The book you are reading is an artefact. It is not quite a record of those conversations that took place in Cambridge in May 2015, in and around the formal conference session. It is not an exact record because the editors/organisers have worked hard – thanks are due to both – and also because many contributors, thinking back a year to the actual presentations and comparing them to the papers gathered here, have had second thoughts, strengthened some arguments, withdrawn others, expressed themselves with more nuance or determination, and in some cases have responded (as I do now) to both written versions and their own memories of the talks they heard on those two days. Conventionally we write about Conference Proceedings or Acts; but for a discussion in which agency loomed so large we should perhaps be up front about the fact that the proceedings have moved on, and our agency today is already producing different acta to those produced on the day. This artefact is not, of course, the only product of those discussions (even if is the most measurable output). Those of us who attended are also products, because in little ways and perhaps some large ones we exercise our agency (in writing, speaking, teaching, engaging in peer review etc.) in slightly different ways because of our co-presence at that meeting. The Laurence Seminar was a node, a moment in time at which our trajectories intersected, and what came out the other end – people and papers – were transformed by the encounter. All good conferences should do this, but not all are as successful as this meeting was.
My opening is not intended as a modish exercise in reflectivity, but to introduce my main theme, the relationship between the experience of people and that of things over time. Our focus was on artefacts, and several papers allude to the ‘material turn’: but people lurked constantly in the wings. Even the commitment to go beyond instrumentalism and representation evokes those who used objects as instruments, and that special form of instrumentality in which objects were used by one person (maker, user, giver) to represent one immaterial thing to another person (recipient, user or viewer). No paper spends much time on objects circulating in unpeopled worlds, despite the fact that most archaeologically preserved objects have in fact spent part of their cultural biography out of touch with society. Our material turn gives us a different way into conversations about people and things: it does not take us away from people.
The term ‘representation’ is used in this volume in a range of ways but essentially it deals with two families of relationships. The first family is located firmly in the past, evoked in discussions of how artefacts were used, both deliberately and with less conscious intent, to mediate relations between different human actors. Sometimes a gift of an object materially changed a relationship, as in the presentation of an auxiliary diploma to mark the grant of citizenship, or the dedication of a childhood doll to mark the transition to womanhood. This is a clear instrumental use of specific items of material culture, a use of things to change people. A vast ethnography and a mass of archaeological theory informs our understanding of this species of deliberate and customary artefact-use, and it has been incorporated into Roman studies for a long time. More often we have been discussing how the use of a particular item of clothing or tableware signified adherence to a particular set of social values. Mingled with these issues are questions of intent and self-consciousness about the use of artefacts: what was going through the mind of an adult Roman subject when he put on a toga for a particular occasion? That sort of question is in the end unanswerable, but several papers here make use of ideas like Bourdieu’s habitus to describe the problem. These questions also have a long currency in archaeological theory, especially in the debate over the communicative functions of style, as well as in relation to a proposed ‘ontological turn’.
The second family of representational relationships are those between modern researchers and relics of the Roman past. Specifically, these papers discuss the way we are accustomed to use a particular set of artefacts as proxies for particular identities (Roman, British, military statues etc.) or else as indicators of phenomena such as literacy. As the editors make clear, there can be no real objection to this general procedure. Natural scientists resort to proxy data all the time, as when exoplanets are detected by their effects on the light coming from the stars they orbit or when the pace of global warming is inferred from directly measurable phenomena such as changes in sea levels or the retreat of glaciers. Some applications have turned out to be misleading, like the culture-historical equation of particular kinds of material culture with particular ethnic groups. But there is nothing wrong in principle with asking what particular artefact classes (or their use patterns) represent. Ceramic studies naturally loom large, but other artefact types are brought into the debate. The questions posed are related to larger ones about inference and what might constitute an interpretative archaeology.
What is common to the analysis of both these families of relationships, is that artefacts are treated as secondary to humans, as passive objects in the control of humans, rather than as subjects possessing their own agency. Most of the contributors to this volume are engaged to some degree or another in a critique of such views. Most are agreed, I think, that artefacts do have some degree of agency, that they have effects beyond those intended or programmed by their human makers and users, and that the properties of artefacts include affordances that elicit, facilitate or stimulate particular responses on the part of humans who encounter them.
That agency in the Roman case turns out to be both more various, and less predictable than is sometimes emphasised. The centuries before and after the turn of the millennium saw the production of many new kinds of artefact, and also some new kinds of material. Just as in the industrialised world an explosion of new materials made possible much more than their inventors could have intended. The history of early imperial architecture is very largely a working out of the new possibilities created by access to new materials. A number of papers take care to assert that representational approaches of one kind or another still have some utility. Historians of Roman art are still trying to assimilate the impact of arguments about the agency exercised by works of art, and the boundary between a work of art and an artefact is a difficult one to establish. Many things we would not consider artworks clearly played a representational role. But then as now objects impacted on people in many other ways. Put bluntly, there is more to Roman things than what they meant or even what they were meant to mean.
What follows if all this is correct? That question returns us again to the practice of archaeology. It makes a difference whether one is curating or studying a single object, considering a category of artefacts (and several contributors have much to say about categories and taxonomies), or whether one is trying to write history from artefacts, as the editors encouraged us to do. The remainder of my comments relate to this last topic, since it is very much of the moment.
Several suggestions have been made about how we might envisage the big picture, and generate macro-scale narratives about ancient Rome that do not begin and end with human agency. One of these ideas is the notion of a ‘consumer revolution’, perhaps beginning in the second and last centuries BC, and attested by the proxy data of a vast expansion in the quantity of Roman-style manufactures, most produced first in central Italy, and then in other Italian and provincial locations. Some of us have found this an attractive formulation, one originally borrowed from studies of early modern Europe (Woolf 1998; Wallace-Hadrill 2008). But it is of course a perfect example of an account in which human agency does all the work, from stimulating the demand in the first place to developing a series of production and distribution systems to satisfy it. If that approach is to be saved, we need to ask what certain kinds of artefacts did to make human agents desire them. What were the ‘technologies of enchantment’ (Garrow and Gosden 2012) possessed by some Roman-style goods? Gosden (2001) developed the idea of a sensorium, a material environment which elicits particular desires and habits of behaviour: it draws on the work of the art historian W.J. Mitchell (1996) and the anthropologist Alfred Gell (1998) and coheres well with ideas about the extended mind, symbolic storage and attempts by Giddens and Bourdieu to reconcile individual agency with the predictability of human behaviour en masse. The relationship between human and material agencies is recursive and dialectical. ‘We build the house and the house builds us’, as it is sometimes put. Miguel John Versluys proposes in this volume the idea of an ‘object-scape’ to describe a dispersed class of objects that brings with it a series of human-thing entanglements. Martin Pitts road tests this idea in relation to Gallo-Belgic pottery, with thought-provoking results.
The potential of many of these ideas remains to be worked through, but it is possible to imagine at least one objection to all of them. Implicit in all these formulations (and others such as entanglement theory) is the sharp dichotomy drawn between people and things. Perhaps the best way to approach this is through examining the category ‘artefact’. Romanists rarely have difficulty identifying artefacts since most of their characteristics are the results of processes of manufacture that completely transform the raw material. Objects made of iron, glass, concrete or ceramic could not have come about without transformations that only humans could organise. Even materials that were not physically transformed to a significant degree, such as granites, basalt and marbles, are usually found as parts of complex objects which have no possible natural origin. For what it is worth both Renaissance and ancient ideas about manufacture stressed the transformative power of human agency. But the question of what an artefact is, is not always so clear. Archaeologists often speak of biofacts or ecofacts, meaning organic finds in archaeological assemblages that have not been converted into tools or furniture, yet are relics of human collection, agriculture or digestion. Primate archaeology raises even more sharply the question of when a stick or stone used as a tool becomes an artefact. And once we populate the Roman sensorium or object-scape with plants and animals, where will we draw the line? Does stock-breeding turn an animal into an artefact, or must that wait for butchery? Gardens are arguably as complex artefacts as atria. If a cultivated landscape is an artefact, is an orchard an object-scape?
Perhaps most crucial is our habitual reluctance to treat human bodies as artefacts. Yet when one of Rome’s subjects begins to depilate or shave, to grow or cut hair into a particular style, to abandon ancestral practices of tattooing, or when a Roman brands his or her slave, is there really such a difference from transforming other biofacts into artefacts? It would be convenient too if this stopped with the materiality of the body, or with its external surfaces and organs. But it is common enough now to discuss self-fashioning (in Greenblatt’s phrase) or le souci de soi, Foucault’s ‘care of the self’. And if we, and our animals and plants and homes and dreams and ambitions are all artefacts, how can we justify a distinction between human and material agency? The material turn, in other words, risks perpetuating Romantic and anthropocentric ideals about the separation of man from nature, and perhaps also a sense that the essence of being human is an immaterial thing, whether we call it soul or self. Yet our bodies’ material components have affordances as well, and they have an agency that is not entirely under our conscious or deliberate control.
Much of this has little special applicability to the Roman world. But the huge geographical and chronological extent of that world, and the fact we know its people and other artefacts so well, makes it an excellent place to try out some of these ideas. And it does have its peculiarities. As several essays suggest, one of these peculiarities is the appearance over the last centuries BC of relatively high levels of standardisation in many categories of object. There are many early precedents for this sort of phenomenon from the Uruk phenomenon of fourth millennium BC Mesopotamia to the spread of Linearbandkeramik in the early European Neolithic. But Roman standardisation is more comprehensive – more materials and artefact types were involved – and easier to contextualise. At a stylistic level that standardisation diminishes gradually over the first centuries AD but in many media – tableware for instance, and domesticated animals – there remained recognisable family resemblances well into the middle ages. It is as if across the vast area controlled by Rome patterns of resonance began to intensify over the last centuries BC, and then to diminish into more regionally fragmented patterns.
Patterning like this is observable because we can study manufactures en masse in a way that is rarely possible. But we lack good analytical tools for examining the phenomenon. Wengrow’s (2014) interest in the implications of mechanical reproduction might contribute to this picture, but most approaches are more easily applied to single objects or classes of complex artefacts with few members. I am thinking of Gell’s approach to style, in which an artefact category becomes an ‘object’ and particular artefacts ‘indices’ of it, and Kopytoff’s (1986) notion of cultural biographies which allow us to describe the changing social valences of particular artefacts from their creation to their deposition. These approaches seem better suited to studying artefact categories such as archaic Greek temples or ‘Celtic’ metalwork than to Roman period brooches or tableware.
If we are to examine agency in such a way to include humans and animals and plants within our super-category of Roman period artefacts – in what some might call a symmetrical archaeology (Witmore 2007) – we do not need a cultural biography or artefact biographies so much as a cultural ecology. We might imagine the Roman world as a vast ecosystem in which different categories of actors play the role of species, each with its own agency. The key questions then become, as in the study of any ecosystem, how are its components related, how do the dynamics of competition, predation, symbiosis, and reproduction intersect to sustain or transform the order of the system?
Put this way, it is immediately apparent that the material world of the Roman period was a relatively stable system. The main species of artefact, of animal and plant and human hardly changed between the second century BC and the sixth AD or even later. There were some technological innovations of course. But they hardly disrupted the cultural ecology in the way that earlier innovations had done, innovations such as the domestication of the horse, the generalisation of iron working, or maritime transport. When the Arab armies of conquest took over half the Mediterranean basin in the seventh century they found more or less the same techniques of agriculture, building materials, industrial techniques, and domesticated species as those that the Romans had found in their own period of expansion around a millennium before. Compared to the thousand years before that expansion (from 1300–300 BC, say) or to the thousand years after (700–1700 AD), the Roman millennium was relatively stable. Stability is not necessarily a good thing of course, and nor was the standardisation which several essays discuss. In ecological terms standardisation represents a loss of diversity, a thinning of reproductive options, a narrowing of the range of affordances. So, in a different way does the solidification of stylistic canons, the classicisation of particular literary and aesthetic motifs, the entrenchment of the power of particular interest groups. Low diversity systems can endure for long periods, but they are not particularly resilient.
Conventional accounts of the endurance of the Roman order and of its collapse still focus on human agency, that of individuals and that of institutions. Most narratives of decline and fall assume that survival and disaster alike are to be attributed to the conduct of the political, fiscal, and military elite. Changes in Roman material culture are most often seen as consequential to political change, rather than as bound up in it. Once again people are attributed agency, and neither the agency nor the affordances of material culture play much of a part in the story. Yet in many ecosystems those species at the top of the food chain are the most vulnerable to change, and disruption rarely comes from the top. Perhaps we might understand long term trajectories of growth, stability, and decline better if we began from lower down, from Roman things rather than Roman people. Were the standardised forms that emerged and were so widely disseminated in the last centuries BC in some way implicated in the stability of the period that followed? Did they establish a sensorium that elicited a particularly limited range of responses from human agents? Giddens explored the contribution of spaces to the routinisation of behaviour. Might the agency of Roman things have been the very opposite of dynamic? Change happened, of course, but it was slow and it was incremental rather than discontinuous and rapid. What role did the stability of Roman artefacts have in the stability of Roman society?
As we move beyond representation and instrumentality we take a great weight of responsibility off the shoulders of Roman actors. Humans are no longer the sole drivers of Roman success and Roman failure. Romans put human virtue and customs at the heart of their historical causation, but we do not need to do the same. Neither expansion nor collapse is to be explained primarily in terms of human values, ideologies, beliefs or motivations. And in fact, some of the political and military failures seem less cataclysmic when we consider how much of the Roman material order persisted under new management. Taking things seriously allows us to put people in perspective.
* Institute of Classical Studies, University of London.