1When attempting to make sense of intentional marks on Predynastic-Late Predynastic material culture, emphasis typically centres on image morphology and the differentiation of iconographic from semantic functions, or the elucidation of symbolic meanings. The concern to classify a mark, or sets of marks, according to conventional categories of sign and symbol has led to a general neglect of related areas of past meaning production. This paper takes as its point of departure the premise that meaning is not bound up solely in the morphology of graphical marks but also resides in their materiality as part of individual and social practice. It is important to document the materials and techniques used in graphical activities (Kahl 2001: 106-108), but this is only the first step. If we are to understand graphical developments in relation to linguistic and wider cultural practices (e.g. Baines 2010: 146), we also need to move beyond notions of material as playing a passive supporting role as a “substrate” or “background substance” for signs, or treatment of imagery as a distinct “constrate” (Mountford 1969: 630 in Kahl 2001: 106), to see these as mutually constituting.

2As I have discussed elsewhere (Piquette 2007, 2008), theoretical concepts of materiality and technological practice (cf. DeMarrais et al. 2004, Dobres 2000) provide useful frameworks for investigating early graphical media. These enable us to reframe the material-image relationship as one of mutual constitution, with the recognition that the meaning and function of one cannot be fully understood without reference to the other. Technological acts of knowledgeable, embodied producers created particular material, technical and semantic relationships; these in turn informed subsequent sensorial perception, embodied engagement, and thus cognitive processes of understanding and meaning construction. It is all very well to advocate a more prominent role for theories of material practice in research on the development of early Egyptian graphical culture, but it is also the case that meaningful adaptation and application of social and archaeological theory is currently impeded by issues of a more practical and immediate nature – namely direct access to image-bearing objects, or accurate high-resolution images of them, as well as to their associated archaeological data. Some of these issues are a question of publication methods for archaeological evidence classified as ‘(early) writing’, ‘art’ or a similar category and conventional assumptions about what should be documented and visualised in a book or article. Other issues have to do with data capture technology.

3In what follows, I discuss the some of the barriers to technology- and material-focussed studies of early Egyptian graphical culture as presented by variability in documentation and re-presentation, focussing particularly on line drawings and photography. In light of these, I present an emerging advanced digital imaging technology, Reflectance Transformation Imaging (RTI), and present selected results from a recent RTI research project undertaken at the University of Oxford in collaboration with the University of Southampton. My aim is to demonstrate the ways in which this powerful imaging technique overcomes many documentation and visualisation challenges by providing a highly accurate record of an object surface, thereby enabling a more reflexive, theoretically-informed and importantly, archaeological, study of the development of Egyptian writing.

Technology Past and Present: Addressing a paradox

4While much of archaeology is ultimately concerned with transformations in technology, Geoff Carver (Carver 2005: 3-4) observes the paradox that, as investigators we have been negligent in considering the effects of the technologies we use in the present for the documentation and study of technologies in the past. Carver focuses his critique on lacking awareness of the impacts of digital technologies, but his observation is also germane to the unreflexive use of analogue technologies for the documentation of archaeological evidence. Only now that we struggle to keep pace with digital technological change, and to grasp how these are transforming modes of research and knowledge construction (e.g. Taylor et al. 2010), is the formative role of conventional technologies being drawn into our discursive consciousness. Discussion of these issues is now underway, although mainly from within Digital Humanities and Information Science (see Benardou et al. 2010). A recent study¹ of how researchers experience and engage with analogue (i.e. paper-based text-objects such as books and journal articles) compared with digital text-objects (e.g. websites, PDFs, blogs), threw into sharp relief the tremendous impact technological change and choice is having on research processes and their outcomes (Piquette et al. 2010). This research into modern writer/reader experience and practice in academia combined with Carver’s critique has provoked me to reflect on the impact of modern technologies used in research on early Egyptian graphical culture.

5Beyond the need to think more critically about how technologies used in the present inform the ways in which we reconstruct the past, there is an added twist for early Egyptian writing and its progenitors: these are not typically seen as material and technological in the same way that, for example, lithic and ceramic data are. I am therefore interested in assessing direct and indirect modes of documentation, as well as subsequent paper-based and digital visualisations. My interest also lies in identifying where these can be augmented to take fuller account of materiality, and the kinds of insight to be gained from particular documentation techniques for our understanding of the development of early Egyptian graphical culture.

6The extent to which a method enables a certain kind of documentation is not simply an objective act where one simply deploys the best imaging hardware and most up-to-date software or carefully adheres to a particular facsimile technique. The ambiguity, heterogeneity or multiplicity of the material world means that choices must be made with regard to where we direct our ‘gaze’ as investigators (Shanks & Hodder 1998: 77). We may choose to document the ‘natural’ surface and/or surface modifications of an artefact as revealed by the trowel and brush, or – perhaps enchanted by the iconic, indexical or symbolic possibilities of surface marks – record, visualise or otherwise direct study only to the contour of shapes attributed such functions (e.g. Raffaele 2005: figs. 1-2, fig. 4, B13, Anselin 2008), if that (e.g. Ogdon 2006). This selectivity is what Haas (Haas 1996: esp. 32) refers to as the “Technology is Transparent” myth; we effectively looking through the materiality and technology of our endeavours (see also O’Hara et al. 2002: 271), neglecting to recognise the impact these have on analysis and interpretation. Archaeological remains, whether or not they incorporate epigraphic, palaeographic, glyptic, pot mark or other ‘visual’ information are not simply given “data”. Archaeologists may have a raw material, but from the moment of designating it “archaeological” and making something of it through recording, drawing, photographing, publishing, exhibiting, etc. this is nevertheless a mode of social and cultural (Shanks & Hodder 1998: 77) – and as just highlighted, technological production. I now turn to selected methods of (re-)production in the present, particularly photography and line drawing. In particular, I consider the challenges and opportunities these present for studies of early Egyptian graphical development that are also sensitive to the techno-material nature of the evidence.

Technologies of Image Re-production in the Present

7Documentation and visualisation of early Egyptian graphical data goes back more than a century and has involved a variety of methods and practices. Until the advent and widespread use of photography, artefact facsimile and dissemination consisted predominantly of sketches, line drawings and water colours. Photographic and line drawing re-production have been primarily in black and white, with the occasional use of colour tinting (Emery 1938: pl. 18A, Petrie 1900: pl. 17, no. 26). With the decreased cost of colour reproduction, especially with the advent of digital and online publication, the critical data that colour provides about past technological practice is increasingly available.

8Facsimile line drawings are an indispensible component of documentation (der Manuelian 1998: 97, see also p. 100, n. 4 for brief bibliography on epigraphic and palaeographic method). What is striking about the 2D re-production techniques found among the studies on early Egyptian graphical culture, is the diversity of ways in which these are deployed and the different extents to which past materiality is re-constructed. A line drawing might be used as an illustration on its own, or together with its companion photograph for comparison, with or without scales (van den Brink 2001: 66-67, figs. 73-77), or the methods might be combined to highlight particular surface details (e.g. line drawings on photographs, Ibrahim & Tallet 2009: 181, fig. 2). A facsimile of a First Dynasty inscribed funerary label from Saqqara combines artefact morphology with preserved and proposed compositional structure, and transcription and description (Menu 2001: 165, fig. 1). However, other features such as the potentially significant use of red, black/dark grey, green, and yellow/white coloured pastes in the incisions (albeit colour as it appears today; see Piquette 2007) are neglected². Publications may also provide information about production technologies in figure captions and in-text description, e.g. potmarks: “wet incised”, “dry incised”, “lightly incised after firing” (Friedman et al. 1999: 8-9, figs. 7-8). Likewise, on occasion figure captions specify that tracings are made from photographs rather than the original surface (e.g. Storemyr 2009: 126, figs. 8-9). In her palaeographic study of early Egyptian writing Regulski (Regulski 2010: 10-12) helpfully provides an alphabetic indicator of, e.g. whether a sign outline was made by tracing the original object, was a digital tracing (presumably of a photograph of the original object), a published drawing or photograph, or derived from Kaplony’s (Kaplony 1963, III) reconstruction drawings. However, such information is by no means provided regularly in primary source publications, presenting a substantial barrier to comparative research from an artefactual standpoint.

9Inconsistencies in visualisation methods in separate studies is one issue, but within a single-themed publication it would seem that the value of bringing related research together between two covers is partially defeated by variability in re-production. For example, Archéo-Nil 2009, on early rock art, reveals substantial differences within and between individual studies. Imagery from the same site may be differently illustrated: a black and white photograph shows an incised “Serekh of Horus Qa-a” and other marks, while line drawing is used to depict a “Late First Dynasty hieroglyphic inscription [above a boat scene]” (Darnell 2009: 102, figs. 24-25). In contrast to facsimile methods where subtractive surface transformations are inked in completely (Winkler 1938: pl. 32, 2), methods which are more sensitive to past techniques of production, such as shading and line width used to differentiate pecked, engraved and scratched surface marks (e.g. Claßen et al. 2009: 50, fig. 6), facilitate comparative analysis of past techniques and tool use. In addition to separate use, the combination of line drawing and photography can bring together empirical and contextual information to convey important aspects of materiality, e.g. where Ibrahim and Tallet (Ibrahim & Tallet 2009: 181, fig. 2) indicate relative positions of scenes in bas-relief on the same rock face. Photographs can convey certain technological details and contextual information, but quality varies highly, ranging from relatively good raking light to subjects being virtually lost in shadow (e.g. Gatto et al. 2009: 161, cf. fig. 14 with fig. 15).

10Of course many factors impinge on the process of image documentation in the field and what ultimately reaches publication. Nigel Strudwick (in the context of Theban tomb painting, Strudwick 2001) highlights the logistic, training and financial implications for various facsimile methods (see also Caminos 1976, Bell 1987, Traunecker 1987, Malek 1988). Lighting presents a perennial difficulty during photography or first-hand drawing. As just touched upon, surface details and their colours, whether underlying or applied, are revealed, altered, or obscured depending on the illumination source, location and intensity (Gibson 1979: 97-98), and cannot always be predicted or controlled. Although a greater challenge in outdoor contexts, lighting also presents difficulties for portable surfaces, especially where surfaces are poorly preserved. For example, a line drawing of a red and black ink-inscribed label (Newberry 1912: 282, E), presumably based on the photograph made using visible light (Amélineau 1905: 425, pl. 37, no. 3; Louvre E 25.268), differs in many details from the drawing based on photographs made under multispectral light. The latter flesh out additional traces of the imagery and other marks (Godron 1990: 27-28, pls. 1-3).

11These differences in copies bear out an additional point, however, that visibility involving ocular stimulation via a light source is but one factor impacting on how an investigator perceives surface information and chooses to reproduce it. Differences between line drawings³ of a wooden label with the same provenance and dating (British Museum, EA 32650) with preserved incisions, some more ambiguous than others, indicate the extent of interpretation that takes place during the act of copying (also cf. Vikentiev’s (1959) ‘tidied’ line drawing of an ivory label dated to the reign of Djet (fig. 1) with its photograph (pl. 2, tablet B, from Emery 1954: 102, pl. 35b).

12Schäfer remarked on the habit of publication to focus on the content of the image, with other surface features inadequately indicated (Schäfer 2002 [1919]: 74-75). Indeed, outdated notions concerning the immateriality of certain kinds of surface feature have persisted, leading to unevenness in their contextualisation in documentation and visualisation. Assumptions concerning the stand-alone nature or ‘complete’ state of graphical marks have led to limited consideration of evidence for the material acts leading to their expression on a surface (Piquette & Whitehouse2013). Potmark typologies are an important chronological tool (e.g. van den Brink 1996), but in addition to mark shape and meaning content such typologies would also benefit from accounting for palimpsest, evidence for tool type and ductus. As Mawdsley (Mawdsley 2009: 199-200) and van den Brink (van den Brink 1992: 267) remark, a significant problem for retrospective potmarks studies is that these are often illustrated without specific reference to their original pottery carrier, making dating and other contextualisation difficult if not impossible. Different levels of subjectivity are also introduced depending on the nature of the source subject. A facsimile may be created by tracing an original subject or a photograph of it (Regulski 2004: 952, n. 1) or vice versa, although this is rare (see Quibell 1923: pl. 11, no. 3). This is not always made explicit, as mentioned, but can sometimes be inferred. For example, when imagery lost in shadow in an original photograph is omitted in a later drawing, this strongly suggests that the source subject was the photograph rather than the original object (cf. Petrie 1901: pl. 3, no. 1 with Emery 1961: 53, fig. 13, Piquette 2010: 56, pl. 1) – and that checking against the original object probably did not take place.

13That a line-drawing of an object is unavoidably the product of subjective choices (Shanks and Hodder 1998: 77) is also highlighted by the contrasting treatments given to objects which do, or do not, bear imagery. On the one hand, the interior and exterior surface features of ceramic vessels from Umm el-Qa’ab are diligently drawn using pen and ink with details precisely indicated by fine stippling (Dreyer et al. 1998: 131, figs. 22-24). On the other, in the same publication a perforated ivory plaque bearing the niched frame of ‘Narmer’ is rendered more schematically. Its rectangular shape is merely indicated in outline along with the basic contours of the incised imagery, although the remains of the coloured paste infilling the label incisions are indicated in careful detail by a fine hatching, with some damage indicated by sparse stippling (Dreyer et al. 1998: 139, fig. 29, pl. 5).In contrast to facsimiles of the contemporary ceramic vessels and their surface features as just noted from the same site and appearing in the same publication, however, evidence for label surface preparation or other tool marks is not visualised. Here we see that material contextual information is inconsistently provided in facsimile form depending, presumably, on assumptions the investigator imposes concerning object function and meaning. For the labels and other objects seen primarily for their ‘writing’, details of technique, texture, style, mechanical use, and ideal orientation for attachment, viewing/‘readability’, and so on are usually de-emphasised in favour of elucidating sign morphology.

14A range of other issues invariably inform the investigator’s ‘gaze’, and influence what is drawn, photographed or otherwise documented. Observers may ‘see’ object surfaces differently depending on environmental conditions such as access to a stationary or mobile light source of varying intensity, physiological factors, such as quality of eyesight, the presence/absence of colour blindness, and so on. ‘Looking’ is also informed by cultural factors such as training and area of expertise, research interests, methods and theoretical perspective (see Terras 2006). The implied aim(s) of re-production facsimile is usually for preservation and documentation, and to aid the processes of reading, decipherment and interpretation. Because of the subjective nature of observation and the fact that method and motivation are not always declared, it is important to think critically and reflexively about the processes which informed a published facsimile.

15As for line drawings, these are typically used to distinguish intentional from unintentional surface marks, for example to clarify the contours or details of relief which in a photograph may be difficult to pick out from an eroded or damaged back-ground. For Schäfer (writing in the 1956 Preface to his Principles of Egyptian Art) the primary aim of his use of illustration in his text is “...not to show the beauty of Egyptian works of art. They are chiefly intended as examples of renderings of nature which seem strange to us. ...this goal can be achieved in line drawings which demonstrate in an unpretentious way the point I am discussing” (see Schäfer 1919 [2002]: xxvii). Strudwick (Strudwick 2001: 128-129) stresses the importance of the ‘feel’ of the surface the epigrapher acquires as she works in front of the surface being documented, and the ‘feel’ of an original versus a facsimile. Here the goal of artefact and standing monument documentation must be to “do duty to the original”. At the same time, it is also acknowledged that this is an impossible task (Baines 1984: 22, Strudwick 2001: 137), or at least thwarted by imprecisions inherent in axioms such as “nature”, “original” or other under-theorised concepts that situate the subject of documentation outside its context(s) of time and past social space.

16Overall, facsimile methods for early Egyptian graphical culture overwhelming emphasise what is depicted, the nature of the iconographic and repertoire, and design development (cf. Aldred 1980: 32-36). This is not to say that materials and technique are wholly ignored (later evidence for past image-making processes has provoked creative solutions such as the use of overlays (Strudwick 2001: 132). Nevertheless, the how of depiction and graphical symbolling by the early Egyptian artist/scribe and the ways in which the choice of tool and technique informed those areas which have been emphasised (design, appearance, semantic meaning) need to be more fully and consistently accounted for. Early Egyptian graphical culture is not merely a representation of past realities awaiting recovery through diligent recording methods but exists in the present; our reconstructions of it are contingent upon the methods and perspectives we bring to them (see Barrett 2000: 62-65, Jones & Macgregor 2002: 3). Even with this awareness and recognising the importance of accounting for the role of materiality and technological practice in the development of early Egyptian scribal systems, how can we detect, document and explore relationships between tools, materials, embodied practitioners and other aspects of the chaîne opératoire of image production and consumption? Before I turn to one possible solution, I want to briefly consider further imagery and related practices in the past from a theoretical perspective.

Past Graphical Practice and Sensory Experience

17Theories of material practice and the body assert the embeddedness of human action and experience in the material world. However, much research on image, writing and text takes an ocularcentric perspective, even though engagement with and experience of imagery, past or present, is never a purely visual activity (Claßen et al. 2009: 59). Indeed, human actors do more than just look at things, gaze, read or contemplate them, “...they may also touch, smell and taste them; people drag objects around, use, wear, tear, fix, repair and maintain them, grow tired of them, put them away, discard them and rediscover them” (Löfgren 1997: 103). In seeking to re-materialise early Egyptian graphical culture (Piquette 2008), yet avoid “fetishising the thingishness of things” we need to think about materiality as part of everyday skills, and competences as part of interactions with and experiences of graphical objects (cf. Löfgren 1997: 103). Within archaeology and anthropology phenomenology has become an importance source of inspiration for thinking through the way people perceive and experience cultural objects and landscapes (e.g. Olsen 1990, 2003, Gosden 1994, Ingold 2000, Thomas 2006). I argue that phenomenology also presents new and interesting ways of exploring graphical culture and past experience of scribal and artistic spaces. Sensory inputs of sight, sound, touch, taste, smell would have been central to individual graphical practice, albeit influencing engagement and meaning construction to different extents in the course of production and consumption.

18Among the environmental conditions which impinge on sensory perception and embodied techno-material action in the world is illumination. Light is fundamental for the visual perception of surfaces of material substances (Gibson 1979: 16-32, 121). From direct, ambient or reflected sunlight to lamp, torch or candle light, suitable illumination would have been required during acts of drawing, writing, painting, carving and so on (see e.g. Sundstedt et al. 2004, Gutierez et al. 2006, Kider et al. 2009). As illumination stimulates ocular perception, it reinforces other sensory experience and vice versa, such as the ways in which visual appearance corresponds to the tactile sensations of roughly cut wooden or ivory surfaces, or those sanded/smoothed in preparation for inscription. The reflexive practitioner would have sought a suitable relationship between light source and the location and angle of the material surface depending on tools and inscription technique used. Lighting considerations would have also extended to the general work space, and additional materials and paraphernalia (pigments, grinding equipment, adhesives, water, mixing pot, palette, artistic implements, etc.). Illumination would also have been adapted to practitioner’s(s’) bodily requirements of height, size, posture, quality of eye sight, types of movement required for a given task, and overall ergonomic preferences. These factors were also important for consumers of graphical culture during touching and handling, looking, reading or discussing aloud and so forth. Of course we must bear in mind the various possibilities: that primary significance may have resided in the creative act rather than its outcome, and that, where intended, audiences and types of consumption varied across time and space (cf. Baines 2012: 3).

19A techno-material perspective and consideration of the embodied practitioner also requires a broader conceptualisation such that early Egyptian graphical culture is set in relation to lived experience. I am acutely aware of the fact that only very limited evidence survives archaeologically to directly support such reconstruction. Even where evidence does survive in later periods, visualising and otherwise modelling past scenarios of phenomenological experience present challenges which may never be surmountable for both epistemological and technological reasons. However, one area where advanced digital imaging technologies are presenting exciting opportunities is in the documentation and visualisation of artefact surfaces. These developments facilitate study of early Egyptian graphical culture from conventional as well as phenomenological perspectives.

Reflectance Transformation Imaging

20Digital imaging technologies are increasingly enabling greater detail and accuracy in image capture in 2D, 2D+ and 3D. These are presenting easier ways of collating and sharing graphical data, and thus more opportunities for exploring the materiality of semantic meanings. However, due to the high cost of hardware and software (e.g. der Manuelian 1998: 113 on digital epigraphy) and difficulties in accessing the necessary training and expertise, some technologies are inaccessible to researchers within and without Egypt. Also, for digitisation in museum and field contexts, devices are needed that are quickly and easily deployable, have a simple calibration procedure and, ideally, few moving parts as these are vulnerable in field contexts (Willems et al. 2005), for example, where sand and dust easily become airborne. Likewise, digital techniques need to fit into ways of working that already exist; the aim should be to complement and augment those workflows rather than replace them. Reflectance Transformation Imaging (RTI) presents a digital capture system that addresses many of these pragmatic requirements. Moreover, it overcomes perennial problems associated with surface lighting during documentation, and of particular interest here, exploration of evidence for the techno-material issues discussed above through dynamically relit visualisations.

21During first-hand examination lighting proximity, intensity and direction can be adjusted manually but this iterative process is time-consuming, and can be cumbersome and unsatisfactory. With the exception of video capture and relighting on 3D models, the fixity of light location and angle is particularly problematic for convex, concave and other complex surfaces. RTI provides a solution to those problems by enabling the user to take multiple high-resolution detailed digital photographs of surface features with light applied from a different position for each exposure. The multiple captures are amalgamated together using mathematical algorithms such as Polynomial Texture Mapping (PTM, developed by Tom Malzbender (Malzbender et al. 2001) of Hewlett Packard Laboratories). The resulting RTI file allows for photorealistic virtual re-lighting and non-photorealistic rendering of surfaces in real-time.

22There are two main methods of creating RTI visualisations. One requires a lighting dome (fig. 1) to provide illumination during photography and is the most efficient way of imaging quantities of smaller portable objects (e.g. no larger than 17cm² for a lighting dome 1 meter in diameter). The other main method, Highlight RTI (H-RTI), makes use of a hand-held light source during photography (see Mudge et al. 2006; fig. 2). Photography of the target surface is undertaken with a tripod-mounted camera, both of which remain stationary while 24-60 photographs are taken, each exposure being systematically lit from a slightly different direction and position to create a virtual dome of light. Although more time consuming than using a lighting dome, H-RTI is adaptable to a wider range of surface sizes and locations. Similar results are gained with the manual and lighting dome methods, although the two case studies presented below were obtained via the latter technique.

23This particular system and the results derive from the collaborative University of Oxford and University of Southampton Reflectance Transformation Imaging Systems for Ancient Documentary Artefacts project, funded by the UK Arts and Humanities Research Council in 2010 via the Digital Equipment and Database Enhancement for Impact (DEDEFI) scheme (for detailed discussion and interim project results see Earl et al. 2011). The system consists of a 1-meter diameter plastic dome or hemisphere with 76 LED illumination sources evenly distributed and affixed to its underside. The dome disassembles into four quarters for portability. A digital camera is suspended overhead from a frame such that the lens projects through an aperture at the top of the dome (fig. 1). Artefacts are placed under the dome for imaging. Each RTI capture sequence entails taking a series of digital photographs using a different LED light source for each exposure, thus casting light onto the 3D artefact surface from multiple angles. Capture is executed from a computer using software to control the camera, LEDs, image file download and to record camera settings and related capture metadata. The series of images are then fitted using, for example, Polynomial Texture Map (PTM) fitting software (Malzbender et al. 2001). The result can be visualised using RTI viewing software (below).

24Compared with some advanced imaging techniques, RTI (and especially H-RTI) are relatively inexpensive in terms of equipment costs, personnel and time. With a few hours of training, dome-based RTI can be easily undertaken by most any researcher. Since surfaces are lit from multiple locations, RTI does not require nearly as much interpretation at the point of capture compared with fixed light documentation methods. Contact with the surface is not required with RTI, reducing risk of damage from tracing paper or similar, adhesives, pen/pencil pressure, etc.

1. Reflectance Transformation Imaging camera and lighting dome system
Reflectance Transformation Imaging camera et système de dôme lumineux

25Like any documentation method, RTI has its strengths as well as drawbacks. The dome places a restriction in the size and position of objects, but this can be overcome by using H-RTI for larger non-portable surfaces whether positioned horizontally, vertically, overhead or underfoot (fig. 2). Compared with conventional epigraphic methods, H-RTI in particular presents a low-cost solution for documenting surfaces accurately and quickly. Facsimile work normally undertaken on-site can be done thoroughly and accurately elsewhere given the ability to virtually relight surfaces from any position with a click of the mouse, as well as output optimally lit photographs for printing and tracing. Of course, in order to understand some aspects of materiality and embodied experience, especially for more 3D surfaces, first-hand engagement is still important, e.g. for perception of scale and shape vis-à-vis the body, the requirements of head and eye movements, the need to perambulate, bend or twist while viewing (Whitehouse 2013). Likewise, the tactile and muscular experience of shape, texture, temperature, and weight, as well as other sensory inputs associated with objects and spaces cannot (yet) be thoroughly assessed via digital visualisation and reproduction. Thus, as with any new digital tool or method, the intention cannot be to supplant traditional way of working, but to complement them (Bates et al. 1995, Benardou et al. 2010: 20).

2. H-RTI equipment set up for the capture of a limestone ostracon in the Ägyptisches Museum und Papyrussammlung, Staatliche Museen zu Berlin.
Mise en place de l’équipement H-RTI pour la prise de vue d’un ostracon en calcaire à l’Ägyptisches Museum und Papyrussammlung, Staatliche Museen de Berlin.

26As for using PTMS and other RTI file types, these can be viewed using software such as the open access RTIViewer (Bogart 2010). The user views the RTI file on the computer screen and explores the artefact surface by virtually adjusting the position of the light with the mouse for optimal viewing, by zooming in and out, navigating around the surface, and using rendering modes. This latter feature of the viewing software is especially critical for enhancing the appearance of surface morphology and, where graphical marks are concerned, improving ‘readability’. When viewing the surface and moving the light position, an impression of a 3D surface is created, but similar to conventional digital photography the viewing plane cannot be adjusted, hence the viewing experience is best described as 2D+. It should be noted that for surface features which do not cast a shadow (e.g. stains, traces of ink that have been absorbed into the surface), results are best achieved using multispectral lighting with RTI (Happa et al. 2010: 162). As the following examples demonstrate, where surface relief is present the use of visible light RTI captures features in striking detail, enabling systematic and controlled study in far greater depth than is usually possible with the naked eye.

Revealing Processes of Production: The Battlefield Palette and an ivory label

27The larger surviving fragment of the so-called Battlefield Palette or Lion Palette (EA 20791⁴) and an incised ivory label (EA 32668) held in the British Museum are among several early image-bearing objects on which I have conducted RTI⁵. This palette, a slab of mudstone (Stevenson 2005: 150) measuring 28.7 cm in width and 19.6 cm in height and carved in low relief on both faces is dated to c.3100 BCE and may have come from the Upper Egyptian site of Abydos (see Spencer 1980: 79-80, esp. n. 1 for no. 576). Large elaborate palettes of this type, the Narmer palette being the most well-known, appear to derive from the more humble palette traditions of the Predynastic–Early Dynastic periods (c.4100 BCE–c.3000 BCE). Predominantly found in burial contexts, these smaller palettes of rhomboidal, faunal and rectangular shapes bear ground mineral residues for cosmetic if not also ritual purposes (Petrie 1953, Stevenson 2005: 150). The large size of examples such as the Battlefield Palette, the lack of grinding evidence, elaborate nature of the carving, and their find contexts (where provenanced) point to the high status and more socially exclusive ‘ceremonial’ function (e.g. Baines 1989: 475-477).

28Published images of the Battlefield Palette include black and white photographs, colour photographs, line drawings (e.g. Petrie 1953: pls. D-E, Spencer 1993: 54-55, figs. 34-35), and line drawing with added diagrammatic details (e.g. Davis 1992: figs. 33-34). The aim of facsimile among previous studies of this objects has been to highlight image morphology, in order to support hypotheses concerning narrative structure and sequence of reading (Davis 1992), social symbolic roles (e.g. Baines 1989, 1993), and object function (O’Connor 2002). Similar to the lack of attention paid to the material significance as discussed by Alice Stevenson (Stevenson 2005: 150-153), material features such as non-representational surface marks are rarely the subject of illustration or study. That these are visible to some extent in many published photographs (e.g. Spencer1980: pl. 64, 576, lower), and especially apparent upon first hand examination, begs the question of why these technological details have been neglected. As discussed, this may be partly due to traditional ways of ‘seeing’ and ‘looking’ at early Egyptian evidence types designated ‘writing’ or ‘art’ – the tendency to look through technology and materiality in the interest of decipherment or interpretation of symbolism.

29It is therefore important to emphasise that the value of new imaging techniques is not solely their ability to capture data more quickly or with more precision, but also because – as the following results compellingly illustrate – these can provoke us to ‘see’ old evidence in new ways and gain new insight. The Battlefield Palette is shown in figure 3 as it appears under visible light, illuminated from all directions. The shape and form of the figures are quite clear. Some tool marks are visible, but the ground overall looks relatively flat. By zooming in and changing the direction of the light, surface marks become much more articulated (fig. 4). The specular enhancement mode augments the appearance of surface shapes (fig. 5). Upon navigating around the surface and adjusting the light settings, numerous tool marks and episodes of surface removal are revealed, bearing witness to the complex processes that were involved in creating the imagery, a selection of which are explained below.

3. Lower part of the Battlefield Palette
Partie inférieure de la Palette du Champ de bataille

4. Detail of the Battlefield Palette under normal visible light from upper right at an angle of 15˚–20˚
Détail de la Palette de champ de bataille sous une lumière normale située en haut à droite de 15˚–20˚ d’angle

5. Lower part of the Battlefield Palette using the specular enhancement rendering mode
Partie inférieure de la Palette de champ de bataille utilisant le mode de rendu de la RTI

30Study of the tools marks on both faces of the palette shows that figures were blocked out by carving relatively deep channels. In various places the decision was taken to adjust contours along the way as evidenced by cuts around some figures which echo their shape but are carved deeper than the present level of the ground. The ‘bird’ on the viewer’s right in figure 6, for example, exhibits a line running parallel and to the left of the lower wing, indicating that the width of the wing has been reduced, with the ends of the secondary feathers (those closer to the body) shortened or brought forward toward the leading edge of the wing. Other evidence for re-carving is seen below the head and beak of the ‘bird’ on the far left. Shadows reveal traces which, again, parallel the present contour suggesting that the lower beak had been thicker and the beak longer and more downward-curving. On the far left side of the palette in the area between the legs of a ‘human figure’ (fig. 7), a pair of cuts which form an up-side-down V-shape may be traces of previous intentional carving (tool slips are also possible). These marks were then partially scratched away or erased. Distinguishing between ‘erasures’ and marks that have been removed or reduced as part of the normal finishing process is difficult, although each is easier to recognise in softer stones like those from which the palette is carved (Anna Serotta Pers. Comm. 15 August 2011). Another possibly related pattern is characterised by a series of short parallel strokes can be seen below the ‘human figure’s’ foot. These short strokes differ in appearance from the contour recarving, and the ‘erasure’(?) between the legs in their length, angle, and spacing. These marks also differ from surrounding scratch marks which seem to relate to the general lowering and smoothing of the ground.

6. Specular enhancement rendering mode showing, among other details, a faint line running parallel with the back edge of of the lower wing of the right-facing bird
Le mode de rendu de la RTI montrant, parmi d’autres détails, une fine ligne parallèle au bord arrière de l’aile inférieure de l’oiseau de droite

31Careful study of the marks by moving the light position also provides insight into tool types. The deep troughs forming the contours of figures are similar in width and may be diagnostic of the shape and width of the tip of the tool used. At various points it is possible to discern a short abrupt line or cut within, yet running perpendicular to, the trough, e.g. on the hand of the ‘human figure’ being attacked by the ‘lion’ (fig. 8). What appears to be the carving mark from a flat square-cornered chisel tip (or part thereof) is apparent along the external contour of the right hand. A similar tool mark runs parallel and just below the hand. This may have been made with the thin edge of the same tool type or a ine-tipped tool to render the thin flexure lines in the hand. Of course, one must bear in mind that the tips of these tools were likely to have been modified constantly as work progressed. Flint chisels would be subject to fracturing, requiring re-knapping. Copper or copper alloy chisels would need to be re-hammered when they became dull (Stocks 2003: 74-99, Anna Serotta Pers. Comm. 15 August 2011). It is interesting to note what appears to be an apparent lack of tool control (?) or care (?) for these latter interior markings in particular. Cuts travelling beyond the outer contours of figures and into the ground can be found in several places on the palette, such as the detailing on the ‘birds’’ wings (fig. 7).

7. Apparent erasure marks between ‘human figure’s’ legs and below one foot directly ‘behind’ the bird’s upper wing, as clarified using the diffuse gain RTI rendering mode
Stries apparentes entre les jambes de la figure humaine et sous un de ses pieds juste à l’arrière de l’aile supérieure de l’oiseau, visibles grâce au rendu de la RTI

8. Detail from the lower central part of the palette showing, below the isolated hand, what may be the mark of a carving tool with a flat square-cornered tip
Détail de la zone centrale de la partie inférieure de la palette, montrant, sous la main isolée, ce qui pourraient être les traces laissées par un outil de taille à l’extrémité carrée applatie

32In addition to facilitating analysis of surface transformations which are important for understanding production of an object, RTI can also uncover evidence for artistic habit and physical movements which may be indicative of skill level, changing priorities or contrasting attention to detail⁶. RTI presents opportunities for reassessing previous interpretations of early graphical culture. For example, Whitney Davis has proposed a four-part narrative structure of the Battlefield Palette, dividing the composition into four rough horizontal registers or bands (Davis 1992: fig. 34, 122-123). In contrast, David O’Connor (O’Connor 2002: 13) finds the register effect irregularly maintained and composition conforming to a mode of circularity, revolving around the cosmetic cup in the centre of the ‘battle scene’. A fundamental question is therefore whether proposals for compositional understanding are supported by physical evidence for the production workflow. Can we discern a patterin in the blocking out of compositional space and image grouping? What does this tells us about the intentions of designers and/or artisans concerning structure and meaning? In the upper right of the Battlefield Palette, it is possible to discern horizontal lines under the feet of the cloaked human figure and rear foot of the bound ‘captive’ (fig. 9). These lines may point to an interest to ensure precise placement of the feet along a horizontal axis, although similar lines below the feet of other upright figures, such as the bound ‘captives’ in the upper left and the ‘lion’, are not present or, at least initial drafting of the composition with ink or by light engraving prior to carving do not appear to survive. About 1cm above the bottom edge of the palette at least two horizontal lines are scratched into the surface beginning on the left and appear to continue under the feet of the ‘bird’ to the right, under the arm of the fallen ‘human figure’. It is difficult to discern whether this relates to figure alignment or perhaps the cutting of the mudstone slab to shape.

9. Lines under the feet of the cloaked human figures in the upper right and the rear foot of the bound captive
Lignes sous les pieds de figues humaines vétues de manteaux en haut à droite et à l’arrière du pied du prisonnier attaché

33Further analysis of the surface features of all the fragments of the Battlefield Palette is needed. This is underway along with RTI-aided study of the Hunters and other decorated palettes (e.g. Piquette & MacDonald 2012). Additional study will hopefully provide insight into the wider social context of production, its organisation, the extent of its restriction, and mechanisms of knowledge transmission.

34These selected examples nevertheless show the kinds of insight RTI provides into artefact life history, for example, shedding light on compositional intentions and choices, including format, direction, orientation, view, association, relative scale, etc. These enhanced visualisations give access to a micro-scale stratigraphy that constitutes a new and valuable source of evidence for augmenting our understanding of how writing and art, through their physical expression, became increasingly distinct yet interrelated mode of symbolling (Piquette 2007: 275-284, Davis 1976).

35I now turn briefly to a second case study, the RTI results on an incised ivory label in the British Museum (EA 32668; Spencer 1980: 65, no. 465, pls. [p] 49, [d] 54). The inscribed labels, examples of which were mentioned above, are perforated, rectangular plaques of ivory, bone and wood found in funerary contexts during the Late Predynastic and the First Dynasty (e.g. Dreyer 1998, Piquette 2007, Petrie 1900, 1901). The labels were presumably attached to, or otherwise associated with, funerary items placed in and around ‘royal’ and other high status tombs. One or both surfaces bear written and iconic imagery concerning goods, their quantity and quality, with some examples depicting acts of the Egyptian ruler, and other events.

36The RTI results for EA 32668 also demonstrate the ways this technology can aid a techno-material and phenomenological approach to early Egyptian graphical culture. The development of preferences for certain configurations, shapes, proportions, types of curvature or rectilinearity was partially due to tool choice (see Herring 2005: 1, also Baines 2004: 159) and their knowledgeable use. Indeed, the development of particular physical movements or ductus – the direction, sequence and speed with which the strokes make up an image – in relation to tool type can be illustrated by comparing two First Dynasty instances of ‘stripped palm branch’/‘M4’ or the so-called ‘year-sign’. One was made using a thin chisel or burin, illustrated here (fig. 10) and another made with brush and ink on wood.

37The RTI specular enhancement rendering mode clarifies the physical structure of this tall vertical sign or ‘simple graphical object’ (SGO)⁷ which runs parallel to the far right edge (fig. 11a-b). This SGO was created by incising into the surface. At least two associated marks can be understood as tool slips. One extends upwards angling away and slightly to the right of the main vertical line, the direction and angle suggesting that force was applied, to the tool and in turn to the surface, from below (fig. 11a). Another tool slip is visible at the top left of the sign where, rather than following the downward curve of the SGO, it ‘derails’ to the left and away from the main trough (as viewed in figure 11b). Together these slips suggest a ductus of bottom to top, the upper slip perhaps even indicating that the scribe turned the label clockwise as she used the thin chisel to remove surface material to form the curve.

10. Incised and painted elephant ivory label, First Dynasty, reign of Semerkhet‘Royal’ Tombs Cemetery, Abydos
Plaquette d’ivoire incisée et peinte, 1ère Dynastie, règne de Semerkhet. ‘Royal’ Tombs Cemetery, Abydos

11a–d. Reflectance transformation images rendered in the specular enhancement mode, showing details of an inscribed label (see caption above)
Rendu d’images transformées par la réflectance, montrant les détails d’une plaquette gravée (voir la figure précédente)

38A contrasting example of ‘stripped palm branch’/‘M4’⁸ appears on a slightly earlier label probably dating to around the time of Adjib, made of wood in the Ashmolean Museum (E.1233), although only its top edge is preserved. It was inscribed in ink using a rush pen or similar flexible brush-like tool. First-hand study and the line drawing provided in Kaplony’s publication (Kaplony 1963: pl. 143, 845 [which also appears to have been based on first-hand examination)], see also Petrie 1901: 39, pl. 42, 41) suggest a different ductus (fig. 12). The preserved upper portion of this SGO appears to have been made using two strokes, one forming the top curved section and a second long vertical stroke. While it is quite clear that two strokes were used, their sequence and direction is more difficult to determine. The curving stroke may have started on the left, the brush being brought around to the right and as pressure was applied, broadening out its fibres as the scribe brought the arm-wrist-hand-finger-tool motion to a stop upon reaching the vertical axis of the sign. However, the reverse is also possible. It is here that RTI⁹ may augment study of early scribal practice (e.g. Kahl’s (2001) study of scribal hands on NIIIA2 labels), by further clarifying the intimate relationships between material, tool and technique in relation to bodily movement and gesture.

12. Upper part of a wooden funerary label from Abydos, probably from Tomb X attributed to Adjib
Partie supérieure d’une plaquette funéraire en bois d’Abydos, probablement de la Tombe X attribuée а Adjib

39While unfinished objects offer a direct window onto production processes (Schäfer 2002 [1919]: 76), RTI enables us to revisit well-studied, apparently-finished objects and discover additional processual evidence. Returning to incised ivory label EA32668, additional surface features in the lower left corner are of techno-material interest. The line drawing in the excavation report (fig. 10; Petrie 1900: pl. 17, no. 26) distinguishes with hatching surface damage probably due to insect activity, although this is inconsistently visualised. Beyond remarks on the damage, no comment is made regarding non-graphical marks in the lower left corner nor have I yet encountered their discussion or ‘accurate’ visualisation in subsequent publications (e.g. Kaplony 1963, I: 305, 1963, III: pl. 144, 847B, Smr-ḥt a, 1, Spencer 1980: 65, pl. 54, no. 461). Granted the surface of this label is difficult to make sense of; even during my own first-hand study for my doctoral research focussing specifically on materiality, I too overlooked the tool marks in this area (Piquette 2007). Only when exploring the RTI image of the label and repeatedly repositioning the light, and especially through use of the specular enhancement rendering mode, did I finally see what appears to be palimpsest. As shown in figure 11c, the surface bears a series of strokes which are clearly different in their morphology and ductus compared with other tool marks, probably associated with surface preparation prior to inscription. Indeed, their purpose seems to have been one of erasure (see Piquette 2013) given the presence of a vertical line discernible underneath the rear-most part of a neatly incised ‘mammal belly and tail’/‘F32’ or ‘log’/‘M41’ (?), and extending below and perhaps under what appears to be ‘pool’/‘N36/N39’and upwards toward ‘lion forepart ’/‘F4’, as well as the remains of an incision below ‘lion’s’ forepaw. It is also possible to discern a horizontal cut underlying the left-most ‘dots’ above ‘hand with dots emerging from it (or located around it)’/‘D46*¹⁰, as well as above the ‘wrist’. It is tempting to see ‘hoe’/‘U6–U7’ under ‘two arms’/‘D28’ on the left, although placement and relative scale appear odd. A similar set of marks can be found just to the right and below and therefore may be an ‘artefact’ of erasure. Overall, this area of the label was subject to at least four episodes of surface removal: surface preparation (e.g. short marks along left edge running from upper left to lower right), primary incision, erasure (marks running from lower left to upper right), secondary incision. This palimpsest raises interesting questions concerning label use. On many First Dynasty labels the lower area or lower left corner typically bears information understood to relate to a type of grave good; erasure and re-inscribed in this area therefore presents several possibilities. The maker could have simply made an error that was later detected and corrected. A change relating to the product may have occurred between the commissioning of the label and its use, whether prior to or in the course of the funeral and final deposition in the cemetery. Another possibility is that some labels were used more than once and for this reason the product information was changed. These scenarios cannot be substantiated at present, but present tantalising glimpses into concerns that would have been part of the lived experience of early Egyptian artisans and consumers, and thus the wider context of graphical practice (for more on ‘erasure’ see Piquette 2013).

40Analysis of techno-material features with the aid of RTI reveals the complexities of artefact ‘life histories’: production and consumption were not necessarily sequential in time and space but consisted of a process of tacking backwards and forwards. From tool use, technique, and compositional choices, to intentions and changing priorities, the implications of these different aspects of graphical practice are important for how we understand image function and overall meaning. Such evidence stands to provide important insight the wider social contexts of graphical development, such as conventionalisation or renegotiation of norms through artistic practice and scribal knowledge sharing, whether in workshops or more informal settings, whether alone or in dialogue with other artisans, or with patrons, commissioners or other stakeholders.

Summing up and Future Directions

41As I have highlighted previously (e.g. Piquette 2008 and more recently, Piquette & Whitehouse 2013), the artefact:text duality has persisted for decades across much of archaeology and textual studies (see Johnson 1999, Moreland 2006, 2001). Similarly, interest in past writing and its progenitors often centres on iconic and linguistic meaning. In this chapter I considered through the significance of materiality and technology for understanding early Egyptian imagery, with particular emphasis on production. I highlighted the need to address the paradox of technology by considering reflexively image-making technologies in the present and how these impact on our reconstructions of past graphical practices.

42I also directed attention to methodological issues of data quality in documentation and visualisation, outlining some of the barriers to the capture of data needed to support a techno-material approach. RTI alleviates many of these problems, presenting a powerful tool for interrogating early Egyptian iconography in new ways. Some practioners of post-processual archaeology adopted the view that, inasmuch as we read the signs and words of a text, so to can we ‘read’ past material culture (Hodder 1986, Tilley 1990, 1991, see also Frantz 1998: 794). There has also been the recognition of important differences between things and texts (Hodder 2000, see also Buchli 1995); material culture is in the world and plays a fundamentally different constitutive role for our being in this world than texts and language (Olsen 2003: 90). However, inasmuch as things do “more than just speak and express meanings” (cf. Joerges 1988: 224), critiques of textual analogy have neglected to recognise that writing and image also do more than just encode language and other symbolic meanings. We must take the critique further and turn ‘material culture as text’ on its head. Thus, from the perspective of ‘text as material culture’ – or more appropriately for evidence where semantic function is ambiguous, ‘graphical culture as material culture’ – archaeological method and theory become emphatically germane to research of an art historical and philological nature, just as these domains do to archaeology, however defined.

43In addition to enabling detailed study of the artefactuality of inscribed objects, improving readability and aiding epigraphic work, the high quality image files created using RTI and its re-lighting and enhancement tools enable greater consistency in documentation and visualisation. The files can be shared and easily viewed via open access software (see e.g. http://culturalheritageimaging.org/). By building and maintaining an open access RTI community, this data capture method enables users to respond to calls for detailed and synthetic studies that are integrated into wider archaeology (see e.g. Riemer 2009: 44). Projects such as RTISAD (above) have been working toward open access RTI repository has been exploring issues of developing interconnected annotations around these digital objects.

44In addition to presenting a robust imaging technology to add the archaeologist’s repertoire, an area where RTI also stands to play an exceptionally important role is in preservation. Peter der Manuelian (der Manuelian 1998: 97) has highlighted the need to take advantage of new technologies, particularly in light of the rapid rate of deterioration and other threats to Egyptian and Sudanese monuments. The destruction of the earliest monumental rock inscriptions at the south Sinai site of Wadi el-Humur (Ibrahim & Tallet 2008: 154) and elsewhere (e.g. Hendrickx et al. 2012: 1069-1070) highlights the need for easily field-deployable technologies and rapid low-cost recording – something for which RTI, and especially H-RTI, are suitably designed.

45RTI has been applied successfully to a range of documentary evidence (Earl et al. 2010a, b, 2011, Piquette 2014), but beyond the present and a small amount of other Egypt-related work (see brief reference in Willems et al. 2009: 21, n. 60, pls. 4A–B, D on Dayr al-Barsha for jar stopper sealings; Anna Serotta is conducting H-RTI on stone sculpture at the Metropolitan Museum of Art, pers. comm. 20 April 2011), RTI has seen limited use within Egyptology¹¹. I hope that the overview presented above will prompt further interest in RTI for documenting, visualising and studying the development of early Egyptian graphical culture. RTI provides a point for departure, and return, for the reflexive exploration of image-making materials and technologies with regard to past practice and of particular importance, our re-constructions methods and technologies in the present.