The paper develops a comprehensive theory of how linear orders (think of any totally ordered set, like the integers or rationals arranged in some sequence) behave when you add them together end-to-end. If you have two ordered sets A and B, you can form A+B by placing a copy of B after a copy of A. The central question is understanding when this operation behaves nicely, particularly when it commutes, meaning when A+B and B+A produce the same ordered structure. The authors build their theory around a new tool analogous to the Euclidean algorithm for integers, adapted to work with pairs of linear orders that nearly commute with each other.
One of the paper's main achievements is resolving a problem that the logician Alfred Tarski posed in 1956. Tarski asked whether, if certain scaled combinations of A and B are equal as ordered structures (like n copies of A followed by m copies of B equaling k copies of B followed by l copies of A, for whole numbers n, m, k, l), does it follow that A+B and B+A are equal? Tarski and a collaborator had confirmed this for specific values of those numbers, but the general case remained open for nearly 70 years. The authors prove the answer is yes across the board, closing this long-standing open question.
The paper also provides a clean geometric-style characterization of exactly when two linear orders commute under addition. The condition involves checking whether infinitely repeated copies of A embed into infinitely repeated copies of B in a compatible way, both going forward and backward through the ordering. This turns out to be a corrected version of an earlier conjecture by Tarski that had been shown to be false as originally stated. Finally, the authors characterize which abstract algebraic structures can be faithfully represented using this addition of linear orders, showing they all belong to a well-understood class of ordered algebraic objects studied by A.H. Clifford in the mid-20th century.