As long as the linguistics are the issue rather than raw inability to understand the concepts being communicated, then the model is just being a bad communicator.