The paper studies a simple mathematical rule inspired by how binary trees grow: take two numbers, find the larger one, and add 1. This mimics what happens to the height of a binary tree when you combine two subtrees, since the total height is determined by the taller side, plus one level for the new root. The authors treat this operation as the foundation of an algebraic system, where the building blocks are whole numbers together with a special symbol representing negative infinity (standing in for "nothing" or an empty tree). They also allow taking the maximum of two numbers as a second operation, giving the system a rich structure to analyze.
A central result concerns what information actually matters when you evaluate a labeled tree using this algebra. It turns out that the only thing that counts is the greatest depth at which each label appears, not the full shape of the tree. The authors then characterize which combinations of depth values can actually arise from a single tree: they are exactly the combinations satisfying the binary Kraft inequality, a classical condition from information theory related to prefix-free codes. When you allow combining multiple trees using the maximum operation, you can achieve any combination of depth values whatsoever. This leads to a clean conclusion: the collection of all algebraic expressions in a given set of variables forms what is called a free algebra, meaning it is the most general possible algebra of its type with no unexpected simplifications or collapsing of distinct expressions.
Beyond these core results, the paper carries out a thorough structural analysis of the algebra itself. The authors classify all compatible ways to add extra operations, identify all sub-algebras (consistent subsets closed under the operations), describe all structure-preserving maps from the algebra to itself, and catalog all the ways the algebra can be simplified into finite versions by treating certain elements as equivalent. They also connect their work to a concept called the dyadic-composition spectrum, recovering known results about compositions of simple scaling functions at the boundary of a particular parameter range. Taken together, the paper shows that this elementary tree-height rule generates a surprisingly rich and well-organized algebraic world.