Abstract
The notion of "size effect" is shown to contain inherent conceptual difficulties. A new framework to be called "shape effect theory", incorporating "shape" as a key variable in place of "size", is proposed to overcome those difficulties. Several cases of pullout rupture of concrete anchors and flexural rupture of beams are analyzed satisfactorily using this framework. Furthermore, a hypothesis of "hidden dimensions" is proposed to alert to a possibility that an apparent size effect that is really a shape effect is related to some hidden dimensions in the object that are ignored in the theoretical model used for observation.
