The 'Bohrification" program in the foundations of quantum mechanics implements Bohr's doctrine of classical concepts through an interplay between commutative and non-commutative operator algebras. Following a brief conceptual and mathematical review of this program, we focus on one half of it, called "exact" Bohrification, where a (typically noncommutative) unital C * -algebra A is studied through its commutative unital C * -subalgebras C ⊂ A, organized into a poset C (A). This poset turns out to be a rich invariant of A. To set the stage, we first give a general review of symmetries in elementary quantum mechanics (i.e., on Hilbert space) as well as in algebraic quantum theory, incorporating C (A) as a new kid in town. We then give a detailed proof of a deep result due to Hamhalter (2011), according to which C (A) determines A as a Jordan algebra (at least for a large class of C * -algebras). As a corollary, we prove a new Wignertype theorem to the effect that order isomorphisms of C (B(H)) are (anti) unitarily implemented. We also show how C (A) is related to the orthomodular poset P(A) of projections in A. These results indicate that C (A) is a serious player in C * -algebras and quantum theory.