Movement: Hierarchical Fracture Thesis: The computational aesthetic of pristine hierarchical order yielding to lateral connection — the tree and its shadow.
Hierarchical Fracture is the algorithmic study of order becoming network. It begins with a recursive branching structure — a pure rooted tree where every triangle is isosceles, every node has exactly one parent, and distance is measured by depth to common ancestor. This is the ultrametric ideal: clarity, isolation, pure hierarchy.
But isolation never holds. The philosophy's computational heart is the controlled introduction of lateral edges — cross-branch connections that violate the tree's closure. A parameter governs the probability that any two nodes, regardless of their hierarchical relationship, form a lateral bond. At zero probability, the algorithm draws a pristine ultrametric tree, its nodes glowing in cool indigo, its branches tracing perfect recursive arcs. As the probability increases, golden-amber threads begin to weave between branches, creating triangles that are no longer isosceles — the strong triangle condition fractures.
The philosophy demands that this transition be visible as a morphological phase change, not a smooth blend. Between the pure tree and the pure network lies a critical regime where the structure is neither — clusters form, shortcuts bypass hierarchy, and the viewer cannot tell whether the underlying geometry is a tree or a graph. This is the regime where the most interesting visual complexity emerges, the product of a meticulously balanced algorithm where branching order and lateral disorder compete in a dynamical tension refined through exhaustive parameter exploration by a master of computational aesthetics.
Color carries the narrative. The tree's branches are rendered in a cool spectrum — deep indigo at the root, fading to electric blue at the canopy. Lateral edges are warm — amber, copper, burnt orange. The shadow cast by the tree is not a tree at all: it is a projected network graph, all interconnected nodes, drawn in muted earth tones below the main structure. Nodes that participate in many lateral connections grow larger and brighter; isolated nodes remain small and dim. The algorithm thus reveals, through color and scale alone, which parts of the hierarchy have fractured.
Temporal evolution is essential. The algorithm does not render a static frame. It grows — branches extend first, establishing the hierarchical skeleton. Then, after a pause that lets the viewer appreciate the pure tree, lateral connections begin to appear, one by one, each arrival a small violation of the tree's closure. The viewer watches the phase transition unfold in real time, a meticulously choreographed sequence where every frame demonstrates the craftsman's deep understanding of generative pacing.
Parameters are few but powerful: branching depth, branching ratio, lateral edge probability, and the growth delay between tree emergence and lateral invasion. These four parameters span the complete phase diagram from exact ultrametricity (depth=6, ratio=2, lateral=0) to pure chaos (depth=3, ratio=4, lateral=0.8). Every combination is reproducible via seed. Every seed produces a unique morphology — the mark of a master-level generative algorithm where randomness is constrained by mathematical structure.
The conceptual seed woven invisibly into the algorithm is the Parisi ultrametric solution for the Sherrington-Kirkpatrick spin glass — a mean-field system (
This is not a visualization OF the paper. It is a generative artwork that embodies the paper's central thesis — that the tree is a conditional attractor, approached only when isolation is perfect, and that the shadow of lateral connection is the default state of complex systems. The algorithm IS the argument, expressed in color, geometry, and time.
Hierarchical Fracture — an algorithmic philosophy for tree-and-shadow-viz. Version 0.1.