arXiv subjects
Publications and source records attributed to Reed Solomon.
Many statements studied in reverse mathematics can be seen as mathematical problems, formulated in terms of instances and solutions. We develop a framework of typicality encompassing measure and genericity, and we classify the reverse mathematics zoo in terms of which problems admit typical solutions. It turns out that even very weak problems do not admit typical solutions.
A relational structure is \emph{strongly indivisible} if for every partition $M = X_0 \sqcup X_1$, the induced substructure on $X_0$ or $X_1$ is isomorphic to $\mathcal{M}$. Cameron (1997) showed that a graph is strongly indivisible if and only if it is the complete graph, the completely disconnected graph, or the random graph. We analyze the strength of Cameron's theorem using tools from computability theory and reverse mathematics. We show that Cameron's theorem is is effective up to computable presentation, and give a partial result towards showing that the full theorem holds in the $\omega$-model $\mathsf{REC}$. We also establish that Cameron's original proof makes essential use of the stronger induction scheme $\mathsf{I}\Sigma^0_2$.
The Ginsburg--Sands theorem from topology states that every infinite topological space has an infinite subspace homeomorphic to exactly one of the following five topologies on $\omega$: indiscrete, discrete, initial segment, final segment, and cofinite. The original proof is nonconstructive, and features an interesting application of Ramsey's theorem for pairs ($\mathsf{RT}^2_2$). We analyze this principle in computability theory and reverse mathematics, using Dorais's formalization of CSC spaces. Among our results are that the Ginsburg-Sands theorem for CSC spaces is equivalent to $\mathsf{ACA}_0$, while for Hausdorff spaces it is provable in $\mathsf{RCA}_0$. Furthermore, if we enrich a CSC space by adding the closure operator on points, then the Ginsburg-Sands theorem turns out to be equivalent to the chain/antichain principle ($\mathsf{CAC}$). The most surprising case is that of the Ginsburg-Sands theorem restricted to $T_1$ spaces. Here, we show that the principle lies strictly between $\mathsf{ACA}_0$ and $\mathsf{RT}^2_2$, yielding arguably the first natural theorem from outside logic to occupy this interval. As part of our analysis of the $T_1$ case we introduce a new class of purely combinatorial principles below $\mathsf{ACA}_0$ and not implied by $\mathsf{RT}^2_2$ which form a strict hierarchy generalizing the stable Ramsey's theorem for pairs ($\mathsf{SRT}^2_2$). We show that one of these, the $\Sigma^0_2$ subset principle ($\Sigma^0_2$-$\mathsf{Subset}$), has the property that it, together with the cohesive principle ($\mathsf{COH}$), is equivalent over $\mathsf{RCA}_0$ to the Ginsburg--Sands theorem for $T_1$ CSC spaces.
We study versions of the tree pigeonhole principle, $\mathsf{TT}^1$, in the context of Weihrauch-style computable analysis. The principle has previously been the subject of extensive research in reverse mathematics. Two outstanding questions from the latter investigation are whether $\mathsf{TT}^1$ is $\Pi^1_1$-conservative over the ordinary pigeonhole principle, $\mathsf{RT}^1$, and whether it is equivalent to any first-order statement of second-order arithmetic. Using the recently introduced notion of the first-order part of an instance-solution problem, we formulate, and answer in the affirmative, the analogue of the first question for Weihrauch reducibility. We then use this, in combination with other results, to answer in the negative the analogue of the second question. Our proofs develop a new combinatorial machinery for constructing and understanding solutions to instances of $\mathsf{TT}^1$.
We introduce the notion of the \emph{first-order part} of a problem in the Weihrauch degrees. Informally, the first-order part of a problem $\mathsf{P}$ is the strongest problem with codomaixn $\omega$ that is Weihrauch reducible to $\mathsf{P}$. We show that the first-order part is always well-defined, examine some of the basic properties of this notion, and characterize the first-order parts of several well-known problems from the literature.
We show that every countable ideal of degrees that are low for isomorphism is contained in a principal ideal of degrees that are low for isomorphism by adapting an exact pair construction. We further show that within the hyperimmune-free degrees, lowness for isomorphism is entirely independent of computable traceability.
We study the implications of model completeness of a theory for the effectiveness of presentations of models of that theory. It is immediate that for a computable model $\mathcal A$ of a computably enumerable, model complete theory, the entire elementary diagram $E(\mathcal A)$ must be decidable. We prove that indeed a c.e. theory $T$ is model complete if and only if there is a uniform procedure that succeeds in deciding $E(\mathcal A)$ from the atomic diagram $\Delta(\mathcal A)$ for all countable models $\mathcal A$ of $T$. Moreover, if every presentation of a single isomorphism type $\mathcal A$ has this property of relative decidability, then there must be a procedure with succeeds uniformly for all presentations of an expansion $(\mathcal A,\vec{a})$ by finitely many new constants. We end with a conjecture about the situation when all models of a theory are relatively decidable.
We define the notion of a determined Borel code in reverse math, and consider the principle $DPB$, which states that every determined Borel set has the property of Baire. We show that this principle is strictly weaker than $ATR$. Any $\omega$-model of $DPB$ must be closed under hyperarithmetic reduction, but $DPB$ is not a theory of hyperarithmetic analysis. We show that whenever $M\subseteq 2^\omega$ is the second-order part of an $\omega$-model of $DPB$, then for every $Z \in M$, there is a $G \in M$ such that $G$ is $\Delta^1_1$-generic relative to $Z$.
Hindman's Theorem (HT) states that for every coloring of $\mathbb N$ with finitely many colors, there is an infinite set $H \subseteq \mathbb N$ such that all nonempty sums of distinct elements of $H$ have the same color. The investigation of restricted versions of HT from the computability-theoretic and reverse-mathematical perspectives has been a productive line of research recently. In particular, HT$^{\leqslant n}_k$ is the restriction of HT to sums of at most $n$ many elements, with at most $k$ colors allowed, and HT$^{=n}_k$ is the restriction of HT to sums of \emph{exactly} $n$ many elements and $k$ colors. Even HT$^{\leqslant 2}_2$ appears to be a strong principle, and may even imply HT itself over RCA$_0$. In contrast, HT$^{=2}_2$ is known to be strictly weaker than HT over RCA$_0$, since HT$^{=2}_2$ follows immediately from Ramsey's Theorem for $2$-colorings of pairs. In fact, it was open for several years whether HT$^{=2}_2$ is computably true. We show that HT$^{=2}_2$ and similar results with addition replaced by subtraction and other operations are not provable in RCA$_0$, or even WKL$_0$. In fact, we show that there is a computable instance of HT$^{=2}_2$ such that all solutions can compute a function that is diagonally noncomputable relative to $\emptyset'$. It follows that there is a computable instance of HT$^{=2}_2$ with no $\Sigma^0_2$ solution, which is the best possible result with respect to the arithmetical hierarchy. Furthermore, a careful analysis of the proof of the result above about solutions DNC relative to $\emptyset'$ shows that HT$^{=2}_2$ implies RRT$^{=2}_2$, the Rainbow Ramsey Theorem for $2$-colorings of pairs, over RCA$_0$. The most interesting aspect of our construction of computable colorings as above is the use of an effective version of the Lov\'asz Local Lemma due to Rumyantsev and Shen.
We analyze the Dual Ramsey Theorem for $k$ partitions and $\ell$ colors ($\mathsf{DRT}^k_\ell$) in the context of reverse math, effective analysis, and strong reductions. Over $\mathsf{RCA}_0$, the Dual Ramsey Theorem stated for Baire colorings is equivalent to the statement for clopen colorings and to a purely combinatorial theorem $\mathsf{cDRT}^k_\ell$. When the theorem is stated for Borel colorings and $k\geq 3$, the resulting principles are essentially relativizations of $\mathsf{cDRT}^k_\ell$. For each $\alpha$, there is a computable Borel code for a $\Delta^0_\alpha$ coloring such that any partition homogeneous for it computes $\emptyset^{(\alpha)}$ or $\emptyset^{(\alpha-1)}$ depending on whether $\alpha$ is infinite or finite. For $k=2$, we present partial results giving bounds on the effective content of the principle. A weaker version for $\Delta^0_n$ reduced colorings is equivalent to $\mathsf{D}^n_2$ over $\mathsf{RCA}_0+\mathsf{I}\Sigma^0_{n-1}$ and in the sense of strong Weihrauch reductions.
The principle $ADS$ asserts that every linear order on $\omega$ has an infinite ascending or descending sequence. This has been studied extensively in the reverse mathematics literature, beginning with the work of Hirschfeldt and Shore. We introduce the principle $ADC$, which asserts that linear order has an infinite ascending or descending chain. The two are easily seen to be equivalent over the base system $RCA_0$ of second order arithmetic; they are even computably equivalent. However, we prove that $ADC$ is strictly weaker than $ADS$ under Weihrauch (uniform) reducibility. In fact, we show that even the principle $SADS$, which is the restriction of $ADS$ to linear orders of type $\omega + \omega^*$, is not Weihrauch reducible to $ADC$. In this connection, we define a more natural stable form of $ADS$ that we call $General\text-SADS$, which is the restriction of $ADS$ to linear orders of type $k + \omega$, $\omega + \omega^*$, or $\omega + k$, where $k$ is a finite number. We define $GeneralSADC$ analogously. We prove that $GeneralSADC$ is not Weihrauch reducible to $SADS$, and so in particular, each of $SADS$ and $SADC$ is strictly weaker under Weihrauch reducibility than its general version. Finally, we turn to the principle $CAC$, which asserts that every partial order on $\omega$ has an infinite chain or antichain. This has two previously studied stable variants, $SCAC$ and $WSCAC$, which were introduced by Hirschfeldt and Jockusch, and by Jockusch, Kastermans, Lempp, Lerman, and Solomon, respectively, and which are known to be equivalent over $RCA_0$. Here, we show that $SCAC$ is strictly weaker than $WSCAC$ under even computable reducibility.
We consider the strength and effective content of restricted versions of Hindman's Theorem in which the number of colors is specified and the length of the sums has a specified finite bound. Let $\mathsf{HT}^{\leq n}_k$ denote the assertion that for each $k$-coloring $c$ of $\mathbb{N}$ there is an infinite set $X \subseteq \mathbb{N}$ such that all sums $\sum_{x \in F} x$ for $F \subseteq X$ and $0 < |F| \leq n$ have the same color. We prove that there is a computable $2$-coloring $c$ of $\mathbb{N}$ such that there is no infinite computable set $X$ such that all nonempty sums of at most $2$ elements of $X$ have the same color. It follows that $\mathsf{HT}^{\leq 2}_2$ is not provable in $\mathsf{RCA}_0$ and in fact we show that it implies $\mathsf{SRT}^2_2$ in $\mathsf{RCA}_0$. We also show that there is a computable instance of $\mathsf{HT}^{\leq 3}_3$ with all solutions computing $0'$. The proof of this result shows that $\mathsf{HT}^{\leq 3}_3$ implies $\mathsf{ACA}_0$ in $\mathsf{RCA}_0$.
We answer a question posed by Hirschfeldt and Jockusch by showing that whenever $k > \ell$, Ramsey's theorem for singletons and $k$-colorings, $\mathsf{RT}^1_k$, is not strongly computably reducible to the stable Ramsey's theorem for $\ell$-colorings, $\mathsf{SRT}^2_\ell$. Our proof actually establishes the following considerably stronger fact: given $k > \ell$, there is a coloring $c : \omega \to k$ such that for every stable coloring $d : [\omega]^2 \to \ell$ (computable from $c$ or not), there is an infinite homogeneous set $H$ for $d$ that computes no infinite homogeneous set for $c$. This also answers a separate question of Dzhafarov, as it follows that the cohesive principle, $\mathsf{COH}$, is not strongly computably reducible to the stable Ramsey's theorem for all colorings, $\mathsf{SRT}^2_{<\infty}$. The latter is the strongest partial result to date in the direction of giving a negative answer to the longstanding open question of whether $\mathsf{COH}$ is implied by the stable Ramsey's theorem in $\omega$-models of $\mathsf{RCA}_0$.
We show that positive measure domination implies uniform almost everywhere domination and that this proof translates into a proof in the subsystem WWKL$_0$ (but not in RCA$_0$) of the equivalence of various Lebesgue measure regularity statements introduced by Dobrinen and Simpson. This work also allows us to prove that low for weak $2$-randomness is the same as low for Martin-Löf randomness (a result independently obtained by Nies). Using the same technique, we show that $\leq_{LR}$ implies $\leq_{LK}$, generalizing the fact that low for Martin-Löf randomness implies low for $K$.
We divide the class of infinite computable trees into three types. For the first and second types, $0'$ computes a nontrivial self-embedding while for the third type $0''$ computes a nontrivial self-embedding. These results are optimal and we obtain partial results concerning the complexity of nontrivial self-embeddings of infinite computable trees considered up to isomorphism. We show that every infinite computable tree must have either an infinite computable chain or an infinite $Π^0_1$ antichain. This result is optimal and has connections to the program of reverse mathematics.
The notions of almost everywhere (a.e.) domination and its uniform version were introduced and studied in reverse mathematics. This paper studies these notions from a recursion-theoretic point of view and explore their connections to notions such as randomness and genericity. It is shown that if $Z$ is a.e. dominating then each $1$-$Z$-random is $2$-random. In other words, $0'\leq_{\rm LR} Z$ for every a.e. dominating $Z$, where ${\rm LR}$ denotes low-for-random reducibility. Other results and corollaries are also given.
In recent years, there has been a substantial amount of work in reverse mathematics concerning natural mathematical principles that are provable from $\RT$, Ramsey's Theorem for Pairs. These principles tend to fall outside of the "big five" systems of reverse mathematics and a complicated picture of subsystems below $\RT$ has emerged. In this paper, we answer two open questions concerning these subsystems, specifically that $\ADS$ is not equivalent to $\CAC$ and that $\EM$ is not equivalent to $\RT$.
This paper falls within the general program of investigating the proof theoretic strength (in terms of reverse mathematics) of combinatorial principals which follow from versions of Ramsey's theorem. We examine two statements in graph theory and one statement in lattice theory proved by Galvin, Rival and Sands \cite{GRS:82} using Ramsey's theorem for 4-tuples. Our main results are that the statements concerning graph theory are equivalent to Ramsey's theorem for 4-tuples over $\RCA$ while the statement concerning lattices is provable in $\RCA$. Revised 12/2010. To appear in Archive for Mathematical Logic