Anticonfined Singularity Patterns

• Anticonfined singularity patterns are bi-infinite sequences of singular behaviors with a finite window of regular iterates, distinguishing them from confined or unconfined singularities.
• They are identified through analytic continuation and ultradiscrete (max–plus) methods that track local multiplicity growth in discrete dynamical systems.
• Different growth regimes—bounded, linear, or exponential—serve as indicators of integrability, guiding classification and deautonomisation in higher-order mappings.

An anticonfined singularity pattern denotes a specific recurrence structure in discrete dynamical systems—typically birational mappings or lattice equations—characterized by bi-infinite sequences of singular values (such as poles or zeros) on both sides of a finite window of regular values. Unlike confined singularities, which heal after a finite number of iterations, or unconfined singularities, which persist only in one temporal direction, anticonfined singularities persist indefinitely forwards and backwards, with only a finite interval of regular iterates in between. The growth and structure of these patterns are tightly linked to the integrability, degree growth, and classification of discrete equations, and they play a pivotal role in contemporary singularity analysis of higher-order and lattice mappings.

1. Formal Definition and Structure of Anticonfined Singularity Patterns

Let $x_{n+1}=F(x_n,x_{n-1},\dots)$ denote a birational mapping (typically of order $k\geq2$) or a reduction of a lattice equation. A singularity is called anticonfined if there exists a finite index window $[N_-,N_+]$ such that all iterates $x_n$ for $n<N_-$ and $n>N_+$ are singular (e.g., infinite, zero, or undefined), while only the iterates $x_{N_-}\leq n\leq N_+$ are regular (finite and well-defined). This produces a pattern:

$\dots, S, S, S,\, R, R, \dots, R,\, S, S, S, \dots$

where $S$ denotes a singular value and $R$ a regular one. Often, the singular tails are periodic or obey integer growth laws, with the local multiplicity of the pole/zero at each $n$ denoted by $d_n$ (Mase et al., 2015, Willox et al., 4 Feb 2026).

This is in precise contrast to:

• Confined singularities: finite blocks of $S$ followed by full return to regularity (recursive freedom).
• Unconfined singularities: infinite persistence in only one direction.
• Cyclic singularities: finite invertible cycles within the singular locus, with no bi-infinite tails.

For higher-order reductions or maps, the loci supporting these patterns can be of co-dimension 1 (a divisor) or higher, and the nature of the pattern (periodic, linear growth, exponential growth) critically informs the algebraic entropy.

2. Archetypes, Classification, and Growth Properties

Anticonfined patterns can be classified according to the order-growth of their singular multiplicities $d_n$:

Type	Asymptotic growth $d_n$	Implications for integrability
Bounded/Periodic	$d_n = \text{const}$	Trivially integrable or linearisable
Linear	$d_n = O(|n|)$	Linearisable, dynamical degree $1$
Exponential	$d_n \sim C\lambda^{|n|}$, $\lambda>1$	Non-integrable, entropy $\log \lambda$

For example, in the second-order Hénon map, the anticonfined pattern exhibits exponential growth of double, matching the dynamical degree $\lambda=2$ (Mase et al., 2015). In higher-order integrable maps, polynomial degree growth emerges from linear or bounded $d_n$, as shown in the tropical ultradiscrete analysis (Willox et al., 4 Feb 2026).

Different recurrences yield different anticonfined laws. For a third-order integrable map such as

$$y_{n+1} = \frac{y_{n-2}y_n}{y_{n-1}(y_n - a_{n-1}y_{n-1})}$$

the multiplicities $m_n$ in the tail satisfy a max–plus (ultradiscrete) recurrence:

$m_{n+1} = m_{n-2} + \max(m_n, m_{n-1}) - m_{n-1}$

with $m_n\sim n$ in the tails (asymptotically linear growth), leading to quadratic growth in degree $d_n$ (Willox et al., 4 Feb 2026).

3. Methods for Extraction and Analysis

The identification and analysis of anticonfined patterns involve several steps:

• Pattern Detection: By analytic continuation or via introduction of small parameter $\epsilon \to 0$, track the propagation of poles/zeros across iterations starting from a singular initial condition.
• Multiplicity Tracking: Compute the local multiplicity $d_n$ of the singularity at each step, often via explicit substitution (e.g., $x_1 = \epsilon^{-1}$) and asymptotic expansion.
• Ultradiscrete (Tropical) Recursion: Pass to a max–plus limit by substituting $x_n = \exp(Y_n/\epsilon)$ and $\epsilon \to 0^+$, leading to a piecewise-linear recurrence for $Y_n$ that governs $d_n$ (Willox et al., 4 Feb 2026).
• Degree Growth Inference: The sum over positive $Y_n$ in the ultradiscrete tails provides the polynomial law for projective degree $d_n$; exponential growth in $d_n$ flags non-integrability.
• Deautonomisation Consistency: In deautonomizing higher-order mappings, preservation of the multiplicity pattern in the anticonfined tail is necessary for polynomial growth; otherwise, exponential degree growth occurs (Willox et al., 4 Feb 2026).

Table: Max–plus ultradiscrete method for computing tail growth in sample maps

Step	Description	Reference
Identify codimension locus	Detect divisor supporting tail	(Willox et al., 4 Feb 2026)
Ultradiscrete recursion	Derive max–plus rule for multiplicities $Y_n$	(Willox et al., 4 Feb 2026)
Degree calculation	Compute $d_n$ as sum over positive $Y_n$ in tails	(Willox et al., 4 Feb 2026)

4. Anticonfined Patterns in Discrete Integrable Systems

Reduction of discrete soliton equations (e.g., dKdV) and higher-order mappings reveals the universality of anticonfined patterns:

• In the discrete KdV equation, certain reductions on a staircase yield $(q+2)$-point recurrences with singularity loci at $u_2=0$ or $u_{q+1}=0$ (Um et al., 2019). For generic (non-autonomous) parameters, $u_2=0$ leads to a bi-infinite anticonfined pattern: both forward and backward iterations remain singular, and a train of co-dimension 2 strata repeats with period $q+1$.
• For generic two-dimensional lattice equations, anticonfined singularities can appear in non-generic, semi-infinite or periodic initial data, but not for generic finite-propagation data. Thus, presence of anticonfined strata does not contradict integrability (Lax pair, multidimensional consistency), because singularity confinement applies to movable, not fixed, singularities (Um et al., 2019).
• In higher-order integrable maps, as in the examples of Willox–Grammaticos–Ramani, the ultradiscrete (max–plus) approach predicts polynomial ($n^k$) degree growth for $k$th-order maps when the anticonfined tail has only linear local growth (Willox et al., 4 Feb 2026).

5. Implications for Integrability and Deautonomisation

Anticonfined singularity patterns furnish both a diagnostic and a constraint on integrability:

• Integrability Detection: Exponential growth in the tail's multiplicity is a necessary and sufficient certificate for non-integrability. All known examples show that the maximal asymptotic growth rate (dynamical degree) of anticonfined singularities coincides with that of the original mapping (Mase et al., 2015, Willox et al., 4 Feb 2026).
• Linearisability: Linear but nonzero tail growth pinpoints mappings that are not regularisable but remain linearisable, as seen in second-kind linearisable mappings.
• Polynomial Growth and Integrable Maps: Bounded tail growth is compatible with zero or polynomial degree growth, which signals integrability, provided all singularities are confined or cyclic.
• Role in Deautonomisation: In the context of constructing nonautonomous (deautonomised) generalizations, preserving only confined patterns is insufficient for higher-order maps; one must tightly reproduce the entire structure—placement and multiplicities—of zeros and poles in anticonfined patterns to maintain polynomial degree growth (Willox et al., 4 Feb 2026).

A plausible implication is that for genuinely higher-order discrete Painlevé equations, a full theory of deautonomisation will require the classification of singular loci of all codimensions and detailed ultradiscrete analysis of their associated anticonfined growth patterns (Willox et al., 4 Feb 2026).

6. Comparison with Other Singularity Types and Physical Analogues

Anticonfined patterns enrich the taxonomy of singular structures in discrete systems:

• Comparison Table:

Singularity Type	Recovery of Regularity	Pattern Support	Integrability Indicator
Confined	Yes (finite steps)	Co-dimension 1 strata	Often integrable
Unconfined	Never (one direction)	Co-dim 1/2	Non-integrable
Cyclic	Periodic, revisited	Co-dimension 1 cycle	Integrability undetected
Anticonfined	Never (both ways, except finite core)	Co-dimension 1 or 2	Degree-growth lower bound

• Physical Analogies: In nonlinear dynamics and quantum nonlinear systems, patterns structurally akin to anticonfined singularities arise: e.g., "antidark" (or "anticonfined") singular vortex cores in 2D nonlinear Schrödinger problems (Chen et al., 2021). These physical solutions have a central singularity (density $\sim r^{-4/3}$ or $r^{-1}$) surrounded by a regular background—a direct spatial analog of a finite central core flanked by singular "tails," with properties such as integrability of norm and analytical thresholds for stability. This emphasizes that "anticonfinement" is not merely an abstract mapping phenomenon but also appears in nonlinear physics as robust, analyzable structures.

7. Ongoing Challenges and Research Directions

The study of anticonfined patterns remains central to several open questions in integrable systems and discrete dynamics:

• Extension to maps with codimension $>1$ singular loci, including alternating sign patterns in residues and more intricate ultradiscrete recursions (Willox et al., 4 Feb 2026).
• Development of a comprehensive deautonomisation theory for higher-order maps and full characterization of integrable hierarchies via singularity structure.
• Systematic understanding of the interplay between anticonfined growth, algebraic entropy, and the full space of initial conditions in both autonomous and nonautonomous settings.
• Further cross-fertilization between discrete dynamical singularity structure and physical realizations in nonlinear waves, Bose–Einstein condensates, and optical systems (Chen et al., 2021).

Anticonfined singularity patterns, by encoding both the worst-case loss of freedom and the polynomial or exponential growth regime of a mapping, provide essential insight into both the theoretical landscape and practical classification of discrete integrable and non-integrable systems (Mase et al., 2015, Um et al., 2019, Willox et al., 4 Feb 2026).