This paper describes and illustrates a simulation model for the processing of grammatical elements in a sentence, focusing on one aspect of sentence comprehension: the assignment of the constituent elements of a sentence to the correct thematic case roles. The model addresses questions about sentence processing from a perspective very different from the conventional perspective found in computational linguistics. As a role-assignment model, it consists of two sets of units: one for representing the surface structure of the sentence and one for representing its case structure. Words are treated as patterns of activation; knowledge about them is stored in distributed form in the connections in a large network of simple neuron-like processing units. The model shows facility in dealing with such problems as frame selection, role assignment, and disambiguation, and it suggests a natural way to resolve unappealing aspects of the idea that there is a fixed set of individuated case roles. The simulation model is currently able to process only one-clause sentences, but extensions to multi-clause sentences are considered possible. (Author/MSE)