So how do we know what happened at the Battle of Clontarf? Or rather why do historians hum and haw and hedge their bets when you ask them what happened at Clontarf in 1014?
The answer is that historians try to reconstruct events from accounts in historic texts, and historic texts are a bit like newspapers — you shouldn’t believe everything you read in them! Unfortunately, very few manuscripts survive from the Ireland of Brian Boru and often we are reliant on copies of copies of texts for our information (some written several centuries after the events they describe). A good example can be found on folio 55 recto of TCD MS 1282 (a sixteenth-century manuscript on display in the Emperor of the Irish exhibition in the Long Room and in the accompanying online exhibition). This is the account of the battle found in the Annals of Ulster, which is generally considered to be one of the most reliable historic texts from medieval Ireland. Even so, if we look at the image of the manuscript page we can see that this text is not the same as it was when first written down in this manuscript, let alone when it was first composed several centuries earlier.
For example, one of the few facts that everyone can tell you about the Battle of Clontarf is that the Viking chief Brotor killed Brian Boru. But did he? Between the fifth and sixth lines from the bottom of the first column we find a revealing little bit of interlinear writing over Brotor’s name, which translates as who slew Brian. The original author (in the larger text) simply tells us that Brotor was a leader of the Scandinavian fleet and that he fell in the battle. A later scribe clearly thought it necessary to insert the ‘fact’ that Brotor also killed the Irish king; the original eleventh-century author may actually have been unaware of this ‘fact’ and indeed it may not even be true. Another example can be found in the fourth line of the second column, where there are two gaps in the text. It appears that words have been lost in the copying process over the generations, which should serve to remind us that our texts are not static, but subject to change (loss and gain) at every point of transmission.