Let's talk about... damage - representing discontinuity of decorative surfaces

E. Naville – The Temple of Deir el Bahari Part 1, Plate 29. Shrine of Anubis

The desire to indicate the greater context of ancient Egyptian reliefs and/or paintings has been present since the dawn of scientific epigraphic documentation. Due to centuries of decay and abuse of the monuments, there is one apparent feature that stands out immediately when we look at tomb or temple walls today: the tremendous amount of loss of the original surface. There are numerous reasons why a decorative surface would be missing. First and foremost, much damage was caused by natural causes, such as earthquakes, fire, rain or sandstorms. Accidental damage could be caused by human activity in or around the monument, like extending/transforming the structure or selecting the monument for reuse in later periods. We shouldn’t forget that ancient Egyptian monuments provided shelter for many generations to come after losing their original function.

A more interesting type of loss is the intentional damage, occuring when special circumstances in certain periods led to iconoclastic acts. For example Hatshepsut's and Akhenaten's images and names were specifically attacked: in Hatshepsut's case, this might have been az an attempt by his nephew at legitimizing his succession; and with Akhenaten, his images were targeted for his rejection of traditional gods and their powerful priesthoods. Not coincidentally, Amarna agents did a fair amount of destruction to the monuments as well. Later, with Christianity becoming dominant in the Roman empire, its followers began destroying “pagan” monuments, paying special attention to the faces of “demons”, appearing on tomb and temple walls around them (although their destructive role is likely exaggerated). Finally, with the Muslim conquest in the seventh century and the rise of Islam as a major religion in Egypt, the remains of pharaonic past were no longer seen as having power and their inscriptions were no longer understood. Temple walls were often reused as building blocks for new constructions.

The most concerning question for data driven visual documentation is how to know which damage is intentional. In some cases, the chisel marks leave no doubt. But in others it is less clear as chisel marks could be left by ancient tomb robbers or modern looters as well. Nonetheless, recording this type of information highly effects not only our understanding of the decorative surface but the greater history of the monument itself.

E. Naville – The Temple of Deir el Bahari Part 1, Plate 25. Hatshepsut (erased) between Nekheb and Harmakhis.

A. H. Gardiner ed. - The Temple of King Sethos I at Abydos, Volume I, Plate 5. Sethos worships and gazes upon Osiris.

In the early days of scientific documentation, damaged areas were deliberately shown on epigraphic drawings for two reasons: to have a clear indication of missing pharaonic elements, that is to show some context to the missing segments of carved/painted outlines; and, alternatively, to provide an aesthetically pleasing appearance of wall segments represented in color. The former effort was directed towards the understanding of decorative elements, therefore it didn’t put much emphasis on the nature of these obstructions, but rather applied a homogeneous fill to represent surface loss. On the other hand, the latter, mostly born out of the necessity caused by the rudimentary state of color photography, would represent a photorealistic approach, imitating damage as a 3-dimensional surface with the added shadow and depth as it appears on the wall. We must add that these carefully painted reproductions were often based on washed out B/W photographs, as was the case with many of The Epigraphic Survey’s color drawings created in the first half of the last century.

Almost none of the damage is represented (The Epigraphic Survey - Medinet Habu Vol II - Later Historical Records of Ramses III (1932), Plate 89. The town of Tunip attacked by the Egyptians).

Most of the damage is indicated in a photorealistic manner (The Epigraphic Survey - Medinet Habu Vol X - The Eighteenth Dynasty Temple, Part II., Plates 227, 226 and 223 – upcoming).

As we have already mentioned in a previous tutorial, there is still a lot of debate, hesitance and argument pro and contra when it comes to studying, documenting and/or indicating damaged areas of ancient Egyptian wall surfaces. Certainly, with the advent of digital macro photography and especially photogrammetry, we’ll have more and more nuanced surface information other than what we present on a drawing. Nonetheless, we think that even with today’s technological advances, there could be a very convincing case to be made for the artistic representation of this type of information. Let’s continue with demonstrating the various ways in which damaged areas are/were treated by the many epigraphic projects in Egypt. In this regard, The Epigraphic Survey’s drawings certainly stand out, as the most sophisticated (and most realistic) renderings of data indicated by surface loss. Perhaps not immediately apparent, but during the Survey’s century-long history, their damage representation has gone through significant changes, ranging from almost no visualization to creating hyper-photorealistic variants. By looking at the Survey’s publications, we may have a better understanding of the features that were considered important enough to be indicated on their drawings.

The Epigraphic Survey’s practices in representing discontinuity of decorative surfaces

(1) All surface loss is represented by a schematic texture with no visible damage outlines (Medinet Habu Vol I (1930), Plate 10 – Ramses III returning in triumph from the Nubian Campaign).

(2) Later cut-outs, such as deep beam-holes, are indicated with heavier modeling, presenting a photo-realistic view of these features (Medinet Habu Vol I (1930), Plate 29 – Ramses III issuing equipment to his troops for the campaign against the sea peoples).

(3) Largely homogeneous cross-hatched texture is applied over damaged parts of the decorative surface (Reliefs and Inscriptions at Karnak Vol 3 (1954), Plate 6 – Triumphal relief of Shoshenq I).

(4) Schematized texture fill indicates all damage with no distinguishable differentiation between light and dark areas, block lines etc. (Medinet Habu Vol V (1957), Plate 305 – The 'Shadow of the door' on the South wall of the second court).

(5) Almost photo-realistic representation of surface loss with the use of deep black shading indicating block lines (Medinet Habu Vol VIII (1970), Plate 594 – The 'Shadow of the door' Ramses III and Ramses IV offering flowers to Amon-Re).

(6) Long, curvature brush strokes represent the broken surface’s general flow, providing a much more dynamic appearance to the drawings (Medinet Habu Vol VIII (1970), Plate 653 – Princesses in attendance on Ramses III).

(7) The sporadically preserved carved decorative elements are basically lost in the sea of heavily rendered damage (The Tomb of Kheruef Theban Tomb 192 (1980), Plate 45 – Ceremony of towing the Night bark).

(8) Surface loss is represented by overly strong texturing that becomes counterproductive, essentially making the relief hard to read (The Temple of Khonsu Vol II (1981), Plate 170 – Ramses XI being crowned by the Theban triad).

(9) Only the darkest damage areas are indicated by a light and elegant texturing technique, drawing the much-deserved attention to the relief (Reliefs and Inscriptions at Karnak Vol (1986), Plate 5 – King Sety I defeating shasu bedouin on the road to Gaza).

(10) Stone repairs are distinguished from regular surface loss by the application of various light texturing, keeping the overall focus on the decorative elements (Reliefs and Inscriptions at Luxor Temple Vol I (1994), Plate 30 – The barge of Mut).

(11) Once again, heavy texturing is applied, indicating individual chisel marks etc., providing a thorough but overwhelming damage representation (Reliefs and Inscriptions at Luxor Temple Vol II (1998), Plate 186 – The king offering flowers to Amun-Re-Kamutef in the presence of Isis).

(12) A wide variety of surface loss types are indicated in a photorealistic manner, where line drawing and damage representation essentially become equally important (Medinet Habu Vol IX - The Eighteenth Dynasty Temple, Part I (2009), Plate 11 – The king led into the temple by Montu-Re and Atum and received by Amun-Re).

(13) On the Survey’s more recent drawings, damage is indicated in a less photo-like manner, leaving blank spots for highlights while applying only a minimal texture over dark areas. (The Facade of TT 107, The Tomb of Nefersekheru - forthcoming).

Drawing damage texture in such a careful manner takes a lot of artistic talent while being a very time-consuming element of the inking process. Despite the many stylistic variations shown above, the general concept of when and how damage should be indicated on a drawing hasn’t changed. For consistency reasons, the digital Chicago Method renders damage texture in a traditional manner, providing the artist with numerous Photoshop tools to speed up the process. As a rule, surface loss is only shown on the Survey’s drawings when the carved line has been interrupted. With some exceptions, damage should not be shown in any damaged area which has never been carved. Regarding the drawing technique, damage is rendered by a series of short trace weight (6 pixels) brush strokes which are subsequently broken with a digital scalpel brush (4 pixels) after the collation corrections have been completed. Deeper damage can be shown with more closely spaced lines, while shallow damage is drawn with less spacing for a three-dimensional appearance. One can revisit the relevant chapter of the Survey’s drawing conventions for more information about how to draw damage according to the Chicago House method. digitalEPIGRAPHY has also created a short video about applying the Survey’s digital damage texture over large areas.

As effective and visually pleasing as it is, the Survey’s complex damage treatment couldn’t be justified if not for the data that is to be collected and represented throughout their drawings. Chisel marks with clear indication of size, flow and orientation can be clear signs of hacking or re-carving certain elements of the scene. Surface scratches and incisions can imply a graffito or iconoclastic acts, occasionally being the remnants of architectural activity on the monument. Representing patch stones and repairs in the selective manner only a drawing provides can be extremely rewarding when studying wall structure. The appearance of beam holes, medieval chisel blows and damage resulting from deliberate mutilation can be clear indicators of later occupancy of the monument. Fertility gouges, regularly arranged scooped-out portions of the decorative surface, can refer not only to the number of visitors during the later history of the monument, but to the changes of ground level throughout history. Finally, and most importantly, the inclusion of damage on the drawing can be used as a visual explanation as to why certain portions of the decorative scheme are poorly preserved or missing entirely. What makes the artistic depiction of damage far superior to photographic recording is the selective nature of drawing. With the drawing, the artist is capable of selecting and enhancing the surface loss elements that have information value, while irrelevant portions can be left out or rendered in a less obtrusive manner. With that said, not every project is as keen on representing surface loss as the Epigraphic Survey, and some of our colleagues have a very different idea about the data value of damage. One can find a few examples of various damage treatments below. (The list is based on the sample studies appearing in our Reading section.)

Damage treatment variants in epigraphic documentation

(1) Epigraphic Program of the French-Egyptian Centre for the Study of the Temples of Karnak during the 2014 season (Orthophotography and facsimile of the eastern doorjamb of the north face of the eighth pylon and Tuthmosis III doorjamb).

(2) Late antique Hagr Edfu - Area 2b, Tomb D (Hagr Edfu, Area 1, two phases on decoration).

(3) The Chapel of Osiris Nebdjet/Padedankh in North-Karnak (Fig. 7. Facsimile of the inscriptions on the lintel).

(4) Die Kultkammer des Ka-ni-nisut im Kunsthistorischen Museum Wien (Fig. 6. Offering scene).

(5) The New Kingdom Temple of Khnum in Elephantine (Three sides of a pillar from the barque station of Hatschepsut).

(6) Digital epigraphy of the temple of Debod (Comparative views of part of a scene)

(7) The Temple of Tuthmosis III at Deir el-Bahari (Scenes from the relief decoration of the Tuthmosis III temple).

(8) The Tomb of Pharaoh’s Chancellor Senneferi at Thebes (TT 99) Volume I - The New Kingdom (Wall 16 wall plan).

(9) The Tomb of Rekhmire (TT100) - facsimile paintings by Nina de Garis Davies (Pl. XIII. Thutmose III enthroned).

(10) The Tombs of Userhat and Khonsuemheb (Lintel of the tomb of Userhat).

When discussing surface damage variants and their epigraphic treatment, we must also differentiate between surface loss occurring in stone versus plaster, keeping in mind that a significant amount of rock-cut tomb decoration was applied on either lime-based or mud plastered surfaces. Although these fragile wall scenes were even more prone to damage, in most cases, damage texturing wouldn’t add much data value to the epigraphic recording. Nevertheless, indicating damage outlines on the drawing has many benefits and a tremendous data value. The main question remains: how can this visual information be represented without reaching a tipping point when the drawing becomes overwhelmed by data?

Omitting (E. Dziobek, Des Grab des Inieni Theben Nr. 81, 1985 on the left) and adding (D. Polz, Das Grab des Hui und des Kel Theben Nr. 54, 1997 on the right) damages to the line drawing both have their strengths and weaknesses.

In the tomb of Nebamun (TT 179), the amount of wall surface loss was so devastating that differentiating between certain decorative elements became a challenge. One obvious solution to this problem would have been to deliberately omit damage in the drawing process and use only photographs to show contextual information. However, with today’s digital drawing techniques, one can indicate the extended set of damage data on separate layers, providing a lot of flexibility in presentation. In TT 179, damage edges were shown as an outline, using randomized single weight brush strokes with no indication of texture. Representing even the tiniest scratches on every wall (including both natural and intentional surface loss) provided us with the complete damage map of the tomb. In addition to the overly complex scenery of the photograph, these damage outline drawings, with their hue toned down to a less obtrusive level, delivered a tremendous amount of context to the decorative elements.

Damage treatment in TT179

(1) Color photograph of the West Longitudinal Hall in TT 179, the tomb of Nebamun (18. Dynasty, detail).

(2) Damage map shown at 100% opacity indicating the overwhelming amount of surface loss occurring on the wall scenes.

(3) Line drawings of the same area presented at 100% opacity without any damage represented. Note the blank areas with no information value.

(4) Line drawing (80% opacity) and damage (65% opacity) shown together has the benefit of indicating contextual information for both layers.

(5) When line drawing, damage and color code layers are all shown at 100% opacity, the drawing gets extremely busy with data.

(6) Line drawing (80%), damage (60%) and color code (65%) represented at various opacity according to their data value. Note the figure that is only recognizable with all three layers turned on.

Although not mentioned here, there are surface elements other than damage altering the way we recognize elements of a wall scene. Portions of a decorative program can be covered by cement, plaster, soot, bat guano etc., all posing the same question: should or shouldn’t we record this information during the drawing process. While the Survey’s drawing conventions give a definite answer to some of these questions, representing such data should be dealt with on a case by case bases.