The circular birefringence of polycrystalline polymers is invariably obscured by strong linear birefringence. To parse the two mechanisms of light retardation, polycrystalline spherulites of polylactide enantiomers were analyzed by Mueller matrix microscopy. Polymer films are barely optically active in normal incidence, but if illuminated obliquely they become strongly optically active. Opposite hemispheres have oppositely signed circular birefringence. The sign is independent of the enantiomer but dependent on the sense of the sample's tilt. These observations are consistent with light path inhomogeneities resulting from stacked, mis-oriented lamellae. Chiroptical commonalities based on symmetry arguments are discussed among polylactide, a single oriented water molecule, and microfabricated metamaterial arrays, as well as the first physical model of optical activity, Reusch's pile of mica plates. The latter model provides the best explanation of the circular birefringence of polylactide spherulites. The only data on the optical rotation of crystalline polymers to date come from ostensible single crystals of polylactide. The enormous, anisotropic optical rotations observed previously are in quantitative agreement with misoriented lamellae observed here. Limitations of parsing circularly birefringent systems into those showing 'natural optical activity' and those others, somehow 'unnatural', are discussed.