What is a castle? The author's answer to this obligatory question is, in his own words, ‘provocative … challenging received opinions and hopefully, formulating many new ones’. The 500 or so pages that follow certainly do not disappoint. The direction of travel becomes clearer with the advocacy of his definition: ‘a castle is the residence of a lord made imposing through the architectural trappings of fortification’. This book is unashamedly a study in architectural history, which charts the development of the castle through the Middle Ages and significantly beyond, into the mid-seventeenth century.

Some readers may be troubled by Goodall's definition, if they associate castle building with the integration of location, planning and physical strength as well as appearance. But this is not a new approach, although it is a step back from Platt's view (1892) of the castle as ‘an expression of personal aspirations; it was a visible demonstration and perhaps the most visible – of wealth and a man's status in society. Of course, the castle had its purpose in defence.’ Even so, many may find greater affinity with Allen Brown's more holistic and perhaps conventional definition (1976): ‘The castle, then, is the fortified residence of a lord, and as such differs from other known forms of fortification in the history of the West both earlier and later, by its duality as both a residence and a fortress and by its private as opposed to its public or communal nature.’ The key concept here, of course, is duality.

There is consensus in one respect, namely that there are structures often described as castles – among them Henry viii's coastal defences, for example – but which are forts, as their accommodation is for a garrison rather than a lordly household. However, Goodall's definition is purposeful, as it enables him to include buildings of the later sixteenth and early seventeenth centuries which many would regard as great lordly residences with little or no effective defensive capability, but having castle-derived architectural embellishment. The early sixteenth century is the period where opinions about castles diverge, of course; it saw the concurrent development of reliable artillery and the adoption of extensive traceried fenestration, sometimes in locations such as the turreted, machicolated and battlemented outer facades of gatehouses. Any defence of these buildings was at best compromised; duality was lost. A result, as Goodall points out, is that ‘even outdated fortifications that might technically have been obsolete in practical terms rarely were so … on occasion in the Civil War the sheer scale of some medieval fortifications, such as the great tower of Raglan, Monmouthshire, proved impervious to weapons of a power unimaginable when the building was first constructed nearly 200 years previously.’ The reader must judge whether or not Brown's concept of duality is important in defining a castle, and with it the matter of form in relation to function. Goodall's clearly stated intention is to be provocative, but if his arguments are compelling, are they also convincing?

Goodall's wider objective, ‘an architectural study that aims to set this legion of historical buildings in historical context’, is handsomely achieved. Following a lucid introductory chapter, the text is rigorously arranged as a chronological narrative. It effectively contextualizes architectural developments by including a great deal of political historical detail and lengthy digressions into aspects of ecclesiastical architecture and heraldry. Closely and fluently argued, the work is impressive for its continuity and indeed reads almost like an extraordinarily long thesis. It brings into consideration the substantial volume of material published during the past twenty or so years, material that has redirected so much of our thinking about the form of castle buildings. The discussion of the origins and transference of ideas and designs is stimulating and impressive. And there are interesting outcomes, such as the reappraisal of Arnold Taylor's seminal work and conclusions on Edwardian castle design.

The reader is constantly reminded that castles should be seen as a vehicle to express and consolidate the political and social aspirations of the lordly class. We have a chronicle of the relationships of this class with the Crown and among themselves, and with those who conceived and built their castles. It is commendable that the author's depth of research and understanding has enabled him to populate the narrative with so many real people. This is less a work about the detail of changes to defensive and domestic arrangements over time, but rather more it is concerned with the architectural forms and embellishments (the ‘trappings’) of the principal buildings as an outward expression of those aspirations. Among those buildings, the focus over time sharpens increasingly on gate towers, turrets, battlements and, to a lesser extent, great towers.

So, Goodall provides us with what he says he will, and in an exemplary manner. The large format is used to the full, and the text is carefully balanced by photographs and drawings of superb quality, both in choice and reproduction, so that his line of argument can be fully understood. The use of early photographic images, shot before twentieth-century interventions to castle structures, is particularly noteworthy, and it is a delight to see the late Terry Ball's reconstructions, among the work of others, used to such good effect. At the same time this is a large and heavy volume which is not physically easy to handle, and it is a pity that the bibliographical notes, essential in a work of this kind, are grouped at the end of the book rather than chapter by chapter.

Finally, readers should be aware of what the work does not include. It has no gazetteer, no glossary, geospatial referencing based on pre-1970 counties, and no tabulated data. It is best approached as a book to be read in whole or in part, and has limitations as a work of reference. That said, it is a very stimulating and provocative work, and well worth the time and effort it takes to follow its line of argument.