This paper continues the development of a general theory of comprehension, focusing on story understanding the role that syntax plays therein. Partly as a reaction against the almost exclusive concern of some linguists and psycholinguists with syntax, syntactic questions play only a minor role in the current work on discourse processing, text comprehension, and conversation. Maybe this is too strong a statement as far as the field as a whole is concerned, but it certainly characterizes my own work in this area. Syntax plays a role in that it directs the reader how to parse a text into meaning units (propositions), but all the most interesting processes psychologically seem to occur at the level of meaning and language use. This imbalance needs to be redressed. Such an essential feature of language as grammar surely must be more important than we have given it credit for; and a recent paper by Givon (1989), in which he proposes that grammatical cues function as mental processing instructions, may give us some clues as to the manner in which syntactic cues are used in discourse processing. I shall show how Givon’s ideas can be incorporated into a model of discourse comprehension, how such an extended model could account for some salient empirical results that have been reported previously by others as well as myself, and I shall argue that by paying attention to the role of syntactic cues in processing, we can greatly expand the power of our model and throw some light on some currently controversial issues. Specifically, I shall show how only a single verbal signal may determine whether a text representation will be organized linearly or in terms of a topic-subtopic structure, and how syntactic cues may indicate the presence of important causal relations in a text. Thus, it is possible to arrive at rather similar mental representations of a text either via weak but general syntactic strategies, or by more powerful, knowledge-based, but domain-limited processing strategies.