This paper traces a progression of four computer-based methods for studying and fostering both the structure and the on-line development of knowledge. Each empirical technique employs ECHO, a connectionist model that instantiates the theory of explanatory coherence (TEC). First, verbal pro tocols of subjects' reasonings were modeled post hoc. Next, ECHOpredicted, a priori, subjects' text based believability ratings. Later, the bifurcation/bootstrapping method was developed to elicit and account for individuals' background knowledge, while assessing intercoder reliability regarding ECHO simulations. Finally, Convince Me, our reasoner's workbench, automated the explication both of subjects' knowledge bases and of their belief assessments; the Convince Me software per mits contrasts between the model's predictions and subjects' proposition-wise evaluations. These ex perimental systems enhance our understanding of the relationships among-and determinant fea tures regarding-hypotheses, evidence, and the arguments that incorporate them. Human reasoning and argumentation represent some of the most vexing phenomena of cognitive psychology. Whether one is opining about 0.1. Simpson's guilt or in nocence at the local pub, or explaining new logic-puzzle data to colleagues, several difficulties arise. First, the ex tent ofan individual's initial knowledge base is rarely clear: What does a person know, and what does a person not know? Determining the mechanisms by which people add to their knowledge is also difficult: How variable are indi viduals' inference engines? Finally, for a terminal corpus ofbeliefs, some propositions seem evidentiary, some seem more hypothetical, and all have varying confidence levels: How do we assess these features of a person's thinking?