The work towards a human-communication paradigm for the guidance of machines within the frame of the new multidisciplinary approach of machine-mediated communication (e.g. Elsayed, ) gives rise to important subordinate questions. With the development of a pattern language in mind, this paper outlines the author's structural model of content information for second language learning materials and from that frame examines the question of what factors influence the difficulty of comprehension of a media object. This is one approach to the question of how media objects carry information. The discussion of parameters of difficulty is grounded in related issues in second language learning materials, which are germane here: the much examined notion of ‘considerate text’ has recently evolved into looking at ‘considerate multimedia,’ an even more complex matter. A tentative list of parameters of difficulty in media object comprehension is developed here, opening questions of significance and applicability for machine automation.