Nonrecursive ordinal

In mathematics, particularly set theory, non-recursive ordinals are large countable ordinals greater than all the recursive ordinals, and therefore can not be expressed using recursive ordinal notations.

The Church–Kleene ordinal and variants

The smallest non-recursive ordinal is the Church Kleene ordinal, ω 1 C K {\displaystyle \omega _{1}^{\mathsf {CK}}} , named after Alonzo Church and S. C. Kleene; its order type is the set of all recursive ordinals. Since the successor of a recursive ordinal is recursive, the Church–Kleene ordinal is a limit ordinal. It is also the smallest ordinal that is not hyperarithmetical, and the smallest admissible ordinal after ω {\displaystyle \omega } (an ordinal α {\displaystyle \alpha } is called admissible if L α K P {\displaystyle L_{\alpha }\models {\mathsf {KP}}} .) The ω 1 C K {\displaystyle \omega _{1}^{\mathsf {CK}}} -recursive subsets of ω {\displaystyle \omega } are exactly the Δ 1 1 {\displaystyle \Delta _{1}^{1}} subsets of ω {\displaystyle \omega } .[1]

The notation ω 1 C K {\displaystyle \omega _{1}^{\mathsf {CK}}} is in reference to ω 1 {\displaystyle \omega _{1}} , the first uncountable ordinal, which is the set of all countable ordinals, analogously to how the Church-Kleene ordinal is the set of all recursive ordinals. Some old sources use ω 1 {\displaystyle \omega _{1}} to denote the Church-Kleene ordinal.[2]

For a set x N {\displaystyle x\subseteq \mathbb {N} } , A set is x-computable if it is computable from a Turing machine with an oracle state that queries x. The relativized Church–Kleene ordinal ω 1 x {\displaystyle \omega _{1}^{x}} is the supremum of the order types of x-computable relations. The Friedman-Jensen-Sacks theorem states that for every countable admissible ordinal α {\displaystyle \alpha } , there exists a set x such that α = ω 1 x {\displaystyle \alpha =\omega _{1}^{x}} .[3]

ω ω C K {\displaystyle \omega _{\omega }^{\mathsf {CK}}} , first defined by Stephen G. Simpson[citation needed] is an extension of the Church–Kleene ordinal. This is the smallest limit of admissible ordinals, yet this ordinal is not admissible. Alternatively, this is the smallest α such that L α P ( ω ) {\displaystyle L_{\alpha }\cap {\mathsf {P}}(\omega )} is a model of Π 1 1 {\displaystyle \Pi _{1}^{1}} -comprehension.[1]

Recursively ordinals

The α {\displaystyle \alpha } th admissible ordinal is sometimes denoted by τ α {\displaystyle \tau _{\alpha }} .[4][5]

Recursively "x" ordinals, where "x" typically represents a large cardinal property, are kinds of nonrecursive ordinals.[6] Rathjen has called these ordinals the "recursively large counterparts" of x,[7] however the use of "recursively large" here is not to be confused with the notion of an ordinal being recursive.

An ordinal α {\displaystyle \alpha } is called recursively inaccessible if it is admissible and a limit of admissibles. Alternatively, α {\displaystyle \alpha } is recursively inaccessible iff α {\displaystyle \alpha } is the α {\displaystyle \alpha } th admissible ordinal,[5] or iff L α K P i {\displaystyle L_{\alpha }\models {\mathsf {KPi}}} , an extension of Kripke–Platek set theory stating that each set is contained in a model of Kripke–Platek set theory. Under the condition that L α V=HC {\displaystyle L_{\alpha }\vDash {\textrm {V=HC}}} ("every set is hereditarily countable"), α {\displaystyle \alpha } is recursively inaccessible iff L α P ( ω ) {\displaystyle L_{\alpha }\cap {\mathsf {P}}(\omega )} is a model of Δ 2 1 {\displaystyle \Delta _{2}^{1}} -comprehension.[8]

An ordinal α {\displaystyle \alpha } is called recursively hyperinaccessible if it is recursively inaccessible and a limit of recursively inaccessibles, or where α {\displaystyle \alpha } is the α {\displaystyle \alpha } th recursively inaccessible. Like "hyper-inaccessible cardinal", different authors conflict on this terminology.

An ordinal α {\displaystyle \alpha } is called recursively Mahlo if it is admissible and for any α {\displaystyle \alpha } -recursive function f : α α {\displaystyle f:\alpha \rightarrow \alpha } there is an admissible β < α {\displaystyle \beta <\alpha } such that { f ( γ ) γ β } β {\displaystyle \left\{f(\gamma )\mid \gamma \in \beta \right\}\subseteq \beta } (that is, β {\displaystyle \beta } is closed under f {\displaystyle f} ).[2] Mirroring the Mahloness hierarchy, α {\displaystyle \alpha } is recursively γ {\displaystyle \gamma } -Mahlo for an ordinal γ {\displaystyle \gamma } if it is admissible and for any α {\displaystyle \alpha } -recursive function f : α α {\displaystyle f:\alpha \rightarrow \alpha } there is an admissible ordinal β < α {\displaystyle \beta <\alpha } such that β {\displaystyle \beta } is closed under f {\displaystyle f} , and β {\displaystyle \beta } is recursively δ {\displaystyle \delta } -Mahlo for all δ < γ {\displaystyle \delta <\gamma } .[6]

An ordinal α {\displaystyle \alpha } is called recursively weakly compact if it is Π 3 {\displaystyle \Pi _{3}} -reflecting, or equivalently,[2] 2-admissible. These ordinals have strong recursive Mahloness properties, if α is Π 3 {\displaystyle \Pi _{3}} -reflecting then α {\displaystyle \alpha } is recursively α {\displaystyle \alpha } -Mahlo.[6]

Weakenings of stable ordinals

An ordinal α {\displaystyle \alpha } is stable if L α {\displaystyle L_{\alpha }} is a Σ 1 {\displaystyle \Sigma _{1}} -elementary-substructure of L {\displaystyle L} , denoted L α 1 L {\displaystyle L_{\alpha }\preceq _{1}L} .[9] These are some of the largest named nonrecursive ordinals appearing in a model-theoretic context, for instance greater than min { α : L α T } {\displaystyle \min\{\alpha :L_{\alpha }\models T\}} for any computably axiomatizable theory T {\displaystyle T} .[10]Proposition 0.7. There are various weakenings of stable ordinals:[1]

  • A countable ordinal α {\displaystyle \alpha } is called ( + 1 ) {\displaystyle (+1)} -stable iff L α 1 L α + 1 {\displaystyle L_{\alpha }\preceq _{1}L_{\alpha +1}} .
    • The smallest ( + 1 ) {\displaystyle (+1)} -stable ordinal is much larger than the smallest recursively weakly compact ordinal: it has been shown that the smallest ( + 1 ) {\displaystyle (+1)} -stable ordinal is Π n {\displaystyle \Pi _{n}} -reflecting for all finite n {\displaystyle n} .[2]
    • In general, a countable ordinal α {\displaystyle \alpha } is called ( + β ) {\displaystyle (+\beta )} -stable iff L α 1 L α + β {\displaystyle L_{\alpha }\preceq _{1}L_{\alpha +\beta }} .
  • A countable ordinal α {\displaystyle \alpha } is called ( + ) {\displaystyle (^{+})} -stable iff L α 1 L α + {\displaystyle L_{\alpha }\preceq _{1}L_{\alpha ^{+}}} , where β + {\displaystyle \beta ^{+}} is the smallest admissible ordinal > β {\displaystyle >\beta } . The smallest ( + ) {\displaystyle (^{+})} -stable ordinal is again much larger than the smallest ( + 1 ) {\displaystyle (+1)} -stable or the smallest ( + β ) {\displaystyle (+\beta )} -stable for any constant β {\displaystyle \beta } .
  • A countable ordinal α {\displaystyle \alpha } is called ( + + ) {\displaystyle (^{++})} -stable iff L α 1 L α + + {\displaystyle L_{\alpha }\preceq _{1}L_{\alpha ^{++}}} , where β + + {\displaystyle \beta ^{++}} are the two smallest admissible ordinals > β {\displaystyle >\beta } . The smallest ( + + ) {\displaystyle (^{++})} -stable ordinal is larger than the smallest Σ 1 1 {\displaystyle \Sigma _{1}^{1}} -reflecting.
  • A countable ordinal α {\displaystyle \alpha } is called inaccessibly-stable iff L α 1 L β {\displaystyle L_{\alpha }\preceq _{1}L_{\beta }} , where β {\displaystyle \beta } is the smallest recursively inaccessible ordinal > α {\displaystyle >\alpha } . The smallest inaccessibly-stable ordinal is larger than the smallest ( + + ) {\displaystyle (^{++})} -stable.
  • A countable ordinal α {\displaystyle \alpha } is called Mahlo-stable iff L α 1 L β {\displaystyle L_{\alpha }\preceq _{1}L_{\beta }} , where β {\displaystyle \beta } is the smallest recursively Mahlo ordinal > α {\displaystyle >\alpha } . The smallest Mahlo-stable ordinal is larger than the smallest inaccessibly-stable.
  • A countable ordinal α {\displaystyle \alpha } is called doubly ( + 1 ) {\displaystyle (+1)} -stable iff L α 1 L β 1 L β + 1 {\displaystyle L_{\alpha }\preceq _{1}L_{\beta }\preceq _{1}L_{\beta +1}} . The smallest doubly ( + 1 ) {\displaystyle (+1)} -stable ordinal is larger than the smallest Mahlo-stable.

Larger nonrecursive ordinals

Even larger nonrecursive ordinals include:[1]

  • The least ordinal α {\displaystyle \alpha } such that L α 1 L β {\displaystyle L_{\alpha }\preceq _{1}L_{\beta }} where β {\displaystyle \beta } is the smallest nonprojectible ordinal.
  • An ordinal α {\displaystyle \alpha } is nonprojectible if α {\displaystyle \alpha } is a limit of α {\displaystyle \alpha } -stable ordinals, or; if the set X = { β < α L β 1 L α } {\displaystyle X=\left\{\beta <\alpha \mid L_{\beta }\preceq _{1}L_{\alpha }\right\}} is unbounded in α {\displaystyle \alpha } .
  • The ordinal of ramified analysis, often written as β 0 {\displaystyle \beta _{0}} . This is the smallest β {\displaystyle \beta } such that L β P ( ω ) {\displaystyle L_{\beta }\cap {\mathsf {P}}(\omega )} is a model of second-order comprehension, or L β Z F C {\displaystyle L_{\beta }\models {\mathsf {ZFC^{-}}}} , which is Z F C {\displaystyle {\mathsf {ZFC}}} without the axiom of power set.
  • The least ordinal α {\displaystyle \alpha } such that L α K P + ' ω 1  exists' {\displaystyle L_{\alpha }\models {\mathsf {KP}}+{\text{'}}\omega _{1}{\text{ exists'}}} . This ordinal has been characterized by Toshiyasu Arai.[11]
  • The least ordinal α {\displaystyle \alpha } such that L α Z F C + ' ω 1  exists' {\displaystyle L_{\alpha }\models {\mathsf {ZFC^{-}}}+{\text{'}}\omega _{1}{\text{ exists'}}} .
  • The least stable ordinal.

References

  1. ^ a b c d D. Madore, A Zoo of Ordinals (2017). Accessed September 2021.
  2. ^ a b c d W. Richter, P. Aczel, Inductive Definitions and Reflecting Properties of Admissible Ordinals (1973, p.15). Accessed 2021 October 28.
  3. ^ Sacks, Gerald E. (1976), "Countable admissible ordinals and hyperdegrees", Advances in Mathematics, 19 (2): 213–262, doi:10.1016/0001-8708(76)90187-0
  4. ^ P. G. Hinman, Recursion-Theoretic Hierarchies (1978), pp.419--420. Perspectives in Mathematical Logic, ISBN 3-540-07904-1.
  5. ^ a b J. Barwise, Admissible Sets and Structures (1976), pp.174--176. Perspectives in Logic, Cambridge University Press, ISBN 3-540-07451-1.
  6. ^ a b c Rathjen, Michael (1994), "Proof theory of reflection" (PDF), Annals of Pure and Applied Logic, 68 (2): 181–224, doi:10.1016/0168-0072(94)90074-4
  7. ^ M. Rathjen, "The Realm of Ordinal Analysis" (2006). Archived 7 December 2023.
  8. ^ W. Marek, Some comments on the paper by Artigue, Isambert, Perrin, and Zalc (1976), ICM. Accessed 19 May 2023.
  9. ^ J. Barwise, Admissible Sets and Structures (1976), Cambridge University Press, Perspectives in Logic.
  10. ^ W. Marek, K. Rasmussen, Spectrum of L in libraries (WorldCat catalog) (EuDML page), Państwowe Wydawn. Accessed 2022-12-01.
  11. ^ T. Arai, A Sneak Preview of Proof Theory of Ordinals (1997, p.17). Accessed 2021 October 28.
  • Church, Alonzo; Kleene, S. C. (1937), "Formal definitions in the theory of ordinal numbers.", Fundamenta Mathematicae, 28: 11–21, doi:10.4064/fm-28-1-11-21, JFM 63.0029.02
  • Church, Alonzo (1938), "The constructive second number class", Bull. Amer. Math. Soc., 44 (4): 224–232, doi:10.1090/S0002-9904-1938-06720-1
  • Kleene, S. C. (1938), "On Notation for Ordinal Numbers", Journal of Symbolic Logic, 3 (4), Vol. 3, No. 4: 150–155, doi:10.2307/2267778, JSTOR 2267778, S2CID 34314018
  • Rogers, Hartley (1987) [1967], The Theory of Recursive Functions and Effective Computability, First MIT press paperback edition, ISBN 978-0-262-68052-3
  • Simpson, Stephen G. (2009) [1999], Subsystems of Second-Order Arithmetic, Perspectives in Logic, vol. 2, Cambridge University Press, pp. 246, 267, 292–293, ISBN 978-0-521-88439-6
  • Richter, Wayne; Aczel, Peter (1974), Inductive Definitions and Reflecting Properties of Admissible Ordinals, pp. 312–313, 333, ISBN 0-7204-2276-0
  • v
  • t
  • e