A remarkable number of twelfth-century histories from the lands controlled by the kings of England survive in manuscripts that have been identified as autographs, that is manuscripts entirely or partially written by the author's own hand. There is a logic to this. New histories, even when these were heavily derived from existing works, were of unproven value, thus the author was the person with most reason to invest time and resources into the creation of a copy of his work. Indeed it has sometimes been assumed that histories must have begun life as autograph manuscripts, from which copies were made in monastic scriptoria or by professional lay scribes. The idea of an autograph is attractive, containing the promise of a physical connection to the author, and a potential insight into his working process through alterations visible in the manuscript. However, the composition of a text was not synonymous with writing in this period, nor with the making of books. Texts could be dictated to scribes or drafted on wax tablets and scraps of parchment before being given to a scribe to copy. In the context of historical writing, such practices might be driven by the reliance on existing texts for the crafting of new ones. Thus Orderic Vitalis, writing at the abbey of SaintEvroul in Normandy in the first half of the twelfth century, complained in his Historia Ecclesiastica; ‘I have no scribes who can make excerpts of my material for me’. Scribes did not just compile material, they could also be used to produce neat copies of texts and documents. In the 1160s Wace, author of a vernacular history of the dukes of Normandy, the Roman de Rou, grumbled that the nobility were no longer sufficiently generous to ‘present me with enough to employ a scribe for a month’. Similarly, Robert of Torigni, who worked on a chronicle and a version of the Gesta Normannorum Ducum, employed a man named Adam as his secretary (scriba) during his time as abbot of Mont Saint-Michel, an office Robert held from 1154 to 1186. In such circumstances it is extremely difficult to identify authors’ hands with absolute confidence, even in manuscripts that can be traced to their lifetimes and places of work.