Interactive activation accounts of morphological decomposition: finding the trap in mousetrap?

Many accounts of the processing of morphologically complex words have been elaborated within the interactive activation model of word identification. Conceptually, this model adopts a "segmentation-through-recognition" approach to morphological decomposition, which assumes that a complex word activates representations of constituent morphemes as well as the representation of the whole word. However, a detailed consideration of the assumptions of interactive activation frameworks reveals that current implementations of the model are incapable of achieving segmentation-through-recognition.