It has been remarked by several historians and commentators on warfare, whether of ancient, medieval or modern, that descriptions of battles can rarely afford any real idea of what actually occurred during the different phases of a violent confrontation. In the first place, those involved, whether in positions of authority or not, rarely if ever know what is happening away from their own particular field of vision throughout the battle—they receive reports (which may or may not be accurate), they respond to the reports, they hear or experience the results, but the connection between orders and effects is impossible to trace exactly. Equally, there is inevitably a tendency to dramatize, to present things, for whatever reason, in a worse or better light, to exaggerate the results of the actions of particular individuals, together with the inevitably different views of what happened, and why, of those involved in, or observing, the fighting from different vantage points. Finally, there is also a tendency for those who win the battle to see the results in terms of their original assumptions or intentions, so that their version of events, and the causal relationships they assume, will be artificially neat and tidy, in contrast with the reality.
These factors are as pertinent to Byzantine literary sources, whether officially sanctioned (or commissioned) or not, as they are to those of any other period, and it is for the most part impossible to extract from a chronicle account of a battle any idea of what actually went on apart from a crudely generalized picture: “The imperial troops advanced with the cavalry in the centre; the enemy line held firm at first, but eventually gave way; the imperial reserve cavalry were then brought up and the enemy retreat turned into a rout. Many were slain on both sides.” Such accounts are scattered through all the Byzantine histories and chronicles, but tell us very little of what individual soldiers experienced, how the enemy forces appeared to them or the nature of the hand-to-hand combat which was involved. Assuming a certain degree of descriptive accuracy, they may give us an idea of the use by one side or another of a particular formation, which can in turn tell us whether certain