State Sum Formula in Kauffman Bracket for Link Diagrams | Math Tutorial

The Jones polynomial is invariant under Reidemeister moves II and III. For move II, expanding both crossings shows that the algebra collapses when setting b = A⁻¹ and d = -A² - A⁻², making the result zero. For move III, after applying the second move invariance, the calculation simplifies to show invariance. Normalization by multiplying by (-A³) to an appropriate power ensures invariance under all three moves.

The normalized bracket polynomial is invariant under all three Reidemeister moves. Under the first Reidemeister move (adding/removing a curl), the normalization factor cancels the q/q^{-1} factor introduced by the smoothing. Under the second Reidemeister move, the two crossings contribute opposite signs that cancel out. Under the third Reidemeister move, the local configuration change doesn't affect the overall value because pulling a strand over or under doesn't change the state contributions.

The Kauffman bracket is a map from unoriented knot diagrams to polynomials in variables a, b, d, defined by two relations: bracket of a crossing equals a times zero smoothing plus b times one smoothing, and bracket of a circle equals d. To be a valid invariant, it must be invariant under Rydemeister moves. Rydemeister 2 and 3 can be verified systematically, but Rydemeister 1 fails by a factor of (-a³) - (-a^(-3)). The writhe, defined as the sum of crossing signs in an oriented link, fails Rydemeister 1 by adding or subtracting 1. Multiplying the bracket by (-a³) raised to the power of the writhe corrects both failures, yielding the Jones polynomial, a genuine knot invariant.

Khovanov homology is a powerful link invariant that categorifies the Jones polynomial by associating a bigraded complex of abelian groups to a planar diagram of a link, where the graded Poincaré polynomial of its homology yields the unnormalized Jones polynomial evaluated at t=-1; this invariant is proven to be independent up to quasi-isomorphism from the specific planar projection through invariance under Reidemeister moves, with the construction involving a cube of resolutions where vertices correspond to circle counts and edges encode circle splitting/merging operations using a Frobenius algebra structure on the vector space V = k⟨+, -⟩.

Khovanov homology is a powerful knot invariant that categorifies the Jones polynomial by replacing the polynomial with a chain complex whose Euler characteristic equals the Jones polynomial; this homology theory provides a graded vector space associated with each knot diagram, where the graded Euler characteristic recovers the Jones polynomial value, offering a more refined invariant that distinguishes knots that the Jones polynomial alone cannot differentiate.

Khovanov homology is a link invariant that categorifies the Jones polynomial, defined as a covariant functor from the category of oriented links in S³ to finite-dimensional F₂ vector spaces, satisfying key properties including invariance under isotopy, tensor product behavior for disjoint unions, and exact triangles for crossings; this construction naturally recovers the Jones polynomial as the alternating sum of dimensions of its homology groups, and the reduced version provides a simpler framework for computations while maintaining essential structural information about links.

This section transitions from the Jones polynomial to its categorification, Khovanov homology. While the Jones polynomial is a knot invariant, Khovanov homology is a b-graded homology theory whose graded Euler characteristic is the Jones polynomial. It associates chain groups indexed by homological and quantum gradings with a boundary map. The key insight is that Khovanov homology groups can be rewritten as the kernel of the Khovanov Laplacian, a self-adjoint, positive semidefinite operator that can be interpreted as a Hamiltonian. This provides a physical interpretation: computing Khovanov homology is equivalent to preparing ground states of a supersymmetric quantum Hamiltonian.

Knot homology should satisfy: topological invariance, natural behavior under connected sums, Euler characteristic recovering known polynomials (like Jones), providing more refined information, and functoriality for cobordisms. The Jones polynomial has a state-sum formula over crossing resolutions, which Khovanov homology categorifies by building a chain complex whose Euler characteristic recovers the polynomial. This construction provides a powerful invariant distinguishing knots with identical Jones polynomials through its richer homological structure.

The Jones polynomial, discovered by Vaughan Jones in 1984, revolutionized knot theory. Originally defined using von Neumann algebras, Louis Kauffman provided an equivalent definition using the bracket polynomial. This invariant distinguished knots that previous methods couldn't, and resolved a 100-year-old conjecture about alternating knots: for alternating knots, the degree difference equals crossing number; for non-alternating knots, it is strictly less. This elegant application demonstrated the power of algebraic topology in solving classical problems.

The Jones polynomial, constructed from the Kauffman bracket using the formula V_L(T) = (-A)^{-3w(D)}⟨D⟩|_{A=T^{1/4}}, provides a powerful invariant for proving Tait's first hypothesis (that reduced alternating diagrams have minimal crossing number) and Tait's second hypothesis (that amphichiral links have zero writhe). The key insight is that for adequate diagrams, the difference between the maximum and minimum degrees of the Jones polynomial equals the number of crossings, while for non-alternating diagrams, this difference is strictly greater, establishing that reduced alternating diagrams achieve the minimal crossing number and that amphichiral links must have zero writhe.

This video presents a geometric proof of Tait's conjectures on alternating knots, which were originally proven using the Jones polynomial. The proof utilizes re-plumbing operations on spanning surfaces—specifically showing that flight moves (diagrammatic operations that relate different alternating diagrams) correspond to re-plumbing moves on the black chessboard surface of an alternating knot. By establishing that any two essential positive definite spanning surfaces are related by re-plumbing moves, the proof demonstrates that all reduced alternating diagrams of a given prime knot are related by flight moves, thereby proving Tait's conjectures about equal crossing numbers and equal writhes. This geometric approach provides deeper insight into the topological nature of alternating knots beyond the algebraic methods used in earlier proofs.

The Jones-Krushkal polynomial generalizes the classical Jones polynomial to links in thickened surfaces by incorporating homological information from the surface, enabling the proof of analogs of the Tait conjectures for surface links and virtual knots; specifically, any reduced alternating diagram on a surface has minimal crossing number, and adequate diagrams on minimal genus surfaces provide canonical representatives for virtual links.

Knot invariants are properties that remain unchanged regardless of how a knot is twisted or tangled. The crossing number is an invariant but only works for the simplest projection. Tricolorability determines if a knot can be colored with three colors following specific rules at crossings. The trefoil knot is tricolorable while the unknot is not. The Alexander polynomial, discovered in 1923, provides a unique algebraic representation for each knot. In 1984, Vaughan Jones discovered the Jones polynomial, which could distinguish more knots. Shortly after, six mathematicians independently discovered the Homfly polynomial with two variables, providing even greater distinguishing power. This polynomial can distinguish knots that other invariants cannot. In 1973, Ken Perko discovered that two knots in Tait's list were actually identical, correcting the count from 166 to 165 knots with 10 crossings.

The Potts model can be formulated using subshifts of finite type, enabling exact solutions through transfer operator methods. The configuration space consists of infinite strings of spin values from a finite alphabet Q, forming a full shift with a natural shift operator. Cylinder sets define the product topology, analogous to real numbers. Interaction energy derives from continuous functions on this topology, building the Hamiltonian through finite-state approximations. The partition function, defined with beta=1/(kT), generates a probability measure on the Borel sigma algebra, transforming the configuration space into a canonical ensemble. Thermodynamic quantities like Helmholtz free energy emerge directly from the partition function. Topological pressure emerges as the logarithm of the leading transfer operator eigenvalue, providing a bridge between dynamical systems theory and statistical mechanics.

The Ising model describes magnetic systems where each spin can align in two directions (+1 or -1), with the Hamiltonian H = -JΣσ_iσ_{i+1} - BΣσ_i, where J > 0 indicates ferromagnetic behavior (parallel alignment favored) and J < 0 indicates antiferromagnetic behavior (antiparallel alignment favored). The partition function for n spins is Z = [2cosh(βJ)]^n, and in the thermodynamic limit, the free energy per spin is f = -kT ln(2cosh(βJ)). The Potts model generalizes the Ising model by allowing each spin to take q possible states, and for the special case q=2, the total number of ground states for n spins with periodic boundary conditions is 2^n.

The Potts model is a statistical mechanics model defined on a graph G with vertices assigned discrete colors (spins) from Q possible values. The energy of a state is determined by counting coincidences: adjacent vertices with the same color contribute to higher energy. The partition function sums over all states weighted by e^(-E/kT), where E is the energy and T is temperature. When Q=2, this reduces to the Ising model. Phase transitions occur when the system crosses critical temperatures, fundamentally changing its physical properties. Baxter proved that certain two-state systems on rectangular lattices exhibit phase transitions.

The Q-state Potts model generalizes the Ising model by allowing spins to take Q values instead of just ±1. The partition function is Z = sum over spin configurations of product over edges of e^(Kδσ_iσ_j), where δσ_iσ_j=1 if neighboring spins are equal, 0 otherwise. Performing high temperature expansion by writing e^(Kδσ_iσ_j) = 1 + (e^K - 1)δσ_iσ_j yields an expansion in terms of clusters—connected sets of edges and vertices. Each cluster receives a weight of (e^K - 1)^B × Q^C, where B is the number of bonds and C is the number of clusters. This cluster representation connects to loop models through geometric transformations.

The Ising model assigns spins (+1/-1) to graph vertices with probability proportional to exp(-βH), where H counts neighboring opposite spins. Wilhelm Lenz introduced it in 1920, and Ernst Ising investigated phase transitions. On bipartite lattices, ferromagnetic and antiferromagnetic versions are equivalent via spin reversal. In 1D, no phase transition exists due to exponential correlation decay. Lars Onsager proved in 1944 that 2D exhibits a phase transition at critical temperature, with spin expectations decaying polynomially (n^(-1/8)) at criticality. The Q-state Potts model generalizes this by allowing Q spin values, with probability proportional to x^(number of different neighboring pairs). For Q=2, it reduces to the Ising model. The Fortuin-Kasteleyn (FK) random cluster model unifies spin and percolation models: configurations have both spin assignments and edge states, with weight (1-p)^(closed edges) × p^(open edges) × Q^(clusters). The critical point occurs at p = √Q/(1+√Q), where the model is self-dual. Baxter reformulated the Potts model using oriented loops on the dual lattice, with probability proportional to (√Q/2)^(number of oriented loops). This formulation is local, depending only on left and right turns, and complexifying by assigning e^(iθ) to counterclockwise and e^(-iθ) to clockwise loops creates a complex probability measure.