In Context: Meaning, Materiality and Agency in the process of Archaeological Recording

The full-text may be used and/or reproduced, and given to third parties in any format or medium, without prior permission or charge, for personal research or study, educational, or not-for-pro t purposes provided that: • a full bibliographic reference is made to the original source • a link is made to the metadata record in DRO • the full-text is not changed in any way The full-text must not be sold in any format or medium without the formal permission of the copyright holders. Please consult the full DRO policy for further details.


TEXT AND CONTEXT
This chapter is concerned with archaeological 'context sheets', pro forma documents that are central to the way in which most British archaeological sites are recorded 1 .
Whilst ostensibly un-remarkable, they pose a number of interpretive challenges. As with documents more generally, the representational logic they engender seems to direct attention away from their own material and artifactual properties. Conceived as passive representations of the material properties of 'the site', their own material properties are paradoxically obscured. Moreover, as literal embodiments of 'context', they seem to confound the anthropological assumption that context somehow exists outside or beyond the text itself.
In a discussion of the concept of 'the archaeological record ', Edgeworth (2003) highlights the conceptual difficulty entailed in simultaneously imagining texts as both artefact and meaning: A written record can be perceived either as a material object in its own right, or as the vehicle for subjective meanings, depending on the point of view of the perceiver. It is very difficult, however, to hold both points of view simultaneously. Focus attention on the objective shape of the printed letters and their meaning disappears. Shift attention back to meaning of the words and their objective form becomes 'fuzzy', then recedes out of direct consciousness. (2003: 4) 1 The interests that animate this chapter developed during a period of six months spent working for a commercially funded archaeological unit, where I was part of a team of approximately 12 archaeologists who excavated a large Roman and Neolithic site in Norfolk. Through this, I became interested in the role that recording sheets played in coordinating and articulating a diverse range of people and things on site. Whilst my main concern is with the use and understanding of context sheets in the practices of the archaeologists I worked with, my account also draws on subsequent analysis of context sheets taken from other sites excavated by the same unit. For reasons of confidentiality, the site and all quotes remain anonymous. This paradox, I suggest, stems from the assumption, foundational to western thought, that artefacts are an inevitable and natural substrate that has to be related through systems of knowledge and categorisation. In this vein, as Strathern (1990) notes, anthropologists have largely imagined material culture as the concrete counterpoint to the abstractions of culture. Thus conceived, meaning is to be elucidated by recourse to social and cultural contexts in which artefacts are embedded. These contexts in turn make artefacts themselves appear as mere illustrations of more general patterns of thought and behaviour. With this in mind, I suggest that the anthropological (and more generally social scientific) propensity to overlook the capacity of artefacts to cohere and elicit actions, stems from this tendency to reduce the 'material world' to its symbolic or meaningful content.
Documents, have provided a focus of one sort or another for a number of anthropological enquiries (e.g. Cruickshank 1998;Harper 1998;Sarris 1993). In a variety of ways, these have tended to draw attention to the different meanings that may be derived from what is manifestly the same text as it enters different contexts. In this way attention has often focused on how texts are given different meanings in different contexts. Such an approach, which has tended in a post-structuralist vein to stress the endless possibilities of meaning and interpretation, has often overlooked the way in which the material and patterned properties of texts frame and contain such interpretations. Thus the work of the sociologist Smith (1990) introduces a necessary corrective, elucidating how texts may themselves facilitate the construction of particular versions of reality. With reference to scientific texts she describes how: The materiality of the text (the printed paper, the TV screen, the computer monitor) is key to the socially organized transition from paramount reality to the domain of scientific theorizing or other textually grounded domains. (1990: 81) Although this approach conceives texts in more active terms, it shares a common assumption that texts are self-evidently set apart from the social reality in which they participate. Whilst texts may impinge upon 'paramount reality', they are not themselves a part of it.
By contrast, I borrow from the work of the anthropologist Riles who argues for the need to examine documents as a subject of enquiry 'in their own right' (Riles 2004a, b). By this she means to draw attention to the aesthetic, artifactual and patterned qualities of documents as opposed to the capacities they have to convey 'meaning' in a more restricted sense. Crook (forthcoming) makes a complimentary point in his elucidation of the 'textual person'. By demonstrating the person-like relations of texts and the text-like relations of persons, he reveals the limitations of analyses that take for granted a distinction between a writing subject and a textual object. The suggestion here is that texts, themselves composed of persons, animate relationships in much the same way as more conventional kinds of 'person'. In different ways, both demonstrate the importance of taking seriously the forms and relations that texts engender and the capacity for these to animate social and material relations in their own right. From such a perspective, the material and aesthetic properties of texts are not separated from the meanings they may acquire.

CRITICAL PERSPECTIVES
In conventional understandings of fieldwork, context sheets are used to record things after the act of interpretation is complete (e.g. Roskams 2001). In other words, although they record the material form of the site and archaeologists interpretations of this, they do not themselves determine or affect these. By contrast, Lucas (2001) suggests that context sheets determine the way in which archaeologists excavate and consequently the very form that the site takes.
In this analysis, attention is drawn to the normative aspects of these documents. In a Foucauldian vein, this line of enquiry suggests that particular kinds of categorisation lead to the formation of particular kinds of subjects, as well as to particular kinds of objects. Thus, as Lucas has it, context sheets 'control not only the record, but also the bodies who produce it ' (2001: 9). If such documents act to make the objects of archaeology comparable, then they do this by making the actions of the people that use them comparable. Here then, the emphasis is on the strategic functions that these documents perform. Chadwick (1998) andHodder (1999) also note how such forms constrain the actions and hence the creativity of those excavating sites. Although originally and potentially empowering, in the sense that they wrested interpretation and description away from the site director, Chadwick suggests that context sheets have increasingly become 'tools of a new archaeological orthodoxy ' (1997: 5), facilitating the imposition of categories and procedures on the archaeological workforce. In a related way, Hodder (1999) argues that whilst such forms potentially facilitate 'multivocality', in practice they have often restricted interpretation to the specific parameters engendered within the form. Whilst context sheets thus constrain those who use them, they are also represented as curtailing the capacity for the archaeological record to contradict and hence change the thoughts and ideas of those excavating. For both Chadwick and Hodder (cf. Adams et al. 1995), critique of conventional context sheets leads to the call for new methods of recording. Contexts sheets, they stress, should be less prescriptive and give greater emphasis to interpretation.
Important as such critiques have been for the development of new methodologies, my own concern is not with the strategic use of such forms, but with the more ethnographic question of how these are used and understood by a variety of people in the practice of excavation. In eschewing a more critical perspective, my intent is to use context sheets and the documentary practices through which they are authored, to reflect upon a set of debates concerning the relationship between 'materiality' and 'agency'. As my account will demonstrate, both of these terms are analytically problematic precisely because they are ethnographically significant. Miller (1987;  Yet as Ingold (2000) amongst others notes, the very idea of material culture rests on an ontological separation between the material or natural world on the one hand, and the cultural world on the other. Whilst each may shape the other, the assumption of an irreducible difference entails the idea that each are characterised by fundamentally different laws. Ingold suggests that the mistake has been to confuse the many separations between different kinds of materials, with a fundamental separation between materiality and the mind. Thus rather than seeing action as arising out of a dialectic between the material and the social, he suggests the need to focus on the diverse kinds of matter that intersect and shape one another in the course of social life.

MATERIALITY AND AGENCY
Whilst Ingold (n.d.) explicitly distances himself from aspects of the approach, the philosopher of science, Latour (1987;1993;1999), similarly calls into question a distinction between 'the social' and 'the natural' as an analytic starting point. In line with actor-network theorists more generally (e.g. Grint et al. 1997;Hassard et al. 1999;Law 1994) he suggests that rather than pit material and social agents against one another, it is necessary to look at the hybrid networks of people and things in which different kinds of 'actants' are conjoined. Action is thus seen to emerge not in the dialectic of material and social agencies but in the ways in which different kinds of people and things are co-mingled. For example, he argues in the case of the 'sleeping policeman' 2 that various people's wills, acts and ideas are conjoined with those of gravel concrete and paint. As such, some of the characteristics of policeman become pavement, just as some of the characteristics of pavement become policeman.
The point, as Latour notes, is not to extend subjectivity or intentionality to the material world and nor is it to treat humans as if they were objects. Rather the rejection of a subject/object leads to the position in which agency (action, intention, will) no longer has to be located by reference to one or the other of these poles.
Indeed the notion of 'actants' as opposed to 'agents', is intended to signify a shift away from the assumption that intention or will are the exclusive properties of people.
The work of actor-network theorists is useful to the extent that it makes visible the contingent associations of people and things though which various kinds of ideas and actions emerge. Yet the result is both a unifying theory of things and a unifying theory of action. As others have objected (Henare et al. 2005;Strathern 1996), the consequence is to flatten the diverse ways in which other groups of people might themselves imagine what a 'person' or a 'thing' is and to submerge the issues of how agency is indigenously conceived. As these recent critiques have pointed out, issues of agency and materiality cannot be resolved in advance of the ethnographic encounter. In this vein, my own account seeks to interrogate the different ways in which archaeologists themselves perceive their own and others agency in relation to the documents through which excavation proceeds.

ARCHAEOLOGICAL CONTEXT
In order to locate my discussion of the way in which archaeological recording sheets are used and understood on site, it is necessary to understand the archaeological concept of 'context'. Archaeological 'context sheets' are integral to the way in which most modern archaeological sites are recorded in the UK. In a recent archaeological fieldwork manual, Roskams (2001) states that the objective of archaeological recording is: To split the site into its component parts -its stratigraphic units, however defined -and then remove them in the reverse order to which they were deposited, recording their physical, spatial and stratigraphic properties in the process and collecting finds from them to agreed sampling policies as one proceeds. (2001: 110) Excavation thus conceived, is a process of taking the site apart, whilst producing a record that preserves not only the artifacts or finds, but also the various kinds of relationships or 'contexts' through which they are related. As Roskams elucidates, 'context' is an integral part of the archaeological record, and central to the very definition of the discipline of archaeology: In contrast to random digging for treasure, the work of the field archaeologist must take place in controlled conditions, allowing for full recording of the physical character and spatial disposition of the stratigraphic units of the site.
Such conditions do not merely allow better descriptions of the site features and yield large numbers of finds; they also let the excavator understand the latter's context of deposition and position in a sequence of development of the site. In the practice of excavation, the term 'context' refers to physically bounded units which are classified in terms of the actions imagined to give rise to them (Lucas 2001). For example, a 'cut' refers to a context created by the removal of material (as in the case of the excavation of a pit), whilst a 'fill' refers to a context created through the deposition of material (for example, the material that accumulates in a pit). As Lucas notes, this idea of 'context' is supported by a very specific kind of temporality: 'contexts' are conceived as homogenous moments, which may be chronologically connected to others, but which are themselves essentially a-temporal (2001: 158) The properties of such 'contexts' can be recorded in site notebooks, and indeed this remains the normal practice in most university-funded excavation.
However on commercially funded sites this information is almost exclusively collected on pre-printed sheets, known as 'context sheets.' Changes in recording methods both reflect and enable changes in the practice of field archaeology more generally. With the professionalization of archaeological fieldwork during the 1960s and 1970s, the process of recording has been increasingly de-centralized, as excavators record and interpret the features they excavate. Whilst the physical singularity of the site notebook has its counterpart in a system where recording is largely undertaken by a single person (the site-director), the context sheet's literal capacity for division (each is mobile and separable) enables a system in which interpretations are arrived at simultaneously by diverse people on site.
The separation and connection of contexts at the heart of archaeological recording is similarly enabled by the material properties of paper itself: context sheets are detachable from one another and hence can be easily transported to the relevant place for recording. Yet their thinness and flatness mean that they are easily recombinable -usually in ring binders -where they can be unified in a logical sequence. Roskams (2003) hints at another way in which the physical properties of context sheets are more suited to modern excavation. By contrast to the smaller sites that characterized much academic excavation up to the sixties, developer funded excavations are often undertaken on a much larger scale. As site size has tended to increase, this has favored context sheets which, by contrast to the bounded site notebook, are potentially 'infinitely expandable' (Roskams 2003).

FILLING IN
The context sheet is divided into a number of distinct sections, through which the description and interpretation of each context is fragmented (figure 1). The first sections provide the space for information locating the context within the site and more specifically the grid/trench. This is followed by a space for 'description' of the physical characteristics of the context, in terms of criteria such as the texture, stoniness, color and particle size. For example, a typical description of a ditch excavated on the site I worked reads, 'dark gray/black sand and mottled with rust colored iron panning. Moderate gravel inclusions -in some cases cemented by panning' Following the description of each context, the form provides a space for the documentation of each context in relation to others. Below this, a number of boxes enable those excavating to record the different kinds of artifact located within each context. Finally, a space is provided for the 'interpretation' of each context. By contrast to the 'description', which should ideally be phrased in neutral and scientific terms, this enables those excavating to make wider inferences about the likely significance of the feature in terms of the wider processes occurring on site.
The recording of context was part of a set routine. Context sheets were located in the site hut, a metal porta cabin, situated on the edge of the site, along with other recording equipment such as spare pencils and rubbers, hand tapes, 'base-plans' and permatrace for drawing plans and sections of excavated features. Tools used for excavating features such as mattocks, spades, hand-shovels and wheel barrows, by contrast, were kept in a separate store and hence the conceptual distinction between 'excavating' and 'recording' was literalized in the physical layout of the site.
Before recording commenced it was normal practice to fully excavate the archaeological 'features' (usually pits, ditches and post-holes). This was a methodical process in which attempts were made to define different 'contexts' through a variety of means. In some cases, this process of identification was relatively unproblematic and in such instances excavators spoke of the way in which the differential texture, color or 'feel' of the soil made it easy to differentiate the extent of distinct contexts ( figure 2). Yet in other cases differences were difficult to spot and at such times excavators employed a variety of means in order to make these distinctions apparent.
For example, other people on site -specifically the site supervisor, or those imagined to have more experience -were enlisted, existing 'sections' were cleaned back in order to make differences of color and texture more apparent and in some cases people simply left the feature for a while, claiming that differences often appeared over time through the differential ways in which soil eroded.
Once a feature had been excavated and 'contexts' defined, the archaeologists would generally return to the site hut, where they would collect context sheets, and sign out individual 'context numbers' from a book known as the 'site indices'. These unique context numbers meant that whilst the description of different contexts could be literally split apart and documented on separate sheets of paper, it was still possible to trace the relations between these. The context sheet makes it possible to displace the original feature onto a form, which has a number of physical properties that make it much more suitable for the task of embodying the archaeological record. Lucas (2001a) suggests that a key attribute of the archive is its capacity to be iteratively re-visited. As such, the material properties of paper are important to the role that such sheets perform in the creation of a 'record'. Unlike spoken words, paper inscriptions are able to remove descriptions from the situations in which they arise. Because of this, context sheets are able to capture particular interpretive moments, which can then be returned to by other people, at other times.
Although the distinction between 'excavation' and 'recording' was made evident in the spatial and material organization of the site, once excavation was completed, this distinction was effectively collapsed. As Edgeworth (2003) argues, the archaeological concept of 'the record' conflates the material remains and the documents that are made of this. Thus he suggests that the text persuades both the reader and writer of its own 'reality' through a variety of rhetorical strategies. Without denying the importance of such rhetorical conventions, I would add that the physical form of the context sheet was also an important aspect of its persuasiveness as an 'objective' document of the site. Bateman (2006) describes how site plans bear the material essences of the site itself, scarred with process of their production and often stained with dirt. Similarly, context sheets, once archived, appear to retain a very concrete connection with the site, often being tattered and ripped, covered in mud and smudged by rain. In this sense the perception that they contain information that objectively conveys the material properties of the site, is enhanced by the connection made latently but very concretely in the material traces of excavation that such documents retain. Thus as Pellegram (1998) has argued in relation to bureaucratic documentary practices, the physical nature of texts have much to do with the messages they convey.

FORMING THOUGHTS
Whilst recent archaeological accounts of context sheets have tended to stress their instrumental functions on site, a close examination suggests that they are not only or indeed primarily strategic in nature. Flicking through a ream of filed context sheets, the diversity of ways in which people fill them in quickly becomes apparent. For example, whilst some give long and detailed descriptions, others are much more cursory. Moreover, many of the forms leave large sections blank, whilst on others sections are expanded and notes or diagrams added on the back (figure 3). Sometimes the reason for leaving spaces was simply that the criteria was inappropriate to a specific feature that had been excavated, but in other cases more fundamentally important aspects of the context sheet such as descriptions, interpretations, measurements or the excavator's name were omitted. One of the archaeologists on the site on which I worked thus claimed that 'At the end of the day, they are just a piece of paper -you can fill them in how you want.' The importance of the information they carried was also confused by remarks made by a number of site supervisors, who suggested that once the site was finished these were of limited value. Most of the information on sheets, was redundant, giving too much of the wrong sort of detail. In depth description of individual contexts, it was suggested, was of little interest when it came to the interpretation of sites (Yarrow 2006a).
If context sheets are not only, or even primarily, imagined by those who use them as instruments of control, how might we interpret the importance that was nonetheless attached to them? An alternative reading is suggested by the work of the anthropologist Reed (forthcoming), who draws attention to the way in which constraint can itself be experienced as a kind of freedom. Reed's analysis focuses on the use of the 'warrant cover', a form filled out by prison warders, in a high security prison in Bomana, Papua New Guinea, describing such things as the medical status, age and ethnicity of inmates. Whilst such documents might seem ready-made for a Foucaldian analysis, Reed suggests that a concern with their use as mechanisms of control has tended to conceal an indigenous appreciation of how these are assigned agency. From the perspective of those who use them, thoughts are imagined to be anticipated by the document's form. Since every response seems to call up the same basic design, responses simply add concrete detail to the abstract pattern already there. Prison warders thus feel themselves to be passive actors, unable to go beyond the design of the form itself. The perception, here is that agency lies not with them, but with the document's technology. As Reed concludes, then: Action is a part of the document's design, a movement that [people] at Bomana recognise and feel led by' (forthcoming: 18) Despite profound differences in terms of the epistemological schemes that support them, I suggest that context sheets were similarly regarded by many archaeologists as eliciting thoughts and responses which were not straightforwardly their own. Since every response had to conform to a strict set of textual practices, the act of completing such forms seemed to emanate more from the document than from the person as an autonomously arrived at interpretation or description.
Perhaps this in part helps to explain the reluctance that site assistants often expressed when it came to filling them in. Where it seemed obvious what was being excavated, or where the interpretation seemed self-evident, filling in the contexts sheet was regarded as a 'pointless' or 'boring' chore. Since the descriptions always seemed 'the same', and an air of inevitability seemed to surround the responses that they made. Although sometimes filled in during excavation, this was often done after the event, hurriedly and with relatively little conscious thought.
Yet if this self-generating capacity often resulted in the feeling of inevitability and in the perception that form filling simply amounted to replication, then at other times, site assistants exploited this capacity was as a tool to extend their thoughts and actions. This was perhaps the case in situations where it was difficult to tell what was being excavated and where people consequently became unsure of their interpretations. In such instances, issues of 'meaning' or 'interpretation' were foreclosed through the attention that was given to 'description'. Thus, rather than considering how a particular feature was formed, site-assistants would focus on the particular shape the feature took, or the kinds of fill that it contained: was it 'ovoid' or 'sub-ovoid'; was it 'flat bottomed'; was a particular fill a 'light brown sandy silt' or a 'mid-brown silty sand'?
Sometimes those excavating would say that the context sheets 'helped them to think' about what they were digging as, for example, when it made them compare the particular feature they had excavated with others on site. In this way the context sheet sometimes elicited particular kinds of conversations and relationships on site.
Attention in these moments was often directed to the pattern or form that language took and to the comparative possibilities that it elicited, rather than to its 'meaning' as such. Form filling, I was told, had a rhythm to it as with other aspects of excavation. 3 The form, seemed to pull the answers out of itself, so that the response was felt to be part of the documents design as much or more than a part of their own actions. In this sense, context sheets were often seen to make visible the agency of those excavating.
As one site assistant once remarked to me, 'Sometimes the fact that there's a piece of paper -that the piece of paper is actually there -reminds me that I've done something.'

FORMING IDENTITY
Context sheets were often imagined to make evident aspects of the identities of the people who filled them in. The site supervisor, who collated and interpreted the context sheets on the site on produced on the site on which I worked suggested that: You can always tell from the sort of person they are how they will fill [context sheets] in. Context sheets also tell you about the people who fill them in.
In this way, he explained, aspects of people's character that are normally hidden sometimes become apparent by looking at the context sheets they produced. By the same token, he described how people who are careful and meticulous produce careful and meticulous context sheets.
This perception that context sheets concretise the actions and thoughts of those who create them was more widely shared. Thus, it was claimed by some that unnecessarily detailed and precise descriptions were characteristic of those forms created by 'youngsters', a fact that was related variously to the way in which archaeology has become 'increasingly professionalized' and the idea that 'they've been brought up in a more objective world.' One site supervisor suggested that variation in the way in which such forms were completed was partly related to differences in terms of knowledge of the archaeological process. Those who know how context sheets are interpretatively used, she suggested, may leave sections blank not through ignorance, but because they know that certain kinds of information will not be helpful. Yet unnecessary lengthy descriptions and interpretations were not always attributed to lack of knowledge. Thus she suggested that some people write at length because they become bored with the job and feel that they are unable to express their opinions. Such people, she argued, feel empowered by the thought that their ideas are made concrete and tangible as part of a record that will endure.
Whilst the length and content of the descriptions people provided were thus taken to evince aspects of their identity, the definition and identification of discrete contexts was also seen as a measure personal identity. In this vein, a distinction was sometimes asserted between 'splitters' and 'lumpers': whilst the former referred to those who tended to separate contexts on the basis of small difference (leading to the identification of numerous contexts), the latter applied to those more willing to aggregate differences under a small number of contexts. The difference between 'splitters' and 'lumpers' was taken to demonstrate not only differences of approach, but different personal attributes more generally. Whilst 'splitters' were imagined by some as careful and meticulous, others saw this attention to detail as 'unnecessary', or 'anal', and spoke of the way in which superfluous description was employed as a way of avoiding making interpretations, or avoiding responsibility. Conversely, 'lumpers' were said by some to employ a crude and un-thinking approach, whereas for others the strategy represented a willingness to interpret what was being excavated, and hence the ability to 'see the bigger picture.'

CAPTURING THE MOMENT
Context sheets (as elucidated above) are widely imagined to form part of an enduring 'record'. This vision is the counterpart to the conception of excavation as destruction (Bateman 2006;cf. Lucas 2001b): the obliteration of the relations and artefacts that constitute the initial site of excavation is justified by the understanding that the record of this will endure indefinitely.
Whilst context sheets were thus formally conceived as a record of the spatial and material relations inherent in the site itself, those excavating sometimes used these sheets to capture more intimate or personal thoughts. For example, a site supervisor recounted how on one context sheet she came across the inscription 'Dom's my best friend'. The motivation and intention behind such acts of inscription are difficult to interpret, and indeed it is partly this opacity that declarations of this nature themselves exploit: is it ironic or literal; flippant or serious? If interpreted as a joke, as it was by the supervisor who later came across it, then the humour seems in part to inhere in the way that an apparently highly personal expression is captured in such a manifestly objective document. The inscription seems also to flout the overt conception of context sheets as records of 'the site', by turning them into records of the people and relationships through which they are created.
Yet the supervisor who told me of this comment suggested that whilst humorous, such inscriptions were nonetheless profound. In making these fleeting and incidental details concrete as part of a record that endures, people are able to in a sense step back and see their own actions as if through the eyes of future generations.
Whilst she suggested that the reality of excavation is often experienced as being tedious and monotonous, such inscriptions enable people to imagine what they do as belonging to something bigger than themselves.
On the back of one context sheet I came across, the site assistant who filled out the form had written: Remember when you find yourself in the interminable expanse of time that stretches out like vast deserts, the glorious oasis of life, the cup of tea. There will always be a brew for you with us.
It is unclear who this reflection is intended for, and perhaps it is simply a quirk of the material properties of pen and ink that has led an incidental detail to endure (much in the way that archaeological sites themselves preserve actions and traces, not conceived by those responsible for them as memorialization). Yet the almost deliberate wording and multiple crossings out in the original text, suggest a degree of care and thought that hints at an alternative reading. The author seems to have intended that their reflection be read. If this is the case, it could be suggested that the very idea manifested in this statement was elicited by the context sheet itself and the putative audience it concretises for the site-assistant. The context sheet does not simply capture a thought that already existed: it makes that thought conceivable.

MEANING, MATERIALITY AND AGENCY
Anthropological and social scientific theorizing on the subject of materiality, has often led to relatively general propositions about the nature of 'things'. In such a view, theories may illuminate the various meanings that different people might attach to artefacts in any given context, but the essential nature of those things is regarded as independent of both their theoretical and ethnographic designations. Anthropology, in other words, has often been guilty of reproducing a specifically western vision of the world, in which meaning and materiality are self evidently independent.
By contrast, if we take seriously the recent suggestion that 'things are meanings' (Henare et al. 2005), then a theory of things -or, indeed of 'materiality'becomes a pointless goal. If things and meanings are inextricable from one another, then it is impossible to assert in universal terms what a 'thing' is. The issue of how artefacts are defined as against 'people' and, indeed, as against one another will always be a matter of ethnographic interpretation.
To take the context sheet as an example, how and in what sense this could be considered 'an artefact' independent from the various people and things that contribute to its creation? Is 'it' only the paper and ink that constitute any given sheet, or the more general Platonic form that is reproduced for use on different sites? In what sense can we assert a separation between the sheets that are filled out on site, and the circuitry of the computer and printer, through which these are created? And how, if at all, can we disentangle such recording mechanisms from the actions and thoughts through which recording takes place? In one sense, we cannot, and the conclusion to be drawn is that the assumption that human subjective agents are self evidently distinct from the material world must always be precisely that: an assumption. Analytically speaking, it is impossible to divorce 'the context sheet' from the wider networks of human and non-human 'actants' (Latour 1999) in which it is enmeshed. In this sense both action and meaning do not reside self evidently in the human 'subject' but are distributed in relation to a variety of people and things.
Yet archaeological recording itself attempts to bring precisely such an opposition into being (see also Yarrow 2003Yarrow , 2006b. The creation of 'the record' requires that context sheets be considered as artefacts independent from the people who produced them. As such, archaeological practice attempts to resolve the issue of what is 'material' and what is only 'mental'; what is 'objective' and what 'subjective'.
In doing so, it reproduces a specifically western conception of the individual, corporeally bounded by the extent of the body. The process of excavation is imagined by the archaeologists who engage in it to generate thoughts and feelings as well as stories anecdotes and knowledge, yet these are not regarded as altering the fundamental biological or physiological attributes of the person. Hence in creating objects, archaeological recording creates the subjective agent as the largely nondiscursive side effect.
The idea that things and meanings should not be separated at an analytic level, does not therefore equate to an assertion that this opposition does not exist ethnographically. Rather I suggest that it is only possible to appreciate the ways in which people and things are combined or separated, if such oppositions are not assumed at the outset. Whilst archaeologists frequently assert and demonstrate the objectivity of the artefacts and contexts they unearth as distinct from their own subjective interpretations, the work required to achieve this distinction is not reducible to the opposition itself.