Consider a normal operation assigning $t_\alpha$ to each ordinal number $\alpha$, and assume that $\lambda$ is a limit ordinal.

Show that cf $t_\lambda$ $=$ cf $\lambda$.

I want to show that the smallest cardinal $\kappa$ such that $\lambda$ is the supremum of $\kappa$ smaller cardinals is equal to the cofinality of the operation. Help would be appreciated here. I know the cofinality of any ordinal $\alpha$ is the least cardinal number $\kappa$ such that there exists a subset $S$ of the ordinal $\alpha$ having cardinality $\kappa$ and that $\alpha$ is the least ordinal $\textit{strictly}$ greater than every member of $S$ (the strict supremum of $S$).

  • $\begingroup$ What is meant by "normal operation"? What is the relationship between $\lambda$ and $\alpha$? $\endgroup$ – Jalex Stark May 8 '18 at 12:37
  • $\begingroup$ Normal operation means the operation is both monotone and continuous and what do you mean the relationship between them? $\endgroup$ – AustereTiger May 8 '18 at 12:43
  • $\begingroup$ That's it! Thank you. I meant to write $cf t_{\lambda}$. $\endgroup$ – AustereTiger May 8 '18 at 12:58
  • $\begingroup$ Okay. Well as a starting hint, you need to show both that the left-hand-side is $\leq$ the right-hand side, and that the right-hand-side is $\leq$ the left-hand-side. Which one seems easier to start with? $\endgroup$ – Jalex Stark May 8 '18 at 12:59
  • $\begingroup$ The left hand side, I suppose? $\endgroup$ – AustereTiger May 8 '18 at 15:26


Pick a cofinal subset $S \subseteq \lambda$. Then $$ T = \{ t_s \mid s \in S \} $$ is cofinal in $t_\lambda$. This shows that $\mathrm{cf}(t_\lambda) \le \mathrm{cf}(\lambda)$.

Conversely, let $T \subseteq t_\lambda$ be cofinal. Then $$ S = \{ \alpha \mid t_\alpha \in T \} $$ is cofinal in $\lambda$. Hence $\mathrm{cf}(\lambda) \le \mathrm{cf}(t_\lambda)$.

  • $\begingroup$ This is not as much as a hint as it is an outline of a solution. $\endgroup$ – Asaf Karagila May 8 '18 at 14:20
  • 1
    $\begingroup$ This was very helpful, Stefan. Thank you. The proof looks to be much shorter than I thought it would be! $\endgroup$ – AustereTiger May 8 '18 at 15:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.