The castle had a dominant presence in medieval society, both physically and ideologically. Controlled by elites, castles towered over medieval villages and towns and were sites of judgement and administrative control. However, castles were also depicted over and over again in the medieval arts as heraldic devices, as pastry or paper table decorations, on seals (see Plates 1, 2 and 3) and as large-scale props in pageants. They featured figuratively in sermons, theological treatises and religious lyrics and in manuscript marginalia (see Plate X), as well as in the more familiar contexts of romance and chronicle. To the modern understanding, there is a wide gap between these ephemeral, miniature and symbolic castles and the imposing stone-and-mortar fortresses scattered over medieval Europe and beyond. This book sets out to show that medieval thinking on these matters was very different.

Numerous studies have demonstrated that medieval ecclesiastical architecture was fully integrated into the intellectual and aesthetic culture of its time. Deliberate resonances were created between the soaring spires of the Gothic cathedral and the miniature pinnacles represented on tombs and reliquaries and in paintings and sacred texts inside the church. Indeed, it is recognized that microarchitecture (as the diminutive, decorative form is known) was an important adjunct to ecclesiastical architecture, expressing its symbolic properties and associations in a particularly concise way. It is precisely this kind of relationship I am seeking to explore in the medieval castle and its artistic and literary representations. I set out to establish the medieval castle as a meaningful architecture, involved in a sophisticated series of ideological relationships with its cultural context. This necessarily demands a rethinking of traditional approaches to the castle in modern scholarship.

The castles of the medieval landscape are, by definition, defensive architectural forms. They are usually considered to have been built with military functions in mind. It is from this point of view that they have most often been approached in modern scholarship. Because defence is such a practical consideration, rooted in engineering, technology and military strategy, it has not often occurred to scholars to treat medieval castle architecture as ideological. By contrast, medieval ecclesiastical architecture has long been understood as a meaningful architecture, which operates at an ideological as well as a practical level. It is appreciated as the highest technological achievement of its period, but also as a vessel for the most important religious ideas and beliefs.

In the previous chapter I showed that the medieval understanding of castle words allowed for a wide degree of overlap between private fortifications and fortified communal and urban enclosures. I used linguistic arguments to explore this link mainly at the level of verbal usage and understanding. However, in this chapter I concentrate on the ways in which the relationship between castle and town was explored symbolically, in medieval literature and art, and in the spatial and political juxtaposition of urban castles and town defences.

The quotation cited above, from Caxton’s preface to his Eneydos, serves as introduction to a number of key ideas about the relationship between the medieval castle and town. Phrases such as Wyclif’s ‘litil tounes . . . wallid’ invoke the affinities between castles and towns. However, Caxton gives a broader and more nuanced appreciation of this relationship. The subjectmatter of the text – the siege of Troy – is given an exemplary function ‘to enstructe’, directed towards dwellers in both ‘townes and castellis’. Both communal and private defences are thus united in their joint duty ‘to kepe and defend’; but this collective responsibility is defined by social divisions. Two groups, respectively of the city and of the castle, are contrasted socially as ‘citizens and habytaunts’ and as ‘smale and grete’. Their common purpose is expressed through carefully differentiated hierarchies of person.

Yet Troy is ultimately an example of failure as well as of heroic joint endeavour. It was sacked by the Greeks through the treachery of one of its own citizens, in a war brought about by the selfish lust of a member of its aristocracy. The knowledge of this ultimate failure lies behind the exhortations of Caxton’s preface, adding poignancy to the exemplary united efforts in its defence. Troy might, then, be interpreted as a negative exemplar: a proof of the ultimate incompatibility of the interests of commons and elite and of the futility of struggling for the common good.

Castles and churches are without doubt the most impressive architectural achievements of the Middle Ages, yet they have traditionally been studied from different points of view and by different scholars. This approach has inevitably emphasized the contrasts between them, especially in terms of their ideological connotations. The whole of this book is an attempt to show that defensive architecture could communicate meaning in the same ways as ecclesiastical architecture.

A close correlation exists in British castle studies between theories about the origins of the castle in England and the question of the proper meaning of the term castle. No thorough linguistic study has yet been made of the meaning and development of the word, despite its great significance for the understanding of the medieval castle. This chapter cannot provide an exhaustive survey, but it sets out a summary of the word’s origins and development in English usage in order to clarify this point.

The passage quoted above, from the entry for 1052 in British Library MS Cotton Tiberius B.iv, the Worcester manuscript of the Anglo-Saxon chronicle, has been central to this debate, as it is identified as one of the first occurrences of the word castle in English. However, none of the discussions of this text by castle scholars has looked to recognized linguistic methodologies, or set out to investigate work by historical linguists on the early meaning of the word castle. This has led to several inconsistencies and circularities in a debate that is crucial to the modern and medieval idea of the castle.

In June 1283 work began on Caernarfon Castle, part of Edward I’s massive castle-building campaign designed to consolidate the English position in Wales by fortifying newly acquired territory. The castle at Caernarfon alone cost over £20,000, a huge amount of money in contemporary terms, and took nearly fifty years to complete. It was built at the mouth of the River Seiont, site of the ancient Welsh centre of Gwynedd; its thirteen polygonal towers and its exterior wall surface were given decorative treatment through coloured banding in the stonework, achieved by the alternation of dark and light stone courses (see Plates XIII and XIV). The castle was built on and around an older work, probably of Norman origin; also in the year of the new castle’s foundation a body was found on the site and re-buried in the nearby church.

In one of the most celebrated pieces of research in castle studies, Arnold Taylor transformed these facts into legends. Through medieval Welsh chronicles he found that the Roman site of Segontium, on the hill above Caernarfon, was connected in legends to Constantine the Great. Taylor was also aware that Nennius, the author of the ninth-century work known as Historia Brittonum, had referred to a tomb at Segontium inscribed with Constantine’s name, providing further evidence of imperial connections. Taylor concluded that the incorporation of the Norman motte into the new castle was a material expression of continuity with the area’s past, acknowledging the powerful symbolism of the ruins. He also found documentary evidence to show that Edward I believed the body which had been discovered was that of the Roman emperor Magnus Maximus, the father of Constantine the Great. For Taylor this, again, was evidence of the strong symbolic connections invoked by Edward to strengthen his relationship with the illustrious ancient rulers of the place. These resonances were confirmed for Taylor by the medieval Welsh text Breuddwyd Maxen (The Dream of Maxen), part of the cycle of the mabinogi, or Mabinogion, as the collection of works is usually called. Magnus Maximus appears in this text as the emperor Maxen (or Macsen), who is associated with the beautiful castle of Aber Sein, situated at Arfon in Wales. According to this legend the emperor marries Elen, the daughter of this castle’s lord, and by this act the fortress at Arfon becomes the chief stronghold of the Island of Britain.

The aim of this book has been to establish the medieval castle as a meaningful architecture, involved in a sophisticated series of ideological relationships with its cultural context. I have set out to trace the architectural iconography of the castle through references in visual and textual sources, and to retrace this iconography back to the physical architecture of the medieval buildings themselves. The conclusions presented here summarize those reached in each chapter of this book, but also point to wider implications and further possibilities for research.

Linguistic analysis shows that many modern definitions of the medieval castle do not match the understanding of the word and concept in medieval use. The castle was not perceived in medieval England as an essentially feudal, private form of architecture imported by the Normans, as historians so often see it now. The word castle seems to have indicated a defensive enclosure of a much more general kind, applicable to urban fortifications, small houses and ecclesiastical foundations as well as the private defences with which the word is exclusively associated in modern use. The castle also had important historical connotations for medieval readers, via the Latin word castellum which appears in Biblical and Classical writings. The principles of historical linguistics suggest that a medieval reader would have understood this word in accordance with the general medieval meaning of castle words. It follows that castles of the medieval type were believed to have existed in important ancient contexts long before the introduction of the form to Britain around the Norman Conquest.

This linguistic reappraisal of the medieval castle has far-reaching implications which can be traced in depictions of castles in medieval art and literature and in the architecture of medieval castle buildings themselves. Through civic imagery on seals and in city descriptions and foundation legends, and in the spatial arrangement of city and castle defences, the urban castle became an important component in the imagery of the city. It was appropriated to represent both the harmony of the ideal city and the social and political tensions of everyday urban experience. These qualities could be expressed in visual and textual depictions, but they could also be built into the fabric of cities in the harmony of their spatial layout, or taken away from cities when these boundaries and hierarchies were disrupted by conflict.