The battle had reached a crisis point. Hard-pressed by the English, the Normans and their allies were convulsed by a rumour: that William had been killed. The left wing broke and ran. The duke, William of Poitiers wrote,

In his work on the twelfth-century birth of heraldry, Adrian Ailes uses this episode as an example of the difficulties that the knights of the pre-heraldic eleventh century had in being recognized on the battlefield. He cites similar examples: the rallying of the English by Edmund Ironside in 1016, when the king also lifted his helmet to show his face, Robert Curthose's near-slaying of his father at Gerberoy in 1079, and Rufus's near-miss in 1091; on the two latter occasions the assailant's hand was stayed only when he recognized his intended victim's voice. Ailes argues that before the development of heraldry it was almost impossible for the individual warrior to be identified. In the first third of the twelfth century, he suggests, shield designs became more systematized and ordered, but only after 1150 were the designs on a knight's shield used consistently, inherited by sons from their fathers, and followed a set of rules as to their shape, position, and colour, which are the defining characteristics of heraldic display.