The paper reports results and discusses implications of three studies on captioning films and television programs for deaf audiences. Study I dealt with the short-term visual memory required in recognition of fingerspelled letters, while study II focused on serial position effects and forward making (interference produced by first stimulus on the second), and study III concerned the generalizeability of inferences from results with fingerspelling across systems of linguistic symbols. Implications of the research for short-term memory strategies and letter recognition are discussed in terms of individual differences in coding strategies, letter recognition and task demands, and the interaction of the task and individual differences in coding strategies. Practical implications for captioning are seen to cover future research study design and to indicate the possibility that because of differences in cognitive modes there is no one optimal means of captioning material for deaf students. (CL)