Full normalization for transfinite stacks

Abstract

We describe the extension of normal iteration strategies with appropriate condensation properties to strategies for stacks of normal trees, with full normalization. Given a regular uncountable cardinal $\Omega$ and an $(m,\Omega+1)$-iteration strategy $\Sigma$ for a premouse $M$, such that $\Sigma$ and $M$ both have appropriate condensation properties, we extend $\Sigma$ to a strategy $\Sigma^*$ for the optimal-$(m,\Omega,\Omega+1)^*$-iteration game such that for all $\lambda<\Omega$ and all stacks $\vec{\mathcal{T}}=\left<\mathcal{T}_\alpha\right>_{\alpha<\lambda}$ via $\Sigma^*$, consisting of normal trees $\mathcal{T}_\alpha$, each of length ${<\Omega}$, there is a corresponding normal tree $\mathcal{X}$ via $\Sigma$ with $M^{\vec{\mathcal{T}}}_\infty=M^{\mathcal{X}}_\infty$. Moreover, if there are no drops in model or degree along the main branches of these trees then the overall iteration maps $i^{\vec{\mathcal{T}}}:M\to M^{\vec{\mathcal{T}}}_\infty$ and $i^{\mathcal{X}}:M\to M^{\mathcal{X}}_\infty$ agree. The construction is the result of a combination of work of John Steel and of the author. We also establish some further useful properties of $\Sigma^*$, and use the methods to analyze the comparison of multiple iterates via a common such strategy.