T. E. Lawrence’s Undergraduate Dissertation – Page 100

It has been hailed by Rey, and by Professor Oman following him, as an untouched example of Latin military architecture. This claim seems a little dangerous, when it is remembered that Kerak was a Byzantine fortress before it became Crusader, and that after this it was the seat of a powerful Arab principality, and that finally Beibars' presence is shown by his name on one of the towers. It may well be that the share of Payn of Nablous in the building of it is infinitesimal. At least, until there is better material to work upon, elaborate deductions from it as to the state of Latin military architecture in 1140 are quite out of place. Beibars' buildings in particular are very easily confused with Crusader work, for it seems most likely that they are a partial imitation. To consider the Crusading castles in their chronological order is extremely difficult: they are mainly a series of exceptions to some undiscoverable rule. To begin with, the castles of the Antioch and Edessa principalities means beginning with a string of nearly untouched Byzantine fortresses. At Antioch the walls have no signs of Latin interference: and the castle, with its long, flimsy wall with the ridiculous buttresses, is evidently residential and not a post of military importance. In the Antioch province the celebrated castle of Harenc (Harim ...) is Byzantine and Arab. It has a gigantic dry ditch cut in the rock on which the castle stands, and this alone would be (argument) for a Greek origin : when the case is supported by a ground plan (14)