Stability: Difference between revisions

643 bytes added ,  10 months ago
no edit summary
No edit summary
No edit summary
 
(2 intermediate revisions by the same user not shown)
Line 1:
Stability is a notion and very wide range of types of nonrecursive ordinals, inspired by the weaker notion of reflection. In general, stability is defined via ranks of \(L\) being similar to each other. The weakest type of stability is \((+1)\)-stable, i.e. \(L_\alpha\) being a \(\Sigma_1\)-elementary substructure of \(L_{\alpha+1}\). In general, \(\alpha\) is \(\beta\)-stable, or stable up to \(\beta\), if \(L_\alpha\) is a \(\Sigma_1\)-elementary substructure of \(L_\beta\). The \((+1)\)-stable ordinals are precisely the ordinals which are \(\Pi^1_0\)-reflecting.<ref name="RichterAczel74">Wayne Richter & Peter Aczel, “Inductive Definitions and Reflecting Properties of Admissible Ordinals”, in: Jens Erik Fenstad & Peter G. Hinman (eds.), Generalized Recursion Theory (Oslo, 1972), North-Holland (1974)</ref><sup>Section 6</sup> The \((+2)\)-stable ordinals are \(\Pi^1_0\)-reflecting on the class of \((+1)\)-stable ordinals,<ref>To be(see provedsection on this page</ref>"[[#(+η)-stability|(+η)-stability]]") and more, but not yet \(\Pi^1_1\)-reflecting.<ref>To be proved on this page</ref> In general, \((+\eta+1)\)-stable ordinals are significantly greater than \((+\eta)\)-stable ordinals, and then \((\cdot 2)\)-stable ordinals diagonalize over this whole hierarchy. This makes the fine structure of stability useful in [[Ordinal collapsing function|OCFs]] for ordinal analyses of systems beyond [[Kripke-Platek set theory]] with full reflection.
 
This continues on endlessly. \(\Pi^1_1\)-reflection is reached by an \(\alpha\) which is \(\alpha^+\)-stable.<ref name="RichterAczel74" /><sup>Section 6</sup>
Line 7:
Theorem: Let \(n<\omega\). If \(\alpha\) is \((+n+1)\)-stable, then \(\alpha\) is \(\Pi^1_0\)-reflecting on the class of \((+n)\)-stable ordinals below \(\alpha\).
 
Proof: Assume \(\alpha\) is (+2)-stable and \(\phi(\vec x)\) is a first-order formula with parameters from \(L_\alpha\) such that \(L_\alpha\vDash\phi(\vec x)\). \(L_{\alpha+2}\) satisfies \(\exists\gamma(\phi^{L_\gamma}(\vec x)\land L_\gamma\prec_{\Sigma_1}L_{\gamma+1})\) with \(\alpha\) as a witness of such a \(\gamma\) (is stability well-behaved in \(L_{\gamma+2}\), a successor stage of \(L\)?), and this is a \(\Sigma_1\) formula, so by \(L_\alpha\prec_{\Sigma_1}L_{\alpha+2}\), \(L_\alpha\) satisfies this as well. Then there is a \(\gamma<\alpha\) such that \(\phi^{L_\gamma}(\vec x)\) and \(\gamma\) is \((+1)\)-stable. QED
Proof:
 
Corollary: \(\Pi^1_1\)-reflecting ordinals \(\alpha\) are \(\alpha^+\)-stable, therefore they are \(\alpha+3\)-stable. So each \(\Pi^1_1\)-reflecting ordinal is \(\Pi^1_0\)-reflecting on the \((+2)\)-stable ordinals below it.
160

edits