16 The Medieval Manor House and the Moated Site

Since the 1990s, there has been growing interest amongst archaeologists in medieval manor houses. Modern scholars recognize that these buildings are central to furthering our understanding of the relationships between the medieval manor, economy, landscape, social, and cultural history of late medieval England (Bailey 2002, 2). The houses of the elite not only provided the basic functions of any shelter, but were designed by a self-conscious class with purpose and meaning: to display status and act as symbols of authority within a designed landscape. They were truly the ‘power houses’ of medieval England and a reflection of a society expressing standing, taste, and style (Emery 2007, 5; Girouard 1993; King 2003, 104). Yet the medieval manor did not just serve as the lord’s home; it was also the administrative heart of their agricultural estate and represented the jurisdiction of the lord, exercised through the manorial courts.

The earliest studies on the medieval manor house were driven by architectural historians, among them T. Hudson Turner (1851), Faulkner (1958), and Wood (1965) who were principally concerned with the fabric, features, and chronology of the architecture they observed. The outcome was an expansive body of extremely detailed building surveys which paid little attention to the social and cultural influences in their design, the way in which they were used at the time of construction and in subsequent years, and the surrounding landscape. On the whole, despite the excavation of a number of sites such as Penhallam (Cornwall) and Joydens Wood (Kent), early investigations also focused on furthering our understanding of architectural history by recording the development of the buildings (Beresford 1974; Colvin 1948; Hogg 1941; Tester and Caiger 1959). While these approaches provided a useful ‘snapshot’ of the manor house at a particular time, they failed to address critical questions about how space was created and used (Emery 2006, 3; King 2003, 104). Even today, scholars tend to focus on their own specialist fields rather than consider the topic of the manor house more holistically. The result is detailed research into the composition of the manor (Bailey 2002), the lives of the gentry (Given Wilson 1996; Goldberg 2004; Radulescu and Truelove 2005) and architectural analyses of buildings, and yet a failure to capture the wider relationship between the medieval manor and society and to understand how the buildings themselves were used to display identity, status, and wealth. Recent exceptions to this are the comprehensive studies produced by Anthony Emery, who recognized that medieval houses were not built within a ‘vacuum’, but were a consequence of external factors such as the economy, fashion, politics, and society (Emery 1996; 2000; 2006).

Generally speaking, archaeological interest in the medieval manor house has not been as extensive as it has for other major medieval monuments such as castles, churches, and cathedrals. Very few manor house or moated sites have been excavated in their entirety or near entirety. Notable exceptions include the Manor of Hextalls (Surrey), Sydenhams Moat (Warwickshire), East Haddesley (Yorkshire), Old Abbey Farm (Cheshire), and the manor houses at Wharram Percy (North Yorkshire) and Goltho (Lincolnshire), which were located within the settlement site (Beresford 1987; Beresford and Hurst 1990; Heawood et al. 2004; Le Patourel 1973; Poulton 1998; Smith 1990). A number of individual sites have also been examined as part of research excavations or developer-led investigations such as Old Abbey Farm (Cheshire), Southchurch Hall (Essex), Mount House (Oxfordshire), Acton Court (Gloucestershire), Tempsford Park (Bedfordshire), Barentin’s Manor (Oxfordshire), and the Manor of Hextalls (Surrey) (Heawood et al. 2004; Helliwell 1975; Brown 2006; Allen and Hiller 2002; Rodwell and Bell 2004; Maull and Chapman 2005; Shotliff 1996; Page et al. 2005; Poulton 1998). Nevertheless, synthesis has been lacking and this chapter will address key areas of debate within the field as well as possible directions for future research.

Despite the acknowledgement that manorial residences, and in particular moated sites, formed an essential part of the settlement pattern of England (Creighton and Barry 2012, 65), there has been almost no discussion regarding the architectural language of display in manor houses of the late medieval period. This includes how these buildings were intended to be seen and how they portrayed contemporary attitudes, social organization, and tastes, particularly as their form, function, and meaning changed over time. These statements may have been clear at the point of construction, ‘but to whose language we may have lost the code’ (Cooper 2002, 28).

Much like the manorial complex which varied in form, size, and composition, the manor house of the post-Conquest period was itself fluid in form and character. The earliest surviving domestic buildings in England date to the twelfth and thirteenth centuries, and this period is considered as to be of increasing standardization and integration in domestic planning (Blair 1993, 15; Emery 2007, 44). Prior to this, manorial sites were made up of free-standing components grouped around a kitchen, hall, and sometimes a chapel, often constructed of timber. Spearheaded by architectural historians such as Wood and Faulkner, early interpretations of surviving buildings centred upon ground-floor open halls (aisled or unaisled) and first-floor halls, which were considered to be an alternative to the open ground-floor hall of earlier and later houses (Blair 1993, 1; Faulkner 1958; Wood 1965). During the 1990s, John Blair challenged this long-held view after analysing the results of a number of excavated manorial sites, coupled with detailed analysis of the documentary evidence. Blair proposed that the terminology used by scholars to describe two-storeyed stone-built blocks was ‘anachronistic and misleading’. Instead he suggested that first-floor halls were in fact heated chamber blocks, once associated with timber-framed or stone-built ground floor halls which had themselves not survived (Blair 1993, 1–2). More than twenty two-storeyed stone-built blocks still stand from the Anglo-Norman period including Donington-Le-Heath (Leicestershire), Old Soar (Kent), Burton Agnes (Yorkshire), Hemingford Grey Manor House (Cambridgeshire), and Boothby Pagnell (Lincolnshire) (Blair 1993, 8; Emery 2000, 180; 2006, 381–2; 2007, 30; Impey and Harris 2002). Blair used Boothby Pagnell, which had long been celebrated as a typical example of a first-floor hall, as his case study. Drawing upon documentary sources and re-interpreting the plan and design of the building, he convincingly argued that buildings such as Boothby Pagnell were in fact free-standing chamber blocks (Blair 1993, 8). Blair’s argument is now widely accepted by scholars and has emphasized the need to study manor houses using every available strand of evidence and not just to focus on architectural features.

These two-storeyed chamber blocks did not stand in isolation. Rather, they were accompanied by a free-standing open hall, of which fewer examples have survived. Notable exceptions are the early thirteenth-century examples at Balsall (Warwickshire) and Barnack (Nothamptonshire) (Alcock 1982; Blair 1993, 8–9). A resistivity survey undertaken in 1995 at Boothby Pagnell to test Blair’s argument discovered a large rectangular building measuring approximately 24.5 x 17 m to the east of the surviving structure, located within the presumed position of the moat. The well-constructed stone footings and deep foundation trench suggested that they once carried a substantial stone structure, almost certainly the missing hall dating to c.1200 (Impey and Harris 2002, 252–3). Blair points out that no examples with unambiguous remains of both elements (hall and chamber) survive together on one site in England, with the possible exception of Leicester Castle (Leicestershire), but mentions comparable examples in France such as Beaumont-le-Richard and Briquebec (Alcock and Buckley 1987; Blair 1993, 9; Impey 1993). Despite the small sample size, it is now widely accepted that two-storeyed stone-built blocks were not first-floor halls, rather, they were chamber blocks that once co-existed alongside ground-floor halls.

Only a small number of ground-floor open halls survive from the twelfth century and only a few of these from a secular manorial context. All are partial or fragmentary in their survival such as Penhallam (Cornwall), Farnham Castle (Surrey), and Old Sarum (Wiltshire) with one notable exception, Oakham Castle (Rutland) (Beresford 1974; Emery 2000, 178–80; 2006, 614–15; Hill 2013). Oakham Castle (Figure 16.1) dates to the late twelfth century and is a remarkably complete example. The hall, which is defined by an exceptional stone-built aisle arcade, consists of four bays and measures 19.9 x 13.2 m. The gable ends have two prominent buttresses crowned by carved finials, and the eastern gable has two blocked doorways which would once have connected to the services (Hill 2013, 168–70; Holland Walker 1924). During a recent study, Hill argued that the hall had supported a timber lean-to structure at either end, perhaps a forerunner to the cross-wing, which likely housed the services and other lesser rooms (Hill 2013, 172). Oakham Castle stands at an important juncture in the development of the tripartite plan (see the following section) in English medieval houses. Despite the fact there was no cross passage at the east or low end of the hall, something which became a fairly standard feature in the following centuries, the presence of original blocked doorways indicates that a service end which had once been attached is a development of the later twelfth century.

Hill’s assessment of Oakham Castle was the first scholarly interpretation of the site since Wood’s examination in 1935. His analysis demonstrated that Oakham Castle is an example of an early English hall and fits into the model proposed by Blair (Blair 1993; Hill 2013, 197). The development of lavish halls and chamber blocks, such as Boothby Pagnell and Oakham Castle, demonstrate the wealth of the growing gentry class under the Plantagenet kings. Blair and Hill’s reassessment of these buildings highlights the need for archaeologists to re-examine manor houses as part of a wider assessment of how medieval buildings contributed to a symbolic language used to display identity, and this must be considered in future research projects. It is clear that one of the key factors driving developments in the style, plan, and materials used in manorial residences was an increased desire to display wealth, power, and status (Blair 1993, 13). The study of late medieval manorial sites has been little debated in the literature since the Blair’s challenge nearly thirty years ago, and as a consequence has advanced little. This reinforces the importance of revisiting sites, particularly those of the lesser gentry such as the Manor of Hextalls (Surrey) or the medieval moated manor site at Boreham Airfield (Essex) which are often only examined as a consequence of developer-led excavations (Clarke 2003; Poulton 1998).

Between the thirteenth and fifteenth centuries, building design from the grandest of houses to relatively impoverished rural cottages drew upon the same elements of a tripartite plan: the central open hall, the chamber or solar, and services (Figure 16.2). This ‘standard’ plan was flexible and easily adapted to reflect the status and personal taste of the owner while working to the physical constraints of the site. It was a means of organizing physical space while adhering to a formal arrangement which held hierarchical and symbolic meaning (Gardiner 2008, 37; Grenville 1997, 89–92; King 2003, 105; and see Chapter 15 in this Handbook). Through the size and disposition of rooms, their architecture and embellishment, a very particular ‘spatial syntax’ was communicated which expressed social distinctions, identities, power, and status. This also established a convention for managing both external appearance and internal space. A visitor moving towards a medieval manor house could ‘read’ the building during their approach using architectural features. For example, large windows such as the oriel windows at Athelhampton (Dorset) and the heraldic stained glass windows at Ockwells Manor (Berkshire) were placed at the high end of the hall (Figure 16.3) to illuminate the lord and to emphasize his status in comparison to those who were only admitted to the low end of the hall (Cooke 2010; Emery 2006, 488; Wood 1965, 113–14). Once inside the hall, a medieval peasant understood the function and layout of the space they had entered and the behaviour that was expected of them: indeed, this was probably part of the appeal behind using the same design at all social levels (Johnson 2010, 68).

Despite this standardization of the plan, however, there was no rigid blueprint for medieval builders to follow; no two houses are in fact precisely the same (Emery 2006, 2). This period was one of increasing social mobility amongst the upper levels of society and, as a consequence, established families needed to reinforce their position, while the nouveaux riches wished to demonstrate their new wealth. The study of late medieval building contracts reveals that house-builders strove to equal, if not to exceed, their neighbours and deployed new houses as deliberate visual references to their status. The building accounts of 1506 and 1509 from Little Saxham (Suffolk) mention payments to skilled workmen, the carpenter, and glazier, to enable them to travel 30 miles (48 km) to Horham Hall (Essex), to view work already completed there (Airs 1978; Emery 2000, 114–16; Howard 1987, 12–15). Presumably this was work to be imitated or bettered.

The entrance to the hall was through opposing doorways into a cross passage, commonly separated from the hall by a screen or wall. From the passage, two doorways led into the service rooms known as the buttery (for the storage of wine and ale) and to the pantry (for the storage of dried goods); if there were three doorways, the central door led to the kitchen which was commonly detached from the main house (Gardiner 2008, 40; Grenville 1997, 89). The hall was a large open space, heated by a central hearth the smoke from which escaped through a louvre. At the opposite end of the hall to the services was a high table, often raised on a dais with a moulded beam or hood to frame those sitting there (Gardiner 2008, 37). Beyond the dais was the entrance to the lord’s private chamber or ‘solar’, a term used to describe the ground-floor and first-floor rooms, the latter of which is presumed to be the ‘best chamber’. Gardiner (2000, 162) has suggested that the chamber served not only as a space for sleeping but for storing personal items and valuables. He refers to the Paston letters which illustrate that chambers in manorial houses accommodated chests containing deeds, money, and account books as well as a counting board (Davis 1971; Gardiner 2000, 162). The tripartite plan was not just a practical way of organizing space, it was also full of symbolism and meaning and clearly defined social grading. It comes as no surprise that the service area was known as the ‘low’ end and the opposite end with the dais and chamber was the ‘high’ end, demarcating social distinctions. Despite a common entrance, there was a clear hierarchy with servants operating in the low end of the hall and the lord operating at the high end, from which access to the chambers was controlled. Guests were expected to negotiate their way from a common entrance, through the low end of the hall to the high end to the chamber beyond if permitted (Grenville 1997, 89).

During the late medieval period, features which were to become fixed in the tripartite plan were consolidated and this development can be traced in surviving buildings. For example, at the late twelfth century manorial sites Oakham Castle and Penhallam (Cornwall), the cross passage did not have an opposing door, something which became standard by the second half of the thirteenth century. The early thirteenth-century Nassington Prebendal manor house (Northamptonshire) had opposing cross passage doors, though conversely only had one door leading to the services rather than the more common two or three doors (Emery 2000, 279–82; Hill 2001). One of the earliest houses to favour the full tripartite plan was Southchurch Hall (Essex), a moated timber-framed manor house which dates to the mid-fourteenth century (Brown 2006, 3; King 2003, 108). The open hall here is a two-and-a-half bay timber-framed building with a chamber at one end and a cross passage at the other which contained a pair of central doors that led to the buttery and pantry (Emery 2000, 9; Rackham 1986, 43).

By the end of the late medieval period architecture had reached levels of excess not previously seen before. Political and social status both demanded visible public expression whether it was through great houses, clothing, or the number of liveried retainers in service (Cooper 2002, 293). Maddern (2005, 31) suggested that, in order to be part of this society, the gentry were ‘acting out a role’. This could mean showing the appropriate hospitality or building a house that displayed the right kind of message. From the fifteenth century, society witnessed some of the most important architectural and structural developments of the manor house, with a growing emphasis on domestic comfort. There was a gradual abandonment of the typical medieval plan, which was viewed by now as old-fashioned and restrictive, and a corresponding expansion in the space devoted to private living accommodation (Cooper 1999, 16, 55; Thompson 1987, 44–5). Houses became larger. No-one of gentry-standing would be expected to sleep in a hall, and the provision of new rooms coupled with architectural devices such as contrived symmetry, the creation of courtyards, and the addition of large expanses of windows all helped to demonstrate status. At the same time, while the hall remained central to spatial organization, its form and function were updated. The aisled hall was no longer fashionable, with the preference for a large uninterrupted space. Other changes included the abandonment of the central hearth and the introduction of large bay windows, wall fireplaces, and screens (Emery 2007, 47; Wood 1965, 56–62).

A post-processual framework enables archaeologists to explore the idea that the gentry invested in their houses and the surrounding landscape in order to display their identity, social prestige, and wealth (Cooper 1999; Creighton 2009). It is precisely this interaction with the physical environment and material culture which enabled the medieval elite to present their position in society by imitating patterns of display exhibited by those at the top of the social scale. It is clear that the gentry shared similar ideologies with the nobility, with both groups striving to display those qualities that made them ‘gentle’, such as refinement, good manners, and honourable behaviour. During the late medieval period therefore space acquired meaning and identity, and this was emulated and displayed through material culture and buildings. The design of buildings, the use of architectural features, the choice of materials, and the creation of a designed setting which surrounded the house all contributed to a symbolic language which was easily read during the late medieval period. These ideas are accepted for the post-medieval period, yet have been rarely discussed for houses and landscapes of the preceding centuries. There are essential similarities which must be considered as an area of future research within the field.

So what does the future research in to the study of the late medieval manor house hold? The construction or modification of existing manorial sites, and the use of material culture were used deliberately to develop a cultural identity and reinforce status, and this is something that future investigations must consider. Perhaps the way to move forward is for archaeologists to consider buildings as artefacts. This will enable scholars to investigate their design, style, and use during the period in which they were built as well as to understand how people responded to buildings over time, something which many previous approaches have failed to consider. Identity and status has always involved the consumption and display of material culture, which also shines a light on how people of the late medieval period viewed themselves (Grassby 2005, 596). The examination of artefacts and environmental evidence must also be included in all future research into the medieval manor house and moated sites. Moated sites have the potential to produce a number of high quality artefacts and to provide an important insight into day-to-day life, as well as recording the changes to the site over the centuries. At Wood Hall (Yorkshire) excavations produced high-quality leather items such as shoes and finely decorated gloves, as well as wooden items such as lathe-turned bowls and items of glass such as an enamelled and gilded glass goblet, possibly originating from Bohemia (Metcalf 2001, 28; 1993, 19). At Acton Court (Gloucestershire) important amounts of pottery and building materials were recovered as well as a large quantity of vessel glass including many imported examples, tobacco pipes, textiles, leather, including six shoes, and two wooden staves from longbows (Rodwell and Bell 2004, 294–414). Timbers also survive well in waterlogged sites. At Southchurch Hall (Essex) several pieces of fourteenth-century boat planking were discovered below the trestle of a timber bridge (Hutchinson 2006, 136). Remains of bridges also survive at sites including Penhallam (Cornwall), Acton Burnell (Shropshire), Lewmote (Kent), Southchurch Hall (Essex), and Hen Gwrt (Monmouthshire), among others (Rigold 1975; Thornhill 1973; Wilson and Hurst 1964, 272–4).

In the last twenty years, the number of archaeobotanical samples analysed has slowed. This is perhaps due to budgetary constraints which can slow the post-excavation process and the rise in developer-led investigations which are focused on being cost-effective (see Chapter 8). In many instances there is little attempt to compare results with the national archaeobotanical dataset or to address questions that are at the core of research-led excavations. Recent papers such as Rippon et al. (2014) indicate the advantages of using zooarchaeological and archaeobotanical evidence on a much broader scale to investigate regional variation in landscape character, particularly in undocumented periods. Unfortunately, much of the valuable information gathered from archaeobotanical reports remains as ‘grey literature’, and may not feature fully in published excavation reports (Van der Veen et al. 2013, 169–70).

Botanical data from excavations can provide an invaluable insight into the diets, agriculture, trade practices, and environment of the period; however, reliable reconstructions are dependent on the quality of datasets and their recovery from a secure context (Van der Veen et al. 2013, 151). The sediment from moat ditches has often remained wet, making it appear ideal for radiocarbon dating and environmental study. Yet the fact that flora and fauna assemblages are preserved in a moat does not mean that they can necessarily produce the desired picture of an environmental record of the period. This is because the ditches of medieval moats were regularly cleaned and drained not only at the time they were in use, but also during subsequent centuries. This was demonstrated at Perceton (Ayrshire), for example, where the preservation of deposits and features was poor due to horticultural mixing and bioturbation and this proved a challenge for the construction of reliable stratigraphic and chronological sequences (Stronach 2004, 148).

Analysis of the 11,105 animal bones and 2265 shells of marine molluscs recovered at Barentins Manor (Oxfordshire) produced significant insights into the economy and change of husbandry at the manor. Management of cattle for meat was initially the principal use of animals on the site; however, there was a change to an arable economy as dairying become more important. An abundance of pig and deer bones suggested the exploitation of local woodland and scrub, unusual in a Thames Valley context. The number of sea fish and shellfish, including edible crab, is also a clear indication of trade (Wilson et al. 2005). On the other hand, the preservation of organic remains from the moat was poor, and they were only recovered from one sample from the very bottom of the moat. This may have resulted from the drainage of the moat after the site was abandoned (Robinson 2005, 153).

Our understanding of space must also be carefully considered in future research. In recent years scholars have began to consider the process of seeing, particularly how past spaces were used to portray identity, power, and wealth. Giles (2007, 106) has argued, however, that current approaches to the relationship between landscapes, buildings, and material culture impose a modern way of thinking and seeing. She believes that archaeologists must address the debates occurring in other disciplines regarding the historicity of visuality and spatiality, in order to understand how communities retained a connection with the past (Giles 2007, 117–18; Chapter 14 in this Handbook). Sensory perception and connotations of the late medieval household must also be critically considered by archaeological investigations in the future, in order that we can begin to understand experiences of the day-to-day life such as eating and drinking (Brears 2008; Woolgar 2006, 273; see Chapter 43 in this Handbook). Material culture such as furniture and furnishings and the reorganization of space were delicate tools that individuals deliberately and interactively used to develop their own cultural identity and standing (De Clercq et al. 2007, 1). These complex ideas must be an essential part of all future research and reiterate the point that studies of the late medieval manor house must adopt a comprehensive approach and incorporate different disciplines and strands of evidence.

Another key theme emerging in the study of medieval archaeology is the role of the landscape. For example, a recent investigation of forty-five pre-thirteenth century nucleated settlements has challenged the idea that manor houses and parish churches lay at their focal points (Dyer 2003, 11; Jones and Page 2006, 79). Some 47 per cent of manor houses were found to be over 250 m from the nearest settlement and 91 per cent were over 100 m away (Campbell 2013, 274). Generalizations should not be made based on such a small sample size, but this suggests that where the manor house was positioned within the landscape and the resources which surrounded it were both an important part of display. Manors were not isolated country houses; rather, they were agrarian and administrative centres of the lord’s estates and surrounded by buildings which were essential in running a farmstead such as barns, stables, dovecotes, bakehouses, and brewhouses (Gardiner 2007, 170–3). Farmyards were an important part of a medieval manorial house and were displayed, not hidden away, as at many post-medieval gentry house sites.

It is clear that those constructing manor houses found landscape features such as fishponds, deer parks, and moats appealing and used them as part of a ‘designed medieval landscape’ (Creighton 2009; Liddiard and Williamson 2008, 533; see also Chapters 23 and 25 in this Handbook). Dominance over the landscape and resources showed a considerable degree of social control and authority and, despite challenges in dating landscape features and the requirement for more rigorous testing, perhaps through GIS analysis, further investigation is required. Moats in particular have largely been ignored in the literature in recent years despite the fact that over five thousand sites were recorded in England by members of the Moated Sites Research Group during the late 1970s and 1980s. This is perhaps due to the fact that the group approached these monuments typologically, categorizing moats by their size and shape. These observations failed to capture the symbolism of these monuments and did not adequately question why so many of them were constructed, and why their construction largely ceased after c.1350.

Moated sites in England are spread across the country with the majority constructed in the more fertile region of central England (Aberg 1978, 2). Conversely, research into moated sites in Scotland and Wales is much more limited in scope and size. To date, approximately 120 moated sites have been identified in Scotland and only a handful have been subject to archaeological investigation. Early studies such as Dunrod and Bombie (Kircudbrightshire) did not advance the subject as excavations primarily focused on a small area of the moat. More recent interventions, as at Wallaces’s House, Elderslie (Renfrewshire), and Perceton (Ayrshire), examined the site in its entirety while also considering the role of moats as a status symbol in Scottish medieval society (Alexander 2000; Anderson 1946; Burdon-Davies 1966; Stronach 2004, 147, 159–60). A similar number of sites, approximately 130, have been recorded in Wales, primarily concentrated in the low-lying areas and Marches. Like England, this area contained some of the most fertile land and moats were likely used as part of the display of wealth and political and cultural status, perhaps related to the Edwardian conquest in the mid-1280s (RCHAM 1982, 69–71). Much like Scotland, there has been very little scholarly research, perhaps because very few moats actually contain standing remains of structures. Notable exceptions include Horseman’s Green and Althrey Hall (both Flintshire) as well as Llay Hall and Lower Berse (both Denbighshire) which were associated with high-status aisle-trussed halls (Cole 1991; Emery 2000, 667; Hubbard 1986, 110, 248–9; Smith 1988, 96–126, 321). Medieval structural remains have been found at Hen Gwrt (Monmouthshire), Horseland and Highlight (Glamorgan) with post-medieval buildings replacing earlier remains at Llay Hall, New Hall (both Denbighshire), and Halghton Hall (Flintshire) (Craster and Lewis 1963; Hurst 1988, 940; RCHAM 1982; Spurgeon and Thomas, 1978, 30). Overall, sites in Scotland and Wales are difficult to date, but it is likely that the majority were built during the long thirteenth century, as in England. Excavations at Llys Edwin (Flintshire) and Highlight suggest a construction date of the late twelfth or early thirteenth century, while excavations of the moat at Hen Gwrt suggest the primary occupation was from the first half of the thirteenth century, although construction of the moat did not take place for another century (Craster and Lewis 1963). Moated sites in Wales are often associated with the place-names ‘Llys’ or ‘Plas’ (mansion, hall, or palace), such as Llysworney and Gadlys (Glamorgan) and Llys Edwin (Flintshire), the latter of which is one of three possibly fortified sites.

Moats were used in the medieval period for a variety of functions, but most were likely to have enclosed an island upon which a house was built, the location of which varied. It has been suggested that by the time the hall and chamber had become integrated the house was usually located towards the centre of the enclosure (Rigold 1978, 33). Moreover, the house was often orientated so that the length of the building ran across the longer axis of the island, creating a forecourt while also leaving room for a back yard, as is seen at Lower Brockhampton or Southchurch Hall (Brown 2006; Emery 2000, 558; Rigold 1978, 33). Later houses may be planned so that three or four sides of the building are built on the edge of the moat, with each range appearing to rise up out of the water, as at Acton Court (Gloucestershire). This reflective quality magnified the structure and created the illusion that the building was floating as at Acton Court (Gloucestershire) and more famously Bodiam Castle (Sussex).

Moats were not just a feature of high-status sites, but were also found on manorial sites of the lesser gentry and indeed non-manorial sites (Platt 2010, 115). The traditional interpretation of the role of moats is that of defence or drainage but, as Creighton has argued, this can only take us so far (Creighton 2009, 88–115); moats vary in shape, size, and depth and are found in variety of topographical locations. Moats were used to contain garden compartments and orchards and house livestock and dovecotes so it is likely that some moats also had an ornamental role. The moated manor house of Lower Brockhampton (Herefordshire) has a second moat filled by a small drain which carried the overflow from the principal moat (British Library Add. MS 36415; Morriss and Hoverd 1994; Poulton 1998). The subsidiary island itself is too small to have supported a domestic building, but may have had room for an orchard, garden, or possibly a small banqueting house as at Kenilworth (Warwickshire) (Creighton 2009, 79; Henderson 1992, 117; Johnson 2002, 137; Woodhouse 1999, 10). The moat may also have been used for breeding fish, for fishing, and as a place for boating. A secondary moat attached to the principal moat was also recorded at Barentin’s Manor (Oxfordshire). Excavations on the second island produced no evidence for any structures and it too may have served as an enclosure for animals or as a garden or orchard (Page et al. 2005, 16–17).

Water has a unique ability to transform the human experience, particularly through its reflective qualities. Creighton (2009, 77–90) has suggested that we must look beyond the functionalist approach to watery landscapes traditionally undertaken in medieval landscape studies and consider the meaning. Water is symbolic and was used to seclude or create access to buildings. It was also used to magnify and have a stunning visual impact. A watery backdrop impacted the way in which the building and landscape was experienced. Future research must seek to examine the symbolism of these monuments and place moats into their European context. It must re-assess how landscape features such as moats, dovecotes, fishponds, and deer parks were used to display lordly dominance as part of medieval designed landscapes, not just in lowland England but across the British Isles (see also Chapters 23 and 25).

In conclusion, houses and landscapes were a conspicuous statement of elite identity in the late medieval period. This is not something that was exclusive to Britain; direct parallels can be drawn with manorial sites in France, Belgium, Scandinavia, and elsewhere. Future research must look now beyond national borders and explore commonalities and cultural influences in sites across Europe.