Interesting article that discusses mottes as "information gathering points and checkpoints" and the reason some are placed on river crossings. Mary Higham does make the strong argument that during the medieval period rivers were a much more difficult barrier to movement that our current canalised and tamed modern rivers appear. She also adds six possible sites not then on the SMR. Unfortunately many of these 'mottes' are very doubtful natural features and suchlike. This does tend to weaken the argument that she is making. However, the evidence of small castles that acted as checkpoint of river crossings is rather better in the welsh and northern marches though, in my view, these may have had more of a function in making it difficult for raiders to make off with cattle (which require even more safe river crossings) than as information gathering points, although there were taxes due for travelling through certain areas and the collectors of such taxes must have had bases to work from where they could safely keep the collected money.