I tried to prove this proposition, if $\alpha \preceq \beta$ then there is a unique ordinal $\gamma$ such that $\alpha + \gamma = \beta$, by order type not by transfinite induction.
My attempt:
- $W = \{ \xi \in \beta \mid \alpha \preceq \xi \}$ is a set of ordinals so, $W$ is a woset.
- Every woset is order isomorphic to a unique ordinal, therefore, exist a unique ordinal $\gamma$ such that, $W$ and $\gamma$ are ordered isomorphic.
My problem: I strongly believe that, $\gamma$ is the solution but I cannot prove that, $\alpha + \gamma = \beta$.
Def: $\prec := \in$. The ordinal addition is defined by induction.
I have read some similar questions. For example, (1) just talk about transfinite induction (limit case) and (2) just talk about the uniqueness. I hope your helps.
Your demonstration that $W$ has the order type of an ordinal is correct. Once you have that $W$ is isomorphic to a unique ordinal $\gamma$, you're basically done, as $\alpha + \gamma = \beta$ by the definition of ordinal addition:
By your definition of $W$, $$W := \{ \xi \in \beta | \alpha = \xi \text{ or } \alpha \in \xi \},$$ it's clear that $\beta$ is the disjoint union of $\alpha = \{ \xi \in \beta | \xi \in \alpha \}$ with $W$.
But the disjoint union of $\alpha$ with $W$ is order-isomorphic to the disjoint union of $\{ 0 \} \times \alpha$ with $\{ 1 \} \times \gamma$ equipped with the dictionary order, since $\gamma$ is the unique ordinal with the order type of $W$. This is the definition of what it means for $\alpha + \gamma$ to equal $\beta$.
If you do it this way, no induction argument is required. The heart of the question is whether there is an argument that works--and I quote--
and this would seem to be what the original post is asking for.