Like their modern counterparts, medieval historians operated under epistemological and methodological constraints governing evidence. They perceived critical differences between real events (res factae) and fictional narratives (fabulae), and deemed certain forms of testimony and modes of representation more appropriate to historiography than others. Nevertheless, a wide gulf separates medieval scriptores rerum from modern documentary historians. Whereas the latter are concerned with verifiable, referential propositions about the past and hope to restrict as much as possible the ‘distortions’ of aesthetics and style, the former would have been baffled by the notion of a strictly factual presentation of events and emphasised instead rhetorical techniques of figuration, persuasion, and moralisation. In this respect, they take their lead from Roman predecessors, especially Sallust, whose speeches, though attributed to historical actors, were legitimately understood by medieval readers as models of oratorical invention. Cicero justifies rhetorical history in the De oratore, arguing that historians must be ‘exornatores rerum’ (ornamenters of things done), rather than ‘tantummodo narratores’ (mere relaters of facts) (De oratore, 2.12.54). Quintilian likewise argues that historians should alleviate the tediousness of factual reporting with a liberal, vivid use of figures (Institutio oratoria, 10.1.31). The seventh-century encyclopaedist Isidore of Seville augurs a change in Western historiography when he argues for the priority of eyewitness testimony over second-hand sources, content over form. Yet like all medieval thinkers, he also believes that sublime truth (veritas) takes precedence over the merely true (verus), metaphysical beliefs over empirical observation.