Notes on Sacks' splitting theorem (Q7033945)

In the early exploration of c.e. sets and degrees, the Friedberg-Muchnik theorem and Sacks' splitting theorem played very important roles, in particular for introducing the finite injury argument. The proof of the Friedberg-Muchnik theorem, which shows the existence of two Turing incomparable c.e. sets, has a characteristic of ``computably bounded mind-changes'' and the constructed c.e. set is totally \(\omega\)-c.a. On the other hand, the proof of the Sacks' splitting theorem, which shows that any non-computable c.e. set can be split to two incomparable c.e. sets, has the future of ``unbounded but finite injury''. The paper under review tries to quantify the difference between the Friedberg-Muchnik theorem and Sacks' splitting theorem within the framework of the Downey-Greenberg hierarchy.\N\NIt is shown that, while the Friedberg-Muchnik theorem is valid in the \(\omega\)-c.a. level, Sack's splitting theorem is valid in the \(\omega^2\)-c.a. level: Any c.e. set \(A\) can be split to two c.e. sets which are totally \(\omega^2\)-c.a. And this result is optimal due to another result of the second author and \textit{K. M. Ng} [Ann. Pure Appl. Logic 169, No. 8, 803--834 (2018; Zbl 1469.03117)].\N\NFurthermore, this paper also shows that, for any \(\alpha < \epsilon_0\), there exist non-computable c.e. sets \(A\) and \(C\) such that for any c.e. splitting \(A_0 \sqcup A_1=A\) of \(A\), if \(A_0\) is totally \(\alpha\)-c.e., then \(C\le_T A_1\). This means that no level of the Downey-Greenberg hierarchy suffices to capture the original full version of the original Sacks splitting theorem [\textit{G. E. Sacks}, Ann. Math. (2) 77, 211--231 (1963; Zbl 0118.25104)]: For any noncomputable c.e. set \(A\) and any noncomputable \(\Delta^0_2\) set \(C\), there is a splitting \(A_0\sqcup A_1 =A\) such that both \(A_0\) and \(A_1\) of low degrees and \(C\not\le A_0\) and \(C\not\le_T A_1\).