Abstract
Abstract
A theory T is tight if different deductively closed extensions of T (in the same language) cannot be bi-interpretable. Many well-studied foundational theories are tight, including
$\mathsf {PA}$
[39],
$\mathsf {ZF}$
,
$\mathsf {Z}_2$
, and
$\mathsf {KM}$
[6]. In this article we extend Enayat’s investigations to subsystems of these latter two theories. We prove that restricting the Comprehension schema of
$\mathsf {Z}_2$
and
$\mathsf {KM}$
gives non-tight theories. Specifically, we show that
$\mathsf {GB}$
and
$\mathsf {ACA}_0$
each admit different bi-interpretable extensions, and the same holds for their extensions by adding
$\Sigma ^1_k$
-Comprehension, for
$k \ge 1$
. These results provide evidence that tightness characterizes
$\mathsf {Z}_2$
and
$\mathsf {KM}$
in a minimal way.
Publisher
Cambridge University Press (CUP)