Extending CL-reducibility on array noncomputable degrees

From MaRDI portal





The paper studies refinements of Turing reducibility for left-c.e.\ reals in algorithmic randomness, focusing on bounded-use reducibilities of the form \((\mathrm{id}+g)\)-bT. These generalize computably Lipschitz (cl) reducibility by allowing a computable nondecreasing overhead function \(g\).\N\NPrevious work showed that two classical cl-reducibility theorems characterize array noncomputable c.e.\ degrees: the existence of maximal pairs of left-c.e.\ reals and the existence of a left-c.e.\ real not reducible to any random left-c.e.\ real. The first theorem has been extended to \((\mathrm{id}+g)\)-bT whenever \(g\) is slow-growing. This paper proves that the second theorem can also be extended analogously, thereby completing the corresponding extension program. This yields a clear dichotomy between slow-growing and fast-growing use bounds.\N\NA technical contribution of the paper is a simplified construction method for left-c.e.\ reals based on global ``loading arguments, replacing more involved game-based priority constructions from the literature. This leads to shorter proofs and a unified treatment of several known results.\N\NThe paper is technically solid and contributes to a better structural understanding of bounded reducibilities and array noncomputable degrees. The exposition is clear for readers familiar with computability and algorithmic randomness, and the discussion of open problems is useful.











This page was built for publication: Extending CL-reducibility on array noncomputable degrees

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7021348)