But there was a problem. The translator had chosen to render the Syriac term for the Eastern Romans, ‘Rom’ as ‘Greeks’. This makes sense in 1300; does it make sense in 500? He had also rendered the name of the city of Edessa as ‘Urhay’, which is the name of the modern town on the ruins of Edessa; and Amida as ‘Diyarbekir’ (where the bombing took place recently, where there is a substantial library of Syriac texts, and where there is also, I believe, a US airbase). Again, do these names make sense at this period? Finally there was the usual profusion of Jacobean English: “what befel”, “thou”, etc, which the reader must mentally translate as he goes. The footnotes were studded with Syriac, which I could not sensibly transcribe, so much had to be changed to put the text online.
What should we do? There is always a case for leaving the text alone, and this is the course that I normally prefer. But in this case I chose otherwise; I fixed all three of these things. Was I right to do so?