0
$\begingroup$

Let $A$ be a set, then there is a function $c :P(A)\setminus\lbrace ∅\rbrace→ A$ such that $c(X) ∈ X$ for every nonempty subset $X$ of $A$.$(*)$

The axiom of choice is stated as:

Let $\mathcal A$ be a set of nonempty sets. Then there is a function $f : \mathcal A→\bigcup \mathcal A$ such that $f(A) ∈ A$ for all $A ∈\mathcal A$.

I need to prove that the first statement implies the axiom of choice:

Here is an answer I found in this site:

3 Statements of axiom of choice are equivalent

I expand it as:

Let $\mathcal A$ be a set of non-empty sets,then $\bigcup \mathcal A$ is also a set by MK4. From $(*)$ we know that there is a function $c :P(\bigcup\mathcal A)\setminus\lbrace ∅\rbrace→ \bigcup\mathcal A$ such that $c(X) ∈ X$ for every nonempty subset $X$ of $\bigcup\mathcal A$.

And then the author says "restrict to $\mathcal A$" then we have a choice function.

But I don't think $\mathcal A$ is a subset of $\bigcup\mathcal A$.

So what does the restriction means. How may I deal with that?

Or other way to prove it would be also appreciated. Thanks!

$\endgroup$
  • 1
    $\begingroup$ You do not need $\mathcal A$ to be a subset of $A,$ or a subset of $\bigcup\mathcal A.$ What you need is that every element of $\mathcal A$ is a nonempty subset of $\bigcup\mathcal A.$ $\endgroup$ – bof Jul 17 '17 at 7:18
3
$\begingroup$

But I don't think $\mathcal{A}$ is a subset of $\bigcup \mathcal{A}$.

But certainly $P \left( \bigcup \mathcal{A} \right) \setminus \{ \varnothing \}$ is a superset of $\mathcal{A}$ (check it), so restriction $c \upharpoonright \mathcal{A}$ makes sense.

| cite | improve this answer | |
$\endgroup$

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.