ remains open and is one of the major unsolved problems in knot theory.

Remark

Implications if true:

Algorithmic decidability of sliceness (ribbon property is combinatorially checkable)
Classification of slice knots via ribbon structures
Deep connections to 4-manifold topology

Recent progress (Piccirillo, 2020) showed the Conway knot is not slice, using sophisticated Khovanov homology arguments, demonstrating modern techniques attacking classical problems.

Theorem

Gabai's Fiber Detection Theorem: A knot $K$ is fibered if and only if it admits a Seifert surface $F$ that is monic in the sense that the Alexander polynomial satisfies $\Delta_K(t) = \det(tV - V^T)$ where $V$ is the Seifert matrix of $F$ and the polynomial is monic (leading coefficient $\pm 1$ ).

This characterizes fibered knots purely in terms of Seifert surface properties, providing a practical criterion.

These theorems establish Seifert surfaces and genus as fundamental tools bridging knot diagrams, algebraic invariants, and deep topological properties of knot complements and 4-dimensional topology.