TRAC 1991: ‘Villas as a Key to Social Structure’? Some Comments on Recent Approaches to the Romano–British Villa and Some Suggestions Towards and Alternative, by Robert Rippengal

Download file in PDF format: TRAC 1991: ‘Villas as a Key to Social Structure’? Some Comments on Recent Approaches to the Romano–British Villa and Some Suggestions Towards and Alternative (pp. 79–101)

TRAC 1991:
By Robert Rippengal

The paper makes a critical examination of one of the more influential pieces of ‘social archaeology’ to have come out of Roman<;-British studies - J. T. Smith's 'Villas as a key to social structure'. Smith's suggestions that many villas were occupied by 'extended families' is found wanting under a more detailed examination of the evidence. Instead, while there may sometimes be more than one clomestic strncture on a site, or even indications that individual buildings were occupied by more than a single nuclear family, it is suggested that it is the nature of lhe relationships involved that should be under scrutiny. To this end, a brief outline is given of an alternative approach to the material. This takes up Bordieu's notion of habitus with its emphasis on the role of material culture, and some examples are given to illustrate how it might expand OUT perceptions of the material from Romano-British sites. INTRODUCTION In his recent and important volume The Rural Settlement of Roman Britain, Richard Hingley has felt able to state that 'considerable evidence exists to indicate that extended families were typical of the rural areas of the province' (Hingley 1989, 149). That he did so is perhaps a testament to the influence of J.T. Smith, for although Hingley's discussion is itself an elaborate one, it is in essence an extension of the paper by Smith published ten years earlier and entitled 'Villas as a key to social structure' (Smith 1978). It is the propositions set forth in that formative paper that are of primary interest here. J.T. Smith's Unit Theory Written explicitly as an antidote to the perceived stagnation of villa studies, Smith's paper sought to argue that 'a great deal can be learnt by comparative study of ... known house plans', and in so doing to provide a new framework for the examination and understanding of a large part of the body of known villas. Followed up by a series of papers dealing mainly with the interpretation of individual sites (e.g. Smith 1985; 1987), this paper has come to form one of the most important of villa studies in the last 30 years. Indeed, in some quarters his position represents almost a standard interpretation, adopted or at least acknowledged in a series of recent works (Hingley 1989; Reece 1987; Miles 1986; Millet 1990). It is by no means denied that his ideas represent a novel departure, one that appears to enliven what had become a rather sterile area of research, and this is surely where their appeal lies. However, it will be suggested that for all their impact, acceptance of these ideas has tended to be somewhat uncritical. There exists a perception that certain points have been demonstrated in this paper when in fact they clearly have not; it has become one of those many archaeological myths, widely cited without detailed reference as one might cite a fact. The essentials of his argument are relatively straightforward. He suggests that an examination of the plans of a great many villas reveals evidence of joint occupancy. Sometimes this involves the presence of more than one domestic structure on a site, but as often as not it is discerned in the divisions of what is termed the 'unit system' within a single building (after Hemp and Gresham 1943, 98). This, it is claimed, represents an expression of 'existing social relations' based on the 'kindred or extended family' and indicating the survival of a pre-conquest, so-called 'Celtic' society (Smith 1978, 170-2). It is then briefly contrasted to what are termed 'hall villas'. These do not show indications of multiple occupation but instead stand in a position of superiority to various other buildings, 'the houses of social inferiors, of people who were dependants rather than equals' (ibid., 170). CRITIQUE In the course of elaborating such a position, a considerable number of illustrations are offered and these will be considered more fully below. However, before doing so a number of more general points are perhaps worth making. At the outset, then, Smith outlines several basic principles that underpin the subsequent argument. Thus, he starts with the proposition that: 'every villa should be regarded as an expression, however humble, of classical architecture' and that 'any deviation from the accepted tenets of that architecture is significant' (ibid., 149). This represents a bland statement, as of fact, yet it is surely an assertion that demands some further elaboration, not to say justification. Implicated is the whole process of Romanisation. This is something that has indeed been identified as central to the future development of Roman studies (Burnham and Johnson 1979; Slofstra 1983), yet nowhere is the nature of that process discussed here. Elsewhere it is claimed we are dealing with an essentially Celtic people, one that was actually very resistant to change, yet here we are told that classical architecture was taken on board lock-stock-and-barrel and that the slightest deviation is of profound significance. In contrast, it might be suggested that if native culture was so persistent it should be those classical elements that were adopted that require explanation, not deviation from classical forms. In fact, Smith is not even consistent in his application of this precept. Thus, for the purposes of the paper he lists 'symmetry of elevation and plan' (Smith 1978, 150) as one of the basic tenets of classical architecture to be assumed, yet elsewhere a considerable number of the sites he lists are denoted as examples of the unit system simply by virtue of their symmetry! Elsewhere, it is suggested that a 'winged corridor villa' is more Romanised than an 'aisled house' (ibid., 153), or that it is fallacious 'to regard small villas of the type of Lockleys or Park Street as vernacular architecture', that they are really 'modest manifestations of classical architecture'(ibid., 170). Once again, however, we might ask why, what are the distinctions implied here and what is the nature of 'Romanisation' they embody? It is these issues that should surely be addressed in a paper of this kind whereas they are left unsaid, in the air. It might be suggested that we know classical forms were adopted simply by virtue of the fact that we find houses with features such as the collonaded, winged facade. Therefore, there is something to explain if they do not also have an axial approach with central entrance into the main room of the house. In contrast, however, I would suggest that it is hardly surprising that in the middle of the British countryside, hundreds of miles and at least one sea crossing from Italy, on the fringes of the north-western part of the Empire, we do not find an exact transplantation of the Pompeiian villa, and therefore to assign any differences to profound social considerations needs considerably more justification. Indeed, Percival (1976, 58-9) notes that even in Italy when we are dealing with working villas as opposed to the limited class of luxury retreats archetypal forms are less important, layout being 'determined much more by the requirements of farming practice than by the dictates of architectural fashion'. To this it might be added that the pedigree of Romano-British houses has been a subject of continuous and vigorous debate since the nineteenth century, within which tradition the nature of classical influence is generally seen as rather more equivocal than Smith allows for. Indeed, as long ago as 1911 Ward noted of Romano-British and Pompeian houses that 'the one set ... is about as unlike the other set, in that most important point, the planning or arrangement, as both are unlike the houses of the moderns'(Ward 1911; see also Haverfield 1912; Collingwood 1923; Collingwood and Myers 1937; Richmond 1969; Hingley 1989). Of interest in this respect is the second of Smith's presuppositions, the notion that 'a clear distinction must be drawn between a house plan which corresponds with the reality of contemporary social organisation, and the facade or other classical trimmings' (Smith 1978, 149). It is notable that Smith's own use of the term 'trimmings' would seem to assign a rather less over-bearing role to classical influence. Indeed, it would tend to bring him more into line with the bulk of thought on the matter, for there is, after Swoboda, a widely accepted contrast between the typical facade and the houses behind it (Swoboda 1919). However, the question becomes, how do we elucidate the 'reality of contemporary social organisation' these houses are supposed to represent. While Smith's attempt to elucidate such questions is certainly unusual in the context of Romano-British archaeology it is perhaps undermined by a lack of theoretical elaboration and the consequent failure to bridge the gap between theory and data. Towards illustrating this, we may move on to the third and final of the initial conditions, the notion that (Smith 1978, 149-50): "[those villas] in the provinces north of the Alps are particularly relevant to one another because the native peoples there are likely to have had more in common with one another than any of them had with Roman life, and consequently comparison of their villas should reveal those elements common to the social structures found by the conquerers. And from all this it follows that changes in type of plan or elevation correspond not only to variation In wealth but, more importantly, to changes in social structure." Although one might be sympathetic to the basic notion expressed here, as stated it represents a series of non-sequiturs, leaps that are never addressed more fully. Perhaps two points may be made. Firstly, the notion of 'the kind of social relations characteristic of Celtic society', viz. the 'extended family', is simply stated, presented almost as a fait accompli. In fact, although the basis of the paper is clearly the notion that social questions may be addressed within the archaeology, what becomes apparent is that such issues are not really addressed at all. Instead, a largely meaningless term is thrust onto the material without further comment or elaboration. In contrast, it may be suggested that the notion of a Celtic social form, particularly the 'pan-European' form hinted at, is one that demands rather more discussion. Kinship is of course notoriously complex, yet this merely makes the lack of such discussion all the more telling. Smith himself concedes how 'surprising' the idea of joint occupancy is, yet never takes up the challenge this represents. Indeed, this aside, if the social form at issue is essentially a pre-<:onquest one, it seems particularly surprising that no reference is made to any of the prehistoric material or to any of the work by prehistorians on this subject for whom the notion of extended families would certainly be controversial. For all this, the very attempt to look at such questions as archaeological problems is an important one, for the all too limited horizons of RomanoBritish archaeology have for long tended to reduce its impact. However, even the basic proposition that villas indeed represent 'a key to social structure'is undermined by a failure to examine the relationship between 'society' and 'architecture', to provide a rationale for why we might be able to talk of kinship when we are simply looking at the floor plans of long derelict buildings. Early in the paper it is implied that there are 'fundamental elements' of villa plans that 'express social structure', yet nowhere is it clearly stated what these elements might be (ibid., 150). Instead, Smith's conception of the link between society and architecture would seem to be rather limited. In effect, he seems to suggest that house plans in some way equate with society such that we may simply look at the plans and 'read off the social form. Lacking is any account of how this relationship operated in practice and, thereby, of why house plans should equate to social form. Equally, there is no concession that there may be other formative and potent influences on architectural form. Thus, in discussing the placement of certain water shrines, he notes, without elucidating why, that it was dictated 'by some profound social consideration', and that, somehow, 'architecture tells us what it was' (ibid., 157). In contrast, it is stressed here that architecture does not tell us anything per se. We might attempt to elucidate meanings within architectural forms, but Smith's simple equations are fundamentally misleading. Central in this respect is the supposition that two or more domestic units equates with an extended family. In essence this rests on the notion that where such discreet units are encountered they are of equal status: 'it appears not only that some villas comprised two houses where the orthodox view would lead us to expect only one, but that far from there being any suggestion that one house was subordinated to the other .. . both houses, though they may be of different sizes, are of equal status'(ibid., 153). This assumption is fundamental, yet nowhere is it adequately examined, let alone justified with reference to the material. At this point we must turn to the examples for it is here that this issue must be pursued. Initially, then, Smith suggests that there are a number of sites on which we may identify two such houses: Newton St Loe (Fig. 8); Gayton Thorpe (Fig. 9); Beadlam; Arquennes; and Paulton. As an initial observation, it is perhaps to be conceded that most of the examples given do appear at least plausible as two houses. However, what is so striking is that even if there are two domestic units on these sites, they are buildings of different character whose equality of status must be open to some question. In criticising Smith's notion of equivalent status, it is conceded that while he is somewhat ambivalent on the point he does not mean that each house or 'unit' is literally exactly the same as the other(s). Instead, differences are taken to be of degree rather than of kind. This, one assumes, is where we may see the contrast with hall villas where differences were somewhat more profound. However, what is stressed here is that if we are claiming to work from the archaeology, the material differences seen in the archaeological remains are often such that to start with an assumption of equality seems inadequate. Instead this is something that Smith needs to demonstrate as being the case in spite of the obvious material differences. Indeed, so far from assuming equality and working from this basis, it is surely the nature of the relationships between the various elements on these sites that should be the very subject of investigation. Are they such that we should talk in terms of kinship or differences of degree, or do they suggest differences of kind, perhaps better discussed in terms of dominance and dependence and/ or mode and relations of production? Figure 8. Newton St Loe (redrawm by Rob Rippengal). Newton St Loe To return to the examples, at Newton St Loe while both buildings are of similar size they nevertheless appear quite different: one is made up of a series of large apartments, the other of a great many smaJl rooms. Clearly this means very little as an observation on its own. However, it does suggest that there may be more to them than is encompassed in Smith's description. Unfortunately, this site is only poorly reported (IRS 26 (1936), 43-6; VCH Somerset i, 302), yet even these limited descriptions tell us that the westerly building was considered the more interesting of the two and was the only one with baths, hypocausts or tessellated floors, including severaJ mosaic pavements. What this means might be open to some question, and again it is stressed that the standard of reporting is poor. However, what is clear is that to taJk simply of two equivaJent houses begins to look rather inadequate, particularly if this is to be cited as a primary example. Figure 9. Gayton Thorpe (redraum by Rob Rippengal). Gayton Thorpe We might look next to Gayton Thorpe, for to say here that the two buildings are equivalent seems a nonsense on the basis of size alone - one has five rooms, the other thirteen. Smith, however, specifically notes: 'Both have the standard type of facade with the portico and wing rooms which proclaim a house of some standing', using this to reject the possibility that the smaller one could have housed farm labourers (Smith 1978, 153). This is an interesting statement, and might be contrasted with his earlier suggestion that all may not be what it seems hidden behind such a facade. Once again, the standard of reporting is rather limited. However, it is sufficient to tell us that the facade on the southern building might be a case in point - it does not seem to denote a 'house of some standing' as Smith suggests, but stands in front of and masks a building that is in stark contrast to the one next to it. The latter, as well as having a far more elaborate ground plan also has baths, other heated rooms and a large central room that produced a mosaic pavement, painted plaster and fragments of Italian marble. The southern building, in contrast, is rather different. The facade is, we are told, well constructed and the wing rooms contain a geometric mosaic pavement and a tessellated floor respectively. However, even here it is to be noted that the facade does not quite match that of the building next door. While the latter includes evidence of a columned entrance and dwarf wall for a colonnaded varandah, here, while there is some evidence for a verandah in the form of a concrete floor, the lack of any more substantial remains indicates an altogether less imposing edifice. This aside, behind the facade the excavator notes that the standard of construction and setting out were clearly inferior. What is more, most of these rooms were furnished with simple rammed chalk floors, and nowhere in this building were there the baths, heated rooms or other signs of comfort and ostentation found next door. The one surprising exception to this is that both buildings have a mosaic, something on which Smith places some little emphasis. However, without purporting to offer a 'solution' to this apparent enigma, it might be noted that in the southern building the mosaic is placed in a wing room. In contrast, the mosaic in the main building is laid in the large central room, perhaps a significant difference. The latter, then, marks the heart of a building built to a high degree of sophistication throughout. The former, in contrast, is perhaps the one room that affords such a degree of comfort in the whole of this otherwise inferior building. Once again, the meaning of this observation is open to some question. However, if we were actually considering the nature of the relationship between the occupants of the two buildings, might not this kind of thing be a starting point? The aim of all this is not to try to demonstrate that there were not two domestic units on these sites. Instead, several points are intended: that to indicate that some sites have what appear to be two 'houses' and elaborate this no further is insufficient; that to assume a simple equality between the two elements is particularly inadequate to the point of being misleading; and that the evidence recovered on most of these sites provides far more useful and relevant information than Smith makes use of, a lack that can only do his argument harm. In sum, it is to be stressed that although Smith raises an interesting possibility he does not elaborate it in a way that would make it convincing. Montacute-Ham Hill Moving on from here, it is assumed in Smith's paper that he has thus far demonstrated that certain sites have two houses, a notion that is then expanded by drawing in a number of other sites where the evidence seems more equivocal. The first of these is Montacute-Ham Hill (Fig. 10), where one relatively typical domestic block is linked by a wall to two further elements: a single separate room with a tessellated floor and independent entrance; and what is described as a work-hall, on the Wirtschaftshalle model (i.e. including living accommodation). On this basis Smith compares it to those sites just described, where he assigns the separate houses equal status, and suggests 'something similar must underlie [this] curious villa'(ibid., 153). Without going into elaborate detail, the plan alone must surely make this a rather remarkable claim. Figure 10. Montacute-Ham Hill (redrawn by Rob Rippengal). It might be countered here that, as noted above, Smith is not really suggesting literally equal status for these buildings. However, in this vein it might be noted that the differences between the various structures on this site appear to be differences not simply of degree but of kind and that if we are to believe the contrary somewhat more forceful arguments need to be brought forward. He attaches some importance to the fact that they are aligned together, but is this really an adequate ground for assigning them equal status? It is of course possible for there to be considerable inequalities within a kinship system. Thus, a group linked through a common ancestor and living together might display such inequalities. However, if Smith intends such a scenario, several points might be made. Firstly, why all the talk of equal status? Secondly, the failure to explore these relationships leaves the notion of extended families, even if we take it on board, rather ephemeral: a bare starting point for more meaningful discussion. Finally, it is hard to see why sites like Montacute do not fall into the category so briefly described at the end of the paper where other buildings on a site are regarded as being subsidiary to the main house. Indeed, there is a widely held notion that such buildings housed farm labourers, a suggestion that seems on the face of it rather more convincing than Smith's ideas. Maillen-Al Sauveniere Among other sites with 'work halls' that are drawn into this schema is Maillen-Al Sauveniere (Fig. 11). Here a single extended porticus forms a unified facade behind which are two separate structures. Smith notes that the eastern one, a bipartite hall, is likely to have functioned as a work-hall. Fig;ure 11. Maillen-Al Sauveniere (redraum by Rob Rippengal). However, he suggests that 'because the two are joined by a corridor the dwelling aspect of the work-hall is likely to have been important' (ibid., 153). As an initial reaction, it might be countered that the separation of the two buildings behind their facade is perhaps as significant as the link it forms between them. However, even if we accept Smith's rather tenuous piece of logic, once again it is to be stressed that these are quite clearly buildings of very different character, of nothing approaching equivalent status and certainly unlikely as the separate dwellings of two parts of an extended family. From here, in a key extension of the argument, Smith goes on to suggest that: 'the social situation thus implied appears to have been embodied from the first in a few villas of unified architectural form' (ibid., 154). Towards illustrating this, he picks out as most obvious those sites where there are two entrances and where the obvious central approach is blocked. In particular, a number of sites have water shrines located where we might otherwise expect a central entrance. Given that there are sites, such as Rapsley, where water shrines were not so placed, it follows, as noted above, 'that the placing of a shrine just where it blocked the most obvious point of access was prompted by some profound social consideration'(ibid., 157), that, in other words, these buildings housed more than one domestic unit. Downton Of the sites used to illustrate this point, Downton (Fig. 12) is particularly interesting. Here the central room has no frontal access, 'so we are left with a hall common to two flanking houses' (ibid.). Ifwe could show this to be the case it would indeed be remarkable, for it comprises one relatively small and apparently coherent block. However, once again a more considered examination of the evidence would seem to leave Smith's interpretation rather lacking - it is all very well to suggest that there might be two houses here but he fails to develop this idea and demonstrate such a situation. Figure 12. Downton (redrawn by Rob Rippengal). In contrast to his claims there clearly is, for example, a central entrance into the porticus itself. Thus, although there may not have been a central access to the main 'hall', there clearly was a single main entrance that split to run into the two halves of the corridor. Equally, the notion of two houses leaves the central room, the only one with a mosaic and the largest in the house, in a kind of limbo, as a no-man's land between the two blocks, surely a rather unlikely proposition. Indeed, if we take this to be the main room, is it really so surprising that there should not be immediate access from the outside world? Such direct access may be one of the canons of classical architecture and it may be the norm in the more temperate climes of the classical world, but is it really to be expected in the rather colder climate of Britain? Indeed, it might be pointed out that the atrium, with which Smith is drawing a parallel, is not by any means an obvious equivalent to the central room of villas like Downton either in terms of form or function. Of the latter, the atrium is above all a public room and is therefore quite naturally linked to the outside, though not in fact directly but via the Jauces. In contrast, the private apartments, including the tablinium, are set rather deeper in the house. Regarding Downton, Smith's division would leave only the southern 'house' with either a hypocaust or baths. The baths, in particular, are not sited where they would be accessible from both halves of the building, but are offset at one end and the corridor that gives access to them provides access only for the southern half of the building, surely an unlikely arrangement if we are really to believe that there are two such carefully laid out 'houses' here. This is important, for it indicates that while there is a degree of symmetry in a strictly geometrical sense, there are not simply two equivalent units here. Given that mere symmetry is regarded as good reason to identify a unit system site, it cannot be too highly stressed that the failure to take account of the more detailed information is positively misleading. Nowhere is this more apparent than in the way in which Smith attempts to see the kind of units distinguished at Downton in a host of other sites, many of far larger and more elaborate design. As presented, these 'unit system' villas appear as empty shells. In contrast, if he used accompanying detail to demonstrate that they can be regarded as equivalent and discrete units in terms more meaningful than their geometry, his argument might be more convincing. Figure 13. Eccles (redrawn by Rob Rippengal). Eccles What is needed is at least one more elaborate example that could stand as a model from which to view the others. Of the examples given, Eccles (fig. 13) stands out, recently and comprehensively excavated. Thus, it might be possible to go further than simply indicating possible suites of rooms and actually demonstrate them to be complementary sets. In this vein, one thing that seems particularly striking about these sites is the question of scale. Eccles is an extremely large and complex site, quite clearly the product of considerable recources. However, if we compare it to Gayton Thorpe, where the different families at least live In sizeable 'houses', it seems odd that here its occupants effectively live in small apartments in the manner of a grandiose 'housing association', which somewhat belies the imposing nature of the site as a whole. Figure 14. Bradley Hill (redrawn by Rob Rippengal). Bradley Hill This is of course a rather superficial reaction. However, we might take it further, for if we compare the Eccles units to a site such as Bradley Hill (Fig. 14) the compalison becomes more telling. Ifwe look at the accommodation here, we seem to have units that would pass muster within a site like Eccles, yet these are sites of radically different character, wealth and status. It might be countered that this is simply an indication of the potency of the kinship relations involved. In answer, however, we might return to the fact that elsewhere the combination of wealth and status with kinship ties produces separate houses and certainly did not produce units that by their size and configuration suggest they could as well be occupied by the humblest of farmers. TOWARDS AN ALTERNATIVE Much has been said thus far that is rather negative. Although the examples could undoubtedly be taken further, a mort: immediate question becomes, what is the alternative? How may we look, not just as villas but at any houses in order to answer the kind of questions raised here? As a starting point, we might look to Heidegger, for he has suggested that dwelling is a fundamental human characteristic, that 'we do not dwell because we have built, but we build because we dwell, because we are dwellers.' Indeed, in pre-industrial societies most people will be conceived, born and die in houses; the house provides the focus of their daily lives both as a centre of domestic routine and as an important site for a range of productive tasks. Furthermore, most will build the house in which they dwell and as such their houses represent close indicators of how they feel this milieu of activities should be set and how they ought to proceed. In other words they are bound to what Bourdieu has termed a habitus (Bourdieu 1977). This is a key notion. On one level it refers to what we might term a world view, a distinct orientation 'in the world'. However, it is rather more dynamic than this, including a key concept of doing. Thus, it refers not so much to a world view as to the very way in which we live in the world - a competance or, as Bourdieu himself puts it, a modus operandi (ibid., 79). In attempting to understand this concept, material objects are seen, as Douglas and Isherwood point out, to have a very positive role; they are used 'to constitute an intelligible universe' (Douglas and Isherwood 1979). Thus, material culture represents far more than merely an embodiment of ideas and meanings but becomes an active force, an enabling element, a medium not merely for expression but for construction, working to produce reality, to order an otherwise chaotic world. This reification lends material culture a particular potency. We are all of us framed in meaning, set within a vivid and tangible matrix of signification. Indeed, we do not simply construct this matrix, but, by virtue of its essential durability, are in turn constructed through it. Thus it is that Bourdieu talks of 'the mind born of the world of objects', and here we might return to the house, for he notes of 'inhabited space' that 'through the ... divisions and hierarchies it sets up between things, persons and practices, this tangible classifying system continuously inculcates and reinforces the taxonomic principles underlying all the arbitrary provisions of ... culture' (Bourdieu 1977,89). In other words, the house literally represents an embodiment of the habitus. What is more, once built it acquires an enduring role in the direction of future practices and, thereby, in the very process of social reproduction. Thus, to return to Smith for a moment, houses or villas must of course relate to social structure. However this is but one element, for the meanings embodied in the house are infinitely richer than he allows for. Rather than the formulation 'house = social structure', the house is bound to a whole way of living, of 'being in the world'. Thus, although things like the placement of entrances or the division of space may relate to kinship, before we begin to make claims of this kind, we need to develop a rather more profound understanding of the particular houses involved, of the 'way of living' they embody. Towards developing such an understanding, we can perhaps begin by viewing the house first and foremost in its universal context, as a landmark, a manifest symbol of all that is cultural standing against the vastness of the world outside. Immediately, we have a link to the fundamental Levi-Straussian opposition of nature and culture and following it we might draw out a number of implications. The Latin domus is the stem for a whole series of words - words such as domicile and domestic, but also domain, dominion or dominate. In this reading, then, the house speaks to us as a symbol of ownership, mastery and power. To dwell, in this sense, is to have conquered, to have made the world one's own. Thus, the house as domus is a product of our will-topower wherein we impose ourselves on the world, locating ourselves within it, claiming possession and transforming the heretofore untamed into the means of habitation. In contrast to this image of human potency, we might draw out a rather different series of associations. Thus, the word house carries us back to the verge of history, to the Dark Ages and beyond. In Old English hus means simply shelter, after the Germanic huden, to hide, conceal or cover, and it comes down to us not simply as house but also as hut, huddle, hoard and husband. It speaks, in other words, of our essential vulnerability, of the fragile, exposed side of our nature as dwellers. We have, then, two contrasting and equally powerful images held and joined in a tension that is resolved in that most mundane of structures, the house. As described in etymological terms this relates specifically to our own culture. However, some negotiation of these two elements, of nature versus culture, of humans in nature versus humans set up over nature, is common to all houses, and if we return here to Roman Britain, an immediate potential for comparison is apparent. We might begin by contrasting, albeit somewhat crudely, the 'native'round house and the most common of villa forms, the 'winged corridor villa'. The aim is not to provide a definitive analysis but a starting point, for these two forms perhaps exemplify the distinction drawn out above. The classic round houses of Iron Age and post-conquest Britain seen, for example, at Whitton (Fig. 15), represent in many ways an essential acceptance of nature. It is often suggested that the circular form is of itself somehow natural or organic. However, without being committed to quite such a bold supposition here, it is suggested that their form, though undoubtedly the realisation of an ideal, was largely unconstrained by either rigid notions of geometry or of 'accuracy' and precision. The straight line, the regular angle and so on, those ideals of systematised construction, are entirely alien to this tradition of building. Figure 15. Whitton (redrawn by Rob Rippengal). Similarly, these structures were built almost entirely of natural materials: posts for the basic framework were generally full trunks or limbs left 'as they came' rather than square sawn or hewn; over this, wattle and daub walls, thatched roofs and trampled earthen floors all represent nature essentially untransformed. What is more, there is no attempt to hide these elements or the details of construction - they exemplify what, more recently, modernist architects have dubbed 'honesty'. Finally, by virtue of their use of materials these structures are rather impermanent marks on the landscape, rapidly and inevitably subject to all the natural processes of decay and degeneration. If we look next to the winged corridor villa, as exemplified by a site such as Newport (Fig. 16), we see a somewhat different picture. Smith places particular emphasis on the importance of classical influence without elaborating what it is that this expressed. Thus, it is here that we shall start, for in combining the Greek idealism of the classical orders with a typically heavy practicality, Roman building produced an architecture that expressed perfectly the unshakeable self-<:onfidence that sprung from a steady assurance of their place in the world. It is an architecture informed by a very definite concept of the relationship between man and nature: a belief that the world has a basic immutable order; that men by powers of reason can discover what that order is; and that, discovering it, they can control their environment as they will. Figure 16. Newport (redraum by Rob Rippengal). Outside official and public buildings, this is an architecture seen only in rather attenuated form in Roman Britain. Nevertheless, there was clearly exposure to it and it was certainly aped to some degree in the vernacular architecture of the period. Thus, while few villas produce clear expression of the classical orders per se, might they nevertheless indicate a similar shift in perceptions? On one level, such basic qualities as the straight line, the level or 'true'surface, the right angle and the use of regular units of measurement, together with their expression in a language of geometric forms, symmetry and so on speak immediately of rationality and order, and if we look further this is mirrored in the changing nature of construction and materials. Gone is the acceptance of nature taking its course, along with the 'organic' form, unworked materials and exposed construction that expressed it. Thus, what is immediately striking about these buildings is their solidity, their permanence - these are structures that might stand for centuries, strong, upright, impervious to all that the elements might throw at them. This solidity is exemplified in the use of stone. However, what is so not able about stone in this particular tradition is the transformation wrought upon it to produce, in the 'best' examples, carefully squared faces laid in regular courses. Similarly, the firing of clay to make brick and tile marks an even more striking transformation of the natural substance and contrasts directly with the earlier use of daub. Similarly, we see a host of new flooring materials. Whether concrete or mosaic, what is important is that the bare earth is hidden, shut away, the floor regular and true. So too the coating of walls; everything is hidden under an even, uniform covering of plaster. Even in the use of hypocausts to provide heat 'as if from nowhere', we see once again an isolation of the 'raw', 'natural' process of burning wood. In all, this new architecture, with its substantial, permanent structures, neat, square rooms, decorated floors, carefully painted walls and impressive, well-proportioned facades exemplifies not man in nature, but man's conquest of nature. It is, I would suggest, not simply a change in degrees of comfort or even of 'manners', but something altogether more profound, a semantic shift, a key transformation of the way in which tllese people saw themselves and, indeed, lived in the world. This is, of course, merely a starting point from which to develop a broader analysis. The aim of such an analysis would be to demonstrate how, by articulating a host of different forms of evidence, relating even to the most mundane of activities, we might hope to gain a fuller understanding of this way of living, of the habitus. Categories of Information To this end, I have drawn up two lists. The first gives, perhaps rather obviously, the kind of evidence available. The second is intended to illustrate how, by means of a comprehensive series of comparisons, the kind of articulation I referred to may be achieved: A. Categories of information available for examination: Site layout including position, orientation and plan of: 'main house', including features such as hearths, baths, hypocausts, mosaics, facade(s) and shrines; and the disposition of reception rooms, private quarters, bedrooms, kitchens and work rooms 'secondary' or other domestic units, including same features as 'main house' subsidiary buildings enclosures andj or yards other features - threshing floors, corn-dryers, hearths, pits, 'activity areas', boundaries and burials fields, roads and other outlying features Constructional details of walls floors roofs facade(s) doorways boundaries and enclosures Building materials: stone (including e.g. origins and whether dressed) marble brick tiles timber wattle and daub thatch plaster render floors in particular: flags, other paving, gravel, cobbles, beaten earth, rammed chalk, concrete, opus signinum, tesserae, tegulae, mosaic Decoration: painted plaster (inside or out) mosaics statuary Faunal remains: PotterY: species representation body part representation food residues 'economy' distribution fine/table-coarse ratio forms/uses fragmentation/trampling origins; distribution quantity Small finds: tools utensils toiletries weapons votive objects other exceptional small finds coms distribution/find-spots B. To shed light on (more formally): in:out front:back centre:periphery (up:down) (these including consideration of: boundaries, doors/thresholds, liminal areas, orientation, separation, inclusion, exclusion etc.) culture :nature male:female human:animal animal:plant clean:unclean sacred:profane fertile:infertile living:dead public:private (more difficult, though not necessarily impossible, e.g. day:night; light:dark) and relating to and cross-cutting these: dirt and hygeine including: personal cleanliness, especially related to the practice of bathing different categories of rubbish and other waste, such as domestic waste:farm waste; excrement:other rubbish/waste; human excrement: animal excrement; animal waste:plant waste processing of: animal products (e.g. butchery); grain (e.g. threshing, drying, milling); other (e.g. pottery production, metal working, woodworking) storage of: food; water; oil/wine; animals; grain; tools/utensils use of water. for drinking; for washing; sacred water/in shrines use of fire: for warmth; for cooking; for drying grain; for craft/industrial purposes burial of: animals; infants; adults decoration: of walls (in or out); of floors; of ceilings; furnishings activities such as: cooking, eating, sleeping, receiving guests, work (domestic:farm) The whole is to be related to such conventional indicators of wealth and status as: fittings, features, decorative and architectural. details; pottery; small finds; diet; economy; and contacts with the outside world. CONCLUSION Finally, and by way of a conclusion, it is important not to give an impression of uniformity here. The contrast between round houses and the winged corridor villa makes a point but is not the most meaningful com100 parison that could be made. The variety of domestic arrangements is quite striking and it is important to maintain this variety in our analyses for this is where Smith falls down. Thus, comparisons within and between round houses, simple one, two or three roomed houses, 'cottage' villas, winged corridor villas, courtyard villas, aisled buildings and even non-domestic sites such as shrines are central to the approach. Do they encompass the elements outlined above in the same ways; do they, in other words, express the same habitus? Thus, in relation to Smith's argument, I suggested earlier that we ought to be addressing the relationships between the various domestic units on many of the sites he discusses. In particular, I suggested that there is good reason to doubt his notion of equality, that there may be differences not simply of degree but of kind . Here, then, we have a means of assessing differences of 'kind', for in comparing different expressions of habitus there are inevitably implications for the social conditions that lie behind them for, as Bourdieu notes, 'all the products of a given agent, by an essential overdetermination, speak inseparably and simultaneously of his class - or, more precisely, his position in the social structure' (Bourdieu 1977,87). Bibliography Bourdieu, P. 1977 Outline of a Theory of Practice. Cambridge; Cambridge University Press. Brandt, R. and J. Slofstra (eds) Roman and Native in the Low Countries. Oxford. Branigan, K. and D. Miles (eds) 1986 The Economies of Romano-British Villas. Sheffield; Sheffield University, Department of Prehistory. Burnham, B. and H. Johnson 1979 'Introduction' in Burnham, B. and H. Johnson (eds) . Burnham, B. and H. Johnson 1979 Invasion and Response: The Case of Roman Britain. British Archaeological Reports, British Series; Oxford. Collingwood, R. G. 1923 The Archaeology of Roman Britain. Oxford. Collingwood, R. G. and J. N. L. Myers 1937 Roman Britain and the English Settlements. Oxford. Douglas, M. and B. Isherwood 1979 The World of Goods: towards an anthropology of consumption. London. Haverfield, F. 1912 The Romanisation of Roman Britain. Oxford. Hemp, W. J. and C. Gresham 1943 'Park, Llanfrothen and the unit system'Archaeologia Cambriens is 97, 98-112. Hingley, R. 1989 The Rural Settlement of Roman Britain. London. Miles, D. 1986 'Villas and variety: aspects of economy and society in the Upper Thames landscape' in K. Branigan and Miles (eds). Millet, M. 1990 The Romanisation of Britain. Oxford. Percival, J. 1976 The Roman Villa. London. Reece, R. 1987 My Roman Britain. Richmond, I. 1969 'The plans of Roman villas in Britain' in A. L. F. Rivet (ed.). Rivet, A. L. F. 1969 The Roman Villa in Britain. Slofstra, J. 1983 'An anthropological approach to the study of Romanisation processes' in R. Brandt and]. Slofstra (eds). Smith, J. T. 1978 'Villas as a key to social structure' in M. Todd (ed.). Smith, J. T. 1985 'Barnsley Park villa; its interpretation and implications'Oxford Journal of Archaeology 4 (3), 341-51. Smith, J. T . 1987 'The social structure of a Roman villa: Marshfield-lronmongers piece' Oxford Journal of Archaeology 6 (2), 243-55. Swoboda, K. M. 1919 Romische und TOmanische Palasta. Vienna. Todd, M. 1978 Studies in the Roman-British Villa. Leicester; Leicester University Press. Ward, J. 1911 Romano-British Buildings and Earthworks. Abbreviations in text: JRS Journal of Roman Studies. VCH Victoria County Histories. [/toggle]

Add a comment

TRAC is the organisation behind the annual Theoretical Roman Archaeology Conference, TRAC Workshops, TRAC Proceedings (1991-2016), and the TRAC Themes in Roman Archaeology series. This site is maintained by the TRAC Standing Committee. Contact us for further details or inquiries.

Follow us: Facebook | Google+ | Twitter

Copyright © 1991–2017 TRAC and individual authors. All rights reserved.