Definition
A structural transformation that expands a relational model into a tree-like (unfolded) model by duplicating nodes along paths so that local relations become tree branches, used to simplify semantic arguments and decidability analyses.

Principle

Principle
Preserve satisfaction of the target fragment (often modal, guarded, or bisimulation-invariant formulas) while converting arbitrary relational structures into tree-like forms that are easier to analyze.

Demonstration

Demonstration
Given a Kripke frame with cycles, produce its tree unfolding by taking all finite rooted paths as nodes and connecting a path p to p·a when a is a successor; the resulting tree satisfies the same bisimulation-invariant modal formulas as the original frame.

Misapplication

Misapplication
Uncritically applying the technique to logics that are not bisimulation-invariant or to languages requiring global cardinality constraints, which can break equivalence of truth and yield false decidability claims.

Consequence

Consequence
Yields tree-model properties, simpler bisimulation arguments, and often a path to decidability or complexity bounds by reducing to tree automata or infinite-tree constructions.

Reversal

Reversal
Collapsing the unfolded tree by identifying nodes that correspond to the same original element recovers the original relational structure but may reintroduce cycles and global features lost in the tree form.

Boundary

Boundary
Applies to relational structures and fragments that respect locality and bisimulation (modal, guarded fragments); it does not generally preserve second-order properties, global counting constraints, or theories that depend on automorphism groups.

Semantic Tension

Semantic Tension
Tension between gaining locality and loss of global structure: unraveling makes local model-checking easier but can obscure or destroy global invariants such as finite-cardinality relations or symmetry.

Synthesis

Synthesis
The Unravelling Technique systematically turns complex relational models into tree-like models that preserve local, bisimulation-invariant truth, enabling simpler semantic proofs and algorithmic analyses while excluding phenomena that require global identification.