Practice, Power and Place: Southern British Perspectives on the Agency of Early Medieval Rulers’ Residences

This paper advances understanding of rulership over the fifth to the ninth centuries AD, drawing upon a category of elite settlement from southern Britain known as the great hall complex. Guided by a practice-based conceptual framework, we connect these sites with the embodied regimens, rituals, habits, and activities through which rulership was constituted in the early medieval world. Harnessing recent expanded datasets, we generate insights in three key areas. First, by documenting the significant and sustained antecedent occupation attested at great hall sites, we reveal how rulers exploited the complex multiple pasts of these places to advance symbolic and worldly agendas. Second, we reframe understanding of hall construction as a strategy of elite legitimation by focusing attention on the agency of the skilled practitioners who created these innovative architectural statements and, in doing so, recognize these hitherto neglected specialists as ‘crafters’ of rulership. Third, we use proxies from recently investigated great hall complexes to reconstruct the networks of dependency and interaction which enmeshed these centres. A concluding comparative discussion of southern Britain and Scandinavia contributes shared perspectives on rulers’ residences as a prime arena for the orchestration and creative renewal of early medieval sovereignty.


RULERSHIP IN THE EARLY MEDIEVAL NORTHERN WORLD: DEFINITIONS AND APPROACHES
Understanding of early medieval rulership has advanced rapidly in recent years under the influence of transnational interdisciplinary research on the dialectics between power and place (e.g. De Jong et al. 2001, Davies et al. 2006, Sánchez-Pardo and Shapland 2015, Rollason 2016, Caroll et al. 2019a. While conceptually and analytically varied, this work has nurtured cross-cultural perspectives on the nature, practice and materialization of rulership that transcend national scholarly traditions and the specificities of institutional labels derived from historical sources.
Our approach draws particular inspiration from recent examinations of 'rulership ideology', which have explored the behaviours, practices and social relations of rulership within the early medieval Nordic arena (e.g. Hedeager 2002, Sundqvist 2002, Steinsland et al. 2011a). This work informs our agenda in two ways. First, it espouses a flexible definition of rulership that is applicable to a diversity of cultural contexts while also acknowledging the role of female actors and broader social collectives and agencies in the performance of power (Sundqvist 2002(Sundqvist , 2012. Second, while conceptually rooted at the intersection between religion and ideology, it provides an analytical framework that is broad-based and practice-orientated, focussing on the ways in which the symbolic, mythic and material worlds were creatively manipulated through, and in response to, different strategies of rulership. This analytical focus is well suited to investigating the systems of rulership that defined much of the early medieval northern world, in which heavy emphasis was placed on the personal charisma of leaders and the ritualized theatre of assemblies and other staged gatherings of the militarized aristocracy, or warband, as the building blocks of legitimacy and authority (Gautier 2009, Pössel 2009, Steinsland et al. 2011b, Sundqvist 2012, Roach 2013, pp. 14-20, Nordberg et al. 2019. This literature underlines the importance of the aristocratic hall, and the landscapes of elite residence of which they formed part, in the promulgation and ritual orchestration of rulership ideology. The wealth of archaeological data from 'central place complexes' in early medieval Scandinavia has provided a rich contextual basis for exploring the embodied and emplaced practices of rulership. In particular, the recognition of extensive ceremonial landscapes with a range of ritual traces at places such as Lejre, Tissø, Uppåkra, and Old Uppsala has greatly enhanced understanding of the varied fields of performance and practice that defined politico-cosmological centres (Herschend 1998, Hedeager 2001, Sundqvist 2012, Fabech and Näsman 2013, Rood 2017, Jörpeland et al. 2018. The emphasis of these studies is rather different from the substantial body of work which has examined such sites from a more overtly economic, territorial and developmental stance in relation to 'central place' paradigms. (Callmer 2001, Ludowici et al. 2010, Stidsing et al. 2014, Skre 2020). However, we see these two broad approaches as necessarily complementary in drawing out the multi-faceted and deeply entangled complexity of ruler's residences as the focal places of early medieval worlds.
This paper originates in an academic network initiated to harness the collective momentum of recent and current archaeological projects investigating early medieval rulers' residences in Britain. 1 The network's principal aim was to develop comparative perspectives on British sites and situate them within the wider North Sea zone. The substantial and ever-growing literature generated by the archaeological investigation and interdisciplinary contextualization of Scandinavian central place complexes has had a strong influence on the interpretation of sites of rulership around the North Sea basin, particularly in mainland Britain where there are also highly elaborate expressions of aristocratic hall-culture. Our intention here is to shine a light back on the Scandinavian scene by drawing upon expanded archaeological datasets and fresh interpretative perspectives from southern Britain in a form which mutually enriches both contexts while also linking with broader crosscultural research agendas. 2 We start by outlining the potential of theories of practice and performance as a conceptual framework for advancing the social interpretation of early medieval rulers' residences. We show how this framework can both bridge divisions of perspective in previous scholarship and be used to make new connections between these sites and the orchestration of rulership, with a particular focus on collective agency and the relationship between power and innovation. We next turn attention to southern Britain, commencing with a brief characterization of great hall complexes, followed by a comparative interrogation which builds interpretation around the themes of temporality, monumentality and centrality. We conclude by comparing southern Britain and Scandinavia to contribute shared perspectives on rulers' residences as a prime context for the orchestration and creative renewal of sovereignty in the early medieval North Sea arena.

TOWARDS A PRACTICE-BASED FRAMEWORK FOR THE STUDY OF EARLY MEDIEVAL RULERS' RESIDENCES
This paper seeks to look beyond the generalized institutional structures of early medieval rulership to gain insight into the agentive intricacies of its practices: 'the embodied regimens, rituals, habits, and activities that reproduce … sovereignty in interactions from the spectacular to the everyday' (Smith 2011, p. 419) Previous comparative archaeological and interdisciplinary studies have generated valuable perspectives on this theme in relation to places of assembly (e.g. Barnwell and Mostert 2003, Semple and Sanmark 2013, Caroll et al. 2019b, funerary landscapes (Semple and Williams 2007) and elite residences (Gleeson 2012, Loveluck 2013, Semple 2013, pp. 207-212, Sundqvist 2016, but there has been limited engagement with, or explicit consideration of, practice theory as a conceptual tool for enhancing understanding of the material and symbolic strategies of early medieval rulership. 3 With a well-developed literature spanning sociology (Bourdieu 1977, Giddens 1979, 1984, Reckwitz 2002, Shove et al. 2012 anthropology (Geertz 1977, Bell 1997 and archaeology (Dobres 2000, pp. 136-44, Inomata and Coben 2006a, Berggren and Stutz 2010, Robb 2010, Fewster 2013, theories of practice and performance have a strong mutual concern with power and politics. Archaeologies of performance emphasize the importance of actions, rituals and spectacles in the construction and negotiation of new identities and power relations, encouraging close attention to the specifics of how time, space and movement interact to shape the experience of actors and audiences (Inomata and Coben 2006b, p. 19). This framework offers fresh analytical and interpretative opportunities for bridging the detailed archaeology of excavated rulers' residences with the ritualized drama of aristocratic hall culture, complementing interdisciplinary studies that have worked at the interface between material and textual sources (Herschend 1998, Bazlemans 1999, Bintley 2020, pp. 83-92, Price and Mortimer 2014, Sundqvist 2016.
This approach also creates conceptual space for thinking about rulers' residences as arenas of creativity and innovation. Classic sociological formulations lay stress on power as the intrinsic ability to transform, with an attendant focus on the ways in which rulers use resources creatively to maintain the status quo (Giddens 1979, pp. 88-94, 1984, pp. 14-16, Roscoe 1993. As a defining locus for the practical regimes of authority, rulers' residences played a central role in positioning 'elites between novel [our emphasis] techniques of power and embedded traditions of social order' (Smith 2011, p. 419). The link between ruler's residences and processes of change and transformation can be seen, for example, in their role as focal points for the Christianization of early medieval polities, a process that was intimately bound up with, and ultimately flowed from, the sacral realignment of rulership itself (Cheney 1970, Oakley 2006, pp. 89-107, Beuermann et al. 2011, Sundqvist 2002, Sundqvist 2016, Andrén 2013, Thomas et al. 2017b. But this conceptual framework has much broader relevance and applicability to the sites in question. As developed in relation to archaeologies of technology and making, practice theory encourages a focus on the skills, knowledge and resources -the 'communities of practice' -that converged at rulers' residences and how this created the conditions for creativity and innovation (Dobres 2000, p. 1, Shove et al. 2012, pp. 63-68, Ingold 2013. This perspective has clear relevance to the evidence for skilled crafting frequently found within or adjacent to elite residential compounds of the period (Hjärther-Holdar et al. 2002, Hedeager 2011, pp. 137-148, Axboe 2012, Wright 2019. But, as we demonstrate below in relation to southern Britain, it is equally germane to the ostentatious timber architecture defining these sites and, more broadly, the innovative systems of resource extraction and consumption channelled through them. Ultimately, this approach offers an enhanced framework for capturing the emic qualities of these sites -the specifics of how power was performed and experienced in space and timealongside the creativity, innovation, and improvisation invested in the praxis of rulership during a sustained and widespread period of socio-political transformation.

INTRODUCING A SOUTHERN BRITISH PERSPECTIVE
The analysis which follows draws upon a distinctive class of early medieval elite settlement known as the 'great hall complex'. Characterized by timber halls of monumental scale arranged in formalized spatial configurations, fifteen such sites are currently known across southern Britain from Kent to the Scottish border, although most have been identified from cropmarks and only four have been excavated on a systematic basis (Fig. 1, Hamerow et al. 2010, Blair 2018, pp. 114-124, Thomas 2018, McBride 2020. Some variation is apparent, but within a fairly consistent range of attributes. Some sites have a dozen or more individual buildings extending over areas up to 1.5 hectares (Yeavering, Sprouston), others are much more tightly clustered with as few as four to six buildings occupying an area of less than 0.25 hectares (McBride 2020, pp. 83-86). Within each complex it is usually possible to discern one principal hall accompanied by smaller, subsidiary structures, often sharing a similar constructional style to the main hall. Although there are exceptions, most sites display evidence for cyclical rebuilding, often applied to a whole suite of buildings.
The current dating evidence indicates that complexes of this form appeared in the final decades of the sixth century and remained current throughout the succeeding century, possibly with some continuing use beyond this (Scull and Thomas 2020). In a minority of cases, notably at Yeavering, which has assumed the status of a type-site, they may be identified convincingly with a historicallyrecorded royal vill (villa regia or vicus regius) where kings are known to have stayed and held court.
Great hall complexes have enjoyed an enduring prominence in early medieval studies since the 1950s when they were first identified through aerial photography and subsequent excavation at Yeavering in Northumberland (Hope-Taylor 1977). Since then they have been extensively quoted by historians as a material correlate for poetic evocations of aristocratic mead-hall culture (Alcock 2003, Cramp 1957, Webster 2002, Bintley 2020 and models of early lordship reconstructed from charters and other documentary sources (Sawyer 1983). Their treatment by archaeologists has followed wider trends within the discipline, with Yeavering recently provoking post-processual and phenomenological readings (Ware 2005, Walker 2010. In the last few years a new level of interest in the great hall complexes has been fuelled by fresh schemes of archaeological research and investigation that have significantly enriched the evidence base, ranging from targeted investigation of cropmark sites focusing on individual buildings (Sutton Courtenay, Long Wittenham), to area excavation of hall arrays (Lyminge) and landscape-scale investigation (Rendlesham). This has informed a major synthetic account of early medieval building culture (Blair 2018) and generated more focused studies that have defined hitherto unrecognized levels of chronological and regional complexity within the great hall tradition (Austin 2017, Thomas 2018, McBride 2020.

ANTECEDENTS AND AFTERLIVES: RETHINKING THE TEMPORALITY OF GREAT HALL COMPLEXES
Concepts of time and temporality have been central to readings of the great hall complex as a social phenomenon and continue to fuel debate and critical reflection (Scull and Thomas 2020). The temporality of these places stands out as distinctive against the wider settlement repertoire of sixth-and seventh-century southern Britain: great hall complexes are the earliest settlements to display cyclical programmes of structural repair or replacement on fixed building plots (Hamerow 2015, pp. 102-109, see further discussion on this below) and also present evidence for overt and sometimes theatrical forms of spatial remembrance in the re-use of prehistoric funerary monuments (Bradley 1987, Crewe 2012, Semple 2013, p. 97, pp. 207-212, McBride 2020. Discussion has focused on perceptions of the remote past in structuring the commemorative, monumental and ideological practices converging on great hall complexes but the antecedent early medieval activity at these places has been less well understood. Investigation at Lyminge and Rendlesham, however, has now identified extended and continuous sequences of prior occupation. In what follows we document the relevant evidence and consider its implications for understanding the relationship between the temporality of great hall sites and the contemporary practice of rulership.
At Rendlesham, metal-detecting, geophysics and trial excavation have defined an extensive polyfocal settlement in use from the early or middle fifth century (Fig. 2, Scull et al. 2016, Scull 2019. There is an elite element to the material culture from the outset but this massively intensifies from the late sixth century and is accompanied by a spatial shift in the focus of activity. A high-status residential area has been identified on a hanging promontory in the southern part of the complex, indicated by concentrations of gold and silver coinage and elite metalwork, including precious-metal jewellery and weapon fittings, a major boundary ditch, middens, and a probable monumental-scale timber hall. Metalworking debris and unfinished items, attesting production in copper alloy and precious metal, have been recovered across the settlement, but a concentration on the southern edge of the high-status residential zone may indicate the location of a workshop and suggest elite patronage of specialist craft workers. Animal bone from the middens indicates lavish consumption of meat from young animals and is consistent with a degree of provisioning from a wider hinterland. The coinage and material culture sequence indicate a sudden change of status from elite centre to unremarkable farming settlement in the second quarter of the eighth century. A similar longevity is indicated at Lyminge by the excavations on a centrallylocated plateau overlooking the source of the River Nailbourne (Fig. 2, Thomas 2017, Thomas 2018. By the time a great hall complex was established here around AD 600 the site had witnessed a least a century of sustained occupation represented by significant building remains (post-built structures and Grubenhäuser) accompanied by massive, finds-rich midden deposits attesting the lavish consumption of material and animal resources together with an impressive array of craft activity spanning iron smelting, nonferrous metalworking and possibly the manufacture of glass vessels. Moreover, interventions elsewhere in and around the village demonstrate that the Tayne Field focus, as at Rendlesham, formed part of a much wider constellation of fifth-and sixth-century activity consistent with the identification of Lyminge as the centre of a wider folk district.
There are strong indications from older excavations and cropmarks that other great hall complexes shared similar developmental trajectories (Scull and Thomas 2020). Recent discussion has emphasized public assembly and cult as the most persistent antecedent strands in the life histories of these places (Blair 2005, pp. 54-57, 2018, pp. 114-125, McBride 2020). This does not, however, explain the diversity and intensity of antecedent occupation seen at Lyminge and Rendlesham and inferable elsewhere. When evaluated on a holistic basis, the evidence suggests that a key attraction of these sites as targets of monumentalization was that they were long-established foci of lordship with permanent populations steeped in the mechanics of extracting and mobilizing landed surplus. While by no means incompatible with the notion that these places enjoyed a prior significance as focal-points for local cult and assembly, nor that their subsequent monumentalization reflects the elite appropriation of these roles (Fabech 1994, Fabech and Näsman 2013, Sundqvist 2016), our conclusion encourages a more nuanced view of the varied sources of power and legitimation that were channelled through the antecedent phases of great hall complexes.
There is no evidence for a settlement hierarchy -as defined by clear differentiations in building size, scale and ordered layout -in post-Roman England before the appearance of great hall complexes in the later sixth century (Ulmschneider 2011, Hamerow 2015. The archaeological signature of the activities transacted at fifth-and sixth-century Lyminge and Rendlesham, however, suggests prior centralities and hierarchies of place that were materialized in other ways than investment in monumental building. This serves as a reminder that expressions of rulership in the immediately post-Roman landscape need to be defined and interpreted in their own terms rather than through anachronistic back projection of later circumstances (e.g. Reynolds 2019, Scull 2019). The foundation of great hall complexes at long-established settlements can be read as monumentalizing acts by which rulers sought to lay physical and symbolic claim to places of dynastic and ancestral significance. The re-use of prehistoric monuments was one strand in this, but our analysis suggests that practices of spatial remembrance bound up with elite hall culture were motivated as much by the legitimating rhetoric of the immediate and genealogically-relevant past as by that of the distant and mythic (Gosden and Lock 1998).
No great hall complex for which there is good dating evidence was in use beyond the earlier eighth century (Scull and Thomas 2020). Some sites were permanently abandoned (Yeavering, Cowdery's Down); others, where there was a significant afterlife, saw substantial reconfigurations of space and character (Lyminge, Rendlesham). How rulers' residences developed after this is difficult to assess because of the lack of evidence for high-status secular settlements of the eighth and ninth centuries (Hamerow 2015, p. 109), but the abandonment of the great hall complex tradition appears to mark a genuine transformation in practices of elite residence and the inscription of authority on the landscape.
The factors behind this are likely to have been complex and we would caution against overly-reductive arguments. The establishment of monastic houses on former royal vills is a recurrent theme in historical accounts of elite-sponsored Christianization (Blair 2005) and archaeological discoveries from Lyminge now provide detailed insights into the nuances of this specific trajectory of development (Thomas 2013(Thomas , p. 2017. As in other parts of the early medieval northern world, monastic establishments in England became important arenas for the performance of rulership, not least because some of the roles enacted -aristocratic hospitality, communal feasting, the perpetuation of dynastic memory -complemented or indeed overlapped with those of secular elite residences and households (Fletcher 1997, Berend 2007). However, contra Blair (2018), we would distance ourselves from 'monasticisation' as a totalizing narrative for the demise of great hall complexes. The disappearance of this monumental idiom sits at the cusp of a series of transformations that reshaped systems of rulership in profound and enduring ways. Fundamental among these was the transition to more delegated forms of rulership predicated on the growing power and influence of the landed aristocracy whose private dominions came to play a key role in the extraction of surplus (Faith 1997, pp. 153-164, Hooke 1997, pp. 76-81, Lavelle 2007. Great hall complexes may thus have become obsolete simply because they failed to fit the contours of this new territorial geography with its more localized centres of political gravity (Scull and Thomas 2020).

BEHIND THE MONUMENTAL FAÇADE: KNOWLEDGE, PRACTICE AND INNOVATION IN GREAT HALL ARCHITECTURE
Defined first and foremost through their striking architecture, the study of great hall complexes has been dominated by the analysis of their constructional and spatial attributes (Millett and James 1983, James et al. 1984, Marshall and Marshall 1993, McBride 2020. While such typological examinations have their place, they offer limited insight into how these extravagant, costly and technically innovative schemes of monumental aggrandizement were constituted in relation to rulership itself and the practices and agencies through which it was reproduced. Theories of technology as social practice offer this new perspective and we start by briefly reprising the key architectural characteristics of these sites with a particular focus on evidence for innovation. While rooted in the shared idioms of early medieval timber architecture, great hall complexes display distinctive features that distinguish them from the mainstream repertoire of buildings seen in sixth-and seventhcentury southern Britain. The two most obvious are formalized planning, characterized by metrical precision and an emphasis on axial and co-axial symmetry, and sheer size -with floor areas in excess of 150 square metres and lengths above 20 metres, the great halls of this period would dwarfed contemporary domestic dwellings which typically have a floor area in the range of 50-60 square metres and a length of 10--12 m (Hamerow 2015). They are also distinguished by styles of construction requiring lavish consumption of timber and highly skilled labour. Walling consists of squared or rectangular timber planks set into deep and carefully cut foundation trenches with distinctive adaptations -external raking timbers and internal aisle posts -designed to deal with the increased stresses and loads of building at this scale (Fig. 3). In tune with their monumental character, many great halls are distinguished by unusually wide and deeply founded entrances and a proportion, notably those from Midland and northern sites, have projecting 'annexes' that display a level of spatial complexity rarely seen in buildings of this period (McBride 2020, pp. 27-35).
Great halls also marked a new threshold in the aestheticization of the built environment through costly and labour-intensive forms of architectural finishing and embellishment intended to communicate grandeur and power. We can only catch glimpses of  Hope-Taylor 1977, Millett andJames 1983). such investment in the archaeological record but what does survive, including plastered wall-renderings from Yeavering, Sutton Courtenay, Lyminge and Eynsford and the opus signinum flooring from Lyminge and Dover, demonstrates that that the physical and sensory experience of these buildings was defined as much by special materials, colours and textures as it was by scale (Thomas 2018).
The impression that great hall complexes constitute a coherent architectural tradition, subject to its own compositional rules, gains further emphasis and definition when sites are examined on a diachronic basis. It is invariably the case that the style described above, or a variation thereof, is carried across all of the component buildings within a complex, whether smaller structures ancillary to the main hall or, as with the 'theatre' at Yeavering (Hope-Taylor 1977, pp. 119-122), constructions performing specialized roles. As explored further below, it is sometimes possible to recognize architectural horizons within the diachronic development of a great hall site which represents the renewal -sometimes following destruction by fire -of an entire suite of buildings in a common style. Hamerow and Brennan (2015, p. 346) have drawn a link between great hall complexes and other strands of contemporary elite culture that display strong conformities in style and technique -feasting paraphernalia, high-status jewellery and personal regalia -as mutual expressions of a common 'court culture'. Indeed, the materialization of elite identities in later sixth-and seventhcentury England can be seen as representing a 'high culture' (Baines and Yoffee 1998) deployed to promote social cohesion, group solidarity and common identities around a constellation of leading families interconnected through marriage and other dynastic relationships. Great hall complexes have been identified in all of the major kingdoms of southern Britain (Kent, Wessex, Mercia, East Anglian and Northumbria, Fig. 1) and their geographical distribution can be viewed as a material echo of the social and dynastic networks that linked the ruling kindreds of these regional polities (Yorke 2006, pp. 61-66). But how was the tradition of the great hall complex disseminated via elite networks?
In contrast to most material expressions of high culture at this time, which are portable and difficult to source to a particular manufacturing site or place of origin (Thomas 2011), great hall complexes provide parallel sequences of elite-sponsored craft production at known locations. They therefore offer an opportunity to investigate the spread and uptake of a tradition in a more nuanced way than that afforded by models of elite emulation that simply presuppose the propagation of elite styles when leading families are locked in intense competition (McBride 2020, p. 107). This top-down notion is problematic on two grounds: it denies agency to the skilled makers that brought rulers' residences into existence; and it carries the implicit assumption that they were constructed to a predefined template carried in the minds of actors (Dobres 2000, pp. 109-110, Conneller 2011, pp. 27-29, Ingold 2013, pp. 20-31, Preston 2013. Theories of making and technology as social practice offer a more fruitful approach for interpreting the evidence at hand. Although developed for Gothic cathedrals, aspects of Turnball's sociological analysis of medieval building practice are relevant to the current context (Turnbull 1993, see also Ingold 2013, pp. 56-59). Central to the approach is the concept that each cathedral should be understood as a kind of laboratory 'in which the local, the tacit, and the messy knowledge and practices of groups of practitioners are transformed through collective work into a coherent tradition … that acted as powerful loci of social transformation, absorbing large amounts of capital and concentrating resources, skills, and labour' (Turnbull 1993, pp. 321-322). Although involving an entirely different architectural medium and smaller, more loosely organised groups of practitioners, great hall complexes can be similarly conceptualized as a 'living tradition' reproduced through the collective action of skilled makers working creatively and collaboratively to achieve a desired outcome (Ingold 2013).
The cyclical programmes of rebuilding characteristic of great hall complexes provide archaeological documentation of developments in constructional technique and architectural tradition (McBride 2020, p. 119-135). In most cases each subsequent phase of rebuilding exhibits progressively modified constructional techniques, whether applied to a single structure or a suite of buildings. Yeavering provides the best example of the latter: Hope-Taylor (1977, pp. 150-154) identified four main structural styles at the site, some representing incremental steps on an evolutionary ladder 'towards full mastery' and others variant branches of development. Lyminge, by contrast, illustrates a similar process of sequential innovation within the biography of a single building. Hall C, the largest within the complex, was rebuilt three times using a modified constructional technique on each occasion (Thomas 2017a, pp. 107-108). This tendency is seen elsewhere, including at Cowdery's Down and Dover (Thomas 2018, McBride 2020. The evidence shows that far from being a static tradition, great hall architecture continually evolved and mutated as a consequence of on-the-ground adaptations -or 'experiments' -effected through the working practices of makers. This perspective can be developed further by taking into consideration the fact that constructional adaptations occur synchronously across multiple sites. Most widespread is the transition from construction in individual post-holes to continuous foundation trenches, a sequence attested at Yeavering, Cowdery's Down and Lyminge and suggested at a number of the cropmark sites (Fig. 4, Hope-Taylor 1977, Millett and James 1983, Thomas 2017a. Although the available dating evidence precludes certainty, there is a strong possibility that the use of post-in-trench foundations, subsequently deployed across a much wider social spectrum of seventh-and eighth-century settlements, was pioneered at great hall sites (Marshall and Marshall 1993). Further evidence for this general tendency can be found in the evolving repertoire of walling techniques, most notably the transition from double to single plank timber configurations which occurs synchronously across several great hall sites in southern counties including Lyminge, Dover and Cowdery's Down (Fig. 4, Millett and James 1983, Thomas 2017a, McBride 2020. Such synchronicity is also seen in the distinctive layouts of great halls, specifically the construction of narrow end chambers or 'annexes', an adaptation that appears in the secondary phases of a number of Midland and northern sites (McBride 2020, p. 128, fig. 3.19).
Several inferences can be drawn from this. First, the rate of innovation in great hall architecture was comparatively rapid with change discernible over periods of 20-30 years -the estimated use-life of earthfast timber buildings before they need substantial repair or replacement -and thus within a practitioner's working life (Darrah 2007, pp. 54-55, Hamerow 2015. Second, the fact that the constructional adaptations occurred synchronously at multiple sites across different regions and polities indicates that innovation was communicated through inter-regional networks of knowledge and practice. Third, while conformities in style and sequence can be recognized across the corpus of great hall sites these always exist in tension with site-specific diversity and variation: great hall architecture was created at the interface between cosmopolitan and local networks of knowledge and practice (Shove et al. 2012, pp. 49-50).
The concept of professional architects is anachronistic in the current -or any medieval -context (Turnbull 1993) but the evidence is compatible with a body of highly skilled practitioners forming a shared -if periodically constituted -'community of practice' that transcended the boundaries of any one kingdom. The term 'master crafter' (cf. Kuijpers 2018, pp. 230-233, 263) would seem entirely commensurate with the degrees of skill and virtuosity embodied by great hall architecture. Of further relevance here is the possibility that great halls sites were sometimes laid out using sophisticated surveying techniques (Hope-Taylor 1977, Grave AX, Figs 25, 62, 94, Blair 2014, p. 23, 2018. For some at least, apprenticeship into this skilled domain may have involved the mastery of complex computational skills as well as the haptic technology of timber. Labour and logistics, especially the sourcing of timber, were largely embedded within the structures of the rural economy that focused on great hall complexes but such master craftsmen and their skilled assistants may have enjoyed greater autonomy, entering the service of ruling families when needed on the basis of reputation and travelling from place to place as their skills were required. Their buildings were the setting for diplomatic and other encounters between elite households and retinues and so would have been experienced directly by potential patrons. The release of a call upon their skills from one ruler to another, or from ruler to leading follower, can be envisaged within the matrix or reciprocities that governed lordship and society. Such dynamics would explain the second-tier aristocratic great hall sites identified by McBride (2020, pp. 91-98) and the subsequent wider adoption of architectural and constructional innovations noted above.
It is interesting to consider comparisons with the skilled metalworkers who produced portable elite material and who were also attached or periodically embedded in elite households or residences (Hjärther-Holdar et al. 2002, Hedeager 2011, Axboe 2012, Wright 2019. Master smiths would have crafted within a closed workshop groupskilled assistants and apprentices -whereas master builders would have worked in sustained and creative collaboration with a wider community of estate dependents including carpenters capable of turning their own knowledge and skills to elite purpose. Such collaborative interactions of this type would have sparked adaptions and mutations in the great hall tradition giving rise to the variability manifest at a localized scale. The virtuoso precious metal artefacts of this period are celebrated as representingquite literally -a golden age in the jeweller's art (Coatsworth and Pinder 2002, Webster 2012, Fern et al. 2019), our contention is that the practitioners behind the remarkable seventh-century fluorescence of ruler's residences in England deserve similar recognition.

RULERS' RESIDENCES AS CENTRES: CONSTELLATIONS OF AUTHORITY IN EARLY MEDIEVAL BRITAIN
When considering how great hall complexes were situated in wider social and economic networks, and how they functioned in the orchestration and maintenance of rulership, there is an important distinction to be drawn between central person and central place. Networks of social and economic relationships might focus on elite individuals regardless of where they are at any one time, inherent to a person's social identity and roles; they might focus on elite individuals but in ways that are enacted only at specific places such as residences, assembly sites and cult sites; or they might focus on a specific place or places in ways which may or may not require the presence of the central person. These overlapping social geographies are further complicated by the fact that where the roles of central persons are linked to specific places these may or may not be at the same site, and that some central place functions, notably agrarian administration and the gathering of dues and renders, were the province of delegated authority. A further useful distinction may therefore be drawn between the practice and materialization of rulership, and the social and economic infrastructures that supported it. To the former belong evidence for elite lifestyle, culture and contacts, patronage and conspicuous consumption, and for assemblies and the theatre of rulership; to the latter farming, the collection and processing of a landed surplus, and routine craft production and exchange. While these categorizations inevitably include elements of cross-over, ambiguity and overlay, they offer a framework within which to address the material evidence and elucidate the networks and hierarchies of social, economic and political power that it represents.
As noted above, approaches to centrality in early medieval Northern Europe have been heavily influenced by models developed in Scandinavian archaeology that place emphasis on the entanglement of social, political, economic and ideological spheres (Brink 1996, Fabech 1999, Näsman 2000, Nielsen 2014). This has given rise to the model of the 'central place complex', an aggregation of foci that intermesh to varying degrees settlement, elite residence, surplus extraction, jurisdiction, craft production, exchange and cult activity. Such phenomena can be identified at a range of scales and the model has been applied in England both to circumscribed places (Scull et al. 2016) and to locales displaying unusual concentrations of wealth and elite investment in funerary displays and other dimensions of the cultural landscape (Dickinson et al. 2011, Noble et al. 2013, Behr et al. 2014.
Another influential concept in recent literature on rulers' residences in southern Britain, which overlaps with landscape-scale readings of the 'central place complex', is that of the core zone or heartland (Wood 2008, Blair 2018. Drawn initially from documentary sources for the geography of seventh-and eighth-century kingship, the central argument -that itineraries and places of royal interest focused in areas where ruling dynasties had ancestral territorial interests -may find archaeological support where great hall sites cluster within landscapes that display long-term trajectories of centrality and elite investment and which can be associated with named early polities. Thus, sites in the upper Thames Valley can be associated with the Gewisse, those at Yeavering, Milfield and Sprouston with Bernicia, and those in Warwickshire with Mercian rulers. It is however unclear how the sites within these clusters interoperated if they were in use at the same time or to what extent the clusters themselves are products of archaeological recognition and retrieval biases at a national scale. Yeavering has been the subject of attempts to reconstruct territorial units around great hall sites (O'Brien 2002(O'Brien , 2005 but exercises so heavily reliant on back-projection from later historical and topographic sources must be treated with caution (Scull 2019, p. 397, Reynolds 2019). There are, however, other ways to model systems of dependency with which great hall complexes were enmeshed, one being to consider their articulation with systems of surplus extraction. Recent work on Yeavering has shown that the site occupied a strategic position on seasonal transhumance routes between lowland and upland (Semple et al. 2017). This provides a new understanding of the 'great enclosure' and similar constructions on other northern sites as manifestations of the importance of cattle to the wealth economy (McBride 2020, pp. 35-48). More generally, it emphasizes how the location and physical structuring of great hall complexes could be governed by their place in the landed economy.
A complementary approach, where suitable datasets exist, is to use fine-grained analyses of artefactual, environmental and other material proxies to model the range of activities and transactions undertaken at great hall sites, and their economic and social reach, thereby overlaying constructs of territorial geography with characterizations of these places by social behaviour, practice and performance. Until recently, southern British sites offered only very limited opportunities for such analysis: few great hall complexes have seen excavation on any scale, and the two most extensively-excavated, Yeavering and Cowdery's Down, were poor in material culture and environmental evidence (Hope-Taylor 1977, James et al. 1984, Hamerow 2015. The rich assemblages from Lyminge and Rendlesham, however, have altered the case and provide material evidence for aspects of these places that have been assumed in the past but have not previously been seen in the archaeology. They also document the role of these places in driving innovations in the extraction and conversion of landed surplus and as centres of monetization. Lyminge occupied a strategic position at the head of the valley of the River Nailbourne, which formed a key communication artery between the south coast and the Kentish capital of Canterbury while its place-name denotes that it had assumed the identity of a district capital by the sixth century (Brookes 2011, Thomas 2013. The biological and artefactual assemblages recovered from the site shed light on the internal mechanics of the Lyminge microterritory. A gravitational influence over outlying resource zones and their dependent communities finds clear expression in these proxies. A high incidence of pig consumption combined with iron production, including smelting, shows that in the sixth and seventh centuries the royal focus at Lyminge was exploiting the heavily wooded upland of the Weald, implying established routeways and patterns of transhumance. Strong and sustained connectivity with coastal territories is signalled by Lyminge's unusually large marine fish bone assemblage which provokes attention given that fish appear to have played a relatively minor role in diet in southern Britain prior to the 'Fish Event Horizon' of AD 1000 (Barrett et al. 2004, Sykes 2007. The nearest coastline is 10 km south of the site so the exploitation of marine resources at these levels must be seen as a deliberate choice used to signal the status and ideological affiliation of its inhabitants (Thomas 2013). Charter evidence demonstrates that coastal fisheries, presumably operated by estate dependents, were supplying Lyminge as a monastic community by the eighth century (Brooks 1988, Brooks andKelly 2013, pp. 28-34) and similar provisioning networks must have been functioning during the site's pre-Christian phases. While the emphasis at this earlier period appears to have been on coastal and migratory as opposed to deep water species that dominate the larger monastic-phase assemblage, it nevertheless represents significant investment in infrastructure combined with the deployment of new skills, knowledge and conceptual understandings of the marine environment.
Innovation is also apparent in the agricultural regime that operated under Lyminge's authority, focused on the fertile lower slopes of the Nailbourne valley in the immediate environs of the settlement. This is manifest in the discovery of a plough coulter from a stratified seventh-century context which demonstrates that continentally-inspired heavy-plough plough technology was being deployed in east Kent at a precociously early date most likely via elite networks spanning the English Channel (Thomas et al. 2016).
Taken together, these and other material indicators suggest that Lyminge drew on, and exercised lordship over, a resource territory broadly equivalent to lathe of Limenwara recorded in later sources. It is possible to envisage a core area of jurisdiction embracing the southerly stretch of the valley of the River Nailbourne forming the immediate catchment of the settlement, surrounded by a halo of more weakly and periodically constituted relationships and networks of dependency by which outlying resources were channelled to the centre (Fig.  5). Lyminge's jurisdiction and authority relied on deeply-rooted networks of communication and resource exploitation inherited from earlier post-Roman territorial configurations. At the same time, the conspicuous consumption that defined its later sixth-and seventh-century existence fuelled innovations in the scale and intensity of resource extraction from composite parts of its extended territorial dominion. This dynamic phase of reconfiguration laid the foundations for the more tightly integrated system of estate management witnessed in its subsequent existence as a monastic enterprise.
A layering of social and economic centralities can also be discerned in Rendlesham's consumption profile. There was a mixed farming regime with an emphasis on stock raising and the evidence is consistent with additional food renders from a wider hinterland. Lavish consumption of meat suggests episodes of feasting associated with periodic elite residence, and an elite presence is confirmed by finds of gold-and-garnet dress jewellery and weapon fittings. The evidence for both elite metalworking and manufacture on a considerable scale of low-value utilitarian items raises the question of whether a single workshop -in the sense of a master craftsman and assistants -undertook both. If so, and if attached to the retinue of a peripatetic magnate, this would afford elite patrons access to their skills at all times, and a local population access when the household was in residence. Without arguing that this was the exclusive mode of production, it does illustrate how ties of dependence and patronage might act to align aspects of economic with social and political centrality enacted periodically at specific places. In the elite sphere controlling access to craft skills, and so to the material trappings of elite identity, could be seen to reinforce the relationships of service and reward fundamental to personal lordship (Hedeager 2011, pp. 145-147, Wright 2019.
Continental gold coinage and fragments from hanging bowls and Byzantine copperalloy basins show that in the later sixth and earlier seventh centuries people here were acquiring the inter-regional imports that made up part of the contemporary elite cultural package: with these, and the gold-andgarnet jewellery, we see in the settlement context at Rendlesham elements of the suite of elite markers that were selected for burial at Sutton Hoo. Byzantine material in England is usually explained as the product of socially-embedded gift-exchange (Richards 1980, Harris 2003, pp. 64-69, 164-167, Drauschke 2007) but contemporary Byzantine copper coinage at Rendlesham suggests some more direct commercial contact with the Mediterranean. Any long-distance trade in high-value items would be directed at elites or their agents, and trade contacts may have had a diplomatic dimension. The reach and complexity of elite-focussed networks implied by this material chimes with the broader picture derived from written sources in, for example, accounts of fosterage, dynastic marriage and political exile (e.g. Härke 1997, p. 126, Yorke 1990.
This coincidence of social and economic centralities, and their links to political authority, is also seen in the evidence for coin use from the later sixth century. The coin assemblage represents transactions over a period of a century or more that spanned the circulation of continental gold issues, the production of the first English gold coinages, and the transition from gold to silver in the third quarter of the seventh century. Transactions in gold would conventionally be seen as social and jurisdictional payments such as tribute, fines and gifts, but given the evidence for trading contact with the Mediterranean world more commercial transactions -albeit socially-restrictedshould not be ruled out. Conversely, the lower-value silver coinage is usually seen as indicating an increasingly monetized market economy, but we should also envisage its use in jurisdictional and administrative payments. It seems likely that from the middle years of the seventh century first gold shillings and then early silver pennies were struck here under royal authority. Bullion and coinage flowed disproportionately to Rendlesham, and as an early centre of coin use it must have acted to promote wider and deeper monetization in its hinterland.
Like the coins, much of the ploughsoil assemblage, including status items such as harness and weapon fittings, appears to be material dropped on the old ground surface around the main foci of occupation and can be seen as the aggregate loss from periodic gatherings as well as permanent occupation. At least some of this material, therefore, is likely to be a residue of the actions and transactions of rulership at a theatre of power: public hospitality, gift-giving and tribute-taking, gatherings of local leaders and their armed manpower, assemblies for justice and jurisdiction. A permanent administrative function, the periodic presence of an elite household, and assemblies of the social elite would also be powerful attractions for directed trade and might over time foster a periodic market or fair.
Rendlesham sits within a river territory comprising the catchments of the rivers Deben and Alde where the combination of topography, soil and early medieval human geography indicate an agricultural core focused on the river valleys with peripheral zones of wood and wood pasture on the marginal higher ground of the interfluves (Scull and Williamson 2018). Far from being environmentally liminal (cf. Blair 2018, p. 112), it was centrally situated within a rich farming territory and concentration of population, and accessible from the sea via the estuaries of the Alde and Deben. Its economic and social centrality is matched by its physical location as a central pace, and the establishment of monumental elite cemeteries at Snape and Sutton Hoo -overlooking the main waterborne approaches to the core territory -can be seen as complementary elements of a ceremonial landscape asserting new levels of social distance and regional lordship. In its immediate valleyside location Rendlesham, like other contemporary farming settlements, is optimally placed to exploit the range of resources afforded by the landscape (Fig. 5). Like Lyminge, we can envisage direct control of a nearby territory and a looser network of dues and obligations over a wider area. Entangled with this were the transactions of rulership at regional and inter-regional scales, around which focused local, regional and inter-regional networks of production and exchange. It is probably not coincidental that the catchment territory closely corresponds with the area of the 'Wicklaw Hundreds', a jurisdictional Liberty granted to Ely Abbey at some time before the 11th century (Warner 1996, pp. 152-157, Scull andWilliamson 2018).

CONCLUSIONS
In this paper we offer new and enriched perspectives on great hall complexes as a key source of archaeological evidence for understanding the practice of rulership in southern Britain over the fifth to the ninth centuries AD. At various points in the discussion, we have turned to early medieval Scandinavia as a source for broader contextualization of the southern British scene. By way of conclusion, we pursue this comparison further to advance common research agendas, again with reference to the analytical themes of temporality, monumentality and centrality.
Our evaluation of the enlarged datasets now available for great hall complexes shows that the practice of prehistoric monument re-use often emphasized in the discussion of these sites -while certainly significant in some cases -needs to be balanced by the temporal perspective provided by continuous phases of antecedent occupation. Invariably constructed in places of persistent human activity with complex multiple pasts, great hall complexes embodied strategies of legitimation, and spatial and monumental remembrance, that referenced both genealogical and mythic time (Gosden and Lock 1998). This situation is closely echoed by the multilayered commemorative practices recognized at Lejre and other elite ceremonial landscapes in southern Scandinavia (Carlie 2006, Lund andArwill-Nordbladh 2016, p. 417). Such a comparison supports the view that complex notions of time, lineage and ancestry pervaded rulership ideology across the North Sea zone, constituting a shared repertoire of commemorative behaviours and practices (Steinsland et al. 2011b, Semple 2013. It is also important to recognize that laying claim to pre-existing settlements, particularly those with prior central place functions, also had a tangible, real-world significance as a contemporary strategy of rulership and political authority. Deeply embedded in the social fabric of 'small worlds' and offering privileged access into extractive and jurisdictional networks, these places were valuable anchors for consolidating larger, supra-local and regional hegemonies that emerged in the transition to more centralized systems of lordship. If a temporal prism helps to bring interregional similarities into focus, it also highlights significant differences. There is nothing in the corpus of great hall complexes in southern Britain to compare with the extended, in some cases centuries-long, sequences of hall construction seen in Scandinavia (Stenholm 2006, Eriksen 2016. 4 While 'spatial remembrance' is reflected in both regions through cyclical schemes of hall rebuilding on the same footprint, in southern Britain this practice rarely exceeds three constructional generations with some sites such as Sutton Courtenay displaying only a single building phase (Hamerow and Brennan 2015). The very different socio-political trajectories of the two regions must in part be responsible for this general distinction. The emergence of great hall complexes in southern Britain around AD 600 represented the fruits of less than two centuries of post-Roman sociopolitical transformation whereas in Scandinavia the underlying structures of rulership and its monumental rhetoric developed over a considerably longer period (Jørgensen 2009, Skre 2020; Skre, in press).
Considering the monumentality of great hall complexes, we demonstrated how a practice-based framework can be used to move beyond static typological comparisons of these sites to illuminate the skills, competences and webs of agency that lie behind these innovative architectural manifestations. We argue that great hall complexes should be understood as a living architectural tradition reproduced and continually adapted through interactions between highly skilled specialists attached to mobile elite retinues and local dependents embedded within the social hinterlands of rulers' residences (Blair 2018, p. 39, McBride 2020. Studies of hall culture in Scandinavia commonly place emphasis on the conservatism of architectural practice down through the Late Iron Age and Viking periods (e.g. Carstens et al. 2015). Yet, as argued by Beck (2014), this may well be a product of the tendency of traditional typological approaches to create artificially homogenized visions of early medieval building culture. This supposition is supported by striking indications of architectural innovation revealed by recent excavations of magnate complexes occupying the upper -royal and sub-royal -echelons of the settlement spectrum: from the novel walling technique and exuberant decorative embellishments featured by the largest of the grand halls at the core of the seventh-century ceremonial complex of Old Uppsala (Ljungkvist and Frölund 2015), through the atypical buildings and spatial configurations displayed by various iterations of the magnate complex at seventh-to tenth-century Tisso (Jørgensen 2003), to the unique design and constructional attributes of the recently recovered palisaded enclosure at late tenth-century Jelling (Jessen et al. 2014). This highly selective appraisal may of course create distortions of its own by giving the impression that innovations were chronologically and spatially concentrated rather than being more continuously distributed across time and space. Nonetheless, in both Scandinavia and southern Britain the seventh century stands out as a genuinely dynamic and inventive period in the development of elite architectural and monumental practices. Overall, rulers' residences across the North Sea region emerge as a prime arena of creativity and innovation: laboratories of skilled practice where novel techniques and representations of power were conjured by renegotiating and reinventing the rules of tradition.
We concluded our analysis by reviewing the application of concepts of centrality to the interpretation of rulers' residences. Using Lyminge and Rendlesham as case studies, we showed how artefactual and ecofactual proxies can be used to develop a more nuanced understanding of the different levels authority and reach exercised by great hall complexes and to recognize innovation as an essential feature of the resource and jurisdictional systems that operated through them. We argued that the complex materializations of centrality seen on such sites is a product of the interplay between permanent entanglements on the one hand -those embodied in their role as the hubs of long-standing economic, social and jurisdictional networks -and periodic entanglements on the other, activated in relation to the public presentation of the ruler.
Research growing out of the investigation of central place complexes in southern Scandinavia has resulted in myriad perspectives on the territorial manifestations of early medieval rulership. Alongside Brink's (1997Brink's ( , p. 1998) fundamental work, Callmer's (2001) 'solar system' model has proven influential in offering a spatial conceptualization of the extensive hinterlands of authority and interdependency surrounding major central place complexes characterizing this cultural zone. While aspects of his model have received criticism, it has formed an important catalyst for further studies driven towards reconstructing the territorial architecture and dynamics of rulers' residences. Strong points of contact exist between this recent research and our examination of the evidence from southern Britain, most notably detailed analytical studies of ecofactual proxies from the site Uppåkra (e.g. Larsson 2015, 2018) which have shed new light on the the mechanics of food supply within its socio-economic hinterland. This paper has been necessarily selective in its application of a practice-based framework to the interpretation of rulers' residences: what wider potential does this approach hold for future research agendas? There is clear scope for placing the performative and experiential dimensions of rulers' residences under a comparative analytical lens, drawing upon evidence for ritualized symmetry characterising spatial articulation of these sites, combined with the theatrical manipulation of wider landscape settings (e.g. Reynolds 2003, Hamerow 2015. This focus offers rich potential for interdisciplinary engagement with the fields of architecture, science and technology studies (STS) and construction management, which possess powerful methodologies for modelling the embodied experience of architectural space and manipulated landscape settings. Interaction with these same disciplines would also offer scope for refining understanding of the socio-technical practices and networks involved in the construction of these sites, for example by providing tailored methodologies for sequencing and measuring resources and materials used in complex construction projects and delineating steps in the innovation and diffusion of architectural traditions. Alongside new avenues of interdisciplinary engagement, comparative work on the detailed cultural and environmental proxies available from a large and growing number of sites would repay dividends, especially if directed towards new questions concerning the temporal rhythms of rulers' residences and their interplay with systems of itinerancy. draws on spatial analysis undertaken for the Leverhulme-funded project Lordship and Landscape in East Anglia CE 400-800 (Leverhulme RPG-2017-172), and to Matthew Austin and Sarah Lambert-Gates for the preparation of Figs 1 and Fig 2 and to Professor Roberta Gilchrist and the anonymous reviewers for their helpful comments. Funding for this research was provided by the Arts and Humanities Research Council Networking Scheme AH/N000218/1.

This work was supported by the Arts and Humanities
Research Council [AH/ N000218/1].

NOTES 1 Monumentalizing Kingship: Places of Royal
Residence and the Making of Early Medieval British Kingdoms AD 500-800: AHRC Research Networking Scheme AH/N000218/1. 2 In this paper we mean by 'southern Britain' the areas of mainland Britain associated with English-speaking polities that emerged around the end of the sixth century AD and were subsequently consolidated into a unified 'Anglo-Saxon' state. 3 For other applications of practice theory in medieval archaeology see, in particular, Gilchrist (2012, p. 7-13). 4 The significantly larger scale of Scandinavian magnate halls in comparison to southern British equivalents is another clear distinction. This is not sufficient, as argued by Loveluck (2013, p. 130), to deny the latter an assembly role, for experimental reconstructions demonstrate that these buildings were still large enough to accommodate gatherings of 30-40 guests. Perhaps more significant is the fact that within their respective cultural contexts, magnate halls were roughly two-three times bigger than typical domestic dwellings: scale was used to proclaim social distance.