I have this problem from Enderton's Elements of Set Theory which says, "Assume that $\lt$A, < $\gt$ is a well-ordered structure with ordinal number $\alpha$ and that B$\subseteq$A. Let $\beta$ be the ordinal number of the well-ordered structure $\lt$B, <$^\circ$$\gt$ and show that $\beta \in \alpha$ or $\beta=\alpha$".
My first approach was to let $a$ be the least element of A-B and say that E(seg($a$))=$\beta$ (where E is the epsilon-image function) and continue from there, but now I realize this won't work, as B isn't necessarily an initial segment of A; we could have A={a,b,c,d,e} (where a< b< c< d< e) and B={b,c,d,e}. Then B$\subseteq$A, but E(seg($a$))$\neq\beta$. So now I'm not sure what to try. It seems obvious that it's true, but I'm having trouble turning that into an actual proof.
I'm thinking now that, using trichotomy of ordinals, I really just have to show that it's NOT true that $\alpha\in\beta$, so I might be able to do a proof by contradiction. If anyone might be able to give me a hint as to where I should look next, I'd appreciate it.
Edit: Could I argue something like this?
Suppose $\alpha\in\beta$. Then $\lt\alpha, \in_\alpha\gt$ $\simeq$ $\lt A$, <$\gt$ $\simeq$ $\lt$ seg(a), <$^\circ\gt$ for some a$\in$B, but B$\subseteq$A so a$\in$A, and we have A is isomorphic to a segment of itself, which is a contradiction?
Hint: Can there be an order-preserving embedding $\beta \to \alpha$ when $\alpha \in \beta$ ?