โ Scribed by Richard Mansfield
No coin nor oath required. For personal study only.
The definition of the ultrapowers operation can be so phrased as to use an arbitrary complete Boolean algebra in place of the usual 2 I. This more general rotion of ultrapower shall be called Boolean ultrapower while the world "ul;rapower" by itself will have only the usual meaning. I do not know of a single proof about ultrapowers which cannot be translated directly to a corresponding proof about Boolean ultrapowers.
At the same time the extra flexibility of the more general notion pays great dividends; we will be able to construct isomorphic Boolean ultrapowers for elementarily equivalent models, good ultrafilters, and saturated ultrapowers. This can all be done without recourse to any false axioms for set theory (e.g. the G.C.H.). The concept of Boolean ultrapower is implicit in the work of Scott and Solovay on Boolean valued set theory and is explicit in Vop~nka's treatment of the same subject. What we do ;,s to develop Boolean ultrapowers as a serious and useful tool in Model Theory. ยง 1. Boolean ultrapowers ?1 R(fl, "", fn) = V { /~ f/(mi) โข (m s) ~ M n A R(ml, ..., m n) } i=l 299 el v ft(mi) R(ml ..... m n) i=l
๐ SIMILAR VOLUMES
The theory of stochastic processes has been traditionally developed in terms of random variables and their joint distributions. This is not surprising since the definition of a stochastic process is abstracted from numerical statistical data that are empirically observed and to which the notion of a