We implement a computational model of first and second language reading over the lifespan. The model simulates individual differences in effects of psycholinguistic factors. Individual differences in frequency effects are due to exposure, not vocabulary size. The model provides causal explanations of correlations in theories of word processing.