User blog comment:Tetramur/BEAF above SVO - comparison/@comment-5150073-20200219121934/@comment-37993808-20200219155617

> But as Plain'N'Simple pointed out, explanation by Bowers conflicts the alternative approach by Saibian, doesn't it?

I don't think so. Approach by Bowers shows how many entries array has, but approach by Saibian shows how structure is built.

> Doesn't it mean that Bowers original description does not characterise a single expansion?

No. I understood the logic of Bowers: we don't begin with one expression, but rather with a plethora of it with all being one order in ordinals, e.g. X^^X, X^(X^^X), X^X^(X^^X)... And Bowers is transcending from one expression to another without second being equal to first. (X^^X)^X is not equal to X^(X^^X*X), but they are of the same order. Simply look at this: first is the expansion of X^^X, while second is the expansion of X^(X^^X).