Exercise 6.5.1

Prove the associate law ( α β ) γ = α ( β γ ) .

Answers

Lemma 1. 0 α = 0 for all ordinals α .

Proof. We show this by transfinite induction on α . For α = 0 we have 0 α = 0 0 = 0 by Definition 6.5.6a. Now suppose that 0 α = 0 so that we have

0 ( α + 1 ) = 0 α + 0 = 0 + 0 = 0

by the induction hypothesis, Definition 6.5.6b, and Definition 6.5.1a. Lastly, suppose that α 0 is a limit ordinal and that 0 β = 0 for all β < α . Then by Definition 6.5.6c we have that

0 α = sup { 0 β β < α } = sup { 0 β < α } = sup { 0 } = 0

by the induction hypothesis. This completes the proof. □

Lemma 2. Ordinal α is a limit ordinal if and only if β + 1 < α for every ordinal β < α .

Proof. ( ) We show this by contrapositive. So suppose that there is a β < α such that β + 1 α . Then by Lemma ?? we have also that β + 1 α , from which it follows that α = β + 1 so that α is a successor ordinal and not a limit ordinal.

( ) Suppose that β + 1 < α for every β < α . Suppose also that α = γ + 1 is a successor ordinal. Then clearly γ < γ + 1 = α so that also γ + 1 < α , which is an immediate contradiction. Hence it must be that α is a limit ordinal.

Note that the bi-conditional is vacuously true for α = 0 . □

Lemma 3. Suppose that α 0 is an ordinal and β 0 is a limit ordinal. Then α β is a limit ordinal and α β 0 .

Proof. Since α 0 we have that α 1 . Now consider any γ < α β . We claim that γ α δ for some δ < β . Suppose to the contrary that γ > α δ for all δ < β . Then γ is an upper bound of the set { α δ δ < β } . But from this it follows that

γ sup { α δ δ < β } = α β

by Definition 6.5.6c, which contradicts the definition of γ . Hence the claim is true so that γ α δ for some δ < β . Then we have by Lemma 6.5.4a and the fact that 1 α that

γ + 1 α δ + 1 α δ + α = α ( δ + 1 ) < α β

by Exercise 6.5.7a since δ + 1 < β since β is a limit ordinal. Note that we also used Definition 6.5.6b. Thus since γ + 1 < α β and γ was arbitrary it follows from Lemma 2 that α β is a limit ordinal.

We also have that α 0 and 0 < β (since 0 β ) so that by Exercise 6.5.7a above and Definition 6.5.6a

0 = α 0 < α β .

Thus α β 0 . □

Main Problem.

Proof. We show this by transfinite induction on γ .

First for γ = 0 we have

( α β ) γ = ( α β ) 0 = 0 = α 0 = α ( β 0 ) = α ( β γ ) ,

where we have used Definition 6.5.6a repeatedly. Now suppose that ( α β ) γ = α ( β γ ) for ordinal γ . Then

( α β ) ( γ + 1 ) = ( α β ) γ + α β (by Definition 6.5.6b) = α ( β γ ) + α β (by the induction hypothesis) = α [ β γ + β ] (by the distributive law, see Exercise 6.5.2) = α [ β ( γ + 1 ) ] . (by Definition 6.5.6b)

Now suppose that γ 0 is a limit ordinal and ( α β ) δ = α ( β δ ) for all δ < γ . First if β = 0 then

( α β ) γ = ( α 0 ) γ = 0 γ = 0 = α 0 = α ( 0 γ ) = α ( β γ )

where we have used Definition 6.5.6a and Lemma 1. So assume that β 0 . Then we have ( α β ) γ = sup { ( α β ) δ δ < γ } = sup { α ( β δ ) δ < γ } by Definition 6.5.6c and the induction hypothesis. Then, since β 0 , by Lemma 3 we have that β γ is a limit ordinal and β γ 0 . From this and Definition 6.5.6c we have that

( α β ) γ = sup { α ( β δ ) δ < γ } = sup { α δ δ < β γ } = α ( β γ )

as desired. This completes the inductive proof. □

User profile picture
2024-07-15 11:42
Comments