Word-length Curve Counting on the Once-Punctured Torus

Basic Information

• Paper ID: 2404.09372
• Title: Single self-intersection words on the once-punctured torus and their counting
• Authors: David Fisac, Mingkun Liu
• Classification: math.GT (Geometric Topology), math.CO (Combinatorics)
• Publication Date: April 2024 (arXiv v3: November 10, 2025)
• Paper Link: https://arxiv.org/abs/2404.09372

Abstract

This paper provides a combinatorial classification of closed curves with a single self-intersection point on the once-punctured torus (Σ₁,₁). The authors determine the exact number of closed curves with zero, one, and arbitrary numbers of self-intersections for a given word-length.

Research Background and Motivation

Research Questions

This paper studies the counting problem of closed curves on the once-punctured torus Σ₁,₁, with particular focus on:

1. How to characterize curves with specific self-intersection numbers from a combinatorial perspective
2. The number of primitive curves with 0, 1, or arbitrary self-intersections for a given word-length L

Significance of the Problem

1. Fundamental problem in geometric topology: Curve counting is central to surface topology research and closely related to geodesic theory and hyperbolic geometry
2. Continuation of Mirzakhani's pioneering work: Mirzakhani proved asymptotic formulas for simple closed geodesics and extended them to fixed self-intersection numbers
3. Bridge between combinatorics and geometry: Studying curves via word-length rather than hyperbolic length provides a purely combinatorial perspective
4. Probabilistic interpretation: Results can be interpreted as "the probability that a random curve has a specific self-intersection number"

Limitations of Existing Methods

1. Mirzakhani's approach: Based on hyperbolic geometry and Weil-Petersson volumes, which is overly "heavy-handed" for low-complexity surfaces like the once-punctured torus
2. Lack of combinatorial characterization: While previous work (Chas, Phillips, etc.) provided bounds and formulas for specific cases, a complete combinatorial classification was missing
3. Self-intersection number 1 unresolved: Although simple curves (self-intersection 0) were completely characterized by Buser-Semmler, the case of self-intersection 1 had not been systematically classified

Research Motivation

To provide a complete combinatorial classification of curves with self-intersection number 1 on the once-punctured torus and give exact counting formulas, filling a theoretical gap.

Core Contributions

1. New proof of simple curve counting (Theorem 1.1):
   • Proves that the number of primitive simple curves of word-length L is 4φ(L) (φ is Euler's totient function)
   • Provides a new purely combinatorial proof method, distinct from homology-homotopy correspondence
2. Complete classification of self-intersection 1 curves (Theorem 1.4):
   • Provides combinatorial characterization of all primitive curves with self-intersection 1
   • Introduces the concept of "2-variation" as a generalization of "small variation"
3. Exact counting formulas (Theorem 1.5):
   • For L odd: |{γ ∈ PC(Σ₁,₁) | i(γ) = 1, ℓω(γ) = L}| = 8φ(L-4)
   • For L even: = 8(φ(L-4) + φ(L/2)/2)
4. Probabilistic results (Corollary 1.7):
   • Proves that a random curve (with at most one self-intersection) has exactly one self-intersection with probability 9/13
5. Counting for arbitrary self-intersections (Theorem 1.8):
   • Provides counting formula for all primitive curves: (1/L)∑_{d|L} μ(d)3^{L/d}
   • Discovers a numerical coincidence with "3-colored necklace" counting

Detailed Methodology

Task Definition

Input: Word-length L and self-intersection number k
Output: Number of primitive curves satisfying the conditions
Constraints: Curves on the once-punctured torus, represented as conjugacy classes in the fundamental group F₂

Core Methodological Framework

1. Correspondence between Curves and Necklaces (Section 2)

Necklace definition: Cyclic equivalence classes of positive integer sequences, denoted n₁, n₂, ..., nᵣ

Small variation condition (Definition 2.2): A necklace nᵢ satisfies small variation if and only if for all s ≥ 1 and all i₁, i₂: $\left|\sum_{j=1}^s n_{i_1+j} - \sum_{j=1}^s n_{i_2+j}\right| \leq 1$

Buser-Semmler theorem (Theorem 2.5): Each simple closed curve can be represented in one of the following forms:

1. a
2. aba⁻¹b⁻¹
3. ab^{n₁}ab^{n₂}...ab^{nᵣ}, where n₁,...,nᵣ has small variation

2. Rigidity Theorem for Necklaces (Proposition 2.8)

Core result: Given m ∈ ℤ_{≥1} and x, y ∈ ℤ_{≥0}, there exists a unique small variation necklace containing exactly x copies of m and y copies of (m+1).

Proof strategy:

• Define operation A: reduction of necklace (removing a digit and replacing with consecutive segment lengths)
• Define inverse operations Bₘ and B̃ₘ
• Key lemma (Lemma 2.10): Operation A preserves uniqueness and strictly decreases min(x,y)
• Prove existence and uniqueness by induction

Action of automorphisms: Define αₘ: a^m b → b, a^{m+1}b → a (and α̃ₘ) These automorphisms preserve self-intersection number (Lemma 2.9), corresponding to operation A at the necklace level.

3. Classification of Self-Intersection 1 (Section 3)

Cohen-Lustig algorithm: Compute self-intersection number by finding linking pairs via cyclic lexicographic order

Classification result (Theorem 1.4): Primitive curves with self-intersection 1 must be one of the following:

(1) Short exceptional cases:

• a²b², aba⁻¹b, ab⁻¹a⁻¹b²

(2) Commutator-type (Proposition 3.2):

• ab⁻¹a⁻¹b·a^{m₁}b···a^{mᵣ}b
• ab⁻¹a⁻¹b·a^{-m₁}b···a^{-mᵣ}b where a^{m₁}b···a^{mᵣ}b represents a primitive simple curve

(3) 2-variation type (Definition 3.4, Proposition 3.6):

• a^{n₁}b···a^{nₖ}b, where n₁,...,nₖ has 2-variation

2-variation definition: nᵢ ∈ {m, m+1}, with exactly one pair of "essential blocks":

• One block: (m, x₂,..., xₖ₋₁, m)
• Another block: (m+1, x₂,..., xₖ₋₁, m+1)

(4) Special type:

• a^m ba^{m+2}b

Key lemma (Proposition 3.8): When gcd(x,y) = 2, there exists a unique 2-variation necklace containing x copies of m and y copies of (m+1); otherwise none exists.

Technical Innovations

1. Necklace rigidity theorem: Connects the balancedness of Sturmian words with geometric self-intersection, proving uniqueness
2. Inductive reduction method: Systematically reduces necklaces via operation A while preserving topological properties
3. 2-variation concept: Precisely characterizes "almost balanced" necklaces, corresponding to self-intersection 1
4. Linking pair and block correspondence: Establishes 1-1 correspondence between Cohen-Lustig algorithm and necklace structure

Experimental Setup

Note: This is a pure mathematics theoretical paper with no experimental section. All results are rigorous mathematical proofs.

Verification Methods

• Verification of self-intersection numbers for specific examples using Cohen-Lustig algorithm
• Verification of formulas using known number-theoretic identities (Möbius inversion, Euler function properties)
• Consistency checks with existing literature (Chas's Proposition 3.2)

Main Theorems and Results

Theorem 1.1 (Simple Curves)

For L ≥ 4: $|\\{γ ∈ PC(Σ_{1,1}) | i(γ) = 0, ℓ_ω(γ) = L\\}| = 4φ(L)$

Cumulative (Corollary 1.2): $|\\{γ | i(γ) = 0, ℓ_ω(γ) ≤ L\\}| = 4Φ(L) + 2 = \frac{12}{π²}L² + O(L(\log L)^{2/3}(\log\log L)^{4/3})$

Theorem 1.5 (Self-Intersection 1)

• L = 4: 8 curves
• L odd (L > 4): 8φ(L-4)
• L even: 8(φ(L-4) + φ(L/2)/2)

Cumulative (Corollary 1.6): $|\\{γ | i(γ) = 1, ℓ_ω(γ) ≤ L\\}| \sim \frac{27}{π²}L²$

Corollary 1.7 (Probabilistic Interpretation)

$\lim_{L→∞} \frac{|\\{i(γ)=0, ℓ_ω≤L\\}|}{|\\{i(γ)=1, ℓ_ω≤L\\}|} = \frac{4}{9}$

That is: A random curve with at most one self-intersection has exactly one self-intersection with probability 9/13 ≈ 69.2%

Theorem 1.8 (Arbitrary Self-Intersections)

For L ≥ 3: $|\\{γ ∈ PC^*(Σ_{1,1}) | ℓ_ω(γ) = L\\}| = \frac{1}{L}\sum_{d|L} μ(d)3^{L/d}$

Non-primitive case: $|\\{γ ∈ C^*(Σ_{1,1}) | ℓ_ω(γ) = L\\}| = \frac{1}{L}\sum_{d|L} φ(d)3^{L/d} + \frac{3+(-1)^L}{2}$

Asymptotic (Corollary 1.9): $|PC^*(L)| \sim \frac{3^L}{L}$

Numerical Coincidence (Remark 1.10)

Number of primitive curves = Number of aperiodic 3-colored necklaces + δ_{1,2}(L)

However, the authors have not found a natural bijection!

Key Proof Techniques

Simple Curve Counting (Theorem 1.1)

Step 1: Via Proposition 2.11, prove that the number of general-type curves of word-length L is 4(L-1)

Step 2: Solve the Diophantine equation (Proposition 2.12): $x(m+1) + y(m+2) = L$ Prove the number of solutions is ⌊L/2⌋ (by constructing bijections Λ₊ and Λ₋)

Step 3: Möbius inversion: $P(n) = \sum_{d|n} μ(d)M(n/d) = 4φ(n) - 4δ_{\\{1\\}}(n)$

Self-Intersection 1 Counting (Theorem 1.5)

Step 1: Classify into P₁(L) + P₂(L) + P₃(L)

P₁(L) (Commutator-type): $P_1(L) = 2·|\\{\text{aperiodic small variation necklaces with sum } L-4\\}| = φ(L-4)$

P₂(L) (Special type): $P_2(L) = \begin{cases} 1 & L \text{ even} \\ 0 & L \text{ odd} \end{cases}$

P₃(L) (2-variation type): Need to solve x(m+1) + y(m+2) = L with gcd(x,y) = 2

Proposition 3.9: Via recursive relations and Möbius inversion, prove:

• L odd: P₃(L) = 0
• L even: P₃(L) = ⌈φ(L/2)/2⌉ - 1

Step 2: Total = 8(P₁ + P₂ + P₃) (factor 8 from generator renaming)

Arbitrary Self-Intersections (Theorem 1.8)

Generating function method: Let wₙ be the number of reduced words of length n, construct: $W(t) = \sum_{n=1}^∞ w_n t^n$

By analyzing word structure (even/odd segments): $W(t) = \frac{4t - 12t³}{(1-t)(-3t²-2t+1)} = \sum_{n=1}^∞ (2 + (-1)^n + 3^n)t^n$

Apply Möbius inversion to obtain the primitive case formula.

Related Work

Hyperbolic Geometry Perspective

1. Prime geodesic theorem (Delsarte, Huber, Margulis, Selberg): $|\\{γ | ℓ_X(γ) ≤ L\\}| \sim \frac{e^L}{L}$
2. Mirzakhani breakthrough (2008): $|\\{γ | i(γ)=0, ℓ_X(γ) ≤ L\\}| \sim C_{g,n}·B_X·L^{6g-6+2n}$
3. Extensions (Mirzakhani 2016, Erlandsson-Souto 2022):
   • Fixed self-intersection number k
   • Independence of length functions

Word-Length Perspective

1. Chas, Phillips, Lalley: Provide bounds and formulas for specific length-self-intersection differences
2. Chas Proposition 3.2: Counting for self-intersection 1 (this paper provides new proof)

Sturmian Word Theory

• Balanced words (Vuillon, Glen-Justin): Equivalent to small variation
• Cutting sequences: Correspondence between lattice paths and necklaces

Conclusions and Discussion

Main Conclusions

1. Complete characterization of curves with self-intersection 0 and 1 on the once-punctured torus
2. Exact counting formulas provided, verifying asymptotic behavior
3. Established deep connections between curve geometry and combinatorics (necklaces, Sturmian words)
4. Discovered interesting numerical correspondence with 3-colored necklaces

Limitations

1. High self-intersections: Methods not extended to i(γ) ≥ 2
2. High genus: Techniques heavily depend on rank-2 free group structure
3. Bijection problem: Correspondence with 3-colored necklaces lacks natural explanation
4. Computational complexity: Algorithm for determining 2-variation not discussed in detail

Future Directions

1. Extension to i(γ) = k: Seek definition of "k-variation"
2. High genus surfaces: Combinatorial classification for Σ_{g,n} (authors believe necklace bijection may help)
3. Algorithm implementation: Efficient enumeration algorithms
4. Geometric applications: Connections with moduli spaces and Teichmüller theory

In-Depth Evaluation

Strengths

1. Methodological Innovation (⭐⭐⭐⭐⭐)

• Necklace rigidity theorem: The uniqueness proof in Proposition 2.8 is elegant and profound, systematically handling seemingly complex combinatorial structures through inductive reduction
• 2-variation concept: Precisely generalizes small variation, capturing the essence of "almost balanced"
• Multi-perspective fusion: Unifies geometry (self-intersection), algebra (automorphisms), and combinatorics (necklaces)

2. Theoretical Completeness (⭐⭐⭐⭐⭐)

• Complete chain from classification to counting
• All major theorems have rigorous proofs
• Multiple independent methods verify each other (e.g., consistency with Chas's results)

3. Technical Depth (⭐⭐⭐⭐)

• The 1-1 correspondence between Cohen-Lustig algorithm and block structure (Proposition 3.6 proof) demonstrates deep understanding
• Combinatorial solutions to Diophantine equations (Propositions 2.12, 3.9) are technically sophisticated
• Generating function method (Theorem 1.8) is concise and powerful

4. Writing Clarity (⭐⭐⭐⭐)

• Clear structure, progressing systematically from simple to complex
• Abundant illustrations aid understanding (Figures 2-7)
• Rich examples (e.g., specific curves in Figure 6)

Weaknesses

1. Limited Generalizability (⭐⭐)

• Methods highly dependent on rank-2 structure of π₁(Σ₁,₁) ≅ F₂
• No obvious extension path for i(γ) ≥ 2
• High genus cases (g ≥ 2) completely untouched

2. Computational Complexity Not Analyzed

• Algorithm complexity for determining 2-variation?
• How to efficiently enumerate all curves for given L?
• Numerical stability for large L?

3. Insufficient Geometric Intuition

• Geometric meaning of 2-variation not sufficiently intuitive
• Why does gcd(x,y)=2 correspond exactly to self-intersection 1? (Proof given but geometric explanation lacking)
• Quantitative relationship with hyperbolic length unexplored

4. Open Problems

• 3-colored necklace bijection: The numerical coincidence noted in Remark 1.10 lacks explanation, potentially a breakthrough opportunity
• High self-intersection classification: Definition of k-variation unclear

Impact Assessment

Contribution to Field (⭐⭐⭐⭐)

1. Fills theoretical gap: First complete classification of self-intersection 1 curves
2. Methodological contribution: Necklace rigidity theorem potentially applicable to other surfaces
3. Interdisciplinary connection: Bridges topology, combinatorics, and number theory

Practical Value (⭐⭐⭐)

• Primarily theoretical, but provides foundation for algorithm implementation
• Probabilistic result (Corollary 1.7) has statistical significance
• May inspire research on curve complexity

Reproducibility (⭐⭐⭐⭐⭐)

• All proofs complete and verifiable
• Formulas explicit, directly computable
• No experiments or data required

Applicable Scenarios

1. Pure mathematics research:
   • Surface topology researchers studying curve counting
   • Combinatorialists studying necklaces and Sturmian words
   • Number theorists studying Diophantine equations
2. Algorithm development:
   • Curve enumeration algorithms
   • Self-intersection number determination algorithms
3. Related problems:
   • Mapping class group actions
   • Moduli space geometry
   • Quantum topology curve counting

Technical Highlights in Detail

Highlight 1: Invertibility of Operation A

By defining Bₘ and B̃ₘ, the authors cleverly prove invertibility of the reduction operation, which is key to the induction. Key observation:

• A preserves "small variation" property
• min(x,y) strictly decreases until divisibility case
• gcd remains invariant

Highlight 2: Combinatorial Characterization of Linking Pairs

The proof of Proposition 3.6 establishes: $\{\text{linking pair classes}\} \leftrightarrow \{\text{essential block pairs}\}$ This completely transforms geometry (self-intersection) into combinatorics (block structure).

Highlight 3: Generating Function Decomposition

In the proof of Theorem 1.8, decomposing words into even/odd segments: $W(t) = \sum_{k \text{ even}} + \sum_{k \text{ odd}} + \frac{4t}{1-t}$ Each term corresponds to clear combinatorial structure.

Potential Applications and Extensions

1. Algorithm implementation: Develop efficient enumeration algorithms based on classification
2. Statistical research: Study properties of "typical" curves
3. Quantum invariants: Connections between curve counting and Jones polynomial
4. Dynamical systems: Orbit counting under mapping class group action

Selected References

• Mir08 Mirzakhani, M. (2008). Growth of simple closed geodesics. Ann. Math., 168(1), 97-125.
• BS88 Buser, P., & Semmler, K.-D. (1988). The geometry and spectrum of the one holed torus. Comment. Math. Helv., 63(2), 259-274.
• CL87 Cohen, M., & Lustig, M. (1987). Paths of geodesics and geometric intersection numbers.
• ES22 Erlandsson, V., & Souto, J. (2022). Mirzakhani's Curve Counting and Geodesic Currents. Birkhäuser.

Summary

This is an excellent pure mathematics paper that achieves substantial progress on the classical problem of curve counting on the once-punctured torus. Through introducing the necklace rigidity theorem and the 2-variation concept, the authors completely characterize curves with self-intersection 1 and provide exact counting formulas. The methods are novel, proofs rigorous, and results profound.

Core value lies in completely transforming geometric problems into combinatorial ones, providing new tools for subsequent research. While generalizability is limited, the paper achieves theoretical completeness within its scope.

Recommended for: Researchers in surface topology, combinatorics, and Sturmian word theory.

Rating: ⭐⭐⭐⭐ (4.5/5)