The Axiom of Choice constrains every set of cardinal numbers which is linearly ordered by size to be well-ordered. By contrast, does ZF-without the Axiom of Choice (but with the Axiom of Foundation)-allow complete liberty in the ordering of cardinal numbers by size? More precisely: Is the following statement consistent with ZF? "If S is any linearly ordered set, there exists a set of cardinal numbers which is linearly ordered according to size and which is ordinally similar to S". Even without the Axiom of Foundation one can define a partial ordering among the sets of ZF. The set A precedes the set B if there exists an injective mapping of A into B. The set A strictly precedes the set B if there exists an injective mapping of A into B but no mapping of A onto B. One could then ask the same question about the linearly ordered subsets of this partially ordered class of sets.
$\begingroup$
$\endgroup$
5
-
9$\begingroup$ I actually intend to address this (or rather "partial order" instead of "linear order") in my Ph.D., and I have the general layout prepared. Some abstract technical difficulties are currently holding me back, but I hope to overcome them at some point in the foreseeable future. The tentative answer, is yes, this is consistent. $\endgroup$– Asaf Karagila ♦Aug 21, 2014 at 19:05
-
$\begingroup$ Oh @AsafKaragila, you're too modest. Your paper in Fund. Math. already has some results in this direction... $\endgroup$– David Roberts ♦Aug 21, 2014 at 23:43
-
2$\begingroup$ @David: But not more than what was known nearly 50 years ago. Yes, every partial order can be embedded into the cardinals of an extension. Here we reverse the quantifiers. $\endgroup$– Asaf Karagila ♦Aug 22, 2014 at 1:54
-
$\begingroup$ Many thanks, Asaf, for your response. I am amazed because I was sure the answer would be "No". $\endgroup$– Garabed GulbenkianAug 22, 2014 at 20:30
-
$\begingroup$ @Asaf ah, of course, my mistake $\endgroup$– David Roberts ♦Aug 23, 2014 at 22:02
Add a comment
|