System of Transliteration and Dating
Transliteration is a nightmare in a book that covers such a long period, and consistency is impossible. I have tried to combine authenticity with clarity. With Greek names, I have rejected the sometimes absurd latinized forms long used, unless, as with Aeschylus, the alternative is unrecognizable to non-experts.
So I have Herodotos and Sophokles, and Komnenos for the great Byzantine dynasty, not Comnenus. This becomes more complicated in later centuries. Ancient Thessalonika becomes Ottoman Salonika and then modern Thessaloniki, while Epidamnos, Dyrrhachion, Dyrrachium, Durazzo, Durrës are all one place in Albania at different epochs; I have used the name current in the period about which I am writing. Comparable problems arise with Hebrew, Turkish and Arabic names. Along the Croatian and Montenegrin coast, I have favoured Slav forms, since they are now in general use, so I use Dubrovnik rather than Ragusa but (lacking an equally elegant word for the inhabitants) I have called its inhabitants ‘Ragusans’.Another contentious issue is whether to use the Christian labels for dates, BC and AD, or the modern substitutes, BCE and CE, or indeed (as Joseph Needham used to recommend) a simple ‘–’ and ‘+’. Since these variants produce exactly the same dates as BC and AD I am not sure what advantage they bring; and those who are uncomfortable with Before Christ and Anno Domini are free to decide that BC and AD stand for some other combination of words, such as ‘Backward chronology’ and ‘Accepted date’.