5
$\begingroup$

Suppose that $\lambda$ is a cardinal. Let $\mathcal{E}_{\lambda}$ be the set of all elementary embeddings from $V_{\lambda}$ to $V_{\lambda}$. If $j,k\in\mathcal{E}_{\lambda}$, then define $j[k]=\bigcup_{\alpha<\lambda}j(k|_{V_{\alpha}})$. Suppose that $j_{1},...,j_{n}\in\mathcal{E}_{\lambda}$. Let $\langle j_{1},...,j_{n}\rangle$ denote the smallest subset of $\mathcal{E}_{\lambda}$ that contains $\{j_{1},...,j_{n}\}$ and is closed under the operation $(j,k)\mapsto j[k]$. Let $K=\{\mathrm{crit}(j)|j\in\langle j_{1},...,j_{n}\rangle\}$.

  1. Then does $K$ necessarily have order type $\omega$?

If $\gamma<\lambda$ is a limit ordinal, then let $\equiv^{\gamma}$ to be the equivalence relation on $V_{\lambda}$ where $j\equiv^{\gamma}k$ if and only if $j(x)\cap V_{\gamma}=k(x)\cap V_{\gamma}$ whenever $x\in V_{\gamma}$. Then $\equiv^{\gamma}$ is a congruence on $\mathcal{E}_{\lambda}$.

  1. Is the quotient algebra $\langle j_{1},...,j_{n}\rangle/\equiv^{\gamma}$ finite whenever $\gamma$ is a limit ordinal with $\gamma<\lambda$?

An affirmative answer to question 2 will give an affirmative answer to question 1 as well.

$\endgroup$
3
$\begingroup$

I claim that the answer to both questions is yes . Suppose that $A$ is a finite set, $j_{a}\in\mathcal{E}_{\lambda}$ for each $a\in A$ and $\gamma<\lambda$ is a limit ordinal. Then I claim that $\langle\{j_{a}|a\in A\}\rangle/\equiv^{\gamma}$ is finite and the sequence $\{\textrm{crit}(j)|j\in\langle\{j_{a}|a\in A\}\rangle\}$ has order type $\omega$.

Let $M$ be the set of all strings $a_{1}...a_{n}$ in $A^{+}$ so that either $n=1$ or if $m<n$, then $\textrm{crit}(j_{a_{1}}[j_{a_{2}}]...[j_{a_{m}}])<\gamma$. Then $M$ is finite; if $M$ were infinite, then the tree $M$ would have an infinite branch consisting of all substrings of $(a_{n})_{n=1}^{\infty}$, but for such an infinite branch, the sequence of elementary embeddings $(j_{a_{1}}[j_{a_{2}}]...[j_{a_{n}}])_{n=1}^{\infty}$ contradicts the Laver-Steel theorem.

Let $F$ be the set of all maximal strings in $M$ (in other words, $x\in F$ if and only if $xa\not\in F$ whenever $a\in A$) and let $L=M\setminus F$. Then define a binary operation $*$ on $M$ according to the following rules:

i. $x*y=y$ whenever $x\in F$

ii. $x*a=xa$ whenever $x\in L,a\in A$

iii. $x*y=(x*z)*xa$ whenever $y=za$ and $x\in L$.

One can show that $*$ produces a well-defined total operation on $M$ the same way that one can show without resorting to large cardinals that the Laver table operation operation on $\{1,...,2^{n}\}$ is well-defined. Indeed, the structure $(M,*)$ is a generalization of the notion of a Laver table (the structure $(M,*)$ is self-distributive although we shall not use the self-distributivity of $(M,*)$ in this answer).

Define a mapping $\Phi:(M,*)\rightarrow(\mathcal{E}_{\lambda}/\equiv^{\gamma})$ by letting $\Phi(a_{1}...a_{n})=j_{a_{1}}[j_{a_{2}}]...[j_{a_{n}}]$ whenever $a_{1}...a_{n}\in M$. Then one can show that $\Phi(x)*\Phi(y)=\Phi(x*y)$ by a double induction which is descending on the length of $x$ and for each individual $x$ one uses an induction which is ascending on the length of $y$. In other words, $\Phi$ is a homomorphism. However, we have $\Phi[M]=\langle\{j_{a}|a\in A\}\rangle/\equiv^{\gamma}$ since the algebra $M$ is generated by $A$. Therefore the algebra $\langle\{j_{a}|a\in A\}\rangle/\equiv^{\gamma}$ is a finite algebra.

Since each algebra $\langle\{j_{a}|a\in A\}\rangle/\equiv^{\gamma}$ is finite, the set $\{\textrm{crit}(j)|j\in\langle\{j_{a}|a\in A\}\rangle\}\cap\gamma$ is finite. Therefore, the set $\{\textrm{crit}(j)|j\in\langle\{j_{a}|a\in A\}\rangle\}$ has order type $\omega$.

$\endgroup$

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.