Shift-Commuting Maps

• Shift-Commuting Maps are structure-preserving transformations in symbolic dynamics that satisfy strict algebraic and topological constraints, including the *-commutation property.
• They connect combinatorial features with operator theory in k-graphs and ultragraph shift spaces, with criteria like 1-coalignedness and covering maps ensuring classification.
• These maps underpin generalized sliding block codes where regressive block maps ensure local homeomorphism and bijective compatibility with shift operators.

A shift-commuting map is a structure-preserving transformation central to symbolic dynamics and higher-rank graph theory, interacting directly with shift operators on various path or sequence spaces. The characterization of such maps is deeply linked to combinatorial properties of the underlying systems—$k$-graphs and their factorization, ultragraphs, or classical shift spaces—and is tightly constrained by algebraic and topological requirements. Distinctions between commutation and *-commutation (a bijective compatibility condition) lead to rich classification results, connecting shift-commuting maps to properties such as 1-coalignedness, covering maps, and sliding block codes.

1. Definitions and Fundamental Properties

Let $X$ be a set equipped with maps $S,T:X\to X$. They are said to commute if $S\circ T = T\circ S$. The *-commutation property is more stringent: $(S,T)$ *-commute if for every pair $(y,z)\in X\times X$ such that $S(y) = T(z)$, there exists a unique $x\in X$ with $T(x) = y$ and $S(x) = z$ (Maloney et al., 2011). This notion is particularly meaningful when $S$ or $T$ are shift maps acting on infinite path spaces or shift spaces.

• In the context of symbolic dynamics, a shift map $\sigma$ acts by removing the initial symbol of a sequence $x = (x_0, x_1, x_2, ...)$, i.e., $\sigma(x) = (x_1, x_2, ...)$.
• In $k$-graph theory, shift maps $\sigma^{e_i}$ on the infinite path space $\Lambda^\infty$ shift in the $i$th coordinate: $(\sigma^{e_i}x)(m,n) = x(m+e_i, n+e_i)$.

Continuous shift-commuting maps (i.e., $\Phi\circ\sigma = \sigma\circ\Phi$) are often the focus due to both their rigidity and their close connection to key structural features of the underlying system. In ultragraph shift spaces, shift-commuting maps admit a symbol-wise description determined by measurable or topological partitions (Gonçalves et al., 2018).

2. Shift-Commuting Maps in $k$-Graphs: 1-Coalignedness and *-Commutation

A $k$-graph $(\Lambda, d)$ is defined as a small category $\Lambda$ with degree functor $d:\Lambda\to \mathbb{N}^k$ satisfying the unique factorization property: any $\lambda\in\Lambda$ with $d(\lambda)=m+n$ decomposes uniquely as $\lambda = \mu\nu$ with $d(\mu)=m$, $d(\nu)=n$. The standard $k$-graph $\Omega_k$ encodes this via morphisms $(m,n) \in \mathbb{N}^k \times \mathbb{N}^k$ for $m\le n$ (Maloney et al., 2011).

A $k$-graph is 1-coaligned if, for every pair of distinct directions $i\ne j$ and for every pair of edges $\mu\in\Lambda^{e_i}$, $\nu\in\Lambda^{e_j}$ emanating from the same vertex, there exists a unique pair $(\mu',\nu')$ with $\mu'\in\Lambda^{e_i}$, $\nu'\in\Lambda^{e_j}$ such that $\nu'\mu = \mu'\nu$; this forms a commuting square.

The key characterization theorem states: for a source-free $k$-graph, the coordinate shifts $\{\sigma^{e_i}\}$ on $\Lambda^\infty$ pairwise *-commute if and only if $\Lambda$ is 1-coaligned. This establishes a precise combinatorial criterion for the simultaneous *-commutativity of shift maps (Maloney et al., 2011).

3. Sliding Block Codes, *-Commutation, and Regressive Block Maps

A sliding block code is a map $\varphi_b:X\to X$ on a full shift $X=A^\mathbb{N}$ induced by a block map $b:A^m\to A$ via $(\varphi_b(x))_n = b(x_n, ..., x_{n+m-1})$. The -commutation of $\varphi_b$ and $\sigma$ is equivalent to $b$ being **left-permutive*, i.e., for each $(x_1,...,x_{m-1})$, the map $a\mapsto b(a,x_1,...,x_{m-1})$ is bijective.

A broader framework for sliding block codes arises via regressive block maps. A block map $d:A^n\to A$ is regressive if, for every fixed suffix $s \in A^{n-1}$, the map $a \mapsto d(a\cdot s)$ is bijective. Classification of *-commuting sliding block codes states that a code $T_d$ *-commutes with the shift if and only if $d$ is regressive (Willis, 2010). This generalizes the left-permutive criterion and admits a complete description for codes with the *-commuting property.

4. Shift-Commuting Local Homeomorphisms and Covering Maps

A continuous surjective map $p:E\to B$ is a covering map if every $b\in B$ admits an open neighborhood evenly covered by $p$. If all fibers have the same size $k$, $p$ is a $k$-fold covering.

For sliding block codes, a map $T_d:A^\mathbb{N}\to A^\mathbb{N}$ is a local homeomorphism and *-commutes with $\sigma$ if and only if $T_d$ is a $k$-fold covering map defined by a regressive block map (Willis, 2010). Progressive maps (where $a\mapsto d(u\cdot a)$ is bijective for each prefix $u$) imply but are not implied by the local homeomorphism property. Weakly progressive maps further generalize this, ensuring local homeomorphism but not necessarily *-commutation.

Counterexamples, such as specific $d:A^2\to A$ that are regressive but not progressive, demonstrate that the covering map and *-commutation criteria do not coincide with progressivity (Willis, 2010).

5. Shift-Commuting Maps in Ultragraph Shift Spaces

Ultragraph shift spaces generalize finite and infinite alphabet shift spaces, incorporating countable vertex and edge sets with source and range maps, and admit both finite and infinite paths.

A map $\Phi:X_\mathcal{G} \to X_\mathcal{H}$ is shift-commuting (i.e., $\Phi\circ\sigma = \sigma\circ\Phi$) if and only if it has a symbol-wise representation based on a measurable partition of $X_\mathcal{G}$. Extending the classical Curtis-Hedlund-Lyndon theorem, every continuous shift-commuting map between ultragraph shift spaces is a generalized sliding block code, where each output symbol is determined by membership in a finitely defined set (union of pseudo-cylinders) in the source (Gonçalves et al., 2018).

This result unifies the description of continuous shift-commuting maps across finite, infinite, and non-locally-compact shift spaces, provided the partitions are taken to be "finitely defined" in the sense appropriate to the topological basis.

6. Applications and Algebraic Implications

Classification results for *-commuting shift maps have significant algebraic and operator-theoretic consequences:

• For 2-graphs $\Lambda(T,q,t,w)$ arising from basic data, *-commutation of shifts is equivalent to 1-coalignedness, which in turn is equivalent to the simplicity and pure infiniteness of the associated $C^*$-algebra $C^*(\Lambda)$. This occurs precisely when the "three corners" of the rule $w$ are invertible (Maloney et al., 2011).
• In the classical symbolic dynamics context, concrete examples include the flip map $b(0)=1, b(1)=0$ (left-permutive), sum-mod-$n$ codes (regressive and left-permutive), and maps that are covering but not progressive.
• In ultragraph dynamics, the generalized CHL theorem implies that continuous, length-preserving, shift-commuting maps correspond precisely to generalized sliding block codes respecting the stratification by length-zero points.

7. Summary Table: Criteria for *-Commutation of Shift Maps

Setting	Shift-Commuting Criterion	Reference
$k$-graph $(\Lambda, d)$	$\Lambda$ is 1-coaligned	(Maloney et al., 2011)
Full shift $A^\mathbb{N}$, code $b$	$b$ is left-permutive (or regressive)	(1101.37951010.5739)
Ultragraph shift space	Map is generalized sliding block code	(Gonçalves et al., 2018)

In all cases, the structure of shift-commuting maps is determined by deep combinatorial or local bijectivity properties, and the passage to *-commutation introduces strong uniqueness requirements on the preimages under the maps. The connection to $C^*$-algebraic properties in higher-rank graphs and the extension to topologically intricate spaces such as ultragraph shifts demonstrates the broad applicability and significance of these classification results.