A model is presented which details the ways in which children in different language communities acquire the morphophonological structure of their languages. In this model, the processes of rote, combination, and analogy are integrated into a single processing goal stack. The model views learning as a cyclical process in which acquisition leads to application, application leads to correction, and correction leads to renewed acquisition. The processes of acquisition, application, and correction are formulated in relatively precise terms. The empirical aspects of the model are then summarized as a series of 16 claims. The validity of these 16 claims is evaluated in the light of experimental and diary data on the acquisition of Arabic, Chinese, English, Finnish, French, German, Hungarian, Latvian, Russian, and Spanish. Experiment 1 examines the acquisition of nine patterns in Hungarian morphophonology by children between 2 and 6 years of age. It indicates that the sequence of acquisition of morphophonological productions is determined chiefly by applicability as defined by the model. Experiment 2 indicates that preschoolers tend to avoid acquisition of synonymous inflections. Experiment 3 gives some evidence that Hungarian 5-year-olds can use information on allomorphic variations to determine the use of allomorph selection productions. Experiment 4 examines the learning of the German plural formation by children between the ages of 3 and 11. The order of acquisition of strategies in plural formation was determined by the interaction of applicability and correctness, as suggested by the model. Experiment 5 gives evidence for the ordering of rote, combination, and analogy as a goal stack in the formation of German gender. The concluding chapters review the success of the model in accounting for the available data on the acquisition of morphophonology. In general, the model constitutes an advance in our ability to account for the acquisition of complex linguistic structures in terms of specific cognitive processing. Every normal newborn is, in a sense, a subject in a vast experiment of nature. Each of these newborns brings to this experiment some set of abilities that will allow him to master his native tongue. Nature has so arranged things that different groups of subjects learning different languages will have to use these abilities to acquire widely differing sets of target language structures. By studying the ways that children come to control these structures, we can hope to make inferences about the universal set of language-learning abilities common to all children. The present report seeks to provide a detailed characterization of the acquisition of the linguistic system called morphophonology. It attempts to do this by describing a set of universally human abilities which are involved in the acquisition of morphophonology. Some writers (Chomsky 1965; Lenneberg 1967; Stampe 1969) have underscored the importance of specifically linguistic abilities in the acquisition of language. Others (Anderson 1975; Braine 1963) have stressed the importance of general cognitive mechanisms. The present model makes no a priori assumptions regarding the ultimate source of the abilities evidenced by the child. Rather, it attempts to focus on the construction of an explicit description of these abilities. The description to be offered here relies heavily on the many important insights achieved by Slobin (1973) in his investigations of universal abilities in language acquisition. The hope is that, once such explicit descriptions become more generally available, the discussion regarding the ultimate source of these abilities could be rejoined with a new clarity and purpose. The …