Consider the three following large cardinal axioms:

  1. there exists a nontrivial elementary embedding $j:V\to V$.
  2. there exists a n.e.e. $j:V\to M$ such that $M^{j^\omega(crit(j))}\subseteq M$.
  3. there exists a n.e.e. $j:V_{\lambda+2}\to V_{\lambda+2}$ for some $\lambda$.

Kunen's inconsistency theorem states that, in ZFC (with replacement extended to formulas containing $j$), these three statements are false. However there is no such (known) result within ZF (with replacement sufficiently extended) alone.

Clearly (1) implies (2). I remember seeing somewhere that (2) implies (1), but I don't remember where, so I might be mistaken. Does anyone knows how these three statements relates to each other?

EDIT: this question already deals with the case (1)=>(3)

  • 1
    $\begingroup$ I find statement (1) ambiguous without further elaboration about how it is to be formalized. $\exists j$ is a second-order quantifier, so I guess we understand this in GBC. In any case, one intends $\text{ZFC}(j)$, for otherwise it is weak. To express "$j$ is elementary" seems to require a truth predicate, whose existence is not provable in GBC. Perhaps one means "$\Sigma_0$-elementary and cofinal," which is an expressible approximation to full elementarity, for it implies elementarity for standard formulas as a scheme, but this is weaker than elementarity with respect to a truth predicate. $\endgroup$ – Joel David Hamkins Dec 2 '17 at 14:01
  • 1
    $\begingroup$ In the case of the Kunen inconsistency, proved in ZFC, it was fine to be a little sloppy, because the Kunen argument in effect ruled out the first-order assertion of statement (3), which is a clear consequence of whatever one might have meant in (1). But now that set theorists are considering the specific statement (1) seriously, I think it is time for set theorists to be a little more careful in formulating the statement precisely. $\endgroup$ – Joel David Hamkins Dec 2 '17 at 20:43


This is a very recent work in progress of myself with Juan Aguilera.

Definition. We say that $\kappa$ is a Kunen cardinal if there is a nontrivial elementary embedding $j\colon V_{\lambda+2}\to V_{\lambda+2}$ with $\lambda=\sup j^n(\kappa)$.

(Sometimes it is easier to talk about the critical points of the sequence, and sometimes on $\lambda$, like in the case of $I0$ and such.)

Theorem. If $\kappa$ is a Reinhardt cardinal, i.e. the critical point of an elementary embedding $j\colon V\to V$, then $\kappa$ is the limit of Kunen cardinals.

Proof. Note that being a Kunen cardinal is a first-order property in the language of set theory. And note that a Reinhardt cardinal itself is Kunen by a fairly easy verification.

Let $j\colon V\to V$ be a nontrivial elementary embedding, and $\kappa$ its critical point. Now, it is easy to see that the set $A\subseteq\kappa$ defined as $\{\mu<\kappa\mid\mu\text{ is a Kunen cardinal}\}$ satisfies $\kappa\in j(A)$. Therefore it follows that $A$ is in the normal measure derived from $j$, and therefore it is stationary.

In particular, a Reinhardt cardinal is the limit of Kunen cardinals. $\square$

  • 2
    $\begingroup$ We thought about the obvious $I(-1)$ name, but decided it's a silly name and it should be given a proper name. Kunen cardinals seemed appropriate. $\endgroup$ – Asaf Karagila Nov 2 '17 at 13:43
  • 1
    $\begingroup$ What do you know about the relation between (2) and (3)? $\endgroup$ – Yair Hayut Nov 2 '17 at 13:46
  • 1
    $\begingroup$ Well. I have no idea. I'm inclined to say that just ordinal-sequences are not enough here to ensure that (2) implies (3); and it seems to me that cutting the universe at the least critical cardinal above the least Kunen would mean that (3) does not imply (2) either. But again, I don't think there is too much known about how closure under ordinal-sequences works in ZF. $\endgroup$ – Asaf Karagila Nov 2 '17 at 13:50
  • 2
    $\begingroup$ Since the existence of a Reinhardt cardinal implies every set has a sharp, a Kunen cardinal is also strictly weaker than a Reinhardt cardinal in consistency strength. $\endgroup$ – Gabe Goldberg Nov 3 '17 at 23:27

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.