Since the initial work of Daryl McCullough on the subject, the securit
y community has struggled with the problem of composing ''possibilisti
c'' information-flow properties. Such properties fall outside of the A
lpern-Schneider safety/liveness domain, and hence, they are not subjec
t to the Abadi-Lamport Composition Principle. This paper introduces a
set of trace constructors called selective interleaving functions and
shows that possibilistic information-flow properties are closure prope
rties with respect to different classes of selective interleaving func
tions. This provides a uniform framework for analyzing these propertie
s, allowing us to construct both a partial ordering for them and a the
ory of composition for them. We present a number of composition constr
ucts, show the extent to which each preserves closure with respect to
different classes of selective interleaving functions, and show that t
hey are sufficient for forming the general hook-up construction. We se
e that although closure under a class of selective interleaving functi
ons is generally preserved by product and cascading, it is not general
ly preserved by feedback, internal system composition constructs, or r
efinement. We examine the reason for this.