Contemporaneity in the archaeological record: a critique of synchronism

One of the most important goals of archaeological methodology since the 19th century has been the control of chronology, specifically the ability to date sites, structures or finds as closely as possible. There is no doubt that the advent in the 20th century of scientific dating techniques such as dendrochronology and radiocarbon have yielded inestimable benefits for the discipline and the current developments of Bayesian analysis for radiocarbon dates are no exception (e.g. see Aitken Reference Aitken1990; Nash Reference Nash2000). However, I would like to ask us to pause and consider for a moment why we should seek tighter and tighter chronological resolution. It seems so self-evident that we do not even question it. But self-evident truths are often the most dangerous. To help articulate this question, let us go back to the first radiocarbon revolution, which was so eloquently narrated in Colin Renfrew's book Before civilization: C14 suddenly stretched our chronologies, and in Europe especially for the Neolithic (Renfrew Reference Renfrew1978). What this meant was that sites which might have been considered contemporary were now were revealed to be 500 years or more apart. It changed the understanding of prehistory immensely – especially many of the then current orthodoxies about diffusion. The Bayesian revolution of more recent times also expands our chronologies, but not so much outward as inward, in terms of the resolution (e.g. see Baillie Reference Baillie1997; Buck and Millard Reference Buck and Millard2004). Sites that were previously considered contemporary are now recognized as being 50 years or more apart. The concern for controlling time in archaeology is thus arguably about distinguishing the contemporary from the non-contemporary – and, as a consequence, the later from the earlier or vice versa.

In many ways, establishing contemporaneity has been a more important – and urgent – problem for archaeologists than establishing succession or sequence. Childe, writing on the eve of the radiocarbon revolution, argued that the most pressing concern of the day was tying together regional sequences by means of what he called synchronism (Childe Reference Childe1956, 104–10); that is, establishing which part of one sequence was contemporary with which part of another, and so on. The radiocarbon revolution certainly made this possible in a way that cross-dating could never achieve with any reliability. What Childe did not foresee, of course, was the degree to which European prehistoric sequences, previously considered broadly synchronous, were in fact off by hundreds, if not thousands, of years (Renfrew Reference Renfrew1978). But methods of absolute dating are only one means that archaeologists use to establish contemporaneity. The most common has always been typology – that is, using finds of the same type to infer broad contemporaneity. We do this both within a site to establish contemporaneity between features or layers but also, of course, between sites. Indeed, prior to the radiocarbon revolution, cross-dating using find types was the main method of linking separate sequences.

Within sites, establishing contemporaneity is actually something one cannot do using stratigraphy; stratigraphy only gives sequence or succession, even if it can bracket or constrain contemporaneity. Contemporaneity is something one has to infer, either using similar methods to those used between sites, or from morphological aspects of layers and features such as the compositional similarity of deposits or spatial alignment of features (such as rows of postholes; e.g. Carver Reference Carver2009, 281–87). Indeed, on poorly stratified sites – a rather more common occurrence than most field manuals like to admit, especially in arable or deeply ploughed areas – contemporaneity becomes the most important tool one has to establish sequence, for example by grouping features with similar fills or those on a similar alignment, and then using time-diagnostic finds or radiocarbon dates to order these groups into a sequence. Seriation, one of the oldest archaeological dating techniques, uses solely typological methods with frequencies and/or co-occurrence to create sequence; it is a method which essentially works by separating contemporary from non-contemporary material. Indeed, methods which start from the problem of dissecting the contemporary from the non-contemporary form a central part of archaeological inference in various aspects of fieldwork, although they are often marginalized in conventional manuals (see Lucas, forthcoming, for further discussion of this).

Despite the emphasis conventionally placed on establishing succession, it seems to me that Childe was as astute as usual; the real problem, the real challenge in many ways, facing archaeologists is not working out succession but establishing contemporaneity, or what Childe called synchronism. Such an emphasis does not eschew sequence, but rather sequence becomes derivative or secondary to contemporaneity. Indeed, arguably all the attempts to improve our chronological resolution, as suggested above, are really about increasing our ability to distinguish the contemporary from the non-contemporary. It is not enough to know these two burials date to the 9th century B.C.; we want to know which decade or quarter of that century. The quest for tighter resolution is the quest to refine what we mean when we say that two sites or features are contemporary (or not). And this leads me on to the critical question I want to pose about contemporaneity. What do we mean when we say that two things are contemporary?

The Oxford English dictionary defines contemporaneity as something which belongs to a certain time or period; when archaeologists talk about contemporaneity in the context of dating, they usually mean something which belongs to the same time as something else, where the term ‘time’ here is taken to mean a certain period or span of years. These definitions are significant because it is clear that contemporaneity is conventionally and broadly understood as a relation to a temporal period: two objects are contemporary because they both date to the early 16th century, for example. It effectively tries to reduce contemporaneity to an instance of synchronism.

Here, the term Childe used for contemporaneity – synchronism – is highly apposite. Synchronism implies the temporal alignment of two things, like two dancers moving in tandem; the implication is one of perfect alignment or symmetry, such that when one is even slightly out of step, we say they are ‘out of sync’. Implicitly, I would suggest that this is how we routinely use the concept of contemporaneity in archaeology; when we say that two objects or sites are contemporary, we mean that they are ‘in sync’. Yet we know that this is actually rarely the case when we are talking about the temporal relation between two objects; indeed, in some of our oldest techniques, like seriation or find combination, lack of synchronism is actually a necessity. Types have to overlap in order to draw out a sequence. This is an explicitly non-synchronous idea of contemporaneity. Another way of saying this is that contemporaneity is not a transitive relation: just because A is contemporary with B and B is contemporary with C, it does not follow that A and C are contemporary. Transitivity only applies if the temporalities of things all align perfectly, as is implied under synchronism.

Non-synchronism is also recognized in other ways archaeologically. For example, objects in archaeological contexts can be contemporary in some ways and not in others. Two objects might have been made 50 years apart and thus be non-contemporary in terms of manufacture, but have been deposited at the same moment and thus indeed contemporary. This is, of course, something that has been discussed widely by archaeologists, not least in Scandinavia in relation to the definition of closed or secure find combinations from Worsaae and Montelius to the present day, and is understood as critical to correct dating (e.g. Rowe Reference Rowe1962; Gräslund Reference Gräslund1987).

What both of these examples illustrate – examples of some of the earliest archaeological methods in chronology – is a very different concept of contemporaneity to the OED definition. The difference comes down to a very simple point. In the cases of seriation and find combination, contemporaneity is defined as a relation between objects, which is made possible because objects are acknowledged as having variously extended lives. This is perhaps why it is also easier to apply a non-synchronous model of contemporaneity to processes over events: processes such as the emergence of agriculture take place over long periods of time and it is much easier to acknowledge overlap than with, say, a short-term event like a burial. Defining contemporaneity of two burials will perhaps always be done in terms of synchronism, not only because of the limits of our dating resolution, but also because of the contrived separation of the punctual nature of the event against the protracted character of process.

But whether one is talking about events, processes or objects, chronology should always remain purely a scale of measurement here for coordinating the relations between these things. Two objects (or events) are not contemporary because they both date to the late 4th century B.C.; they are contemporary because their life spans overlap or imbricate. Chronology is solely used as a means to establish this overlap (or not). In the case of absolute dating and its tacit synchronism, however, contemporaneity is implicitly defined in relation to a period or unit of time. It is what you might call an enveloping concept of contemporaneity, where sites or objects are forced to fit into pre-given time ranges. The smaller you can make these ranges (i.e. increase resolution), the more precise your ascription of contemporaneity. I would argue that we too easily fall into the second – and arguably inadequate – definition of contemporaneity because of the dominance of chronology as our model of time.

One of the advantages of adopting the first definition, i.e. making contemporaneity a relation between objects, is that it forces us to ask about the meaning of contemporaneity in terms of the longevity of the objects being related. This means that the question of chronological resolution has to be made relative to the temporality of the objects under investigation. Asking, for example, whether two sites are contemporary, where both sites’ occupations span several centuries, why would we need a resolution tighter than a century? Obviously there are many objects whose longevity falls below the threshold of our current dating techniques so I am certainly not arguing that we abandon the goal of refining our chronological tools. My argument is rather that we should not always assume that tighter resolution is intrinsically better; it has to be relative to the temporality of the objects under investigation.

Another advantage of making contemporaneity a relation between objects is that it allows overlap or imbrication, a property which has long been exploited in archaeological techniques such as seriation and find combination. However, this property has subsequently been marginalized in more general interpretive accounts which favour a more segmented view of time. It is no great revelation to state that the basic model of time used in archaeology is linear; whether one is dealing with stratigraphic sequences, radiocarbon dates linked to our calendrical system, or simply period divisions, time is perceived as flowing in one direction (see e.g. Lucas Reference Lucas2005). One of the problems with this idea of linearity is that temporal succession or sequence is conceived as a series of intervals or points derived from divisions of a line. The line is primary, the intervals are secondary or derivative. As a result, there can be no gaps between them and no overlap either. However large or small the segments into which you cut the line (e.g. years, decades, centuries), each segment runs tight up against those on either side. This is the form of the classic timescales we use, both clock time and calendrical time. This model of time consequently dictates the range of temporal relations we employ, of which there are only three: before/earlier, after/later and contemporary – where contemporaneity is, of course, essentially synchronism. These concepts are enshrined in dating tools such as stratigraphy or terminus post quem (TPQ) and terminus ante quem (TAQ). What other tools, like seriation, demonstrate is the paucity of this triad and the need for a more sophisticated model of articulating the temporality of objects and especially the concept of contemporaneity.

There is actually a very useful model already in existence which has only recently been explored by archaeologists working in computing (Binding Reference Binding, Aroyo, Antoniou, Hyvönen, Ten Teije, Stuckenschmidt, Cabral and Tudorache2010). It is a time-interval algorithm developed in the context of research into artificial intelligence by James Allen (Reference Allen1983) in the 1980s and it proposes thirteen different temporal relations (what are now called Allen operators) as opposed to the three most currently in use in archaeology (figure 1). Immediately, one sees how the idea of overlap or imbrication actually covers multiple possibilities and varieties, and some work employing these algorithms has already been explored as part of an English Heritage-funded project on digital technologies called STAR (Semantic Technologies for Archaeological resources; see Binding Reference Binding, Aroyo, Antoniou, Hyvönen, Ten Teije, Stuckenschmidt, Cabral and Tudorache2010). It seems to me that their potential application for developing more sophisticated chronologies in archaeology is quite high. To give just one obvious example, using the full panoply of Allen operators instead of the usual three temporal relations in constructing stratigraphic matrices would enable a much more sophisticated model of site development – and most of the complexity will revolve around the variations of imbrication.

Figure 1

The main temporal relationships under Allen's temporal logic; six of them can be inverted, resulting in a total of thirteen possible relations (source: author, after Allen 1981).

In summary, then, the linear model of time we use has essentially driven us to adopt a very impoverished concept of contemporaneity, one which effectively is reducible to synchronism because contemporaneity is defined as a relation to a unit of time. In this section, I have argued for foregrounding an alternative concept, one which already has a deep pedigree in archaeological method but which has been marginalized, the idea of contemporaneity as a relation between objects. If we draw on this idea, issues of dating and chronological resolution become points to argue for rather than to take for granted; they become relative to our interpretation of objects, rather than vice versa. In the next section, I want to explore another dimension of this issue and I will argue that the same impoverished model of contemporaneity as synchronism has affected the way we look at the archaeological record.

Contemporaneity of the archaeological record: a critique of anachronism

It is almost a truism to say that the archaeological record is a contemporary phenomenon. It has been articulated by Binford (Reference Binford1983) as an epistemic problem and by Shanks (Reference Shanks2012) as the ontological basis for archaeology: we do not have access to the past, but only to what remains of the past in the present. This is a point that has been long appreciated, but its understanding has changed – and changed dramatically, I believe – since the 19th century. Today, the emphasis is given to the gap or distance between present and past, but in the 19th century there was no gap – the past was present in the present. Today, the contemporary nature of the archaeological record is often considered a problem to be solved. For antiquarians, its contemporariness was, in contrast, what made a study of the past possible in the first place. I want to spend some time exploring this distinction, and to do it by connecting the concept of contemporaneity with another related term – anachronism.

Various words have been used to characterize the nature of the archaeological record: ‘remains’, ‘vestiges’, ‘ruins’, ‘relics’, ‘traces’ and ‘fragments’. The list could go on, and while many terms are still current, some have a certain antiquarian ring to them (e.g. ‘vestiges’). What is striking, however, about these antiquarian terms is that they foreground an aspect of the archaeological record that was of great theoretical concern in the 19th century, but one we completely bypass today. What characterizes the terms ‘relic’ or ‘vestige’ is their untimeliness: they are from another era, anachronisms, survivals. Indeed it is no coincidence to observe in texts from the mid- and late 19th century the close relationship that existed between material remains and what came to be called survivals; that is, extant customs or practices which no longer have any obvious purpose, like vestigial organs (see Lucas Reference Lucas2012 for an extended discussion of this issue). In fact, up to the middle of the 19th century, the word ‘antiquities’ was commonly used to refer to both; it was only with the crystallization and separation of academic disciplines in the last quarter of the 19th century that such differences also started to emerge, and with them a loss of sensibility to issues such as the temporality of the archaeological record. It is very instructive to read Edward Tylor's work on survivals in this regard, because he quite explicitly saw survivals and relics (terms Tylor himself used) as the twin pillars of the study of the past (Tylor Reference Tylor1865).

What is important about this characterization of the archaeological record is that it is very much defined in relation to the contemporary; what makes relics and survivals significant is their non-contemporariness to other objects or customs. Their status as anachronisms formed the very basis and possibility of studying the past. However, it is also interesting to see how this concept of anachronism changed orientation at the end of the 19th century and into the early 20th – a change that was very much aligned with the consolidation or professionalization of the academic disciplines of history and archaeology. For with the advent of professionalization came a new concern for systematizing knowledge. This period saw the growth of publications on historical and archaeological methodology, and with it a growing sense of the impact of the scholar on interpretations of the past. Issues of objectivity and the removal of subjectivity, expressed through concepts such as historical distance and hindsight and the curse of presentism (or the fallacy of nunc pro tunc), became central to such methodological reflections (e.g. Ashplant and Wilson Reference Ashplant and Wilson1988a; Reference Ashplant and Wilson1988b). In short, anachronism was no longer about objects or practices from the past, occurring in the present, but rather about the historian or archaeologist projecting their conceptual frameworks or agendas into the past. The arrow of anachronism was reversed, so to speak.

In archaeology, this epistemic concern mostly crystallized around the role of ethnographic analogy. Where relics and survivals had worked alongside each other, survivals now morphed into something else for archaeologists. Lubbock and Sollas's works were written just as this change was happening. Australian Aborigines were both survivals from and analogies for the Palaeolithic. There is a definite ambiguity about which. Over the course of the 20th century, the role of historic and contemporary peoples and material culture as analogies for interpreting the deeper past became central to bridging the gap between past and present. Such a role became consolidated through middle-range research and middle-range theory from the 1970s. However it was expressed, though, the central problem of analogy has always been the same: the danger of making the past too much like the present (Wylie Reference Wylie1985). One solution, of course, is to embrace this through the principle of uniformitarianism: the past is like the present, because the same basic processes are operating now as they did then. This may work for geology, but archaeologists differ greatly on how applicable such a principle is to human society or culture, although it is interesting to note a recent resurgence in this issue in the context of cross-cultural universals (e.g. see Lloyd Reference Lloyd2010; Shyrock and Smail Reference Shyrock and Smail2011).

In summary, then, I would suggest that in the 19th century, the archaeological record itself was the anachronism. Today, it is our own subjectivity, imposed on interpretations of the past, that is anachronistic, which in the context of archaeology is principally expressed though the role of analogy. At the same time, though, we do not tend to worry much about this either any more; indeed, we accept subjectivity and analogy as often necessary elements in the interpretive process, a hermeneutic fusion of horizons. Indeed, we accept that the past is always a contemporary construct – which does not of course mean that it did not happen or that we pluck our interpretations out of thin air. But analogies and the danger of presentism should not be given too much epistemic weight. The playful status of analogy today in archaeology is perhaps well captured in a genre of contemporary images which work with the idea of anachronisms (figure 2). These images are interesting, because in many ways they are negatives of the way 19th-century antiquarians may have viewed past objects. Such modern images perhaps recapture for us how 19th-century antiquarians may have felt about encountering remains from the past – the same sense of matter out of place, temporally speaking. We do not feel like this about archaeology any more. Perhaps we never did, but it is a provocative thought nonetheless because it reminds us of the theoretical status that anachronism once held for our discipline.

Figure 2

Abraham Lincoln with a boom box; this image was posted widely on the Internet in 2011 (original source unknown).

Despite this, I am uncertain whether anachronism itself is a very useful idea. The crux of the concept refers to an object or custom or way of thought which is out of its proper time. And this raises the question of what constitutes the ‘proper’ time for an object or thought. No doubt archaeology itself has helped to create the perceived proper temporal order for things, but more generally this is probably the legacy of modernist thinking insofar as the trope of modernity defines the very possibility of something being untimely. And this brings us back to the idea of contemporaneity as a relation to a period of time, which I discussed earlier. What makes something untimely is an idea of contemporaneity which works off the notion of the present as an era or period. For 19th-century antiquarians, archaeological remains were non-contemporary because they did not belong to the present – that is, the modern era. The present, as distinct from the past, and the modern as distinct from the ancient or premodern, are the classic chronoschisms of modernity, but they also peform the same function as any periodization: they make contemporaneity a question of belonging to a certain unit or stretch of time. This is exactly the same chronotype being played out here, as with my earlier discussion of chronology.

By the same logic, though, we can rethink contemporaneity in other terms. More precisely, we can adopt a concept of contemporaneity where contemporaneity is principally about the temporal relation between things, rather than between things and an abstract measure of time. This is the challenge: to think about contemporaneity without drawing on the tropes of synchronism or anachronism. When we say that the archaeological record is a contemporary phenomenon, what we really mean is that it is a phenomenon of our present time, where the present time is conceived as a temporal envelope, defined by synchronicity and distinguished from the past (or future). But if we eschew this conception, then how should we conceptualize contemporaneity? We could use the same terminology I developed in the previous section, of imbrication or overlapping and even characterize it through those Allen operators. But somehow the stakes appear to change once one of the objects involved in a temporal relation is oneself. The contemporaneity of the present is more than the contemporaneity of two things overlapping for a given interval or unit of time. The difference here is that contemporaneity in the former needs to be articulated through a tensed conception of time (i.e. past–present–future) as opposed to a purely successional one (earlier–later) – that is, it needs to articulate a concept of the present but in such a way that it does not reduce it to a time period or punctum in a successional view of time. This difference is especially felt in discussions regarding the status of archaeologies of the contemporary past, as Harrison (Reference Harrison2011) has recently alluded to, where the concept of contemporaneity is deeply problematic.

This distinction can be related to a rather old philosophical discussion by McTaggart at the start of the 20th century, where he contrasted two types of time series, A (past–present–future) and B (earlier–later), and argued that it was impossible to reduce one to the other (McTaggart Reference McTaggart1908; also see Lucas Reference Lucas2005). Part of the problem with McTaggart's analysis, however, lies in his attributing a serial or successional nature in the first place to the tensed (A series) view of time. When it comes to talking about the past, present and future, I would suggest that we have to dispense with any notion of succession altogether. In order to see how this might affect our understanding of contemporaneity, let me develop this point through an analogy. A few years ago, I had a discussion with an eminent archaeologist about two types of handkerchief user – the folder and the scruncher, depending on how you retained the handkerchief in your pocket for use. The folder always has the handkerchief neatly folded over on itself in a flattened square or rectangle, while the scruncher has a messy ball. Our conversation did not shift into one about time, but the distinction is an interesting one to use when the two sorts of handkerchief are viewed as representing two different chronotypes. In the folded handkerchief, time is successional through the neat layers of the folded handkerchief. What belongs in one fold belongs there and nowhere else. This object is Neolithic, not Bronze Age or Iron Age. With the scrunched handkerchief, time is messy and any two points on it can touch. An object made in the Neolithic can also irrupt into the Iron Age or in fact our own present.

The analogy of the scrunched handkerchief is actually a famous metaphor used by the French philosopher Michel Serres in his conversation about time and contemporaneity with Bruno Latour (Serres and Latour Reference Serres and Latour1995, 60), and one which has often been quoted in archaeology (Witmore Reference Witmore2006; Holtorf Reference Holtorf2002). With the folded-handkerchief chronotype, contemporaneity is clearly defined in relation to an era or period – a fold in the handkerchief. With the scrunched handkerchief, however, contemporaneity is defined by the particular relation between any two or more points on the fabric. It is a point of contact. Serres has also used another metaphor, that of percolation, where time not only folds back on itself, but also is filtered as through a sieve (Serres and Latour Reference Serres and Latour1995, 58). Some events or objects persist in their effects while others cease. These might sound like rather vague metaphors and indeed they perhaps are – but that is also because we still lack an adequate vocabulary to articulate this new conception of time. This is what Serres was searching for. The idea of succession seems more concrete because it has been made so through centuries of use. The idea of percolation seems abstract and vague because it is still an idea; it needs working through. Such work has already begun (e.g. Witmore Reference Witmore2006; Reference Witmore2009; Reference Witmore and González-Ruibal2013), for in reflecting on the archaeological record, percolation seems an eminently appropriate concept: some objects from the past irrupt into the present, others slip away forever, others simply wait or pause – they may or may not re-emerge.

Fundamentally, I would suggest that a central concept here is that of persistence, an issue which preoccupied other philosophers of time at the start of the 20th century, especially Bergson (Reference Bergson1991) and Husserl (Reference Husserl1966). In a way, persistence returns us to the antiquarian notion of the relic and the very possibility of archaeology: the persistence of the past in the present. I believe that this is a very positive idea that we simply have forgotten or take for granted today. What I would do, however, is drop the idea of anachronism which accompanied this notion for antiquarians; our task is rather to step sideways and explore modes of persistence, the reason being that we are talking about the temporality of things in relation to one another, not time per se. It is a temporality much closer to memory in the way it operates than conventional, historical time (see Olivier Reference Olivier2011 here for an extended discussion on this). Thus the contemporaneity of the archaeological record is not about its existence in our present, but rather about its particular mode of persistence that interconnects past, present and future.

In bringing this issue around to the temporality of things, we have in many ways ended at the same place as the previous section. In both cases, the vernacular concept of contemporaneity in archaeology was shown to be inadequate and linked to the ideas of synchronism and anachronism respectively. In both cases, the important question hinges on the temporality of things – their imbrication on the one hand and their survival or persistence on the other. In both cases, we need to expand what these terms mean, to draw out a taxonomy. There are multiple forms of imbrication, as Allen's temporal logic has shown; there are also diverse modes of persistence, as indicated by Serres's metaphors. To develop a more complex and useful concept of contemporaneity in archaeology, ultimately we need to understand how temporality is bound to an object's identity and how it mediates its relation to other objects. To that end, I want to close this paper with a third and final section which explores the idea of consociality.

Consociality and writing archaeology in the contemporary mode

The concept of consociality derives from the writings of Alfred Schutz, a phenomenologist working in the mid-20th century. In his studies on the phenomenology of the social world from the 1930s, Schutz made an important observation that what matters in understanding social relations and time is not contemporaneity but consociality (Schutz Reference Schutz1967). Any two people might be contemporaries, and live at the same time in history, but if their paths never cross and they never affect each others’ lives, this relation is irrelevant. Thus we might question the meaningfulness of being able to demonstrate that a farmer living near Cahokia in the 13th century A.D. was a contemporary with a peasant living near London at the same time. What matters is that they shared the same physical space as well as time, that they constituted a we-relation, as the phrasing goes. This notion of consociates actually lies at the basis of how we often use the term ‘contemporaries’ – of course, people of the same generation or age cohort – but Schutz wanted to emphasize the importance of not separating space from time.

Now although Schutz's concept of consociality is important, we do need to give it a less phenomenological flavour because, as it stands, it entails problematic consequences. For example, teaching to a room full of students would constitute them as my consociates, even if only temporary, but the president of the United States is technically not my consociate, only a contemporary. We have never met, we probably never will, and in no way do we constitute a ‘we’, in the way ‘we’ do when I am present in a classroom with my students. On the other hand, the president is affecting my life – perhaps not to the same extent as he affects others – but nonetheless it is hard to use physical proximity as a primary way to define consociality. We all know the cliche about a butterfly flapping its wings in South America and the weather changing in Central Park; chains of causation need to be integrated into this idea of consociality. We need to de-phenomenologize it. A second and related thing we need to do is to disconnect it from its association purely with people. People can be consociated with things and even things can be consociated with other things.

However, in redefining consociality this way – an actor-network-theory (ANT) version of consociality, if you like – have we not torn down the original distinction between contemporaries and consociates and just replaced one term with another? At one extreme, you could say that everything is connected, and therefore everything is consocial. At a very general level, this may well be true, but it is not terribly helpful. The task is rather to trace gradients of consociality, and identify the more important relations between things. Part of doing this requires that we have good control over the space–time coordinates of these relations. In archaeology, high-resolution dating helps this. But this is only part of the job. If we are to preserve the distinction between contemporaries and consociates, then, it will be a difference of degree rather than one of kind. However, in tying the conception of consociality to agency, we are now faced with a new problem. Let us turn Schutz's original point on its head. Can two people be consociates without being contemporaries? To give a blunt example: can I be a consociate of Abraham Lincoln? Schutz had other terms for consociates who were not contemporaries – ‘successors’ and ‘predecessors’ – but if consociality is defined differently, as I have just done, then such terms become redundant. Indeed, the whole idea of successors and predecessors reverts back to a successional conception of contemporaneity which is what we have been trying to avoid in this paper. Surely agency or affect can be distributed through time as well as space? Surely Abe and I can be consociates? Is not that, after all, the point of Serres's crumpled handkerchief?

I would argue so, but, as with the original distinction, we need to qualify it, especially in terms of reciprocity. If Lincoln and I are to be defined as reciprocal consociates – that is, people who can affect each other – then one argument would be that we both need to be alive at the same time. Clearly we are not; Lincoln died in 1865, 100 years exactly before I was born. This is not to suggest that he or his past actions have not affected me today; they surely have, however weakly. But because he no longer exists and ceased existing before I came into existence, it is difficult to argue that I can affect him in anyway whatsoever. Our consociation is one-way, or asymmetrical. The same is not true of Lincoln's bloodstained coat, which he wore the night of his assassination; this still exists (owned by the Ford Theatre where he was shot), so I can affect it and it can act upon me. Indeed, I can affect objects of all kinds directly or indirectly associated with Lincoln, and as a result alter the present perception of who he was.

Now one could claim that in doing this I am changing the past and thus negating the idea of non-reciprocity with non-contemporaries. However, I feel that this move makes a major leap of abstraction. Our perception and representation of the past is, of course, moulded or constructed in the present, using remains from the past. We can quite seriously change the effect the past has on us by how we choose to interact with its remains. In this way, the present can affect the past – or rather the persistence of the past in the present. But the trouble is, we are not talking about the past – we are talking about Abraham Lincoln. Talking about the past slips us back into the old terminology of past and present as periods or slices of time. The point about using contemporaneity to rethink these temporal relations is that the focus remains on things themselves and their persistence.

But in focusing the attention on the object – in this case Lincoln – we are forced to respond to the issue of how we then choose to define this object we call Abraham Lincoln. At a time when we freely talk about distributed persons and how people and their things are connected (e.g. Strathern Reference Strathern1988; Gell Reference Gell1998; Fowler Reference Fowler2004), can we really say that Lincoln no longer exists if his coat (and no doubt many of his other personal possesions) are still around? If Lincoln was more than his body, then surely there is a case for arguing that, in some small way, a part of him still exists. Lincoln persists, albeit in a very attenuated form. One might counter this by pointing out that I am conflating Lincoln the being with Lincoln the idea or memory; Lincoln died over a century ago, but the idea of Lincoln, our collective or social memory of him, lives on through objects like his coat, but also through other things like the Lincoln Memorial in Washington, DC, the image reproduced in figure 2 or even the 2012 Steven Spielberg film. But why create this division between Lincoln and the idea of him? Why revert to the deep division between world of spirit and world of matter? Can anyone separate their identity from what other people think of them? Can we separate the idea of Lincoln from the hundreds and thousands of objects, images, books and films that relate to him? Can we draw a line between Lincoln's coat (as a distributed part of Lincoln the man) and Lincoln's coat (as part of Lincoln the myth)?

If the idea of a distributed person collapses any ontological distinction between Lincoln and the idea of Lincoln, as I am suggesting here, then surely we can argue that Lincoln and I are in fact reciprocal consociates, and by extension, of course, all the people in prehistory whose remains have survived for us to unearth and study. I think this is a reasonable claim, but once again rather than simply accept it, the real task is to define gradients of reciprocity: what precisely is the agential relationship between myself and Lincoln? How strong are the vectors of influence between us and what form do they take? Just because we cannot draw any ontological divide between Lincoln's coat and the Lincoln Memorial (or the Spielberg film) we are not prevented from still making a distinction. However distributed Lincoln might be, his core or centre resides around his body and that body interacted with that coat – unlike the memorial (which was erected in the early 20th century) or the more recent film. One might even talk of degrees of separation here; the coat lies at one degree, whereas the movie (and myself) lie at the nth degree.

To bring this discussion back to the topic at hand, namely the concept of contemporaneity in and of the archaeological record, it seems clear that there is a much more complex set of issues to be addressed than is often assumed. The concept of consociation, as articulated here, is offered as a way to think about how we articulate the idea of contemporaneity, both in our narratives of the past and in our comprehension of the relation between ourselves and what remains of the past. Indeed, ultimately this is significant because of the consequences it has for how we represent the past and its relation to the present. One of the most important of these consequences, I would suggest, is related to the temporal voice we use in archaeological narratives. We need to write about archaeology in the mode of the contemporary.

It should be clear by this point in the paper that by the contemporary mode, I do not mean something as naively simple as writing with an understanding that the narrative is written in the present. Contemporaneity is not the same as the present, which is merely a period designation, a punctum. A narrative written in the contemporary mode is one which is attendant to the changing interplay between present, past and future tense. And this will vary depending on the objects – or subject – being selected. Perhaps the most critical question here lies in deciding the subject position of the narrative. If everything I have argued here is accepted, then it should be clear that the temporality of any narrative is bound and relative to subject or object positions – there is no general chronology or periodization. This is not to reject chronology as a tool, nor is it to reject the use of periodization as a narrative framework; it is only to insist on its relativity to the subject. Each subject will have its own temporality, which frames the nature of contemporaneity and consociality. The temporal span of that subject will frame the beginning and end of the narrative and it is the interval between these points that determines the configuration of consociality, particularly the gradients of consociation and reciprocity.

A narrative that starts at 500 B.C. and ends at 400 B.C. has a different set of agential possibilities than one which spans 400–300 B.C. This field of possibilities changes as one slides the end point back or forward, and therefore, depending on what we take as our point of reference, history will appear slightly different each time. Gradients of reciprocity will change between objects as some objects either cease to exist or subsist weakly in a distributed manner, like Lincoln's coat. To write in the contemporary mode means acknowledging the relevance of the time frame, it means acknowledging the relevance of the subject; history not only looks different depending on where one chooses to start or end the narrative, it also looks different depending on which subject position is chosen to write the narrative from. This is about embracing fully a multi-subject and multi-temporal archaeology, where the subjects can be anything from humans to horses, pyramids to pots, and where the temporality of these subjects defines the mode of contemporaneity.