‘History’, as it was understood by medieval Christians, was a broad concept with many meanings. It could be defined as a written record, compiled through processes inherited from classical Greek and Roman authors. But even when it proclaimed itself to be ‘factual’, the work of those who shaped the past in the Latin Middle Ages was different from that of their pagan ancestors. Although they often made claims about veracity, even when dealing with the miraculous, they did not ‘absolutize’ history (as de Lubac put it). Although they had a sense of universal history, time instead related to and unfolded within a framework conditioned by the Incarnation of the Word, and by the moral sense that this history-transforming event could impart to listeners and readers. The biblical orientation of history gave time a clear beginning and also predicted an apocalyptic end. But even as time moved relentlessly forward in this cosmic sense, the liturgy made it constantly spiral backward, rendering past sacred events present through ritual commemoration. Such liturgical celebration of time had many layers, mingled and arranged according to the calendar, with its varying and its fixed cycles of feasts, voiced through psalmody, readings from the Old and New Testaments and the lives of the saints.

How the past was known, both by individuals and within communities, varied from one specific local community to another, for the liturgy was ever changing, especially as new feasts, and feasts of new saints, were added to suit particular needs. Some knew about certain past events from what they read, as had long been the case, but in the Middle Ages many more knew about the past from various reenactments: ritual actions, dramatic productions, sermons and tales they heard about biblical characters and saints, encounters with art and architecture and from contact with shrines and relics.

Much of medieval history-making was thus memorial in nature, bringing the past forward, again and again, to recall its individual or communal significance. At the heart of the medieval Christian understanding of the past was a simple, foundational command from Christ: ‘Hoc facite in meam commemorationem’ [‘Do this in remembrance of me’] (Luke 22. 19).

Among the most celebrated historians of twelfth-century England, William of Newburgh (1135/6–c. 1200) has a long-standing reputation for bookishness. According to Rachel Fulton: ‘The world came to William in books, and it was through books that he was most comfortable making sense of it.’ Fulton follows the Rolls Series editor, Richard Howlett, in adducing a long list of sources used by William in his Historia Anglorum, including the histories of Symeon of Durham (d. after 1129) and Henry of Huntingdon (d. c. 1157). Yet there is good reason to doubt whether William's reading was as extensive as Howlett suggested – and this not just because of the apparently negligible size of Newburgh's library. Indeed, John Gillingham has demonstrated that many of William's alleged sources, Symeon and Henry included, almost certainly reached him second-hand, mediated through the compiling efforts of another Yorkshire historian, Roger of Howden (d. c. 1201). This dependence on another northern writer, one whose history could be found in the library at nearby Rievaulx (now London, Inner Temple, MS 511.2), suggests the local and limited nature of William's bibliographical resources. This sense of regional limitation is reinforced by his other major work, a Marian commentary on the Song of Songs. Though he could find in patristic writings ‘quomodo … vel in ecclesiam vel in excellentis meriti animam idem nuptiale carmen intelligi debeat’ (‘how this nuptial song should be understood either in regard to the Church or a soul of distinguished merit’), William was apparently unaware of the more recent Marian commentaries on the Song by such writers as Rupert of Deutz (d. c. 1129) or Honorius Augustodunensis (d. 1154). In light of these limitations, William's elaborate historical and exegetical writings are all the more impressive, revealing that even with a small library one can still be bookish.

Were he to be judged solely on the basis of these two major works, William might seem to present a split authorial personality. At the very least, his Historia and his commentary could almost be read as unrelated productions, each composed in response to a specific request by a different Cistercian official: the Historia for Ernald, abbot of Rievaulx (r. 1189–99), and the commentary for Walter, abbot of Byland (r. 1142–96).