Despite years of castle studies, the strategies, tactics and stratagems that underpinned Norman castle warfare remain largely unexplored. This book outlines a strategic approach to castles, arguing that 'it was not what was built, but where it was built' that was the key to Norman success. Castles are placed into context via landscape archaeology, and a study of contemporary and later sources. Castle positioning in the landscapes of Somerset, Monmouthshire and County Meath are examined in detail. Above all, the author argues, it is necessary to see the Normans as part of a wider European phenomenon that drew its inspiration from the Roman Empire. (Publisher)

Whilst Prior's use of nineteenth century military textbooks to suggest a military location for Norman castles is interesting it is flawed. He fails to demonstrate a difference between castles founded to control administrative 'nodal' points and castle founded as administrative centres. He argues from an a prior position rather than doing a critical analysis of the evidence he collects. The 'metadata' in his appendix is, in my view, statistically flawed and shows a lack of understanding of metadata analysis, although my view on such metadata analysis, influenced by the cautious use of such data in the hard sciences, is not shared by all archaeologists/historians. For those emotionally wedded to a military view of castles the book provides succor. For those interested in a deeper, more complete, understanding of the Norman castle the information provided is useful, the analysis and conclusions are not.
I found much of interest in this book and even more so in the PhD on which it is based. The comments on the management of horse were most useful and the archaeological site descriptions are generally fine. However there is much to contest in the book and it may be a perceived need not to produce a 'boring publication' has somewhat influenced the perception of the Normans, their military strategy and their castle building.
Of the deep flaws an example is the suggestion that the castles of Somerset were sited to protect the boundaries of the county. Actually, given that the centre of the county was inaccessible marsh in the C11/C12, they were clearly cited where the people were and most were probably built to administer a civil population in a post-Conquest phase. The suggestion that castles were sited in association with rivers and roads is an old idea and not unworthy but by including castles within 4-5 kilometers from a river or Roman road the author has defined an area which would include virtually the whole of inhabitable medieval Somerset making the point meaningless. Other flaws that have been suggested to me include the idea of a co-ordinated three prong assault on the county which are just untenable given the communication possible in the C11. (Philip Davis 2009-2013)